Zbirka / Series OPERA INSTITUTI ARCHAEOLOGICI SLOVENIAE 46 Uredniki zbirke / Editors of the series Jana Horvat, Benjamin Štular, Anton Velušček Slavko Ciglenečki Between Ravenna and Constantinople Rethinking Late Antique Settlement Patterns Recenzenti / Reviewed by Rajko Bratož, Ivan Šprajc Urednica / Editor Zvezdana Modrijan Prevod / Translation Andreja Maver Fotografije / Photographs Slavko Ciglenečki Tehnična ureditev in prelom / Technical Editor and DTP Mateja Belak Oblikovanje ovitka / Front cover design Tamara Korošec Založnik / Publisher Založba ZRC Zanj / Represented by Aleš Pogačnik Izdajatelj / Issued by ZRC SAZU, Inštitut za arheologijo Zanj / Represented by Anton Velušček Tisk / Printed by Present d. o. o., Ljubljana Naklada / Print run 500 izvodov / copies Izid knjige sta podprla / Javna agencija za raziskovalno dejavnost RS (Slovenian Research Agency), Published with the support of Znanstvenoraziskovalni center SAZU (Research Centre of the Slovenian Academy of Sciences and Arts) Fotografija na naslovnici / Slavko Ciglenečki: Front cover photo Gradina Žirje, well-preserved wal s and protechisma of Early Byzantine fort Ljubljana 2023; prva izdaja, prvi natis / first edition, first print Prva e-izdaja knjige (pdf) je pod pogoji licence Creative Commons 4.0 CC-BY-NC-SA prosto dostopna tudi v elektronski obliki (pdf) / First e-edition of the book (pdf) is freely available in e-form (pdf) under the Creative Commons 4.0 CC-BY-NC-SA. DOI: https://doi.org/10.3986/9789610507369 CIP - Kataložni zapis o publikaciji Narodna in univerzitetna knjižnica, Ljubljana 904(4)«652« 94(37) CIGLENEČKI, Slavko, 1949- Between Ravenna and Constantinople : rethinking late antique settlement patterns / Slavko Ciglenečki ; [prevod Andreja Maver ; fotografije Slavko Ciglenečki]. - 1. izd., 1. natis = 1st ed., 1st print. - Ljubljana : Založba ZRC, 2023. - (Opera Instituti Archaeologici Sloveniae, ISSN 1408-5208 ; 46) ISBN 978-961-05-0735-2 COBISS.SI-ID 149698051 ISBN 978-961-05-0736-9 (pdf) COBISS.SI-ID 149805827 Slavko Ciglenečki BETWEEN RAVENNA AND CONSTANTINOPLE RETHINKING LATE ANTIQUE SETTLEMENT PATTERNS LJUBLJANA 2023 CONTENTS Preface ...................................................................................................................................................................................... 7 1. Introduction ....................................................................................................................................................................... 9 1.1 Scope of the book ........................................................................................................................................................ 9 1.2 Changes in settlement patterns – recent research ................................................................................................... 9 1.3 Geographical scope ................................................................................................................................................... 10 1.4 Chronological scope ................................................................................................................................................. 11 1.5 Some overviews of Late Antique settlement in the area under discussion ........................................................ 13 1.6 Content of the book .................................................................................................................................................. 14 1.7 Methodological remarks .......................................................................................................................................... 15 1.8 Terminological remarks ........................................................................................................................................... 16 1.9 Open problems, shortcomings ................................................................................................................................ 17 2. Cities .................................................................................................................................................................................. 19 2.1 Altered appearance of Late Antique cities ............................................................................................................. 19 2.2 How to identify a Late Antique city? ...................................................................................................................... 20 2.3 Cities abandoned before the end of Late Antiquity .............................................................................................. 24 2.3.1 Western part ....................................................................................................................................................... 24 2.3.2 Eastern part ........................................................................................................................................................ 38 2.4 Cities with continuity ............................................................................................................................................... 39 2.4.1 Western part ....................................................................................................................................................... 40 2.4.2 Eastern part ........................................................................................................................................................ 59 2.5 Cities newly-founded in Late Antiquity ............................................................................................................... 104 2.5.1 Western part ..................................................................................................................................................... 104 2.5.2 Eastern part ...................................................................................................................................................... 119 2.6. Concluding observations on Late Antique cities ............................................................................................... 149 2.6.1 Transformation of Late Antique cities .......................................................................................................... 149 2.6.2 Characteristics of Late Antique cities ........................................................................................................... 151 2.6.3 Characteristics of groups of Late Antique cities and differences between them ..........................................156 3. Settlement in the countryside ....................................................................................................................................... 167 3.1 Major settlement changes in the countryside ...................................................................................................... 167 3.2 Settlements with continuity ................................................................................................................................... 167 3.2.1 Western part ..................................................................................................................................................... 169 3.2.2 Eastern part ...................................................................................................................................................... 182 3.3 Newly-founded settlements in the countryside .................................................................................................. 187 3.3.1 Newly-founded unfortified settlements .................................................................................................. 187 Western part .................................................................................................................................................... 187 Eastern part ..................................................................................................................................................... 196 3.3.2 Newly-founded fortifications .................................................................................................................... 200 Western part .................................................................................................................................................... 200 Eastern part ..................................................................................................................................................... 265 3.4 Interpretation of the settlement patterns in the countryside ............................................................................ 314 3.4.1 Settlements with continuity ...................................................................................................................... 314 3.4.2 Newly-founded unfortified settlements .................................................................................................. 314 3.4.3 Newly-founded fortifications .................................................................................................................... 315 History of research ......................................................................................................................................... 315 Functional identification of the newly-founded fortifications ................................................................. 322 Categories of fortified sites ............................................................................................................................ 324 4. Diachronic assessment of the settlement changes and of the city–countryside dynamics. General remarks on the chronology of the settlement changes ............................................................................................................ 337 4.1 First changes of the settlement pattern in the second half of the 3rd and first half of the 4th century ......... 438 4.2 Marked changes in cities, gradual abandonment of lowland settlements and the appearance of numerous hil top sites (last third of the 4th and first half of the 5th century) ..................................................................... 340 4.3 Transformation of the settlement patterns after the mid-5th century: landscape of fortifications and declining cities in the 6th and the end of Late Antique settlement in the first third of the 7th century ......... 344 5. Brief outline of the Late Antique settlement patterns in other parts of the Roman Empire ................................ 351 5. 1 Western part of the Empire ................................................................................................................................... 351 5.1.1 Cities ................................................................................................................................................................. 351 5.1.2 The countryside ............................................................................................................................................... 356 5.2 Eastern part of the Empire ..................................................................................................................................... 362 5.2.1 Cities ................................................................................................................................................................. 362 5.2.2 The countryside ............................................................................................................................................... 372 6. Concluding remarks and an attempt to interpret the settlement patterns ............................................................. 383 Bibliography ........................................................................................................................................................................ 389 Index .................................................................................................................................................................................... 421 PREFACE The recent research of the cities and the coun- tenegro, Kosovo and North Macedonia). I also closely tryside in the area between the Late Antique capital followed such investigations in neighbouring Austria of the western Empire in Ravenna and cosmopolitan and Italy. The idea for a publication on the subject was Constantinople has revealed a settlement that has many thus spontaneously born in the 1990s in the face of a layers and one that witnessed extensive processes of rapidly growing number of previously unknown Late transformation. Despite the growing body of evidence, Antique sites that brought a fundamental shift in our it is an area that remains poorly known to a large part perception of the settlement in the eastern Alpine and of the scholarly community and hence often incorrectly Balkan areas. interpreted due to linguistic barriers, cultural differences I set out to first il uminate the phenomenon of and occasional y tumultuous political situations. fortified Late Antique settlements in the area of the In contrast with the well-known settlement of the former Yugoslavia, which were poorly known outside Roman period, its equivalent during the time when the the Yugoslav borders, thereby giving the impression of Roman Empire was in decline, the first Germanic king- a sparsely inhabited area. The first major step toward doms emerged in the West and the Byzantine Empire be- this goal was already taken in 1987, when I published an came a prominent force in the East, was not well-known overview and interpretation of the fortified sites in the until the final decades of the 20th century. This was in eastern Alps and part of the western Balkans. During part the result of the scholarly attention being focused my preparations for the publication, it transpired that on studying the vestiges of the highly-developed Roman this phenomenon was by no means limited to the area civilisation, but also a consequence of the substantial of the eastern Alps as had been believed until the 1970s. changes in the Late Antique settlement patterns that Observing numerous similar investigations in many ar- were not appropriately identified across the Empire. eas that were exponential y increasing as time went on, The trend of research shifted in the 1970s and I expanded the overview to include the whole Balkan 1980s, with an increasing interest in the vanishing Peninsula. I realised that this settlement phenomenon classical world and its gradual transformation to the risked to be poorly understood unless it was integrated Middle Ages. More and more scholars recognised the in the complete settlement of the time. For that reason, significance of the last settlement cores for understand- I included in my work not only urban but also unforti- ing how life changed in different areas of the former fied countryside settlements that allow a comprehensive Roman Empire. This went hand in hand with a growing understanding of the changing settlement patterns in interest in the profoundly altered settlement pattern with this turbulent period in the history of the western world. dwindling cities and other lowland settlements, on the The ongoing work on the book seemed to still be one hand, and fortified settlements at locations selected manageable in the 1990s. The later investigations and with different criteria in mind, on the other; the latter innumerable publications increased the number of revealed a settlement pattern not noticeable in such an known sites, but also explained some of the previously extent in other parts of the Roman Empire. open questions. This required visits to the different and The 1970s and 1980s also witnessed a proliferation in some cases poorly accessible sites, as well as reading of research into the Late Antique fortified settlements in the vast body of publications. Slovenia, in which I actively participated. At the same I was final y able to finish the book thanks to my time or only slightly afterwards, similar research of colleagues at the Institute of Archaeology ZRC SAZU varying intensity was conducted in other republics of in Ljubljana, who enabled me to continue work long then Yugoslavia. This gave me the opportunity to learn after retiring and provided me with all the necessary about the sites in the former Yugoslav republics (today logistical support. I would especially wish to thank Slovenia, Croatia, Bosnia and Herzegovina, Serbia, Mon- Zvezdana Modrijan, my closest colleague, who success- 7 FOREWORD ful y undertook the demanding and time-consuming and Bojan Djurić. Among the currently active genera- role of editor-in-chief, reviewing and editing the nu- tion of archaeologists are also two of my closest col- merous early versions, but also careful y verifying the leagues, namely Zvezdana Modrijan and Tina Milavec, translations. Jana Horvat and Tina Milavec also offered with whom we performed the last major investigations many helpful comments and critical remarks. Mateja of Late Antique sites. I wish to sincerely thank all of Belak was entrusted with arranging the multitude of them for the innumerable insightful conversations and photographs and plans, creating maps, but also with the generous support in my work. desktop publishing tasks. Andreja Dolenc Vičič and The study of Late Antiquity also brought me into Tina Berden are to be credited for careful y reading and contact with numerous scholars from the former Yu- amending the long list of references. Dragotin Valoh goslavia and neighbouring Austria, Italy, Germany and scanned some of the plans, Tamara Korošec made the Hungary. Their list is long and I should limit myself book cover design. Andreja Maver translated the text to mentioning those with whom I col aborated most into English that required great attention with regard to frequently; they are Franz Glaser, Ul a Steinklauber, the sometimes poorly defined terms used for particular Sabine Ladstätter, Volker Bierbrauer, Michael Mack- forms of Late Antique settlement. I am also grateful to ensen, Gian Pietro Brogiolo, Maurizio Buora, Luca Vil a, all three colleagues who succeeded each other as heads Orsolya Heinrich-Tamáska, Vladimir Sokol, Emilio of the Institute during the years the book was being Marin, Željko Tomičić, Miljenko Jurković, Mihajlo written, namely Janez Dular, Jana Horvat and Anton Milinković, Vujadin Ivanišević, Ivan Mikulčić, Viktor Velušček, who attentively followed my work with a Lilčić and Gezim Hoxha. The Alexander von Humboldt wholehearted support. Foundation offered me a fellowship and thereby enabled The five decades of my active research work have an in-depth study of Late Antiquity in 1989/1990 in involved the col aboration with numerous colleagues in Bonn and Frankfurt. Slovenia. I should first mention the preceding generation The book could also not come to fruition without of archaeologists, now deceased, with whom my profes- the unfailing support of my family, Vlasta ( ), Živa and sional path began; they are Lojze Bolta, Jaro Šašel, Iva Jan. My partner Nataša Segulin shared with me the Curk, Vinko Šribar, Peter Petru, Marjan Slabe, Ljudmila arduous last two decades of the work, provided en- Plesničar Gec and Stanko Pahič. Most of my work was couragement and often accompanied me in fieldwork conducted with the members of my own generation and and travels. those slightly younger who include Timotej Knific, Rajko Bratož, Davorin Vuga, Mira Strmčnik Gulič, Dragan Božič, Marjeta Šašel Kos, Peter Kos, Janez Dular, Darja Pirkmajer, Milan Sagadin, Jana Horvat, Ivan Šprajc, Ivan Slavko Ciglenečki Tušek, Andrej Pleterski, Nada Osmuk, Božidar Slapšak In Ljubljana, 14 March 2023 8 1. INTRODUCTION 1.1 SCOPE OF THE BOOK Having said that, the broadened scope of the topic required that I set geographic boundaries, as the work The frequent political, military, economic and social would otherwise prove too extensive. I decided to focus crises of Late Antiquity coupled with a relentless pressure on the area between both capitals of the Late Antique on the limes caused profound changes in the settlement of Empire, namely Ravenna and Constantinople. This is the areas between the Eastern Alps and Constantinople. also an area straddling both sides of the border that, Gradual y, the elaborate network of weakened Roman from the reign of Theodosius onwards, divided the cities, countryside vil as, farmsteads and vici witnessed the Empire into two parts. development of new types of fortified cities, settlements, Such a geographic scope seemed sensible in spite refuges, and military forts. These newly-formed sites are of a slightly more limited knowledge of the eastern half the most characteristic expression of the altered settle- of the selected area. Time and time again, I noted how ment of Late Antiquity and closely reflect the challenging the results of studies were being presented almost exclu- conditions in a time when the Graeco-Roman world went sively within the confines of contemporary state borders, into decline and the Middle Ages began. where typologies and terminologies were also being A great amount of research has been conducted in separately developed. It therefore seemed important to recent decades in the area between Ravenna and Con- adopt a single point of view to present the numerous sites stantinople that go a long way toward explaining these and to use a single set of criteria to il uminate their com- new, previously poorly understood settlement patterns. monalities and differences. For a proper understanding The results offer a better insight into the settlement of of the selected area, it is important to take into account Late Antiquity, but are − mostly due to linguistic and the characteristic settlement patterns in other parts of partly also political barriers − difficult-to-access to a the Empire as well; these parts are briefly presented in large number of archaeologists and historians. The study separate chapter (see Chapter 5). in front of you aims to partly rectify this. The writing process was hindered by a lack of The book, which was initial y intended to only regional overviews and a mass of sites either newly-ex- discuss the newly-founded hil top settlements in the cavated, trial-trenched or detected during field surveys. wider area of the Roman Empire, has its roots in the On the other hand, new techniques and approaches offer previous century. At that time, the number of sites a better quality of data and facilitate site identification, seemed manageable, particularly in view of the fact that and the numerous rescue investigations and consider- I actively participated in the first large-scale investiga- ably fewer systematic excavations rapidly advance our tions across Slovenia and sporadical y also abroad from knowledge of settlement. the late 1960s onwards. The field surveys, trial trenching and systematic excavations in the following decades, however, have brought an enormous amount of new 1.2 CHANGES IN SETTLEMENT data and consequently new questions; these results have PATTERNS − RECENT RESEARCH rendered the substantial changes of the Roman urban and countryside settlements even more apparent. I The settlement pattern of the previous, Roman pe- soon realised that an appropriate understanding of the riod is well known: it was made up of densely spaced cities newly-founded fortifications also needed to take into as the basic elements of habitation and administration of account the transformation of the cities, the different the Roman state, joined in the countryside by a dense net- unfortified settlements and the markedly altered forts, work of differently-sized habitation and economic units, as all these elements are chronological y and functional y i.e. countryside vil as, as well as small hamlets. Comple- interdependent and form an indivisible whole. menting this settlement structure was a series of forts and 9 Slavko CIGLENEČKI fortresses along the limes and other locations of strategic knowledge of the urban transformation came a more importance. In the second half of the 3rd century, such a intense investigation of the Late Antique countryside, structure gradual y began changing, with the civil wars, where the achievements of antiquity were discovered to uncertainty, economic difficulties and Barbarian threats be very much alive in certain places, persisting even into beginning to affect the heretofore stable settlement. This the early 7th century. In architecture, this is visible in the is particularly true of the area under discussion, which lay numerous houses of a high-quality construction in the at the heart of the Empire and suffered greatest pressures newly-established fortifications and, alongside churches, from incursions from the north. in certain buildings of a public character, that in some The transformation of the settlement together with cases reveal modest echoes of urban design. an array of political and social changes is the most read- Literary sources tell us of people migrating from the ily perceptible feature of Late Antiquity. Cities largely endangered parts of the Empire to safer southern regions lost their original significance. Their last peak in the 4th and to islands. However, the density of the settlements century and at some places also later with the construc- proves that a large part of the population remained in the tion of mighty church complexes still gave the il usion of previously inhabited areas, only moving short distances prosperity, but their real role was in decline across most within small territorial units and primarily from the of the Empire. This process did not take place in al areas lowland to higher-lying locations. simultaneously, nor did it have the same intensity. The For a more comprehensive understanding of the differences between the west part, which witnessed the structural changes in the settlement pattern, it is particu- emergence of Germanic political entities, and the east larly important to be aware of the relationship between part, where the Byzantine state developed more or less the periods of a weakening urban fabric and the con- continuously, were enormous. temporary establishment and development of new forms The settlement in the area under discussion in- of habitation in the countryside. In fact, the parts that cluded a series of fortifications in natural y protected have been better investigated show a fairly significant locations. This is a specific phenomenon unknown in correlation between the periods of weakened cities and such a measure elsewhere across the Empire. These set- either simultaneous or slightly delayed intensification of tlements predominantly represent the reaction of the life in markedly differently organised countryside settle- local population to the frequent incursions of foreign ments. There are even instances where we can discern peoples; the local inhabitants gradually abandoned the first outlines of the unfortified settlements, often in their dwellings along the busy roads and settled on combination with nearby fortified sites. natural y protected elevations, most frequently in areas The main aim of this book is to provide an overview removed from the main lines of communication. In a of the varied and extremely numerous Late Antique large part of the former Empire, fortified sites are thus habitation traces between Ravenna and Constantinople, the most characteristic expression of the contemporary which is an area poorly known and hence poorly under- settlement and most clearly mirror the plight of the stood to a large part of the international archaeological population in the face of powerless central authorities. and historical public. Another aim is to highlight the The form, size and often also structures in the interior phenomenon of fortified hil top sites, which is most of these settlements largely depended on the particular numerously represented and also most characteristic terrain, which resulted in an immense diversity that of this very area. contrasts the earlier, more readily identifiable forms of habitation. This diversity causes a fair amount of dif- ficulties when attempting to identify the basic function 1.3 GEOGRAPHICAL SCOPE of a settlement, as it is frequently only a comprehensive analysis of the architectural remains and small finds In discussing individual sites, I have divided the that enables researchers to propose a more definite study area into two parts, roughly equal in size, by taking identification. In the beginning, the fortifications were into account the line that in AD 395 divided the Empire primarily used as places of refuge, later largely as places into two halves ( Fig. 1.1). This line fateful y cut into the of habitation, while certain spots of surveillance or previously common space, with increasing differences strategic importance hosted a combination of civilian subsequently observed in the settlement pattern, the and military inhabitants who protected the key lines of chronology of the sites and the construction of settle- communication, barrier wal s and so forth. ments. On the other hand, the dividing line also served We can therefore speak of tectonic shifts in settle- as an intermediary between the two halves of the Empire. ment, which only became more apparent in the more re- To obtain a balanced impression, I attempted to deal cent stages of the Late Antique research. In earlier stages, with both halves as evenly as possible. researchers mainly focused on the urban settlements The overview thus deals with an important part of that displayed the degradation and gradual decline of the Roman Empire extending across vast areas between the Graeco-Roman civilisation. With the advancing the capital Ravenna in the west and the metropolis of 10 1. INTRODUCTION Fig. 1.1: Geographical area between Ravenna and Constantinople and the present national borders (Map data @2023 Google, GeoBasis DE/BKG (@2009), Inst. Geogr. National, Mapa GISreal). Constantinople in the east. The Danube delimits this 1.4 CHRONOLOGICAL SCOPE part in the north, the Mediterranean in the south. In the west, it includes the Eastern Alps and part of northern It is difficult to delimit the period of Late Antiquity Italy, in the east it reaches to the Bosporus. with precise dates, which is evident from the series of With the exception of the plains of northern Italy different chronological definitions (cf. Demandt 1989, and Hungary, it is an area that mainly covers the moun- XV−XXI; Cameron 2012, 1−7). Considering that the tainous Balkan Peninsula. Considering Diocletian’s book is dedicated to settlement, I opted to use the administrative division of the Empire, it encompassed changes observable in the destruction or abandonment part of Italia Annonaria, the important province of of certain earlier and the appearance of new settlements Venetia et Histria, the whole prefecture of Illyricum as the criterion. The time frame thus begins in the last and the Thracian part of the prefecture Oriens ( Fig. 1.2). third of the 3rd and ends in the first third of the 7th cen- According to the present-day political division, the west tury, when a major part of the area under discussion part covers north-eastern Italy, large parts of Austria witnessed the abandonment of old settlement and the and Hungary, Slovenia, Croatia, as well as Bosnia and onset of a clearly different settlement pattern. Herzegovina, whereas the east half spans a large part of This is a time between the beginning and the end Serbia, Montenegro, Kosovo, Bulgaria, part of Romania of major shifts in settlement in the given area. It corre- along the Black Sea, Albania, North Macedonia, Greece sponds well with the chronological definitions of Late and the European part of Turkey. Antiquity as proposed on the basis of other criteria; this 11 Slavko CIGLENEČKI Fig. 1.2: Late Roman provinces of the area under discussion (from Cedilnik 2004, binding sheet at the back /back cover). correspondence additional y justifies the chronological part of the Eastern Alps and across the Balkans with the delimitation. It is a time frame that I already used in disappearance of masonry architecture. 1987 when discussing the fortifications settlements in Herve Inglebert examined several urban chronolo- the South-Eastern Alpine area (Ciglenečki 1987a, 9) and gies using a similar time frame and also offering similar one that can be extrapolated to the whole area discussed reasons for the observed changes (Inglebert 2006). in this book. Andrew Poulter (2007a, 5) delimited the period for the Dating the beginning of the time frame to the Balkans with the dates of AD 300−600, though some last third of the 3rd century rests primarily on the first argue that Late Antiquity in Athens already began after important changes in the settlement that occur with the Herulian sack of AD 267 (cf. Snively 2009, 42). In the the destruction or abandonment of certain urban and title of his contribution, Archibald Dunn gives the pe- other settlements, and the simultaneous establishment riod between Galienus and Justinian (Dunn 2004). The of new, mostly fortified sites on elevations. Its end comes recent publications on the settlement in Bulgaria delimit in certain places already towards the end of the 6th and the time frame from AD 284 to 610 (cf. Dintchev 2021b, elsewhere in the early 7th century, but is so radical that 273). Among the different periodisations, we should also the break in the continuity of the Roman way of life mention the chronology of Greece as proposed by John cannot be overlooked or ignored. It is manifested as Bintliff, where he refers to the period to the middle of a profound change of the settlement pattern in a large the 7th century as Late Roman and only that of 650−842 as Early Byzantine (Bintliff 2012, 382−383). 12 1. INTRODUCTION Not only to dates, differences also pertain to the Considerably less is known on the countryside set- names authors from different countries and different tlement across a large part of the former Empire. Many historical-archaeological traditions use to designate the archaeologists and historians, particularly those who dealt time frame. For Greece, archaeologists use the terms Late at least in part with the area discussed here, have acknowl- Roman, proto-Byzantine and Palaeochristian to refer to edged that it is only possible to understand the settlement the same period (Dunn 2004, 541). In the West, similar by also exploring and comprehending the contemporary time frames are also known as the Migration period, countryside. Archibald Dunn, for example, examined the Early Middle Ages and other terms (cf. Ciglenečki settlement in northern Greece and wrote that we can only 1999, 287). In this book, I general y use the term Late understand the settlement system of a region as a whole Antiquity, which is most widely used and also contains and the transition from polis to kastron if we understand a reference to the (albeit often modest) heritage of the the rural context. Furthermore, he mentions that no one Graeco-Roman civilisation (cf. Lavan 2003, VII−VIII). has as yet proposed a functioning model of transforma- For the eastern part and for the sites in the west associ- tive factors either for the Empire as a whole or for the ated with the Byzantine state, I also used the term Early Balkans in particular (Dunn 1994, 71, 75). Wolfgang Byzantine. Liebeschuetz observed that very little is known on the dwellings of most of the urban and even more so the non- urban population. He thought it was not possible to offer 1.5 SOME OVERVIEWS a convincing description of the urban transformation, OF LATE ANTIQUE SETTLEMENT and even less to explain why it had occurred in the first IN THE AREA UNDER DISCUSSION place, without first establishing where the majority of the population lived (Liebeschuetz 1998, 819). The overview In many modern historical-archaeological over- of the literature on settlement from 2004 even talks of a views, the settlement of the area between Italy and the ‘relative invisibility of Late Antique settlements’ and the Bosporus in Late Antiquity is either very incomplete, in fact that the archaeologists in the West focused on the some cases distorted and often simply left out. study of vil as, whereas nothing is known of the dwellings The overviews dealing either only with urban set- of the lower classes (Chavarría, Lewit 2004, 4−5). Simon tlements or only with the countryside are discussed in Ellis also admits that the Balkans remain among the most greater detail below. In the introduction, we take a look challenging regions for summarising Late Antique settle- at several major works that discussed the settlement ment and notes, in the introduction, that the greatest gap in the former Roman Empire, i.e. cities and less well- in the Housing in Late Antiquity proceedings lies in the known countryside, but that have also noted certain absence of a synthetic overview of the Balkans and parts shortcomings in the understanding of settlement as a of the northern provinces (Ellis 2007, 5). These proceed- whole. These overviews are more or less limited to the ings state that our knowledge of the dwellings is skewed well-known Near East (primarily the so-called Dead as certain parts of the Balkans and the Danube Basin have Cities of Syria), Turkey, northern Africa and the better- been poorly researched (Uytterhoeven 2007, 80−81). known European West. Broader regional overviews that Even as late as the beginning of the 21st century, the deal with the Balkans as well as the West are rare (cf. Late Antique settlement in the Empire appeared rather Henning 1987; Curta 2001a; Kirilov 2006a; Milinković simple. In the chapter on rural settlement of the prestige 2007; Ciglenečki 2014). publication The Cambridge Ancient History, Brian Ward- Cities justifiably play a major role in these overviews, Perkins writes of the main difference between east and as they also boast the longest tradition of research. Not west being in the existence of compact vil ages in the for- so long ago, researchers attempting to understand the mer and dispersed settlement in the latter (Ward-Perkins continuity of Roman cities only relied on the brief notes 2000, 327−336). Similarly, Yicak Hirschfeld wrote an in literary sources (primarily mentioning bishops even in overview of the residential structures in which he divided the 5th and 6th centuries), which they saw as supporting the rural settlement to farms in the west and vil ages in the hypothesis of cities existing in an almost unaltered the east (Hirschfeld 2001, 268). Particularly apparent is form. A more realistic image of Late Antique cities was the absence of the whole of the Balkan Peninsula with the only revealed with the modern archaeological investiga- exception of some of the most prominent Late Antique tions bringing to light poorly made or even improvised cities in the basic work by Chris Wickham (2005, 5). The buildings that had previously often been overlooked. western limit of the Roman settlements (vil ages) was, in Many novelties have been observed in the transforma- his estimate, somewhere in Illyricum, while further west tion of the urban character such as a different role of there was only more or less dispersed settlement. He public buildings, reduction of city wal s, appearance of takes this hypothetical division of settlement as marking cemeteries intra muros, modest residential and economic a real east-west division in Illyricum! He thus presumed architecture next to rare public buildings and so forth that Bulgaria, North Macedonia and northern Greece (cf. Liebeschuetz 2001; Wickham 2005; Brogiolo 2011a). belonged to the eastern sphere of settlements (vil ages) of 13 Slavko CIGLENEČKI the 5th and 6th centuries, with the only exceptions being 1.6 CONTENT OF THE BOOK the hil top forts in northern Greece and the dispersed settlement in the Lower Danube region to the end of The overview of the Late Antique settlement pat- the 4th century. Among the fortified hil top settlements, terns first presents cities as the basic element of settle- he only mentions rare cases in Spain and France, partly ment. The cities are divided into three clearly separate also in northern Italy, but also states that we should not groups: cities abandoned prior to the end of Late overemphasise the significance of these castra (Wickham Antiquity, cities inhabited to the end of Antiquity and 2005, 465, 479)! newly-established cities. In conclusion, the character- Having said that, a multitude of sites had already istics of each of the three groups are analysed, and their been known and also published when these overviews transformation and duration compared. were written that no longer allowed for such a crudely The discussion on the countryside begins with simplified settlement. At least from the 1980s, there is the forms that enjoyed a continuos development from a whole host of literature with partial translations or at the Roman period and most frequently lasted into the least detailed summaries in one of the main European 4th century, very rarely longer. Here we are primarily languages that enables a wider public to gain a more talking of the numerous and well-known Roman vil as, comprehensive understanding of the settlement. This individual farmsteads, vici and other, not always clearly literature adds a third important element in the unforti- identifiable forms of settlements. These are followed by fied settlement − vil a dichotomy, namely fortified sites similar forms of undefended settlements that only ap- (primarily fortified settlements). These are very densely peared in Late Antiquity and existed for a brief period. distributed across the discussed area and are gradual y Particular attention is paid to the newly-established also being detected to the west and east. In spite of this fortifications, which are the most significant nov- density observed in different areas, there is a fairly com- elty of settlement in Late Antiquity and also the most mon opinion in literature that the fortified sites are local characteristic expression of the altered political and phenomena. In contrast, the increasingly larger-scale socio-economic situation in this area. Their existence and frequent field surveys and excavations have revealed is short-lived in certain areas, while they have a longer they formed an important settlement pattern that is the duration or repeated occupation in the more exposed most characteristic indicator of the transformation that parts. Substantial changes in the living conditions not- Roman settlement witnessed. withstanding, they still exhibit the civilizational achieve- In 2001, Florin Curta published the first detailed ments of Antiquity and play a much more important and comprehensive insight into the settlement of the role in the history of a large part of the former Roman eastern half of the area under discussion (Curta 2001a; Empire than believed until recently. id. 2001b). He took a detailed look at cities, but also The presentation and analysis of the settlement pat- numerous settlements in the countryside that included terns is followed by a diachronic view of the transforma- the newly identified fortified hil top sites as an element tions both in the cities and in the countryside. The more of the settlement. Focusing primarily on the finewares important phases of establishment, as well as abandon- as the consequence of the annona and with it the state ment or destruction of settlements are indicated, as are supply of the garrisons led him to the conclusion that the breaks that delimit the important phases of settlement. many fortified hil top settlements were army posts. He Special chapters are dedicated to the settlement structures upgraded this hypothesis in later papers and saw, in his and individual examples from the parts of the Empire most recent contribution, the majority of such sites only beyond the area that is the focus of this book. as military forts (Curta 2017). Only considering the At the end, I briefly summarise the complex im- fineware and coins without the vast amounts of other age of settlement patterns and their comparison on finds (mainly coarseware and a multitude of metal arte- both sides of the dividing line set in AD 395, but also a facts) when interpreting the fortified hil top settlements comparison with the settlement structures beyond the does not lead to a comprehensive understanding, as most discussed area. This adds a range of newly-established latter finds reveal an autarchic nature and consequently posts, primarily fortifications, to the previous simpli- predominantly civilian population. fied dichotomy of settlement patterns only composed Certain categories of settlements, particularly cities of undefended settlements in the east and villas in and Roman vil as, have been discussed in many publica- the west. tions and their more detailed presentation is not sensible The sites in the overview are arranged geographi- here. For these, I only highlight the more prominent cal y, from west to east and from north to south, by also studies for individual categories. A lengthier discussion taking into account their similarities in character. is offered on the development and the main centres of research in the group of newly founded fortifications, of which least is known. 14 1. INTRODUCTION 1.7 METHODOLOGICAL REMARKS have revealed the existence of wooden buildings at some sites (e.g. Tinje above Loka pri Žusmu, Veliki vrh above Discussing the area separated in AD 395 by a sharp Osredek pri Podsredi). For the sites not archaeological y dividing line has an inherent problem in that the divi- investigated due to financial or personnel constraints, sion brought a myriad of consequences and also greatly these brief observations have significantly contributed to influenced the settlement (cf. Liebeschuetz 2001, 400). a better understanding of countryside settlement. This influence is observable not only in Late Antiquity, An important addition in the research of countryside but is strongly felt even in the work of modern researchers settlement is the intensive field surveys, particularly in who − particularly in broader overviews − deal primarily Greece and Bulgaria. However, the mass of completely with the east half of the area under discussion. An added new results brings the problem of evaluating the vast difficulty is that the research is predominantly limited amounts of new data, which do not necessarily enable an with modern state borders and the results only rarely accurate interpretation of the different settlement traces juxtaposed with those from neighbouring countries or in the absence of additional research (on the results, as even wider regions (cf. Wickham 2005, 4). well as the drawbacks and methodological problems as- This is why it seemed especial y important to offer a sociated with the regional surveys, see Liebeschuetz 2001, balanced presentation of the west half as wel , which has 388−390; Sanders 2004; Poulter 2007b, 41−46; Bintliff previously been treated rather cursorily with the excep- 2007; Veikou 2013, 128). tion of Italy and in part Hungary. The line dividing the Alongside the better known and documented inten- Empire, running almost precisely across the middle of the sive surveys, there are also extensive surveys of varying area under discussion, is considered inasmuch as it makes intensity and duration conducted across large parts of the the discussion clearer. The first to be considered are the discussed area. Given the terrain that is for the most part settlements in the west half and later those in the east half. very difficult and does not always enable results of the The differences between the two are highlighted in the same quality as those in the narrow Mediterranean belt, conclusion of the chapters on cities and the countryside primarily due to the vegetation cover, we need to take into settlement (see Chapters 2.6 and 3.4). consideration the fairly uneven level of investigations in Before presenting individual categories of sites, it is different regions. Moreover, the extent of investigation is sensible to say a few words on the different methods of often the consequence of the zeal of individual research- gathering data. They are the result of systematic research ers who gathered data in frequently treacherous terrain, and trial trenching, as well as numerous field surveys. whereas we lack large-scale systematic projects that would It should be noted that the objective of systematical y use the same methodology to cover areas across differ- investigating many of the sites in their entirety, particu- ent modern states. A particularly prominent example of larly those of modest remains such as refuges or briefly the latter is the vast fieldwork that Ivan Mikulčić and his occupied local fortifications, will remain unattainable colleagues conducted in North Macedonia. Through in- for some time. The limited knowledge on the less typi- numerable extensive surveys, they succeeded in revealing cal forms of sites is also the consequence of the fact that an extremely rich and diverse Late Antique settlement, investigations mainly focus on the more revealing and which offers a good insight into the extent and intensity most exposed elements such as settlements with masonry of the settlement transformation in the central Balkans, buildings, which are also best preserved. Missing in such as well as the possibility of comparing the results with an image is the smaller settlements, refuges, as well as the neighbouring areas. The extent of investigations in North more modest residential architecture, which in turn skews Macedonia in comparison with the neighbouring areas is our perception of the settlement in general (cf. Ciglenečki clearly visible on the map of Late Antique fortified settle- 1994, 240−241; Cameron 2012, 150). ments ( Fig. 3.307) (Milinković 2007, 170, Fig. 6). A fortunate circumstance is that the Late Antique The poor knowledge of many of the sites in the coun- remains in the Eastern Alpine area, the central Balkans tryside is due to their location in remote and uninhabited and the coastal areas of the Mediterranean are frequently areas, but also to the fact that − in contrast with cities and excellently preserved due to their remote locations and lowland settlements − they are less exposed to modern complete abandonment of some of the settlements. Many interventions into the landscape. As a consequence, they allow us to discern the degree of fortification, number of only rarely witnessed rescue excavations, which nowadays buildings and their distribution, often also their shape completely predominate in more readily accessible and and size, already in relief. Examples of such settlements hence more endangered lowland areas. in Slovenia are Gradec near Prapretno, Ajdovščina above In the overview, I strive to present all the important Rodik, Tonovcov grad near Kobarid, Korinjski hrib forms of settlement (cities, settlements and forts) across above Veliki Korinj. Exacavations in these settlements the area under discussion. However, there are inevitably only show minimal deviation from the ground plans gaps caused in some places by poor accessibility, in others drawn on the basis of relief observations. In addition, by a perilous political situation, in part also the (stil ) hin- the different earthworks damaging archaeological sites dered accessibility to local literature. I therefore present 15 Slavko CIGLENEČKI the better investigated, documented and preserved sites to all the major groups of sites and one mentioned in more precisely. In this I greatly rely on the archaeologi- literature several times, but rarely solved in a satisfactory cal field surveys and visits to the sites that I have been manner. In this book, I first acknowledge the problem in conducting since 1975 across former Yugoslavia and general terms in the introduction and tackle it in greater neighbouring countries. detail in the appropriate passages below. Alongside the plans of cities and other settlements, For urban settlements, it is frequently uncertain I pay particular attention to the photographs aimed at whether they should be termed ‘city’ or ‘town’ (for more offering the reader a rough idea of the predominantly details on the problem of the urban character and the poorly known settlements and partly also of their state of notion of cities/towns, see Wickham 2005, 591−596). preservation. The need for accompanying photographs is Next is the unclarity regarding the naming of an urban lesser for cities and undefended countryside settlements settlement in the second half of the 5th and in many that mainly lie in the lowland. In some cases, photographs areas also in the 6th century when the classic urban are only added to il ustrate the text and primarily in the character is completely absent, but a settlement is still cases where there are substantial ruins visible on the sur- marked as a city in literary sources. In the final phase of face. For the fortified sites, it seemed logical to use photo- Antiquity, most commonly in the 6th century, there is a graphs to show their position within the landscape, their problem of how to name the newly-established urban natural y protected location and the remains of buildings settlements, with the term civitas also used for many and defensive wal s wherever they were sufficiently visible fortified settlements. on the surface or the vegetation permitted it. Among the unfortified settlements as wel , there The use of new prospection techniques and geo- are numerous forms that cannot be reliably interpreted physical methods, in some areas even LiDAR images, has given the limited degree of investigation (Roman vil as, in many ways facilitated and accelerated archaeological vici, farmsteads, roadside stations and others). There research. The Late Antique architecture and construc- is extensive literature on the subject, but definitions tion techniques are fairly well known. Numerous studies vary and different authors use terms rather arbitrarily. of the high-quality imported tablewares coupled with Roman vil as in particular are only rarely comprehen- regional analyses of coarsewares already allow a more sively known; in most cases only the main residential precise dating of the recovered remains. Together with complex is excavated and it is unclear whether it was at the rare coin and metal finds, this enables us to identify all associated with a production complex. In this book, the basic habitation phases by working with data from the countryside settlements where evidence indicates surface surveys. large units with both residential and production parts In the recent three or four decades, the results of are called villas, whereas the smaller complexes are these types of investigation have been joined by numerous simply marked as farmsteads. Having said that, the poor metal finds that (mostly unauthorised) metal-detectorists or insufficient research makes it impossible to make a brought to the museums. The problem of unauthorised clear distinction between the two and the distinction is metal detecting is most acute at hil top sites removed from in some cases arbitrary! present-day settlements, not readily accessible and hence The newly-established fortifications pose a par- most exposed and often at the mercy of such ‘explorers’. ticular set of naming problems. In earlier literature, It is a problem that archaeologists across the globe have the term refuge was used (far too) many times without been unable to adequately solve and one that particularly defining the archaeological criteria for such an identifi- affects the area under discussion that is strewn with Late cation. Its overuse caused the permanent settlement of Antique settlements in the remote and largely mountain- Late Antiquity to become blurred or unrecognizable. ous terrain! In addition to refuge, the newly-established fortified The discussion does not include the cemeteries that settlements were identified with a whole host of other serve as indirect evidence of the existence of settlements. terms, which raises another methodological issue (cf. I only mention cemeteries where they offer additional Ćurčić 2010, 10). To il ustrate, the names commonly support to defining the chronology of habitation phases used for the variously-sized fortified settlements in as established through investigations within settlements the Eastern Alps are castra or castel i (Bierbrauer 1985; and to defining the nature of individual sites. Brogiolo, Gelichi 1996, 7−8; Geuenich, Zotz 2008), but also descriptive terms (fortified hil top settlements, mili- tary posts, refuges: Ciglenečki 1987a), those in North 1.8 TERMINOLOGICAL REMARKS Macedonia are known under ancient names oppida, oppidula, frouria, castela, but also as refuges (Mikulčić A problem unto itself is naming the different forms 2002), those across the border in Greek Macedonia are of settlements. This is in part due to the use of ancient termed fortified military sites and communal/civilian terms and in part to the insufficient investigation of walled sites (Dunn 2004, 551, 565), those in Serbia are individual categories of sites. It is a problem pertaining vil ages and military forts (Milinković 2008, 556−557), 16 1. INTRODUCTION those in Bulgaria are descriptively marked as fortified is only implicit, it is necessary to take into account that settlements, military fortifications, fortified refuge sites the decisive factor is the natural protection of a site and (Dintchev 2007). In spite of the varied nomenclature and not merely the high altitude. interpretations, however, there are great similarities and differences in structure and function often difficult to pinpoint. The different names stem from the different 1.9 OPEN PROBLEMS, archaeological traditions that have become ful y ac- SHORTCOMINGS cepted in different areas. All the ancient terms are ap- propriate where they are based on the use of these terms Several decades ago, any attempt at such a synthetic in literary sources. However, there are several instances overview was greatly hindered by a lack of basic litera- in the archaeological publications − particularly in the ture, reports and regional syntheses from the area poor east − of some of the ancient terms indicating different in investigations and rich in settlement remains. This functions. For example, many authors using castra and situation changed in recent times with the exponential y castel a presuppose a primarily military nature, which increasing number of investigations and publications of contrasts with the understanding of these terms in the Late Antique remains, from Italy in the west to Bulgaria west. The problem is highly complex, as not even the and Romania in the east. A balanced overview of the Late Antique authors reliably distinguished between settlement in such a vast area still needs to face a highly different categories of sites (cf. Claude 1969, 201−202). varied state of research, for both fortified and even more In ancient Greek sources, for example, the term kastron so the unfortified settlements that are more endangered marks a fortified settlement that differs from others in and also less well known due to their location in inhab- that it is an administrative and military centre not to be ited lowland areas. Only a planned recognition of their equated with smaller forts ( kasteli, pyrgoi, frouria and remains will give an integral settlement pattern. so forth) (Dunn 1994, 78). In my work, I was greatly hindered by a lack of In this book, I strive to use names (as many authors knowledge of the Hungarian, Romanian, Albanian before me) that are as descriptive and neutral as possible and Greek languages. Hence, I could only consider with the aim of facilitating the recognition of identi- the literature from these countries if published with cal settlement patterns and highly similar settlement translations or at least comprehensive summaries in structures in the differently named settlements across other languages. It is the same linguistic barriers that the whole area under discussion. Wherever the research prevented an even consideration and evaluation of results permit, I add a specific name or a more clearly sites in the earlier synthetic overviews. Regional studies identified function to a site. accompanied by catalogues of sites in the area are rare The problem is compounded by the occasional y and partial y already outdated (cf. Ciglenečki 1987a; ambiguous understanding of the term fortified hil top Brogiolo, Gelichi 1996; Dintchev 1999; Mikulčić 2002; site, which has a double meaning in most modern Mulvin 2002; Busuladžić 2011). languages: on the one hand, it stands for a settlement This book draws attention to the problem of dis- built on an elevation and, on the other hand, it implies tinguishing between the military and civilian functions a location in a mountainous or high-altitude area, i.e. at of many of the newly-established fortified sites on more greater absolute altitudes. Examples of this can be found than one occasion. From the Tetrarchy onwards, this is in the German ( Höhensiedlungen or Höhenbefestigun- an extremely challenging undertaking, as cities, but also gen), most Slavic ( utrjene višinske postojanke, visinska smaller fortifications may have combined both functions utvrđenja, visinske naselbine), in Italian ( siti di altura), (see Claude 1969, 219−222; Ćurčić 2010,137−138; Sar-Spanish ( yacimientos de altura) and partly in the French antis 2016, 190−191). It is only with great difficulty and language ( établissement de hauter). Consequently, some very rarely that we can identify solely military posts in authors use the term primarily to denote locations on Late Antiquity, particularly in the 5th and 6th centuries. tops of natural y protected elevations, whereas others Least is known of the military presence in the newly- use it for settlements in the high-altitude areas. In my established fortifications that had different functions publication of the fortified hil top sites in the Eastern in different phases of Late Antiquity (cf. Ciglenečki Alps, I therefore noted in the introduction that the term 1987a, 110). Even the forts on the limes show a blurring Höhenbefestigungen is used specifically for the sites between purely military posts and civilian settlements that take advantage of their relative height above the that occurred in the 5th and 6th centuries (cf. Popović, surrounding area for defensive purposes, but were not V. 1984, 280; Špehar 2010, 154, 173). The identification always established in areas of high altitudes (Ciglenečki of military posts is made more difficult by the fact that 1987a, 9). Much more appropriate are the terms hilltop these and the fortified civilian settlements are both site in English or sites perchés in French, which mark all constructed in natural y protected locations and hence kinds of natural y protected locations regardless of their no longer display the regular layouts characteristic of absolute altitude. For the languages where this duality the earlier Roman period. 17 Slavko CIGLENEČKI For the newly-established fortifications, there is of Montenegro, while investigations in Kosovo have also the problem of identifying those exhibiting the intensified in recent times. vestigial urban features and a more prominent status With the many known countryside sites comes the within smaller regions; a poor state of research for most problem of presenting them on maps intended to reliably of these sites does not allow for reliable conclusions and show the density of settlement. The problem lies in the size is merely one of the criteria for distinguishing such lack of information on the chronology for many of the ‘cities’. Even in the most systematic and best documented sites, as some were inhabited in all phases of Late An- overview of the Late Antique fortifications in North tiquity and others only in some of them. Furthermore, Macedonia, it is sometimes difficult to discern whether there are numerous cases where sites are only shown individual sites are cities or merely large settlements. on maps under a general designation of Late Antiquity Recent research has confirmed the surface observations and therefore misleading. The situation is slightly better and upgrade them with many other elements of architec- in the eastern part, where several maps already exist, ture that were not recognised during field surveys and featuring more precisely identified sites as Early Byzan- measurements (cf. the plan of the fortified settlement tine fortified posts from the 6th century (Mikulčić 2002, at Vinica, published in Mikulčič 2002, Fig. 147, and the App. 2; Milinković 2008, Fig. 1; Dintchev 2007, Fig. 1). It satellite image1 of the same site). Similarly difficult is the will only be possible to present reliable distribution maps recognition of the newly-established central settlements of individual types of countryside sites when at least the in Bulgaria and Serbia, where future investigation may basic phases of their existence will be clearly identified. reveal an urban status. Transcribing geographical and personal names and The discoveries of ever new fortifications confirm even citing literature for the east part is a problem not the altered Late Antique settlement pattern across a large always solvable to a satisfactory manner (cf. Ćurčić 2010, part of the area under discussion and indicate that we 12)! I transcribed the personal names and place names may expect many new additions in this very segment in the text, but the quotation in the literature remains of settlement. Here, we only need to think of the newly- in the original spelling. found hil top fortifications such as Tonovcov grad near All the above-stated difficulties and reservations Kobarid (1994), Lobor (1999), Veliki Sikavac (2011), notwithstanding, I attempt to offer a synthesis that wil Gornji Bakovci (2013) or Harilaq (2017). In Bosnia and be upgraded and corrected by more detailed regional Herzegovina, a part where reliable Late Antique settle- studies to come that would al ow for new and more ments were among the first to be presented, research comprehensive syntheses to be written. The text below ground to a halt due to political issues and the war fol- presents the settlement patterns in the outlined part of the lowing the breakup of Yugoslavia, but it is also the very Roman Empire in which cities are discussed alongside the area where numerous new findings can be expected. newly-established countryside settlements, with special The archaeological lexicon of Bosnia, published before attention paid to the latter that have been less wel known the war, indicates a multitude of largely unexplored sites in the past, but are crucial for a proper understanding of (Čović 1988). Very little is known on the whole territory the settlement in the final period of Antiquity. 1 Map source: Google earth V 7.3.2.5776 (24th April 2023), North Macedonia 41° 52’ 33’’ N, 22° 20’ 14’’ E, camera 971 m, altitude 446 m, Maxar Technologies CNES / Airbus Data SIO, NOAA, U.S. Navy. NGA, GEBCO. 18 2. CITIES 2.1 ALTERED APPEARANCE mention the monograph on the Late Antique cities in OF LATE ANTIQUE CITIES Bulgaria by Velizar Velkov (1967; reprint 1977). Mate Suić published a concise historical and archaeological The Roman cities in the area under discussion overview of the cities on the eastern Adriatic coast and underwent different fates in Late Antiquity. There is their transformation in Late Antiquity (Suić 1976a; re- a growing interest in this topic, more precisely in the print 2003). These early monographic publications were transformation of the Late Antique urban settlement, followed in the 1980s by numerous articles discussing which is mirrored in the numerous studies publishing Greece (Dagron 1984; Spieser 1984; Bavant 1984; Sodini the results of the systematic and rescue investigations 1984), Serbia and to a certain degree also the rest of the in these cities. Many syntheses have already been writ- Eastern Empire (Popović, V. 1982), North Macedonia ten on the subject (cf. Claude 1969; Liebeschuetz 2001; (Wiseman 1984; Mikulčić 1986) and Slovenia (Petru Curta 2001b; Wickham 2005: Saradi 2006 and others), 1982; Ciglenečki 1987a). Later came a multitude of new which reveal a predominantly historical viewpoint that papers, including those specifical y addressing the altera- takes into account archaeological evidence. There are tions to the Roman city in Late Antiquity (e.g. Poulter also many proceedings, which tackle a variety of aspects 1992; Dintchev 1997; id. 1999; id. 2012; Mikulčić 1999; pertaining to Late Antique cities, though individual Ladstätter 2002; Kirilov 2006a; Snively 2009; Milinković studies in them are largely limited to outstanding case 2015; Ivanišević 2016). Much is known of the fate of the studies (e.g. Rich 1992; Brands, Severin 2003; Ghilardi cities in northern Italy, where the systematic research of et al. 2006; Krause, Witschel 2006; Burkhardt, Stichel Late Antiquity in recent decades has yielded a wealth 2010; Christie, Augenti 2012). Many of them lack the of valuable data, particularly concerning the transfor- examples of cities that are less known, but numerous mation of housing (overview in Brogiolo 2011a). Also and significantly diverse from areas that played a key worth mentioning is the Bulgarian systematic presen- role in Late Antiquity. The eastern Alps, Pannonia and tation of the Roman cities and their changes in Late the Balkan Peninsula, for example, are only represented Antiquity (Ivanov, R. 2012). For the areas forming part to a limited extent with the exception of few prominent of the provinces of Noricum and Pannonia, we should and better investigated centres such as Athens, Corinth, mention the fundamental work on the local autonomous Thessaloniki and Philippi in Greece, Stobi and Heracleia Roman towns, though Late Antiquity is only cursorily in North Macedonia, Salona in Croatia, Butrint and Byl- represented (Šašel Kos, Scherrer (eds.) 2002; iid. 2003; lis in Albania, Nicopolis ad Istrum, Novae and Iatrus in iid. 2004). Bulgaria, declining Sirmium and an exemplary newly- The literature on Late Antique cities uses very dif- founded city at Caričin grad (presumably Iustiniana ferent names for the urban agglomerations that show Prima) in Serbia. Overviews with such lacunae thus a heavily altered structure of the former Roman cities. frequently paint a misleading picture that takes us to We come across terms and phrases such as ‘urban settle- generalised conclusions drawn from the better-known ments’ (Suić 1976a, 239; Snively 2008), ‘cities−kastrons’ centres and often heavily reliant on literary evidence. (Suić 1976a, 241), ‘post-Roman city’ (Christie 2000, The data from literary sources are modest and 293), ‘semi-urban settlement’ (Snively 2008), ‘new unrevealing for most cities in the area under discussion, cities’ (Rizos 2017b), ‘changing city’ (Cameron 2012, mainly limited to references to the large or prominent 152), ‘quasi-urban centres’ (Dintchev 2007, 482) and cities, or lists of the late urban agglomerations (cf. Pro- others. In my book on the Late Antique hil top sites in copius and Hierocles). More reliable are the regional the eastern Alps, I used the term ‘miniature city’ for the syntheses based largely on archaeological evidence. Of densely inhabited large fortified settlements (Ciglenečki the first comprehensive regional syntheses, we should 1987a, 114). Slobodan Čurčić introduced the category of 19 Slavko CIGLENEČKI ‘miniature cities’ for examples such as Split and Romuli- 2.2 HOW TO IDENTIFY ana in his overview of the Late Antique architecture in A LATE ANTIQUE CITY? the Balkans (Ćurčić 2010, 23). This myriad of names mirrors the difficulty in appropriately designating cit- The definition of a city in Late Antiquity and the ies that differ from the classic Roman city in their basic criteria for its identification are tackled in the publica- features and do not always allow for a reliable distinc- tions mentioned above, as well as in many others that tion between a ruralised city and a large countryside discuss what remained of the old cities and what the settlement (general y on the problem of naming cities, newly-established cities looked like. The limited investi- see Wickham 2005, 592–594). gation or outdated data available on this topic frequently Also problematic are the different attempts at prevent us from reliably assessing whether the remains a more detailed differentiation of cities according to of former cities or the large new settlements can be the terms employed in ancient texts. Ancient authors identified as urban settlements. The great diversity of used different terms in a very liberal way, making the Late Antique settlements rather dictates a careful exami- archaeological identification of individual categories nation of the distinctive features. Below, we summarise exceedingly difficult. Using Procopius, Gilbert Dagron the basic observations on a city in Late Antiquity and (1984, 8) indicated different categories of cities and outline the criteria that enable at least a rough identifica- vil ages in the Early Byzantine Illyricum. Even before tion with the emphasis on the characteristics particular him, Ivan Mikulčić attempted to correlate the names to the area under discussion. from ancient texts with the agglomerations established Dietrich Claude (1969, 195–229) noted the main in North Macedonia (Mikulčić 1982a, 58–64; id. 2002, features and difficulties in identifying Late Antique cities 51–52). Having said that, it is very challenging or even early on. His work was dedicated to the Early Byzantine impossible to identify such names in the archaeological cities in the 6th century, though the changes he outlined remains of the different and differently investigated sites, in a large measure also pertain to the earlier cities of the as has been noted on several occasions (e.g. Claude 1969, 5th and partly even the 4th century, and are not limited to 201–202; Milinković 2007, 180–187; Ćurčić 2010, 182). the Byzantine area. As many before him, Claude stressed Particularly in the continental hinterland, it is frequently the absence of curiae, with their role replaced by a new difficult to distinguish between cities, forts and large order with the bishop at the head. Studying the descrip- settlements as the boundaries between them are unclear tions of cities in the ancient texts, he observed that a city (Milinković 2007, 187). Most authors therefore tend to was primarily marked by physical characteristics (size, avoid ancient terms and rather strive to use neutral de- city wal s, number of inhabitants, buildings), which scriptions or terms to refer to cities. Many use the term caused that even large kastra or vil ages could be seen polihinia for small cities, though without specifying what as cities; numerous casteli became poleis. In addition, a exactly a ‘small city’ is. large number of fortified sites hosted a bishop, which Researchers often very tentatively describe in- equated them with cities in ecclesiastical terms. He dividual presumably newly-founded cities that come also emphasised the significant discrepancy between with insufficient evidence for a reliable identification. the status of a city as noted in the ancient texts (mainly An example is Romuliana (Gamzigrad), a palace of the Procopius, Hierocles and the Anonymous Geographer of Emperor Galerius that transformed into a large fortified Ravenna) and the archaeological evidence. A city was no settlement or city after his death. It is described as a ‘large longer defined by its legal status, but rather the buildings vil age’ (Popović, V. 1982, 556), ‘urban settlement inside and defensive architecture. He noted fortifications as the a former palace’ (Wolff 1991, 307), ‘miniature city’ and element best defining a city, which is particularly appar- ‘episcopal centre’ (Ćurčić 2010, 24, 139), ‘independent ent in the descriptions of cities as found in the ancient settlement anticipating the concept of a self-sufficient texts. All this contributed to an ambiguous notion of a medieval city’ (Petković 2011, 126), as well as ‘ruralised city, which led him to posit that already by the first half of artisanal and ecclesiastical centre’ (Milinković 2015, the 6th century there was an absence of criteria to clearly 250). This example clearly illustrates the difficulties distinguish between poleis and kastra. With their solid in defining the status of a presumed Late Antique city fortification wal s, cities also became military centres. caused by a lack of reliable criteria. After Claude, several other authors tackled the subject in a systematic manner and using prominent case studies (primarily Liebeschuetz 2001; Wickham 2005; Saradi 2006; Brogiolo 2011a). They also considered the growing amount of archaeological evidence, which represents the most important source of information for the period. An author who contributed important observations on the transformation of cities on the eastern coast of the 20 2. CITIES Adriatic is Mate Suić (1976a, 227−251, second (updated) entertainment and other public buildings. The state pur- edition 2003, 341–375). In urban areas, he paid great portedly encouraged their construction even in areas of attention to the city wal s, their particular features and a less obvious danger of incursions (Saradi 2006, 469). the novelties of construction (proteichisma). For several Researching Late Antique cities in Serbia led Mi- cities, he presumed an interior line of defence (acropolis). hailo Milinković to present in greater detail the examples He outlined the introduction of the elements of rural of disintegration and ruralisation, while he used the economy and with it rural architecture into the urban large settlement at Jelica as an example to discuss the environment, i.e. ruralisation of cities. In the settlement formation of new cities (Milinković 2007, 179; id. 2010; pattern, he observed that several coastal cities became id. 2015, 143−190). He emphasised the mass influx of both civitas and castrum (phenomenon he called kastri-countryside population into the cities in the second half zam in Croatian). He noted that, as a result of barbarian of the 6th century, where they sought shelter behind city raids, many settlements were affirmed as new urban wal s. He reiterated the caution associated with defining formations, reaching their peak of prosperity after the the settlements of Late Antiquity based on the terms Gothic Wars. He also emphasised that cities in this period used by ancient authors. became refuges for the local population. For many new The text below takes a concise look at as many urban formations, he observed a development within the characteristic, but diverse and sufficiently known cities earlier military forts, the natural and artificial defences as possible that are distributed across the whole area in which enabled a rapid transformation into small cit- under discussion with the aim of offering a balanced ies (e.g. Split, Gamzigrad, Mogorjelo). He indicated picture of the urban transformation based primarily on numerous new settlements of an urban character along archaeological evidence. Even a very small city with its the Adriatic coast, but they are still poorly known due specific character adds to our knowledge on the com- to continuous occupation and insufficient investigation. plex and often poorly understood period that is Late In his first major overview of the newly-formed Antiquity. The boundary separating a city from a large hil top forts around the city of Scupi, Ivan Mikulčić settlement or fort was heavily blurred even in the time (1982a, 60–63) studied the archaeological evidence to of their formation. reveal that the large casteli towards the end of Antiquity The main criteria adopted here for identifying the became small cities – Late Antique oppida ( casteli–cit-urban features of a settlement are the following: size, ies). They were characterised by a division between the defensive and church architecture, elite housing and acropolis and the lower city, as well as the presence of mentions in ancient texts. These criteria can only rarely unfortified suburbs. They were presumably the seats of be clearly defined, hence it is often difficult to distinguish civil and ecclesiastical administration, but also held arti- between urban and non-urban settlements. This is also sanal and commercial facilities. He classified Čučer, Taor the reason why some of the poorly-known potential ‘cit- and Vodno in this category. When later dealing with ies’ are discussed in the chapter on the newly-founded the whole territory of North Macedonia, he expanded fortified countryside settlements (see Chapter 3.3). these findings on several other sites and also attempted Among the above-defined criteria, one of the most to identify newly-founded cities using passages in Pro- prominent and readily observable for the group of copius and Hierocles (Mikulčić 1986, 269–271; id. 2002, newly-formed cities is their size. Examining the lists that 50–58). He termed these newly-formed Byzantine cities Byzantine geographer Hierocles compiled, researchers in oppida and oppidula. Analysing the architecture and Albania, Bulgaria and Serbia established that settlements small finds led him to observe a lack of a clear dividing covering a surface in excess of 3–4 ha are defined as cities line between military and civilian sites. in his work Synekdemos (Baçe 1976, 70; Dintchev 1999). In contrast with large cities, which are clearly In addition to this undoubtedly significant criterion, I separate from their agrarian hinterland, studies in recent considered three others deemed most important by both decades often mention ‘central settlements’ as newly- ancient and modern-day authors, namely the construc- formed cities where a church in their interior serves as tion and maintenance of city wal s, the presence of Early the strongest indication of a central role (cf. Wolff 1991, Christian architecture or ecclesiastical centres, as well 312; Milinković 2007, 180; Ciglenečki 2011a). as the presence of reliably dated and numerous late Using the examples of cities in the dioceses of housing. For the last element, I also took into account Thrace and Illyricum, Ventzislav Dintchev (1999) at- its diversity in both layout and size that in some cases tempted to roughly determine the size that separates a clearly points to social differentiation. This is confirmed city from a large settlement; in his opinion, the size of (bearing in mind the possible ethnic changes) by the 3−7 ha indicates a small city. differences in the construction techniques ranging In her model of new cities, Helen Saradi noted from masonry wal s bound by high-quality mortar to that these first and foremost ecclesiastical and military clay-bonded drywal s, wattle-and-daub, wooden con- centres were without the basic features of Graeco-Roman structions, as well as sunken or semi-sunken buildings. urbanism, i.e. devoid of the agora, colonnaded avenues, Wherever the scale of investigations permits, it is also 21 Slavko CIGLENEČKI Favianis Ovilavis Vindobona Iuvavum Lauriacum Aelium Cetium Carnuntum Aquincum Scarbantia Mursel a Környe Aguntum Savaria Gorsium Teurnia Flavia Solva Virunum Salla Kirchbichl Ságvár Sabiona Poetovio Keszthely-Fenékpuszta Iulium Carnicum Carnium Forum Iulii Celeia Alsóhetény Tridentum Concordia Aquileia Emona Brixia Aquae Iasae Altinum Tergeste Sopianae Castrum Neviodunum Verona Gradense Capris Andautonia Siparis Mursa Siscia Parentium Bosar Cibalae Pola Nesactium Argyruntum Ravenna Blagaj Asseria Iader Varvaria Municipium Riditarum Salona Split Lisina Pharia Cities abandoned before the end of Late Antiquity Cities with continuity 0 100 km Cities newly-founded in Late Antiquity Fig. 2.1: Cities in the western part, discussed in the book. possible to consider the presence of different workshops, periods, while their renovations and adaptations are not which become an important element of the late, largely always precisely dated. Numerous cities witnessed a size self-sufficient cities. The recovered artefacts enable ever reduction in Late Antiquity, though we know very little new observations on the short- and long-distance trade, of the habitation extra muros. In most investigations of local production, military presence, possibility of the city urban structures, particular attention is paid to church dwellers to find employment in the immediate vicinity buildings that are chronological y diagnostic through and so forth. One of the more important criteria for the characteristic architectural elements, decorations, identifying a newly-formed city is their position within donor inscriptions, burials and small finds. However, individual territorial units that included other, smaller they only il uminate one, the most apparent and hence and partly dependent settlements. The observations are most frequently researched facet of urban development. juxtaposed with the passages in ancient texts where the Without understanding the complete city interior, which latter enable an identification of newly-formed cities. frequently hosts architecture of non-durable materi- The sites, for which ancient texts merely mention the als, church buildings alone can offer a skewed picture. presence of a bishop are also taken into account. In a considerable share of the cities under discussion, Defensive wal s and church architecture as im- churches are also a good indicator of their duration; in portant components of Late Antique cities occasional y many cases it is their destruction and abandonment that distort the understanding of cities and do not always indicate a decline of a city! enable a reliable representation of a city as a whole. In the desire to better understand the dynamics and Many defensive wal s were already constructed in earlier extent of the changes that affected cities in Late Antiq- 22 2. CITIES Dinogetia Sirmium Troesmis Bassianae Histria Singidunum Viminacium Tropaeum Traiani Durostorum Zaldapa Novae Ruse Romuliana Asemus Iatrus Abritus Acrae Jelica Nicopolis Shoumen Naissus ad Istrum Carevec Izbičanj Tzoides Mesembria Caričin grad Ulpiana Serdica Deultum Cabyle Doclea Hissar Pautalia Čučer Konjuh Philippopolis Scodra Scupi Taor Vinica Kalaja Salemydessos Lissus Bargala Constantinople Dyrrachium Castra Nicea Stobi Heracleia Philippi Scampis Lychnidos Heracleia Lyncestis Thessaloniki Amphipolis Diocletianopolis Pydna Castoria Byl is Louloudies Dion Buthroton Demetrias Nicopolis Thessalian Thebes Tanagra Thespiae Athens Corinth Sparta Cities abandoned before the end of Late Antiquity Cities with continuity 0 100 km Cities newly-founded in Late Antiquity Fig. 2.2: Cities in the eastern part, discussed in the book. uity, we should consider the rare maintained houses of has been rarely appropriately recognised. Two excep- the elite, but also the frequently overlooked remains of tions are therefore all the more important, namely two modest dwellings. With the focus placed predominantly articles written by Vladislav Popović on Sirmium (1982) on the better preserved and richer church architecture of and Ivan Mikulčić on Stobi (1982b). They brought the central Balkans, the significance of private housing the first reliable results of systematic excavations that 23 Slavko CIGLENEČKI highlighted the radical transformation of the Late An- 2.3 CITIES ABANDONED BEFORE tique housing in this part of the Balkans. These great THE END OF LATE ANTIQUITY changes and subdivisions were also discussed in the first overview of Roman housing (Ellis 2000, 110–212). The This group comprises cities that saw the last peak increasingly numerous modern excavations bring ever of construction in the 4th century, followed by gradual new information on the subject, particularly in the cit- abandonment and end that for most already came ies with continuity (e.g. Brescia, Verona, Scupi, Serdica, in the first half of the 5th century. Only few examples Stobi and others). In this book, I strove to include as persisted longer and only in a highly reduced and im- much as possible the evidence recently acquired on the poverished form – certainly no longer as urban entities. newly-constructed buildings, their encroachment on The question here is how to classify the cities of which public space and earlier buildings, all of which indicates the inhabitants and church leaders in particular moved substantive changes in the character and structure of to a nearby elevation or even a more distant fortified cities. Having said that, I attempted to also consider the location. Bearing in mind the occasional y arbitrary existence and transformation of the road network, as decision, the cities that lost their inhabitants to more well as the appearance of burials intra muros. remote locations are treated in this chapter as abandoned The problem that is better perceptible in most cities (e.g. Aguntum, Demetrias), while their new centres are of the first two groups than in the countryside concerns discussed separately within the group of newly-founded the continuity of habitation in a city versus its mere reoc- cities (e.g. Lavant). The cities that only witnessed shifts cupation. The spaces completely built-up today in many towards peripheral areas are treated as cities with conti- cases prevent us from reliably identifying the character, nuity (e.g. Tridentum, Teurnia, Municipium Riditarum). extent and chronology of the Late Antique cities and blur the distinction between complete abandonment and limited habitation that is combined with a simultaneous 2.3.1 WESTERN PART use of adjacent refuges. For such cities, this problem is particularly discussed. It is also a problem that is more ITALY likely to be researched for the group of newly-formed cities, which were mostly located atop well-defended The province Venetia et Histria, in the northeast elevations later not resettled. In contrast, the newly- corner of Italy, was the best protected part of the area un- founded cities share a difficulty regarding the identifi- der discussion and, consequently, witnessed the decline cation of their urban character as they witnessed fewer of some of its cities last. But even here, several examples investigations and are also less frequently mentioned in shared the same fate (cf. Brogiolo 2010, 76). The study of ancient texts; this issue is highlighted both for the group these Late Antique remains is hindered by the later oc- of newly-formed cities and lower down for the fortified cupation and their fate can often only be traced through settlements that include many potential urban centres. literary sources, very rarely archaeological observations. The latter are slightly more substantial for Altinum and The cities discussed in this book are divided into Concordia (Cantino Wataghin 1996, 254–255; Brogiolo three groups ( Figs. 2.1, 2.2): 2011a; Possenti 2008, 418). - Cities abandoned before the end of Late Antiquity; Concordia, a post with important arms produc- - Cities with continuity; tion ( fabrica sagittaria) even in the 4th and the early - Newly-formed cities. 5th century, does not show significant habitation traces from the later 5th and the 6th century with the exception Most cities can be ascribed fairly reliably to one of the basilica that persisted, was renovated on several of the three groups, others can only be classified with occasions and later became cemeterial in nature (Vil a reservations, which are noted for each case. The first two 2002, 365–427; Croce Da Vil a 2003; Marano 2011, 182). groups comprise ‘classic’ Roman cities, of which those Located at the edge of Italy, at the foot of the Alps, of the first group were largely abandoned in Antiquity Iulium Carnicum (Zuglio) ( Fig. 2.3) underwent an and only exhibit minor changes in the urban fabric in interesting transformation. The city survived to the their final phase, whereas those of the second group mid-5th, in a reduced extent even persisted into the witnessed a radical transformation that brings into ques- 6th century (Corazza et al. 2001, 245–247; Vil a 2002, tion their urban character. The third, most diverse group 342–365). In use at the same time was the hil top site comprises newly-formed centres that in many cases take on San Pietro di Carnia, an elevation high above the over some of the functions of the earlier cities, but have city that revealed an inscription mentioning Bishop a markedly different appearance; they survived in their Ienuarius – this find indicates that the episcopal see may full extent to the end of the 6th and in places persisted have moved to this protected location (Mirabel a Roberti into the first half of the 7th century. 1976, 95; Buora 1992, 107; Ciglenečki 2003a, 265). The 24 2. CITIES Fig. 2.3: Iulium Carnicum. Basilica in the foreground, San Pietro di Carnia behind it (1978). unexplored terraces to the south below the summit allow coins) as archaeological evidence of urban continuity the possibility that a large settlement unit existed there. (cf. Kolšek 1984, 344; Šašel Kos 1994, 102). In the provinces Noricum Ripense and Noricum Mediterraneum, in modern-day Austria, cities did not AUSTRIA survive the final period of Antiquity with the exception of Teurnia. The information that Eugippius provides on Much more is known on the abandonment of the the life of Saint Severinus, from the second half of the 5th Roman cities in present-day Austria, Hungary and Slove- century, is therefore all the more valuable and indicates nia, to the north and east of the defensive belt marked in the existence of cities along the limes in a more modest the literary sources as Tractus Italiae circa Alpes and par- form all to the organised departure of the Romans to ticularly of the Claustra Alpium Iuliarum barrier wal s Italy in 488 (Ubl 1982; Bratož 1982; Ladstätter 2002, (for an overview of the interpretation of both terms, see 340–342; Ubl 2011, 442–448). The longer existence Ciglenečki 2016). Most cities here shared a fate already of cities in the exposed Danubian province Noricum outlined above, with the last peak towards the end of the Ripense may be the consequence of the diplomatic ef- 4th century followed by a rapid decline. Only some show forts of the charismatic Severinus, who succeeded in faint signs of life in the first half of the 5th century, with creating a tolerable cohabitation with the peoples in the some cores of habitation. After the mid-5th century, these immediate vicinity (Wolff 1991, 313). We should note traces become an exception rather than the rule. There that the archaeological investigations do not entirely are modest dwellings in the ruins of some cities, but it is confirm the literary evidence on the late existence of cit- no longer possible to talk of urban continuity complete ies; the final part of Late Antiquity is barely identifiable with an administrative apparatus and episcopal see. The in their archaeological record (overview in Ubl 1982, highly reduced number of inhabitants could no longer 74–85; id. 2011; Christie 2000, 282–284; Ladstätter 2002, defend the extensive city wal s wherever these existed 313–317; Stuppner 2011, 130–141). (overview in Ciglenečki 1987b, 267–271; Poulter 1992; In Iuvavum (Salzburg), most buildings were Ladstätter 2002, 312–320). The recent archaeological abandoned by the third quarter of the 4th century at the investigations are thus fundamentally altering our latest (Kovacsovics 2002); only in the centre were there knowledge of the urban continuity in the 6th century presumably two buildings constructed towards the end that was previously primarily based on literary evidence of this century. The remains of a small fort from the late (mentions of bishops attending synods) and disregarded 4th century were found on Hohensalzburg, a hill above the archaeological evidence (cf. Tóth 1974; Wolff 1991, the Roman city, that presumably formed part of a large 287–311; Liebeschuetz 2001, 76–77); it should be noted fortified civilian settlement. The church that Eugippius that earlier authors often used stray finds (of costume, mentions has not been recovered as yet, finds from 25 Slavko CIGLENEČKI Fig. 2.4: Favianis. Well-preserved Late Antique eastern city wal s (2018). the 5th century are absent. The archaeological data are that already hosted German foederati. There are some rather scarce, mostly due to the later occupation of the archaeological finds that postdate the year 488, but these fortified area. What is certain is that the old city below can no longer be seen as signs of an urban settlement the hill was abandoned (Ubl 1982, 84; Ladstätter 2002, (Ladstätter 2002, 314–316; Ubl 1982, 78–79). 313–314; Winckler 2012, 248). The forts marked as civitates in Eugippius are a The colony Ovilavis (Wels) became the capital of borderline category of cities that can also be found Noricum Ripense in the 4th century (Ladstätter 2002, in other late literary sources (Bratož 1982, 115–118; 314; Scherrer 2011, 112). In the late 2nd or at the begin- Christie 2000, 283; Ubl 2011, 443–444). Of these ning of the 3rd century, it was enclosed with 1.4 m thick civitates, the best known is Favianis (Mautern) ( Figs. city wal s additional y fortified with protruding towers 2.4, 2.5), which is central in the activities of Severinus. and as many as four ditches. A decrease in the archaeo- Eugippius mentions lively trading markets in the city, logical finds has already been observed in the 4th century, the existence of an upper class of inhabitants (widow which suggests a significant reduction in the population Procula, daughter of noble descent who hid supplies number. The city soon lost its prominent place and is no of grain in times of famine), granary and monastery longer mentioned by Eugippius. There are only scarce with a church (Ubl 1982, 77; Bratož 1982, 119). Ar- Early Christian finds, but no church. Habitation traces chaeological remains are well researched and also from the 6th century are absent. comprehensively published (see Groh, Sedlmayer 2002; The administrative and military centre of Lauria- Scherrer 2011, 109; Stuppner 2011, 131–133). The Late cum (Lorch) witnessed great changes after the reforms Antique fort was established with different modifica- of Diocletian (Ubl 2002; Igl 2011). The legion previously tions in the place of an earlier fort in Period 5 (roughly stationed in the camp was divided into smaller units last third of the 3rd century). The small finds reflect a and only a small garrison remained in the city. People civilian component in addition to the military one. gradual y left the civilian settlement, which witnessed In Period 6 (370/380–450), there were considerable a devastating fire in the 3rd century, and moved to a changes in the fort interior. The earlier wooden archi- well-defended legionary fortress (Scherrer 2011, 108). tecture was replaced almost entirely by buildings of a A church was constructed in the abandoned valetudi- different concept, made of unfired brick. The fort was narium. Modest civilian buildings only show partial also extended towards the Danube and the extension habitation. Two wooden edifices are mentioned (one used as a refuge and place for livestock or additional with heating canals) that were built into the former troops. The small finds exhibit Roman, but also Ger- army barracks. Eugippius mentions Lauriacum as op- manic elements and hence a mixing of the population, pidum, urbs or civitas, showing that we may see it as a as well as a strong civilian presence. Period 7, which modest urban (?) settlement at least until 488 and one is the time of Severinus’ activities in the city, shows a 26 2. CITIES . Fig. 2.5: Favianis. Well-preserved tower on the eastern side (2018). reduced number of inhabitants and later abandonment, The Roman municipium Aelium Cetium (St. which corresponds with Severinus’ account. Pölten) witnessed abandonment of some of its parts Vindobona (Vienna), a legionary fortress in the in the 3rd century and renewed prosperity with the Austrian part of the province Pannonia Prima, changed construction of new houses after Constantine, when it its character in the second half of the 4th century to became the supply base in the hinterland of the Danube that more like a city (Stuppner 2011, 137–139; Mosser limes (Ladstätter 2002, 316; Scherrer 2011, 111). It was 2011, 484–498). Instead of army barracks, the heavily abandoned by the end of the 4th century, which reflects fortified interior held densely spaced buildings of dried the end of an organised defence of the limes. In the clay, which were both living quarters and workshops. first half of the 5th century, there are only modest signs The small finds reveal soldiers and civilian population of Roman presence mixed with Germanic elements. It including women and children, as well as an increasing was final y abandoned soon after the mid-5th century. presence of foreign ethnic groups. Final renovations in The city of Aguntum (Lienz) ( Figs. 2.6, 2.7) has a the interior date towards the end of the 4th century and longer existence (Walde 2002, 152–158; Sossau 2018). A the small finds shows that groups of inhabitants lived long wall was built in the 3rd century that protected the here at least to the first third of the 5th century. An in- settlement from the attacks coming from the east. After direct confirmation of this comes from the cemeteries, fires, the city was renovated on numerous occasions that which were created in the area of the former cannabae can reliably be dated to the second half of the 4th and first and span from the second half of the 3rd to the early half of the 5th century. At this time, all public spaces were 5th century. turned into modest dwellings or workshops, similarly as Also in the Austrian part of Pannonia Prima is in many cities across the Eastern provinces. A prolonged the civilian city of Carnuntum (Ladstätter 2002, 317; existence of the city is also visible in the large Early Chris- Kandler 2004; Gugl 2011; Stuppner 2011, 139–141). In tian church constructed on the eastern side of the city the first half of the 4th century, it still functioned as a wal s. There are modest habitation traces recorded even city and received defensive wal s presumably at this very after the fire in the early 5th century. Stray finds from the time. It suffered an earthquake in the mid-4th century 6th century reveal that smal groups of people were present and Ammianus Marcellinus described it as abandoned in this area, whereas the church authorities and part of already in 374. The inhabitants moved within the wal s the population moved to the other side of the River Drau/ of the adjacent legionary fortress, where minor build- Drava, to the natural y excellently protected Kirchbichl ings have been recorded that date around 400. Modest hill (Lavant). The churches excavated there, which include habitation traces have also been dated to the first few a presumed episcopal complex and other buildings, show decades of the 5th century. that the hill now served as the centre of settlement (Lad- stätter 2002, 319; Winckler 2012, 244–245). Evidence of 27 Slavko CIGLENEČKI Fig. 2.6: Aguntum. City plan (Tschurtschenthaler 2018, Fig. 1). Fig. 2.7: Aguntum. Central part of the city and its wal s, view from the south (1975). this is also to be found in the writings of Venantius For- the 4th century (Ladstätter 2002, 317–318; Dolenz et al. tunatus, who mentions in 565 that Aguntum is located 2016). The first reduction of the settlement area followed on a high hill (Egger 1916, 60–61). the incursions in the third quarter of the 3rd century Virunum (Zollfeld) ( Figs. 2.8, 2.246), formerly the and the city witnessed its last peak in the second half capital of the province Noricum, became the capital of of the 4th and first third of the 5th century. At that time, the newly-formed province Noricum Mediterraneum in Insula I was renovated and transformed into a palatial 28 2. CITIES Fig. 2.8: Virunum. Plan of insulae VIII, IX, XI with suburbium (episcopal complex?) (from Dolenz et. al. 2016, Fig. 14). building that presumably housed the seat of the provin- the Pannonian Plain that suffered from waves of hostile cial administration. Also constructed were large church incursions. Partial continuity can be discerned on the buildings (episcopal complex?) at the edge of the city. nearby Frauenberg hil , where the recent discovery of The city was abandoned soon afterwards, which may be an Early Christian church indicates the possibility of the consequence of its exposed location in the lowland, Christian population living there deep into the 5th cen- along major roads, and of the absence of defensive wal s. tury (Ladstätter 2002, 318; Schrettle 2014). The extensive Flavia Solva (Leibnitz) ( Fig. 2.9) witnessed a pe- cemetery unearthed on the slopes of the hill points to riod of great prosperity in the late 3rd and early 4th cen- a predominantly Roman population and presence of turies, which is reflected in lavish residential buildings foederati (Steinklauber 2002). (Hudeczek 2002, 209–210; Groh, H. 2021). Very scarce are the archaeological traces postdating 400, which show barely any human presence. Similarly as Virunum, Flavia HUNGARY Solva had no city wal s and was located at the fringes of The flatland in the hinterland of the exposed part of the limes was under relentless pressure from foreign peo- ples, which in itself makes the late existence of Pannon- ian cities questionable (cf. e.g. Tóth 1987; Poulter 1992, 105–113; Müller 2000; Vida 2007, 322; Christie 2007, 558; Reka 2015, 672). Some do show the odd primitive building of a late date, but these do not make up urban centres of their former importance. In the western part of the Pannonian Plain, in present-day Hungary, there are two cities of significance for our understanding of urban (dis)continuity, namely Scarbantia and Savaria. Scarbantia (Sopron) ( Fig. 2.10), in the province Pannonia Prima, is most frequently mentioned as a city with the continuity of Roman habitation (Tóth 1974; Fig. 2.9: Flavia Solva. Remains of Roman buildings (2014). Müller 2000, 244, 246; Gömöri 2003, 86–87; Vida 2007, 29 Slavko CIGLENEČKI Fig. 2.10: Scarbantia. Plan of the city (from Gömöri 2001, Fig. 1a). 322, 325–326). Péter Tomka (2015) wrote the last exten- sive overview of excavations and recovered small finds in this city that enables a dating and an insight into the ethnic structure of its inhabitants. He particularly ob- served the forum being filled up with deposits early on, Fig. 2.11: Savaria. Plan of the palace complex (Isztin et. al. as well as an early abandonment of public and residential 2014, Fig. 1). buildings. Excavations revealed fairly simple dwellings in the forum and partly along the city wal s that show three phases of existence from the late 4th to the 5th/6th centu- the Roman tradition, but also widely used among the ries. Also mentioned is an entirely investigated wooden Germanic foederati. The fact that there are no reliable building and other buildings, only partly known, that contexts from the 6th century, the existence of an episco- have masonry foundations and a wooden superstruc- pal see in the 6th century and with it the continuity of the ture. One of these was fitted with a Late Roman system Roman population is questionable. The most plausible of heating canals. Tomka observed that the immediate conclusion is that the available evidence does not speak surroundings of the city underwent rapid barbarisation of a continuity of habitation, but merely of a reuse of the and that settlements were abandoned towards the end monumental Roman city wal s and thus the continuity of the 4th or beginning of the 5th century. The original of ruins (Gömöri 2001, 230). inhabitants presumably left the city (dilapidated public The Roman city of Savaria (Szombathely) ( Figs. buildings and houses) and some people previously living 2.11, 2.12) lay on the Amber Route and was an impor- in the vicinity now inhabited the city with strong wal s. tant post of the Italo-Pannonian trade (Scherrer 2003, They are believed to have built simple huts on top of the 54–55). It became the capital of the province Pannonia ruins, with at least some of the huts still exhibiting the Prima in the 4th century. The city wal s are poorly known, Roman tradition. Tomka thus allows for the possibil- only identified in a few places and predominantly dated ity of the habitation traces dating throughout the 5th as early as the 2nd century. An inscription relates the century, but does not see – as opposed to some earlier construction of a horreum on imperial order and in authors – reliable contexts from the 6th century. Endre honour of Constans II. The remains of a Late Antique Tóth posited that we may tie episcopus Scaravacensis, palace have been found at the western city wal s, which bishop known from the records of the Synod of Grado, was most likely built in the time of Constantine the held in 579, with this city (Tóth 1974). The city thus far Great and extended across as many as six insulae, public yielded no church remains, but there are elements that baths and a prestigious house. The centre of the palace point to a Christianised Roman population (Heinrich- was a large hall terminating in an apse in the west. Its Tamáska 2012, 225). In connection with this is the mass wal s were richly painted and the floors adorned with a of burnished ware believed to have been produced in high-quality Late Roman mosaic (Tóth 2011). This out- 30 2. CITIES Fig. 2.12: Savaria. Remains of the Late Antique palace (1982). standing building was surrounded by numerous smaller Fig. 2.13: Aquincum. Plan of the Late Antique fort (Zsidi rooms and an octagon with lavish baths constructed in 2004, Fig. 8). the former peristyle. The city, which was the birthplace of St Martin, yielded several Early Christian finds and epitaphs, while the Passion of Quirinus even mentions a ‘ basilica ad Scarabetensem portam’ near the north gates. A critical assessment of the Early Christian remains, particularly the sacral buildings, has shown that we can as yet not confirm the existence of a church in Savaria (Scherer 2003, 55; Heinrich-Tamáska 2012, 224). The transfer of the reliquiae of martyr Quirinus to Italy in the early 5th century proves the col apse of the ecclesiastical organi- sation in the city (Bratož 2011b, 600; Vida 2011, 635). Following the earthquake of 456, Savaria is believed to have been completely abandoned (Scherrer 2003, 55). The proximity of the Huns makes urban conti- nuity particularly disputable in northern and eastern Pannonia. Inhabitants of Aquincum ( Fig. 2.13) were abandoning the civil settlement during the 4th century and moved to the adjacent fort (Mócsy 1974, 310; Mül- ler 2000, 244; Zsidi 2011). The latter was – presumably Fig. 2.14: Sopianae. Remains of the Early Christian ‘Mauso- under Constantine I – extended towards the river. In leum’ (1998). its older, still well-fortified part, military buildings were renovated and given a primarily administrative function. numerous Early Christian tombs (30 tombs, mausolea, The large baths were extensively renovated and turned to hypogea, cel ae memoriae) beautifully painted with a palace, presumably to house the military commander biblical motifs and Christian symbols. No church of the province Valeria. A religious function is presumed remains have thus far been found in the city. The last in two buildings, but could thus far not be reliable proven excavations show that a new basilica urbana, extending (Heinrich-Tamáska 2016, 126). The last diagnostic finds into a horreum, was constructed in the forum in the from the palace date to the late 4th century, while traces 4th century. Wooden constructions already predomi- from the 5th century are mentioned in connection with nate the city in the first third of the 5th century; these presumed churches. Indirect evidence of habitation in were torn down and destroyed in fire, which scholars this part comes from the necropoleis, in use to the late linked to the arrival of the foederati or even Huns and 4th or early 5th century. Goths. After the arrival of the Huns in 433, the urban Sopianae (Pécs) ( Fig. 2.14), a city in the south of existence of the city was no longer possible without the Valeria, reached its apogee in the 4th century (Fülep existence of the province and the cemeteries also no 1984; Sándor 2001; Gábor et al. 2004, 287). This is longer show continuity; only a few barbarian burials less clear in the city itself, but indirectly proven by the have been recorded (Gábor 2008). 31 Slavko CIGLENEČKI Fig. 2.15: Emona. Plan of the Early Christian centre (fourth construction phase) (Plesničar Gec 1983, Fig. 11). SLOVENIA destroyed, most likely during the Hun incursion in 452 (Plesničar Gec 1983, 30–31; Kos 1983, 102–103). The city The transformation and abandonment of Roman yielded no architecture that could reliably be set to the cities in Slovenia is most clearly il ustrated on the example second half of the 5th century. The literature does mention of Emona (Ljubljana) ( Figs. 2.15–2.17, 2.248, 2.257). It poorly documented remains of modest dwellings and a formed part of the province Venetia et Histria, but its rotunda, though the associated small finds published thus location east of the Alpine barrier wal s resulted in a far do not allow a reliable dating of the building to a time more rapid decline of the urban fabric. The extensive after the mid-5th century (Plesničar Gec 1997, 364–368). excavations conducted in recent decades have provided Very rare objects from the second half of the 5th and the us with reliable evidence on the discontinuity of the city 6th century indicate a brief presence of individuals or (Slabe 1975, 84–86; Plesničar Gec 1983, 29–32; ead. 1997, small groups of people rather than continuous habita- 366–368; ead. 2005). Numerous small finds including tion to the late 6th century (Plesničar Gec, Sivec 1978, 61; coins show that Emona had its last peak in the 4th and Ciglenečki 1997, 192). The most convincing evidence of early 5th centuries. It developed within the perimeter of discontinuity is the destruction and abandonment of the the Augustan city wal s, which were reinforced in Late Early Christian centre in the first half of the 5th century, Antiquity and side entrances walled. In the second half which clearly shows that the clerics headed by the bishop, of the 4th century, rooms were arranged in Insula 32 of and indirectly also the predominantly Roman population which the layout and mosaic decoration indicate the left the city (Ciglenečki 2012, 464–466). first Christian buildings (Djurić, B. 2005, 674–675). In The cities in the south-eastern part of Noricum the next phase, a baptistery and porticus with richly Mediterraneum, along the main road from Pannonia decorated mosaic floors and many donor’s inscriptions to Italy, shared a similar fate. were constructed. They are doubtlessly part of an episco- Celeia (Celje) ( Figs. 2.18, 2.19) still shows diverse pal centre that has as yet not been ful y investigated. Its and intensive signs of life in the 4th century. In the middle construction can reliably be dated to the first few decades of this century, mighty city wal s were constructed with of the 5th century, though it was very soon abandoned or numerous towers (Lazar 2002, 91–92, 96–97; Novšak 32 2. CITIES Fig. 2.16: Emona. City wal s, reinforced in Late Antiquity (2017). Fig. 2.17: Emona. Excavation of Late Antique buildings in the south-western part of the city (1998). 2007, 30; Bausovac, Krempuš 2009, 44, 45; Bausovac, destruction and abandonment of its sacral buildings in Krajšek 2020, 51–56). The significance of the city in Late the first half of the 5th century. Burials from the second Antiquity is reflected in a large Early Christian basilica half of the 4th and first half of the 5th century came to boasting mosaic floors with many donor’s inscriptions light in the immediate vicinity of the city centre. There (Kolšek 1984, 342; Glaser 1997, 67). An octagonal are no known sealed contexts from later times that would baptistery was also unearthed, but at a considerable suggest urban continuity. The only chronological y di- distance from the basilica, which allows for a possible agnostic items were found either in secondary position second basilica located in between (Vogrin 1991). Both (Early Christian epitaph) or are without reliable findspot the baptistery and the basilica date to the late 4th or early (several late coins and other small finds). The numerous 5th century. Similarly as in Emona, Celeia witnessed a bronze chandeliers in the shape of Christograms are 33 Slavko CIGLENEČKI therefore all the more important finds, recovered on the nearby hil top settlement on Vipota and brought there from the destroyed churches – they prove that the remaining inhabitants of the city in the second half of the 5th and the 6th centuries took refuge in remote and natural y well-protected settlements in the hil y hinter- land of Celeia (Ciglenečki, Modrijan 2020). The fate of the important Roman city of Poetovio (Ptuj) ( Figs. 2.20, 2.21), located at an advantageous river crossing of the Drau/Drava, is well-known. The limited extent of the later habitation remains and the intensive rescue investigations in recent decades have brought a considerable amount of evidence on the city in Late Antiquity (overview in Horvat et al. 2003). It shows a gradual y shrinking urban area with a shift and concen- tration of habitation at the foot and on top of both hil s on the left riverbank (Saria 1939, 148; Mikl Curk 1978, 407; Horvat et al. 2020b, 51). A passage in Ammianus Marcellinus mentions that Constantius was captured in a palace outside the Poetovio city wal s, which suggests the Fig. 2.18: Celeia. Plan of the Roman city (from Lazar 2002, Fig. 27). Fig. 2.19: Celeia. Late Antique wal s with the entry gates (from Krajšek, Žerič 2010). city was walled in Late Antiquity. The numerous small chronological y last group of graves dating to the mid- finds of a late date belong primarily to the second half dle or possibly even second half of the 5th century (Mikl of the 4th and first third of the 5th century (Ciglenečki Curk 2005, 246; Horvat, Dolenc Vičič 2010, 73). More 1993). Several elements indicate the existence of Early is known on a burial likely of a Hun warrior, the only Christian churches, which are not known in detail. Many one of its kind in the area under discussion (Lubšina cemeteries on both riverbanks reveal that parts of the Tušek 2004, 76–78; Knific, Nabergoj 2017, 22–23). We city were already abandoned in the second half of the also know that Priscus mentions Poetovio as the starting 4th century. The investigations conducted in Rabelčja point of the diplomatic mission to the court of Attila in vas, albeit not yet integral y published, revealed the 449 (Šašel Kos 1994). The city was abandoned in the 34 2. CITIES Fig. 2.20: Poetovio. Plan of the wider city area (Horvat et al. 2003, Fig. 4). Fig. 2.21: Poetovio. Central part of Late Antique settlement on Grajski grič and Panorama, from the north-west (2020). second half of the 5th and throughout the 6th century; ing to Italy from the east. In analysing the small finds the scarce small finds (mostly without reliable findspot from the city, Peter Petru also tackled the ratio between data) can certainly not be taken as evidence of the city’s the finds from different periods and established that, of existence and probably represent the traces of individual the 1300 dated Roman items, 300 belonged to the 4th groups that occasional y protected the still important century, which was followed by a marked decrease with river crossing (Knific, Tomanič Jevremov 1996). practical y no finds from the 5th century (Petru, P., Petru, The fate of a lowland city is best il ustrated on the S. 1978). He only identified two or three items from the 5th example of Neviodunum, an unfortified city and harbour century, which recent analyses have shown to be earlier as on the River Sava and on the most important road lead- well (Petru, P., Petru, S. 1978, 362; Knific 1993, 521–542). 35 Slavko CIGLENEČKI Fig. 2.22: Aquae Iasae. Thermal complex with remains of the Roman buildings (2011). Fig. 2.24: Aquae Iasae. Interior of the basilica (2011). period, when the pool was completely renovated and a basilica built (Vikić-Belančić 1978, 590–591; Migotti 1997, 25–26). The remains of wall paintings and modifica- tions of certain rooms (baptistery!) show that the basilica Fig. 2.23: Aquae Iasae. Plan of the basilica in the thermal baths, urbana was turned into a church towards the end of the converted into the Early Christian church in the 4th century 4th century. Both the architecture and the recovered smal (Vikić-Belančić 1978, Fig. 1). finds reliably date the abandonment of the buildings to the first half of the 5th century, which is to be expected given that the settlement on an exposed location was unfortified CROATIA (Migotti 1997, 26; Kušan Špalj 2020). The cities Mursa and Cibalae, both located in the For the continental part of Croatia, the bath centre strategical y important hinterland of the limes, shared a of an urban character called Aquae Iasae (Varaždinske similar fate. The colony Mursa (Osijek) is poorly known Toplice) ( Figs. 2.22–2.24) reveals most on the abandon- (Buzov 2010, 142–143; Marin 2018). It was walled in the ment of cities. The systematic investigation here revealed Roman period, but we lack archaeological evidence of the peak of construction activities in the Constantinian its continuity in the 5th century. 36 2. CITIES the 5th century (Gračanin, Rapan Papeša 2011, 13–16; Rapan Papeša, Roksandić 2016, 151–152). Rescue excavations have also noted the existence of houses reusing the still standing parts of Late Roman buildings and new wal s built of clay-bonded stone. The other, more numerous type of dwellings were sunken houses, usual y in small groups. Many pottery kilns were in use until the 4th century. In the 5th and 6th centuries, vessels with characteristic stamped and burnished decoration joined the previous forms. Anita Rapan Papeša and Danijela Roksandić thus posit a continued presence of a part of the Roman population in the 5th and 6th centuries, joined by immigrants. Small finds indicate that Cibalae were under the Gepid rule in the mid-6th century (Rapan Papeša, Roksandić 2016, 151–159). The necropoleis outside the city were in use to the 5th century, while later burials came to light intra muros either individually or in small groups. The absence of city wal s, public buildings and reliably identified church architecture, but also bishopric after the middle Fig. 2.25: Cibalae. Plan of the Late Antique city and necropoleis of the 5th century signify a degradation and downfall of (from Rapan Papeša, Roksandić 2016, Fig. 1). the Roman city. On the other hand, the simple build- ings and sunken houses point to a modest presence of the Roman population and more substantial numbers of immigrants. Together with the late burials intra Much more is known on the colony Cibalae muros, they show partial continuation of life, but do (Vinkovci) ( Fig. 2.25). Similarly as most other cities, not support the claim that Cibalae still functioned as it reached its peak in the 4th century, when many new a city after the mid-5th century. buildings were constructed (overview in Iskra-Janošić At Kamenica, a site outside the city of Cibalae and 2004; Gračanin, Rapan Papeša 2011). The city was of a along the road to Sirmium, a walled Early Christian great strategic significance as it stood on an important complex was found that is believed to have been built crossroads in the immediate vicinity of the limes. It was on the spot where Pollio, primicerius lectorum, was here that Constantius II set up camp in 351, during his martyred. The results of the preliminary, mainly non- conflicts with Magnentius. Cibalae was also the birth- destructive surface surveys have suggested that the place of the Emperors Valentinian and Valens and the complex with a church, baptistery and tombs was cre- city’s prosperity in the second half of the 4th century can ated in the 4th century. Its duration has as yet not been therefore also be due to imperial munificence. What established (Vulić 2016). remains open is the question of the city’s defensive On the east coast of the Adriatic where most cities installations. The existence of earthwork ramparts, show a clear continuity, Argyruntum (Starigrad na which earlier researchers posited, seems questionable moru) met with a different fate. It was an emporium and city wal s have not been detected, which led some built on a small peninsula and enclosed with a defensive to suggest the surrounding marshland offered suf- wall (Abramić, Colnago 1909). Its interior is poorly ficient protection (cf. Buzov 2010, 141; Rapan Papeša, investigated and much more is known of the associated Roksandić 2016, 149). A horreum was built in the 4th cemetery. The latter has revealed that the city reached century and spolia from the capitolium were used in its zenith in the 2nd century, began declining in the 3rd its construction. A basilica was presumably built in and was abandoned in the 4th century (Fadić 1995). the second half of the 4th century, which shows no The reasons for such an early abandonment are unclear, elements of an Early Christian church. At that time, neither do we know where its inhabitants retreated to. the large baths still functioned in the city. The recently Some have suggested they moved to the fortification discovered rich hoard of silverware from the 4th cen- on Sv. Trojica near Tribanj, Šibuljina, some 6 km away tury points to a high standard of the inhabitants (Vulić (see Chapter 3.3). et al. 2017). Earlier researchers presumed that Cibalae was destroyed towards the end of the 4th century and the ruins later inhabited by the Gepidae (Iskra-Janošić 2004, 185). For some buildings, new investigations revealed the possibility of use at least in the first half of 37 Slavko CIGLENEČKI Fig. 2.27: Diocletianopolis (Argos Orestikon). City plan (Rizos 2017b, Fig. 11). century is very much open; it is treated in the group of Fig. 2.26: Bassianae. City plan (Milin 2004, Fig. 4a). cities that declined before the end of Late Antiquity due to the lack of reliable archaeological evidence. GREECE 2.3.2 EASTERN PART The writings of Hierocles reveal that some cities in In the eastern part of the Empire, much fewer cities the southern Balkans still marked as poleis in the mid-3rd were abandoned in the course of Late Antiquity. century lost their urban status in the early 6th century (Dunn 2004, 569). The recently identified Diocletianopolis (Argos SERBIA Orestiko) ( Fig. 2.27) in Thessaly revealed the remains of several buildings and city wal s enclosing a 45 ha large area In Serbia, the attribution of Bassianae (Sremski (Karagiorgou 2001b, 150–156). The partly investigated Petrovci) ( Fig. 2.26) into this group of cities is thus uncity wal s, measuring 2.4 m in thickness, incorporate certain as the earlier and poorly published excavations rectangular towers combined with pentagonal ones in the do not allow for more definite conclusion (overview in corners and semicircular ones flanking the gates. A partly Milin 2004). The city was established on natural y raised known three-aisled basilica is known in the city interior, ground and enclosed with mighty wal s with towers. two were found extra muros. Of the two investigated The urban layout shows a number of public buildings, residential buildings, House B from the first half of the as wel as an Early Christian basilica (possibly even 4th century revealed that many small rooms were built two?) roughly dated to the 4th to 6th centuries (Grbić into the ruins of the original building; the destruction of 1936, 30–32; id. 1937, 3). With the exception of a pos- the latter cannot be precisely dated, though small finds sible church from the 5th and 6th centuries, the available indicate it occurred during the Ostrogothic incursion archaeological evidence is insufficient to posit a late ex- towards the end of the 5th century, possibly even earlier istence of the city, though some literary sources do speak (Karagiorgou 2001b, 154, 156). Procopius informs us that, in favour of continuity. Jordanes is one such source, after the abandonment in 473–483, Justinian moved Dio- mentioning that the city was besieged in the mid-460s, cletianopolis to the island on Lake Castoria (Karagiorgou while Justinian’s Novella XI from 535 states the city was 2001b, 150–156; Saradi 2006, 468). under the jurisdiction of the archbishop of Justiniana The old and vast Hellenistic port of Demetrias ( Figs. Prima. It is also mentioned in Hierocles’ Synekdemos and 2.28, 2.29), in the Pagasetic Gulf in the Aegean, witnessed the Anonymous Geographer of Ravenna (Milin 2004, extensive changes in Late Antiquity (Karagiorgou 2001a, 257). With the current state of research, the question of 197–211). With its centre of gravity then at the north the city’s continuity in the 5th and even more so the 6th harbour, it was walled in the 4th century to enclose only a 38 2. CITIES silica was built in the late 4th century. The abandonment of Basilica of Damokratia in the second quarter of the 6th century proves that the northern city area was abandoned at this time. The city wal s were no longer renovated in the 6th century and the remaining population moved to Iolkos, a hill some two kilometres away. The fort on the hil extended across 6 hectares and revealed a defensive wal and different buildings, including a basilica. It is believed to have been established soon after 482, which has been linked with the Ostrogothic incursion, while Justinian presumably refortified the hill and the new fort became an administrative and Christian centre (Karagiorgou 2001a, 203–210). In a smaller measure, life also continued in the settlement at the foot of the hil . Parts of the population are believed to have inhabited the lowland parts of the city and only sought refuge behind the wal s of the fort in times of danger. A similar situation has been observed in the case of Pydna (Poulter 1998; Saradi 2006, 388; Sodini 2007, 317), a city that the Goths seized in 479. This led to the transfer of the bishopric to Louloudies, a site discussed below as part of the group of newly-founded cities (see Chapter 2.5.2). 2.4 CITIES WITH CONTINUITY Roman cities that persisted in variously altered forms to the end of the 6th, in places into the early 7th century Fig. 2.28: Demetrias. Roman city and the fort of Iolkos (from and some even lasting to the present day, can be found Karagiorgou 2001a, Fig. 10a). across a large part of the area under discussion ( Figs. 2.1, 2.1). Also included in this group are the cities abandoned for a lengthy period after the mid-5th century and later re-established in a more limited extent under Byzantine authorities. The lengthy interruption of their existence does bring them close to the cities abandoned in An- tiquity, but the resettlement and partial reuse of the old architecture on the same spot in the 6th century, explicit mentions in literary sources and scarce archaeological remains nevertheless identify them as cities with conti- nuity. We should bear in mind that they suffered a major break brought on by the Hun incursions in the 5th century, when cities such as Sirmium, Singidunum, Naissus and Serdica were heavily hit; this is particularly apparent in the end of the bishoprics in these cities. Bishops are again Fig. 2.29: Demetrias. In the foreground are the remains of mentioned there later: in Serdica already in 458, Naissus Basilica of Damokratia (2004). in 516, Sirmium in 567/568 and Singidunum in 579. Only the last reports of bishops in the years between 580 and 615 indicate the end of the ecclesiastical organisation and small portion of the former city. The walled part included abandonment of cities (Bratož 2011a, 228–229). a large Early Christian basilica (Basilica of Damokratia), For these cities with a long architectural tradition, built in the early 5th century and later renovated several the presentation below will only briefly mention the times, as well as the luxury ‘House of Damokratia’ not important constructions and changes in the 4th and first far from it. A small residential area extended outside the half of the 5th century and rather focus on the period walled city, at the southern harbour where a smaller ba- after the mid-5th century and the altered structures of 39 Slavko CIGLENEČKI Fig. 2.30: Ravenna. Front of the Late Antique palace (2009). that time. The emphasis is on the characteristics that These changes occurred slightly later at the heart came to light in the last decades during the systematic, as of the empire, which is understandable given the pro- well as the more numerous rescue investigations. These tection of Italy’s north-eastern flank with barrier wal s afford an increasingly realistic insight into the defensive, ( Claustra) in the 4th century; even this protection proved sacral and residential architecture, into the chronology insufficient towards the end of the century. Alongside and the altered economic and social circumstances. other defensive measures (primarily a variety of newly- formed forts and either construction or renovation of city wal s) implemented across the whole of northern 2.4.1 WESTERN PART Italy and eastern Alps, the Romans were forced to find a new and natural y well-protected location in the marshes ITALY of the Po Delta. Via canals, the site was well-connected with the Adriatic Sea and the imperial fleet based there. The two capitals of Late Antiquity need no special The next major shift that mainly affected the west half attention, though we should begin this discussion by of the Empire was the Ostrogothic state, with Gothic mentioning Ravenna ( Fig. 2.30), a capital and its archi-troops settling the well-defended earlier forts and cit- tecture that are symptomatic of the general shifts in set- ies. Their presence in Ravenna can be observed in the tlement (overview in Cirelli 2008). With the exceptional magnificent new constructions (Sodini 1995, 165–166; achievements of Late Antique architecture, it indicates Baldini Lippolis 2001, 251–253). This substantial trans- all the changes observable in more modest settings and formation of the city is rounded off by the renovation on a smaller scale in other cities and settlements. The and embellishment under Justinian, which is particularly clearly identifiable construction and renovation phases marked by the construction and adaptation of churches. corroborated in the literary sources closely mirror the Recent intensive investigations in Ravenna also explore changes in the settlement pattern across the area under the different residential and economic parts of the city discussion in the 5th and 6th centuries. The character and that gradual y shed light on these aspects of the city and appearance of Ravenna changed fundamental y in the its harbour at Classe (Cirelli 2010, 244–253). beginning of the 5th century (402), when the previously In another north Italian city, in Brixia (Brescia) small Roman town became the glittering capital of the ( Fig. 2.31) modern excavations under the leadership of western world (Brogiolo 2011a, 151). The transfer of the Gian Pietro Brogiolo revealed the characteristic signs Western Roman court to Ravenna effectively il ustrates of transformation in a poorly known segment of late the deteriorating conditions of the day, as has already urbanism and housing (Brogiolo 1988; id. 1999). In been discussed in connection with the cities of the first Late Antiquity, the city continued to develop within the group, but is even more evident in the fortifications in confines of the Augustan city wal s, though with the ad- the countryside (see Chapter 3.3.2). dition of another, previously undefended part that was 40 2. CITIES Fig. 2.32: Verona. City plan (Brogiolo 2011a, Fig. 53). Fig. 2.31: Brixia. City plan (Brogiolo 2011a, Fig. 16). now enclosed with double wal s; Brogiolo posits this was a proteichisma (Brogiolo 2011a, 96–97). A cathedral and ducal complex presumably stood in the south-western part of the city, but not much is known of them. The public buildings appear to have been unaltered at least until the late 4th century, followed by destruc- tion, spoliation and construction of modest dwellings. The capitolium thus seems to have been abandoned by the end of the 4th century and remained abandoned for a century to come, after which there appeared the first Fig. 2.33: Verona. Hil San Pietro above the River Adige (2016). traces of modest dwellings. The arrival of the Langobards brings a large-scale destruction of buildings in this area, as well as the instal ation of two pottery kilns and a house (Brogiolo 2011a, 58, Fig. 20). the houses served as places to deposit refuse or bury The degradation of the city is primarily reflected the dead. Surprisingly, these dwellings revealed a fair in the housing, with subdivisioning of the abandoned amount of imported pottery. domus and a considerably lower living standards. Near At Piazza Duomo, a domus was abandoned in the 5th the monastery of S. Giulia, a series of dwel ings from the century and wooden huts built into it, in the Lango bard second half of the 5th to the 7th century came to light era it was used as burial grounds (Rossi 2003, 33–34). in the area of five domus. The primitive buildings were The wal s of domus at Via Alberto Mario were razed partly built into the earlier wal s, partly sunken into to the ground and simple residential buildings of wood the ground or built of wood on drystone foundations were constructed at the same place. In the Ostrogothic (Brogiolo 1999; id. 2011a, 165). An interesting example period, a large two-room building was erected with is that of a large and lavish Augustan domus that for the poorly constructed wal s and a wooden superstructure, most part continued unaltered to the first half of the 5th which suffered a devastating fire in the 6th century that century, after which it witnessed a radical change of its led to the area being abandoned (Brogiolo 1988). interior with postholes dug through the mosaic floors, Research at several sites has revealed habitation hearths arranged on top of the mosaics, the rooms sub- continuity into the following centuries. divided and so forth. The smaller rooms may be seen Verona ( Figs. 2.32, 2.33), the capital of the Ostro- as small family dwellings. The arrival of the Langobards gothic King Theoderic and the Langobard King Alboin, is brought changes such as the ruins of the old wal s reused well-known through different literary sources, medi eval to make dwellings with the addition of wal s made of depictions ( Iconographia Rateriana) and archaeological wood or clay-bonded stone. The empty spaces between evidence, and shows several characteristic signs of Late 41 Slavko CIGLENEČKI Antique transformation (Cavalieri Manasse, Bruno 2003; Brogiolo 2011b, 136−139). The city wal s, dated with the help of an inscription to the Gallienic times, were rein- forced under Theoderic with a new outer wall of reused material and included an amphitheatre in the south, while on the north side an imperial palace was built on the hill San Pietro, on the opposite bank of the River Adige (Brogiolo 2011a, 96). The forum was already abandoned towards the end of the 4th or beginning of the 5th century and a wooden building erected on it in the 6th century that burnt down in the fire of 589/590, as Paul the Dea- con relates. On that occasion, the forum was reportedly covered by an up to 2-m-thick layer of dark earth used at least in part for gardens. The city also revealed houses only built under the Ostrogoths and in use to the 7th century, as well as a domus renovated in the early 5th century and then gradual y abandoned. The excavations at Via Dante unearthed a surprising situation, with partly maintained street-fronts, behind which were humble dwellings and gardens (Wickham 2005, 648). Of church architecture, archaeological investigations revealed an episcopal com- Fig. 2.34: Tridentum. City plan (Brogiolo 2011a, Fig. 32). plex, as well as martyrial and cemeterial churches extra muros. Single or small groups of burials have been found within the city, existing alongside contemporary cemeter- ies outside the city wal s. Evidence thus shows the city at the end of Antiquity boasted mighty city wal s, several imposing church buildings, rare domus and different habitation cores associated with small burial grounds and between them empty spaces with gardens or orchards. Tridentum (Trento) ( Figs. 2.34, 2.35) and its trans- formation is poorly known as the now densely inhabited area of the former Roman city offers few possibilities for investigation (Ciurletti 2003; Bassi, Cavada 2013). In the second half of the 3rd or early 4th century, the old city wal s of a Late Republican date were reinforced to reach the thickness of 3 m and include an amphitheatre Fig. 2.35: Tridentum. Hill Doss Trento from the south-east in front of the wal s. The city presumably had horrea for (1987). supplying the army on the Danube. In the second half of the 4th century, a church was constructed in the area of the forum. At the same time, a cemeterial church was plateau above the river al ow the conclusion that at erected in the vicinity of the Verona Gates that received least part of the population occasional y retreated there the relics of St Vigilius already around 400. With regards (Bierbrauer 1985, 497–498; Cavada 2019). In addition, to the living standard, the city shows similar signs of recent research notes that the city area of Piedicastello, degradation as others, with individual houses and parts protected with wal s and located between Doss Trento of the city abandoned, as well as new and poorly-built and the River Adige, was more intensely inhabited in dwellings constructed. The former domus now hosted the 6th century (Pisu 2019). Given the exposed strategic small primitive huts, of which postholes remain, as well location of the city in the narrow part of the Adige valley as hearths set up on the mosaic floors and new floors and along a major Roman road, we should also expect made of an inferior quality. We should also mention an occasional increased presence of the army; with this small groups or single burials in the city centre. A par- in mind, the habitation traces on both sides of the river ticularly important source for understanding the Late are understandable, albeit not ful y explained. Antique city is Cassiodorus, who reports of Theoderic’s Following Diocletian’s reforms, Aquileia ( Figs. edict that urges the inhabitants of Tridentum to move to 2.36–2.39, 2.254) became the capital of the province the nearby hill Doss Trento. It is unclear to what measure Venetia et Histria. It was a major administrative, eco- this migration took place, but the double church on nomic and ecclesiastical centre, as well as the main trad- the hill and the thick cultural layer on the vast rocky ing post for supplying the army deployed to the Danube 42 2. CITIES Fig. 2.36: Aquileia. City plan (Jäggi 1990, Fig. 1 ) Fig. 2.37: Aquileia. Late Antique wal s in the harbour (2003). Fig. 2.38: Aquileia. Remains of the domus in the city centre (2010). 43 Slavko CIGLENEČKI mid-5th century; this break is visible in the numerous reliably dated layers of burnt debris that have been as- sociated with Attila’s campaigns in Italy. After this time, an effective defensive wall of a demanding construction was built that reduced the urban area. The wall was in the shape of a zigzag line with the tips reinforced with small towers. Recent investigations show this wall dates to the 6th century and is very similar to Justinian’s city wal s in Durostorum (Silistra) (Groh, S. 2012; Atanasov 2013). Outside the reduced urban area, only stray finds have come to light that represent scarce habitation traces (Vil a 2004, 614). In addition to other functions, Aquileia was also the most prominent centre of Christianity in this part of the Empire, boasting the earliest episcopal complex. The imposing and through its transformations also characteristic complex aptly il ustrates the development and main phases of Early Christian architecture from the 310 onwards. It is joined by several suburban churches that include two large ones from the 5th century, namely the basilica at Fondo Tullio del Beligna and the basilica at Monastero. In the 4th century, a large double horreum was erected south of the episcopal basilica , but fell into disuse in the mid-5th century. Presumably dating beyond this chronological boundary are three commercial buildings connected with the River Natisone and arranged round a court (Marano 2012, 579). As for the public buildings, new high-quality floors were installed in the area of the large baths sometime between the 5th and the 6th century Fig. 2.39: Aquileia. Mosaic and remains of the domus in Fondo (Marano 2011, 177). Several prestigious domus were also Cossar (1997). constructed in the city, most often with large reception hal s that were fitted with apses in the 4th and continued to be inhabited in the 5th century, some were even renovated Basin (overview in Jäggi 1990; Cantino Wataghin 2004; in the late 5th century (mosaic in the Domus at Fondo Sotinel 2005). The location along the main road leading Cossar) (Sodini 1995, 167; Baldini Lipolis 2001, 144). to the heart of Italy from the east, not far behind the A large number of these houses was abandoned, with Alpine barriers, dictated the changes in its defensive in- smaller and more modest subdivided rooms arranged stal ations and urban fabric. The sieges, numerous stops within them that revealed hearths and postholes, some of army troops in the city, barbarian incursions, as well even traces of artisanal activities. This subdivision was as lively religious and political activities of the Aquileian mostly investigated during the early excavations, and church coupled with the changed socio-economic and can therefore not be precisely dated (Marano 2011, 177). administrative conditions greatly influenced the trans- After the mid-5th century, a large part of the population formation of the city. Its fate and all its alterations echoed moved to the better fortified south part, while workshops across wide areas, particularly in the more exposed parts and simple huts appeared in the north part. In the second to the northeast. The historic sources make the reasons half of the 5th and the 6th century, Aquileia was a typical for the fate of the city in Late Antiquity clear. They also fortified Late Antique city with grand church complexes, reveal a gradual migration of the population and final y, rare maintained domus and predominant modest and in the second half of the 6th century, the transfer of its primitive architecture with signs of a pronounced rurali- main administrative and church functions to Grado sation of the remaining urban area (Vil a 2004, 561–567). (overview in Bratož 2000; Cuscito 2001; Sotinel 2005). The downfall of Aquileia was in a large measure brought Archaeological evidence reveals the reinforcement of the about by the decline in trading and its intermediary role city wal s, the shrinking settlement area and the reduc- in the distribution of goods, which is also reflected in the tion of most public buildings (Jäggi 1990; Bonetto 2004; weakening river harbour. Concurrently with the shrink- Vil a 2004; Cantino Wataghin 2004). A city that was rich ing urban core of Aquileia, there was a gradual rise of a and enjoyed its full extent and heavy fortifications even settlement some 10 km away called castrum Gradense in the late 4th century, witnessed a sharp break in the (Grado), established on an island in the lagoon. Its name 44 2. CITIES Fig. 2.40: Forum Iulii. On the eastern and southern sides the city was protected with the River Natissone (2015). Fig. 2.42: Forum Iulii. Part of the wal s with a proteichisma and two towers (Bonetto, Vil a 2003, Fig. 1). alone is symptomatic of a different type of settlement, which also received the metropolitan see in the second half of the 6th century (Mirabel a Roberti 1974–1975; Tavano 1976; Cuscito 2001). Located not far from Aquileia is Forum Iulii (Cividale) ( Figs 2.40−2.42), which contrary to Aquileia significantly gained in importance in the 5th and 6th centuries due to its strategic significance and advantageous defensive location (Brogiolo 2010, 71–72; Marano 2011, 180–182). In the 6th century, its wal s were reinforced with proteichisma and enlarged pentagonal towers (Bonetto, Vil a 2003; Vitri et al. 2006). The remains of a lavish domus with an apsidal hall were partial y investi- Fig. 2.41: Forum Iulii. City plan (Vitri et al. 2006, Fig. 1). gated at Piazza Paolo Diacono that was presumably built in the second half of the 5th century in the area where the forum is believed to have stood. The importance of this house is underscored by a rich Langobard burial 45 Slavko CIGLENEČKI Fig. 2.43: Tergeste. Roman remains and Late Antique basilica on the hill of San Giusto (2014). from the 7th century. Partial traces of other domus also came to light at other sites. The octagonal baptistery indicates that an Early Christian church stood where the cathedral currently stands. The continuity and a greatly increased importance of the city in the Langobard era, when it became one of the centres of the Langobard state, is reflected in the numerous and rich cemeteries located in different suburban areas. The city of Tergeste (Trieste) ( Fig. 2.43) in the 4th century spread across the slope and summit of the hill San Giusto fortified with a defensive wall (Maselli Scotti 1990; Marano 2011, 183). In the time of the Byz- antine reconquista, the summit was turned into a fort enclosed with a wall of reused stones and reinforced Fig. 2.44: Teurnia from the east (2014). with pentagonal towers. In the place of the capitolium, the basilica of St Justus was constructed in the first half or middle of the 5th century, and was later remodelled. Traces of oil and wine production show that life also 1992, 39–43, 81–123; id. 2002, 141–143; Ladstätter 2002, continued, albeit in a limited degree, at the foot of the 319–320; Dolenz et al. 2016). Its location on top of a hil . The cemeterial basilica from the first half of the 5th natural y well-protected hill above the Drau/Drava ena- century at Via Madonna del Mare is well-known, located bled a continued existence of the city on the same spot in a suburban area. It was renovated and decorated in that, from the second half of the 3rd century onwards, the mid-6th century, its colourful mosaic has numerous also gradually welcomed the inhabitants previously donor’s inscriptions mentioning a variety of functions of living on the slopes below. Eugippius refers to the city a wel -organised church community from the 6th century as metropolis Norici (Vita Sancti Severini 21, 2). It was (Caillet 1993, 270–271). abandoned in the Middle Ages, which allowed for a good preservation of its remains. Its excavations, however, were conducted in the early 20th century and their poor AUSTRIA documentation does not enable a reliable dating of the city’s individual elements. The only city in both Norican provinces that The urban settlement extending over 10 ha was exhibits signs of some continuity to the end of the 6th most likely enclosed with thick wal s already in the century is Teurnia ( Figs. 2.44–2.46) (Egger 1914; Glaser 4th century. The walls later received additions that 46 2. CITIES Fig. 2.45: Teurnia. Plan of the Late Antique city (Glaser 2014, Fig. 5). Fig. 2.46: Teurnia. Plan of a part of the defensive wal s with a proteichisma and Late Antique buildings in the interior of the settlement (Egger 1914, Fig. 12). included defensive towers and a heavily fortified en- routes leading to Italy from the north, which led past trance. Egger’s plan shows that the outer wal s were the city and is corroborated by the nearby fort at Duel reinforced with a proteichisma in the exposed part of (Ciglenečki 2012, 462). the city. Such reinforcement is ful y understandable An Early Christian episcopal church was con- as part of Justinian’s protection of one of the main structed in the 5th century and extensively renovated 47 Slavko CIGLENEČKI after a fire in the first half of the 6th century (Glaser ing construction work without any documentation. 1997, 133). Right next to it a xenodocheion ( hospitium) According to Branka Migotti, the most likely location was found, which is the first example of a public build- of the episcopal basilica is in the southeast corner of ing with such a function in the eastern Alpine area the city, where Early Christian sarcophagi came to light (Glaser 1992, 93–103, 107–112). Different fragments (Migotti 1994, 47). Late Antique architecture is poorly of Early Christian church furnishings indicate the known, the remains of buildings from this period were presence of at least two other sacral buildings (Glaser only sporadical y found, but even these do not appear 2015, 14–18; Dolenz et al. 2016, 116–117). A cemeterial to form larger complexes. Baths were still active in the church lies outside the settlement (Egger 1916, 12–55). 4th century, as was a horreum. There are abundant smal The settlement interior revealed several independent finds from the 4th and even the first third of the 5th two- or three-room houses of a Late Antique date century, whereas the finds from the second half of the (Egger 1916, 55–56). Of these, more is known of the 5th century are rather rare. The numbers again rise in large house with an apse and heating canals, which was the first half of the 6th century, when Roman artefacts a modest urban vil a (Glaser 1992, 107–112; Ramstet- are joined by those of typical y Gothic, Byzantine and ter 2019). Different renovations of earlier buildings also Langobard costumes (Simoni 1989). The second have also been recorded in the vicinity of the forum half of the 6th century is again poorly represented and (Eitler 2017, 94–102). The distribution of buildings it would appear that Siscia already lost its urban char- in the late phase of the settlement does not indicate a acter in this time. planned urban layout. Numerous small finds, as wel as a partially excavated cemetery prove that life in In contrast, urban continuity is strongly felt on the Teurnia lasted to the late 6th century (Piccottini 1976; east coast of the Adriatic. Although marked with lowered Glaser 1997, 140–141). urban standards in comparison with earlier times, it did boast new constructions in defensive, residential and particularly sacral spheres. CROATIA In the Pannonian part of modern-day Croatia, there is a single identifiable city with continuity, namely Siscia (Sisak) ( Fig. 2.47), the Late Antique capital of the province Savia. The completely built-up modern city offers archaeological and literary evidence that suggests continuity to the end of the 6th century (Gračanin, Bilogrivić 2016, 116–117). It was strategical y located at the junction of several major communications, at the confluence of the Rivers Kupa and Sava and hence naturally well-protected. Siscia was an important starting point for the nearby mines of metal ores in present-day north-western Bosnia. After Diocletian’s division, it became the seat of the province Pannonia Savia and enjoyed a particularly prominent role in the 4th and first third of the 5th century. Sources mention different high administrative and army officials, as wel as an imperial mint and a unit of the Danubian fleet stationed here. In the 6th century, Cassiodorus men- tions curiales heading civil administration and a Gothic comes governing the city. For a brief period, there was also a mint of Ostrogothic coins operational in the city (Demo 1994, 134–136). Two bishops are mentioned for the first half of the 6th century (Bratož 2011a, 241). Recent investigations and the recovered small finds offer only a limited insight into the constructions and development of the Late Antique city (Simoni 1989; Migotti 1994, 47–48; Lolić 2003, 143–144). The course of the city wal s is only outlined. There are no reliably data on the Early Christian architecture. Two apsidal buildings are mentioned, which were destroyed dur- Fig. 2.47: Siscia. City plan (Migotti 2015, Fig. 13). 48 2. CITIES Fig. 2.48: Parentium. Basilica and episcopium from the north-east (2010). Fig. 2.49: Parentium. Interior of the episcopium (2010). Parentium (Poreč) ( Figs. 2.48, 2.49, 2.255) is a Exceptional in several aspects is the episcopal complex city with a strong habitation continuity, established with a double church and an episcopium. Research in on a peninsula that offers excellent natural protec- the current cathedral has revealed a development from tion. The long centuries of later occupation are also a simple church complex built (most likely) in the late the reason for a lack of elements to shed more light 4th century, to a double, ‘pre-Euphrasian’ church of the on the city in Late Antiquity. We know of its extent mid-5th century and finally the Euphrasian basilica and part of the city wal s, layout of communications of the mid-6th century, which ranks among the most and three Early Christian churches (Matejčić 2007). beautiful y decorated churches of the Justinian period 49 Slavko CIGLENEČKI (Matejčić 2014). The completely preserved episcopal palace is an excellent example of prestigious housing, the construction of which has also been posited in the time of Bishop Euphrasius (Matejčič, Chevalier 2012). Its ground floor has a large central hall surrounded by smaller utilitarian rooms, while the first storey boasts an aula palatina with an apse and a triumphal arch in the shape of a tribelon. The building represents completely preserved housing of the elite from the mid-6th century, whose layout of rooms, multi-storeyed construction and decoration reveals in details the appearance of the last prestigious buildings of Antiquity. The colony Pola (Pula) ( Figs. 2.50–2.52) gained in significance in Late Antiquity due to its exceptional geographic and strategic location at the tip of the Is- trian Peninsula on the navigable route to Ravenna. In this time, it witnessed changes visible primarily in the numerous newly-built churches (Marušić 1967, 9–11, 22–26; Matijašić, Buršić Matijašić 1996). As a prominent stronghold on the navigable route along the coast and a point of quickest crossing of the Adriatic, the city also served as a maritime base for the Byzantine conquest of Fig. 2.50: Pola. City plan (Marušić 1967, Fig. 1). Ravenna. In the 5th/6th century, the Late Republican city wal s with defensive towers were renovated and added a new exterior facing. Early excavations unearthed three round towers, which were later made polygonal along the exterior (Starac, A. 2001, 65). A defensive structure built of spolia was found on the interior side at the Gate of Hercules that presumably supported an earthwork rampart. A large episcopal centre was built in the 4th/5th century. The imposing, lavishly decorated church of Maria Formosa was constructed in the middle of the 6th century, on the order of Bishop Maximianus of Ravenna, who was born in nearby Veštar (Vistar). In addition to the churches inside the fortified area, there are also numerous extramural churches, both in the immediate vicinity and farther away (Marušić 1967, 11, 15–16, 23–26). Recent rescue investigations have revealed a large part of the Fig. 2.51: Pola. City wal s and defensive towers by the gates. city’s interior, known as St Theodore’s quarter (Starac, A. Fig. 2.52: Pola. Remains of Early Christian church of Maria Formosa with a ful y preserved chapel (2014). 50 2. CITIES Fig. 2.53: Nesactium. City plan (Rosada 1999, Fig. 20). Fig. 2.54: Nesactium. Ruins of the double Early Christian church (2011). 2009), with numerous constructions dating from the it a modest building on flimsy foundations. Part of the beginnings of the Roman city to the modern period. In abandoned space yielded traces of economic activities Late Antiquity, the area hosted a luxury domus and public including oil presses. Numerous archaeological finds baths in the vicinity of a temple. The domus, extensively support the continuity of this part of the city throughout renovated already in the early 2nd century, was destroyed the 5th and to the 7th century. together with the baths in a fire in the second half of the The smal Roman city of Nesactium ( Figs. 2.53, 5th century. Soon afterwards, a small church of St Lucia 2.54), located on the spot of the former important hil - was built in the area of the destroyed baths and next to fort of the Histri, later transformed into a typical Late 51 Slavko CIGLENEČKI Fig. 2.55: Iader. Plan of the wal s on the southwestern part of the city with Late Antique proteichisma and the gates (marked with number II) (Suić 2003, Fig. 61). Antique city with reliable evidence of life in the 5th and 6th centuries (Rosada 1999). Towards the end of the 4th century, it was enclosed with wal s, 1.6 m thick, that fur- ther increased its defences on a natural y well-protected location. A protruding rectangular tower was unearthed in one part of the wal s. In addition to small finds, the chronology and character of the city is revealed by the well-researched double Early Christian church with a baptistery, which was built in the 5th century at the edge of the Roman forum and remained in use to the end of the 6th century. Nesactium is thus a typical example of a prehistoric hillfort that was an important local centre also in the Roman and Late Antique periods and was final y abandoned at the end of Antiquity. The major port of Iader (Zadar) ( Figs. 2.55, 2.56) boasts a location on a natural y well-protected peninsula, in the central part of the navigable route along the east Adriatic. Along the exterior of the city wal s from the Augustan period, where access from land was easiest, an equal y thick proteichisma was built in Late Antiquity, the previously wide entry gate was closed and only a narrow passage left (Suić 1976b). Extensive traces of Early Christian buildings from the 5th and 6th centuries came to light here, which largely respected the earlier urban layout (Suić 1976a, 246; id. 1981; id. 2003, 350, 367; Uglešić 2002, 11–33). Standing out is the church complex from the late 4th century, sited at the edge of the Roman forum. The city also hosted numerous small Fig. 2.56: Iader. Remains of the forum and basilica (2012). intramural churches of the 5th and 6th centuries. Wine and oil presses in the urban centre are evidence of a heavy ruralisation of the city (Suić 1976a, 248). Early Byzantine reinforcement of defensive wal s with a proteichisma can also be observed in two smaller cities, namely Asseria and Varvaria in the hinterland 52 2. CITIES Fig. 2.57: Asseria with well-fortified western wal s (2020). Fig. 2.58: Asseria. South-western corner of city wal s with a Fig. 2.59: Varvaria. City wal s reinforced in Late Antiquity tower and a proteichisma (2004). (2020). of Zadar, both of which revealed extensive reparations ers and buttresses, and adding a drystone proteichisma of earlier wal s. They aptly il ustrate the construction that incorporates numerous spolia and traces the main efforts during Justinian’s reconquista in cities and – as wal s at a distance of up to 10 m. The proteichisma can we shall see below – even more frequently in coun- be attributed to the Justinian period, as is the case in tryside forts. numerous forts along the eastern Adriatic navigable route that are discussed below. A tomb from the first Not much is known of the interior of Asseria half of the 6th century located along the city wal s and (Podgrađe) ( Figs. 2.57, 2.58, 2.251), a Roman city estab-covered by one of the buttresses supports a Justinian lished on the spot of an old Liburnian settlement. Early dating. In the city interior, the foundations of an Early investigations focused on the remains from the 1st and Christian church with an apse came to light in what re- 2nd centuries. The city’s last peak presumably dates to mained of the forum. The furnishing remains of another the times of Diocletian and Constantine (Liebl, Wilberg sacral building, presumably of a cemeterial nature, were 1908). The modern, large-scale systematic excavations found outside the city wal s (Uglešić 2002, 65–66). The have revealed a typical Early Byzantine fortification of extensive reinforcement of defensive instal ations in the the city wal s (Fadić 2001, 78–79). The old Roman wal s, Justinian period raises the possibility of strong garrisons then (and even today!) still excellently preserved, were stationed here; we may even posit the influx of people cleverly adapted by reinforcing the exterior with tow- from surrounding areas. 53 Slavko CIGLENEČKI Fig. 2.60: Varvaria. Late Antique city wal s on the south-western part (Suič 2003, Fig. 5). Fig. 2.61: Varvaria. Hill with the city from the south (2006). A similar transformation of an old prehistoric Ostrogoths in Dalmatia (Suić 1976a, 234; id. 1980, 40). centre and Roman municipium has been observed at Modern excavations conducted in the city interior have Varvaria (Bribirska glavica) ( Figs. 2.59–2.61). The ad-revealed an unusual church building – rotunda with ditions to the Tiberian city wal s are less characteristic, eight apses – dated to the 6th century (Milošević A. 2017; but also include thick buttresses and towers protecting Ghica et. al. 2019). entrance gates. The small finds and a coin of Justin- Municipium Riditarum (Danilo) ( Figs. 2.62–2.64) ian led Mate Suić to date this phase of the wal s to the lay on the major Roman road from Scardona to Salona in time of renovations following the victory against the its first few centuries, while its Late Antiquity inhabitants 54 2. CITIES Fig. 2.62: Municipium Ridi- tarum. Hill Gradina from the south (2020). Fig. 2.63: Municipium Ridi- tarum. Remains of Late An- tique wal s (2006). Fig. 2.64: Municipium Ridi- tarum. Remains of a Late An tique building cut into the bedrock (2006). 55 Slavko CIGLENEČKI Fig. 2.65: Salona. City plan (Jeličić-Radonić 2006, Fig. 9). Fig. 2.67: Salona. Defensive wal s with a tower (2021). (Zaninović 1978, 17–29; Suić 2003, 114; Brajković et al. 2013; Ciglenečki 2020, 263–264). Adding to the appear- ance of an agglomeration on the hill is a system of several cisterns and a stairway hewn into the bedrock. The limited investigations prevent a more definite description of the Late Antique city. Neither is its status clear; it is possibly merely a settlement with a fort, devoid of urban features. Coins and other small finds confirm its existence into the 6th century. An Early Christian church has been investi- Fig. 2.66: Salona. Episcopal complex and well-fortified wal s in gated below, at the Šematorij site within the old city. It is the northern part of the city (Chevalier, Mardešić 2006, Fig. 6). as yet not possible to date the phases of abandonment of the old city or explain the relationship between both parts; it is also not possible to exclude the possibility of moved to the hill Gradina above the city. The latter is the their simultaneous existence. site of a prehistoric hillfort that today hosts the remains In the area under discussion, a city that was particu- of Late Antique buttressed wal s (acropolis, fort?). The larly important in Late Antiquity is Salona (Solin) ( Figs. slope below has the remains of roughly fifty houses built 2.65–2.70), a colony that increased in size during the in the drystone technique and sunken into the bedrock Tetrarchy to encompass an urbs orientalis, and persisted 56 2. CITIES Fig. 2.68: Salona. Episcopal complex in Urbs orientalis (2021). Fig. 2.69: Salona. Manastirine from the west (2021). in this extent to the early 7th century (surface of ca 50 ha?). thick defensive wal s erected under Marcus Aurelius. An As the capital of the province Dalmatia, it witnessed extensive renovation of the city wal s occurred under reparations of its defensive wal s and lively construction Theodosius II, when epigraphic evidence reveals the activities reflected in numerous Early Christian churches renovation of all the towers and probably also individual both in and outside the city wal s (cf. Marin 1994; Suić sections of the wal s. A particularly valuable source for 2003, 365; Marin 2006). It represents one of the greatest our understanding of the fortifications in Salona is Pro- Early Christian centres of the Balkans that also ranks copius, who mentions that the wal s were fortified during among the best researched of its kind across the Empire. the Gothic War ( De bello Gothico V, 7.9 and 7.26–31). Salona was positioned so that several branches of the River Archaeological research has confirmed reparations of the Jadro ran through the eastern part of the city, supplying city wal s with towers, as well as a wide ditch dug in front water and affording additional protection. of the wal s in this period. The multitude of amphorae Much research has been dedicated to Salona’s defen- from the 5th/6th centuries filling the core of the city wal s sive features, particularly those in the exposed northern clearly shows the haste of the reparations. part (Jeličić-Radonić 1998, 30–32). It has shown gradual From the 4th to the late 6th century, the centre of the addition of new towers and different reparations of the city was at the episcopal complex with a double church, 57 Slavko CIGLENEČKI Fig. 2.70: Salona. Remains of the so-called ‘Five Bridges’. (2021) large baptistery and episcopal palace. The city boasted other intramural churches, as wel as large cemeterial complexes at Manastirine, Marusinac and Kapljuč. This exceptional church architecture clearly mirrors the intro- duction and significance of Christianity in the province. Much less is known of the housing in Salona, as the whole first hundred years of archaeological investi- gations were focused on the Early Christian buildings. Recently, small parts have been investigated in the city interior (Chevalier, Mardešić 2006) that revealed aban- doned parts of the city and a ruralisation of the urban area (e.g. torcular for pressing grapes found at the edge of the forum). Sparser habitation is posited in the eastern part of the city that holds several branches of the Jadro. There are even known examples of burials intra muros. The remains of luxury architecture with mosaic floors indicates habitation cores in the interior of the fortified city, similar to those observed in other cities. Coins and other small finds prove that Salona was gradual y abandoned and the hoard buried in 639 indicates its final abandonment (Marović 1984, 298). The city revealed no traces of a violent destruction. It rapidly declined in the final decades of its existence, the inhabitants left and most of the remaining population emigrated to the Palace of Diocletian, some 5 km away. The old Greek city on the island of Hvar, Pharia (Stari grad) ( Fig. 2.71), is an example of a city on an island Fig. 2.71: Pharia. Remains of double church with baptistery that enjoyed the status of a municipium in the Roman (2005). period (Katić 2003, 453–454). In Late Antiquity, defensive wal s were constructed on the north side that decreased the size of the former city by half; according to Miroslav double church with a baptistery decorated with mosaics Katić, this represents the only example of a reduced urban and wall paintings that represented the cult centre of the fabric on the Dalmatian coast. The interior revealed a Late Antique city (Jeličić-Radonić 1994, 18–87). 58 2. CITIES 2.4.2 EASTERN PART new administrative functions brought exceptional development in the 4th century with a hive of construc- SERBIA tion activities that produced a palace, hippodrome, horreum, large urban vil as and several baths (Popović, Sirmium (Sremska Mitrovica) ( Figs. 2.72–2.75), V. 1982; Jeremić, M. 2006; Popović, I. 2008; Heinrich- one of the most important Late Antique cities in the Tamáska 2015). Martyrial churches were constructed in Empire, underwent a particularly characteristic trans- the 4th century in the suburbs, in the early 5th century formation and fate that is paradigmatic of a large part an Early Christian church in the centre of the city of the area under discussion (Popović, V. 1982, 549; (Mirković 2011, 87–90). Jeremić, M. 2006; Mirković 2011; Milinković 2015, It was Vladislav Popović who conducted the fun- 111–120; Heinrich-Tamáska 2015, 58–66; Popović, I. damental research that il uminates the continuity of the 2017). Under Diocletian, it became the capital of the city in Late Antiquity (primarily Popović, V. 1982, also new province Pannonia Secunda and of the Pannonian col ected papers in Popović, V. 2004). He established that diocese. Several emperors resided in the imperial palace, a church was constructed on a levelled layer of debris the city boasted a major mint, as well as a manufacture of towards the end of the 4th or beginning of the 5th century, weapons and a state weaving mil . Christianity became following a barbarian attack. A small necropolis dating prominent here very early on, bishops are mentioned to the first half of the 5th century was created around all to 441, in the middle of the 4th century it was even this church. Nothing reliable is known on the existence considered the centre of Arianism. of the church after this time, particularly after the Hun The city was protected with the River Sava in the incursion in the mid-5th century. Shabby buildings were south and marshes in the north; it received defensive unearthed in its vicinity, made of debris material and wal s in the 3rd century. Its geostrategic location and bound by clay; Popović allows for the possibility that Fig. 2.72: Sirmium. City plan (Jeremić, M. 2006, Fig. 7). 59 Slavko CIGLENEČKI Fig. 2.73: Sirmium. Traces of later additions to the Roman vil a with baths (Popović, V. 1982, Fig. 9). Fig. 2.74: Sirmium. Late Antique buildings (2016). some better constructed earlier buildings remained in to this time are modest huts constructed in the courts of use up to the Hun attack. an urban vil a and the lavish building interpreted as the After the mid-5th century, during the presence of the imperial palace from the 4th century. The huts were made Ostrogoths and Gepidae, the urban area was reduced to of light materials or brick fragments bonded with clay. the southernmost part along the Sava. Earlier architecture The only complete residential complex was that had already been destroyed by this time and was uninhab- created in the 6th century inside the hippodrome. These ited. Cores of habitation formed in the open spaces that modest dwellings had rectangular hearths and presum- were associated with contemporary cemeteries. Dating ably lasted from the late 4th century to the end of the 60 2. CITIES city’s existence. Recent research has brought to light several new primitive buildings, among them also a large building with baths from the 6th century, which reveals the hippodrome as an especial y protected habitation core (Jeremić, M. 2006, 146; Heinrich-Tamáska 2015, 65; Milinković 2015, 117–118). The utterly reduced and impoverished settlement of the 5th and 6th centuries is in stark contrast with the former splendour of the city. Even more than elsewhere, this situation raises the question of whether it is at all pos- sible to speak of urban agglomerations in this time. The question of the church organisation and the ratio between the Roman and non-Roman population remains open. What is certain is that Sirmium of the 6th century was only a pale reflection of its former glory as an imperial capital and was only composed of small habitation cores, with a strong non-Roman component in its population. As for many other cities, the current state of research does not allow a confirmation that Singidunum (Beo- grad) ( Fig. 2.76), a city at the confluence of the Danube and the Sava, was continual y inhabited to the end of the 6th century or whether a previously completely abandoned city was modestly re-inhabited in the Justinian period (Milinković 2015, 120–126). Archaeological evidence indicates a long hiatus following the Hun incursion, Fig. 2.75: Sirmium. Remains of the Late Antique palace (2016). which is partial y filled by the small cemeteries showing the presence of groups of foreigners (Ivanišević, Kazanski 2002). Presumably dating to the Justinian period are the remains of a horreum within the former legionary fortress, while scarce traces of a small settlement from the 6th cen- Fig. 2.76: Singidunum. City plan (Milinković 2015, Fig. 76). 61 Slavko CIGLENEČKI Fig. 2.77: Viminacium. City plan (from Milošević, G. 2002, Fig. 4). tury came to light along the Danube outside the fortress. as the newly-established post resembles a small fortified Mihailo Milinković observed that habitation continued settlement or refuge rather than continuation of the for- sporadical y after the Hun incursion, but that it was not mer city (Popović, M. 1988; Milinković 2015, 126–136). possible to speak of urban continuity. Singidunum then This settlement is interpreted as a military post established resembled a partial y destroyed fort that hosted groups in the time of Justinian and later also settled with foederati. of rural inhabitants (Milinković 2015, 125). In contrast, Considering that the late sources mention the location Procopius reports that Singidunum with its Justinian of the new city on an island of the Danube, the problem wal s became ‘a very important city’, which is indirectly of the city’s continuity into the 6th century, from which confirmed by the mention of a bishop from the 6th century bishops are also mentioned, remains open. (Claude 1969, 200; Bratož 2011a, 228). The importance of Naissus (Niš) lies in its location A disparity between the accounts in literary sources along the main road leading from the west to Constan- and the unearthed archaeological remains is particularly tinople, on a major crossroads (Milinković 2015, 137; noticeable in the case of Viminacium (Kostolac) ( Fig. Ivanišević 2016, 93–94; Jeremić, Čerškov 2021). As the 2.77), hence it is only conditional y included in the group birthplace of the Emperor Constantine, it gained in of cities with continuity. Sources mention Viminakion on significance in Late Antiquity and witnessed numer- a large island in the Danube (Milinkovič 2015, 126–136; ous prestigious houses constructed in the suburban Ivanišević 2016, 91–92). The main and also largest city Mediana. In a very vivid account, Priscus reports that of Upper Moesia underwent major settlement changes the Huns captured Nais in 441 or 447 and razed it to the already in the first half of the 5th century. The last habita- ground. In 473, the Goths headed by King Theodimer tion layer only yielded the remains of poorly built houses and Prince Theoderic set up temporary base here. Pro- with wal s constructed of reused stones bound by clay. copius mentions the city in connection with Justinian’s There are no archaeological traces postdating the Hun renovations that also included the former residential incursions and only few from the 6th century. quarters in Mediana. Several bishops are known from Procopius reports that Justinian renovated the this city. There is almost no archaeological evidence on torn-down Viminacium, suggesting that we could expect the city and its structure, though there is mention of a significant archaeological remains from this time. These part of defensive wal s and other small finds of unreliable have as yet not been found, raising the possibility that date. In contrast, the associated cemeteries do confirm the small settlement of simple huts established on the continuation, particularly the one at Jagodin Mala that river peninsula at the Svetinja site and protected by thick points to a Roman population. The cemetery revealed defensive wal s could represent part of these Justinian numerous graves, 65 beautiful y decorated Early Chris- renovations. Such a hypothesis is questionable, however, tian tombs and five churches. 62 2. CITIES MONTENEGRO One of the most prominent cities in Dalmatia, Doclea (Podgorica) ( Figs. 2.78–2.80), came under the province Praevalis following Diocletian’s reorganisation. This Late Antique city is poorly known; sources from the 5th and early 7th centuries mention bishops. It was located at an important junction of Roman roads and at the confluence of the Rivers Zeta, Morača and Širalija, hence natural y excellently protected from three sides. It was additional y protected with 2–2.3 m thick defensive wal s reinforced with towers on the most easily acces- sible, eastern side. The most important public buildings were already excavated in the second half of the 19th and beginning of the 20th century, but the excavation records Fig. 2.78: Doclea. City plan (from Sticotti 1913, Plan). do not allow for a reliable assessment of the city in Late Antiquity (Sticotti 1913; Mijović Kovačević 1975; Geli- chi et al. 2012; Tufi 2012). The archaeological remains include two Early Christian basilicas (double church) that were presumably part of an episcopal palace. They have been dated to the 5th and 6th centuries, renovated under Justinian and abandoned in the face of Avaro- Slavic incursions in the early 7th century. Of interest is their location in the centre of the eastern part of the city, away from the main public buildings of the pre-Christian era, 200 m from the forum. We may posit other con- temporary buildings in the eastern part, similarly as in the urbs orientalis in Salona, but the area has as yet not been investigated. Two necropoleis are known, the later of which is dated to the 2nd to 5th centuries. Fig. 2.79: Doclea. Remains of the buildings in the forum (2009). Fig. 2.80: Doclea. City wal s preserved to a substantial height (2009). 63 Slavko CIGLENEČKI KOSOVO roughly 36 ha, archaeologists found traces of a rectilinear street grid with a differing orientation of buildings that Ulpiana (Iustiniana Secunda) (Gračanica) ( Fig. suggests at least two construction phases. The large two- 2.81) went through an interesting transformation in Late aisled building may be a basilica in the forum, while an Antiquity (Parović-Pešikan 1981; ead. 1989; Teichner episcopal centre may have existed next to a polyconchal 2015a; id. 2015b). The Roman city, which was the centre of baptistery. a mining region, was established at Gradina. Early excava- Another fortified area was discovered in the im- tions here unearthed part of the city wal s and the north mediate vicinity of the original city, at the Bedem site. entrance gates, as well as an Early Christian basilica in its It covers the surface of only 16.5 ha and was enclosed vicinity. The Late Antique remains show three phases. The with a 3 m thick wall reinforced with semicircular and first was a renovation that took place towards the end of pentagonal towers (Teichner 2015a). Its interior re- the 3rd or beginning of the 4th century following a large vealed a building with a triconch, a horreum and other fire and destruction of the city. The next renovation is simple rectangular buildings. Considering the note of believed to have occurred towards the end of the 4th or Procopius, who mentions the renovation of Ulpiana beginning of the 5th century. The last phase is dated to the after an earthquake and Gothic Wars, as well as renam- 6th century and shows a partial abandonment of the city ing to Iustiniana Secunda, the remains at Bedem may and a destruction of the entrance during the earthquake represent the Early Byzantine city. However, parts of the in 518; in this time, a variety of artisanal facilities was set old city (Gradina) continued to function in this period; up in the abandoned buildings, also found was a smal the presumably martyrial basilica was enclosed with a group of graves (Parović-Pešikan 1989, 118–120). wall with towers ( quadriburgium) (Teichner 2015b). If The recent German-Kosovar investigations that Dietrich Claude wrote in 1969 that only the construction applied modern methods of prospection produced a of defensive wal s speaks in favour of changing the name great amount of data that shed new light on the Late to Iustiniana Secunda (Claude 1969, 203), the recent Antique remains. In the city interior, on a surface of investigations entirely corroborate Procopius’ note. Fig. 2.81: Ulpiana. City plan (Teichner 2018, Fig. 1). 64 2. CITIES BULGARIA Bulgaria − as a part of Byzantine territory − offers an excellent insights into the continuity of cities which recived strong support from Byzantine central authori- ties. The numerous investigations in these cities also allow us to identify their basic Late Antique elements. Serdica (Sofia) ( Figs. 2.82–2.84), the capital of the province Dacia Mediterranea, lay on the most important land route connecting East and West of the Empire. In Late Antiquity, it witnessed two major transformations. The first one occurred in the Tetrarchic or Constantin- ian period, when the city was enlarged from 18 to 84 ha and enclosed with new wal s (Dintchev 1999, 42–43; Fingarova 2015; Dintchev 2021a, 239–240). The size of its new part with as many as eight horrea suggest it was a strong military base and supply centre (Rizos 2017b, 24–25). It also went down in history as the place of an important ecclesiastical council in 342/343. The main public buildings were concentrated in the southern part and included a bouleterion, in its imme- diate vicinity also baths or, in the opinion of some, the Fig. 2.82: Serdica. City plan (4th century) (Rizos 2017b, Fig. 5) Fig. 2.83: Serdica. Baths and Rotunda of St George (2012). 65 Slavko CIGLENEČKI Fig. 2.84: Serdica. City plan (6th century) (Fingarova 2015, Fig. 2). palace of Constantine, as well as part of prestige housing garova 2015, 112). After the mid-5th century, two large (Ćurčić 2010, 51, 64). In the northern part, there were churches were built right next to the old city centre, as springs of mineral water and the remains of baths, as well as two Early Christian basilicas extra muros, in the well as different houses that included three spacious eastern cemeterial area. Of particular interest is one of peristyle domus. Also found was a great apse, which has the rare epigraphic monuments, which relates a concern been interpreted as the remains of an episcopal basilica, for the city in the late period and reveals that the local and further north a three-aisled basilica constructed on bishop had the city aqueduct renovated as late as 580. the foundations of an earlier domus located in proximity An important city, Novae (Svištov) ( Figs. 2.85– to the mineral springs. 2.87) was also located on an exposed part of the limes In the middle of the 5th century at the latest, the in Lower Moesia. It grew on the spot of a legionary city again shrunk to its original size. The thoroughly fortress and witnessed numerous changes in Late renovated city wal s are dated to the Justinian period, Antiquity that have been well-documented during the reinforced with triangular and pentagonal towers, as long Polish-Bulgarian investigations (Čičikova 1994; well as a proteichisma unearthed in some places (Fin- Poulter 1994; Biernacki 2005; id. 2013; Dintchev 2008, 66 2. CITIES Fig. 2.85: Novae. City plan (Dintchev 2018, Fig. 7). Fig. 2.86: Novae. Episcopal complex (2012). 407, 409–412). The heavily defended fortress from the the wal s. This led to an extension of 10 ha in the east 1st century extended over 18 ha and was associated with (Novae II); this extension initial y only served as a ref- adjacent canabae, 2.5 km further east also a vicus that uge and gradual y came to host civilian population. The was given the status of a municipium in the early 3rd refugees were first housed in the vacated valetudinarium. century. Following a Gothic raid in the second half of Later, part of the old city was intended for civilian use. the 3rd century, when houses outside the fortress were Modest buildings were constructed nearby and used as destroyed, numerous inhabitants sought refuge within workshops. This first civilian settlement was inhabited 67 Slavko CIGLENEČKI Fig. 2.87: Novae. Remains of the valetudinarium (2012). to the opening decades of the 4th century, when it was destroyed in a fire. An urban vil a was erected here in the mid-4th century. Both parts of the city joined together during the 4th century, a forum was built and other ci- vilian architecture. Research shows that the principium was only destroyed in the middle of the 5th century. The development of the city was brought to a halt by the Gothic Wars in 376–382 and later by Hun incursions. After the victory over the Huns, the city prospered into the Justinian period. The earliest mention of a bishop of Novae dates to the transition from the 5th to the 6th century. A large episcopal complex grew at the edge of the former forum and comprised several interconnected buildings: a large and a small basilica, episcopium, baptistery and auxiliary buildings. The complex stood on the spot of the legionary baths torn down in 376–382, hence the construction of the complex has been dated to the first quarter of the 5th century. As many as five renovations of the basilica have been established, the last one in the third quarter of the 6th century (Biernacki 2005). The Late Antique transformation of Nicopolis ad Istrum (Nikiup) ( Figs. 2.88–2.90) is well-known (Poulter 1995; id. 2007c, 51–82). Most of the old city, located in the hinterland of the Danube limes, was destroyed around 447. In the second half of the 5th century, strong defensive wal s with protruding rec- tangular towers were constructed in the natural y most protected part. The 5.74 ha large interior revealed two Early Christian churches and large, well-constructed Fig. 2.88: Nicopolis ad Istrum. City plan (Poulter 2007c, Fig. 1 ). buildings that Andrew Poulter presumes to have been army barracks and storehouses. This strongly 68 2. CITIES Fig. 2.89: Nicopolis ad Istrum. Plan of the Early Byzantine city (Poulter 2007c, Fig. 10). Fig. 2.90: Nicopolis ad Istrum. Ruins of the forum (2012). defended and well-built part of the city is believed as gardens. Ventzislav Dintchev notes the prospect to only have hosted the church elite and the army, of wooden buildings standing in the areas where possibly also storehouses for the imperial army. The geophysical surveys detected no masonry buildings. empty space in the interior was presumably intended He also allows for the possibility that the centre of for temporary army unit, as a civilian refuge or even ecclesiastical organisation moved from Nicopolis to 69 Slavko CIGLENEČKI Fig. 2.91: Pautalia. City plan (Dintchev 2018, Fig. 6). present-day Carevec, presumed Zikideva (see Chapter 2.5) (Dintchev 1997, 100–103). The data available on Durostorum (Silistra) in Late Antiquity are insufficient to allow a comprehensive reconstruction. Recent work did investigate the defen- sive wal s of the Byzantine fort constructed immediately next to the Danube, which only protected a small part (ca 5 ha) of the formerly large city (Atanasov 2013). The Justinian wal s are of a complex construction, built in a zigzag line with additional reinforcement of the triangular tips. This feature is closely paralleled with the wal s of Aquileia. The large (29 ha) Roman city of Pautalia (Kjus- tendil) ( Fig. 2.91) was already fortified under Marcus Aurelius. In Late Antiquity, new city wal s were built that considerably reduced its size (Dintchev 1999, 46). Several churches were also constructed, with some of the earlier buildings exhibiting Late Antique modifications. A 2.1 ha large fort was built on the nearby Hissarlaka hill already towards the end of the 4th or beginning of the 5th century. A passage in Procopius suggests that this fort on elevated ground was refortified and built-up in the 6th century (Kirilov 2006b, 68). Fig. 2.92: Philippopolis. City plan (from Dintchev 2018, Fig. 2). Philippopolis (Plovdiv) ( Fig. 2.92) is one of the largest cities in the area under discussion (ca 80 ha) and the capital of the province Thracia (Topalilov 2012). In this city of continuity, the elements of particular interest 70 2. CITIES first and foremost include the strong city wal s (Dintchev 2021a, 242–244). In the second quarter of the 6th century, the walled area was limited to the acropolis that extended across three elevations and their slopes. Even the episco- pal basilica alongside numerous other buildings of the old city in the lowland remained outside the new wal s. Among the Late Antique buildings, we should mention the lavish vil as from the late 3rd or the 4th century with magnificent mosaics, which were inhabited to the end of the 6th century (Valeva 2011, 28–41). An unusual example of a reduced city area in the Early Byzantine period can be observed in Deultum (Debelt) ( Fig. 2.93) (Vagalinski 2008; id. 2018). In the north, its Late Roman city wal s with a tower were constructed on top of the ruins of earlier Roman buildings; the wal s were finished prior to 383 and renovated in the Early Byzantine period, more precisely in the last third of the 5th or beginning of the 6th century. At this time, the city was radical y narrowed to an elongated rectangle that, for defensive purposes, reached to the terrace in the north and encompassed the river harbour in the south that enabled communication with the Black Sea. After a fire in the early 6th century, the north wal s were repaired and masonry buildings added that leaned onto the wal s. The city was destroyed soon after 574 Fig. 2.93: Deultum. City plan (from Vagalinski 2018, Fig. 2). (Vagalinski 2018, 87). The city of Mesembria (Nesebar) ( Figs. 2.94–2.96, 2.256) stood on the coast of the Black Sea, on an excel- lently protected peninsula (Velkov 1988, 223; Dintchev 2016, 313; Preshlenov 2012; id. 2018). Already in Hel- lenistic times, it had strong wal s in the part that offered easiest access. New wal s were constructed in the time of Marcus Aurelius. In the second half of the 5th cen- Fig. 2.94: Mesembria. City plan (Dintchev 2018, Fig. 12). 71 Slavko CIGLENEČKI Fig. 2.95: Mesembria. City wal s on the western side (2012). Fig. 2.96: Mesembria. Church of St Sophia (2012). tury, thick wal s were constructed in the opus mixtum the reason for Procopius not mentioning the city among technique on top of the destroyed wal s and reinforced Justinian constructions. The interior revealed three at either end with a round tower. The entrance was also better preserved large Early Christian basilicas and the renovated and fortified with a pair of pentagonal towers remains of two other basilicas. The large baths from the erected on the spot of earlier Roman rectangular towers. 6th century together with an oval square and the remains Traces of the city wal s also came to light in other parts of a street grid show that life and a high living standard of the city. The earlier creation of the wal s is probably persisted in the city. 72 2. CITIES ROMANIA towards the end of the 4th century and turned into a private vil a (Sodini 1997, 451–453). Part of the city in Following a Gothic destruction in the mid-3rd the west appears to have hosted economic activities ( hor- century, the city of Histria ( Fig. 2.97) in the Danube reum, smithies, bakeries). In the last quarter of the 6th or Delta was renovated, but reduced from 30 to mere 7 ha even first decade of the 7th century, the well-constructed and reinforced with strong city wal s (Angelescu et al. domus show minor subdivisioning (Munteanu 2011). 2017). The wal s date to the late 3rd or early 4th century and girdled the entire acropolis of the former city. It underwent several repairs and renovations, both under ALBANIA Anastasius and later in the 6th century. The readily acces- sible west side was additional y fortified with towers. In Scodra (Shkodër) ( Figs. 2.98–2.100), the capital of the 6th century, the earlier Hellenistic wal s on the west the province Praevalitana, is poorly known, but there are side were renovated and now represented a proteichisma some elements that offer an insight into its appearance that also served as shelter for refugees and their herds. between the 4th and the 6th centuries (Hoxha 2003; So- The city had three basilicas and another one extra muros dini 2007, 319; Dintchev 2021a, 237). It encompassed the (Achim 2012). Unearthed under the central, episcopal acropolis on a large and natural y excellently protected church that stands out in size and décor was its prede- Rosafa hill and the city in the lowland below that had cessor datable between the last quarter of the 4th and rivers protecting it on two sides. Between the rivers was the early 6th century. Located at the city wal s, near the a transverse wal , 2.1–2.4 m thick and reinforced with ‘Parvan basilica’, archaeologists were able to date the first wooden beams, dating to the early 5th century. The next church to the time of Anastasius and its enlargement to thick cultural layer after this time indicates a renova- the first half of the 6th century (Angelescu et al. 2017, tion of the city wal s in the Justinian period, which is 149). In the 6th century, several domus were constructed corroborated by the note of Procopius that Justinian on the east side that include one with an apsed audi- renovated the city. This renovation can be identified in ence hal . These are characteristic houses with rooms the addition to the earlier wal s, which took the form of arranged around a central court. Similar houses came 2.1 m thick wal s of a poorer construction and bound to light in the southwest, where baths were abandoned by firm mortar. Semi-oval towers were leaned on these Fig. 2.97: Histria. City plan (Achim 2012, Fig. 2). 73 Slavko CIGLENEČKI Fig. 2.99: Scodra. Hill Rosafa with the acropolis (2009). wal s from the exterior that were constructed either contemporaneously or only slightly afterwards. In the east, an earlier flood wall was presumably reused as a proteichisma in the 6th century. The interior revealed simple houses erected close to one another, indicating a densely built-up interior, as well as buildings of a better construction and the remains of mosaics dated to the 4th and early 5th centuries. The latter had firm concrete floors, on top of which a thick layer of burnt debris was Fig. 2.98: Scodra. City plan (Hoxha 2003, Fig. 22). found that held numerous coins and other small finds Fig. 2.100: Scodra. Lower city at the confluence of Rivers Drim and Buna (2009). 74 2. CITIES Fig. 2.101: Lissus. Remains of lower part of the city from the south-west with the acropolis at the background (2009). Fig. 2.102: Lissus. The entrance to the lower part of the city is protected by towers (2009). indicating a fire in the early 5th century. The later cultural obtain a 260 × 60 m large protected area in the lowland layers inside the buildings prove intensive life in them at the foot of the hil . The walled area was well-protected throughout the 6th and in the early 7th century. in the west, north and partly south with the River Drin. Municipium Lissus (Lezhë) ( Figs. 2.101, 2.102) The Late Antique wal s incorporate reused stones in witnessed substantial changes towards the end of the some parts and have interior pilasters in the distance 4th and beginning of the 5th century (Anamali 1993, of 5 m. Procopius does not mention the city among the 455–456; Hoxha 2014). It disintegrated into two sharply forts that Justinian renovated. Literary sources reveal divided parts: the lower city near the river and the up- that the city held an episcopal see, which may be iden- per city in the former acropolis where sections of Late tified in the large church complex under the church of Antique wal s prove occasional use also in the 6th century St Nicholas (Skenderbeg Tomb), of which parts of the (Hoxha 2014, 513). In the lower part, the old Hel enistic atrium and the baptistery have been unearthed (Hoxha wal s were partly reused and partly extended so as to 2014, 517–518). Outside the wal s of the lower part, 75 Slavko CIGLENEČKI traces of adaptation were observed in the rooms of the former baths; a Christian cult place was set up in the first half of the 5th century, but soon abandoned (Lehner 2004, 10–11). A large three-aisled basilica stood outside the wal s, on the other side of the river. Dyrrachium (Durrës) ( Figs. 2.103–2.106), the ter- minus of Via Egnatia and the birth place of the Emperor Anastasius, is densely built-up in the modern period and thus offers few elements for the reconstruction of its Late Antique appearance (Hoti 1996; Chevalier 2015, 229). There are certain incongruences in the dat- ing of the Late Antique wal s, clearly visible in certain places, but most authors agree that brick stamps indicate construction under Anastasius (Karaiskaj, Baçe 1975, 29–31; Karaiskaj 1998, 868; Baldini, Bazzechi 2016, 702–703). The wal s were built of brick and reinforced with towers that include those of a characteristic pen- tagonal form. Jean-Pierre Sodini posits the city enjoyed a triple line of defence, similarly as Caričin grad (Sodini 2007, 318–319). Standing out among the Late Antique constructions in the interior is a building decorated with Proconnesus marble and initial y interpreted as a macel um (Hoti 1996, 176–177), whereas recent research rather suggests a forum (Hoti et al. 2008). The strong Christian character of the city is reflected in different decorative elements of church architecture, as well as in an Early Christian chapel surviving within the amphi- Fig. 2.103: Dyrrachium. City plan (Hoti et al. 2008, Fig. 1). theatre. Local martyrs are depicted in the mosaics dated to the 6th to 8th centuries (Neri et al. 2017). Fig. 2.104: Dyrrachium. Early Byzantine city wal s (2009). 76 2. CITIES Fig. 2.105: Dyrrachium. Remains of the forum (2009). Fig. 2.106: Dyrrachium. Amphitheatre with the Early Christian chapel in the foreground (2009). 77 Slavko CIGLENEČKI The city of Byllis ( Figs. 2.107–2.111) on a natural y well-protected high plateau shows strong traces of Late Antique transformation (Chevalier, Beaudry 2018). The old Hel enistic wal s were extensively renovated and bound with mortar in the early 5th century. Under Justinian, the city was renovated, which is corroborated in as many as four inscriptions (Feissel 1988; id. 2000, 92). These relate that the architect, Victorinos, radical y reduced the city from 30 to 11 ha and constructed new wal s on the most easily defensible ridge on top of the hill; we may in fact only speak of a fortified acropolis (Anamali 1993, 451–455). The up to 2.2 m thick wal s incorporated a mass of reused stones and were rein- forced with six protruding three-storeyed and rectan- gular towers. It seems, however, that at least part of the old city outside the wal s remained inhabited even after this reduction. From the 5th century onwards, there were five churches in the city, two of which remained outside the defended area after the construction of the new wal s in the 6th century (Raynaud et al. 2003). Not much is known of the housing; research was conducted in the episcopal palace at the central basilica that boasts prestigious spaces and many rooms intended for eco- nomic activities. A particular feature are public baths, the construction of which has been dated with the help of an inscription to the middle of the 6th century (Feissel 2000, 92; Chevalier 2015, 232). More extensive investigations have been conducted in Buthroton (Butrint) ( Figs. 2.112–2.115), where sev- eral characteristic features of the Late Antique city can be observed alongside the earlier architecture from the Hellenistic and Roman periods (Bowden 2003, 85–88; Sodini 2007, 318). In Late Antiquity, the natural y well- protected city on a peninsula witnessed the renovation of its Hellenistic wal s that has recently been dated to Fig. 2.107: Byllis. City plan (Chevalier 2015, Fig. 6). Fig. 2.108: Byllis. Episcopal complex (2009). 78 2. CITIES Fig. 2.109: Byllis. Victorinos’ wal s (2009). Fig. 2.110: Byl is. Victorinos’ wal s with a tower (2009.) Fig. 2.111: Byllis. Baths (2009). 79 Slavko CIGLENEČKI around 525 (Greenslade 2019, 197). Two Early Chris- tian churches are known, one on the acropolis and the other, larger one in the lower part of the city. The latter is well-preserved and boasts a large baptistery from the 6th century decorated with figural mosaics. The system- atic investigations of the Triconch Palace have afforded a detailed insight into the development and decline of a Late Antique domus (Bowden, Mitchell 2007). They revealed that a peristil was added to the original Roman house around 400. Between 420 and 440, the domus was added a large three-apsed triclinium that opened onto the peristil. It has been presumed that the renovation works were not concluded due to water flooding the building on several occasions. No major activities have been observed in the second half of the 5th century. Postholes suggest subdivisioning in the first half of the 6th century, while a staircase shows that at least part of the building had an upper storey. Fishermen used the other parts for cleaning shellfish. In the second half of the 6th and in the 7th century, the south wing was used for artisanal activities, the rest for burial. Another group of buildings was researched west of the domus; from the late 3rd to the early 5th century, this group wit- nessed several modifications and extensions that created masonry rooms surrounding a rectangular court. The group was presumably intended for commercial activi- ties with living quarters in the upper storey. Traces of Fig. 2.112: Buthroton. City plan (Bowden, Mitchel 2007, Fig. 1). activity cease here in the second half of the 5th century; renovation ensued towards the end of the 6th century. A two-storey building was erected in the corner of the city wal s. The area of the former palace was abandoned in the mid-7th century. The research in the Vrina Plain has provided im- portant new data on the Late Antique habitation in the city. In the western suburbia, a large domus was built in the second half of the 3rd century (Greenslade 2019). It was temporarily abandoned towards the end of the 4th century, again inhabited in the early 5th and abandoned in the second third of the 5th century. In the early 6th cen- tury, a large apsidal basilica and small baths were built in a part of the domus, while the rest was transformed into small dwellings and workshops. This complex of Fig. 2.113: Buthroton. Triconch Palace from the west (2009). buildings persisted to the end of the 6th century, when it was completely abandoned, most likely following a fire. NORTH MACEDONIA The city of Scupi (Skopje) (Figs. 2.116, 2.117) was long considered largely abandoned after the earthquake in 518. Ivan Mikulčić only mentioned an Early Byzan- tine vil age and the flight of the population to the fort Fig. 2:114: Buthroton. Triconch Palace from the east (2009). 80 2. CITIES Fig. 2.115: Buthroton. Interior of the Early Christian church (2009). on Markova kula (Mikulčić 1982a, 47–48); this was repeated by other authors (e.g. Saradi 2006, 467; Sodini 2007, 331). The extensive systematic investigations in recent years, however, have brought to light numerous remains from the time after the earthquake that offer more reliable evidence of Late Antique continuity (Jo- vanova 2015a; ead. 2015b; Ončevska Todorovska 2017; Jovanova, Ončevska Todorovska 2018). The significance of Scupi is reflected in two visits of the Emperor Theo- dosius I and the presence of an army unit, the pseudo comitatenses Scupenses, mentioned in Notitia Digni- tatum (Jovanova 2015b, 238–242). The remains of the early phase of colonia Scupi are poorly known, with investigations primarily unearthing evidence of disruption in the urban development (layers of burnt debris, ruins) from the late 3rd century. The city gained in significance when it became the metropolis of the newly-founded province Dardania following Diocletian’s division. In the 4th century, it witnessed new constructions (civil basilica, horreum, urban vil a) and renovations (large baths, domus under the southeastern city wal s). The earlier city walls are not well-known; they presumably existed to the late 4th or even the mid-5th century and had a prestigious rather than a defensive Fig. 2.116: Scupi. Plan of late buildings by the cardo (Jovanova character. The city was then enclosed with thick wal s, 2015b, Fig. 18). constructed in part of spolia (th. 3.5 m), that markedly decreased the size of the city; in the east they were erected on top of earlier housing ( domus). The wal s revealed several construction phases. A large horreum also had a semicircular tower with 2.6 m thick wal s. of the 3rd and 4th centuries was built on top of an as yet Research in the central part of the city, on the unidentified earlier building and was in turn covered west side of a cardo (presumably cardo maximus) has towards the end of the 4th century by the city baths 81 Slavko CIGLENEČKI Fig. 2.117: Scupi. Cardo and the adjacent buildings from the north-east (2013). that remained in operation into the 6th century. In the one is marked by a fairly planned positioning of different south, the civil basilica was erected in the 4th century buildings, while buildings in the second phase show a on the ruins of buildings from the 2nd and 3rd centuries. haphazard distribution in proximity and across earlier After destruction towards the end of the 3rd century, a buildings. The wal s of the first-phase buildings are still peristyle domus in the eastern part was renovated and fairly well-built of rubble and brick fragments partly ar- enlarged, remaining inhabited into the second half of the ranged in lines, the wal s are mostly bound by clay and 4th century. A major break has been observed towards only rarely mortar. The wal s of the second phase are less the end of the 4th century, when the theatre, horreum careful y constructed, slightly thinner and associated and the peristyle domus were abandoned. This was soon with postholes. In addition to hearths, several refuse and followed by another expansion (renovation of the baths, storage pits were also unearthed in and next to houses. civil basilica and construction of new city baths). These recent investigations importantly contribute On the east side of the cardo, large baths existed to the earlier knowledge that relied on the writings of from the 2nd to the 5th century. In the 6th century, a large Marcellinus Comes, who reported that an earthquake three-aisled Early Christian basilica with three apses, in 518 completely destroyed the city. It would seem that atrium and court was constructed in the south part of the city recovered and persisted to the end of the 6th and the former baths (Ončevska Todorovska 2018). A large presumably into the initial decades of the 7th century. basilica with a baptistery came to light in the north of Evidence of this primarily comes from the large city the city that functioned in the late 5th and throughout basilica, the late phase of the city baths and numerous the 6th century, constructed in part on a luxury building simple dwellings incorporated into older houses that from the 4th century. have been recorded in all the investigated parts of the After the abandonment of the public buildings city. All new constructions and renovations respected on both sides of the cardo, their ruins were subdivided the existing street grid; the research of the cardo has into simple dwellings and workshops. The destroyed shown numerous repairs and additions spanning from and abandoned parts of the horreum also saw reuse in the late 3rd to the late 6th century. We can see a well- the 6th century; it revealed two phases of interconnected fortified, but impoverished and ruralised provincial primitive dwel ings (Jovanova 2015a, 33–34). Buildings centre, the urban landscape of which was dominated by were still well-constructed, with some wal s erected in at least two large churches and, for some time, also the combination of courses of stones and horizontal y laid city baths, while a multitude of small and poorly built bricks, though already clay-bonded; there is also men- houses and workshops developed in the ruins of other, tion of brick roofs. once grand buildings. Traces of Late Antique buildings and a group of Late Antiquity was an important period for another burials also came to light in several places inside the city of North Macedonia, namely municipium Stobi theatre (Ončevska Todorovska 2017). Based on nu- ( Figs. 2.118–2.122, 2.253). The long years of research merous small finds, the dwellings have been reliably have offered a fairly reliable insight into its structure and dated from the late 5th to the initial decades of the 7th chronology, although we are still missing the publica- century and exhibit two construction phases. The first tions of several important buildings and recent investi- 82 2. CITIES Fig. 2.118: Stobi. City plan (Mikulčić 2002, Fig. 337). gations (Wiseman 1984; Mikulčić 2003; Talevski 2018). The city has an excellent strategic location at the junction of two major roads and at the edge of a fertile plain, on a natural y protected terrace above the confluence of the Rivers Crna reka and Vardar that protect the city on three sides. In the Roman period, the city wal s enclosed a ca 20 ha large area. Its interior revealed several public buildings such as a theatre, baths, synagogue and several private vil as. After a devastation in the late 3rd century, the city was extensively renovated. To counteract flood- ing, the city wal s in the east were transferred deep into the interior, which reduced the city to 14 ha. In the west, city wal s were renovated with protruding rectangular towers and well-fortified double gates in the Valentinian period. Irregularly shaped insulae were arranged along both main streets that followed the configuration of the terrain. Investigations revealed five Early Christian churches intra muros and three extra muros, as well as a Fig. 2.119: Stobi. Plan of Late Antique buildings in the south-synagogue, three baths and seven domus. The episcopal western part of the city (from Talevski 2018, Fig. on p. 432). 83 Slavko CIGLENEČKI Fig. 2.120: Stobi. Remains of Late Antique buildings (2013). Fig. 2.121: Stobi. Palace of Theodosius (2013). basilica was erected along Via Sacra close to the entrance early 4th century and primarily used as luxury homes to into the city, partly on top of the ruins of the theatre; the the end of the 4th, some possibly into the first half of the first church was built there in the second half of the 4th 5th century as suggested by the renovations of mosaic century, followed in the first half of the 5th century by floors in some of them. After being abandoned, they a large three-aisled basilica with an atrium constructed hosted simple dwellings with wal s without mortar that on considerably raised ground. Other churches date to predominated in the second half of the 5th and through- the 5th/6th centuries (Aleksova 1997, 111–152). out the 6th century. Standing out among the vil as in size The investigations paid particular attention to the and decoration is the House of Parthenius, also known architecture of the rich peristyle domus (Mikulčić 1999, as the Palace of Theodosius, which also revealed a col ec- 249−253; Baldini Lippolis 2001, 298–302; Gerasimovska tion of sculptures arranged in niches around a fountain 2012, 108–110). They were constructed in the late 3rd or (Nikoloska 2015). 84 2. CITIES Fig. 2:122: Stobi. Ruins from the east (2013). Of particular importance for understanding the investigations shed light on this very last habitation late habitation in the former domus are the systematic phase (Gerasimovska 2012, 109; Talevski 2018). investigations of Domus Fullonica (Mikulčić 1982b, Modest dwellings also came to light in the area of the 536). This large house reached its greatest extent and theatre, where they were presumably built in the early importance even prior to the mid-4th century, when 5th century (Wiseman 1984, 295, 303). The housing deep apses were added to both triclinia. The abandon- thus exhibits an omnipresent disintegration of earlier ment of the domus can more precisely be dated to the buildings and the appearance of considerably simpler late reign of Theodosius I, when four new families dwellings, on the one side, and the renovation of the created new housing units of coarse clay-bonded wal s episcopal church with lavish decoration, on the other. and an ironworking workshop in the court. This phase, Traces of habitation in the city cease towards the end reliably dated with coins and other artefacts, lasted of the 6th century (Talevski 2018, 430–438). roughly to the mid-5th century. Mikulčič links its end Recent research has revealed Lychnidos (Ohrid) with the Hun raid of 447, as the burnt debris layer held ( Figs.2.123–2.125) as another city of an important status numerous coins from the first half of the 5th century, in Late Antiquity (Mikulčić 1999, 101–106; id. 2002, trilobate arrowheads and the remains of a reflex bow 476–479). Enjoying a well-protected location on a hill (Mikulčić 1982b, 537). After the destruction of the and slopes above Lake Ohrid, it was an important stop on mid-5th century, most of the former domus remained the Via Egnatia. The course of its Late Antique defensive unused, only small huts with clay-bonded wal s were wal s has not been ful y established, though we do know found in two of its corners. Several ground surfaces in of some parts constructed in the 5th century, as well as their interior prove that the huts were repaired. A simi- the sections where the earlier wal s were merely repaired lar situation has been recorded in other urban vil as (Bitrakova-Grozdanova 2009, 28). The size of the city as wel . We may surmise the influx of people from the has been estimated at 40 ha. There is a concentration of countryside who abandoned their old homes to seek Late Antique architecture on the west summit, where shelter inside a better protected city and continued, at the acropolis and the church buildings at the Plaošnik least in part, to practise crop cultivation and animal (Imaret) site were especial y fortified. husbandry (Mikulčić 1982b, 536; id. 1999, 253). Of the Most is known on the Early Christian architecture, domus, only the Houses of Peristerius and Polyharmus with the quality of construction and rich interior fur- saw continuation, albeit in a more rustic form. They nishings speaking of the contemporary importance of may be seen as the living quarters of priests and state the city (overview in Aleksova 1997, 203–222; Mikulčić officials (Gerasimovska 2012, 110; ead. 2015, 262). 2002, 476–479; Malenko 2008). Standing out in its Buildings in the vicinity of the western necropolis, architecture and beauty of mosaics is the basilica with dating from the 4th to the 6th century, shows a shift of a polyconchal plan from the late 5th century, which the population from areas around Stobi close to within was constructed on an earlier cult place (Bitrakova- the city wal s (Gerasimovska 2012, 110); the extensive Grozdanova 1975; Aleksova 1997, 210–216). The in- 85 Slavko CIGLENEČKI Fig. 2.123: Lychnidos. City plan (Mikulčić 2002, Fig. 395). Fig. 2.124: Lychnidos. Slope of the city with the acropolis and theatre in the foreground on the right (2013). tensive excavations of recent years have also unearthed hil . Large Early Christian buildings were also found in a large longitudinal three-aisled basilica with an atrium immediate proximity to the city. and baptistery, which dominates the southern part Less is known of the residential architecture. We of the Plaošnik (Imaret) hil (Bitrakova-Grozdanova only know of the remains of small buildings found inside 2009, 29–36). The remains of at least five contemporary the partial y investigated Roman theatre, similarly as in intramural churches were found during the rescue inves- Heracleia, Stobi and Scupi. In Late Antiquity (dating tigations in the modern-day city, in the saddle between to the Justinian period is particularly mentioned), it is the two elevations and on the slopes of the east Devoj believed to have hosted numerous new buildings as part 86 2. CITIES Fig. 2.125: Lychnidos. Atrium of the large church at the Plaošnik hill (2013). of a settlement that persisted in the medieval and post- medieval periods (Malenko 2008, 478–482). Our knowledge mainly relies on the rich Early Christian architecture on the acropolis, though the distribution of the churches indicates habitations clusters in the wider area of the urban agglomeration where – similarly as in other cities – modest dwellings and economic facilties predominated. Much more varied and better documented is the appearance of the small (ca 9 ha), but rich and well- researched city of Heracleia Lyncestis (Bitola) (Figs. 126–2.130) (Mikulčić 1999, 326–332; id. 2002, 263–268). It lay on the important Via Egnatia in the province Macedonia I and encompassed the acropolis, as wel as the slopes and plain along the Siva voda stream. The course of the up to 2.5 m thick city wal s has been re- constructed and several protruding rectangular towers excavated. The city wal s are believed to have been first renovated in the early 4th century, in the lowland there Fig. 2.126: Heracleia Lyncestis. Plan of the buildings from the are also observable reparations and additions from the 6th century in the theatre (Popović, V. 1982, Fig. 16). Early Byzantine period (protruding semicircular tower). In the late 5th/mid-6th century, a new defensive wall was built in the east side that reduced the urban area to 5 ha constructed extra muros east of the city probably in the and gave the city plan a pentagonal shape. Justinian period, and associated with a necropolis (Ma- A sacral complex was created below the theatre neva 1989; ead. 2008). The different elements of Early in the main lowland part. A large episcopal basilica Christian church furnishings indicate the existence of was built towards the end of the 5th and a smaller one other, as yet undiscovered churches. next to it in the 6th century, both fitted with mosaics of In contrast with the imposing church buildings of very high quality (Aleksova 1997, 235–242). They were the 5th and 6th centuries, there is modest housing that joined in the west by an episcopal palace, where mosaics has been increasingly identified in recent investigations. were gradual y being fitted from the early 6th century to Throughout the 4th century, small houses were being the 560/570s (Cvetković-Tomašević 2008). The church built into a Roman porticus and a marble statue of a architecture in the city further comprises Basilica D, dignitary reused in its construction (Mikulčić 1999, 87 Slavko CIGLENEČKI Fig. 2.127: Heracleia Lyncestis. City plan (Mikulčić 2002, Fig. 157). Fig. 2.128: Heracleia Lyncestis. City from the south-east. Defensive wal s in the foreground, acropolis and theatre in the background (2013). Fig. 2.129: Heracleia Lyncestis. Late Antique and Early Medi- Fig. 2.130: Heracleia Lyncestis. Decorated fountain from the eval buildings on the slope under the former theatre (2013). 6th century (2013). 88 2. CITIES Fig. 2.131: Thessaloniki. City plan (Whitby 2000, Fig. 22). 255; id. 2002, 268). Earlier buildings of clay-bonded GREECE wal s were also found under the floors of the large Early Christian basilica from the late 5th century (Popović, V. Thessaloniki ( Figs. 2.131–2.135) is a city that 1982, 562; Gerasimovska 2010, 123). gained in importance in Late Antiquity; it first became The complex of poorly constructed buildings with- the capital of the province Macedonia and in the early in the former theatre has been more comprehensively 4th century also hosted the imperial residence. Following identified (Janakievski 2001); they were constructions the Hun conquest of Sirmium, it became the seat of the of clay-bonded stone. They represent a small housing praetorian prefecture of Illyricum. Its significance was unit created in the second half of the 6th century on a not diminished even after the seat moved to Justiniana thick layer of levelled debris belonging to the abandoned Prima. The greatness of the city is reflected in the mighty theatre. Smal dwel ings were arranged around a slightly ruins of the city wal s, the Early Christian churches and larger central building. Recent research has shown that the imperial palace, while archaeological excavations such late buildings were made not only in the theatre, but continuously il uminate other settlement details (Curta also in the area between the theatre and the episcopium, 2001b, 136–138; Ćurčić 2010, 19–22). as well as higher up on the slope of the acropolis and in The most clearly visible element of Late Antique the eastern suburbs. The trial trenching on two terraces Thessaloniki is the several kilometres long city wal s next to the northern city wal s revealed buildings of a that enclosed a vast space of 260 ha. The Hellenistic residential and economic nature, their wal s bound by wal s were presumably first renovated in the mid-3rd clay or a very small amount of mortar. All this indicates century using the opus mixtum technique, after which dense settlement in the last period of Antiquity, which they witnessed several repairs and additions. Opinions in part continued into the Middle Ages (Ǵorǵievska, on these additions vary, but most scholars agree they Nasuh 1997–1999; Todorovski 2010; Gerasimovska took place in the early 4th century and there is general 2010, 123–124). Contemporaneously with this modest consensus that complete renovation took place in the housing, a beautiful y decorated fountain was set up next first half of the 5th century (Crow 2001; Dunn 2002, 709; to the theatre. The inscription on this fountain reveals Sodini 2007, 318; Rizos 2011; Baldini, Bazzechi 2016, it was built on the request of Bishop John to honour 35 698–699). The Arch of Galerius, part of an imposing years of reign of the Emperor Justinian in 561 (Đidrova tetrapylon, partial y excavated remains of the imperial et al. 2011; Mihajlovski, Rospendovski 2011, 417). palace south of cardo maximus and impressive rotundas 89 Slavko CIGLENEČKI Fig. 2.132: Thessaloniki. City wal s (2013). Fig. 2.133: Thessaloniki. Remains of tetrapylon, rotunda in the background (2013). 90 2. CITIES Fig. 2.134: Thessaloniki. Imperial palace (2013). Fig. 2.135: Thessaloniki. Agora (2013). to the north afford a glimpse into the wealth of the Late in rescue excavations. In the early 5th century, the agora Antique city. A church with a baptistery and funerary was abandoned and changed its function similarly as in chapels was built into the rotunda that once formed part most other Late Antique cities (archaeologists unearthed of the Palace of Galerius; a similar alteration regards pottery kilns). Shops were found around it, which the octagonal triclinium, which was transformed into persisted with certain repairs into the 6th century. Also a mausoleum with added chapels (Rizos 2012, 333). unearthed were twelve houses with apsidal terminals and Literary sources relate the existence of different peristyles. One of these revealed two phases, one in the public buildings such as baths, theatre, hippodrome late 4th and the other in the early 5th century. In the 6th and praetorium in the 6th century. Housing is poorly century, they witnessed the usual subdivisioning with known, with individual elements only coming to light wal s of inferior quality (Sodini 2007, 327). 91 Slavko CIGLENEČKI The city of Amphipolis ( Fig. 2.136), located on the Via Egnatia, shows an interesting, but rather incomplete Late Antique picture. The old Greek city spread across a wide area and was protected on two sides by a river and also had extensive city wal s. The natural y most protected elevation held the acropolis, where several different and lavishly decorated Early Christian basilicas were built in Late Antiquity. The acropolis was enclosed with defensive wal s in the 6th century; these wal s only protected a small part of the city, primarily the area with church buildings (Sodini 2007, 320). The acropolis was later further reduced with a transverse wal . Intensive surveys in the immediate vicinity of the acropolis have revealed six sites with small concentrations of Late Antique habitation traces outside the acropolis wal s (Dunn 2004, 543). We may surmise that, in times of danger, the people living in these unfortified settlements retreated to the acropolis that provided physical, as well as spiritual shelter. The city of Philippi ( Fig. 2.137–2.139) had a differ- ent development, retaining the large surface of protected area into Late Antiquity. The earlier, presumably Hel- lenistic wal s were renovated in the late 3rd or early 4th century, while a proteichisma was added in the flatland below in the second half of the 5th or the early 6th cen- tury (Provost 2001; Baldini, Bazzechi 2016, 699–700). The 70 ha of fortified surface stands out in comparison with most other cities, which witnessed a reduction of the defended area; this was particularly apparent in neighbouring Amphipolis. Fig. 2.136: Amphipolis. City wal s (2002). The street grid with insulae continued into Late Antiquity and more prominent changes are only visible in the monumental newly-built churches in the centre. Standing out is an octagon (Basilica D) with an episcopal palace and rooms for pilgrims who arrived to the city in conection with the visit of St Paul. The octagon was constructed in 340 in the place dedicated to his memory. Large Basilica A and Basilica B flanked the forum on the north and south, symbolical y embracing it. The complex as a whole gave the impression of a great pilgrimage centre. The churches occupied spaces formerly of a dif- ferent use and partly even encroached on streets (Saradi 2006, 393–395; Sodini 2007, 327–328). The research in some insulae and the theatre has shown that houses were already abandoned in the second half of the 4th century and a variety of workshops was set up in them. One building in the southwestern part of the city was torn down towards the end of the 4th, subdivisioned in the 5th and reused as modest dwellings that persisted to the early 7th century. All these changes point to a transformation of the urban core similar to those observed in the cities of nearby North Macedonia. In Dion ( Fig. 2.140–2.142), the Hel enistic city wal s in the opus mixtum technique were presumably already renovated in the middle of the 3rd century. More substantial changes in the fortifications are noticeable Fig. 2.137: Philippi. City plan (from Poulter 1998, Fig. 1). 92 2. CITIES Fig. 2.138: Philippi. City wal s with a proteichisma (2002). Fig. 2.139: Philippi. Late Antique buildings (2002). towards the end of the 4th century, when the former church built in the mid-5th century (Basilica B). The urban area was reduced from 37 to ca 16 ha with the large basilica just behind the city wal s, of a cemete- construction of new, thinner and less well-built city rial nature, was constructed roughly at the same time wal s. Opinions differ as to the dating of these changes (Mentzos 2002, 334). Among public buildings, Ćurčić (Mentzos 2002, 333; Sodini 2007, 319, 320; Baldini, particularly mentions a large dodecagonal building at Bazzechi 2016, 700–701). The fortified urban area still the junction of the main streets, which he interprets as included important public buildings that reveal a con- a church building and dates to the 5th century (Ćurčić siderably prosperous city in the 4th and 5th centuries. 2010, 134). Towards the end of the 4th century, the first church was The capital of the province Epirus Vetus, Nicopo- constructed in the vicinity of the agora (Basilica A). After lis (Nikopolis) ( Figs. 2.143–2.146, 2.259), was greatly an earthquake, earth was deposited here and another reduced in Late Antiquity when thick wal s were con- 93 Slavko CIGLENEČKI Fig. 2.140: Dion. Late An- tique walls incorporated spolia (2013). Fig. 2.141: Dion. Early Christian basilica (2013). Fig. 2.142: Dion. Remains of a wealthy domus (2013). 94 2. CITIES Fig. 2.143: Nicopolis. City plan (Bowden 2007, Fig. 11.1). structed that placed the core of the Late Antique city sculpture. A similar building may have stood near in the northeast corner of the earlier Augustan wal s Basilica B. Another domus with a large apsidal hall lies (Bowden 2003, 89; Sodini 2007, 320; Baldini, Bazzechi near the ‘nymphaeum’ (Bowden 2003, 50–51). A series 2016, 701–702). The new wal s were constructed in the of smal , mostly single-room dwellings was found on the opus mixtum technique with a mass reuse of stones. interior side of the southern wal s that show an inferior Scholars agree that these wal s were constructed towards construction and measure roughly 5 × 5 m. Their clay- the end of the 5th century. The densely spaced towers bonded stone wal s commonly incorporated spolia. They are still well-preserved today, they are rectangular with were presumably in use from the construction of the round ones in the corners, the west entrance is flanked city wal s towards the end of the 5th to the 8th century with horseshoe-shaped towers. Several important and (Bowden 2003, 166). In the letter of Saint Gregory the beautiful y decorated churches were erected inside the Great, Nicopolis is described as the third largest city of fortified area. These largely respected the street grid Illyricum, which suggests that the well-fortified core of the earlier city. Mentioned among the significant only represent part of the Late Antique city. This is cor- residential buildings is the ‘episcopium’, original y a roborated by two vil as from the 5th/6th century found domus that was enlarged and modified in Late Antiquity outside the city wal s. The fortified Late Antique part and furnished with high-quality mosaics and marble thus far revealed no groups of burials. 95 Slavko CIGLENEČKI Fig. 2.144: Nicopolis. Epis- copal church from the south (2002). Fig. 2.145: Nicopolis. City wal s with towers (2002). Fig. 2.146: Nicopolis. Well- preserved city wall from the south (2002). 96 2. CITIES Fig. 2.147: Thebes. City plan (from Karagiorgou 2001a, Fig. 2). Thebes (Nea Anchialos) ( Figs. 2.147–2.149) was palace presumably constructed in the late 4th century. an important port in Late Antiquity and one that has The building was lavishly decorated with mosaics and been archaeologically well-researched (Karagiorgou wall paintings, and had an audience hall in the centre. 2001a, 184–197; ead. 2013, 156–167). The city wal s, Another richly decorated house stood near Basilica A, believed to have already been constructed by the mid-3rd which is believed to have belonged to a local nobleman century and renovated under Justinian, were reinforced (Sodini 1984, 367). Other buildings are mentioned such with rectangular towers and enclosed a 25 ha large area. as several baths, a hospice and a school. A large complex Traces of numerous late buildings also came to light with a court surrounded by shops has been investigated outside the wal s. between the two basilicas, as well as commercial build- The significance of this episcopal city is clearly mir- ings along the main street. Domestic buildings plots rored in the large and beautiful y decorated churches – consisted of enclosuers which surrounded the house and four inside the wal s and as many as five outside (Kara- the open space around it; some had stone-built wal s on giorgou 2001a, 187–194). Traces of the earliest church the ground floor and wal s of brick in the first storey, from the 4th century lie under the cathedral (Basilica C). their rooms were decorated with wall paintings and fitted The renovation of this and the construction of other with window panes (Karagiorgou 2001a, 196–197). Late churches mainly occurred in the 5th and 6th centuries. Antique cemeteries were located outside the fortified In the vicinity of Basilica C, living quarters were also area and only a few small groups of burial were found unearthed and tentatively interpreted as an episcopal in the interior (Karagiorgou 2001a, 194–196). The 97 Slavko CIGLENEČKI Fig. 2.148: Thebes. Episcopal church (2002). Fig. 2.149: Thebes. Church by the wal s (2002). research thus far shows a medium-sized fortified Late Tanagra revealed Late Antique activities inside Antique city with rich Early Christian architecture and the city and an intensive use of the surrounding coun- vast unfortified suburbs. tryside (Bintliff 2007, 664–665). The old city wal s were In Boeotia, systematic field surveys and geophysi- renovated in this period. The city plan shows substantial cal measurements have been conducted in two cities changes to the Classical grid of streets and insulae, with (Bintliff 2012, 361–362). Thespiae revealed city wal s baths and churches constructed on top of earlier public constructed of spolia in the 4th or early 5th century, buildings and insulae. At a distance of two kilometres enclosing a 12 ha large area inside a much larger city from the city, a large settlement (Aghios Constantinos) from the Classical era. To the east outside the wal s was was fortified that was to serve as a satellite settlement and a fairly large settlement with several small churches. may even have replaced it due to its superior defences Extramural settlements with associated cemeteries also and shorter defensive wal s. came to light on several other sides of the city. 98 2. CITIES Athens ( Figs. 2.150–2.154, 2.258) witnessed sweep- which was a complex of buildings with peristyle courts ing changes in Late Antiquity that included a marked and defensive wal s; Eudocia, wife of Theodosius II has reduction of the defended area. The post-Herulian wal s been identified as its owner (Ćurčić 2010, 119–123). The only enclosed the acropolis and part of the slope with the building was already abandoned around 530 and turned Roman forum and Hadrian’s Library, though numerous into a monastery. Other private vil as were erected in its Late Antique buildings were constructed both inside and vicinity and below Aeropagus that have as yet not been outside the wal s, most numerously on the Classical agora comprehensively investigated (overview in Sodini 1984, (Saradi 2006, 238–239; Baldini, Bazzechi 2016, 707–709, 344–354; id. 1997, 463–465). Several domus date to the Tsoniotis 2016). The new Early Christian churches were 5th and 6th centuries and exhibit a high standard of living. most frequently built on the spot of earlier pre-Christian The House of Sculptures, for example, had new baths temples from the second half of the 5th century onwards. added in the first half of the 6th century. Alongside these The late private buildings are also well-known. On the high-quality private buildings, the agora hosted shops and agora, parts of the city infrastructure and private build- workshops. A rapid ruralisation occurred as early as the ings began to be renovated already in the second half mid-6th century, which is corroborated by water mil s and of the 4th century. A large private residence or Palace olive presses, followed towards the end of the 6th century of the Giants was probably constructed around 421, by a fast decline of the urban area. Fig. 2.150: Athens. Agora from the west (2007). Fig. 2.151: Athens. Remains of the post-Herulian wal s (2011). 99 Slavko CIGLENEČKI Fig. 2.152: Athens. Plan of the city inside the post-Herulian wal s (Tsoniotis 2016, Fig. 1). Corinth ( Fig. 2.155), a city advantageously sited on fortified Acrocorinth, on the one hand, and a gradual a well-protected peninsula, developed a unique appear- decline of the city and its fragmentation, on the other ance in Late Antiquity. This is clear from the research (Robertson Brown 2018). Numerous investigations that shows the city existed to the end of the 6th century, notwithstanding, many questions regarding the Late after which there was a partial shift to the natural y Antique topography of the city remain open. Even the 100 2. CITIES Fig. 2.153: Athens. Plan of the agora (Sodini 1997, Fig. 45). Fig. 2: 154: Athens. Facade of the Palace of the Giants (2011). Fig. 2.155: Corinth. Plan of Late Antique buildings in the area of basilica of Hagios Leonidas (Sodini 1984, Fig. 24). course of the Late Antique city wal s is problematic, one fitted with a bathroom in the 6th century, as well as presumably only built around 400 and enclosing a third a long building from the mid-6th century (Sodini 1984, of the former city (Ćurčić 2010, 126; Baldini, Bazzechi 394; Warner Slane, Sanders 2005). 2016, 701). Many important public institutions, as well as Habitation cores appear at several locations within the agora, remained outside the wal s. Evidence shows a the large former city. The most important one was that similar picture as in most other cities with continuity in in the Lechaion port, where Hagios Leonidas, one of that the 5th century brought an abandonment of public the largest Early Christian basilicas was built. Alongside buildings, a gradual reuse of abandoned buildings for other known or surmised basilicas, Hagios Leonidas small and modest dwellings and workshops, as well as underscores the Early Christian significance of the city smal groups of burials appearing within the former city. that hosted a large Christian community already in Certain large private houses do persist, however, such as Apostolic Age. Simple and irregularly arranged houses 101 Slavko CIGLENEČKI Fig. 2.156: Gortyn. Praetorium (2020). Fig. 2.157: Gortyn. Odeon below the acropolis (2020). 102 2. CITIES began to be erected in the area of this basilica soon after TURKEY the mid-6th century (Sodini 1984, 370–373; Warner Slane, Sanders 2005). The construction technique and At the eastern end of the area under discussion, in small finds reveal it as a completely rural settlement. In the province Europa in modern-day Turkey, most traces contrast with most other cities, Corinth witnessed strong of settlement continuity come from two coastal cities. continuation in the 7th century. One is the old Greek colony of Perinthus, in the south- Gortyn (Gortyna) ( Figs. 2.156, 2.157), a well- ern part of Europa at the Sea of Marmara, that in the researched large city on Crete shows all the Late An- 3rd century changed its name to Heracleia. The city was tique elements characteristic of a major island city and located close to the junction of two most important west- provides an opportunity of comparison with cities on wards routes (Via Egnatia and a military road through the continent (Di Vita 2010). The acropolis is poorly the Danube Basin) and enjoyed a natural y excellently known, but we known that an Early Christian church protected location on a peninsula; in Antiquity, it was was constructed there. The defensive wal s around the also separated from the mainland by an isthmus. It had acropolis have not been reliably dated and may even strong double wal s (Crow 2002, 342–343; Dintchev be slightly later. What is certain is that the whole city 2021b, 245–246). The outer wal s were reinforced with was not enclosed with wal s in Late Antiquity. Several U-shaped towers and located in the lowland part of the churches stood in the city, predominantly built between city. The stamped bricks show it is attributable to the first the 4th and the 6th century; the buildings of the Justinian half of the 5th century. The inner wal s, which protected renovation stands out in size and decoration. a vast acropolis, were built of brick. Surviving of these The Praetorium is the most complex and best- wal s are two pentagonal towers that suggest the wal s investigated building that represents important ad- were constructed under Justinian. A large Early Chris- ministrative architecture from the late 4th century tian basilica stood at the outer wal s in the lower part, (Sodini 1984, 343–344). After the earthquake of 365 which indicates this part of the city was still inhabited in had destroyed the Hellenistic architecture later used as a the 6th century (Crow 2002, 343). The later inner wal s large Roman bath complex, the building was extensively encircling the acropolis certainly show a compromised renovated under governor Dositheus. Roman baths security of the 6th century in the immediate vicinity of were still in operation, while a basilica was constructed the Byzantine capital. Other elements of the city are not in another part that functioned as a court of justice. A known; Procopius mentions a palace and an aqueduct. large part of earlier rooms was abandoned. Housing was The city of Salemydessos-Medeia (Midye or Ku- later arranged in the wider area of the complex, the baths jikoy), on the coast of the Black Sea, also took advantage were abandoned by the 6th century. The basilica was of its convenient location on a raised plateau above a renovated after the earthquake of 618 and continued to well-protected sea port. Not much is known of other Late have an administrative function for some time. Antique remains, while the city wal s were thick and partly Italian researchers focused their attention on the still survive on the accessible west side (Crow 2002, 345). dwellings in the Quarter of Pythion (Baldini Lippolis Literature mentions rectangular towers and the remains of 2001, 208; Zanini 2009). After a devastating earthquake a large pentagonal tower at the entrance. The wal s were (presumably in 365), an area of Late Antique housing not investigated to the degree that enables a more precise developed there in the second half of the 5th century dating of the different Late Antique phases. that represented the main residential quarter. Another To conclude the discussion of the cities of this major renovation was detected in the 6th century and group, we should mention Constantinople (İstambul), the complex was only abandoned in the 8th century. a metropolis of the ancient world and the capital of the Researchers observed all the features of late urban Byzantine Empire. Its monuments and its continuity from construction including buildings of inferior quality a Greek to Roman city and further on to the Byzantine built into earlier architecture, subdivisions and the ap- capital on the border between two continents certainly pearance of different workshops. In other areas as wel , reach beyond the scope of this discussion (overview in there is visible degradation of earlier public buildings Ćurčić 2010, 54–58, 77–100, 187–203). Of the character- that now hosted simple dwellings. Such dwellings and istic Late Antique elements that created the magnificent agricultural tools indicate an influx of people from the capital, we should mention first and foremost the land countryside who took advantage of the urban infra- wall constructed in the early 5th century, the design and structure and existing buildings. efficiency of which became the role model for many other cities across the Empire. The main wall had numerous towers and forts, as well as a proteichisma that appears in cities and large forts as a characteristic element of defence already during the second half of the 5th century and spreads far westwards in the time of Justinian. 103 Slavko CIGLENEČKI Fig. 2.158: Sabiona. Hill from the south (1989). 2.5 CITIES NEWLY-FOUNDED 2.5.1 WESTERN PART IN LATE ANTIQUITY ITALY The section below discusses the urban formations established anew in Late Antiquity that constitute a Literary sources for the 6th century relate that sev- varied group and differ in many aspects from the earlier eral episcopal sees were transferred to small and largely Roman cities. Some were founded completely ex novo, naturally protected locations, the features of which for example Caričin grad, others developed in the places hardly speak of a city. At the northwestern edge of the of earlier settlements of a different, non-urban character. area under discussion, in the province Raetia II, Sabiona Their diversity and the different local factors influenc- (Chiusa/Klausen) ( Fig. 2.158) is one example of a mod- ing their creation are a frequent cause of unease in est newly-formed episcopal see. The research conducted classifying a city into this group. It is a group that most thus far does not allow reliable interpretations of the poignantly raises the methodological question of what settlement and its defensive elements, archaeologists constitutes a Late Antique city; this issue has in part have only been able to confirm a prominent ecclesiastical already been tackled in the introduction and is discussed centre and scarce traces of a small settlement. North of in greater deal at the end of this chapter. the modern-day city, a dominant rocky hill rises high In addition to settlements with several urban char- above the valley and the confluence of the Rivers Isarco acteristics, this group includes sites, for which archaeo- and Tinne. It is natural y excel ently protected with pre- logical evidence, size and sometimes literary evidence cipitous slopes on three sides and can only be accessed indicate important agglomerations or merely episcopal from the south. Because of its excellent defensive loca- centres. They represented religious and administrative tion, the hill was inhabited in several earlier periods, in centres of individual territories associated with several the Roman period the Via Claudia led past it. The long small contemporary settlements. systematic excavations and trial trenching campaigns The newly-founded cities occur in greater numbers have established that the hill was only partly occupied in in limited periods, primarily from the late 3rd to the Late Antiquity (Berg 1985, 89–97; Bierbrauer, Notdurfter first third of the 4th, as well as in the 6th century, as is 2015; Kuhnen 2020). Modest remains of buildings from presented in more detail in the interpretative part and the 4th to the first half of the 6th century have come to the diachronic overview of the emergence and abandon- light near the present-day Marian church, while a large ment of cities. First, however, the cities are presented in Early Christian church was constructed around 400 a geographical order. slightly lower down the slope, which is associated with a vast cemetery spanning from the 5th to the 8th century. There is no evidence of a settlement existing throughout the 6th century and beyond, but we do have habitation 104 2. CITIES Fig. 2.159: Castrum Gradense. City plan (Brogiolo 2011a, 51). Fig. 2.160: Castrum Gradense. Church of Santa Eufemia (1991). remains around the church on the slope, as well as the 5th century, which were soon joined by a third one. The remains of a building probably used for artisanal activi- defensive wal s, which protected a 3 ha large interior, ties. Also not entirely clear is the dating of the second and new church buildings and additions are believed to presumed Early Christian church on the top of the hill have taken place during the 6th century, most likely after (Glaser 1997, 152–154). Literary sources mention the the Gothic Wars, when Grado also became an important episcopal see here from the mid-6th century onwards. port (Brogiolo 2011a, 128–129). The most characteristic example and one also well-documented in literary sources is the transfer of the episcopal see from endangered Aquileia to Castrum AUSTRIA Gradense (Grado), a small fort in the Grado Lagoon ( Figs. 2.159, 2.160) the name of which is indicative of the The size, the existence of an ecclesiastical centre and significant change in the perception of the newly-formed the mention of an episcopal see in the area of Aguntum settlement (Cuscito 2001; Bierbrauer 2004). Recently, its suggest that the fortified settlement on Kirchbichl near foundation has been set to the mid-6th century (Brogiolo, Lavant ( Figs. 2.161, 2.162) can also be identified as a Cagnana 2005, 103–106; Marano 2012, 583). The natu-newly-founded settlement. It lies on the south side of ral y well-protected island was inhabited even before. the River Drau/Drava, roughly 4 km from the Roman Two churches were built there in the late 4th or early city of Aguntum. The hill has an excellent defensive loca- 105 Slavko CIGLENEČKI Fig. 2.161: Kirchbichl near Lavant. Plan of the settlement (Grabherr, Kainrath 2011, Fig. 3). Fig. 2.162: Kirchbichl near Lavant. Double church from the south-east (2003). tion with high precipitous slopes in the south and very which the first explorations presumed to have protected steep slopes on all other sides. The settlement covers a the settlement in the north side of easiest access, have surface of 350 × 150 m. The first systematic excavations not been confirmed in the recent investigations; it is pos- were already conducted in 1948–1956, while occasional sible that the steep slopes provided sufficient protection. rescue investigations have been taking place in recent The towers at the presumed entrance may also not be times (Grabherr, Kainrath 2011). The defensive wal s, chronological y linked to this settlement. Found beneath 106 2. CITIES the east tower was part of a building with a hypocaust horrea, outbuildings, stables and baths. Most authors and several phases observable above the hypocaust. The agree that their construction was a military project with steep slope to the east of here revealed a long building economic considerations, as they also served as supply with several rooms that partly closed the access to the centres for the army or its mobile units ( comitatenses). interior of the settlement. Having said that, the plans and small finds of the build- An Early Christian church with a baptistery and ings in the interior speak of a civilian population. After extensions, or ‘episcopal church’ as the first excavator the province Valeria was relinquished to the Huns in the Miltner described it, stood in the middle of the slope 430s, all except Keszthely-Fenékpuszta were abandoned. on a long terrace (Grabherr, Kainrath 2011, 23). In its A good and well-investigated example is Keszthely- vicinity on the north side, a complex of several rooms Fenékpuszta ( Figs. 2.163–2.165) (Müller 1988; Christie came to light, including those fitted with heating, which 2007, 558–560; Heinrich-Tamáska, Müller 2009; Hein- appears to be the most prominent living quarters within rich-Tamáska 2011b; ead. 2015, 48–58; Visy 2018). This the settlement. Another church was found at the south large Late Antique fortification with ca 15 ha of interior edge of the settlement and also associated with a resi- surface lay on a peninsula of Lake Balaton that was well- dential complex, which is not known in detail. The entire protected with the lake and marshes. A 2.4–2.6 m thick slope has revealed densely spaced traces of only partial y defensive wall was constructed here in the second third known buildings. The numerous coins and other small of the 4th century that enclosed a 377 × 358 m large area. finds (coarseware and imported pottery, glass, differ- The wall had four fortified entrances reinforced with ent metal and bone artefacts) indicate the settlement interior propugnacula, as well as 44 round protruding existed from the mid-3rd to the late 6th century, though towers. In the north, a defensive ditch was found and a detailed chronology of the buildings and cultural lay- another rampart with a ditch was set up 760 m in front ers is hindered by their position on the slope and hence of the fortification spanning the whole width of the pen- marked erosion, but also by the incomplete data of the insula and thus additional y defending the access to the early excavations. Venantius Fortunatus, who visited settlement. Two roads led through the fortification and a Aguntum in 565, mentioned it was located on a high tetrapylon was erected above their junction. The interior hill – this certainly corresponds with the remains of the revealed 29 buildings, which were excavated at least in vast settlement on Kirchbichl. part; geophysical surveys indicate the existence of many more. They comprise different forms, spanning from very simple, single-room buildings to vil a-like complexes HUNGARY with a peristyle or a corridor, as well as a horreum, baths and an Early Christian church. The fortification revealed Large fortifications of Late Antiquity represent a several different phases and the dating of the buildings settlement novum in the vast Pannonian expanses of and their renovations has not yet been ful y explained. western Hungary, for which it is difficult to distinguish between semi-urban centres, fortified settlements or army posts (overview of the functional and chrono- logical identifications in Heinrich-Tamáska 2011b). The size, heavy defences, as well as numerous and varied buildings in the interior show that some of them can be seen as newly-founded cities (cf. Rizos 2017b, 26–27). Such examples are the fortifications at Fenékpuszta, Környe, Ságvár and Alsóhetény, which are similar to each other in their lowland location close to water and even more so in the architecture, both the defensive in- stal ations and the buildings in the interior (Tóth 2000; Heinrich-Tamáska 2015). The group further includes Gorsium – Tác, where strong wal s and interior build- ings completely altered its character. These settlements are most frequently marked as ‘inner fortresses’, which emphasises their location in the hinterland of the Dan- ube limes. Most are irregularly rectangular in plan and cover a surface of 7 to 21 ha. The form of the defensive towers reveals at least two construction phases: the first one in the second third of the 4th century and the second one in the Valentinian period. Their interiors revealed Fig. 2.163: Keszthely-Fenékpuszta. Plan of the fortification main buildings that resemble vil as, in several cases also (Heinrich-Tamáska 2017, Fig. 2c−d). 107 Slavko CIGLENEČKI Fig. 2.164: Keszthely-Fenékpuszta. Remains of the entrance gates (2003). Fig. 2.165: Keszthely-Fenékpuszta. Horreum (2003). There are persisting differences of opinion regarding the thus appears as a complex of military architecture with a function of individual buildings; the identification of the strong civilian presence. The necropoleis unearthed thus presumed principium and praetorium is still disputed. Re-far largely correspond with the chronology of the settle- cent research indicates that most buildings had masonry ment, which persisted to the middle of the 5th century foundations, while the superstructure used a combination when it was torn down. It was renovated most likely in of bricks and wooden beams, which researchers have asso- the Ostrogothic times, the extant houses modified and ciated with the numerous traces of fires. The fortification numerous sunken and other modest buildings added. A 108 2. CITIES continuity of the settlement throughout the 6th and partly into the early 7th century can be surmised from the burials, small finds and an Early Christian church. Florin Curta presumes that a bishopric existed here in the 6th century (Curta 2001b, 58). Less is known of the fortification at Alsóhetény (Fig. 2.166) (Mulvin 2002, 73–74; Tóth 2009, 34, 36). It is comparable in size with Fenékpuszta and its interior also indicates similar buildings. The smaller fortification at Ságvár has three large horrea, which underscore the role of the site as an army supply base (Mulvin 2002, 101–102). Also sufficiently revealing is the small ‘inner fortress’ at Környe (Szabó, Heinrich-Tamáska 2011). The fortified area extended across roughly 7 ha and its architecture shows a fortification of the Valentinian phase characterised by round towers. The fortification was constructed on top of an up to 1 m thick levelled layer that contains the remains of the civil settlement from the Principate, which the coin finds suggest was destroyed in the time of Gallienus. Parallel with Fig. 2.166: Alsóhetény. Plan of the fortification (Rizos 2017b, Keszthely-Fenékpuszta, its construction has been dated Fig. 6). to the second third of the 4th century. After a devastation in the 260s, the city of Gorsium witnessed renewed prosperity in the Tetrarchic times, towers and fan-shaped ones in the corners. The interior when it was renovated and renamed Herculia ( Figs. thus far mainly revealed public buildings including a large 2.167–2.169) (Fitz 1976; id. 2004); we can even speak palace in the northwest, which presumably served as the of the establishment of a new city. Its central part was residence of the governor of Valeria. Two buildings with enclosed with thick wal s reinforced with rectangular apses in the north have been interpreted as churches, Fig. 2.167: Gorsium (Herculia). Remains of the Late Antique palace (2011). 109 Slavko CIGLENEČKI Fig. 2.168: Gorsium (Herculia). City plan (from Fitz 2004, Fig. 7). Fig. 2.169: Gorsium (Her- culia). Remains of the fan- shaped tower in the south- western corner (2011). 110 2. CITIES Fig. 2.170: Carnium. Plateau of the large Late Antique settlement from the west (2003). Fig. 2.171: Carnium. Partly excavated city wal s with a tower (2003). though their sacral function cannot be confirmed (Hein- SLOVENIA rich-Tamáska 2012, 229). The buildings and small finds from the 4th century show a time of prosperity for the city Continental Slovenia revealed many fortified set- that only faded after 378, when clay-bonded houses of tlements from Late Antiquity, of which the significance an inferior quality began to be built. Jeno Fitz presumes and size only allow Carnium (Kranj) ( Figs. 2.170, 2.171) these were the dwellings of the suburban population to be included in the category of newly-formed cities. Its who sought shelter in the protected interior and used al exceptional strategic location at the junction of several the free spaces to create modest dwellings by using the roads and pathways, on a natural y excellently protected debris material (Fitz 2004, 42). The palace was turned high terrace enabled the formation of the largest forti- into artisanal facilities in the first half of the 5th century fied Late Antique settlement in Slovenia (ca 10 ha). Its (Müller 2000, 244). The last phase of the city’s existence existence has long been surmised on the basis of the rich is corroborated by the adjacent cemetery, where the last finds from the V Lajhu and Križišče Iskra cemeteries reliably dated burials belong to the 430s. After the mid-5th (Stare 1980; Vinski 1980; Sagadin 1988; Knific 2005). In century there are no more observable traces of intense the last few decades, investigations have also offered an habitation in the city interior, only individual barbarian insight into the defences and habitations of the settlement graves (Heinrich-Tamáska 2012, 229). (Sagadin 1991; id. 1998, 715–717; id. 2004; Ciglenečki 2005, 265–267; Sagadin 2008, 176–178; Urankar 2021). 111 Slavko CIGLENEČKI Fig. 2.172: Capris. Plan of the Late Antique remains in Kapucinski vrt (from Cunja 1996, App. 3). The roughly 250 × 90 m large plateau at the confluence of the Rivers Sava and Kokra provided an effectively de- fensible spot additional y fortified with wal s and towers. High-quality masonry buildings were constructed next to the interior face of the wal s, while wooden buildings predominate in most other parts of the city (Sagadin 2020, 208–210). The recovered artefacts include those that indicate different workshops, presumably includ- ing glassworks. The remains of a large Early Christian church with an octagonal baptistery were found under the present-day parish church. The central place of Carnium in a wider region is indirectly also perceptible through the smaller Late Antique settlements in the surrounding area (Puštal above Trnje, Hom above Sora, Sv. Lovrenc and Fig. 2.173: Capris, south view (2020). Gradišče above Bašelj, Sv. Jakob above Potoče, Štefanja gora), which enclose in an arch the plain in the middle of which Kranj is nestled ( Fig. 2.263). Indirect confirmation 6th century (Šašel 1974, 452–454; Bratož 2001). Jaroslav of the archaeological evidence can be found in the writ- Šašel presumed that the city in the 5th/6th centuries wel- ings of the Anonymous Geographer of Ravenna, whose comed a large number of refugees from Pannonia and Cosmography mentions the land of Carneola and in it was centred around the fortification called Justinopolis castel Carnium that is mentioned first. located on the highest point of the island. The rescue In northwestern Istria, the Late Antique agglomera- investigations of 1986–1987 systematical y unearthed tion Capris / insula Capritana (Koper) was established a large part of this city with architectural remains and on an island close to the coast ( Figs. 2.172, 2.173), for numerous small finds (Cunja 1996). The architectural which sources mention the presence of a bishop in the remains comprise one large building and the wal s of 112 2. CITIES several others that continued beyond the excavation area. Results show that the island was first more intensely inhabited in the Late Roman period, practical y ex novo, with the small finds also indicating the possible exist- ence of an army post (Cunja 1996, 130). An even more intensive settlement has been established for the 5th and 6th centuries that continued into the Middle Ages (Cunja 1996, 130). Earlier masonry structures have been found under the current cathedral of Koper that can as yet not be positively identified as an Early Christian basilica (Župančič et al. 2007; Hofman, Župančič 2008). CROATIA In the Croatian part of Istria, the establishment of three new bishoprics (Pedena, Neapolis and Cissa) was Fig. 2.174: Siparis. Plan of the eastern part of the city (Milošević noted for the 6th century, though they have not been reli-Zakić 2019, Fig. 4). ably identified (Bratož 1992, 301–302). The Anonymous Geographer of Ravenna also mentions civitas Siparis ( Figs. 2.174–2.176). Its identification is not disputed, located on the long and narrow peninsula near Umag that becomes an island during tide; the small finds show its existence over a long period of Antiquity to the 9th century (Marušić 1975, 338–340; Milošević Zakić 2019). Excavations in the exposed part where access is easiest have revealed a heavily fortified entrance with a pen- tagonal tower. A cistern was found in its interior, while the tower was associated with thick defensive wal s on both sides. The peninsula also revealed different build- ings arranged on both sides of the narrow central street. The small finds show that habitation was most intense in the second half of the 6th and first half of the 7th century, with the dating to the 6th century corroborated by the Fig. 2.175: Siparis. Peninsula with Late Antique remains (2010). Fig. 2.176: Siparis. Remains of the tower on the approachable part of peninsula (2010). 113 Slavko CIGLENEČKI Fig. 2.177: Bosar. Peninsula Sokol with Vela Luka and Mala Luka (2018). Fig. 2.178: Bosar. Double church with a narthex, pre- served at a considerable height, the fortification of Korintija in the background (2018). Fig. 2.179: Bosar. The cen- tral part of the ruins lies at the bottom of the Mala Luka, a narrower strip is vis- ible along the eastern coast (2018). 114 2. CITIES construction of the pentagonal tower (Milošević Zakić above a part of a large rectangular defensive wal , the 2019, 218–219). function of which has as yet not been ful y explained. Large, unfortified, but well-constructed settle- Leaning on the church complex from the south are the ments are very rare. An exceptional example and one residential remains that reach to the coast. These may be that deserves particular attention is Bosar near Baška in direct connections with the church (monastery?). To ( Figs. 2.177–2.179, 2.264) on the island of Krk (Faber the north of here, the remains of less well-constructed 1988, 123–126; Tomičić 1988, 148–151; id. 1990; Šiljeg drystone buildings are visible in plan along the coast, 2008). It has been poorly investigated, but the excellent which Tomičić presumed to be the dwellings of the preservation and remains visible on the surface allow autochthonous population (Tomičić 1990, 35). a reconstruction of the last large Early Byzantine set- The small finds mentioned in earlier reports, con- tlement in the eastern Adriatic. It is a veritable time sisting of gold coins of which only one of Justinian has capsule, as it was not inhabited neither before nor after. been identified, as well as numerous shards of amphorae Together with the fort of Korintija, which rises above, and glass goblets, date this settlement to the 6th century. it offers a comprehensive picture of an Early Byzantine The field surveys also yielded traces of a glassworks post with several functions that survived exceptional y (Šiljeg 2008, 85–86). The existence of a vast complex well due to its remoteness and mild climate. Most of settlement with at least four churches was closely tied the buildings are still buried under heaps of rubble and with the maritime route, which is clear from the am- investigations would certainly offer an excellent insight phora shards and the difficult communication lines into the life and function of an Early Byzantine settle- with the rest of the island as a result of the undulating ment far in the west. It lies in the southern part of the rocky terrain in the hinterland. The small patch of fertile island of Krk, where two harbours (Mala Luka and Vela plain in immediate proximity could provide modest Luka) are nestled at the root of the Sokol peninsula. In subsistence, but it would appear that trading, as well as Mala Luka, in an area called Bosar, lie the remains of a port sheltered from winds and hostile views was more tens of masonry buildings in different conditions of prominent in that sense. The fortification Korintija on preservation and visibility that already attracted atten- the hill above the settlement, which is an excellent addi- tion in the 18th century (Fortis 1774, 283–284). The core tion to the unprotected residential complex, is presented of this settlement was located at the bottom of the bay below among the fortified newly-founded settlements that measures ca 150 × 100 m. Three church buildings (see Chapter 3.3). stand out. The largest is a basilica measuring 30 × 12 m The Palace of Diocletian in Split ( Figs. 2.180–2.184, in plan and oriented towards the east that survives in 2.250) is an impressive example of a fortified imperial tal ruins in the southwest of the settlement area and has vil a later transformed into a city, and it is certainly one already been identified during field surveys (Ciglenečki of the most significant monuments from Late Antiquity. 1987a, 104, Fig. 153). Two smaller sacral buildings are The palace needs no description here, we only offer a visible in the east and north sides of the main part. Trial selection of the vast literature that presents, interprets and trenching has been conducted in the east church with attempts to reconstruct it (Adam 1764; Niemann 1910; a triconchal terminal (Šiljeg 2008, 84–85). Most of the Hébrard, Zeiller 1912; Wilkes 1993; Marasović 1994; Mc- residential buildings in this area are hidden beneath Kay 1998; Belamarić 2003; Ćurčić 2010, 33–38; Piplović piles of debris, only in some places can we observe the 2016). The fortified complex from the early 4th century remains of mortar-bound wal s. In spite of a lack of in- combines the features of an army camp and a luxury vestigations, the orientation and outlines of individual imperial residence, but also indicates the difficulties of buildings indicate a settlement with differently designed the following centuries when this part of the Empire was houses. This main part of the settlement is associated turned into a fortified landscape. Irregularly rectangular with a roughly 20-50 m wide strip of buildings, more in plan, it is oriented with the south side towards the sea than 200 m long, with well-visible foundations along and measures 216 × 175 and 180 m. With the surface of the north part of the bay. Roughly in the centre of this 3 ha, it was the size of a small city, the layout of which strip is a large walled area, named ’Klošter’ in the earlier reflects all the qualities of Roman urbanism (Suić 2003, literature. Aleksandra Faber described it as a courtyard 358–360; Ćurčić 2010, 26–29). It has a rectilinear grid of complex surrounded by buildings, while Željko Tomičić streets with the residential part clearly separate from the sees an fort at the entrance to the settlement (Faber 1988, private part reserved for the emperor and his retinue. The 124; Tomičić 1990, 35). Bartul Šiljeg saw the dressed monumental, 2.10 m thick wal s were reinforced with good-quality stones and a wall with pilasters on the sixteen rectangular and square protruding towers placed better-preserved south side as evidence of a church at the corners and along the sides. Three main entrances (Šiljeg 2008, 86). A more detailed examination of aerial were flanked with octagonal towers additional y fortified photographs and of the architecture on the ground re- with propugnacula in the interior. vealed it was a double church with a narthex, of which The two main communications crossing the palace the facade with pilasters survive. The church was built perpendicularly to one another are reminiscent of the di- 115 Slavko CIGLENEČKI vision of army camps and the orthogonal urban design. It is presumed that the northern part housed the army and servants. A peristyle forms the centre and leads to the temples and the imperial residence. The recent inves- tigations in the underground level of the palace revealed the layout of the emperor’s private quarters. The different architectural elements and surviving decoration shows a number of elements of Eastern architecture. Joško Belamarić hypothesises that the gynaeceum mentioned in Notitia Dignitatum (Occ. XI) represents a component of the palace’s original design (Belamarić 2003); most other authors set its creation to a time after the death of Diocletian. At an undefined time during Late Antiquity, the palace transformed into a small urban settlement that inherited the layout with a rectilinear grid of streets and a separation between the official, cult and residential ar- eas. The densely inhabited modern-day city prevents us from understanding the development of the palace after the death of Diocletian. For a while, it certainly hosted local dignitaries, possibly even the Emperor Julius Nepos who fled Italy. In the southern part, there were radical alterations interpolating buildings into previously empty spaces. Archaeologists have been able to confirm the ex- istence of two large baths that were fitted into the rooms Fig. 2.180: Split. City plan. (Suić 2003, Fig. 180) of the already existing buildings in the palace’s south part after Diocletian’s death (Piplović 2016, 276). Their crea- tion can be brought into connection with the numerous springs of sulphurous and mineral water in the area of the palace and its immediate vicinity. Their abandon- ment has been dated to the 5th century. The area of the former palace was continuously inhabited and the pot- tery finds show lively trading in Late Antiquity (Dvoržak Schrunk 1989). The gradual influx of people from both Salona and the hinterland, who sought shelter behind the well-preserved wal s, resulted in the development of a small city with excess openings walled and public areas transformed. The cardo and decumanus remained the main axes of the newly-formed city and guided further growth (Suić 1976a, 238–240). Not much is known of the residential architecture, while the remains from the 5th and 6th centuries include different elements of Early Christian churches that exhibit a transformation from Roman cult places and a strong presence of a Christian population. The Anonymous Geographer of Ravenna mentions it as civitas (Cosm. IV. 16, V. 14). The urban agglomeration Lisina (Hvar) on the island of Hvar also gradual y developed in the 5th and 6th centuries (Katić 2003, 453–454). In addition to the remains of housing and defensive wal s, archaeological investigations unearthed huge amounts of imported goods and mention traces of Early Christian churches. As a safe port, it also played an important role in the Gothic Wars; Procopius mentions it as the spot where the Byzantine fleet gathered before attacking Salona. Fig. 2.181: Split. Porta argentea in eastern wal s (2017). 116 2. CITIES Fig. 2.182: Split. The present appearance of the west front of the Palace of Diocletian (2017). Fig. 2.183: Split. City wal s, preserved at a considerable height, and a tower in the south-western part of the palace (2017). Fig. 2.184: Split. Peristil of the Palace of Diocletian (2017). 117 Slavko CIGLENEČKI Fig. 2.185: Blagaj. Plan of the settlement and the metal urgic complex (Basler 1993, Fig. 1). BOSNIA buildings from the 6th century predominantly stood on the completely destroyed remains of a settlement from Đuro Basler identified Blagaj ( Figs. 2.185, 2.186) in the 1st to 3rd and possibly the 4th century, preventing the val ey of the River Japra as the only urban settlement us to speak of continuity. Basler dates the resettlement of Late Antiquity in Bosnia (Basler 1972, 37, 67–70; and renewed metal urgic activities possibly as early as id. 1977; id. 1993, 21–22). The rescue investigations the time of the Ostrogothic occupation, but certainly conducted in the 1960s in the Japra valley succeeded in throughout the 6th century. salvaging several buildings. It was the site of iron ore The building remains can be divided in two parts, extraction and ironworking ever since the beginning namely habitation remains at Crkvina and Bare in of Roman domination. Excavations revealed several the northeast and metallurgic remains to the south construction phases that could not be sufficiently clearly at Majdanište. More is known of the former, chrono- identified due to the heavily damaged remains. The site logical y reliably identified by the large Early Christian is classified into the group of settlements newly-founded basilica with an atrium, mosaics and decorated interior in Late Antiquity based on the observation that the furnishings. Two large contemporary houses were con- 118 2. CITIES 2.5.2. EASTERN PART SERBIA The significance of bishops in a Late Antique city is well il ustrated in three epigraphic monuments found in the eastern part of the area under discussion. The most telling is certainly the inscription mentioning the construction of a large episcopal residence (fortified settlement, vil a?) from the time of Justinian that came to light at Izbičanj near Prijepolje, at the eastern border of the province Dalmatia (cf. Stričević 1961, 181–182; Suić 1976a, 249−250; id. 2003, 373; Dagron 1984, 16, Note 71; Ćurčić 2010, 216). The inscription in verses was engraved on a 2 m long stone monument and records the construction of a city, fortified settlement or vil a on the initiative of Bishop Stefanus and (probably) financed by the Church. It relates the defensive wal s, sanctuar- ies, residential buildings, fountains, baths and stables as component parts of such a centre. Some identified a Late Antique vil a in this description, but the time of construction and the mountainous terrain far in the hinterland of the Adriatic coast make this unlikely. It is rather similar to some of the fortified hil top settle- ments that revealed prominent church complexes such as Hemmaberg, Kučar near Podzemelj or the lowland site at Louloudies near Katerini. The inscription stone was found in secondary position, though its weight suggests it could not have been brought from afar. A potential original location may be the rocky elevation of Kovingrad, some 2 km away, which holds the remains of a medieval fort and a terraced slope just above the River Lim that has as yet not seen archaeological investiga- tions. Its location is typical of Early Byzantine fortified settlements. If such an identification were correct, inves- tigations would enable a comparison between the Late Fig. 2.186: Blagaj. Plan of the Early Christian church (Basler Antique poetic description of a manor and the reality of 1993, Fig. 30) a modest fortified hil top settlement that hosted a bishop. The systematic research of recent decades has shown that the large Late Antique complex on Gra- dina on Jelica ( Figs. 2.187, 2.188) also ranks among the group of newly-founded cities (Milinković 1995; structed of spolia on the spot of earlier buildings and id. 2002; id. 2015, 143–190; id. 2020). It is a complex parallel with the basilica, but on the other side of the fortified settlement built on the summit and slopes of Roman road (Basler 1977; Busuladžić 2011, 165–166). an 846 m high hil . The results gained thus far reveal a The metal urgic part was enclosed with a tall wall fortified acropolis with several important buildings and most likely from the 3rd century onwards. It is here that flanked to the north and south by a walled area that has we may observe new constructions and additions to only been partial y investigated. Three Early Christian earlier buildings. The whole complex together with the churches were located to the west outside the fortified remains of contemporary cemeteries shows a revival of settlement. Earlier data suggest that this part was also metal urgic activities and construction of the settlement fortified. The wal s that enclose all three parts of the in the 6th century. settlement are of a different width in different sections, measuring from 0.96 to 1.65 m. The exposed part of the acropolis and the southern part of the settlement have revealed three protruding defensive towers of irregular shapes. The acropolis, girdled with its own wal s protect- 119 Slavko CIGLENEČKI Fig. 2.187: Jelica. Plan of the fortified settlement (Milinković 2010, Fig. 7). ing a surface of 1.25 ha, hosted a simple Early Christian basilica at the top accompanied by several houses of a solid construction of mortar-bound stone rubble. These were fitted with window panes and contained a cultural layer that held copious amounts of amphorae and im- ported pottery. Standing out among the buildings is a large multi-room edifice of a prestigious appearance, the finds from which include a figural column capital. Mihailo Milinkovič allows for the possibility this was the residence of a bishop (Milinković 2015, 155). Other buildings were living quarters, one held a workshop for metal working. The south walled area was roughly 1.28 ha large and only partly investigated, but revealed scarce remains of less well-built houses predominantly of clay-bonded wal s. The roughly 3.3 ha large north walled area revealed a beautifully furnished church with side rooms and a large house in its vicinity. Lean- ing on the exterior of the acropolis wal s was a large building with several rooms in at least two storeys, of which some were used for storage. It revealed rich finds that include items indicating the presence of barbarian newcomers. The unfortified west part held two more Fig. 2.188: Jelica. Plan of the upper part of the settlement (Milinković 2010, Fig. 9) 120 2. CITIES churches with individual graves, and a small cemeterial church at the far west surrounded by more than thirty graves. The construction of the settlement in the remote mountainous area is dated to the Justinian period and its end, indicated by strong traces of fire established in most buildings, to the late 6th or early 7th century. Some artefacts may suggest limited habitation at the site in the 4th/5th centuries (Milinković 2010, 284). Only slightly later than the Palace of Diocletian was another mighty fortified palace, namely Romuliana (Gamzigrad) ( Figs 2.189, 2.190), constructed by the Tetrarch Galerius, which also shows elements of a city newly-founded in Late Antiquity (Srejović, Vasić 1994; Ćurčić 2010, 139; Büllow 2011b). In contrast with the continuously inhabited Diocletian’s Palace, Romuliana was completely abandoned after the 11th century, which allows modern researchers to thoroughly investigate its basic features and chronology. The remains of a partial y investigated early defensive wal s measuring 200 × 180 m point to an fort presumably dating to the last quarter of the 3rd century. Under Galerius, considerably thicker wal s were built that enclosed a 234 × 195 m large area. These wal s were 3.6 m thick and reinforced with twelve polygonal protruding towers, the largest one positioned in the corners and measuring an impressive 26 m in diameter. We should emphasise that the wal s were not merely defensive in nature, but primarily had the func- Fig. 2.189: Romuliana. Plan of the settlement. Newly discov- tion of manifesting the power of the emperor and of ered buildings in the vicinity of the palace (Bullow 2011b, the Tetrarchic order of the Empire. A road divided the Fig. 3a). interior to the north and south halves. The north part held official buildings, lavishly furnished vil as with peristyles and audience hal s, as well as a small temple. of simple dwellings that existed alongside storehouses The south part had a large temple and baths, as well as and workshops, Basilica I was replaced by a considerably other public buildings that have as yet not been clearly larger Basilica II with a sizeable baptistery, and another identified with the exception of a likely horreum. This church was constructed, possibly even a third one. magnificent residential-memorial complex was final y The varied small finds show that this Late Antique identified in 1984 with the discovery of the inscription settlement was economical y self-sufficient, but heavily Felix Romuliana. For the purposes of our discussion, we ruralised, mostly inhabited by autochthonous popula- should focus on its transformation into a fortified set- tion with a barbarian presence. In addition to numerous tlement with certain urban features (Popović, V. 1982, buildings visible on the surface and partly also excavated, 555–556; Petković 2011; Milinković 2015, 249–257). geophysical surveys revealed others outside the wal s, in The palace with its high and thick wal s was gradual y the immediate vicinity of the palace. These include the and pragmatical y used as a well-protected settlement complex of three churches dating from the 4th to the 6th that hosted a prominent ecclesiastical centre, an array century corroborated during trial trenching (Milinković of workshops and fairly simple housing characteristic 2015, 251–257), as well as a rectangular walled area with of both former cities and many newly-founded fortified large elongated buildings, the function and chronology hil top settlements. Recent investigations revealed a clear of which have as yet not been explained (Bülow 2011, division into two post-palatial phases. The first one, dat- 162, Fig. 8). There are different interpretations regard- ing from the last quarter of the 4th to the mid-5th century ing the nature and significance of the former imperial or just beyond, revealed traces of smithing furnaces, palace in the 5th and 6th centuries, ranging from fortified workshops for dying textile, glassworks, the baths were vil age to urban settlement. Gerda von Bülow ties the still partly in operation, people were living in the tow- location with the numerous rich metal deposits in the ers. The first church was built in this phase – Basilica I region and the good lines of communication (Bülow located in the west part of the palace. The fort also hosted 2011, 159–160). Considering the concentration and a cavalry unit. The next phase, dating from the late 5th size of the churches, Slobodan Ćurčić sees the closest to the early 7th century, witnessed an increased number analogy in the episcopal centre at Louloudies and 121 Slavko CIGLENEČKI Fig. 2.190: Romuliana. Plan of the settlement (Büllow 2011b, Fig. 8). presumes an episcopal centre at Romuliana as well 2007; Ivanišević 2016; Ivanišević et al. 2019). Recent (Ćurčić 2010, 139). Having said that, the defence of research that applied numerous modern approaches this settlement exposed on a gentle slope required a has been able to provide a comprehensive picture of considerable number of people in spite of its mighty the settlement’s centre, but also suburbs, fortifications, wal s, which means a fairly densely inhabited interior housing of inferior quality, aqueduct and smaller forts and immediate proximity. associated with the city (Ivanišević 2011; Ivanišević, A newly-founded and reliably-dated city that Stamenković 2013; Ivanišević et al. 2016b; Ivanišević deserves special attention is Caričin grad ( Figs. 2017; Ivanišević et al. 2019). It importantly contrib- 2.191–2.196, 2.252, 2.262 , 3.306), which is increasingly utes to our understanding of the history of Byzantine convincingly identified as Justiniana Prima. The long urbanism. years of international systematic investigations have Caričin grad lies on a plateau slightly raised above been unveiling this Byzantine city that exhibits all the the Rivers Caričinska reka and Svinjarička reka, in an characteristics of contemporary construction, as wel area removed from the main roads, and is ful y adapted as the rapid decline of the urban fabric. More than a to the terrain. The centre, extending over roughly 8 ha, century of investigations has provided a good insight was protected with a strong, 2.2–2.8 m thick walls into its history and development (Kondić, Popović built in the opus mixtum technique. It was additional y 1977; Duval 1996; Bavant, Ivanišević 2003; Bavant fortified with numerous protruding rectangular towers 122 2. CITIES and a pair of impressive pentagonal towers flanking the with an imperial statue at the centre and surrounded by entrance to the upper city. These wal s were associated shops stood at the junction of the main lines of com- with thinner wal s of an inferior quality and without munication. New research shows that the areas previ- defensive towers that enclosed four suburbs. In the ously believed to have been uninhabited were actual y south, investigations established palisades and two deep filled with housing radial y arranged along the north ditches, in the east an artificial lake with a dam (also slope below the acropolis. The Lower city held a series used as a bridge) surviving in the length of 70 m. The of churches, as well as a large cistern and small baths, city and its suburbs extend across ca 20 ha. while the remaining space was taken up by houses with The wal s of the city divided it into three struc- atria, extensions and courts; geophysical surveys also tural y and functional y different parts: the Acropolis, revealed numerous other, smaller houses. The housing the Upper city and the Lower city. The Acropolis was in this area was built using a fairly simple technique located on the highest, north part of the hil , protected with wal s bound by clay or timber framework filled with wal s and towers, and held an episcopal church with cob and displaying several construction phases. with baptistery and a large three-part prestigious build- The city was surrounded by suburbs. In them, the ing in the north (presumably the episcopal palace), as only public buildings were baths and a hospice at the well as a small residential building in the south near entrance to the lower city. Monasteries are presumed the baptistery. The Upper city extended across the to have stood in the southern suburbs and on several, gentle slopes below the acropolis. Its main porticated more remote locations. Geophysical surveys in the streets boasted public buildings: three churches, a large suburbs also revealed numerous houses joined into administrative building ( principia) and horreum, as small units, to the west below the town also different well as a richly decorated ‘vil a urbana’. A round square workshops lining the river. Fig. 2.191: Caričin grad. Pla- teau of the Acropolis (2012). Fig. 2.192: Caričin grad. View from the Acropolis to Upper city and Lower city (2012). 123 Slavko CIGLENEČKI Fig. 2.193: Caričin grad. City plan (Ivanišević 2017, Fig. 2) 124 2. CITIES Fig. 2.194: Caričin grad. Plan of the northern slope of the Upper city (Ivanišević et al. 2016b, Fig. 8) Fig. 2.195: Caričin grad. Baths outside the wal s on the eastern side (2012). 125 Slavko CIGLENEČKI Fig.2.196: Caričin grad. Ruins of the modest dwellings, leaned against the northern wal s of the Acropolis (2012). The city was constructed according to a uniform Novel XI that speaks of the seat of the praetorian pre- concept with elements of classic urban design visible fecture of Illyricum being transferred from Thessaloniki primarily in the two main axes joining in a forum, public to Justiniana Prima and also of the episcopal see being buildings, a 20 km long aqueduct and a sewage system. transferred to a city built in the vicinity of the birthplace It clearly displays all three main functions: ecclesiastical, of the Emperor Justinian. administrative and military. The high number of church- The architecture of the city shows great diversity, es in the city, the suburbs and in the immediate vicinity which is characteristic of other newly-founded cities raise the possibility of a pilgrimage destination; the im- as wel . All the important public buildings, churches portance of the Church is also reflected in the observed and defensive wal s were constructed in opus mixtum. monasteries. Of all the newly-founded contemporary The largest portion of housing was constructed of clay- sites, Caričin grad boasts the most elements reminiscent bonded wal s of inferior quality, of unfired brick or of a classical city, which indirectly reveals its important even a timber-framed construction. In the second half administrative and ecclesiastical role (finds of seals and of the 6th century, there is a rapid construction of small bul ae) in the hierarchy of the last of the Late Antique dwellings in the empty spaces between public buildings, cities. The monumental architecture of a uniform and while the material culture reflects a strong ruralisation. complex design indirectly points to imperial patron- The density of the buildings was so great that some were age. There are very different opinions as to whether the even leaned against the wal s enclosing the Acropolis and archaeological evidence from Caričin grad reflects the thereby undermined its defensive function. Towards the enthusiastic, undoubtedly embellished description of end of the 6th century, urban life began to decline and Procopius (cf. Poulter 2007b, 20; Whittow 2007, 377). the settlement was being abandoned, only some houses Recent investigations have revealed many of the urban were renovated using spolia. The last modest dwellings features mentioned by Procopius that were previously with hearths date to the beginning of the 7th century. either not investigated or not identified (Ivanišević The city was final y abandoned in the first few decades 2016) and increasingly accurately il ustrate the note in of the 7th century. 126 2. CITIES BULGARIA decades of the 4th century, when it likely only covered the upper two terraces, i.e. 0.4 ha. The enlargement The greatest number of settlements with structural presumably came about when the fort transformed elements and literary references that allow us to identify into a fortified settlement after the col apse of the Late them as newly-founded Late Antique cities comes from Roman military organisation on the Lower Danube in present-day Bulgaria. This is largely due to the dangers the late 4th to early 5th century. Among small finds he this area faced, but also to the efforts of Byzantium to use mentions dolia, diverse pottery and coins from the 4th its imperial power to protect the areas on the doorstep to 6th centuries. The date of abandonment is unknown. of Constantinople. Located east of the city Novae and on the limes was The fortified settlement at Anasamus or Asemus the fort Iatrus (Krivina) ( Figs. 2.198–2.201), marked as ( Fig. 2.197) is marked as a city in ancient literary sources polis in the late literary sources, which had a ful y military (Torbatov 2016). It was founded at a strategic location nature in the beginning (Bülow 1995a−c; Bülow et al. at the confluence of the Rivers Osam and Danube, in 2007; Bülow, Wachtel 2015). It lies close to the confluence immediate proximity to a Roman roadside station. A of the Rivers Yantra and Danube, on a well-protected Roman garrison, mentioned in Notitia Dignitatum, natural terrace additional y secureded by the river. The was stationed here in the second half of the 4th century. long years of systematic excavations conducted by a team After the mid-5th century, it presumably lost its mili- of German and Bulgarian archaeologists have revealed tary significance and gradual y changed into a densely five habitation phases. These uncover the establishment populated settlement. Literary sources describe it as a and decline of the army fort, as well as its transformation stronghold that still hosted well-trained soldiers and into a large fortified vil age or small city that continued one that resisted the Huns. In the second half of the to be of great military importance. Its archaeological 6th century, Asemos was a flourishing city with all the record il ustrates well the fate of the forts along the Moe- features of urban life. It is the location of the well-known sian limes, which were exposed to incessant barbarian event from 594, when its city folk rebelled against Peter’s threats and incursions during Late Antiquity. The fort campaign and did not allow the departure of their city was constructed in the early 4th century and enclosed troops. Sergey Torbatov convincingly locates the Late with massive, up to 3.5 m thick wal s with numerous Roman and Early Byzantine Asemus to Osamsko kale, a protruding U-shaped towers combined with fan-shaped rocky elevation rising 70 m above the plain and the River ones in the corners. The wal s traced the configuration Osam. The remains here have not been investigated in of the terrain and are therefore irregular in plan. The detail, we only have field observations. The walled area incomplete preservation only allows an estimate of the on the top of the hill is just over 1.15 ha large and divided protected interior – roughly 3 ha. A single entrance has with wal s into three different parts of unknown dating. thus far been established, which was reinforced with a pair Torbatov states that the complete length of the defen- of external y protruding towers and a propugnaculum in sive wal s measures roughly 470 m and also mentions the interior. The main road led from the main entrance a proteichisma. Based on literary sources, he posits the to the principium and was lined with different buildings Late Roman fort was already constructed in the first few for the garrison, as well as shops, workshops and other Fig. 2.197: Anasamus. Plan of the fortified settlement (Torbatov 2021, Fig. 46). 127 Slavko CIGLENEČKI Fig. 2.198: Iatrus. Plan of the two phases of the fort and the fortified settlement: b – around 400, c – mid-6th century (Heinrich-Tamáska 2017, Fig. 5b−d). Fig. 2.199: Iatrus. Low terrace with ruins ot the city and its wal s (2012). buildings. A fairly large praetorium and two small private workshops. Other buildings were also transformed and baths were built slightly later. Towards the end of the 4th received more modestly built dwellings of clay-bonded and in the first half of the 5th century, both the principium stones. The new buildings no longer adhered to the former and praetorium were already partly in ruins and used for use of space within the fort and were haphazardly distrib- 128 2. CITIES Fig. 2.200: Iatrus. Remains of the Early Christian church Fig. 2.201: Iatrus. Remains of the praetorium (2012). (2012). Fig. 2.202: Carevec. City plan (Dintchev 2018, Fig. 15). uted in the interior. There is a marked ruralisation of the ian renovated the strategical y significant fort; the scarce fort that has been linked with the presumed presence of habitation traces show simple dwellings erected within foederati. The construction of two large horrea and the the remains of earlier buildings. The church was again first church also dates to this time. A thick layer of burnt renovated and enlarged; this and the heavy defensive debris and ruins indicates that the Huns burnt the fort in instal ations to some degree justify the identification of the middle of the 5th century, and it was abandoned for the fort as a city. In this form, the settlement persisted to several decades. Limited habitation is again perceptible the end of the 6th century. in the late 5th or early 6th century, in the shape of simple The Late Antique city at Carevec (Veliko Tarnovo), buildings erected within earlier edifices, but also a new presumed Zikideva ( Figs. 2.202–2.205) was established and larger church constructed on top of the earlier one. on a natural y exceptional y wel -protected location on a After a second fire in the first half of the 6th century, Justin- high rocky plateau surrounded by the meanders of the 129 Slavko CIGLENEČKI Fig. 2.203: Carevec. View from the top of the city towards the entrance in the west (2012). Fig. 2.204: Carevec. Walls and ruins of the buildings on the western side of the city (2012). Fig. 2.205: Carevec. Episco- pal complex from the south (2012). 130 2. CITIES Fig. 2.206: Abritus. City plan (Heinrich-Tamáska 2017, Fig. 10). Fig. 2.207: Abritus. Wal s on the west and apsidal terminal of the church (2012). Fig. 2.208: Abritus. View of the praetorium from the north Fig. 2.209: Abritus. Heavily fortified entrance on the north (2012). (2012). River Yantra (Dintchev 1997; id. 2018; Kirilov 2006b, Late Antique literary sources mention the fortress 70–72). The transformation of a small settlement into Abritus (Razgrad) as a city and episcopal centre ( Figs. a city has been dated to the late 5th or early 6th century, 2.206–2.209). The interior enclosed within thick wal s when defensive wal s were erected, as were the first and sited on a relatively high elevation has revealed public and numerous private buildings. Its final appear- several buildings of a public nature (Ivanov, T. 1980; ance was created in the middle of the 6th century. The Radoslavova 2011; Carrie, Moreau 2015). The wal s are northernmost part, which was natural y best protected, well-known, 2.7 m thick, reinforced with 35 towers and hosted the houses of the wealthy inhabitants, while other constructed in the early 4th century. They encircled a parts mainly revealed simple buildings with one or two 15 ha large area and had four entrances, two of which rooms. Standing out is a church complex comprising a with propugnacula. In the early 4th century, four large basilica, episcopal palace and monastery. The city inte- buildings stood in the east side of the interior and were rior revealed two other churches, one of them on top torn down in the second half of the same century. To- of the fortified hill and the other close to the entrance. wards the end of the century, a large residential complex Army barracks are mentioned along the defensive wal s. with a peristyle court was constructed on top of their It would appear that the city also had its suburbs – traces foundations and persisted to the late 6th century. An- of dwellings came to light on the slopes of the hill and other large house was found next to it, as well as smaller on the adjacent hil s of Devin grad and Trapesitsa. The houses and different workshops. Two large edifices stood fortified city proper presumably measured 15 ha, while at the western wal s, of which only the horreum has the whole inhabited area extended across some 30 ha. been positively identified. A large three-aisled basilica This well-protected settlement is believed to have re- was erected in the 5th century, under Justinian also a placed the abandoned Nicopolis ad Istrum and has been smaller three-aisled church at the western wal s. The identified as Zikideva, a city mentioned by Procopius. baths from the 4th/5th centuries were also changed into a 131 Slavko CIGLENEČKI sacral building. The abandoned horreum and its vicinity revealed simple dwel ings with clay-bonded wal s dating to the 5th and 6th centuries. The inhabited interior shows the city’s continuity at least to the end of the 6th century. The large and completely built-up fortified settle- ment above the city of Shoumen ( Figs. 2.210–2.212, 2.265) lies on an impressive rocky elevation and is thus natural y well-protected on three sides (Antonova 1978; ead.1981; Dintchev 1997, 48; id. 2007, 490; Ćurčić 2010, 214). Roughly in the mid-3rd century, a new settlement presumably grew on a previously inhabited spot. Habi- tation intensified in Late Antiquity and the settlement spread across 2 ha. In the more readily accessible part it was protected with an initial y thin and later, in the Early Byzantine phase, thicker wal s with numerous pentagonal Fig. 2.210: Shoumen. City plan (Dintchev 2007, Fig. 10) Fig. 2.211: Shoumen. Densely inhabited interior of the fortified settlement (2012). Fig. 2.212: Shoumen. Part of the settlement with the church in the foreground (2012). 132 2. CITIES Fig. 2.213: Zaldapa. City plan (from Moreau, Baudry 2021, Fig. 2). and semicircular towers. The densely built-up interior and individual buildings in the exposed parts of the city. had many narrow streets lined with long multi-room Early reports mention a large public edifice ( horreum?) buildings. The interior also revealed a large three-aisled and an Early Christian basilica, new research has ad- church and another one outside the defensive wal s. Both ditional y revealed two other Early Christian basilicas. underline the significance of the well-organised large set- The buildings detected on the surface do not allow tlement and its central position within the region. conclusions to be drawn, as neither their chronology Zaldapa ( Fig. 2.213) has only been investigated in a nor function are clear, but they do offer a good insight limited extent, but the small-scale research and satellite into the spontaneous urbanism of the final centuries images nevertheless offer an interesting insight into the of Antiquity. The lines of communications are not ar- structure of the newly-founded city from the 4th century ranged in an orthogonal grid, some empty spaces may (Torbatov 2000; Beaudry et al. 2018; Dintchev 2018, Fig. represent small squares (particularly those on both sides 9). Most buildings inside the 25 ha large city have not of Basilica 3). Most edifices can be interpreted as small been investigated, though the absence of later habitation one- or two-room buildings, akin to those predominat- has enabled the size and distribution of the buildings to ing in the contemporary fortified hil top settlements. The be clearly visible. The military architecture in the interior city that is also mentioned in Hierocles’ Synekdemus was did not play a major role and was likely limited to towers abandoned in the late 6th/early 7th century. 133 Slavko CIGLENEČKI The settlement of Acrae (Cape Kaliakra) ( Figs. wal s on the only accessible side. The last of the wal s 2.214, 2.215) gradual y developed into a city during dates to the second half of the 4th century and enabled the 5th century (Dintchev 1999, 50; id. 2009, 426–429). a considerable enlargement of the settlement. It lies on a natural y excellently protected peninsula on The city most frequently cited as a new construc- the coast of the Black Sea that was protected with three tion from the time of Diocletian is the mighty fortress Fig. 2.215 Acrae. View of the easternmost part of the city on the peninsula (2012). Fig. 2.214: Acrae. Wal s with a tower (2012). Fig. 2.216: Hissar. City plan. (Dintchev 2018, Fig. 11). 134 2. CITIES Fig. 2.217: Hissar. Wal s on the western side of the city (2012). Fig. 2.218: Hissar. Well-preserved entrance in the southern Fig. 2.219: Hissar. Ruins of the urban vil a (2012). wal s (2012). of Hissar (Hisarya) ( Figs. 2.216–2.219, 2.249), which between the 4th and the 6th century. The city interior was presumably named Diocletianopolis after the em- revealed traces of a wide avenue connecting the south peror (Madzharov 2012; Rizos 2017b, 28–29; Dintchev and north entrances with a tetrapylon in the centre. 2018, 359). It lies on an important road leading across Investigations explored two large peristyle houses the Stara planina mountain range and is believed to and public baths (Kirilov 2006a, 76). The remains of have been established for the local thermal springs. The army barracks along the southern wal s show a strong different design and different role of the Late Antique military presence. city is clearly perceptible in the visible military and Hierocles also mentioned the city of Tzoides, which civilian elements. Particular attention was paid to the has been identified in the large fortified post at the imposing defensive wal s that still mark the city today edge of the modern-day city of Sliven (Dintchev 1999, and enclose an area of 30 ha. It presumably dates to the 54–55; Rizos 2013, 685; Borisov, Coičev 2019) ( Figs. Diocletianic times, but later witnessed several additions 2.220, 2.221). It has strong defensive wal s of a roughly and modifications. The walls are 2.5 to 3.5 m thick rectangular layout and enclosing an area of 4.5 ha on a and up to 12 m high. Constructed in the opus mixtum small elevation above the modern-day city. It was pre- technique, they incorporate numerous protruding sumably fortified in the early 4th century. The interior rectangular towers, as well as fan-shaped ones in the held a large Early Christian basilica and many houses corners. A proteichisma protected the city on the most and workshops along the wal s. A long horreum appears exposed, north side. As many as nine Early Christian to have stood in the southwest corner and there is also churches were constructed in and around the city mention of extensive suburbs with a church. 135 Slavko CIGLENEČKI Fig. 2.220: Sliven (Tzoides). Church on the highest point of the plateau (2012). Fig. 2.221: Sliven (Tzoides). Densely distributed houses along the north wal s (2012). Fig. 2.222: Cabyle. Wal s and baths in the north-west cor- ner of the city (2012). 136 2. CITIES Fig. 2.223: Cabyle. Large Early Christian church with the acropolis in the background (2012). On the spot of an earlier Thracian settlement and Roman camp, a city was founded at Cabyle near Yambol ( Figs. 2.222, 2.223) that literary sources relate held an episcopal see already in the middle of the 4th century (Dintchev 2000, 75–77; Kirilov 2006b, 40–41). Archaeo- logical finds indicate times of prosperity in the late 3rd and first half of the 4th century. The camp on the Hissarluka hil was then transformed into a settlement covering a surface of 3 ha, while two basilicas, a horreum, baths and other buildings grew outside this area. The question that is stil largely open pertains to its fortifications. In the 5th and 6th centuries, two Early Christian churches were constructed, as well as numerous small and poorly built houses in the ruins of earlier buildings, both within the fort and in the ruins of the horreum and baths. According to Dintchev, Cabyle did not become a proper city in Late Antiquity; it was an important local centre to the early 7th century he marks as a semi-urban fortified settlement. ROMANIA The province of Scythia, now in Romania, holds several important newly-founded cities that show a strong military presence alongside their civilian role. The fortress Tropaeum Traiani (Adamclisi) ( Fig. 2.224) was constructed in 315/317 (Poulter 1998, 331; id. 2007b, 35; Saradi 2006, 291; Ćurčić 2010, 49; Poulter 2010, 25–27; Rizos 2017b, 27). It is marked by strong defensive wal s with numerous U-shaped towers enclosing a considerably large surface of 10.5 ha. Alongside other archaeological evidence, the founding inscription is of particular impor- tance that mentions both a military and civilian compo- Fig. 2.224: Tropaeum Traiani. City plan (Dintchev 2018, Fig. 6). 137 Slavko CIGLENEČKI Fig. 2.225: Dinogetia. City plan (Döhle 1995, Fig. 7). nent. A large horreum was built in the city centre, at the holds numerous buildings including a large basilica built junction of both main streets, other public buildings are into the former principia. missing. Already in the early 5th century, private buildings presumably encroached on streets and portici. Four Early Christian basilicas, constructed towards the 5th and in the ALBANIA 6th century, underline the city’s civilian nature. Dinogetia ( Fig. 2.225) is another excellent exam- A fortress also transformed into a city at Kalaja ple of a newly-founded small city with a double nature near Grazhdani ( Fig. 2.226) (Baçe 1976, 70; Bunguri in the exposed part of the Empire: it was an important 2007; Rizos 2017b, 31). The city formed at a strategical y military post and a settlement of civilians (Popović, V. important site and was intended to control the commu- 1982, 557–559; Barnea 1986; Torbatov 1999; Dintchev nications below. It lies on slightly inclined terrain and 2006, 14, 16; Ćurčić 2010, 29). Under Diocletian, extends across 34 ha. The defensive wal s in opus mixtum a fort measuring 1 ha in size was constructed on a trace the configuration of the terrain and were reinforced small elevation on an island above a tributary of the with forty protruding rectangular and U-shaped towers. Danube, in the northeasternmost part of the Danube The interior has been poorly investigated and the dura- limes. It is particularly marked by strong, 3 m thick tion of the city cannot reliably be determined. walls with fourteen characteristics horseshoe- and fan-shaped towers, a heavily fortified entrance with a propugnaculum and a natural y protected location on an island. In addition to the presumed praetorium ( burgus?) and a large private villa (presumed home of the commander?), a three-aisled basilica was also constructed already in the 4th or 5th century that was later twice renovated and enlarged. In the second half of the 5th century, the interior of the fort was filled with housing. The fort was destroyed in the mid-6th century, the Byzantines renovated it and constructed simple dwellings with clay-bonded stone wal s. It was abandoned during the Avaro-Slavic incursions. A similar example of a military – civilian settlement is Troesmis (Turcoaia, Igliţa) (Poulter (ed.) 2007a). At the transition from the 3rd to the 4th century, the legion- ary fortress here was moved from the area between two plateaus to the east plateau. Its wal s with U- and fan- shaped towers girdle a surface of 2.2 ha. The interior Fig. 2.226: Kalaja. City plan (Rizos 2017b, Fig. 13) 138 2. CITIES A better known example of a fort-to-city trans- NORTH MACEDONIA formation is Scampis (Elbasan) ( Fig. 2.227) (Karaiskaj 1972; Amore et al. 2005, 348−351; Hobdari, Cerova Several fortifications of a size and elements indica- 2014). The excellently preserved and strong defensive tive of an urban settlement have been found in North wal s are the remains of a large rectangular fortress meas- Macedonia, a country rich in settlement remains and uring 308 × 348 m, through which led the important also archaeological-topographic investigations. We Balkan communication of Via Egnatia. The wal s were should particularly mention four fortified settlements reinforced with 26 towers, fan-shaped in the corners that certainly fall into the group of newly-formed cit- and roughly U-shaped elsewhere. The beginning of the ies. Other sites, also large but less well-known, will fortress has been set to the first half of the 4th century and be discussed among the countryside settlements (see the renovation observable in the wal s from the mid-5th Chapter 3). century to the beginning of Justinian’s rule. It is believed The settlement on Davina, Kula in the vicinity of to have been abandoned in the 6th century. In the late 4th Čučer ( Figs. 2.228–2.229) was constructed above the and early 5th centuries, pseudocomitatenses Scampenses valley that held the road from Scupi to Ulpiana, just were stationed here, after which the fortress is believed before it entered the Kačanik Gorge (Mikulčić 1982a, to have gradual y transformed into a city. In the interior, an Early Christian church was found near the southeast corner. Two other basilicas with numerous fragments of interior furnishings came to light extra muros. For 458, a bishop is mentioned in this city that is identified as such in Hierocles’ Synekdemus and presumably also in Procopius. Fig. 2.227: Scampis. City plan (Rizos 2017b, Fig. 14). Fig. 2.228: Čučer (Davina, Kula). City plan (Mikulčić 2002, Fig. 45) 139 Slavko CIGLENEČKI Viktor Lilčić, who excavated the upper church, sees the numerous finds from the 3rd century as evidence of the first fort constructed on the hill already in the second half of the 3rd century (Lilčić 2004, 180). The low, but natural y well-protected hill of Taor ( Figs. 2.230–2.232) just above the fertile plain and the River Vardar holds the remains of a large, densely inhabited and fortified settlement. As early as the second half of the 19th century, Artur Evans hypothesised that the phonetic similarity of this site and the adjacent site of Bader suggest it can be identified with Tauresium, the birthplace of Justinian located close to the Bederiana fort. The topographic and surface surveys began in the Fig. 2.229: Čućer (Davina, Kula). Settlement from the east 20th century (overview in Mikulčič 1982a, 103–106), (2013). while modern systematic investigations have been taking place over the last two decades (Ristov 2012; id. 2019). The site has three prominent parts: central plateau, 107–116; id. 2002, 101, 153–158; Lilčić 2004, 180–181; acropolis on the south side, as well as a large habitation Ćurčić 2010, 182). Ivan Mikulčič distinguishes between terrace and another one below it on the north (Mikulčić two parts of the settlement. The earliest is the acropolis 2002, 187–189). Recent systematic excavations focused on the highest part of the hill that held a Late Roman on the settlement’s central part and trial trenches also fort (size 185 × 110 m). On a slightly lower terrace below partly investigated the surrounding area and the area is the presumably civilian settlement measuring 260 × below the settlement (Ristov 2019, 89). All this offers a 110 m. The settlement is encircled with wal s with tow- good insight into the simple urban-rural unit that was ers and a proteichisma, it presumably has a forum in most intensively inhabited in the 6th century and the the centre. Both parts have a three-aisled church and many small finds from which tell of the everyday life differently-sized housing. This reveals a uniform urban (Ristov 2012; id. 2019). design from the 6th century. Below the wal s is a 4–5 ha The central part of the settlement and the acropolis large unfortified contemporary settlement, in its vicinity are girdled with 1.6 m thick wal s reinforced with sev- also five Early Christian churches. Mikulčić interprets eral towers. The most important and also largest was the highest part of the post as a Late Roman fort that the tower – bastion in the south where the settlement later expanded and grew into a settlement, in the 6th cen- was most readily accessible. This rectangular tower was tury then into a small city with an ecclesiastical centre. supported on two sides by additional wal s and the in- Fig. 2.230: Taor. Excavated buildings in the north-eastern part of the settlement (2009). 140 2. CITIES terior probably filed with stones. Access on this side was architecture and strong defensive wal s. A church has additional y protected with a deep ditch. Other towers as yet not been found, but a Christian nature of the were also rectangular and protruding, only the one on settlement is reflected in numerous small finds with a the west side was constructed from the interior side. The Christian symbolism. Also numerously represented are settlement thus far revealed fifteen densely spaced build- the finds of militaria, which indicate a military nature ings. Of these, a large public building in the northeast of the site already in the 3rd and 4th, even more so in corner stands out, showing a high quality of construction the 6th century. and composed of two units separated by a colonnade. The site was occasionally inhabited already in Its wal s of stone and brick were bound by mortar and earlier periods, with the first reliably identified artefacts the building was decorated with marble columns with indicating a military post as early as the 3rd century capitals, bifore and the facade partly ornamented with (Ristov 2019, 50–52). More intensive habitation appears bricks; it as well has a spacious atrium. There are also to date from the second half of the 4th century onwards, simple rectangular houses, most often with two rooms, when a large portion of the buildings on the hill was of an inferior construction and clay-bonded wal s. The constructed. Kiro Ristov emphasises that the defensive suburbs were populated as wel , but subsequent agricul- wal s with towers were only erected in the 6th century, tural activity there destroyed the architectural remains. presumably after the devastating earthquake of 518. The The buildings known thus far suggest the existence of defence of the settlement prior to this time is unknown. a ruralised urban settlement that had its own aqueduct The small finds of both civilian and military nature and and sewage system, differentiated public and private particularly the hundreds of coins indicate that habita- Fig. 2.231: Taor. City plan (Ristov 2019, Fig. 12). 141 Slavko CIGLENEČKI Fig. 2.232: Taor. Southern edge of the settlement (2004). tion was at its peak here throughout the 6th century or which has been interpreted as an episcopium. It had possibly slightly earlier (Ristov 2019, 258–262). an interior courtyard with colonnade (Snively, Sanev A large and unusually fortified settlement was 2013, 160–167). detected at Golemo Gradište near Konjuh ( Figs. 2.233, A ‘south residential complex’ came to light on the 2.234) that shows compelling urban features of cities highest part of the north terrace; it is a cluster of several newly-founded in the 6th century. This complex site has rooms of irregular plan in which excavators see several a long rocky ridge at the centre that rises some hundred buildings that have one to three rooms and some even metres above the surroundings and is flanked north and an upper storey. The rooms held numerous pithoi. Two south by terraced slopes. The north terrace is larger and phases have been established, both dating to the 6th cen- more suitable for habitation, and has in part already tury. Identified as a particular feature was an irregularly been investigated. The whole complex is protected with shaped city block defined by streets (Sanev et al. 2012). defensive wal s with towers and the River Kriva offers The wal s of these houses were clay-bonded, only the additional protection in the north (Mikulčić 2002, ‘north residence’ shows a higher quality of construction 128–133; Snively 2002; ead. 2011; Snively, Sanev 2013). with the wal s, albeit still clay-bonded, of more precisely The systematic investigations in the last two made faces and decorated with brick courses. decades have provided details of the residential and Outside the wal s, a church in the shape of a ro- church architecture, as well as a reliable chronology of tunda was already found before World War II. With the site. Trial trenching on the acropolis has shown a its centralised form, it is an important addition to the densely inhabited area and small buildings sunken into church architecture in the wider area (Snively 2011, the bedrock with rock-cut-foundations for the wooden 195–197). superstructure. Also found was a large masonry building Caroline Snively posits that the city on the north with three rooms, an apse and mosaic floors, which was terrace already began in the second half of the 5th cen- presumably of a residential or administrative function tury and witnessed substantial additions in the Justinian (Snively, Sanev 2013, 158–159). period, more precisely in the second quarter of the 6th A large Early Christian basilica with three aisles century, when the fort on the acropolis with numerous and a narthex stood in the centre of the north terrace. sunken buildings and defensive wal s around the whole Its rich furnishings date it to the 6th century (most likely settlement is believed to have been constructed (Snively its middle) and its central position suggests it was the 2011, 188). She also observes that Golemo Gradište is the main church and likely the episcopal see (Snively 2011, first example of cities such as are only known in greater 197). The north terrace also revealed several houses, number in the 6th century, when semi-urban settlements shops and workshops. The most prestigious house were being constructed on natural y well-protected loca- was the ‘north residence’ directly below the basilica, tions (Snively 2008). 142 2. CITIES Fig. 2.233: Konjuh (Golemo gradište). City plan (Snively 2017, Fig. 11). Fig. 2.234: Konjuh (Golemo gradište). Plan of the acropolis (Mikulčić 2002, Fig. 22). 143 Slavko CIGLENEČKI The city of Bargala (Kozjak) ( Figs. 2.235–2.238, and four fortified entrances. In the east, it is addition- 2.261) was established on a major road leading to al y protected with a proteichisma and a ditch, but also Thessaloniki from the north (Aleksova 1997, 155–166; an exterior propugnaculum at the entrance gates. The Mikulčić 1999, 262–264; id. 2002, 391–396; Beldedovski, remains of an episcopal basilica were found in the north- Nacev 2006; Nacev 2016; Nacev, Atanasova 2017). The east corner of the city that dates to the late 4th century fortress was likely already founded in the early 4th cen- and is associated with a baptistery, it was also renovated tury, while the inscription from 371 that celebrates the in the 5th century. A bishopric in Bargala is confirmed construction of city gates points to a civilian city. The in literary sources, while the church itself revealed a city of a trapezoidal, 4.7 ha large interior heavily fortified column capital bearing the name of Bishop Hermias. with wal s and towers lies on a gentle slope above the Leaning to its south aisle is a smaller and later church River Bregalnica. It boasts numerous rectangular towers built of reused stones, presumably medieval in date. Near the basilica was a large cistern and a presumed episcopal palace with a spacious apsidal reception hall and several smaller residential rooms. Baths and economic facilities (granary, wine presses and furnaces) have been found between the presumed episcopium and the defensive wal s. Some partial y investigated buildings were leaned against the defensive wal s, others are presumed in the central part of the largely uninvestigated interior. The remains of a small three-aisled basilica, named ‘city basilica’, were excavated at Via Principalis, while a large three-aisled basilica with Late Antique burials came to light extra muros, in the vicinity of the main entrance. Fig. 2.235: Bargala (Kozjak). City plan (Mikulčić 2002, Fig. Fig. 2.237: Bargala (Kozjak). Residential and economic facili-298). ties in the northern corner of the city (2013). Fig. 2.236: Bargala (Kozjak). Episcopal complex (2013). Fig. 2.238: Bargala (Kozjak). Entrance with propugnaculum at the north-western wal s (2013). 144 2. CITIES Archaeologically clearly identifiable is a small An Early Christian basilica was erected in the city founded on a low elevation called Kale, Gradište upper part of the interior and another, smal church in the southwestern part of Vinica ( Figs. 2.239–2.242) with a baptistery beside it. Investigations in immediate (Mikulčić 2002, 256–260; Balabanov 2011; Dimitrova proximity revealed a large edifice with numerous sunken 2012, 91–105). Already in Hellenistic times, this spot dolia for grain, as well as a complex of rooms of a resi- hosted a settlement with a temple and drystone wal s that dential nature. The proximity to the two churches, the enclosed a roughly 3.5 ha large area. The different and layout of the complex, the hypocaust and the clay relief not always consistent information in literature reveals ‘icons’ have led authors to suggest this was the seat of that a fort stood on the highest part in the 4th century ecclesiastical administration (episcopium?) (Manaskova, (according to Balabanov already in the mid-2nd century) Angelovski 2017). In addition to these buildings on the girdled with 1.1 m thick wal s with towers and a 2.5 ha highest, south part of the settlement, investigations in large interior; not much is known of this fort. In the recent decades unearthed numerous others that indicate time of Justinian and his renovatio imperii Romanorum, the walled area was densely built-up. The houses in the it was renovated and added 1.5 m thick defensive wal s prestigious part show a high-quality construction with with new towers along the exterior. The settlement was mortar-bound wal s, while the lower north part appears also protected with a wide ditch on the side of easiest to host a complex of different, densely spaced and poorly access, while the old Hellenistic wal s encircling the built houses with clay-bonded walls; these contain Early Byzantine acropolis from three sides were used numerous sunken dolia and storage pits. Directly as a proteichisma. below the settlement is the outlines of 4–5 ha large Fig. 2.239: Vinica (Kale, Gradište). Panoramic view of the settlement from the east (2013). Fig. 2.240: Vinica (Kale, Gradište). Plan of a section of the city wal s with a tower and presumed episcopium (Mikulčić 2002, Fig. 148). 145 Slavko CIGLENEČKI Fig. 2.241: Vinica (Kale, Gradište). Northern part of the settlement with churches in the foreground (2009). Fig. 2.242: Vinica (Kale, Gradište). Main communication with neatly distributed buildings in the south part of the city (2013). suburbs with an Early Christian church. Another, but ond half of the 4th and the 6th century indicates a gradual unprotected Late Antique settlement has been discov- development from an fort to an urban agglomeration at ered 1.5 km away that also revealed traces of a church. least in the early 6th century. The urban character of the Most activities connected with the excavations at settlement is perceptible in the uniform concept behind Kale were directed towards interpreting the numerous, the distribution of buildings with churches, a cistern, but functionally unusual clay relief tablets (‘icons’), granary and military edifices in the top part, as well as whereas very little is known of the housing and the a small rectangular and paved open space (presumably contexts enabling a reliable interpretation of the ar- forum), a straight street leading across the centre of the chaeological remains. It is thus challenging to write of settlement, aqueduct and sewage system. The beauti- the development the city witnessed between the 4th and ful y decorated basilica in the city and numerous others the 6th century. The concentration of coins from the sec- in immediate proximity may indicate an episcopal see 146 2. CITIES (Dimitrova 2012, 261). The newly-founded agglomera- tion can be interpreted as one of the last urban centres of Antiquity in the rich and fertile valley of the River Bregalnica, in the vicinity of important mining regions. The cities newly-built on the remains of earlier set- tlements and forts may include the fort of Castra Nicea (Dolenci) located at the Kale site below the mountain pass of Diavato, at an altitude of 820 m ( Fig. 2.243). It lies on the Via Egnatia and is referred to as a civitas in Synekdemus (639, 3) (Mikulčić 2002, 53, 271–273; Ćurčić 2010, 47). The ruins of a 122 × 125 m large army fort enclosed within 2.5 m thick wal s are clearly visible on the ground. The wal s had round towers in the corners and U-shaped ones in between. Two entrances are vis- ible, in the centre also traces of a large edifice with the remains of mosaic floors, presumably belonging to a basilica. Above it were the remains of modest dwellings with sunken dolia. A roughly 3 ha large Late Antique settlement with a church and mosaic floors has been detected east of the fort, slightly further away another Early Christian basilica. The fort was transformed into Fig. 2.243: a city in the 5th/6th century and presumably received its Castra Nicea (Dolenci). Plan of the fort (Mikulčić 2002, Fig. 165). own bishop (Mikulčić 2002, 272). GREECE There is little archaeological evidence on the Castoria. Today, this natural y excellently protected is- newly-formed urban settlements in Greece, while more land shows no archaeological traces that could be linked information on the topic can be gained from literary to the Justinian period (Karagiorgou 2001b, 57−161). sources. Archibald Dunn notes that Hierocles mentions Much better investigated is the unusual example of cities in northeastern Greece in places that only hold the a newly-founded city of Louloudies near Katerini ( Figs. remains of differently-sized fortified settlements (Dunn 2.244, 2.245, 2.247), where rescue excavations revealed 2002, 707). Procopius reports that Justinian transferred a 90 × 80 m large quadriburgium (Poulter, Marki 1998; the city of Diocletianopolis to a small island in Lake Saradi 2006, 388–390, 467–468; Sodini 2007, 317). It Fig. 2.244: Louloudies. Fortified episcopal centre from the south-west (2002). 147 Slavko CIGLENEČKI Fig. 2.245: Louloudies. Plan of the fortified ecclesiastical centre (Saradi 2006, Fig. 41). lies 6 km south of ancient Pydna, on a plain close to the facilities constructed. After the earthquake in the mid- coast. Its well-defended interior revealed a large Early 6th century, the ecclesiastical centre was abandoned, Christian church, parallel to it a presumed episcopal while artisanal activities continued in the fort. Under- palace and along the defensive wal s other auxiliary standing the functioning of the city must also take into buildings and army barracks. It is suggested that it account the 4 ha large fort located mere 150 m away, can be identified as an episcopal see, created roughly which was observed during the intensive surveys and in 480 after the Emperor Zeno had surrendered the electromagnetic measurements and dated using the city of Pydna to the Goths in 479. Under Justinian, the recovered pottery to the 6th, possibly even the early 7th basilica and episcopium were enlarged and economic century (Poulter 1998, 343). 148 2. CITIES 2.6 CONCLUDING OBSERVATIONS 2.6.1 TRANSFORMATION ON LATE ANTIQUE CITIES OF LATE ANTIQUE CITIES The overview of the characteristic Late Antique This is a phenomenon observable in al parts of the cities has revealed great differences between them, but Roman Empire, but changes were particularly radical be- also the different ways in which they were transformed tween Ravenna and Constantinople. The major changes in different geographic settings and phases of Late An- include a shift of the urban population – both temporary tiquity. Some underwent changes within their original and permanent – to new and natural y better protected design, others reduced in size, still others were moved locations, as well as a reduction of the urban area. Shifts partial y or completely to better protected locations. The are characteristic of the cities that were abandoned and cities that developed more intensely in Late Antiquity partly also of those with continuity, while reduction were those that went on to become provincial centres, affected the cities with continuity and is discussed in but also those located on major land and maritime connection with this group. routes – i.e., those with a prominent imperial support; The gradual abandonment of old city centres is this is particularly apparent in the eastern half of the Em- associated with shifts of the population to nearby loca- pire and is reflected in a number of important buildings. tions that offered better natural or man-made protection. In the western half of the area under discussion, From the late 4th century onwards, the inhabitants of there is a marked disparity between the cities in Italy and the civilian towns that grew next to former legionary on the eastern Adriatic coast, on the one side, and the fortresses in Lauriacum, Carnuntum and Aquincum cities in the interior, on the other. The former do show gradual y moved to the fortresses that had, at least par- extensive degradation of housing, but also a persisting tial y, been abandoned by then. There they built simple presence of high-quality architecture in churches and dwellings, in Lauriacum also a church. The fortress in elite housing. This has been observed from Ravenna, Aquincum was enlarged and shifted to a better pro- Brescia, Verona and even greatly reduced Aquileia to tected part near the Danube. Castel um Favianis was the large cities of the eastern Adriatic such as Parentium, also greatly expanded, to host the refugees from the Pola, Iader and Salona. The continental cities, on the neighbouring areas who sought shelter in the urban other side, exhibit a rapid decline and abandonment with settlement. only some provincial centres that indicate (Teurnia) or Where there was a suitable hil top location in the allow us to expect (Siscia) prominent buildings erected vicinity, this shift already began in the last third of the even in the 5th and 6th centuries. 3rd century in a limited measure and more intensely The reasons for the decline or transformation of from the late 4th century onwards. An example of such a cities have been tackled by a number of authors who migration was already noted in the early 20th century for provided a complex explanation of the underlying po- Virunum; Rudolf Egger wrote in 1916 that the remains litical, economic and social changes coupled with the of two churches on the adjacent, only 40 m high fortified increasing role of Christianity. In the area under discus- hill of Grazerkogel ( Fig. 2.246) should be identified as the sion, this process was greatly accelerated by numerous shelter of the bishop and the inhabitants of the lowland incursions that frequently had disastrous consequences. city (Egger 1916, 105–109). The poorly investigated Fig. 2.246: Virunum, Grazerkogel. Hill from the south-west (2016). 149 Slavko CIGLENEČKI Fig. 2.247: Louloudies. Fortified ecclesiastical centre from the north-west (2002). Grazerkogel does not allow us to verify this hypothesis, 41), though a complete lack of investigations at the site but the Early Christian churches and the habitation does not allow us to confirm this hypothesis. area enclosed with massive wal s doubtlessly hosted Population shifts occurred later in the cities in the the Virunum bishop at least for a short while. In addi- south, primarily Greece, compared with those in the tion to this fortified location closest to Virunum, other western part. After the arrival of the Goths in Pydna, potential shelters located at a greater distance include which took place in 479, a new episcopal see is believed Ulrichsberg, possibly even the major Christian centre to have been established in Louloudies ( Fig. 2.247). The on Hemmaberg. old Diocletianopolis was abandoned and its inhabitants Similar shifts to adjacent hil top fortifications have moved to the island of Castoria only under Justin- also been recorded for Iulium Carnicum and Aguntum. ian (Karagiorgou 2001b). The city of Demetrias was The inhabitants of the former presumably moved to an abandoned in the same period and its administrative elevation above the city that has not yet been investi- functions and inhabitants relocated to the Iolkos hil gated with the exception of the church. In the case of (Karagiorgou 2001a). Aguntum, the migration may have been to Kirchbichl, the site of a large Late Antique settlement near Lavant, There are also several examples where most in- though there are also other possible shelters, but these habitants moved either to the periphery or to a safer so far only revealed churches from the 5th/6th century location farther away and some population persisted (Walde 2002, 160–161). The population of Iuvavum in the old fortified city centres. Particularly well-known presumably moved to a hil top location just above the is the migration from the endangered Aquileia to the city, a site that is poorly known due to the large medi- insular Castrum Gradense. More precisely, the strongly eval castle now occupying the spot; it is possible that reduced Aquileia fortified with Byzantine defensive the church mentioned in Eugippius may refer to one wal s persisted into the 6th century, while a new forti- that stood on this very hill (Ladstätter 2002, 313–314; fied centre of limited accessibility gradual y developed Kovacsovics 2002, 199). in the nearby lagoon that also welcomed the head of the Most cities on the eastern coast of the Adriatic had Church towards the end of the 6th century (Mirabel a a continued existence, but there are some exceptions. An Roberti 1974–1975; Tavano 1976; Cuscito 2001). example of a city abandoned before the end of Late Antiq- Late Antique habitation traces in Tridentum came uity is Argyruntum. The reasons for its abandonment are to light both in the city, on the rocky elevation of Doss unclear. It has been presumed that the civitas Argerunto Trento and in the fortified area at Piedicastello below mentioned in the Anonymous Geographer of Ravenna it (Bierbrauer 1985, 497–498; Cavada 2019). Without refers to the hil top settlement on Tribanj, Šibuljina, sev- systematic investigations, it is not possible to establish eral kilometres away (see Chapter 3.3) (Dubolnić 2007, in which periods any of the three habitation areas were 150 2. CITIES Fig. 2.248: Emona. Walled up gates in the southern part of the Roman wal s (2018). intensely inhabited. The advice that Theoderic gave to 2.6.2. CHARACTERISTICS the cityzens, that they should seek shelter on Verucca OF LATE ANTIQUE CITIES (interpreted as present-day Doss Trento) in times of danger, certainly points to a shift to the more easily DEFENSIVE STRUCTURES defensible hill – this is confirmed by archaeological evidence. Because of its location on one of the major The Roman cities not protected with defensive arteries, the city must have come under threat on several wal s were abandoned soon after the 4th century (e.g. occasions. For the same reason, we may also expect that Virunum, Sopianae, Neviodunum, Andautonia), with substantial army units were stationed here. exceptions in remote areas such as Gortyn on the island In Nicopolis ad Istrum, a small fortified complex of Crete. There are, in contrast, also rare new unforti- was created at the edge of the former Roman city in fied agglomerations established in this time that can the second half of the 5th century, on a natural y most be linked to the presence of a new authority, either protected spot. Of building remains, it only revealed Ostrogothic or Byzantine. Examples of these are Blagaj a church and another large building thus far; other in the valley of the Japra, in Bosnia, and Bosar on the buildings may have been made of wood. The limited island of Krk. investigations do not al ow for an estimate of the share of Several previously unfortified cities were enclosed the population that remained in the old city. It has been with defensive wal s in Late Antiquity (Verona, Teurnia, presumed, however, that it was mainly the ecclesiastical Gorsium, Celeia, Nesactium, Carevec, Abritus, Dyr- and administrative elites, as well as the army who sought rachium, Taor, Heracleia/Perinthus). Most commonly, shelter here (Poulter 2007c, 78–79). however, the already existing walls were renovated. The population of Tanagra is believed to have fled The renovation work included a variety of repairs, the to the fortification of Aghios Constantinos (Bintliff previously numerous openings, mainly gates, were now 2007, 664–665). The lack of investigations at this site walled up (Iader, Emona, Fig. 2.248), the wal s were prevents us from drawing any conclusions with regard raised and made thicker (Pola, Scodra, Lissus, Vinica), to the migration, but we can presume a fate similar to towers added (Forum Julii, Asseria, Salona, Serdica, that described for the cities above. Mesembria, Scodra, Thebes). The work was carried out The comparison of these short-distance population at different times, from the second half of the 3rd century shifts clearly shows the greater endangerment of the onwards, and in all three groups of cities. western part that can largely be attributed to barbarian New towers were built into the existing city wal s that incursions across the limes towards Italy. In the eastern were of a shape better suited to new weapons. Alongside part, such shifts are sooner an exception than a rule, the old round and rectangular towers, fan- and U-shaped which can be associated with a lower degree of threat, forms (Hissar; Fig. 2.249) appeared in the late 3rd and but also greater imperial support. early 4th century, joined by particularly characteristic 151 Slavko CIGLENEČKI Fig. 2.249: Hissar. City wal s with a fan-shaped tower in the corner (2012). Fig. 2.250: Split, Palace of Diocletian. Propugnaculum at the north entrance gates (Porta Aurea) (2017). triangular and pentagonal towers towards the end of the adjacent areas. It would appear that the cities thus forti- 5th and the 6th century. Entrances were heavily reinforced fied also served as temporary camps for large army units; and protected with different towers, with pentagonal several cities revealed army barracks on the interior towers most common in these spots in the 6th century. side of the wal s, but they have only rarely been more Frequently, entrances were also added propugnacula in extensively investigated (cf. Hissar, Tzoides, Bargala). the interior (Keszthely-Fenékpuszta, Split ( Fig. 2.250), The defences of some cities, particularly in places Iatrus, Dinogetia) or the exterior (Bargala). of easier access, included deep ditches (cf. Caričin grad, Particularly common and well-known are the Zaldapa, Vinica, Taor, Bargala) and earthwork ramparts defensive wall reinforcements that the eastern cities (cf. Keszthely-Fénekpuszta). made in the 6th century. As only a considerable number After its introduction in Constantinople, pro- of inhabitants was able to defend the extensive wal s, teichisma as an important element of the Byzantine these reinforcements indicate an influx of people from defensive technique quickly spread across the Byzantine 152 2. CITIES Fig. 2.251: Asseria. Late Antique tower with a proteichisma (2004). Fig. 2.252: Caričin grad. Episcopal basilica (2012). Empire from the 5th century onwards (Ovčarov 1973; CHURCH ARCHITECTURE Lawrence 1983, 185–186). It was frequent in the eastern part (cf. Hissar, Bargala, Vinica, Philippi), whereas in Construction of churches flourished in the cities of the west it only made its appearance after the Justin- all three groups. It lasted from the first church buildings ian reconquista (Ciglenečki 2003a, 269–272). It has in Aquileia in the 310s, to the slightly more numerous been established in Brixia, Asseria ( Fig. 2.251), Iader, churches in the second half of the 4th century and the it is clearly identifiable in Forum Iulii and presumed in intensive construction in the late 4th and early 5th cen- Teurnia. It is an element of defence often ignored as it turies. The development of church architecture was then was long unknown in the west. interrupted across wide areas of the western part and 153 Slavko CIGLENEČKI Fig. 2.253: Stobi. Semicircular square in front of the episcopal basilica (2013). there are no known churches built in the continental the Quarter of St Theodor persisted into the second half part after the mid-5th century with the exception of of the 5th century, as did the domus in Stobi that were Teurnia. Church architecture is all the more present in largely already modified. Such houses were very rare northern Italy, along the eastern Adriatic coast and in in the 6th century and only documented in some major the whole eastern part, both in the group of cities with centres (e.g. Philippopolis, Histria, Thessaloniki, Athens) continuity and those newly-founded in Late Antiquity. and on the islands. The Palace of the Giants on Athenian Here, the peak of church building comes in the second agora was erected in the first half of the 5th century and half of the 5th and even more so the 6th century, under converted into a monastery already in the first half of Justinian. These churches are evidence that the techni- the 6th century. The last known peristyle houses in Argos cal know-how and great skil s of a variety of masters were presumably constructed around 530 (Sodini 1984, continued into the second half of the 6th century. The 354–356; Liebeschuetz 2001, 370). In addition to the episcopal sees, composed of a complex of basilicas with rare examples of town houses in the 6th century, several baptistery and episcopium, are sometimes associated episcopal palaces were built at this time that show a high with small squares that become the new city centre (cf. quality of construction and rich decoration (especial y Caričin grad ( Fig. 2.252), Stobi ( Fig. 2.253), Vinica). In Parentium ( Fig. 2.255), Salona, Caričin grad, Carevec, addition to the main church buildings, there is a variety Byllis, Stobi, Heracleia Lyncestis, Bargala, presumed of martyrial and cemeterial churches in the suburbs that episcopia in Vinica, Konjuh and Thebes). are surrounded by residential buildings. Public baths were predominantly abandoned after the 4th century, though some were still constructed in the 6th century (e.g. Mesembria, Fig. 2.256), Byllis, URBANISM AND HOUSING Scupi, Bargala). Also in the 6th century, the construc- tion of small private baths has been recorded in some The features of urban transformation can best be cities (Athens, Buthroton, settlement in Sirmium’s hip- observed in the large cities with continuity. Alongside a podrome). proliferation of church architecture, a prominent feature Degradation of housing began in the second half of is also the numerous high-quality and even luxury domus the 4th century and escalated in later times. Subdivision- that witnessed a revival after the crisis in the 3rd century ing proliferated, with the large houses incorporating a and continued even into the first half of the 5th century. wide array of simple dwellings. The latter either reused Many of these houses were built anew, though often the the shell of the old house or were built completely anew. already existing houses were enlarged and embellished, The interiors of the partial y dilapidated houses suffered added new apsidal hal s (e.g. Aquileia, Fig. 2.254). After improvised interventions, subdivisioning into smal er the mid-5th century, luxury domus, episcopal palaces and units and changes of function (artisanal activities, ru- individual administrative or military edifices were stil ralisation). Residential quarters began loosing their being constructed, but this was mainly limited to major orthogonal layout, insulae were no longer maintained centres, coastal areas and islands. In Pola, the domus in and gradual y disintegrated. A particular feature is the 154 2. CITIES Fig. 2.254: Aquileia. Late Antique apsidal hall (1979). Fig. 2.255: Parentium. Ground floor rooms in the episcopium (2010). Fig. 2.256: Mesembria. Baths from the 6th century (2012). 155 Slavko CIGLENEČKI compaction of houses and economic facilities into smaller observed for the 4th and 5th centuries (Mikulčić 1982b, units separated from one another with empty spaces – a 536–537; Gerasimovska 2012, 110), at Gortyn on the feature known in literature as ‘island city’ or città a isole. remote Crete only in the second half of the 6th century These residential units were frequently associated with (Zanini 2009, 118–119). small groups of intramural burials. The building remains With the alterations presented above, the Late inside former theatres and hippodromes may be seen as a Antique cities lost the characteristics of the classic variant of such compaction, as has already been presented cities though retaining the earlier urban layout. Their in detail in the group of cities with continuity (see Chapter largely degraded interiors took a variety of forms with 2.4). Refuse began piling up in the empty spaces of the an emphasis on church buildings accompanied by rare abandoned city areas, in some cases these empty spaces high-quality constructions primarily intended for the were also used for agricultural purposes. This reveals a ecclesiastical and administrative apparatus. Many cities frequently ful y haphazard transformation. had army barracks along the city wal s. Predominantly, The construction technique used to build the new however, the city interiors hosted a myriad of simple dwellings was of a low quality, often completely impro- dwellings that partial y reused earlier buildings and vised. Drystone construction predominated, using clay elsewhere created a new, but modest and unpreten- and only rarely mortar as bonding material. Spoliation of tious architecture. earlier buildings was common. There was an increasing use of timber, most frequently in combination with dry- stone, some cities also reveal semi-sunken houses inside 2.6.3. CHARACTERISTICS OF GROUPS earlier buildings. However, the late dwellings built into OF LATE ANTIQUE CITIES AND earlier ones are not always of inferior quality and purely DIFFERENCES BETWEEN THEM utilitarian; in Scupi, for example, clay-bonded wal s were unearthed that were embellished with alternating CITIES ABANDONED courses of stone and brick, i.e. the rustic opus mixtum. BEFORE THE END OF LATE ANTIQUITY In Teurnia, one of the rare continental cities with continuity, several buildings in the old city area (pri- The cities of this group largely respected the tradi- marily the forum) were modified and new buildings tions of Roman urbanism and architecture. The note- constructed within them. At the same time, simple two- worthy changes include housing degradation, in some or three-room masonry buildings were put up in the places subdivisioning with dwellings of inferior quality previously uninhabited part of the plateau such as were built into earlier buildings, partial abandonment of characteristic of fortified settlements. It is an example public areas and minor reparations of city wal s. From of the ‘urbanism’ of small cities with the combination of the late 4th century onwards, a small number of cities modified earlier buildings and new constructions that is hosted the first Christian buildings, but these were soon devoid of a pre-conceived plan and rather incorporates abandoned or destroyed. buildings without a detectable order. In the western part of the area under discussion, A similarly haphazard construction can also be abandonment of cities occurred at different times and observed in the intensely populated, but much larger depended on local factors. There are certain differences newly-founded city of Zaldapa, where small modest between cities in the eastern Alps and those in Pannonia, buildings no longer display the regular layout of the with the natural y protected locations away from major former cities. communications most frequently allowing a prolonged Streets were partly maintained, in places en- existence. The more exposed and unfortified small cities croached on by different new buildings. A fairly isolated along the fringes of the Pannonian Plain, such as Mur- example is Scupi, where research has show regular sel a (Szőnyi 2004), Sal a (Redő 2003, 211), Flavia Solva repairs of the cardo to the late 6th century. (Hudeczek 2002, 210–211), Neviodunum (Djurić, B. The process of urban fabric transformation pro- 1976, 566; Lovenjak 2003, 96) and Andautonia (Nemeth- ceeded at a different pace in different areas and cities. Ehrlich, Kušan Špalj 2003, 125), already ceased to exist These changes are hardly noticeable in the cities aban- towards the end of the 4th century. Significantly, these doned before the end of Late Antiquity, while they sig- cities revealed no traces of Early Christian architecture. nified a fundamental change in the appearance of cities In the much less exposed western fringes, habitation with continuity. The influx of people from the country- remains and large church buildings in Iulium Carni- side has been observed in most cities and ruralisation cum and Aguntum date even after the mid-5th century was corroborated by the numerous economic facilities, (Corazza et al. 2001, 245–247; Sossau 2018). such as oil and wine presses, mil s, pens for domestic The most severely afflicted were the Norican and animals, as well as the presence of agricultural tools, in Pannonian cities located on the major roads towards city centres (cf. Suić 1976a, 248; Popović, V. 1982, 565). Italy, none of which (with the exception of Teurnia and In the more endangered Stobi, this has already been Siscia) show clear indications of urban continuity after 156 2. CITIES Fig. 2.257: Emona. Ecclesiastical centre from the end of the 4th and beginning of the 5th centuries (2011). the mid-5th century. This il ustrates the great impact of before drawing conclusions as to their continuity to the incursions from behind the limes to areas immedi- the late 6th century. ately before the gates of Italy. A great majority of cities was gradual y abandoned during the first half of the 5th century, whereas the time of CITIES WITH CONTINUITY the Hun raids in the middle of the 5th century represents the upper limit of existence for the cities that survived in The most numerous and also the best-known group a limited capacity. This period of abandonment is clearly comprises the cities that kept some of the traditional mirrored in the destroyed Early Christian churches, achievements of classic urbanism while at the same which indirectly indicate the date at which most cities time undergoing gradual change. These changes are were abandoned. In Emona ( Fig. 2.257), for instance, the predominantly visible in the disintegration and rurali- ecclesiastical centre was presumably constructed in the sation of the urban fabric, which became particularly 420s and abandoned even before the mid-5th century, intense after the mid-5th century. There were also new possibly already in the 430s (Kos 1983, 102–103). The constructions, as is clear from several buildings reliably shift from cities to nearby hil top settlements is also dated to this time, as well as from the two inscriptions corroborated by the hoard of large bronze Christograms from Serdica and Heracleia that mention the construc- found on the natural y well-protected settlement on Vi- tion of an aqueduct and a fountain in the second half pota above Pečovnik, which could only be brought there of the 6th century. by the refugees fleeing the nearby Celeia (Ciglenečki, This group includes cities for which continuity is Modrijan 2020). not without reservations, as they were largely abandoned Literary sources indicate that some cities persisted in the second half of the 5th and reactivated in the 6th into the second half of the 5th and the 6th century, but century. An example in Italy is Opitergium (Oderzo), this has as yet not been archaeological y confirmed. A which lost in significance in Late Antiquity, but was typical example is the archaeological y poorly-known fortified anew under the Byzantine rule (Brogiolo 2011a, city of Bassianae, which is mentioned in as many as four 105). A similar observation holds true of Singidunum, different Late Antique sources. It would appear that which Procopius mentions as a city of great importance, these texts often provide pretentious accounts of cities while archaeological remains are scarce. After destruc- that only existed as the ruins of former buildings. The tion by the Huns, Naissus was completely desolate for a bishops mentioned in these texts also moved episcopal while with insignificant archaeological traces, though it sees to safer locations, but retained their old titles. is still considered a city in literary sources (Milinković Such incongruities between literary and archaeological 2015, 136–137). The difficulty in such interpretations evidence show that further investigations of the urban is twofold; on the one hand, we cannot be certain what structures and reliable material evidence is needed the ancient authors had in mind when writing of a city 157 Slavko CIGLENEČKI Fig. 2.258: Athens. Post-Herulian wal s under the acropolis (2011). and, on the other hand, these cities are currently built-up The final major reductions occured in the 6th cen- and their research insufficient to reliably prove (non-) tury, under Justinian, when small parts of some cities existence in this period. Even in the case of Viminacium, were protected with massive wal s. Reliably dated wal s which was not built up in later times, we can observe that with characteristic towers from this period are known the city was deserted and presumably continued (?) in in Aquileia, Tergeste, Philippopolis, Durostorum, the modest settlement on the Svetinja peninsula, which Amphipolis and Heracleia/Perinthus. This group may rather ranks among the abandoned cities. also include Ulpiana, even though it did not witness a An important change in the group of cities with proper reduction. Under Justinian, a new fortified city continuity is the reduction of the urban area. Alongside was constructed here that was located next to the larger, the above-mentioned shifts of population to other loca- earlier one and was named after the emperor as Iustini- tions, reduction is a reliable indication of the radical ana Secunda. In spite of the new construction, several interventions in the urban fabric. It is the consequence buildings remained in use in the old city. of the dangers cities faced and the gradual depopulation The defensive wal s of the reduced cities incorpo- (overview in Kirilov 2007; Sodini 2007, 320). rated large public buildings of an earlier date. An exam- Reduction occurred at different times in different ple of this is the theatre in Sparta ( Fig. 2.260) that was areas. First signs are already observable in the second half incorporated in the defences of the acropolis (Lawrence of the 3rd century. In Stobi, it came about presumably due 1983, 180; Saradi 2006, 323). to floods in the part along the river. In Athens ( Fig. 2.258), There are also examples of cities where insufficient the Herulian sack of 267 led to the construction of new research does not allow us to either reliably confirm city wal s that greatly reduced the urban area. Histria was (Lychnidos) or more precisely date (Pharia, Pautalia, also reduced in the second half of the 3rd century. Gortyn) the reduction. The next prominent wave of urban reduction came Not much is known of the habitation remains in the late 4th and early 5th centuries. It is most apparent outside the reduced cities. These remains were certainly in the Albanian cities of Scodra and Lissus, as well as in less prestigious and include modest dwellings and work- the Greek Dion, Thespiae and Corinth. shops (in Aquileia, Nicopolis in Epirus, Amphipolis, Another wave dates to the second half and espe- Athens, Thessalian Thebes and others). In times of cial y the late 5th century. In this time, Sirmium shrunk danger, the people living in the surroundings sought to the better protected part at the River Sava, although shelter in the fortified area, which could not even be there is no evidence of the city wal s that would de- successful y defended without these people. Particularly limit the reduced extent of the city. Defensive wal s that for the eastern and southern parts of the area under dis- contracted the urban area were reliably established in cussion, there appears to have been intensive life outside Nicopolis in Epirus ( Fig. 2.259), Dyrrachium, Scupi, the reduced urban area (for example the lavish domus in Heracleia Lyncestis and Deultum. Philippopolis, Nicopolis in Epirus, Thessalian Thebes, 158 2. CITIES Fig. 2.259: Nicopolis in Epirus. Late Antique wal s with towers (2002). Fig. 2.260: Sparta. A theatre included in the reduced urban area (2002). Athens), which suggests there was less need for taking In Serdica, the city was substantial y enlarged in the 4th shelter within protected areas as in the western parts. century, but soon reduced again to its former extent. There are also rare cases when the opposite has The enlargement of Novae was of a longer duration; the been observed, namely an enlargement of the walled enlarged area first hosted refugees and later gradual y urban area. The city of Verona, for example, was enlarged became permanently inhabited. under the Ostrogothic administration and incorporated Predominantly in the 6th century, in some places an amphitheatre in the south and a palace on the hill on possibly even slightly earlier, we can observe a con- the other bank of the Adige. In Brixia as wel , a small centration of modest dwellings within theatres and fortified part was added to the earlier walled area. The hippodromes. The best investigated example of this is wal s in Tridentum now included an amphitheatre that in Heracleia Lyncestis, which appears to have hosted was previously located outside the city, as well as the a small organised community or ‘micro settlement’ area of Piedicastello on the opposite bank of the river. (Janakievski 2001). Also well investigated is such a 159 Slavko CIGLENEČKI community in another Macedonian city, in Scupi, Cities – fortresses where two construction phases have been discerned. from the late 3rd and early 4th centuries The 6th-century buildings in the hippodrome of Sir- mium also represent a residential unit that included in Cities – fortresses were set anew in the late 3rd and its centre a large building with baths. Less is known of early 4th centuries. They were established on the incentive such complexes in Stobi and Ohrid. Further examples and with the help of imperial authorities primarily in the are also mentioned in Messenia (Saradi 2006, 321). area of the limes and its hinterland, individual y also at These appear to be the last attempts at organised important strategic locations further inland. To a certain habitation in cities, which have parallels elsewhere in degree, they still have a regular layout, they are sited on the Empire, for example in Carthage, Leptis Magna slightly raised ground or at a river confluence and show and Valencia (Quiroga 2016, 80). It is conceivable that more commonalities with earlier cities in comparison it was the strong and partial y surviving wal s of these with the cities established later. They boast strong city former buildings that offered a protected environment wal s with fan- or U-shaped towers. Their interiors lack for human habitations while at the same time enabling monumental public buildings and the organisation of effective defence in times of danger. space no longer traces an orthogonal street grid. Their character has long been known and they are frequently discussed in literature (overview in Rizos NEWLY-FOUNDED CITIES 2017b). The inscription from Tropaeum Traiani, dated AD 314, clearly relates a combination of a fortress and This is the most complex and problematic group a civilian settlement. They are most common in the of cities that brought about, because of the late date provinces Lower Moesia and Scythia. Their limited size and difficult circumstances, a complete change in the raises the question of whether part of the inhabitants concept of a city. Selecting the criteria that justify the lived in the surrounding area and only sought shelter identification of the last urban settlements was hardest behind city wal s in times of danger. Efthyimios Rizos for this very group. As mentioned in the introduction, believes the main reason for establishing these cities there is no general consensus on the criteria and the was to serve as a support to the authority and defence, urban character has to be assessed for each settlement rather that a rich backdrop for the imperial ceremonial separately. (Rizos 2017b, 20, 25). The newly-formed cities range from those con- This subgroup may also include the inner fortifi- firmed by epigraphic evidence to large central set- cations in Pannonia, which have been discussed under tlements with the last signs of an urban character. different terms – either as military posts, cities, supply Newly-founded cities differ from those of other groups centres and so forth (overview in Heinrich-Tamáska in function and appearance, consequently also interior 2011a; Rizos 2017b, 25–27). Their fortifications, structure. In recent times, these cities have received structure of buildings in the interior and size indicate much attention because they are a novum in the set- a combination of functions; as such, these cities are tlement pattern and also generate discussion on which characteristic representatives of the new concept of a settlements can be identified as cities. Late Antique city. In newly-founded cities, the classic city trans- Slobodan Ćurčić, who also notes that new cities formed into a fortified settlement with an ecclesi- in the Balkans already formed in the early 4th century, astical centre, a small army garrison (possibly only includes the enigmatic category of miniature cities such urban militia) and several functions that constitute a as Split and Romuliana in this subgroup (Ćurčić 2010, central settlement. Literature has already marked this 23), although these two cities attained the city status at process as a transition from polis to kastron (overview a later date. In the former imperial palace of Romuliana, in Dunn 1994), which corresponds with reality in a a strong civilian character coupled with a military pres- great measure. ence can only be felt towards the end of the 4th century. We should distinguish between two subgroups At that time, the fortress of Iatrus was transformed into of newly-founded cities that differ in the date of their a civilian-military settlement. Its architecture is similar beginning, the function, location and appearance. The to the contemporary dwellings in the classic cities. The first subgroup is earlier and consists of cities – fortresses, subgroup further includes Ságvár in Hungary, where the the other comprises the newly-founded cities of the 5th shape of the towers points to the establishment in the and 6th centuries. second half of the 4th century, as well as Bargala in North Macedonia ( Fig. 2.261), where the inscription dated to 371 reveals that the army camp was already transformed into an urban settlement. In North Macedonia, the fortress Dolenci (Kale) is another similar example of transformation from army camp to city. The site is not 160 2. CITIES Fig. 2.261: Bargala. Wal s and buildings in the north-eastern corner of the city (2013). sufficiently known and we must allow for the possibility literature, these suburbs are only rarely mentioned as that this transformation occurred at a slightly later date. parts of the newly-founded settlements and are poorly What is certain is that the churches intra and extra muros known. Geophysical investigations have revealed them point to the altered nature of the fortress. at Caričin grad, Romuliana, Louloudies and elsewhere. Even before these, field surveys and trial trenching have detected several such instances in North Macedonia: at Cities newly-founded Čučer (Davina), Taor and Vinica (Kale). More knowl- in the second half of the 5th and in the 6th century edge on the modest residential and economic architec- ture in these border areas will allow us to gain a more The discussion above revealed that, in spite of their complete picture of the cities. altered appearance, the cities from the early part of Late The final forms of newly-founded cities have all Antiquity were still clearly recognisable and their status the defensive elements used in Late Antiquity, namely corroborated by epigraphic evidence. Much more chal- wal s with towers, proteichisma, ditches and earthwork lenging is the identification of cities newly-founded after ramparts. Their interiors are dominated by the church the mid-5th and especial y in the 6th century. complex or a large church with a baptistery, while there A number of fortified and natural y well-protected are often additional churches in other parts of the cities settlements was established in this period that hosted the and outside the city wal s. They hold extremely rare endangered civilian population. Some of these were large public buildings, but a multitude of simple dwel ings, the and had residential, as well as modest administrative, size and layout of which indicates social stratification. military and especial y church buildings, all of which These cities were inhabited by large organised com- speaks of an urban character. These cities lack the attrib- munities with local autonomy who performed religious utes characteristic of the former cities and are primarily and administrative functions. Also present were small distinguished from the contemporary settlements in the units of the army or local militia. The population was vicinity by their size and dominant location. largely self-sufficient. Additional criteria for an iden- Only rare have been investigated well enough to tification as a city are the social differentiation of the allow us to verify the criteria listed above. The first cri- inhabitants, different crafts and commerce. Alongside terion is size. Most authors estimate that the settlements a central position within a region, the presence of the covering a surface of more than 2–3 ha represent major ecclesiastical apparatus appears to be another indication settlement centres that stand out from those measuring of a late urban character. 0.5–1.5 ha on average. Having said that, we should not Particularly numerous are the examples of such forget several smaller fortified settlements with suburbs cities in the most intensely investigated North Mace- or smaller settlements in their vicinity, the inhabitants donia. Ivan Mikulčić used passages in ancient texts of which occasional y sought shelter in the centres. In to term these agglomerations as oppida and oppidula, 161 Slavko CIGLENEČKI Fig. 2.262: Caričin grad. Lower city in the foreground, Upper city with the Acropolis in the background. View from the south (2012). which presumably replaced the abandoned Roman cit- establishing smaller cities and central settlements that ies and were seen as small Byzantine cities (Mikulčić could be more easily managed (Suić 1976a, 242; Bratož 2002, 269–271). He mentions citadels among their 2011a, 224–225). basic features, which were separate from the lower city In the Alpine area, such a city can be seen in Sa- as the presumed autonomous area of an army garrison. biona, though this is a problematic site in many respects Their interior also hosted churches and even remains of and one that lacks several basic identification elements. other public buildings. Some cities revealed traces of a Literary sources clearly mention it held an episcopal see, street grid and urban layout. These settlements played while investigations thus far yielded very scarce settle- a primary role in the defence of the area, they were the ment remains. In the mentions of bishops, this site is seats of the local administration and the last remains similar to frourio Meridio, located near the city of Aque of artisans and merchants at the end of Antiquity. They on the Danube and only known from literary sources. include Čučer (Davina), Vodno, Kalata near Kamenica It also received a bishop in the 6th century, but the site and others. The best example is Čučer (Davina), the itself has as yet not been reliably identified (Janković layout of which includes a city square, church complexes, 1981, 52). It would appear that cities of the 6th century communications and suburbs (Mikulčič 2002, 101). were no longer identified by an episcopal see as was the The fortress of Vodno is less likely ascribable to case for the 5th century (Dagron 1984, 8). this category, though it seems the very settlement that In Late Antiquity, there are other instances of Mikulčić used to define the appearance of such cities. episcopal sees transferred to better protected locations What seems problematic is the absence of a church on (cf. Berg 1985, 90; Bratož 2009, 39–43). These locations the acropolis; neither has a church been found in the have not been reliably recognised, which is additional lower part of the fortress. It is certainly a post of several evidence to look for them in the heavily transformed functions, but its heavy fortifications suggest a primar- centres of settlement closely comparable with large ily military function, possibly also serving as a refuge. fortified hil top settlements (presented below). Rajko The tendency for establishing new cities is particu- Bratož observes that ephemeral episcopal sees were larly pronounced in the 6th century. Literary sources not exceptional in the declining period of Antiquity reveal the request of the metropolitan of Salona for regardless of the fact that such sees were not permitted dividing the bishopric of Salona expressed on the syn- or hindered in the ecclesiastical legislation (Bratož 2009, ods in the 530s. This can be understood as a desire for 39). In Dalmatia, there was little success in locating the 162 2. CITIES Gradišče (Bašelj) Sv. Lovrenc Sv. Jakob Gradišče (Pivka pri Naklem) Štefanja gora Britof CARNIUM (Kranj) Puštal Mengeš Hom 0 10 km Fig. 2.263: Map of Late Antique sites surrounding Carnium. episcopal sees mentioned in the records of the synods Some of the elements known from Caričin grad – held in Salona, there were only theoretical attempts not primarily the square (forum?) and regularly distributed confirmed by archaeological investigations (cf. Škegro buildings on the Acropolis and in the Lower city – 2008; Periša 2009; Bratož 2009, 27–43). This only cor- are also detectable in the poorly investigated Čučer roborates the altered appearance of the last episcopal (Davina). Individual urbanistic elements, such as an sees, which were devoid of external characteristics that aqueduct, sewage system, main axis, important edifices would enable at least a rough identification. and suburbs, are also visible in Vinica and Taor. The investigations of Caričin grad (presumed Ius- Sufficiently convincing are the elements indicating tiniana Prima), a city of exceptional status, shed light an urban nature of Gradina on Jelica. In addition to its on the newly-founded cities and indirectly also allow us size, five churches and three habitation areas separated to identify other similar religious-administrative urban by wal s, these elements comprise several large and well- formations. In its concept, Caričin grad was divided into built edifices, the absence of ruralisation and disintegra- functional y clearly separate parts (Acropolis, Upper tion of the settlement, as well as the paucity of poorly- and Lower city, different suburbs; Fig. 2.262) and thus constructed buildings (Milinković 2015, 188–189). certainly represents a model imperial design. The city in Alongside the above-mentioned newly-formed its initial phase was a capital with a strong ecclesiastical settlements of a fairly readily identifiable urban char- and administrative apparatus, but it soon ruralised in acter, there is a series of other settlements that could the face of the massive influx of people from the coun- also be ascribed an urban character considering the tryside, with the small finds from the second half of the lower standards of cities in Late Antiquity. They are 6th and the early 7th century showing an increasingly of a considerable size, but only show individual urban autarchic character (Ivanišević 2017, 103–108). Caričin elements. An example in the western part is the forti- grad is the only city newly-founded in the late period to fied settlement of Carnium (Kranj) which undoubtedly exhibit substantial remains of classic urbanism, though had a central role, but the continuous later habitation of its exceptional status paradoxical y makes it an atypical the spot hinders an assessment of its urban character. representative of this subgroup of cities. The latter makes Not much is known of the organisation of space in the it impossible to compare it with other urban formations interior, though the settlement on an elongated plateau of a more modest design and a character that is more seems to have been organised similarly as the modern- rural from the onset. day city, i.e. along a main longitudinal communication 163 Slavko CIGLENEČKI Fig. 2.264: Bosar. Unfortified settlement from the 6th century from the south-east (2016). with transverse streets leading off of it, with a church and times, where archaeological traces with the exception of a large baptistery in the centre. The size and distribution churches are all but imperceptible. of the numerous smaller fortifications along the edge In Bolgaria, the city of Shoumen ( Fig. 2.265) also of the plain that holds the city indicate that Carnium corresponds with the notion of a Late Antique city, hold- was their administrative and ecclesiastical centre. The ing a large number of densely spaced buildings and heav- smaller, roughly 1 ha large settlements are located at ily reinforced defensive wal s. A multitude of buildings, Gradišče and Sv. Lovrenc above Bašelj, Sv. Jakob above several churches and strong defensive elements point to Potoče, Puštal above Trnje, Hom above Sora, Štefanja a semi-urban character (a miniature Late Antique ‘city’ gora and Gradišče above Pivka pri Naklem ( Fig. 2.263). as the author termed it) of the Late Antique settlement None has as yet revealed a church, although they could on Golemanovo Kale, which is reliably attributable to the be presumed to have existed in the larger settlements. Justinian period (Ćurčić 2010, 215); it is discussed in the Also dating to Late Antiquity are the poorly known group of fortified settlements because of its small size. lowland sites located at nearby Britof and Mengeš (Mi- In interpreting newly-founded cities, some authors lavec 2020, 160). rely on the well-investigated Nicopolis ad Istrum (cf. Standing out among the presented newly-founded Saradi 2006, 467; Whittow 2007, 387; Snively 2009, 41). cities are Bosar on the island of Krk ( Fig. 2.264) and Only administrative and church buildings are known Blagaj in the Japra Val ey, which are the only known here and it would appear that it is an exception rather large unfortified settlements that display some urban than a model Late Antique city. Given the absence of elements. They indicate the possibility that such set- housing in the interior, it is also not certain whether we tlements existed in other areas as wel , primarily in know enough of the city. areas removed from the main communications. Their The description of the as yet unidentified episcopal existence is thus far only indicated by large church see on the inscription from Izbičanj is in many respects buildings. An example of the latter is the large and similar to the description of the estate and vil a of Sidonius beautiful y decorated basilica in Zenica, where Đuro Apollinaris in Avitacum, as well as the description that Basler presumed a central settlement in the place of Venantius Fortunatus provides of the palace of Nicetius the earlier municipium Bistue (Basler 1972, 126–127). along the River Mosel (Egger 1940; McKay 1998, 186). The Similar observations have been made for most Late high expectations created by the poetical y embellished Antique settlements in areas intensely populated in later descriptions are certainly to blame that their remains 164 2. CITIES Fig. 2.265: Shoumen. Entrance and wal s with towers in the north-western part of the city (2012). have not yet been archaeological y detected. At Izbičanj, different agglomerations, for which only systematic in- we may therefore only expect a large fortified settlement vestigations would allow the final urban characteristics with parallels in Sabiona, Kirchbichl, Louloudies and to be identified. frourio Meridio mentioned in literary sources. For the Peloponnese, Chronicle of Monemvasia Final y, we must also mention a large number of reports that the inhabitants migrated to Italy, Sicily, ‘cities’ established in the second half of the 6th century. islands along the coast and the isolated refugium of These are the last, utterly degraded and impoverished Monemvasia, on which a bishopric was established urban agglomerations mainly located on the coast and (Whitby 2000, 727; Veikou 2012, 184–186). the islands that offered better defensive possibilities and In the western part of the Empire as wel , the civitates a maritime supply. Many were also established on the that the Anonymous Geographer of Ravenna mentions in Adriatic coast, but the lack of research does not allow the continental part are frequently small fortified settle- us to reliably attribute them to this group (Suić 1976a, ments. It would appear that they can be perceived as min- 235–238, 241–242; Katić 2003, 524–526; Saradi 2006, iature cities; this is the term I used for the Late Antique 464–470); the already mentioned examples are Capris, fortified hil top settlements in the eastern Alpine area that Sipar and Lisina, but possibly also Dubrovnik, Budva had an ecclesiastical centre and densely spaced housing and others that have revealed insufficient evidence for inside strong fortification wal s (Ciglenečki 1987a, 114). a detailed identification. In the late 6th or early 7th century, a large part of the area under discussion witnessed a rapid decline of Many of the civitates mentioned by the late authors cities, with partial continuity of urban structures only are not cities proper, as they are fortified settlements observable in some coastal cities, in Ravenna, Constan- minor in size and significance. The use of the term tinople and Thessaloniki, as well as in several cities of civitates for Late Antiquity thus suggests a myriad of northern Italy. 165 3. SETTLEMENT IN THE COUNTRYSIDE 3.1 MAJOR SETTLEMENT CHANGES The difficulties and open questions associated IN THE COUNTRYSIDE with the research of the Late Antique countryside are briefly presented in the introduction (see Chapter 1). Of In the area under discussion, the the Late Antique these, we should again emphasise the endangerment of countryside settlement is one of great complexity, simi- settlement remains in lowland areas, which are highly larly to that of cities but with even more radical changes, as exposed to damage or destruction through agricultural already noted in the introduction (see Chapter 1). Most of or construction activities. Modern investigation tech- the settlement forms from the Roman period (vil as, vici, niques (primarily geophysical investigation and aerial roadside stations and others) continued into the initial photography) have been used in many areas, but the part of Late Antiquity; most persisted to the first half of poor preservation of sites often prevents a reliable iden- the 4th, many to the second half of the 4th century and tification of the remains as a vil a, settlement, farmstead, only rare ones survived to a later time. A substantial break individual farm buildings or other. came in the middle third of the 5th century, when most The situation is different for the newly-founded of the unfortified lowland settlements were abandoned. fortifications, located on natural y protected elevations People ever more frequently sought shelter in remote and often with well-preserved remains, which make areas and natural y well-protected locations. At first, these them more readily identifiable and consequently better locations only served as refuges in the face of unstable known. Here, field surveys and the use of new tech- circumstances, but many later turned into permanent niques (primarily LiDAR or surface laser scanning and settlements. This process of settlement shifts began to be aerial photography) frequently offer an insight into the felt in the last third of the 3rd, it intensified in the second structure of the sites even without major interventions half of the 4th and lasted to the end of the 6th century. into the subsoil. The vast body of countryside settlements can rea- sonably be divided into two groups: 3.2 SETTLEMENTS - Settlements with continuity from the Roman period WITH CONTINUITY ( Figs. 3.1, 3.26). Most survived to the mid-5th cen- tury, in rare cases also beyond. Contemporaneously with last peak of cities in - Settlements established in Late Antiquity. These com- the 4th century, prosperity is also observable in the prise rare unfortified settlements and predominant countryside. The area under discussion also witnessed fortifications, the latter representing the most densification of estates and consequently fewer but larger prominent habitation form. vil as (Thomas 1964, 389; Henning 1987, 34; Dintchev 1997, 125; Mulvin 2002, 23). For a long time, not much was known on the set- This chapter presents a selection of characteristic tlement of the countryside. It was discussed in several settlements that already existed in the Roman period and regional overviews, but completely absent in synthetic continued to develop in Late Antiquity. The selection works with the exception of Roman vil as. Many authors includes settlements that changed their function at the noted this shortcoming. The knowledge only real y began end of the Roman period and transformed into a differ- advancing from the 1970s onwards through the numer- ent type, i.e. one of an economic or even religious nature. ous modern rescue excavations (primarily in advance of Roman vil as are the best known form of countryside motorway construction) and intensive field surveys in the settlement, discussed in regional (see Thomas 1964; Vasić lowland and in remote areas primarily through extensive 1970; Dintchev 1997; Begović, Schrunk 2001; Marensi surveys that revealed a multitude of fortifications. 2001; Busuladžić 2011; Kastler et al. 2017) and wider over- 167 Slavko CIGLENEČKI 10 8 11 9 12 13 14 15 16 17 1 18 19 3 2 33 4 20 34 35 36 5 6 7 21 22 23 24 30 37 31 25 26 27 28 29 32 38 42 44 39 43 45 40 41 46 47 48 49 50 51 52 54 55 53 56 57 0 100 km 1 Castel Antico 15 Slovenska Bistrica 29 Vižula 43 Crkvina near Otres 2 Castelletto di Brenzone 16 Blagovica 30 Poje in Njivice 44 Bisko near Knin 3 Nuvolento near Pieve 17 Trojane (Atrans) 31 Cickini near Sršiči 45 Crkvina near Halapić 4 Manerba del Garda 18 Grobelce 32 Martinščica 46 Grušine 5 Borgo Regio in Desenzano 19 Nauportus (Vrhnika) 33 Drenje near Zaprešič 47 Strupnić near Livno 6 S. Pietro in Cariano 20 Ulaka 34 Donja Glavnica 48 Proložac Donji 7 Sovizzo 21 Školarice 35 Virovitica, Kiškorija South 49 Lovrečina Bay 8 Loig near Salzburg 22 Predloka 36 Kövágószőlős 50 Novo Selo Bunje 9 Immurium (Moosham) 23 Červar, Porat 37 Ivandvor 51 Proboj near Ljubuški 10 Bruckneudorf 24 Vrsar 38 Ograja in Putovići 52 Sv. Klement near Hvar 11 Tatárszál áson 25 Barbariga 39 Debelo Brdo in Sarajevo 53 Ubli 12 Csúcshegy 26 Dragonera Jug 40 Ilidža in Sarajevo 54 Višići near Čapljina 13 Balácapuszta 27 Veliki Brijun 41 Mušići near Višegrad 55 Dračeva strana near Bileća 14 Stari trg (Colatio) 28 Stancija Peličeti 42 Muline 56 Panik near Bileća 57 Bihovo near Trebinje Fig. 3.1: Countryside settlements that continued from the Roman period, western part, discussed in the book. views (Henning 1987; Mulvin 2002). It is often difficult resettled. Without meticulous excavation, it is not easy to to identify the time when the continuity of Roman vil as distinguish between the two; the discussion thus includes ceases and the time when a completely new and (largely) the buildings for which archaeological evidence suggests different use begins. Therefore, the discussion below takes an uninterrupted existence in an unaltered function and into account the continuity both of the original building those that (possibly with a minor chronological hiatus) and of the spot of the original building that was later occur within the earlier buildings. 168 3. SETTLEMENT IN THE COUNTRYSIDE The term ‘vil a’ is used here for a wide formal array were highly standardised across the Empire and that the of countryside settlement and the term ‘ villa rustica’ transformation of vil as already began in the 3rd century. is only reserved for the sites clearly identified as such Initial y, changes are primarily visible in the function in publications. The book also discusses a few cases of of individual rooms, which became production units, poorly identifiable buildings or farmsteads, as wel as set- whereas from the 5th century onwards vil as show de- tlements such as vici and roadside stations. In addition crease of habitation and beginning of artisanal activities. to these, the countryside was dotted with a large number Gian Pietro Brogiolo dates the end of the vil as around of settlements of timber or semi-sunken buildings, but Lake Garda from the late 5th to the late 6th century, noting they have only rarely been examined and presented in a degradation of earlier buildings and subdivisioning with detail; some are known to have combined masonry and simple constructions, as wel as the vil a interiors used for timber architecture. Such remains have been coming burial (Brogiolo 1997, 300). to light in recent rescue excavations, which revealed individual small buildings and larger settlements con- Traces of a vil a from the 1st and 2nd centuries came structed of less durable material (cf. Jelinčić Vučković to light at Nuvolento near Pieve. The brief reports reveal 2015; Horvat et al. 2020a). the vil a was extensively renovated on several occasions An important subject with regard to vil as is their and torn down in the 4th century (De Franceschini 1998, fortification, which is a feature that investigations often in- 109–111). On top of levelled ruins, small rooms with dicated, but almost never confirmed. The wal s enclosing hearths were constructed of wood in a phase dated to the most farmhouses appear to be delimiting the habitation 5th/6th centuries (Brogiolo, Chavarría Arnau 2014, 234). and production areas rather than serving a defensive pur- A specific example is the vil a at Castel etto di Bren- pose. For Pannonia, Edith Thomas (1964, 389–390; 1980, zone ( Fig. 3.2), on the east coast of Lake Garda, where 312–317) presumes a general practice of vil a fortification one of the buildings revealed layers from the 6th and 7th already from Aurelian period onwards. It is believed that centuries (Brogiolo, Chavarría Arnau 2014, 234). These vil as also served as refuges for the farmers from the sur- show a radical change of the interior that only contained rounding areas. However, most of the detected wal s did traces of economic and artisanal activities; this is certainly not enclose the vil as in their entirety and, moreover, most a case of new use rather than continued habitation. vil as lie on natural y poorly protected locations (most The large and lavish vil a at Borgo Regio in Desen- frequently terraces and gentle slopes). Fortifications on zano was completely renovated in the first half of the 4th such locations only provided protection against animals century (De Franceschini 1998, 147–151). Architectur- and small robber bands; they would less likely be success- al y, the renovated vil a is marked by baths and a hall ful against barbarian incursions. with a triconch terminal, which had rich polychrome Research in Pannonia, the eastern Alps and the mosaic floors and wall paintings. The last repairs of the interior of the Balkan Peninsula has pointed to an exten- vil a date to the late 4th or early 5th century. sive abandonment of vil as and other forms of lowland Manerba del Garda yielded the remains of a small settlement already towards the end of the 4th century. vil a that persisted to the 4th/5th centuries (Brogiolo Only rare cases persisted into the first half of the 5th 1997, 307; De Franceschini 1998, 155–156). In Late century or beyond (Thomas 1964, 395–398; Henning Antiquity, a small apsed church dedicated to St Syrus 1987, 35; Dintchev 1997; Mulvin 2002, 42; Milinković was constructed on top of its ruins in the east part. 2008, 545; Begović Dvoržak et al. 2020). Most lowland settlements in Serbia came to an end in the late 4th or first half of the 5th century (Milinković 2007, 162), those in Bolgaria were presumably deserted until 450 (Dintchev Fig. 3.2: Castelleto di Brenzone. Plan 1997, 205). In Greece, most vil as were already aban- of the vil a (Brogiolo, Chavarría Arnau doned in the 4th century, while some were reoccupied 2014, Fig. 4). in the 6th century (Curta 2014, 39). 3.2.1 WESTERN PART ITALY The systematic work conducted in Italy over the re- cent decades has produced a wealth of data that il uminate the fate of vil as in Late Antiquity and partly even later (cf. Brogiolo, Chavarría Arnau 2014, 233). Researchers have observed that both the evolution of vil as and their end 169 Slavko CIGLENEČKI Two simple buildings, dated between the 4th and the 7th century, were constructed on the other side of the vil a. The vil a at S. Pietro in Cariano, already built in the 1st century, witnessed its last phase with primitive buildings at the transition from the 5th to the 6th century (De Franceschini 1998, 175–177). Part of a settlement with simple and predomi- nantly single-room houses came to light at Castel Antico ( Fig. 3.3), on the south side of Lake Idro (Brogiolo 2018, 137). The settlement was presumably inhabited from the 1st to the 5th/6th century and comprised two types of houses. The first was sunken houses with drystone wal s, the other masonry houses with tiled roofs. Only the economic part of the vil a at Ponte Lam- bro has thus far been investigated (Airoldi et al. 2016). After being abandoned around the mid-3rd century, some of the rooms were reinhabited in the 4th–6th cen- tury; timber buildings were constructed on top of the ruins and fitted with hearths and two water cisterns. The rectangular multi-room building at Sovizzo was erected in the 1st and remained in use to the 5th cen- tury (De Franceschini 1998, 221–222). In its last phase, the rooms were rearranged and changed in function by adding ironworking facilities. AUSTRIA The large palatial villa at Loig near Salzburg was Fig. 3.3: Castel Antico. Plan of the settlement (Brogiolo 2018, mostly excavated already two centuries ago, but it is the Fig. 3). recent investigations that provided sufficient evidence to interpret its function, the use of individual rooms and its gradual development from the 1st to the late 4th century (Ladstätter 2002, 325; Winckler 2012, 255; Gruber 2017). The vil a is one of the largest north of the Alps. It is divided into several parts that comprise the prestigious, economic and garden areas. The main part is the residence fitted with hypocausts and mosa- ics; it even incorporates a horreum. The vil a had four construction phases, the last two attributable to Late Antiquity. From the Severan to the Constantinian pe- riod, the vil a witnessed numerous modifications and repairs, as well as the instal ation of precious mosaics. It is at this time that it reached its maximum extent of 339 x 256 m, covering an 8.98 ha large surface. It is also a time of increased activity of the numerous workshops operating inside the vil a. The numismatic analysis, in- terestingly, revealed a diminished use of the central part of the vil a after the Severan period and a long hiatus, which has been linked with the instabilities brought about by Germanic incursions in the third quarter of the 3rd century. For the last, fourth phase dated from Constantine to the late 4th century, coins again show an even distribution across the whole of the vil a. The Fig. 3.4: Bruckneudorf (Parndorf). Plan of the vil a (Thomas last peak of prosperity is visible in the renovations of 1964, Fig. 99). individual rooms and the instal ation of mosaics. The 170 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.5: Bruckneudorf (Parndorf). Central residential building from the north-east (2009). to the 4th century. Spread across an estimated surface of 3 ha, it held masonry houses and the roadside station proper, as well as baths and a mithraeum. HUNGARY The vil as in Hungary are well-known and dis- cussed in numerous studies (cf. Thomas, 1964, Bíró 1974; Mulvin 2002). Of the latter, we should mention the pioneering work by Edith Thomas, who collected the vast body of evidence, identified the types of vil as and their function (Thomas 1964). She also considered Fig. 3.6: Bruckneudorf (Parndorf). Apsidal hall from the the Hungarian ‘inner fortresses’, which differ from the north (2009). Late Antique vil as in their impressive defences (not present in vil as) and a different function; in this book, vil a was then abandoned and stray finds indicate oc- the ‘inner fortresses’ are discussed in the chapter on casional habitation to the mid-5th century. newly-founded cities (see Chapter 2.5). The vil a complex at Bruckneudorf (Parndorf) ( Figs. 3.4−3.6) enjoys a long history of research (Thomas A large and luxuriously decorated vil a came to 1964, 177–192; Ladstätter 2002, 323; Zabehlicky 2011). light at Balácapuszta ( Fig. 3.7) (Thomas 1964, 73–107; A timber building under the residential part of the vil a Mulvin 2002, 75–76). It is a complex of fifteen edifices was already constructed in the 1st, while the first ma- enclosed with a wal . At its heart is the main residence sonry building there appeared in the 2nd century. This with an apsidal hal . Also identifiable are baths and a was followed by several minor additions. The vil a was horreum, while the rest are smaller living quarters and extensively renovated (presumably following an earth- economic facilities. The vil a and its perimeter wal were quake) after 355, when a large apsed hall was constructed mostly constructed in the 2nd and partial y renovated and rooms decorated with high-quality mosaics. After after the 3rd century, the latter focusing on the main this renovation, the vil a may be marked as a palace residential building. The vil a was abandoned in the late that housed a high-ranking military or administration 4th century, though there are modest and chronological y official. It persisted into the 5th century. Considering the undiagnostic habitation traces from later times. enormous granarium and the location near the limes, Only two buildings, dated from the 2nd to the 4th it presumably served as a supply base for the army sta- century, are known of the vil a at Csúcshegy, in the tioned at Carnuntum. vicinity of Aquincum (Thomas 1964, 216–227; Mulvin The roadside station of Immurium (Moosham) 2002, 77). One is a large rectangular residential build- has been extensively investigated, as was the associated ing with several heated rooms and a decorated apse, but settlement interpreted as a vicus (Gassner et al. 2002, without a peristyle. Next to it is a smal , incompletely 143–144; Groh, Lindinger 2008). The settlement was known building with an apse that presumably held a located at a major crossroads and inhabited from the 1st bathroom. 171 Slavko CIGLENEČKI Fig. 3.7: Balácapuzsta. Plan of the vil a (Thomas 1964, Fig. 40). Fig. 3.8: Tatárszál áson. Plan of a Late Roman building of the roadside station (from Kelemen 2014, Fig. 1). The building at Kővágószőlős, in the vicinity SLOVENIA of Pécs (Sopianae), had the appearance of a fortified farmhouse (Mulvin 2002, 88–89). The large rectangular Early excavations in Slovenia revealed numerous residential building with a peristyle had a pair of semi- Roman vil as, but their remains are for the most part circular towers on the front. Small baths leaned onto it poorly published and do not provide clear evidence in the east, in the west the building was associated with of continuity into Late Antiquity. In addition to these, a porticus. It was presumably already constructed in the recent investigations in the countryside brought to light 2nd century, renovated in the 3rd and survived into the small settlements and roadside stations that have been 4th century. A chapel with three apses and a mausoleum more extensively researched and more precisely dated inside a crypt was erected in the immediate vicinity of (cf. Horvat et al. 2020a). the main building. The Late Roman building at Tatárszálláson The vicus in Nauportus (Vrhnika) developed on ( Fig. 3.8), along the Roman road between Aquincum and the road from Emona to Aquileia and measured roughly Brigetio, formed part of a small roadside station (Kele- 600 m in length and 200 m in width. It was continual y men 2014). The spot was initial y occupied by a timber inhabited from the 1st century onwards. In the 4th cen- building that burnt down in 260/270. In the first third of tury, the settlement was joined by a pentagonal army the 4th century, a 20.6 x 12.4 m large masonry building fort positioned at the beginning of the Claustra Alpium was erected, composed of two habitation rooms and an Iuliarum barrier system (Horvat 2020). The systematic L-shaped court. The building was modified in the Valen- investigations inside the vicus have revealed building re- tinian period, added new rooms and fitted with a late form mains in three areas. The recovered coins and other smal of central heating. The coins and other small finds reliably finds prove that the settlement and the fort existed to the date the masonry building up to the early 5th century. Its late 4th, in a small measure even the early 5th century. end came in a fire shortly afterwards. The stamped bricks The rescue investigations at Blagovica ( Fig. 3.9) indicate that the building was erected and maintained by unearthed parts of a settlement that measured roughly the army and that it functioned as mutatio. 300 m in length (Plestenjak 2020). Large masonry buildings lined the major Roman road from Emona 172 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.9: Blagovica. Plan of the 4th (last) phase of the settlement (Plestenjak 2020, Fig. 12). to Celeia and were used, with several modifications, sumably used for cult purposes. Investigations did not from the 1st to the second half of the 4th century. The establish the main housing area, but did reveal part of settlement is presumably the roadside station of Ad a large building of an unidentified economic function. Publicanos. The strong wal s enclosing the complex indicate the Slightly further east, the remains of a customs post were partial y investigated on the pass of Trojane (Atrans), the border between Italy and Noricum, which was inhabited throughout the 4th century (Železnikar, Visočnik 2020). Traces of a large settlement and postal station were investigated in Stari trg near Slovenj Gradec (Colatio), on the Roman road from Celeia to Virunum, which were in use to the early 5th century (Djura Jelenko 2020). Six buildings of the settlement in Slovenska Bis- trica lined the Roman road from Celeia to Poetovio at the spot where a minor road forked off to lead towards the Pohorje Hil s. The initial y timber buildings from the 1st and 2nd centuries were replaced in the 3rd and 4th centuries by large masonry ones. The size and stru c ture suggest that the settlement was associated with transport and storage of goods (Strmčnik Gulič 2020). The remains of two vil as, lying 1.5 km apart, were partial y investigated in the area of Šmarje pri Jelšah (Lorger 1936; Thomas 1964, 344–351; Mulvin 2002, 102–103). The larger of the two was located at the Grobelce site ( Fig. 3.10) and ranks, with a surface of ca 3.5 ha, among large vil a complexes dating between the 1st and 4th century. It revealed a small dwelling, richly Fig. 3.10: Grobelce near Šmarje pri Jelšah. Plan of the vil a decorated baths and a large two-room building pre- (Thomas 1964, Fig. 167). 173 Slavko CIGLENEČKI Fig. 3.11: Školarice near Ankaran. Pars rustica of the vil a, view from the east (2002). Fig. 3.12: Virovitica, Kiškorija South. Plan of a part of the settlement (Jelinčić Vučković 2015, Fig. 7). possibility that people from the vicinity sought shelter CROATIA here in times of danger. A settlement developed in the large prehistoric Roman vil as in the Pannonian part of Croatia are hillfort on Ulaka already in the 1st century. It comprised rare and poorly known. The excavations in advance of buildings arranged in concentric rows around the main motorway construction did, however, unearth previ- open area (Gaspari 2020). The buildings had stone ously unknown settlements with simple timber buildings foundations and a timber superstructure. The settlement spread across large areas, which have been predomi- developed continual y with its last peak in the second nantly identified as vici or settlements. half of the 4th century, though some artefacts also speak of life in the 5th century. At Drenje near Zaprešič, the central building of The rescue investigations at Školarice near An- a farmhouse with outbuildings has been found, dated karan ( Fig. 3.11) brought to light large parts of an ar- to the 1st–4th centuries (Leleković, Rendić Miočević chitectural complex (Žerjal, Novšak 2020). It is a large 2012, 281). vil a, complete with pars urban a and pars rustica, located The integral y investigated vil a at Donja Glavnica at a major junction of land and maritime routes. Only was constructed in the vicinity of a mineral spring and the baths were found of the pars urbana, while the pars can broadly be dated between the 2nd and the 4th century rustica comprised areas for the production of olive oil (Sokol 1981, 178–184). It comprised thermal complex and a large storage house. The vil a was created in its and living areas. greatest extent already in the 1st century and persisted An example of a vicus was found at Ivandvor, in with minor modifications to the mid-5th century, when the suburbs of Đakovo (Leleković, Rendić Miočević it was abandoned following a fire. 2012, 292–293, 300–301). It presumably extended across A vast farmhouse was constructed in the 1st or ca 20 ha and was composed of three groups of largely 2nd century at Predloka and renovated in the 4th or timber, in a small measure also brick buildings. Most of 5th century (Boltin Tome 1986). The renovation work them had long rectangular rooms associated with dif- encompassed restoring and raising the semicircular ferent sunken buildings in the vicinity. The settlement apse of the large reception hall with a wall in the opus has been dated from the late 2nd to the mid-4th century. spicatum technique. Coins (solidus of Justinian) and Similar to it is a large settlement complex at Vi- other small finds show that the vil a was inhabited to rovitica, Kiškorija South ( Fig. 3.12), predominantly the late 6th century. The associated cemetery dates to composed of sunken timber houses and dated between the 6th century. the 2nd and the 5th century (Jelinčić Vučković 2015, 351–358). Individual units were composed of a hut or semi-sunken dwelling, work area and refuse pits. The architecture and small finds reveal a ful y self-sufficient vil age built in a prehistoric tradition. 174 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.13: Dragonera Jug. Plan of the vil a (Starac, A. 2010, Fig. 89). In Istria, numerous vil as were continual y inhab- The residential-production complex at Dragonera ited (cf. Jurkić 1981; Matijašić 1988; id. 1998; Starac, Jug near Peroj ( Fig. 3.13) is an excellently preserved A. 2010). In Červar, Porat near Poreč, a building was Roman vil a with Late Antique architecture that offers found that served for pottery production from the 1st a good insight into the Late Antique renovations. It was century onwards, later also oil production (Džin, Girardi continual y inhabited from the 1st to the 7th century, Jurkić 2005). Residential rooms, hypocaust and baths when it was final y abandoned. It has been integral y were added in the 3rd century, in the 4th century also investigated and also published (Starac, A. 2010). The a large apsidal hal . The 6th century witnessed gradual complex has four wings, the south one being residential, decline with a large part abandoned and only some as well as a large inner courtyard. Surviving above the rooms adapted, while hearths were set up on top of the first construction layer from the 1st–4th century are the former mortar floors towards the end of the 6th century. remains of the Late Antique residential area from the A Roman vil a at Vrsar was modified in the 4th 5th–7th century, when four rows of rooms were built century to incorporate an apsidal hall decorated with connected by a corridor, as well as a prestigious apse in mosaics long interpreted as an Early Christian church. the central part. Recent investigations, however, have revealed that the Two Roman vil as on the island of Veliki Brijun dis- latter was a Late Antique vil a that only lost its original play characteristic signs of Late Antique transformation function in the 7th century, when an oil production facil- (Begović, Schrunk 2001, 159–162). One is a prestigious ity was installed in its interior (Tassaux 2003, 383–390). vil a built in the Verige Bay ( Fig. 3.14) in the 1st century; In the vicinity of a large peristyle vil a at Barbariga, some of its luxury parts were abandoned in Late Antiquity archaeologists found a production centre that is the and others transformed, primarily into storerooms, the largest one on the east coast of the Adriatic. The partly baths into a Christian church. The other Roman vil a, with known complex hosted at least twelve pairs of oil presses a fullonica and located in the Madona Bay (Kastrum; see in function simultaneously (Matijašić 1982, 58–59). Chapter 3.3.2), was turned into a densely built-up settle- The oil production plant was already created in the 1st ment protected with strong defensive wal s. century, underwent several modifications and decreased The vast Vižula Peninsula near Medulin holds the in size through centuries, but remained in operation at remains of buildings distributed on several terraces on least to the late 6th century. all sides of the peninsula (Džin, Miholjek 2019). The 175 Slavko CIGLENEČKI Fig. 3.14: Verige Bay at Veliki Brijun. Remains of the prestigious Roman vil a (2003). continuity of construction here spans several phases In Cickini near Sršiči, also on the island of Krk, a from the 1st to the 6th century. Particularly prominent is complex of buildings next to an Early Christian church a lavishly decorated maritime vil a from the 4th century. with a baptistery was created through several modi- Also well-documented is the phase from the 6th and fications of a Roman vil a (Starac, R. 2005). The vil a early 7th century, when earlier rooms were adapted and was renovated during the 5th century and its corners hearths set up on top of mosaic floors. The vil a then lost reinforced with rectangular additions; the recovered its original character and was transformed into a modest small finds date to the late 4th and the 5th century. In settlement (Džin, Girardi Jurkić 2008). the Early Byzantine period, parts of its porticus were In the interior of the Istrian Peninsula, a farmhouse subdivided with drystone wal s to create a row of resi- has been integrally investigated at Stancija Peličeti dential rooms. These were home to a poor population (Džin 2006). It was enclosed with a wal , comprised who stored liquids in the stone urns collected from a both residential and production parts, and was inhabited Roman cemetery. from the 1st to the 6th century. In the third construction The large Roman maritime vil a at Martinščica, phase, dated from the 3rd to the 6th century, the existing on the island of Cres, witnessed a variety of changes in rooms were subdivided with drystone wal s, and three Late Antiquity (Čaušević-Bul y, Bul y 2020, 279–281). water cisterns and production-storage faculties were These include a part turned into a reception hall with constructed. The vil a of this phase offered markedly a large apse. As yet unexplained is its relation with the lower living standards and hosted an increased number large Early Christian church built at the edge of the of inhabitants. vil a roughly in the 5th/6th century; the vil a presumably The Roman vil a at Poje in Njivice, on the island belonged to the bishop from the nearby city of Osor of Krk (Vodička Miholjek 2015) reached its peak in (Apsorus). the 3rd century. In the 5th and 6th centuries, two rooms Particularly important for our understanding of the with wal s of inferior quality and a slightly different settlement continuity in Dalmatia is the large farmhouse orientation were constructed that disregarded earlier at Muline ( Figs. 3.15−3.17), on the island of Ugljan constructions. Some elements recovered in the vicinity (Suić 1960, Suić 1976a). Its remote location enabled indicate the existence of an Early Christian building an uninterrupted development. Mainly its production part has thus far been investigated, while trial trenching 176 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.16: Muline. Outbuildings from the north-west (2005). Fig. 3.15: Muline. Plan of the vil a with a church and memoria (Begović, Schrunk 2001, Fig. 8). Fig. 3.17: Muline. Memoria from the north (2005). and underwater remains indicate the existence of other a large building (church?), a Late Antique tomb and buildings. Mate Suić briefly mentions several construc- burials from the 6th century (Milošević, A. 1989; id. tion phases and a proliferation of production facilities 1998, 288–290). from the beginning of the Roman period to the end of The remains of a large Roman vil a with a lavish Antiquity. These were associated with the production of residential part and small associated buildings came to olive oil (room for storing olives, mil , press, oil tanks, light at Grušine and the nearby church of St Lawrence cel ar with storage pithoi), as well as water cisterns. The (sv. Lovre) near Šibenik (Laszlo Klemar, Zeman 2010, vil a is particularly significant because of the Early Chris- 150–153). It is believed that a rural agglomeration de- tian complex ( memoria), built in immediate proximity in veloped here in Late Antiquity that included an Early the second half of the 4th century and comprising several Christian church and several dispersed buildings. rooms and later additions, a basilica from the 5th century The Lovrečina Bay on the island of Brač ( Fig. 3.18) and a small mausoleum, the construction of which has holds the remains of Roman outbuildings from the 2nd– been dated to the initial decades of the 5th century (Suić 3rd century, on top of which a building (presumably vil a) 1996, 660–665; Uglešić 2002, 86–92). developed in Late Antiquity and later an Early Roman- In the Dalmatian hinterland, the remains of a esque church (Kovačić 2006a). Located in immediate Roman vil a were found at Crkvina near Otres that proximity are the ruins of an Early Christian church was renovated in the middle of the 4th century, though and the associated cemetery (Fisković 1982; Begović, in a more modest form and with economic functions Schrunk 2001, 164). (Alajbeg et al. 2019). It persisted to the mid-6th century. Numerous and rich artefacts date the farmhouse The site of Bisko near Knin revealed approximately at Novo Selo Bunje on the island of Brač from the ten small isolated buildings (Milošević, A. 1986). Three 1st to the 6th century. Excavations are ongoing and of these were partial y or completely excavated. They only preliminary reports have thus far been published were constructed in the 2nd or 3rd century, later reno- (Jelinčić Vučković et al. 2019), revealing residential and vated on several occasions and inhabited to the late 6th production parts with oil and wine presses, baths and in century. Found not far from them were the remains of one room a sarcophagus. It seems to have been among 177 Slavko CIGLENEČKI settlement church Fig. 3.18: Lovrečina. Position of the settlement (from the west) with church wal s preserved to a great height in the foreground (2018). the largest farmhouses on Brač, where a stonemason’s workshop is also believed to have been in operation. In the Dalmatian hinterland, the remains of a Roman and Late Antique settlement have been found at Proložac Donji, associated with an Early Christian church from the 5th/6th century, the latter with a clearly identifiable renovation phase in the 6th century. Rising above the settlement is the Kokića glavica hill that holds a prehistoric settlement refortified in the Justinian pe- riod (Gudelj 1998; id. 2000). The maritime vil a on the island of Sv. Klement near Hvar revealed parts of a building with corners subsequently strengthened with rectangular additions, the artefacts from which indicate a dating to the late 4th and the 5th century (Begović et al. 2012; Teichner, Ugarković, 2012). The hidden and protected bay of Ubli on the island of Lastovo ( Figs. 3.19, 3.20, 3.301) hosted a large set- Fig. 3.19: Ubli. Plan of the settlement (Jeličić-Radonić 2001, Pl. 2). 178 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.20: Ubli. Position of the settlement (from the west) with the church foundation in the foreground (2017). tlement already in the Roman period (Jeličić-Radonić BOSNIA AND HERZEGOVINA 2001; Begović, Schrunk 2001, 164). A series of houses and economic facilities lined the main communication Very little is known of settlements with continuity and revealed different additions and modifications, in in Bosnia and Herzegovina. Their existence can primar- some places with the reuse of Roman stones. An Early ily be surmised from some Early Christian churches Christian church was built there in the 6th century. The presumably built in the area of their ruins. Also poorly settlement was agrarian in nature and revealed several researched are the unfortified settlements near Late An- presses for oil and wine, as well as facilities associated tique forts that thus far revealed no elements other than with their processing and storage. The housing is poorly churches (Fekeža 1998, 256–257). Some settlements do built; the remains of stairs in some indicate the existence display continuous habitation from the Roman period of upper storeys. A single house, with painted wal s, was into Late Antiquity. found to have changed in function from residential to economic, though possibly only on the ground floor. The villa at Ograja in Putovići near Zenica Jasna Jeličić-Radonić presumes that the living quarters ( Fig. 3.21) has three phases with continuity from the 1st were located in the upper storey. The large quantities of to the 6th century (Paškvalin 1980; id. 1990, 69). In the amphorae and other ceramic containers for exporting second phase, dated to the 4th century, the main hall wine and oil, possibly also grain, show the settlement was extended with an apse with a piscina, while the reached its peak in Late Antiquity. The exact date of third phase (from the 4th to the 6th century) brought an abandonment is not known, though the repairs to the end to the use of the pool and praefurnium. There were church indicate it was in use over a long period, suggest- also minor changes in the interior of the vil a; the pool ing that the settlement was continual y inhabited into was rather used as dwelling space for new inhabitants. the Early Middle Ages. The Roman vil a at Debelo Brdo in Sarajevo was situated below a fortified Late Antique settlement. Inside the vil a, individual Ostrogothic artefacts were found; the existence of an Early Christian basilica is also very likely (Fekeža 1998, 255). 179 Slavko CIGLENEČKI Fig. 3.21: Ograja in Putovići. Plan of the vil a (Paškvalin 1990, Pl. 1). The Roman settlement at Ilidža in Sarajevo re- vealed several buildings including a luxury peristyle villa with several modifications (Mulvin 2002, 86; Busuladžić 2011, 158). Three phases are mentioned, the last two dating to the 2nd–3rd centuries, when the vil a was extended in the east. The buildings persisted to the 4th century. The complex was undoubtedly connected with the nearby thermal springs. Two large residential buildings, one next to the other, were found at Mušići near Višegrad ( Fig. 3.22) (Busuladžić 2011,159). The recovered small finds broadly date them from the 2nd to the late 4th century. Traces of a large Roman settlement that persisted into Late Antiquity came to light at Crkvina near Halapić, in Glamočko polje (Sergejevski 1942, 126–131). The settlement is primarily known for Roman stone monuments and coin finds, while Frane Bulić and Ejnar Dyggve also excavated parts of a large basilica they dated to the 6th century. Located just above the settlement is the Late Antique fort on Gradec, which has not yet been investigated. The vil a at Strupnić near Livno remains enigmatic in several respects (Bojanovski 1970, 508–509; Mulvin 2002, 104; Busuladžić 2011, 152; Turković, Zeman 2011). The poorly documented excavations do not allow a more precise dating and it is hence not known whether it is an earlier construction renovated in the 3rd or 4th century or one that was only created at that time. We primarily know the central residential part of the vil a, while the remains of two other buildings served as outbuildings or economic facilities. It is dated to the mid-4th century. Ivo Bojanovski reasonably presumes a mutatio here, located at the point where the Roman road began its ascent across the high mountain pass of Čatrnja (Bojanovski 1974, 68–69). The complex of buildings at Proboj near Ljubuški was constructed in the middle of the 2nd and destroyed in the second half of the 4th century (Busuladžić 2011, 151). The finds of tools, as well as the distribution and function of rooms point to a typical farmhouse. The large luxury villa in Višići near Čapljina ( Fig. 3.23) (Čremošnik 1965; Busuladžić 2011, 150) was built in the 1st and survived to the late 4th century. It boasts a large peristyle building with richly decorated rooms. Reports mention no later constructions or reno- vations, only a variety of workshops. A small group of burials from the 5th/6th centuries was dug into the ruin- ous remains of the vil a and indicate a small settlement Fig. 3.22: Mušići near Višegrad. Plan of the vil a (Busuladžić 2011, Pl. 19b). 180 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.23: Višići near Čapljina. Plan of the villa (from Busuladžić 2011, Pl. 2b). existing in immediate proximity (Čremošnik 1965, 200; Mulvin 2002, 76). One of the largest vil as was unearthed at Panik Fig. 3.24: Panik nar Bileća. Plan of the vil a (Busulađić 2011, Pl. 11b). near Bileća ( Fig. 3.24) (Čremošnik 1976; Busuladžić 2011, 154). Its central part with lavishly furnished rooms is composed of two buildings facing the river and con- nected via a porticus. This part is associated with other residential and economic spaces. The vil a has been Presumably associated with the luxury villa at dated from the 2nd to the 4th century and presumably Panik was the agrarian settlement at Dračeva strana peaked in the late 3rd or early 4th century. It holds the near Bileća (Busuladžić 2011, 156). It is a fairly compact mosaic with the depiction of Orpheus that is believed to group of four buildings constructed around a courtyard, have expressed the Christian identity of the owners in a as well as one building further away. Small finds show time before Christianity became an official y recognised that it lasted from the 1st to the 4th century. religion (Čremošnik 1974). In the 6th century, an Early The small residential-production building at Bi- Christian church was presumably built in the north hovo near Trebinje ( Fig. 3.25) has been broadly dated part of the vil a, succeeded later by a medieval church from the 2nd to the 4th century. An oil press or torcular (Čremošnik 1974, 243). was found in one of its rooms (Busuladžić 2011, 153). Fig. 3.25: Bihovo, Trebinje. Plan of the farmhouse (from Busuladžić 2011, Pl. 10b). 181 Slavko CIGLENEČKI 3.2.2 EASTERN PART (presumably horreum) and a building with a hypocaust, in the vicinity also an associated cemetery. It appears SERBIA to have been an important production and distribution centre that supplied settlements in a wider area. It existed A large roadside settlement lies at Mala Kopašnica from the 2nd to at least the late 4th century (Stamenković near Leskovac. Exacavations revealed a large outbuilding 2013, 54, 148–157; Ivanišević et al. 2016b). Madara Montana Nicopolis ad Istrum Skobelevo Mala Kopašnica Podvrh-Crkvine Cherven breg Chatalka Lambata Mirište Armira Diaporit Strongyli Frangoklisia H. Thomas/Agios Thomas 0 100 km Fig. 3.26: Countryside settlements that continued from the Roman period, eastern part, discussed in the book. 182 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.28: Madara. Plan of the vil a (from Mulvin 2002, Fig. 35). of them came to light in the lowland part of the Rosica Valley. With the growing distance from Nicopolis, the vil as show an increasingly irregular distribution, occur farther apart and are predominantly larger. Most are a cluster of houses at a distance of 500 m from the main building of the vil a. All these settlements were aban- doned towards the end of the 4th century. Fig. 3.27: Montana 1. Plan of the vil a (Dintchev 1997, Fig. 45). The large vil a at Madara ( Figs. 3.28, 3.29, 3.316) was constructed in the second half of the 1st century and likely represented the centre of an imperial domain (Dintchev 1997, 74–79; id. 2008, 406; Mulvin 2002, BULGARIA 90−91). The large peristyle building was associated with other, mainly outbuildings. The central peristyle build- Three large vil as in the area of the city of Montana ing was completely renovated in the late 3rd or early 4th ( Fig. 3.27) show a similar appearance and development; century, when it was also enclosed with a strong wal . they are a complexes of houses and outbuildings already The destruction of this complex has been set to the constructed in the 2nd century (Dintchev 1997, 32–41; time of the Gothic incursions in the last third of the 4th Mulvin 2002, 95–97). They were torn down in the late 3rd century. Subdivisioning and small finds show that the century, then renovated and inhabited to the last third of buildings were later at least partial y reinhabited. The the 4th century. In one of them, an Early Christian church many uncertainties in the excavations and their publica- was constructed in the late 4th or early 5th century on tions hinder an interpretation of the remains of the last top of the ruins of the earlier vil a. Remains of modest phase, but they do suggest a modest settlement rather dwellings indicate life even after this date, but they no than a large vil a (Dintchev 1997, 77–79). longer speak of the continuity of the vil a, but rather The remains of a large unfortified settlement were of small rural settlements (Dintchev 1997, 34, 37, 40). found at Cherven breg, below the Rila mountain range. The systematic field surveys in the hinterland of It was created in the Roman period. An Early Christian Nicopolis ad Istrum have identified three hundred church was constructed in its centre in the middle of Roman to Early Byzantine sites (Poulter 2007c, 79–82). the 4th century (Petrova 2019). Numismatic evidence Most investigated sites revealed rectangular peristyle reveals intense occupation from the 2nd to the last third vil as with small outbuildings that were the home of re- of the 3rd century. Habitation traces cease for three or spective owners and their families. The greatest number four decades following a fire associated with the Gothic 183 Slavko CIGLENEČKI fort Fig. 3.29: Madara. Roman vil a from the west with the Early Byzantine fort behind it (2012). invasion in 270/275. The settlement was reinhabited in the first quarter of the 4th century and persisted all to its destruction in the 480s. Svetla Petrova presumes the settlement was the vicus of the army fort in the city of Germania. Three vil as constructed close to one another and known as Chatalka near Stara Zagora lie in the imme- diate vicinity of a river confluence that offered a good natural protection to two of the vil as, namely Chatalka 1 and Chatalka 2 (Dintchev 1997, 60–67; Mulvin 2002, 77–79). All three were already built in the Early Roman period and numerous coins show they were still inhab- ited, at least partial y, in the 5th century. Particularly interesting is Chatalka 1, which had its pars urbana on the natural y best protected part and was encircled with a 2 m thick wal . Chatalka 2 was completely economic in nature ( pars rustica). The third vil a is located at the nearby site of Lambata and was added a large apsidal residential building in the second half of the 3rd cen- tury. After it was abandoned or destroyed in the 370s, a modest dwelling was constructed in its ruins that was presumably inhabited to the mid-5th century (Dintchev 1997, 68–70). Traces of a contemporary settlement ( vi- cus?) were found in the immediate vicinity of Chatalka 1 and 2 (Harizanov 2020, 109–110). Located around Skobelevo is a settlement of at least twelve separate farmsteads spread across a 40 ha large area (Katsarova et al. 2018). The settlement devel- oped continuously from the second half of the 1st to the 360s/370s. The largest farmstead was composed of sim- ple buildings arranged around a walled courtyard. Other farmsteads were separate from each other and associated with fields. The inhabitants are believed to be the tenants Fig. 3.30: Armira. Plan of the vil a (Kabakchieva 2011, Fig. 2). 184 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.31: Mirište, Petrovac. Plan of the vil a (Zagarčanin 2014, Fig. 7). of the estate belonging to a large, as yet undiscovered MONTENEGRO farmhouse. Geophysical surveys revealed four other settlements in the vicinity. One of these, spread across a The large Roman villa at Mirište in Petrovac ca 60 ha large surface, presumably encompassed fifteen ( Fig. 3.31) initial y had a ful y residential character and farmsteads. They also include a large building complex was abandoned in the 4th century (Zagarčanin 2014). that may be a farmhouse possibly associated with other, Soon afterwards, it was modified into a facility for the smaller farmsteads. production of olive oil and likely also wine. The building The Roman villa at Armira near Ivailovgrad was reinforced with thick buttresses and added another ( Fig. 3.30) is one of the extensively excavated and outbuilding. The numerous pithoi and amphorae reveal a restored vil as in Bulgaria (Kabakchieva 2011). It was complete transformation into a farmhouse. It is believed already constructed in the 1st century and later renovated to have been in use from the 2nd to the early 7th century. and enlarged on several occasions. Following the last The last phase, dating to the 6th century, reveals levelling extensive renovation in the early 3rd century, it remained of rooms and the use of materials of inferior quality. unchanged to the first half of the 4th century when it The villa at Podvrh-Crkvine, at the foot of witnessed minor subdivisioning and additions to the Martinička Gradina (see Chapter 3.3.2), has only been mosaic decoration. It is believed to have been destroyed partial y identified and documented, and is believed to in a fire after the Battle of Hadrianopolis that took place have been a rural aristocratic abode. It also comprised in 378. Only scarce habitation remains in a limited part an Early Christian church (Stevović 2014, 102). of the building postdate this battle and even these cease towards the end of the 4th century, when thick layers of loam slid down the slope after an earthquake, covering the remains of the vil a. 185 Slavko CIGLENEČKI Fig. 3.32: Diaporit. Plan of the vil a (Phase 7, c. 490) (Bowden, Përzhita 2014, Fig. 15). ALBANIA side settlements that presumably existed between the 2nd and the 6th century on the territory of Philippi. Un- The large vil a near Diaporit ( Fig. 3.32), in imme- specified Late Antique settlements were also discovered diate proximity to Butrint, is one of the rare systemati- during the extensive surveys conducted in the Campania cal y investigated vil as in the area (Bowden, Përzhita region (Dunn 2004, 539–542). 2014). It is sited on the terraces of Lake Butrint, on a major Roman road. The rich vil a underwent several Several cases of well-preserved ruins in Epirus en- renovations and was abandoned as a luxury residence able an insight into such countryside architecture. The in the middle of the 3rd century. Next to the ruins of its vil a at Frangoklisia shows clearly discernible buildings buildings, a three-aisled basilica was constructed that dating to the 2nd and 3rd centuries, while the surveys also reused the stone material from the vil a, which indicates yielded Late Antique pottery (Bowden 2003, 63–64). A the vil a was at least partial y ruinous at this time. The closely similar situation has been observed in the Roman church may roughly be dated to the late 5th century, vil a at Strongyli, with the small finds recovered during when a pilgrimage centre with associated buildings was field surveys indicating continuity of the site (Bowden established here; the whole complex was abandoned in 2003, 64–65). the mid-6th and only reactivated in the late 7th century. Considerable amounts of Roman and Late An- Located close to Diaporit is also a hil top fort on Kalivo, tique finds were unearthed close to the settlement at which may have served as refuge in times of danger. H. Thomas (Agios Thomas), north of Preveza, around The intensive surveys in the hinterland of Butrint the Early Christian church of St Minas. The site, meas- revealed several sites with an overlap of Roman and Late uring 225 × 110 m, appears to be a small settlement or Antique remains, but none can be identified and dated a large vil a (Bowden 2003, 70). with any precision (Bowden 2003, 67–68). Most vil as in Greece are believed to have already been abandoned in the 4th century. Those that show signs GREECE of life in the 6th century were likely reinhabited in Late Antiquity (Curta 2014, 37). In the 6th century, several There is a wealth of field survey data on the coun- vil as were presumably still active around Corinth, in tryside constructions in Greece, but not much excavated Argolis and in Messenia (Morrison, Sodini 2002). Florin remains, hence the settlement forms are general y not Curta believes that Greece was a land of isolated single- well-known. In his overview of the countryside in Greek family farmsteads, i.e. of a pattern not present elsewhere Macedonia, Archibald Dunn notes unfortified country- in the Balkans (Curta 2014, 39). 186 3.3 NEWLY-FOUNDED SETTLEMENTS slope that would not provide the conditions to withstand IN THE COUNTRYSIDE a major attack. Villas are not a novelty of this period. With The previous chapter shows that only rare set- the exception of those of earlier origins and higher tlements in the continental part survived the process quality, they have a more modest design and are also of settlement disintegration, whereas those in better poorly known, receiving limited attention on the part protected areas, primarily along the coast, continued in of archaeologists as their remains are less exposed and greater numbers. This was not a one-way process without often poorly identifiable on the surface. It was therefore interruptions. Observable in parallel from the last third necessary to look even more careful y for the rare ex- of the 3rd century onwards are increasing attempts at amples that would offer a more comprehensive insight local self-protection, but also state interventions that into the settlement of the period. It is these modest promoted the establishment of new settlements, most traces that frequently and inconspicuously fall victim frequently in natural y well-protected locations. These to the modern construction euphoria. Unfortunately, new settlements form a particularly important part of even the modern systematic rescue interventions the book, as they represent the most prominent settle- (primarily in advance of motorway constructions) ment phenomenon in the part of the former Empire sometimes fail to detect these settlements, as they only under discussion. rarely take place in the mountainous areas beyond the There were also other forms of dispersed set- main valleys. tlement, primarily newly-constructed Roman villas, The text below tackles some of the more character- farmsteads and other unfortified settlements. The mod- istic vil as and other residential buildings constructed est remains and the location in areas that often hinder in the second half of the 3rd and the 4th century, which successful intensive surveys are the reasons why these were abandoned towards the end of the 4th or in the are not well-known. But even in areas where the favour- first half of the 5th century at the latest. To this group able climatic, vegetation and habitation conditions allow with very few examples, we can add the extremely rare intensive surveys, their remains are scarce (cf. Poulter cases of unprotected dwellings that persisted or were 2007b, 41–46; Bintliff 2007) and only offer a partial and even established after the mid-5th century. None of this anything but general insight into the settlement patterns. continuing settlements reveal major changes in the This, more limited segment of Late Antique settlement architecture and wil therefore only be discussed briefly. will be presented first, before the fortifications. The overview below aims to present all the diversity of the newly-established settlements. Their defences WESTERN PART are emphasised as one of the most important elements; they show great variety and reveal the entire range of Italy possibilities, from an effective use of the natural environ- ment to the different elements such as defensive wal s, The part of Italy discussed in this book (province protechismas, ditches, towers and combinations of these. of Venetia et Histria) revealed a great number of vil as The other important element, which often also explains and other forms of countryside settlement, but most the function of settlements, is buildings in the interior. were already constructed in the Roman period and These range from simple casemates as soldiers’ accom- continued into Late Antiquity or display a succession modations to simple dwellings and private residences of of abandonment, modification and new construction a higher quality. There are few known public buildings; on the same spot that is often difficult to date with any the most prominent among them are churches that often precision. With modifications, these vil as have impos- add another function to the forts and fortresses. ing dimensions and also a luxuriously furnished pars urbana. There are certain examples where available evidence (primarily coin finds) only shows a late phase, 3.3.1 NEWLY-FOUNDED UNFORTIFIED but this evidence originates from field surveys and brief SETTLEMENTS rescue investigations that cannot conclusively prove their construction in the Late Roman period and not This group ( Figs. 3.33, 3.49) comprises villas, previously (cf. sites at Padenghe sul Garda 1, Maniago 11: farmsteads and other forms of unfortified settlements De Franceschini 1998, 169, 326). The late modifications mostly established in the second half of the 3rd or in the and extensions to vil as have already been discussed in 4th century, very rare ones later. The group also includes the chapter on the settlements with continuity from the wal ed Roman farmhouses, as their wal s are considered Roman period (see Chapter 3.2). In areas protected by as delimiting the residential-economic unit rather than the Alpine barrier wal s, some vil as that have survived as fortifications against enemy attack; such an interpre- from the Roman period reach an imposing size and tation is based on their location in the lowland or on a boast a highly luxurious pars urbana. 187 Slavko CIGLENEČKI Höflein Szentkirályszabadja-Romkat Löffelbach Kékkút Radvanje Komló-Mecsekjánosi Volpago Hosszúhetény Montorio Rim near Roč St Chrysogonus in Glavotok Ljusina Založje Brodac Rankovići Tišina Mali Mošunj Stup Skelani Lisičići Vučipolje Miri near Ostrvica Mirje (Mirine) Polače 0 100 km Fig. 3.33: Newly-founded unfortified settlements in the countryside from the western part, discussed in the book. Despite the predominance of settlement forms was constructed in the second half of the 4th century that continue from the Roman period, there are several and partly persisted into the 5th century. Only the main new constructions here as wel . We should mention the residential building is known, the centre of which was remains at Montorio, where mosaics and perimeter a small heated hall with an apse and several side rooms. wal s with towers indicate the existence of a large Late The associated economic facilties have not been inves- Antique vil a (De Franceschini 1998, 157–158). The tigated. other example is the vil a at Volpago, dated to the 3rd The smal , but architectural y very complex peri- and 4th centuries (De Franceschini 1998, 286–287). style vil a from Löffelbach ( Fig. 3.34) is an example of a prestigious new construction (Schrettle 2007, 262–264; Marko 2011; earlier interpretation in Mulvin 2002, 90). Austria It is marked by a peristyle, as well as a large apsidal aula and several adjacent room of a polygonal plan. New There are several vil as in Austria that were erected interpretations show that it was built in the early 4th in Late Antiquity, located in the two newly-formed century, later witnessed only minor adaptations and was provinces of Noricum. abandoned within the same century. An example of a modest, but still well-built farm- house was found at Höflein (Ladstätter 2002, 324). It 188 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.34: Löffelbach. Plan of the vil a (Marko 2011, Fig. 2). Hungary Vil as were an important element of the Roman countryside economy in the fertile plains of Hungary and most had a continuous development into the Late Roman period. However, there are several examples that were built anew in the Late Roman period and, interestingly, display reliably identified Early Christian architecture. Distinguishing between the villas with continuity and those built anew is not always apparent and we are only able to identify few reliable examples. One of these is the vil a at Hosszúhetény ( Fig. 3.35), in the vicinity of Sopianae (Pécs) (Thomas 1964, 35; Mulvin 2002, 85). Only the central residential build- ing has been investigated, with a peristyle and a large apsidal hal . A particular feature is a pair of hexagonal towers that guard the entrance to the building. In plan, the vil a is similar to that at Polače on the island of Mljet (see below). The wall paintings and the design with a protected entrance suggest a dating to the 4th century. Also in proximity to Sopianae are the remains of a partially investigated peristyle villa at Komló- Mecsekjánosi, which has a large apsidal hall (Mulvin 2002, 88) that general y characterises vil as built in the 4th century. Fig. 3.35: Hosszúhetény. Plan of the vil a (from Mulvin 2002, Fig. 22). 189 Slavko CIGLENEČKI Fairly similar in plan is the vil a in Kékkút at Lake Balaton (Thomas 1964, 54–55; Bíró 1974, 31; Mulvin 2002, 86–87). The chronology of the site is poorly known, but the partial y investigated vil a shows features characteristic of the late vil a constructions. The main residential building has an elongated symmetric plan with a peristyle and small towers at the front. Two other buildings were found next to it, one with a large three- aisled Early Christian basilica with a narthex identified in plan; a bronze Christogram corroborates the Early Christian character of the complex. The remains of a villa at Szentkirályszabadja- Romkat at Lake Balaton, are poorly known (Thomas 1964, 118–122; ead.1980, 292; Mulvin 2002, 104–105). The cluster of buildings indicates a vil a with a pars urbana and a pars rustica. Its construction has been tentatively dated to the 3rd and its destruction to the 4th century. The scarce data from the excavations and a poor preservation of the vil a, with only the outer wal s surviving, do not allow a more detailed interpretation, though the long rectangular building protected with a pair of towers at the entrance can be seen as the main Fig. 3.36: Radvanje. Plan of the vil a (Strmčnik Gulič 1990, residential building. Fig. 10). Slovenia Poštela Several vil as and smaller buildings were excavated in Slovenia that correspond with the usual layouts of contemporary countryside architecture. An integral y investigated villa is that at Radvanje near Maribor ( Figs. 3.36, 3.37) (Strmčnik Gulič 1990, 139–143). This large vil a was constructed in the second half of the 3rd century on the spot of earlier constructions of an unknown function. It measured 98 × 90 m and was enclosed with a 0.7 m thick wal , onto which most of the buildings leaned from the interior. A large residential Fig. 3.37: Radvanje. Foundations of the vil a in the foreground, building, with a hypocaust heating system in all rooms, prehistoric and Late Antique fortification Poštela in the back- stood in the northwest corner. This building was sur- ground (1988). rounded by a small court closed off by a long building with several rooms. The south corner held outbuildings, while other buildings were situated along the perimeter wal . Modest finds indicate the vil a continued only in Part of an unfortified Late Antique settlement came the first half of the 4th century. to light at Rim near Roč, in the continental part of the Istrian Peninsula (Marušić 1987). Its remains, damaged by arable farming, show a settlement with sacral and profane Croatia buildings. The former include a hall church from the 5th century with mosaics and architectural decoration, as wel In Croatia, there are major differences between the as an apse added at a later date. A small room beside the countryside settlement in the continental, Pannonian church holds the remains of a baptistery, which is presum- part and that along the Adriatic. The countryside vil as ably referred to in an inscription from the 6th century that and other settlements in the former were abandoned speaks of embellishing a small church building (Marušič after the end of the 4th century and no new ones have 1987, 237). West of the church was a residential building, been detected. The coastline, in contrast, revealed a attributable to the 4th and partly the 5th century, that was myriad of new constructions ranging from simple partial y destroyed during the construction of the church. dwellings to palaces. A granarium with a roof of brick tiles, dated from the 4th 190 3. SETTLEMENT IN THE COUNTRYSIDE to the 6th century, was found north of the church. It would appear these are the remains of a Late Roman settlement that persisted, with different modifications, to the late 6th century; this dating is corroborated by the Late Antique burials unearthed in the vicinity. Excavations at the church of St Chrysogonus in Glavotok near Milohnići, on the island of Krk, revealed the remains of a multi-period site beginning in Late Antiquity (Janeš 2015). The earliest remains are of a rural complex. It comprises several buildings, of which a small square edifice survives best, while all others can Fig. 3.38: Miri near Ostrvica. Plan of the vil a (Rapanić 1984, only be surmised on the basis of the recovered floors. Fig. 2). Also preserved is part of a modest wall that enclosed this rural unit. The small finds and radiocarbon analyses date the settlement to the 4th–6th centuries, with most finds belonging to the later part of this time frame. The site at Vučipolje near Dugopolje yielded several habitation elements dating from the second half of the 4th to the 6th century (Borzić, Jadrić 2007). This site lies in the hinterland of Salona, at the junction of several Roman roads and at the edge of a fertile plain. It is presumed that a farmhouse existed here in the second half of the 4th century, which included a complex of cisterns with mosaic floors. After its abandonment in the 5th century, the toponym of Crkvina (meaning ‘church grounds’ in Croatian) and two masonry tombs indicate that an Early Christian church was built here. At the same time, inhabitants presumably dispersed, as several simultaneous agglomerations appeared at different sites in the vicinity and lasted into the 6th century. Most likely connected with this site is also the one at Banjače, some 500 m away, where two drystone buildings were unearthed, dated to the 5th and 6th centuries. Revealing no habitations traces, they are rather believed to have been primarily intended for storing foodstuffs and can certainly be seen as part of a larger settlement (Ožanić Roguljić et al. 2018). The vil a at Miri near Ostrvica ( Figs. 3.38, 3.39) lies on a remote and fertile high plateau of Poljice near Omiš (Rapanić 1984). A complex of three buildings was unearthed here, of which only the ‘palace’ is known better, which is a 33 m long two-storeyed building with thick buttresses and a central hal . In the absence of systematic research, Željko Rapanić used parallels and Fig. 3.39: Miri near Ostrvica. Remains of the vil a from the the construction technique to date the building to the 5th north (1987). or 6th century (for an earlier dating, see Turković 2011, 226). An important chronological indication of the life in this vil a is the unusual and beautiful y decorated church Jeličić-Radonić 2015). The dilemma associated with from the Justinian period, found in Gata roughly 2 km this vil a is whether we should see it as a fortified or an away (Jeličić-Radonić 1994). These elements define the unfortified complex; it was built on higher, but still ac- Poljice plateau as a settlement area where, in addition cessible and fairly flat ground and was visible from afar. to the ‘palace’, i. e. residence of a prominent dignitary, It is encircled with a thin perimeter wall that does not other settlement cores from the Late Antique period seem to have had a primarily fortifying nature, but was could be assumed. rather intended to delimit the complex and protect it The Late Antique villa at Mirje (Mirine) near from wild animals and thieves. The numerous pieces of Postire ( Figs. 3.40, 3.41), on the island of Brač, is stone church furnishings initial y led to it being inter- extremely well-preserved (Kovačić 2006b; ead. 2010; preted as a monastery (Fisković 1982, 165, 202), while 191 Slavko CIGLENEČKI Fig. 3.40: Mirje near Postire. Plan of the vil a (Kovačić 2010, Fig. on p. 27). Fig. 3.41: Mirje near Postire. Remains of the vil a from the south-east (2021). rescue excavations indicated the residence of a wealthy and lavish Late Antique vil a with a series of defensive landowner or manager of the Imperial quarry at nearby elements. It was constructed on the north coast of the Škrip. The pieces of church furnishings may belong to island, in a hidden bay sheltered from the winds. The an as yet undiscovered Early Christian church. residential part comprises a large apsed audience hal Holding a pride of place within the group of luxury with adjoining rooms, which is partial y fortified, most residences is the ‘palace’ at Polače ( Figs. 3.42−3.44), on ostentatiously with a pair of sizeable polygonal towers that the island of Mljet (Suić 1976a, 240; Fisković 1999; Mulvin guard the main seaward entrance. The opinions on the 2002, 101; Cambi 2002, 234–235; Suić 2003, 360; Turković vil a’s beginnings span from the early 4th to the 6th century. 2011). It is a complex that combines the design of a large Early studies dated the construction to the second half of 192 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.42: Polače. Luxury Late Antique residence (from Fisković Fig. 3.43: Polače. Wal s preserved to a great height from the 1999, Fig. 7). eastern part of the vil a (2005). Fig. 3.44: Polače. Vil a view from the west (2005). the 5th century based on the literary sources mentioning a is known on two Early Christian churches from the 5th deed of gift with which Odoacer gave the island of Mljet and 6th centuries, as wel as recently discovered baths to his comes Pierius. In contrast, the recent analyses of with hypocaust heating and mosaic floors from the late both towers and architectural parallels rather point to 5th century. Located in the immediate vicinity, on a hil , a luxury vil a erected in the early 4th century (Turković is a small Late Antique fort that has not been investigated 2011). Tin Turković proposes the Emperor Licinius as the in detail and its relationship to the vil a remains unclear. person who commissioned the vil a. The secondary wal s The bay was presumably a protected harbour for ships in the interior indicate its further development. Standing passing along this route, possibly even a winter base for next to the vil a is a complex of buildings, of which most the fleet (Fisković 1998, 275). 193 Slavko CIGLENEČKI Bosnia and Herzegovina paintings dated from the late 3rd to at least the mid-4th century. Two construction phases were established for The remote valleys of Bosnia host several settle- one of the buildings. This vil a is evidence that lavish ments of a late date. One is the three-room villa at architecture was constructed not only along the coast, Založje near Bihać ( Fig. 3.45) that is similar to the but also in the interior of the province of Dalmatia. vil a from Rankovići (see below), only differing in an Another vil a of a late date was found at Rankovići additional corridor around the building. Central heating near Travnik ( Fig. 3.47) (Čremošnik 1955a, 122–132; canals were found in one of the rooms. The rare small Basler 1972, 45–46; Busuladžić 2011, 160). It includes a finds date the building to Late Antiquity (Busuladžić long rectangular building with three rooms that had a 2011, 164). It was situated between two Early Christian porticus lining one of the longer sides. Stairs indicate an churches (Čremošnik 1958; Basler 1972, 120). upper storey. The pilaster capitals are formal y similar A vil a of a prestigious character is located at Lju- to those found in sacral buildings across Bosnia and sina near Bosanska Krupa, showing a design charac- date the building at least to the 5th century or later. An- teristic of late vil as, with a large apsed hall (Čremošnik other building stands in the vicinity; its apsidal design, 1959; Mulvin 2002, 90; Busuladžić 2011, 104). The hypocaust and praefurnium indicate baths. Traces of existence of an estate with economic facilities is posited several unresearched buildings were also recorded on in proximity. The vil a is believed to have been built in the slope of a nearby hill hosting an uninvestigated Late the first half of the 4th and lasted to the early 5th century. Antique fortification. These buildings point to a small Surviving of the vil a at Brodac near Bijeljina is Late Antique hamlet existing here, possibly associated primarily the part that incorporated a large room – with a refuge on the hil . workshop – and an apse of a similar, economic function, The building at Stup near Sarajevo is an example while the residential quarters were presumably located in of simple rural architecture. A wooden partition wall the destroyed southern part. Small finds date the build- divided the main room into two parts, one for dwelling ing largely to the 4th century (Busuladžić 2011, 166). and the other for economic activities. The floors were The modest building at Mali Mošunj near Vitez made of beaten loam. There was a small extension in has been dated roughly from the second half of the 3rd the east. Small finds date its construction to around 300 to the second half of the 4th century (Busuladžić 2011, (Busuladžić 2011, 159). 160). Frequent reports of other finds in the vicinity As many as three residential complexes were suggest this was most likely part of a small settlement partly investigated at Lisičići near Konjic ( Fig. 3.48), (Mulvin 2002, 92). in a length of several hundred metres. Vil a 1 was the Complex residential architecture composed of most prestigious, decorated with wall paintings. Irma three buildings with several heated and apsed rooms Čremošnik, who excavated the vil a, sets it to the 3rd came to light at Tišina near Zenica ( Fig. 3.46). Remains of timber architecture are mentioned in proximity (Busuladžić 2011, 161). The three buildings form a single luxurious complex fitted with mosaic floors and wall Fig. 3.45: Založje near Bihać. Plan of the vil a (Busuladžić Fig. 3.46: Tišina near Zenica. Plan of the vil a (Busuladžić 2011, Fig. 27b). 2011, Fig. 23b). 194 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.47: Rankovići near Travnik. Plan of the vil a (from Busuladžić 2011, Fig. 21b). century. Coins and other small finds indicate that the other two vil as (Nos. 2 and 3) date from the late 3rd to the late 4th century (Čremošnik 1957; Basler 1993, 23; Mulvin 2002, 89; Busuladžić 2011, 157–158). Surviving best is Vil a 3, which had a simple multi-room residential building and a building of an economic function inside the walled area (Čremošnik 1957, 149–155). A child skeleton was found beside the southern wal s enclosing Vil a 2, buried with items that include a cross-shaped brooch datable to the 6th/7th century (Čremošnik 1957, 147–148). The domus at Skelani (presumably forming part of municipium Malvesatum) is a well-researched building with mosaic floors and wall paintings broadly dated to the first half of the 4th century (Gavrilović Vitas, Popović 2015). Of a later date is a series of poorly-constructed and narrower wal s that have not been more precisely dated. Two Early Christian churches were also found in the vicinity (Basler 1972, 113–115). There are several other sites in Bosnia where we could posit either a continuation of Roman buildings Fig. 3.48: Lisičići near Konjic. Plan of the Villa 3 (from into Late Antiquity or buildings constructed anew in Busuladžić 2011, Fig. 16b). this period. This is most apparent in the case of the Early Christian churches where reports mention the ruins of Roman buildings in the vicinity without giving any details; such are the examples at Crvenica in Duvanjsko polje, Prisoj near Duvno and at Šiprage (Basler 1972, 75, 112, 115). At Majdan near Mrkonjić Grad, the remains of an Early Christian church were excavated in the midst of the ruins of an uninvestigated Roman mining settle- ment, for which Basler presumes continuity, similarly as for Blagaj in the Japra valley (Basler 1972, 92). 195 Slavko CIGLENEČKI EASTERN PART mansio) from the 4th century that is believed to have lasted to the mid-5th century (Petković 2016). There is Serbia also an associated cemetery nearby, while a Late Roman and Early Byzantine fort is posited above. The vast rescue investigations at Davidovac - The Late Antique settlement of Mediana, the sub- Gradište, a site on the road connecting Naissus and urbs of Naissus, revealed a large complex of residential, Scupi, revealed the remains of a settlement (presumed prestigious and economic buildings (Mulvin 2002, Mediana Pleven Pančićev vrh Kostinbrod Davidovac Obelija Bela voda Donje Nerodimlje Kruče Pešterica Trpčeva Crkva Thespiae Akra Sophia Berbati valley 0 100 km Fig. 3.49: Newly-founded unfortified settlements in the countryside from the eastern part, discussed in the book. 196 3. SETTLEMENT IN THE COUNTRYSIDE 92–94; Milošević, G. 2011). The different component Bulgaria parts, such as a large prestigious peristyle vil a, two small vil as, traces of a vicus, a horreum, a water tower Research shows that a relatively large portion of the and army barracks indicate a residential – economic vil as in the north-eastern and central northern parts of compound of great complexity. Researchers identified Bulgaria were abandoned in the mid-3rd century, though three main construction periods spanning from the first there are examples inhabited from the mid-3rd century third of the 4th to the mid-5th century when the settle- to the 370s that Dintchev attributes to his Period C ment was abandoned. It is interpreted as an imperial (Dintchev 1997, 123–125). vil a that hosted occasional visits of the emperors and their retinue between 330 and 378. The vil a near the city of Pleven ( Fig. 3.50), near A special example is the settlement with a church the Roman fortress of Storgosia, is believed to have been below Pančićev vrh (Pančić’s Peak, also Nebeske sto- inhabited from the late 3rd/early 4th to the 6th century. lice), located at 1800 m asl (Tošić, Rašković, 2007, 34–39; It is a large complex of buildings of different functions. Ivanišević 2016, 96). Just below the peak is an Early Standing out is an apsed building and a hexagonal tower Christian basilica with a double apse, wal paintings and that was probably added later, just before the 6th century. mosaic floors. It stands next to the ruins of buildings The items recovered from the tower include a hoard of 26 that small finds date to a time between the 3rd and the coins of Justinian I (Dintchev 1997, 94–95; Mulvin 2002, 6th century. Horizontal stratigraphy shows separate early 100). Dintchev sees this large residential complex as an (3rd and 4th centuries) and late parts, the latter dating to urban vil a of a high administrative officer from Storgosia. the 5th/6th centuries and incorporating the church. The The large peristyle complex near Kostinbrod was buildings represent an Early Christian centre linked built in the early 4th century. It was interpreted as an with adjacent settlements, which were mainly associated imperial residence or at least the home of the provincial with the particularly rich ore deposits on this mountain. governor of Dacia Mediterranea. A vil age presumably developed within it in the 5th century and a military fort towards the end of the 5th or in the 6th century (Dintchev Montenegro 1997, 83–94). The vil a at Obelija ( Fig. 3.51), in the vicinity of The bay of Kruče near Ulcinj holds a partial y in- Serdica (Sofia), was built in the late 3rd and persisted vestigated vil a with mosaic floors (Mijović 1984–1985, to the mid-5th century (Dintchev 1997, 70–72; Mulvin 79–80). The brief publication relates that two large rooms 2002, 97–98). It had an enclosed courtyard plan, with and baths were investigated, though building traces were residential parts surrounding the courtyard. visible across a wider area. The mosaic floor, coins and The villa at Bela voda near Pernik presum- other small finds date the vil a to the 6th century or the ably dates from the late 3rd/early 4th to the 7th century Justinian period. Late Antique remains also came to (Dintchev 1997, 53–54; Mulvin 2002, 99–100). Dintchev light on the nearby island of Stari Ulcinj, which could writes that the unearthed remains belonged to a fortified have served as the refuge for the inhabitants of the vil a. central building of a vil a that developed into a vil age after the second half of the 5th century. Kosovo The remains of a luxurious Late Antique vil a were found at Donje Nerodimlje (Kovaljev 1990). The vil a is poorly known, with only four rooms with mosaic floors investigated. The mosaic in the central part of the main hall represents the Seven Sages of Greece. A round pool was found in the centre of the reception hal , while the surviving stairs indicate an upper storey. The construction of the vil a was initial y dated to the 4th–5th century, though recent analyses show the mosaic was most likely made in the Justinian period (Djurić, S. 1994; Sodini 1997, 440). The lavish and high-quality interior furnishings of the rooms point to a vil a or even palace belonging to a member of the elite. Fig. 3.50: Pleven. Plan of the vil a (from Mulvin 2002, Fig. 54). 197 Slavko CIGLENEČKI Fig. 3.51: Obelija. Plan of the vil a (from Mulvin 2002, Fig. 49). North Macedonia The site of Trpčeva Crkva near Dunja ( Fig. 3.52) revealed the remains of a large residential complex composed of living quarters and economic facilities (Kepeska, Kepeski 2006). The wal s were built of clay- bonded stone and the building roofed with tegulae and imbrices. The rooms were arranged around the central paved court. The economic facilities held grain pithoi, later also two water basins. The complex is believed to have been constructed in the late 5th century and pre- sumably persisted to the late 6th century. The remains of an Early Christian church and baptistery came to light nearby. Evidence indicates the unearthed remains at this site formed part of a small vil age. The rescue investigations at Pešterica near Prilep unearthed a characteristic Roman villa, most likely built towards the end of the 3rd century and comprising residential, economic and thermal areas. Following a presumed fire in the mid-4th century, it was remodelled and subdivisioned with clay-bonded wal s. The vil a was destroyed most likely in the last third of the 4th century (Kepeski 2008). Greece The knowledge of the countryside settlement in Fig. 3.52 : Trpčeva Crkva near Dunja. Residential and produc-Greece primarily comes from several systematic field sur- tion parts of the complex (Kepeska, Kepeski 2006, Fig. 3). veys (overview in Bintliff 2012, 352–382). Some authors (Morisson, Sodini 2002, 177; Bintliff 2012, 359) posit that 198 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.53: Akra Sophia. Plan of the vil a (Gregory 1985, Fig. 3). individual vil as persisted into the 6th century or were The intensive survey in Boeotia has shown that only constructed at that time, whereas Florin Curta sees rural settlements in the hinterland replaced Thespiae in Greece merely a landscape of individual farmsteads in its role of a city in the 6th century (Poulter 2007b, 41; and presumes that all the vil as dated to the 6th century Bintliff 2012, 357, 360). were actual y constructed earlier and later reinhabited as smal and modest habitation units (Curta 2014, 40). John A large number of Early Christian basilicas were Bintliff believes the vil a landscape in Greece survived detected in Greek Macedonia, without associated settle- until the devastating Slavic incursions (Bintliff 2012, 358). ments in the vicinity. The distribution of these basilicas only makes sense in connection with as yet undetected The intensive field surveys at Akra Sophia unfortified and hil top settlements (Dunn 2005, 274). ( Fig. 3.53), south of the Isthmus of Corinth, revealed the In the area of the civitates Pel a and Europos, field remains of a luxury vil a from the Early Byzantine period surveys detected as many as nineteen rural settlements, (Gregory 1985; opposing view in Curta 2014, 38). The c. of which five are believed to have been established in Late 80 × 35 m large edifice also held the remains of a mosaic Antiquity (Dunn 2004, 542). The plain at Philippi also from the 5th or 6th century. The recovered pottery and reportedly revealed at least six unfortified settlements amphorae date to the mid-6th century, though the vil a is associated with better-known Early Byzantine churches. believed to have persisted to the early 7th century. In its proximity, there were several sites with concentrations The research in Greece shows very lively country- of small finds but almost without architectural remains; side settlement, which is primarily due to its position archaeologists only found two cisterns or wel s. Timothy deep in the south of the area under discussion that was E. Gregory believes this was not a vil a of an agrarian less and later exposed to barbarian incursions. However, character, but rather a residence, possibly of a high of- this settlement is mainly known through field surveys ficial or rich shipowner from Corinth. that do not enable more detailed conclusions as to the The intensive field surveys in the valley of Berbati nature, size and duration of individual sites; these can in Argolis unearthed a large vil a with mosaic floors only be gained through systematic excavations of the surrounded by a cluster of sites seen as worker accom- characteristic settlement units. modations and industrial areas (Bintliff 2012, 358). 199 Slavko CIGLENEČKI 3.3.2 NEWLY-FOUNDED cal y tackle the question of whether the large and most FORTIFICATIONS significant sites should be seen either as newly-founded cities or merely large settlements that literary sources In the face of numerous civil wars, the increas- mark with different terms. This division is often arbitrary ingly frequent incursions of peoples from beyond the and some of them have already been discussed above, borders and the inability of the central government to as newly-founded cities (such are the examples of both tackle these issues, people in Late Antiquity reacted by large imperial palaces and several large settlements with transforming their settlements and also resettling; this a series of urban elements). Their typological diversity is a phenomenon observable across large parts of the and the definition of their function will be discussed in areas under discussion. They gradual y abandoned set- the closing part of this chapter, after the overview of the tlements in the lowland and along major roads to move characteristic sites. to natural y protected elevations, most frequently in The dating of the fortified settlements is largely remote areas. Across a large part of the former empire, based on coins and other diagnostic artefacts. Related to fortifications (particularly those on well-protected hil s) this is the question of residuality, which is insufficiently are thus the most typical expression of the contemporary considered in early literature and overemphasised in settlement pattern and most aptly il ustrate the plight more recent publications, none of which is helpful. of the population. This category of Late Antique settle- In recent studies, analyses of pottery (fine tableware, ment is most numerous ( Figs. 3.54, 3.201) and also best coarseware and transport vessels) allow us to enhance investigated, and comprises sites ranging from the very the previous dating and provide a more reliable dating rare fortified palaces and small fortified vil as to the of individual phases. more numerous military forts, some newly-founded The discussion below will use examples of settle- ‘cities’ and by far the most numerous fortified hil top ments to present the great diversity of the settlement settlements and refuges. forms, with the examples ranging from major and well- Fortifications are a highly diverse category as investigated to minor and poorly investigated sites with their form, size and often also interior layout is greatly functional y and chronological y significant elements; adapted to the terrain. Each one is practical y unique, settlements exhibiting already known patterns are not in contrast to the earlier settlements in the countryside included. that come with a readily identifiable set of common features. This diversity is the cause of some difficulties for the researchers in identifying their basic functions, WESTERN PART as a more detailed interpretation can only be given by performing a comprehensive analysis of the geographic Italy setting, architectural remains, small finds, economic resources and other aspects. Research shows that some In recent decades, research in northern Italy fo- of these settlements began as early as the second half cused on identifying the role and significance of the of the 3rd century and in some places persisted, with fortified sites mainly located in the Alpine and sub- minor or major interruptions, to the late 6th century or Alpine regions. even later. Interestingly, those in strategic or otherwise significant locations served not only as refuges and The southernmost examples are those around Lake settlements, but also hosted garrisons of soldiers. Their Garda. Particularly complex is the settlement and de- great diversity in both time and geographic location is fensive features of the slender Sirmione Peninsula ( Figs. the root cause for a major part of the professional public 3.55, 3.56), which reaches far into the lake that provided poorly understanding their appearance, functions and excellent natural protection (Roffia 1997, 141–169; Bro- chronology. Most of the syntheses of Late Antiquity giolo 2018). In the Roman period, two large vil as were (particulary those written from a historical standpoint) constructed on the peninsula, at Grotte di Catulo and provide a generalised and simplified account, in which Via Antiche Mura. Both were destroyed in the mid-3rd the fortifications – if at all mentioned – are presented century, though the latter was extensively renovated in as a generic type of hil top settlements, which in reality the late 3rd or early 4th century using the material from only corresponds with a very limited segment of the the abandoned vil a. In the second half of the 4th cen- known sites (cf. division in Wickham 2005, 479–480). tury, Sirmione was transformed into a large fort with The complexity of this issue, with a combination of dif- its centre in the vil a of Via Antiche Mura, while a vast ferent types of sites in different periods of time, is further cemetery grew in the area of the other vil a, at the tip of exacerbated by confusing terminology (see Chapter 1). the peninsula. Three-metre thick wal s were constructed Most of these fortified sites have seen only limited in the late 4th or early 5th century that reached from one investigations, which hinders us in defining reliable vil a to the other and enclosed a surface of 26 ha. They criteria for their classification. We will therefore specifi- appear to have protected a strong army post associated 200 3. SETTLEMENT IN THE COUNTRYSIDE 6 7 43 44 42 45 13 12 14 15 16 17 5 9 18 10 19 20 30 21 4 8 22 31 3 11 24 25 32 33 34 23 29 35 36 2 26 37 1 27 38 39 28 40 47 41 46 48 57 58 49 59 50 51 60 52 61 62 53 63 66 54 64 65 67 55 56 68 69 71 70 72 0 100 km 1 − Sirmione 16 − Tscheltschnigkogel 32 − Gradec (Prapretno) 47 − Veli grad 61 − Gradina 2 − Rocca di Garda 17 − Hemmaberg 33 − Tinje 48 − Korintija (Vašarovine) 3 − Sant’Andrea di Loppio 18 − Hoischhugel 34 − Marija Gorska 49 − Tribanj Šibuljina 62 − Gradac 4 − Riva del Garda 19 − Tonovcov grad 35 − Ajdovski gradec 50 − Veliki Sikavac (Potočani) 5 − San Martino di 20 − Ajdna (Vranje) 51 − Gradina (Modrić) 63 − Gradina (Vidoši) Lundo/Lomaso 21 − Kranj 36 − Veliki vrh 52 − Čuker (Mokro polje) 64 − Gradina (Kovaći) 6 − Georgenberg 22 – Mali Njivč 37 − Kuzelin 53 − Gradina (Vrgada) 65 − Gradac (Lepenica) 7 − Knal wand 23 − Castelazzo (Doberdo) 38 − Velike Malence 54 − Toreta 66 − Debelo Brdo 8 − Castelraimondo 24 − Sv. Pavel 39 − Zidani gaber 55 − Gradina (Žirje) 67 − Gradac (Ilinjača) 9 − Invillino 25 − Hrušica 40 − Kučar 56 − Split 68 − Gradina (Biograci) 10 − Artegna 26 − Ajdovščina (Castra) 41 − Crkvišće Bukovlje 57 − Gradina Zecovi 69 – Mogorjelo 11 − Attimis 27 − Ajdovščina (Rodik) 42 − Velem – Szent Vid 58 − Gradina (Bakinci) 70 − Majsan 12 − Kappele near Jadersdorf 28 − Pasjak 43 – Tokod 59 − Grad 71 − Koštur 13 − Lug ins Land 29 − Korinjski hrib 44 − Sibrig domb (Gornji Vrbljani) 72 − Gradac (Todorovići) 14 − Duel 30 − Ančnikovo gradišče 45 − Sümeg 60 – Gradina 15 − Ulrichsberg 31 − Rifnik 46 − Kastrum (Veliki Brijun) (Podgradina Kamenska) Fig. 3.54: Newly-founded fortifications in the countryside from the western part, discussed in the book. 201 Slavko CIGLENEČKI est timber housing was erected in its ruins. The small finds from the 5th and 6th centuries show that the vil a at Grotte di Catullo was partly reinhabited. Two forts were built on peninsula in the second half of the 5th century: a small one with high-quality housing that is believed to have hosted the military headquarters and a larger one that included a church. At this time, Sirmione presumably hosted an Ostrogothic garrison of soldiers and their families. After the Gothic Wars, the Lango- bards chose Sirmione as an important base, which is indirectly reflected in the rich burials unearthed in the ruins of the vil a at Grotte di Catullo that continue to the late 7th century. The hill of Rocca di Garda high above Lake Garda holds the traces of a multi-period settlement intensely occupied in Late Antiquity (Brogiolo, Gelichi 1996, 88–92). On three sides, the summit is protected with precipitous slopes and only readily accessible from the east, where thick wal s were put up. Not much is known of the settlement’s interior, though excavations did un- earth a large Late Antique building. Mentioned on the slope below are the remains of a cemetery from the 7th century, which indicates the Langobards were living in the settlement upon arriving to Italy. Available evidence allows us to hypothetically include Monte San Martino – Riva del Garda (San Martino di Campi), on the north shore of Lake Garda, Fig. 3.55: Sirmione. Plan of the fortified complex (Vil a 2020, in the group of fortified settlements (Possenti 2013, 28, Fig. 4). 35; Brogiolo 2014, 149–151; Pisu, Possenti 2020). This high and natural y excellently protected peak ( c. 800 m asl) and its southern slope revealed a settlement that with the harbour of the lake fleet and also offered shelter grew on the spot of an earlier sanctuary. Several of its for the local population. Archaeological evidence points buildings have been investigated, as well as a drystone to a fire in the mid-5th century, which may have been wall that served defensive purposes (Brogiolo 2014, 151). the consequence of a Hun raid. The vil a at Via Antiche The houses forming a compact row on the slope may also Mura was also destroyed in this time, after which mod- have served as defence. The mass of small finds from Fig. 3.56: Sirmione. Fortified complex on the peninsula from the west with Rocca di Garda in the background (2003). 202 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.57: Monte San Martino di Lundo/Lomaso. Plan of the Fig. 3.58: Sant’Andrea di Loppio. Plan of the fortification fortification (Zagermann, Cavada 2014, Fig. 5). (Maurina 2016, Fig. 536). the 4th to the 6th century shows an intensely inhabited century. The function of the fortification can as yet not settlement with different trading contacts. In addition be ful y explained. The absence of the pieces of female to a civilian function, the large building occupying a and children’s costumes, as well as a short duration of dominant position on the eastern slope and interpreted the buildings suggest a strategic significance of the fort as a horreum, but also the location in a strategical y manned with a garrison of soldiers, but also the pos- significant area along the communications leading from sibility it served as a protected storehouse on the road Lake Garda to the heart of the Alps show that the set- from the Alps towards major cities in northern Italy. tlement may also have had logistical and military roles. Following a fire in the 6th century, the settlement was partly renovated and a small church constructed in it. The systematic investigations at Monte San Mar- tino di Lundo/Lomaso (985 m asl; Fig. 3.57) revealed a 125 × 80 m large fortification (Zagermann, Cavada 2014). It was encircled with 0.8–1 m thick wal s, with a pair of towers additional y protecting the entrance. Built together with the defensive walls were several buildings that leaned onto the wal s from the interior. An Early Christian church was also built at the wal s. It was a single-aisled apsidal edifice with a narthex. It witnessed numerous modifications, which hinder us from establishing whether the narthex and side rooms were constructed simultaneously with the main part. Small finds and radiocarbon dates limit the construc- Fig. 3.59: Sant’Andrea di Loppio. Fortification from the south tion period to the second half of the 5th and the 6th (2016). 203 Slavko CIGLENEČKI Fig. 3.60: Sant’Andrea di Loppio. Remains of the buildings in the north-western part of the fortification (2016). The fortification at Sant’Andrea di Loppio ( Figs. tioned here after the end of the Gothic Wars. The study 3.58−3.60) is another typical example of a well-fortified of the numerous ceramic finds, primarily amphorae, has strategic post with military assignments (Maurina 2016; shown state provisioning of the army ( annona). After ead. 2020). Its particular feature is the location on the the fort had been abandoned in the second half of the island of Lake Loppio, which provided excellent natural 6th century (Period IV), the newly-constructed modest protection, but also offered the garrison the advantage dwellings point to a different population, presumably the of quick reactions in control ing the traffic along the Langobards (Period V). It would appear, however, that important route that connected the Adige valley with the the fort retained its military character and persisted as northern part of Lake Garda. The walled surface on the such to the second half of the 7th century. top of the island measured c. 0.64 ha. The defensive wal s The systematic investigations on Colle Santino were partial y investigated, found to be 0.7–0.8 m thick above Invillino in Carnia ( Figs. 3.61, 3.62) have un- and buttressed. The entrance was in the north and well- earthed the remains of a Late Antique settlement iden- protected with a large tower. The remains of the earliest tified as the fortification Paul the Deacon mentions as Late Antique buildings of stone and timber (Periods I Ibligo (Bierbrauer 1987; id. 1988). A team of German and II) date to the second half or even late 5th century archaeologists conducted extensive archaeological and presumably accommodated those who were build- excavations on the hill in the 1960s that explored the ing the later fort. Proper defensive wal s were likely con- entire settlement. Invillino remains to this day the only structed towards the end of the 5th century, reportedly comprehensively excavated Late Antique settlement on the orders of the central administration (Period III). in the wider area of the eastern Alps, which is of great This period also witnessed gradual construction of hous- value for our understanding of the settlement pattern. It ing along the wal s in both investigated parts of the fort. lies on the natural y very well-protected, isolated hill of Best researched are the houses in Sector A, where a row Colle Santino, which rising above the plain of the River of densely spaced buildings of a high-quality construc- Tagliamento. The remains of earlier habitations are two tion was detected next to the entrance. The small finds complexes of buildings presumably dating from the 1st place the construction of the fort to the time of either to the mid-4th century, but cannot be reconstructed due Odoacer or Theoderic, with its strategical y significant to later modifications (Bierbrauer 1987, 291–292). They location indicating a primarily military role. The finds represent the remains of a small countryside settlement, of weapons confirm such a role, as do the coins dating to which could have been established as late as the second the time of the Gothic Wars. Some small finds indicate half of the 3rd century (Martin 1992, 261). In the second the presence of Byzantine troops with their families sta- phase, dated to the second half of the 4th and first half 204 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.61: Colle Santino (Invillino). Plan of the settlement in the Roman period and Late Antiquity (Bierbrauer 1988, Fig. 39). of the 5th century, the buildings retained their exterior a baptistery, constructed in the first half or middle of the appearance, but changed in function. In addition to 5th and persisting to the first half of the 7th century, when numerous finds of tools, there were also substantial it was burnt down. Excavations also unearthed part of the traces of ironworking and glassworking, i.e. artisanal associated burial grounds, which was in use even before activities. The imported pottery points to connections the construction of the church complex. The results of with the Mediterranean. The end of this phase brought the investigations on both hil s show this to be a major radical changes in the settlement structure signalled by settlement and religious centre, which is visible in the timber architecture dating from the first half of the 5th architecture and exposed location, but also in some of the to the second half of the 7th century, but also a masonry smaller settlement cores investigated in a limited extent tower in the area of easiest access. The buildings of this in the surrounding area, primarily those at Cuel Budin phase are simple, long and rectangular with drystone and Verzegnis (Rupel 1997; Vil a 2001, 828–829, 858–860; foundations and a timber superstructure. They have Vannacci Lunazzi 2001, 85–86; Ciglenečki 2003a, 268). been interpreted as dwel ings, while Bierbrauer sees The fortification at San Giorgio near Attimis lies two as workshops. A cistern has also been attributed on the road that led from Cividale to Artegna, Gemona to this phase. Excavations on the nearby hill of Colle di and Osoppo. It is only known from preliminary reports Zuca revealed an associated large church complex with (Vil a 2003; id. 2006, 164–169; Buora 2010). Excavations 205 Slavko CIGLENEČKI Fig. 3.62: Colle Santino (Invillino). Plateau above the River Tagliamento from the south-east (2002). revealed part of the defensive wal s and several buildings, as well as small finds and Ostrogothic coins that date its existence to the first half of the 6th century. Publications reveal that the wal s were made of local y available stone, mortared and 0.8–0.9 m thick (Vil a 2003, 305). Dwel - ings survived as clay-bonded wal s that were likely the foundations of timber buildings. Researchers see the military finds and Gothic coins as an indication of a garrison stationed in the fort. More is known on the fort at Artegna ( Figs. 3.63, 3.64). Similarly as San Giorgio, it lies on the fringes of the Friuli plain, on a low, but natural y well-protected hil . It has defensive wal s that incorporate two towers, one Fig. 3.63: Artegna. Wel -protected hil hosting a fort, view pentagonal and the other triangular, which are unusual from the south (2010). for the area (Vil a 2006, 162–164, Figs. 5, 6). Towers of these forms were frequently erected by Byzantine builders in the Eastern Empire and can be found in all major Early Byzantine forts and settlements (Lawrence 1983, 177–200). Close parallels are also known at nearby Cividale (Vitri et al. 2006, 103). The heavily fortified wal s and a large, well-preserved cistern in the interior show this was one of the most important Late Antique forts in Friuli (also mentioned by Paul the Deacon) and one that also had a special y protected acropolis. A well-researched, but more modest is the fortified hil top settlement at Castelraimondo above Forgaria ( Figs. 3.65, 3.66), which was an exception in that it was constructed on an earlier predecessor (Santoro Bianchi 1992). The lower ridge has three peaks separated from one another with in part artificial y made ditches. In Fig. 3.64: Artegna. Cistern in the interior (2010). the Roman period, a square tower was constructed on a dominant location on top of prehistoric remains and several other buildings were renovated. A major break in 206 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.65: Castelraimondo. Plan of the hil top site (Santoro Bianchi 1992, Fig. 25). Fig. 3.66: Castelraimondo. Remains of the multi-period main building (2003). habitation is perceptible around 275, when the tower and is then believed to have been a refuge and persisted as two earlier buildings were damaged. This event is well- such throughout the 6th and part of the 7th century. dated with the coins of Quintil us, Florinus and Probus. The fortification at Castelazzo above Doberdo The buildings were soon repaired and the tower served del Lago ( Fig. 3.67) is an excellent example of reuse as a military fort throughout the 4th century. It was torn of a prehistoric hillfort (Furlani 1969; Maselli Scotti, down around 430, presumably in a military assault and Montagnari Kokelj 1989). The natural y well-protected fire that destroyed a large part of the buildings. In the hill guarded in the south by precipitous rocky faces was next phase, in the second half of the 5th century, the the site of a strong fort in Late Antiquity that was located ruins were reused for modest housing. The settlement in close proximity to a major Roman road. The fortified 207 Slavko CIGLENEČKI Fig. 3.67: Castelazzo above Doberdo del Lago from the south (2015). plateau measuring 260 × 120 m suffered heavy damage mostly conducted here in the 1960s and again recently, in World War I, but archaeological investigations nev- with the results being jointly published in a monograph ertheless managed to find several interesting elements. (Lang 2019). The top of the hill was inhabited or visited One is the three-metres-thick defensive wal s, the cross in many periods. In Late Antiquity, more precisely the section of which shows the prehistoric rampart twice 460s and 470s according to the written source, it hosted reinforced with a mortared wal . Earlier authors report a settlement with a church that Severinus visited on two of several towers and other buildings, which together occasions. Excavations on the summit, in the vicinity of with the recovered small finds and the location along the present-day church, unearthed a building with a heat- the road indicate a strong military post from the 4th ing channel and another Late Antique edifice on the other and early 5th centuries. Standing out among the many side of the church. The interior of the present-day church small finds is a multitude of coins (scattered hoards?) also revealed several earlier constructions, but these could from the 4th and first half of the 5th century. Small finds not be more reliably identified. The burials inside this and individual coins allow for the continuation of the church include a skeleton radiocarbon dated to the 5th/6th fort into the second half of the 5th and the 6th century. century. A Late Antique origin of the defensive wal s could not be reliably confirmed, though it is likely, according to trial trenches at the point of easiest acces. Part of a Late Austria Antique cemetery was found at the foot of the hil . Smal finds recovered from the settlement paints an unusual The Late Antique hil top settlements in the eastern picture: the monetary circulation peaks in the second half Alps have been first identified as an important settle- of the 4th and early 5th century and even other small finds ment factor in Austria, in the early 20th century. The do not reach beyond the early 5th century. Finds from the many early excavations joined by several more recent 6th century are completely absent. As for the metal finds, systematic investigations provide a good insight into parallels from other sites in the eastern Alps only show the Late Antique settlement structures in both Norican the settlement existed to the mid-5th century. provinces of Late Antiquity. The mountainous part of the Enns valley hosts several Late Antique sites. The most revealing is at Knall- A pride of place among the Late Antique fortifica- wand near Ramsau ( Figs. 3.69, 3.70), located at 944 m tion north of the Alpine arch goes to Georgenberg near asl below the Dachstein Mountains (Steinklauber 2005, Kuchl (Cuculis) ( Fig. 3.68). It is one of the rare such sites 142–157). This hill is natural y protected from three sides in the area under discussion that is debated in greater and only accessible from the north, where 0.6 m thick detail in an ancient written source (Vita Sancti Severini, wal s were constructed. Leaning onto them from the Chapters 11 and 12; cf. Ubl 1982, 85). The settlement interior are two rooms that held a layer of burnt debris. lies on a 400 × 150 m large elongated plateau natural y The 75 × 30 m large fortification interior held traces of a protected from all sides and located near the Roman road wall that suggest a bastion-like building. Investigations that connected Iuvavum and Teurnia. Excavations were revealed no other buildings. The many recovered coins, 208 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.68: Georgenberg near Kuchl from the north-west. Fig. 3.69: Knal wand near Ramsau. Plan of the fortification Fig. 3.71: Kappele near Jadersdorf. Plan of the fortified settle- (Steinklauber 2005, Fig. 10). ment (Felgenhauer 1993, Fig. 2). pieces of costume, iron tools and pottery date the site to the 3rd–5th century. The traces of a fire are believed to have signified the end of the settlement in the mid-5th century at the latest. The question that remains open is whether the site was a refuge of a long duration, a short-term settlement from the second half of the 4th century, the post of a small garrison stationed here in the second half of the 3rd and again in the second half of the 4th century or even a combination of all three. Located between the valleys of the Gail and the Drau/Drava are the remains of a fortified settlement at Kappele near Jadersdorf ( Fig. 3.71) (Felgenhauer 1993). It covered a surface of 90 × 90 m with remains of Late Fig. 3.70: Knal wand near Ramsau. Hil top site from the north Antique buildings in part visible on the surface. Excava- (2007). tions unearthed well-constructed masonry edifices, but 209 Slavko CIGLENEČKI also timber buildings, which suggests a social differentia- tion of the inhabitants. Certain metal and ceramic items point to lively trading and diverse artisanal activities that include metalworking. The fortification at Lug ins Land ( Fig. 3.72) lies near Molzbichl, high above the Drau/Drava valley (Gostencnik 2000). It is 90 × 30 m large and natural y excellently pro- tected. Two Late Antique phases have been established. In the first phase, drystone buildings with loam floors were erected on the plateau and later burnt down; the quantity of the recovered pottery suggests permanent settlement. In the second phase, a rampart was built on the north side where access was easier, while terraces with stone retain- ing wal s were created on the plateau. The summit was girded with substantial, 1.2 m thick mortared wal s. This phase also yielded only Late Antique pottery that cannot be more precisely dated. Lug ins Land has a good visual communication with Duel and Teurnia, hence could have served as a signalling post and also offered permanent shelter for a small group of people. The systematic investigations conducted between 1928 and 1931 on the hill of Duel near Deutsch Feistritz ( Fig. 3.73) revealed a large Late Antique fort on a strategic location on the right bank of the Drau/Drava (Egger 1929; Petrikovits 1985, 236–237; Steinklauber 2013). The investigation results are only partial y published and do not offer a more integral insight into the site. The fort was constructed on a roughly 250 × 110 m large plateau with very steep slopes on all sides and additional y protected Fig. 3.72: Lug ins Land. Plan of the hil top site (Gostencnik with a stream. The 0.9 m thick defensive wal s have two 2000, Fig. 1). Fig. 3.73: Duel near Deutsch Feistritz. Plan of the fortification (Glaser 1996, Fig. 35). 210 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.74: Hoischhügel near Maglern. Plan of the fortification (Glaser 1996, Fig. 31). phases that correspond with two thick layers of burnt debris. A particularly interesting construction feature is the triangular west terminal of the second-phase wal s. Such terminals are a characteristic of Early Byzantine fortifications. Access to the fort from the north runs along a raised ramp and the entrance is reinforced with a large tower. An Early Christian church occupies the highest point of the fort interior, flanked by another large building with a baptistery (presumably also a church). Three smaller buildings were also unearthed, the function of which could not be reliably identified. Many masonry buildings leaned against the defensive wal s from the interior, most of them interconnected; Egger marked some of them as towers on the plan of Fig. 3.75: Hoischhügel near Maglern. Substantial wal s of the the site. A vast part of the interior was not built-up, fortification (2023). though postholes suggest makeshift buildings. The fort is believed to have been established in the 5th century, while the form of the defensive elements suggests the The poorly known site at Hoischhügel near second phase could be attributed to the time of Justin- Maglern ( Figs. 3.74, 3.75) can also be seen as a military ian’s reconquista (Egger 1929, 211–212; Petrikovits 1985, fort (Egger 1916, 97–104; Steinklauber 2013, 55–62). 236–237). The later analysis of the small finds does not It was constructed along a major Roman road and is oppose such dating (Steinklauber 1990, 124; ead. 2013, similar in size to the fort on Duel (210 × 90 m). It has 63–64). The fort controlled the traffic along the Drau/ substantial wal s running in fairly straight lines and in- Drava valley and provided shelter for the local popula- corporating towers and buildings leaning on them from tion in times of danger. the interior. The fort hosts the remains of a large Early Christian church. Excavations were mainly conducted here in the early 20th century, later there was only lim- 211 Slavko CIGLENEČKI Fig. 3.76: Ulrichsberg. Plan of the hil top settlement (Egger 1950, Fig. 6). Fig. 3.77: Ulrichsberg from the south-west (1999). ited trial trenching conducted in 1982 and 1983, hence fortification boasts a highly exposed strategic location on many a question remains open. The time of the fort a major Roman road, it is natural y well-protected and construction can also not be identified more precisely; affords unimpeded view over vast areas. The defended similarly as for Duel, Egger presumes this occurred in surface measured 270 × 180 m and was only protected the late 4th century (Egger 1929, 214) and the settlement with 0.9 m thick wal s in two areas where access was likely lasted to the late 6th century (Steinklauber 2013, easier. In the east, defensive wal s protected the main 64). Considering its strategic position, the garrison in entrance to the summit that was flanked with a pair of the fort certainly performed demanding tasks in the 6th towers. A single building was excavated in the interior; century, which can be indirectly inferred from the hoard it had two rooms, the larger of which had heating. Two finds of solidi from the late 6th century. timber buildings presumably leaned onto the defensive Investigations in the fortification on Tscheltschnig- wal s on the west. The interior revealed a multitude of kogel (Kadischen) were conducted before World War II coins and other small finds that roughly date the site to and only published in preliminary reports (Dolenz, Gör- the 4th–6th centuries. Pieces of church furnishings also lich 1935; Glaser 1996, 66–68; Steinklauber 2020). The indicate the existence of a church. Found below the set- 212 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.78: Hemmaberg above Globasnitz. Plan of the hil top settlement (Glaser 2008, Fig. 10). Fig. 3.79: Hemmaberg above Globasnitz from the nort-west (1995). tlement was an Early Christian cemeterial church with 59–60). It is a clustered settlement on a protected terrace several burials. just below the summit, composed of different houses and The high hill (1022 m asl) of Ulrichsberg above an Early Christian church. The settlement is excellently Zollfeld ( Figs. 3.76, 3.77, 3.303) holds the remains of natural y protected in the east, but lacks defensive wal s a Late Antique settlement (Egger 1950; Glaser 1996, in the west. A later survey has shown a chronological y 213 Slavko CIGLENEČKI Fig. 3.80: Hemmaberg above Globasnitz. Remains of the western double church (2017). undetermined rampart located at a greater distance Hungary from the central part of the settlement and enclosing a much larger area (Glaser 1996, 58, Fig. 24). The area The eastern foothil s of the Alps at the western has not been investigated and it is not clear whether the fringes of the Pannonian Plain hold the remains of a settlement was actual y larger or was the additional area typical, but poorly known example of a fortified hil top intended for sheltering people from the lowland. Only settlement at Velem – Szent Vid (Quast 2011; Filipec et a rough dating can be given, to the 4th–6th centuries. al. 2020, 201–205). This is a large multi-period site with The hilltop settlement on Hemmaberg above several horizons of a prehistoric fortified settlement, but Globasnitz ( Figs. 3.78−3.80, 3.311) is located at the also habitation remains from the 4th and 5th centuries. north foot of the Karawanken/Karavanke Mountains, This high (569 m asl) and natural y well-protected hill where long years of investigations unearthed an ec- was already fortified with a rampart in prehistory, when clesiastical centre that was presumably also a pilgrim- it hosted numerous habitation terraces used and modi- age site (Glaser 1991; id. 1997, 96–120; id. 2008, 617; fied also later. The summit plateau revealed a timber Ladstätter 2000; Eitler, Seidel 2022). The hill rises some building from the second half of the 4th century. Publica- 300 m above the surrounding plain and is well-protected tions mention Late Roman finds and stray finds from the with rocky slopes, while the more readily accessible 5th and 6th centuries. The area below the summit plateau south side is protected with wal s. The vast summit, revealed several inhumation burials with artificially measuring roughly 10 ha, has so far revealed as many deformed skul s, dating to the 5th century. as six Early Christian churches and several residential The long frontier of the Roman Empire along the buildings, the function of which has been related to the Danube ( Ripa Pannonica) is marked by a series of for- ecclesiastical centre. Housing is believed to have been tresses exposed to incessant pressure from barbarian arranged along the defensive wal s. The first church was peoples from the north and east (Soproni 1978; id.1985; presumably erected in the early 5th century, two double Visy 2003). Research has shown an extensive renovation churches in the early 6th century and another church of the limes under Diocletian and again under Valen- under the Frankish rule. The settlement remains largely tinian, which was mirrored in the contemporary con- unexplored and further research may shed light on the struction of forts in the hinterland. The reinforcement function of the complex; the site may even have func- and renovation of these fortifications in the 4th century tioned as an ecclesiastical administration centre. The represent the efforts of the central government to protect associated necropolis dates to the 6th century. this vital part of the Empire. The group of forts built in 214 3. SETTLEMENT IN THE COUNTRYSIDE Danube, well-protected on three sides and affording an excellent view over a wide stretch of the limes in the Danube knee (Soproni 1978, 55–58). The fort was built in the 310s, partial y modified under Constantine II and again under Valentinian, and final y abandoned around 378. Buildings were only observed along the defensive wal s, while the interior was found to be empty. In the second phase, a large square tower was constructed at the entrance, which presumably retained its function as a watchtower long after the fort had been abandoned. It is not clear whether some of the Pannonian vil- las with fortifications features (primarily towers) also belong to the group of newly-founded fortifications; the difficulty is that it is unclear whether these features are of a merely prestigious or also a defensive nature (Thomas 1980, 293–296). Considering their location in the lowland and often in the vicinity of other unfortified complexes, their defences seem rather symbolic, particularly if com- pared with the contemporary and well-protected ‘inner Fig. 3.81: Sibrig domb near Visegrad. Plan of the hil top site fortresses’. As a typical example of a fortified Pannonian (Soproni 1985, Fig. 27). vil a, we will therefore only present that at Sümeg, north of Lake Balaton (Thomas 1964, 111–116; ead. 1980, 312; Mulvin 2002, 104). The vil a was presumably already built in the 3rd century, with its fortified design pointing to the the 4th century shows a different choice of location with late period of vil a construction. Its defensive wal s are regard to the earlier ones, now sited in higher-lying reinforced with a protruding rectangular tower in each and natural y well-protected locations, similarly as we of the four corners. The interior holds houses arranged observe across the hinterland of the eastern Alps and along the defensive wal s and around a vast courtyard. the Balkan Peninsula. A large rectangular building of an east-west orientation An example of a fairly typical military fort with pro- was found in the vicinity; this was presumably an Early truding towers is at Sibrig domb near Visegrad ( Figs. Christian church dated on the basis of the construction 3.81, 3.82). It was constructed on an elevation above the technique into the 4th century. Fig. 3.82: Sibrig domb near Visegrad. Tower in the western wal s with the Danube River in the background (2011). 215 Slavko CIGLENEČKI The well-researched fort at Tokod ( Figs. 3.83, 3.84) was created at a major road junction, only 3 km from the Danube (Mócsy 1981; Soproni 1985, 58–60). The 142 × 118 m large fort was constructed on a gentle slope and enclosed with 1.6 m thick wal s reinforced with horseshoe-shaped towers in the corners, semicircular ones along the sides and two rectangular ones flanking the entrance. The interior held a large horreum. The recent GPR surveys has revealed an apsidal building next to the horreum, but also several other buildings (Kocsis 2020). The creation of the fort has been set to the Valentinian period. Some decades later, timber housing and drystone buildings were erected in the interior. The late phase of the fort correlates well with the associated inhumation cemetery from the 5th century. Fig. 3.83: Tokod. Plan of the fort (Kocsis 2020, Fig. 8). Fig. 3.84: Tokod. Remains of the wal s (2011). Slovenia (Ciglenečki 1994; id. 2008; Ciglenečki et al. 2011; Modri- jan, Milavec 2011). This large and complex settlement is Research into Late Antique fortifications in Slo- located in the valley of the River Soča, next to a major road venia has been ongoing since the late 1960s. Many are connecting Carinthia and Friuli, and only some 20 km only poorly known, but there is a fair number of settle- from the city of Forum Iulii (Cividale). The first reliable ments that have been systematical y investigated and human traces from the Roman period can be dated to the also comprehensively published. second half of the 3rd century. They come in the shape of well-preserved bronze finds and coins, while we could not Systematic investigations over the last three decades identify any contemporary architectural remains and the have gradual y shed light on one of the most important existence of defensive wal s is questionable as the prehis- Late Antique centres at the western edge of Slovenia, toric rampart may have been reused for the purpose. The at Tonovcov grad near Kobarid ( Figs. 3.85−3.88) small amount of finds suggests this habitation phase was 216 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.85: Tonovcov grad near Kobarid. Plan of the hil top settlement (Ciglenečki et al. 2011, Fig. 1.7). Fig. 3.86: Tonovcov grad near Kobarid. Main part of the settlement with the excavated residential building (2010). 217 Slavko CIGLENEČKI Fig. 3.87: The position of Tonovcov grad in the gorge behind Kobarid, above the Soča river (1996). Iuliarum defensive system that effectively controlled ac- cess to Italy from the north. The main habitation phase at Tonovcov grad began in the last decade of the 5th century, when a heavily fortified settlement with numerous buildings was established here. The 150 × 90 m large summit was protected with precipitous slopes and defensive wal s. Excavations unearthed roughly twenty masonry buildings, most of them partitioned into smaller rooms. Housing concentrates in the best-protected northern and western parts. Rising above these is a small rocky plateau that holds the remains of three well-preserved Early Christian churches and a memoria. Near the Fig. 3.88: Tonovcov grad near Kobarid. Cistern in the eastern churches are traces of two cisterns, while a third one part of the settlement (2010). was investigated in the eastern, highest part of the set- tlement. In the part of easiest access, the settlement was protected with 0.8 m thick wal s reinforced with a series very short-lived. More intense habitation can be observed of buildings leaning on them from the interior. The site in the second half of the 4th and into the first third of the defences include another wal , running perpendicular 5th century. Attributable to this phase are some of the to the defensive wal s steeply down the slope and en- modest remains of wal s under later buildings. The con- closing a large area in front of the settlement, which siderable amount of military finds indicates the presence provided shelter for people and livestock in times of of an army garrison. In this time, the fort on Tonovcov danger. The large Late Antique building with an ex- grad was among the key points in the Claustra Alpium tension, located in the vicinity of the entrance to the 218 3. SETTLEMENT IN THE COUNTRYSIDE settlement, joined residential, storage and partly even The hilltop settlement on Ajdna above Potoki economic functions and is thus in many respects typi- ( Figs. 3.89−3.92, 3.308) is, in its location, a paradigmatic cal of the period. The multitude of artefacts indicates example of a naturally defended settlement (Leben, this settlement existed from the end of the 5th to the Valič 1978; Sagadin 1987; Vidrih Perko, Sagadin 2004, initial decades of the 7th century. It was a large forti- 219–221; Sagadin 2006). It is sited on a 1048 m asl high fied settlement with an important ecclesiastical centre crag in the Karavanke Mountains that rises as much as that served as a central settlement for a wide area of 500 m above the River Sava. The approximately 120 × the Soča valley and also played a major military role. 10–30 m large settlement spreads on several terraces be- Fig. 3.89: Ajdna above Po- toki. Plan of the hil top set- tlement (Sagadin 1997, Fig. on the cover). Fig. 3.90: Ajdna above Po- toki. Crag from the south- west (2021). 219 Slavko CIGLENEČKI Fig. 3.91: Ajdna above Potoki. Main part of the settlement with residential buildings and the church (2009). Fig. 3.92: Ajdna above Potoki. Ruins of the residential buildings high above the Sava Valley (2009). low the summit. Its excellent natural protection required one shows that it was destroyed in a fire in the second no additional, man-made protection. The ruins of more half of the 6th century. than ten buildings are visible on the surface, of which The natural y sufficiently protected rocky shelter only the church and four houses have thus far been exca- on Mali Njivč above Novaki yielded a narrowly dated vated. The rich and varied small finds date the settlement range of coins and other items that point to two habita- to the second half of the 5th and the entire 6th century. tion phases (Istenič 2015). The coins from the terraces The different iron tools indicate an autarchic way of life show this ridge was used as a refuge in the third quarter and the rare imported items point to modest trading of the 3rd century, while most other finds reveal the pres- contacts. Excavations on the lower terrace revealed two ence of a garrison stationed here in the last quarter of the solidly constructed houses that both presumably had a 4th century, when it protected the road from Emona to cistern in the interior. A thick layer of burnt debris in Forum Iulii as part of the Late Roman defences of Italy. 220 3. SETTLEMENT IN THE COUNTRYSIDE The dominant rocky ridge of Sv. Pavel above verse wal , bound with mortar, divides the approximately Vrtovin ( Figs. 3.93−3.96, 3.312) holds the remains of a 4 ha large settlement in two differently fortified halves. vast Late Antique settlement (Svoljšak 1985; Ciglenečki The north half is natural y less well-protected and holds 2021). Trial trenching, field surveys, LiDAR data and houses on several-metres-wide terraces. Trial trenching aerial imagery have revealed a settlement with substan- on one of these terraces confirmed the existence of a tial residential, defensive and sacral architecture. The house with solid masonry wal s. The houses took up ridge is natural y protected from three sides and has the entire width of the terraces, while their length has defensive wal s in the exposed parts. A 2.5 m thick trans- not been established. The habitation terraces reach to a small saddle that holds visible remains of buildings and a cistern, on a raised plateau also the remains of a double church. Detectable in the south half are the outlines of numerous buildings of diverse and complex plans, in contrast to those in the north half. This part terminates in the remains of another church with a baptistery. The southern end of the settlement is a rocky area where early sources (Rutar 1886) mention numerous inhumation burials and tombs. In the south, the area protected with thick wal s and precipitous slopes terminates in a large tower erected above a water source and accessed from the settlement via a set of rock-cut stairs. Most small finds date to the 5th and 6th centuries while the thick transverse wall was constructed later in the 6th century as revealed by an early layer below it. Iron tools point to an agrarian nature of the settlement, while the finds of weapons reveal a military presence at a major road leading to Italy from the east and north. The settlement Fig. 3.94: Sv. Pavel above Vrtovin. Tower above the spring on Fig. 3.93: Sv. Pavel above Vrtovin. Plan of the hil top site (from the southern foothil s of the hil top site preserved to a great Ciglenečki 2021, Fig. 2). height (2021). 221 Slavko CIGLENEČKI Fig. 3.95: Sv. Pavel above Vrtovin. Settlement from the south-east (2021). Fig. 3.96: Sv. Pavel above Vrtovin. Wal s on the south-eastern part of the settlement (2021). was important for a wide area and represented a semi- with substantial wal s with fourteen round towers. The urban formation of the 6th century. wal s were 3.85 m thick at the foundations and thinned Castra (Ajdovščina) ( Figs. 3.97−3.99) was the in two steps towards the top. The west side was addition-largest fortress of the Claustra Alpium Iuliarum barrier al y protected with a ditch. Investigations of the wal s, system and presumed military headquarters (Osmuk towers and interior buildings have provided a reliable 1997; Žerjal, Tratnik 2020; Urek, Kovačič 2020). It was dating of construction to the 270s or 280s (Osmuk 1997, established in the lowland, at the confluence of the Rivers 122, 126–127; Kos 2014, 36; Urek, Kovačič 2020, 57). Hubelj and Lokavšček, in the place of the earlier Roman The interior thus far revealed parts of five large build- station Fluvio Frigido. It was roughly polygonal in plan, ings of a predominantly residential nature and smal measured 220 × 160 m (surface of 2.6 ha) and enclosed baths. The central part was taken up by a large complex 222 3. SETTLEMENT IN THE COUNTRYSIDE kanal / sewer Lokavšček Gregorčičeva ul. 1 Lokarjev park aja 8 ta 5. mes 10 C Lokarjev drevored 11 Goriška cesta obzidje / fortification skeletni grob / inhumation zgodnjesrednjeveški ostanki / arhitektura, stavbe / architecture, buildings žgani grob / cremation Early Middle Age remnants antične najdbe / Roman period finds 0 50 m raziskave / investigations 2006–2007 Fig. 3.97: Ajdovščina ( Castra). Plan of the fortress and the Fig. 3.98: Ajdovščina ( Castra). Interior of the fortress with settlement extra muros (Žerjal, Tratnik 2020, Fig. 2). excavated buildings (from Urek, Kovačič 2020, Fig. 1). Fig. 3.99: Ajdovščina ( Castra). Eastern wal s with a tower (2015). that faced the main street and its monumental entrance contemporary pottery and metal items (predominantly opened onto a vast porticated inner court. Only one of weapons) clearly outline the duration of this fortress at its wings has been partial y investigated, revealing a long the foot of the Alps, but also show that life in a limited development with four phases. Numerous coins from extent continued in the first three decades of the 5th the last third of the 3rd and the 4th centuries, as well as century (Kos 2012, 285, 289). 223 Slavko CIGLENEČKI The fortress of Ad Pirum (Hrušica) ( Figs. 3.100−3.102) stands on the highest point of the Roman road across the Hrušica Pass (867 m asl) that connected Emona and Aquileia (Ulbert 1981; Kusetič et al. 2014; Kos 2014; id. 2020). It was elongated oval in plan and en- tirely adapted to the terrain, extending across two peaks and the saddle in between. It covered a 250 × 35–86 m large area. The lower part, in the saddle, was more in- 910 tensely inhabited and also hosted the Roman road. It was girded with 2.7 m thick wal s and divided roughly 900 along the middle with an equal y thick transverse wal . Trial trenches showed it was uninhabited in the upper 890 part, which likely served a refugial purpose in case the 880 lower and more easily accessible part was captured, or served as a place to host a larger army unit. The fortress 870 has three polygonal towers and a substantial square 870 tower at the highest point, which served as a watch post. 870 Two other polygonal towers guarded the east entrance; as much can also be surmised for the opposite, west entrance. Late Antique habitations concentrated in the lower part along the Roman road, where investigations revealed several buildings that included one with an apsidal room and hypocaust heating, a cistern, a square building and a log house. The site is reliably dated with 860 850 numerous coins and other finds (Ulbert 1981; Kos 890 0 50 m 880 870 2020). Some 400 coins have been recovered, most of which fall to the 3rd and 4th centuries. Research of the Fig. 3.100: Ad Pirum (Hrušica). Plan of the fortress (Kos 2020, southwest tower revealed that it was initial y square and Fig. 1). constructed simultaneously with the defensive wal s, destroyed in a massive fire, which coin finds date to the mid-4th century, and then renovated under Valentinian in a sub-pentagonal shape (Svoljšak 2015). The fortress was abandoned in the early 5th century. It is associated Fig. 3.101: Ad Pirum (Hrušica) from the south-east (2022). 224 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.102: Ad Pirum (Hrušica). Northern part of the fortress and the transverse wall with two towers and the guarded entrance (2017). with three differently long walls with towers of the Claustra system. The large hil top settlement at Ajdovščina above Rodik ( Figs. 3.103, 3.104) was established on top of a prehistoric hillfort, on a dominant hill (804 m asl) close to the Roman road connecting Emona and Tergeste (Slapšak 1978; id. 1997). The impressive prehistoric stonework rampart, enclosing a c. 260 × 200 m large area, was reused to defend the fortification. The interior held substantial ruins of large and well-built edifices of complex plans (Mušič 1999) that greatly differ from the contemporary and later Late Antique buildings known Fig. 3.103: Ajdovščina above Rodik. Plan of the hil top site (Slapšak 1997, unpaginated). Fig. 3.104: Ajdovščina above Rodik. Dominant hill with the settlement from the south-west (2022). 225 Slavko CIGLENEČKI Fig. 3.105: Ančnikovo gradišče near Jurišna vas. Plan of the fortification (Ciglenečki 2015, Fig. 10). in the eastern Alpine area. Excavations also unearthed part of a two-phase building with an apse. The rich smal finds indicate the settlement was inhabited from the 4th to the mid-5th century. The imported goods from the first half of the 5th century could be linked to the annona (Vidrih Perko 1997, 349; Vidrih Perko, Župančič 2003, 463–467). Its location just above an important road junc- tion in the immediate hinterland of the Alpine barrier wal s and the excellent view over the surrounding area indicate that we might see its role in connection with military events and also explain its increased significance after the main road from Emona to Aquileia across Hrušica had been abandoned (Slapšak 1978, 546–547). It revealed high-quality constructions and a rich cultural layer. Little is known of its Late Antique defences. Trial trenching of the prehistoric rampart unearthed no later defensive elements and suggested that the old rampart must have sufficed, while the absence of stronger defen- sive wal s was the most likely reason for the settlement being abandoned in the 6th century. The well-investigated hil top fortification from the second half of the 4th and first half of the 5th century include Ančnikovo gradišče near Jurišna vas ( Figs. 3.105−3.107) (Strmčnik 1997, 272–279; Strmčnik Gulič, Ciglenečki 2003a; Modrijan 2017; ead. 2020). It lies on a ridge on the eastern foothil s of the Pohorje Mountains, at 700 m asl. In the Late Roman period, the most exposed parts of its summit were protected with 1.5 m thick wal s and the entrance strengthened with a pair of parallel Fig. 3.106: Ančnikovo gradišče near Jurišna vas. Wal s and the wal s. Two large masonry and several timber buildings strengthened entrance in the northern part of the fortification leaned onto the defensive wal s from the interior, which (2016). was largely empty. The fortification was initial y marked as a refuge, but the defensive architecture and charac- 226 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.107: Ančnikovo gradišče near Jurišna vas. Masonry buildings from the south-western part of the fortification (1998). teristic military finds unearthed later suggest a military nature. The location offers effective control over a vast area, from Poetovio to the many Late Roman posts in the mountainous hinterland of Celeia. The masonry buildings along the most exposed part of the fortifica- tion undoubtedly housed a small garrison and the empty space could have sheltered a large number of refugees and livestock, while the timber buildings could be the dwellings of farmers and shepherds. Rifnik near Šentjur ( Figs. 3.108−3.110) hosts one of the fortified hil top settlements in the eastern Alpine area that was investigated earliest and is today best-known (Schmid 1943; Bolta 1981; Pirkmajer 1994, 46–47; Bierbrauer 2003; Bausovac, Pirkmajer 2012, 33; Ciglenečki et al. 2020, 179–180, 235–236). The domi- Fig. 3.108: Rifnik near Šentjur. Tower and wal s from the 6th nant hill rises high above the surrounding area and is century above the earlier wal s (2003). natural y well-protected from all sides. The 221 m long and 1 m thick wal s protected the settlement in the more easily accessible southern and western parts. The first of the defensive wal s. A large single-aisled church with a wal s without towers were most likely constructed in the baptistery was excavated in the highest part of the settle- late 4th or early 5th century, while rectangular protruding ment, while a smaller apsed church came to light in the towers were added in the 6th century. Investigations at west part of the summit plateau. A large part of the 1.65 ha Tower 3 provided indirect evidence for dating the late large interior, protected with wal s and precipitous slopes, phase of the defensive wal s, showing that a large four- was empty and may either have welcomed people seeking room building from the early 6th century was torn down shelter or was intended for agricultural purposes (gardens, when the defensive walls were modified (Bausovac, pasture). The large quantity of small finds allow us to date Pirkmajer 2012, 33). The interior revealed houses of a the houses and churches from the late 5th to the late 6th well-built and mortared construction. Simple buildings century. The defensive wall reinforcement postdates the with one or two rooms concentrated in the east and only Byzantine victory over the Goths, when the church on the above-mentioned long building – undoubtedly the the highest point was also renovated and enlarged; both dwelling of a small garrison – was located in the most of these features are characteristic of Justinian period vulnerable location. There are also several buildings less renovations of Late Antique forts across a large part of the well-built of poorly durable materials along the interior Byzantine Empire. Of great importance was the discovery, 227 Slavko CIGLENEČKI Late Antique wal s towers added in the mid-6th century 5 4 3 2 0 20 m 1 Fig. 3.109: Rifnik near Šentjur. Plan of the fortification (Ciglenečki 2020, Fig. 12.7). Fig. 3.110: Rifnik near Šentjur. Arial view from the south-west (2023). 228 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.111: Gradec near Prapretno. Plan of the hil top site (Ciglenečki 2011b, Fig. 5.9). Fig. 3.112: Gradec near Prapretno. Panoramic view from the Fig. 3.114: Gradec near Prapretno. Rocky plateau from the north (1975). south (1977). at the beginning of modern excavations, of the large as- sociated cemetery with diagnostic goods, which proves continuous habitation from the late 5th to at least the late 6th century. Numerous grave goods correspond well with the material culture known from the settlement. The Late Antique settlement at Gradec near Prapretno ( Figs. 3.111−3.114) lies in the hinterland of Roman Celeia, in an area removed from major Roman roads (Ciglenečki 1976; id. 1981; Bausovac 2003). Hid- den in the mountains at 732 m asl, it is a characteristic example of a permanently inhabited Late Antique hil top settlement. The rocky plateau with building remains is Fig. 3.113: Gradec near Prapretno. Remains of the Late Antique sufficiently natural y protected and needs no man-made buildings are seen in the relief (1977). fortification. The field surveys and trial trenching on the 229 Slavko CIGLENEČKI Stari grad 464 sonda 5 sonda 4 Na cerkvi sonda 3 sonda 6 V riži sonda 7 objekt 4 sonda 2 objekti objekt 3 5 6 7 objekt 2 sonda 1 objekt 8 objekt 1 grobišèe Loka pri Žusmu 26 0 25 m Fig. 3.115: Tinje above Loka pri Žusmu. South-eastern part of the settlement with partial y excavated houses (Ciglenečki 2000, Fig. 71). 100 × 90 m large plateau revealed the remains of at least bulk of recovered artefacts indicate a dating to the second 21 well-preserved buildings. They had one, two or more half of the 5th, the 6th and possibly the early 7th century. rooms and some were fitted with cisterns. Their size and The fortified settlement on Tinje above Loka distribution indicates social differentiation within a smal , pri Žusmu ( Figs. 3.115−3.118) stands out among but highly effectively organised settlement that made good the densely spaced Late Antique settlements in the use of every inch of protected space. The highest part of hinterland of Celeia in its location and architecture the interior holds the vestiges of a large building with an (Ciglenečki 1987a, 43–44; id. 2000). The limited rescue eastward orientation – undoubtedly an Early Christian investigations revealed a settlement mostly covering church. Also standing out is a long building with five the vast southern slope of the hill and only partly rooms located next to the main entrance, which probably extending onto the summit. In spite of this unusual accommodated a small garrison. A few artefacts from location, the settlement was natural y well-protected the second half of the 3rd, second half of the 4th and early with steep slopes on three sides and two parallel ditches 5th century allow for the possibility that people occupied guarding the eastern slope. In combination with a this natural y protected location already in this time. The slightly raised rocky ridge behind them, the ditches mortared masonry buildings visible on the surface indi- protected the settlement in the most easily accessible cate a planned layout with the church in the centre. The part and also separated it from the adjacent burial 230 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.116: Tinje above Loka pri Žusmu. Slope with the settlement from the south (1991). Fig. 3.117: Tinje above Loka pri Žusmu. Remains of the build- Fig. 3.118: Tinje above Loka pri Žusmu. Small building of a cult ings cut into the rocky slope (1981). nature from the south-eastern part of the settlement (1981). grounds. The defences appear highly improvised, which one of the wal s and in it a small stone altar, suggest- is understandable as the hidden location provided pro- ing a building of a cult nature. Its proximity to a smal tection in itself. Surveys indicated that the buildings Late Antique necropolis with predominantly children’s were distributed evenly across the artificial terraces burials may point to a cemeterial sacral building. The hewn into the bedrock across a sizeable portion of the beginnings of permanent habitation here date to the roughly 300 × 200 m large protected hil side. Houses second half of the 4th century. The small finds and were constructed of timber and some were found to stratigraphic evidence show that the settlement may have partition wal s and hearths. A large building in have been continual y inhabited into the 6th century the centre of the settlement had channels, which most and possibly even beyond, though later finds can only likely represent the remains of a modest Late Antique reliably be dated to the 8th and 9th centuries. Evidence heating system. The lower edge of the settlement held also shows that two buildings burnt down and were a small masonry building, which had a shallow apse in abandoned roughly towards the end of the 6th century. 231 Slavko CIGLENEČKI Fig. 3.119: Ajdovski gradec above Vranje. Plan of the hil top site (red: from Knific 1994, Fig. 4; yellow: LiDAR-derived DEM proven by field survey). Source: ArcGIS, Esri, Slovenija Lidar TlaZgradbe D96, © Ministrsvo za kulturo RS). Fig. 3.120: Ajdovski gradec above Vranje. Fortified settlement from the south (2016). 232 3. SETTLEMENT IN THE COUNTRYSIDE The fortified settlement on Ajdovski gradec above Vranje ( Figs. 3.119−3.121, 3.304) was long seen as a typical example of a Late Antique refuge with the focus on its role as an ecclesiastical centre (Riedl, Cuntz 1909; Petru, Ulbert 1975; Knific 1994; Glaser 1997, 73–78). It lies in a remote valley below Mount Bohor, extending across the top of a high domed hill and on its eastern and western slopes. It is natural y well-protected from all sides and measures 162 × 66 m. The first major ex- cavations were already conducted here in 1901–1905, while the 1970s brought revision excavations by a joint Slovenian-German team, which the Slovenian team continued for several years. The summit held an apsed Early Christian church and below it another church with a baptistery. The church complex further comprised three buildings very close to one another that appeared to have functioned as a single unit combining residen- tial, prestigious and economic functions. Particularly interesting is a large house, destroyed in a fire, for which the study of the rich small finds allowed the function of individual rooms to be identified (Knific 1979). Several other houses were excavated on both slopes below the summit, as well as a cistern, with excellently preserved parts of the wooden construction, and two towers. The defensive wal s can today not be traced in their entirety, Fig. 3.121: Ajdovski gradec above Vranje. Double church with though LiDAR-derived DEM does indicate sections of a baptistery; dwelling houses in the background (2023). them. The model revealed an additional large area in the south and a narrow strip with housing in the western extension. Previously interpreted as an ecclesiastical eastern Alpine area, where coins and other finds from centre, the settlement thus shows a much greater pres- the 3rd century were predominantly interpreted as ence of housing. Where investigated, the defensive wal s residual finds. are 0.6–0.8 m thick. The hill was first more intensely inhabited in the second half of the 4th century, which is the dating for the building remains under later construc- tions, as well as numerous coins and other small finds. The majority of finds, and the masonry architecture in particular, however, dates to the second half of the 5th and the 6th century. Layers of burnt debris and well- preserved artefacts from sealed contexts show that the settlement was destroyed and burnt down in an attack (Knific 1979, 752). A small cemetery from the second half of the 5th and early 6th century is located on the north side just below the settlement. The site at Veliki vrh above Osredek pri Podsredi ( Figs. 3.122, 3.123) is important for our understanding and dating of the first prominent occupation of hil top sites in Antiquity (Ciglenečki 1990a; id. 2008, 486). The trial trenching of the prehistoric hillfort here revealed clear traces of habitation of a short duration, which coins and other metal finds reliably place to the late 260s or early 270s. The terraces showed traces of im- provised housing and the old prehistoric rampart being reused for defence purposes. Particularly important is the fact that the site was not occupied after this time, which enabled us to date the initial habitation phase Fig. 3.122: Veliki vrh above Osredek pri Podsredi. Plan of the also in other, multi-period hil top settlements in the hil top site (Ciglenečki 1990a, Fig. 5). 233 Slavko CIGLENEČKI Korinjski hrib above Veliki Korinj ( Figs. 3.124−3.127) exhibits the main architectural elements of a Late Antique fort (Ciglenečki 1985; Ciglenečki et al. 2020). Rectangular towers came to light at the edge of the natural y excellently protected rocky plateau (size 180 × 100 m). The centre of the protected area revealed the remains of an Early Christian church with an ex- tension hosting a baptistery. The trenches exploring the vast rocky area between the church and the tow- ers revealed no traces of masonry architecture. There were concentrations of habitation traces in the towers, in their vicinity and on a several-metres-wide terrace along the edge of the fort. Three towers yielded finds Fig. 3.123: Veliki vrh above Osredek pri Podsredi. Hil top site that point to a habitation function, but also a variety of from the north-west (1986). crafts practised here (processing of bronze, bone and wood), two towers even had a cistern. No defensive wal s have been established and it is presumed that the artificial terrace above the rocky area and in part also the old prehistoric rampart sufficed to defend the fort. The analysis of coins and other small finds indicates a possible short-lived occupation in the 270s. A more prominent military presence has been proven for the Valentinian period, a time when a road in the vicinity became significant and the existence of which can be Fig. 3.124: Korinjski hrib above Veliki Korinj. Plan of the fortification (Ciglenečki et al. 2020, Fig. 13.1). 234 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.125: Korinjski hrib above Veliki Korinj from the south (2022). indirectly surmised from the barrier wal at Rob, not far from the fort, which formed part of the Claustra Alpium Iuliarum barrier system (Ciglenečki 1985, 267–270). Attributable to this phase is a number of coins, pottery finds and pieces of costume, while the housing must have been of non-durable materials. It would seem that the hill was then inhabited only occassional y, either in times of increased danger or when the posited road needed protection. It may also have served as a signalling post and occasional y a shelter. The longest stretch of habi- tations in the fort dates from the late 5th to the late 6th Fig. 3.126: Korinjski hrib above Veliki Korinj. Precipitous slope century, which is also the dating of the above-mentioned from the north (2022). masonry architecture. The small finds from the towers and their parallels from across the Mediterranean date the construction of the towers and the renovation of the church on Korinjski hrib to the time of Justinian’s reconquista, i.e. the second third of the 6th century. In plan, the fort shows great similarities with many Early Byzantine forts in the Eastern Empire and in Dalmatia, underscoring its strategic significance. A considerable number of Late Antique fortifica- tion on the slopes of the Gorjanci Hil s above Šentjernej shows this area gained in importance in Late Antiquity. The most important among these is at Zidani gaber above Mihovo ( Figs. 3.128, 3.129) (Ciglenečki 1990b; Bitenc, Knific 2008; Križ 2021, 23–29). It lies on a narrow ridge, the highest part of which holds the remains of an Fig. 3.127: Korinjski hrib above Veliki Korinj. Remains of the Early Christian church and smaller, residential build-tower with a cistern (1983). ings (possibly also towers?). Access was easier in two areas, both fortified with wal s. An associated cemeterial church with several Late Antique burials came to light 235 Slavko CIGLENEČKI Fig. 3.128: Zidani gaber above Mihovo. Plan of the hil top site (Ciglenečki 2020, Fig. 12.4). slightly further to the north (Breščak 1990). The excep- tional y numerous small finds from the area show this was a major settlement centre inhabited almost through- out the Late Antique period (Križ 2021). Zidani gaber and a group of adjacent sites were inhabited at least in the first settlement wave in the second half of the 4th and early 5th century, but particularly intensely in the late 5th and throughout the 6th century. The recovered artefacts have not yet been studied in detail, but the preliminary examination points to a largely Roman population with noticeable Germanic and nomadic elements, but also the presence of a Byzantine garrison. Numerous precious items of costume and a mass of weapons from both habitation phases reveal Zidani gaber as one of the main Late Antique centres in the eastern Alpine area. The fortified hil top site at Kučar above Podzemelj ( Figs. 3.130−3.132), in the Bela krajina region, has a unique layout and is a rare example of a fortified ec- clesiastical centre (Dular 1978; Ciglenečki 1986; Dular et al. 1995). The hill of Kučar with a massive prehistoric rampart rises above the River Kolpa. Its lower northern summit held a cluster of Late Antique architecture. The hil top is protected with steep slopes on three sides, as well as defensive wal s with towers, guarding a 130 × 90 m large surface. The several years of excavations have revealed two large churches in the central part, beside the upper one also a small building – presumably 0 50 m Fig. 3.129: Zidani gaber above Mihovo. Hil top site from the south-west (2016). 236 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.130: Kučar above Pod- zemelj. Plan of the hil top site (Ciglenečki 1994, Fig. 10). Fig. 3.131: Kučar above Pod- zemelj. Ruins of the lower church at the western edge of the site (1978). a baptistery. Leaning against the defensive wal s to the defensive wal s were 0.7 m thick and reinforced in two south of both churches was a large residential building places with towers constructed on the interior side. The (25 × 17 m) with two rooms fitted with central heating, total number of towers cannot be ascertained as part of the larger of which was the main hal . A smaller house the settlement was destroyed prior to the excavations. stood between this building and the upper church. The Also integrated into the defensive wal s was part of the 237 Slavko CIGLENEČKI Fig. 3.132: Kučar above Podzemelj. Remains of a large residential building from the southern side of the hil top site (1979). narthex of the lower church and the south part of the above-mentioned large residential building, which pro- vided additional protection to the most easily accessible area of the site. The small finds and the architecture sug- gest the settlement dates roughly between 380 and 500. Standing out in size and strategic importance is the fortress on Gradišče near Velike Malence ( Fig. 3.133) (Saria 1929; id. 1930; Ciglenečki 2003b, 586–587). It lies in the vicinity of the confluence of the Rivers Krka and Sava, on a narrow low plateau between the Krka and the Gorjanci Hil s, just above the Roman road connecting Siscia and Emona. It is triangular in plan and covers a surface of 430 × 270 m. Its western and northern sides are natural y well-protected with very steep slopes, while the south side rises only slightly above the surrounding area. Defensive wal s with towers were constructed on top of the rampart of the prehistoric hillfort, with towers more densely positioned along the more exposed southern side. Balduin Saria distinguished between two phases of Fig. 3.133 Gradišče near Velike Malence. Plan of the fortress the fortress. The 2.10 m thick defensive wal s were built (Ciglenečki 1987a, Fig. 146). in the early phase, later reused and strengthened, but also pierced to integrate rectangular towers at a distance of 35–40 m from each other. Saria conducted trial trenching in the interior, detecting no masonry buildings here in phases, setting the first one to the 3rd and the second one to the first phase and presuming the existence of wooden the late 4th or early 5th century (Saria 1939, 145). The plan huts (Saria 1939, 144). In the second phase, he posited the and position of the towers alone does not seem sufficient construction of an Early Christian church. However, the grounds for such precise dating of the second phase (cf. unearthed apsidal building with an unusual, westward Wilkes 2005, 261–264). With its interior poorly known, orientation and very thick wal s is nothing like the Early we can use its size ( c. 8 ha) and fortification features Christian churches known from Late Antique forts, which to draw parallels with the Pannonian ‘inner fortresses’, makes its function and dating questionable. Saria used though the absence of small finds suggests it was only historical more than archaeological data to date the two occasional y used as a military fortress. 238 3. SETTLEMENT IN THE COUNTRYSIDE Croatia via a saddle, where in Late Antiquity a 0.7 m thick wall was built on top of a high prehistoric rampart; there are The fortified s hil top ettlement at Marija Gorska also visible traces of a ditch. Coins and other small finds near Lobor ( Figs. 3.134−3.136) revealed traces from dif-indicate habitation already in the 3rd and 4th centuries, ferent periods (Filipec 2007; Filipec et al. 2020, 211–218). though a large fortified settlement ( c. 230 × 140 m) was The well-protected domed hill is only readily accessible only established in the second half of the 5th and contin- Fig. 3.134: Marija Gorska near Lobor. Plan of the fortified hil top settlement. Remains of the Early Christian church with a baptistery are marked in black (Filipec et al. 2020, Fig. 11). Fig. 3.135: Marija Gorska near Lobor. Fortified settlement from the south-west (2017). 239 Slavko CIGLENEČKI Fig. 3.136: Marija Gorska near Lobor. Prehistoric rampart with a wall on top of it from the north (2017). ued throughout the 6th century. Under the present-day pilgrimage church, the interior of the settlement holds the remains of a large Early Christian basilica with an octagonal baptistery that was destroyed in a fire in the early 7th century. The excavated artefacts reveal intense habitation across the whole settlement, while the size and furnishings of the basilica reveal the site was one of the Early Christian centres in the southeastern part of Noricum Mediterraneum. The fortification on Kuzelin near Donja Glavnica ( Fig. 3.137) lies on a natural y very well-protected hill in proximity to the Roman road connecting Andauto- nia and Poetovio (Sokol 1981; id. 1994; id. 1998). The defended area on the summit measures 200 × 40 m, with buildings concentrated on an 8 m wide and very long terrace a few metres below the summit. The rich assemblage of coins and other small finds indicates two phases. The first one is dated to the last third of the 3rd Fig. 3.137: Kuzelin near Donja Glavnica. Plan of the fortification (Sokol 1994, Fig. 2). century based on numerous coins of emperors from Gal- lienus to Aurelianus. Attributable to this phase are the residential buildings with floors of square bricks in the upper part of the fort. It would appear that a palisade was the only defence at this time. The second phase shows much more intense habitation, which coins and other artefacts date to the last quarter of the 4th and first half of the 5th century. In this phase, a 200 m long and 1 m thick defensive wall was constructed in the east. Erected behind it were 4 × 4 m large buildings, of which only the mortar floors survived. A female burial was found along the interior face of the wal , which the grave goods date to the late 4th century. Numerous weapons and belt parts also date to this phase. The defensive wall and the mod- est buildings behind it reveal a military post, while the multitude of small finds also mirror a civilian presence, suggesting the fort was also used as a refuge. Fig. 3.138: Crkvišće Bukovlje near Generalski Stol. Plan of the hil top site (from Azinović Bebek, Sekulić 2014, Fig. 4). 240 3. SETTLEMENT IN THE COUNTRYSIDE A large Late Antique fort has recently been discov- ered in the prehistoric hillfort at Crkvišće Bukovlje near Generalski Stol ( Figs. 3.138, 3.139) (Azinović Bebek, Sekulić 2014, 167–168; iid. 2019). It lies in the bend of the River Mrežnica and is natural y protected with steep slopes. In the highest part of the hil , excavations so far integral y unearthed an apsed Early Christian church, beside which there may be two rooms of a large and well-built edifice. Found beside the west defensive wal s were two rectangular single-room buildings, one of them dated to the 4th and the other to the 6th century. The side of easiest access shows traces of two protruding towers. It is as yet not possible to determine when the fort was established. The earliest concentration of finds dates the beginning of the Late Antique occupation to the last third of the 3rd century, while the fort is believed to Fig. 3.139 Crkvišće Bukovlje near Generalski Stol from the west (2016). have been constructed in the second half of the 4th, was added a church in the 5th and then persisted throughout the 6th century. The architectural evidence, the important strategic location on the Roman road from Romula to Senia and the recovered small finds indicate the presence of a garrison here in the 4th and the 6th century, though the fort may also have functioned as a civilian settlement. The protruding towers and the modest church exhibit similarities with the military forts from the Justinian period both along the Dalmatian coast and inland. The fortified complex at Kastrum on Veliki Brijun (Brijuni Islands) ( Figs. 3.140, 3.141) lies in a well- protected Madonna Bay (Dobrika) (Mlakar 1975−1976; Begović Dvoržak 2001). The Late Antique settlement was established on top of an earlier vil a and functioned as a refuge for the population of Brijuni and adjacent areas, but was also a major military base. The sea, the marshland and the small fort on the nearby Petrovac hill provided effective protection. The location on the exterior side of the islands also concealed it from the Fig. 3.140: Kastrum on Veliki Brijun. Plan of the fortified site view from the mainland. The defensive wal s were built (Suić 2003, Fig. 188). Fig. 3.141: Kastrum on Veliki Brijun from the south-west (2003). 241 Slavko CIGLENEČKI in two phases. The first phase shows features character- istic for the 4th and 5th centuries; these wal s were 2.6 m thick, 2.9 m at the corners, and enclosed a surface of 1.2 ha. Additions and reinforcements date to the early 6th century, to the time of the Gothic Wars, when the fort served as the starting point for crossing the Adriatic to reach Ravenna. The contemporary church of St Mary was erected outside the fort. The defensive system also comprised watch posts at exposed locations on the outer islands of the Brijuni Archipelago. Because of its short duration, the fort at Pasjak near Klana ( Fig. 3.142), on the Roman road between Tarsatica and Tergeste, serves as a reliable chronological indicator (Starac, R. 1993; id. 2009, 286). This low hill holds the remains of a roughly rectangular and 100 × 80 m large fort. The 3 m thick wal s, bound with solid mortar, thus far revealed three entrances and presumably a tower in the south west. The entrance was protected with two wal s running perpendicularly. A thin cultural layer and Fig. 3.142: Pasjak near Klana. Plan of the fort (Starac, R. 2009, the coin hoard found under the ruins of the defensive Fig. 2). wal s show the fort was destroyed around 270 (Starac, R. 2009, 286). It was never renovated. The fortification at Veli Grad on the island of Krk ( Figs. 3.143, 3.144) lies on a crag above the sea (Faber 1988, 116–119). It was largely protected by precipitous slopes, on the only accessible side with a 0.8–1.1 m thick wall enclosing an interior surface of 100 × 80 m. In the middle of this wall is the entrance fortified with two paral- lel wal s set perpendicularly to the defensive wal . Visible on the interior side of the defensive wall are buttresses or footing for a wall-walk. Aleksandra Faber noted the great similarity of this wall with the barrier wal s of the Claustra Alpium Iuliarum system. The rocky interior of the fortification did not reveal a cultural layer. The archi- tectural remains show that the fort greatly differs from Fig. 3.143: Veli Grad on the island of Krk. Wal s and the forti-the later Justinian period forts in the Kvarner Bay and fied entrance on the western side of the fortification (1996). Fig. 3.144: Veli Grad on the island of Krk. Entrance with a part of the wall (from Faber 1988, Fig. 9). 242 3. SETTLEMENT IN THE COUNTRYSIDE Dalmatia; it is most likely earlier (3rd or 4th century), built structure of buildings without having to conduct system- to oversee the important narrow route between the island atic investigations (Ciglenečki 1987a, 104–105; Tomičić and mainland, but rarely used and rapidly abandoned. 1988, 148–151; Faber 1988, 121–127; Šiljeg 2008; Regan, The large settlement at Bosar is already discussed Nadilo 2009, 1080–1083). The fortification is located on above, in the chapter on the newly-founded Late An- a steep hill offering a good view over the activities in two tique cities (see Chapter 2.5). Located 100 m above it is bays sheltered from the wind, i.e. Vela Luka and Mala a well-preserved fortification at Korintija near Baška Luka. The fort measures 110 × 70 m and is protected with ( Figs. 3.145−3.149) on the island of Krk. Similarly as the well-preserved defensive wal s measuring 0.7–0.8 m in settlement below, the favourable climate and the absence thickness. Some of the buildings along the interior of the of vegetation offer an excellent insight into the extent and wal s can be interpreted as towers, while others housed Fig. 3.145: Korintija near Baška from the south (2021). Fig. 3.146: Korintija near Baška. Plan of the south-eastern part of the hil top site (Faber 1988, Fig. 13). 243 Slavko CIGLENEČKI Fig. 3.147: Korintija near Baška. Eastern part of the fortification from the south-west (2015). Fig. 3.148: Korintija near Baška. Cistern preserved to a great height in the south-eastern part of the fortification (2021). the troops. A large polygonal tower with thick wal s, which was constructed with gradual additions, marks the southern end where the fortification was most read- ily accessible; visible next to this tower is a large cistern. The high ridge in the interior holds traces of an apsed Early Christian church and a side room, while there is a large building with a small cistern further along the ridge. Surface finds of pottery date the fort to the 5th/6th centuries, reaching its peak under Justinian, similarly as other forts along the Dalmatian coast. Small finds and architecture indicate it functioned in association with the contemporary unfortified settlement below. Fig. 3.149: Korintija near Baška. Wal s from the north-western part of the fortification (2021). 244 3. SETTLEMENT IN THE COUNTRYSIDE The small rocky island of Veliki Sikavac ( Figs. 3.150−3.152), which a narrow strait separates from the island of Pag c. 100 m away, recently revealed the remains of an Early Byzantine fort built only 13 m above the surface of the sea (Oštarić, Kurilić 2013, 270–271; Gluščević, Grosman 2015). It stands in an important strategic position that enabled control over the narrow strait between Pag and the mainland to lead to the Velebit Channel and was, as such, crucial in the control of coastal routes. The fort is adapted to the terrain and measures 90 × 50 m. The impressive defensive wal s were 1.8 m thick and built of stones bound by high-quality mortar. There are six towers, of which the two in the highest part were attached to the wal s, while four others integrated in the wal s and were open on the interior side. Buildings cluster in the northwestern and southeastern parts of the fort. No building remains are visible in the central part. Fig. 3.150: Veliki Sikavac. Plan of the fort (Gluščević, Grosman The northwestern part has 23 buildings measuring 2015, Fig. 11). 4 × 5 m on average and arranged in four large groups with clearly discernible passages between them. In the southeastern part, buildings are arranged on six parallel terraces from the defensive wal s inwards. The location on a low and poorly defended island displays charac- teristic Byzantine military architecture with rectilinear defensive wal s, protruding towers and army barracks in the interior. Field surveys unearthed sherds of Late Antique pottery, amphorae and glass goblets. Fig. 3.151: Veliki Sikavac from the north-east (2016). Fig. 3.152: Veliki Sikavac. Impressive defensive wal s and the remains of the buildings on the eastern side of the fort (2016). 245 Slavko CIGLENEČKI The Early Byzantine fort at Sv. Trojica, Tribanj, wal s abutted in the interior on a dry-stone construction, Šibuljina ( Figs. 3.153, 3.154) was built on the site of a presumably prehistoric earthwork rampart. Modest traces Roman settlement, in an important strategic location of drystone defensive wal s are visible on the less exposed control ing the land and maritime routes along the Velebit east side. The interior was divided into four or five terraces Channel (Glavičić 1984, 19–21; Tomičić 1990, 142–143; with remains of heavily damaged wal s. The end of the Dubolnić 2007, 39–40; Ciglenečki 2020, 255–257). The fort in the south has not been established, but it would gentle slope of the hill with the church of St Trinity (Sv. seem that it covered a smal er area than the earlier Roman Trojica) holds the remains of a fortification. Most clearly settlement. The small finds show that the defensive wal s visible on the surface are the excellently preserved remains constructed in Late Antiquity, most likely under Justinian, of the defensive wal s with four protruding towers on the additional y protected the area of the earlier settlement. most readily accessible north side. These wal s survive in The size of the defended area suggests the fort served not the length of some 120 m, height of 6–8 m and thickness only to accommodate army troops, but also as a refuge of 0.8 m. They show putlog holes in some places. The for civilian population. Fig. 3.153: Sv. Trojica, Tribanj, Šibuljina from the south-west (2004). Fig. 3.154: Sv. Trojica, Tribanj, Šibuljina. Excellently preserved wal s on the northern side of the fort (2014). 246 3. SETTLEMENT IN THE COUNTRYSIDE Field surveys identified another Early Byzantine and mortar-bound. Publications mention five c. 7 × 6 fort at Gradina near Modrić ( Figs. 3.155−3.157), at or 6 × 6 m large towers. The interior holds the ruins of the beginning of the Velebit Channel (Tomičić 1990, several buildings, but they cannot be identified or dated 141–142; Dubolnić 2007, 38–39). It lies on a gentle more precisely without investigations. The fort enjoyed slope inside a prehistoric hillfort, just above the coast. an excellent strategic location, controlling the passage The mighty prehistoric rampart was reused to defend through the strait of Masleničko ždrilo and the navigable the whole of the north side, where access is easiest, and route along the north coast of the island of Pag. mortared wal s were added on top. Elsewhere, the earlier The small fort at Toreta (Tureta) ( Figs. 3.158, rampart was reused as a proteichisma. The fort inte- 3.159) on the island of Kornat has been identified as a rior is considerably smaller than that of the prehistoric Late Antique fort very early on (Petricioli 1970; Fabijanić hillfort and its outline takes the shape of an irregular et al. 2012). It is a rectangular tower-like construction on trapezium with straight lines. It measures 162 m in a small hill with steep slopes rising directly from the sea, length and 122 m in width. The wal s are 1.6–1.9 m thick with a 10.4 × 5.4 m large interior. It is reinforced with Fig. 3.155: Gradina near Modrič. Plan of the fort (Ciglenečki 2020, Fig. 12.22). Fig. 3.156: Gradina near Modrič from the east (2011). 247 Slavko CIGLENEČKI Fig. 3.157: Gradina near Modrič. Late Antique wal s on the southern side of the fort (1987). Fig. 3.158: Toreta (Tureta) from the south-west (2019). eight large buttresses organical y tied into the perim- etral wal s. The wal s and buttresses are 1 m thick. The entrance is careful y made and includes a relieving arch. The ground floor has arrowslits. Higher up there are holes for floor joists and above them eight windows. The small-scale excavations in the tower interior revealed a structure, presumably cistern, which held small finds that included Early Byzantine pottery. The ground floor is believed to have served for storage, while the upper storeys were used for defensive purposes. The fort hosted a small garrison that watched over the maritime routes leading through the Kornati Archipelago. Below it in the Tarac Bay, there are traces of a large and repeatedly modified Early Christian church with a baptistery that was presumably already built in the second half of the Fig. 3.159: Toreta (Tureta). Plan of the fort (from Petricioli 5th century (Zornija 2016). The fertile field in the vicinity 1970, Fig. on p. 720). 248 3. SETTLEMENT IN THE COUNTRYSIDE provided sustenance for a small community that could erected on top of a prehistoric rampart. One of the tow- seek shelter inside the fort in times of danger. ers holds a cistern. The slope in front of the defensive wal s shows traces of a roughly 1 m thick proteichisma. Similar small-sized forts (predominantly towers) In the interior there are traces of several buildings. Along also came to light on other islands ( Fig. 3.315; such as the wal s is a complex of at least three rooms with a Svetac, Majsan, Palacol) and on the mainland (Ošlje – cistern, which is very similar to the residential building Gradac), mirroring the diverse strategies employed to investigated at Gradina on the island of Žirje (see below). protect the maritime route along the eastern coast of Smaller buildings also lean against the south rocky edge the Adriatic (Kirigin, Milošević 1981; Badurina 1982; of the fort. Elements of a contemporary Early Christian Gunjača 1986; Tomičić 1996; Baraka Perica, Grbić church came to light in the church of St Andrew (Sv. An- 2019). These forts share a characteristic, natural y well- drija) below the fort (Domijan 1983, 136). protected location associated with a favourable harbour, Gradina on the island of Žirje ( Figs. 3.163−3.166) which clearly shows their function of controlling and holds the most characteristic and also the best re- protecting navigable routes. searched military post in the area under duscussion A small Early Byzantine fort lies at Gradina on the that offers an insight into the construction of the Early island of Vrgada ( Figs. 3.160−3.162) (Suić 1976a, 238; Byzantine forts on the eastern Adriatic coast (Iveković Domijan 1983, 123; Gunjača 1986, 126–127; Ciglenečki 1927, 49–52; Gunjača 1986; Pedišić 2001; Ciglenečki 2020, 185–187, 246–249). Its remains are only known 2011b, 676–678; Karađole, Borzić 2020). The fort was from field surveys. It is sited on the highest part of a built on a low hill of the southeastern cape of the is- rocky ridge that offered visual control over the aquatory land that provided a good control over the access to between the Kornati Archipelago and Zadar. It was most the nearby port and over the traffic along the exterior easily accessible from the north, partly also the east side, side of the island. It measures roughly 100 × 50 m. It is whereas the south and southwest sides were protected pentagonal in plan and whol y adapted to the terrain. with steep terrain. The fort measures 85 × 35 m. There The 0.8–0.95 m thick defensive wal s have protruding are clearly visible remains of five rectangular towers rectangular and one pentagonal tower. The north wal Fig. 3.160: Gradina on the island of Vrgada. Fortification from the south-east (2019). 0 20m Fig. 3.161: Gradina on the island of Vrgada. Plan of the fort (Ciglenečki 2020, Fig. 12.18). 249 Slavko CIGLENEČKI Fig. 3.162: Gradina on the island of Vrgada. Defensive tower with a cistern on the western side of the fort (2019). Fig. 3.163: Gradina on the island of Žirje. Plan of the fort (Ciglenečki 2020, Fig. 12.13). was constructed on the top of the ridge, while the in- of stacked stone blocks. Leaning onto the wal s from terior extended across the gentle southern slope. The the interior was a building with several rooms, two of main entrance led through a tower, while investigations which had heating; the luxurious appearance suggest revealed three narrower entrances ( poternes) next to this was the residence of a commander. Small finds the defensive towers. The fort was additional y pro- reliably point to the Early Byzantine period and date tected with a proteichisma in the shape of a low wal the establishment of the fort to the time of Justinian’s 250 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.164: Gradina on the island of Žirje. Remains of the large building and the wal s from the north (2010). Fig. 3.165: Gradina on the island of Žirje. Tower on the north-eastern part of the fort (2004). Fig. 3.166: Gradina on the island of Žirje. Part of the northern wal s with a pro- teichisma (2010). 251 Slavko CIGLENEČKI Fig. 3.167: Čuker near Mokro Polje. Plan of the fortification (Ciglenečki 2003a, Fig. 12). reconquista. It was destroyed in a fire even before the end of the 6th century. Čuker near Mokro polje ( Figs. 3.167, 3.168) is the site of one of the rare well-investigated prehistoric hillforts in the Adriatic hinterland inside which a fort was built in Late Antiquity (Delonga 1984, 277; Ciglenečki 2003a, 268). Having said that, our knowledge of the fort is very limited and only a plan with some marginal notes sur- vives of the investigations that Ejnar Dyggve conducted in 1930. Examination of the site today is hindered by the lush vegetation, making the remains that Dyggve marked on the plan largely unverifiable. We can better understand the site by comparing the surviving plan with those of the forts built into prehistoric hillforts along the fringes of the karst poljes in Bosnia and Herzegovina. It was constructed on an exposed location above the rocky slopes descending to the River Zrmanja, which blocks acces from the north and west. On all other sides, it is protected with a high prehistoric rampart reinforced with defensive wal s and a proteichisma. A comparison with other fortifications suggests that the buildings lining the defensive wal s were dug into the prehistoric rampart. A small apsidal church stood on a narrow terrace above the Zrmanja. Several small buildings were also erected anew in the southwest. Dyggve’s plan, kept in the Römisch-Germanische Kom- mission in Frankfurt, also shows stairs, which would explain the height differences between the buildings sunken into the rampart and the terrace with the church (cf. Schnurbein 2001, 194, Fig. 18a, top right). Dyggve’s Fig. 3.168: Čuker near Mokro Polje from the north-west. plan suggests the fort measured 170 × 70 m, while satel- lite imagery reveals a much smaller fort, comparable with 1980, 230–233). The monastery reused the buildings of those in the Dalmatian interior ( c. 90 × 30 m). a smal Roman settlement and combined them into a The complex of buildings on Majsan ( Figs. 3.169, single organic unit. The residential area arranged around 3.170), a small island between the Pelješac Peninsula an inner court was separate from the industrial part. The and the island of Korčula, is one of the rare identified complex was fortified with towers. A tomb, which report- Early Christian monasteries in the Adriatic (Fisković edly once held the relics of St Maximus, was built into 252 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.169: Majsan. Plan of the fortified monastery (Cambi 2002, Fig. 393). tower monastery Fig. 3.170: Island of Majsan. The location of the monastery and the tower, view from the west (2005). one of the central rooms and an altar was set up above Bosnia and Herzegovina the tomb, which gave the room a memorial character. Another tomb was found in the complex, which held The early overviews and the multitude of data col- three skeletons, presumably belonging to the founding lected in Arheološki leksikon Bosne i Hercegovine (Čović members of the religious community. Later, four pil ars 1988) reveal a large number of fortifications, though they were erected above this tomb to hold a cupola intended are often too brief to allow more detailed identifications to emphasise its importance. The monastery located on (Basler 1972; Čremošnik 1990; Špehar 2008). an important maritime route would certainly have been a pilgrimage destination. A tower was built behind the Gradina Zecovi near Čarakovo ( Figs. 3.171, 3.172) monastery, on the highest point of the island, which was displays the characteristics of a medium-sized Late An- a watch tower and possibly also provided shelter in times tique fortification with a military function (Čremošnik of danger. This feature places Majsan in the system of forts 1955b; ead. 1956; Ciglenečki 1987a, 93; Chevalier 1995, lining the maritime routes ( Fig. 3.315) of the 6th century. 158; Špehar 2008, 571–572). It has a natural y well-pro- 253 Slavko CIGLENEČKI tected location on the Roman road connecting Salona and Siscia. It is 125 × 50 m large and enclosed with 2 m thick wal s. The buildings at the far ends of the fort interior can be seen as towers or accommodations for the troops. A modest Early Christian church with an apse and narthex was investigated in the centre. The small finds and the concept of a fort with a church suggest a date to the 6th century, with possible beginnings in the second half of the 5th century. In northwestern Bosnia, most likely within the former province of Savia, there was a large Late Antique agglomeration at Gradina in Bakinci near Banja Luka ( Figs. 3.173−3.176) (Vujinović 2013; id. 2014). It lies in the mountains, slightly removed from the Roman road that led from the valley of the Sava towards Salona. Research in the recent decade revealed the remains of this large settlement on a natural y well-protected hil . Its core is the c. 220 × 140 m large fortified area on top of the hil . The sections of easier access were protected with wal s that measured between 1.2 and 1.6 m in thickness, depending on the terrain. The top of the plateau revealed a well-built single-room building. In the north along the wal s, researchers found an unexplored building with high surviving wal s. Other edifices are visible on the surface. The distribution of the small finds suggests that the settlement extended down the slopes. The saddle located just a few metres below and east of the fortified part holds the remains of three Early Christian churches, which the plans, construction manner, architectural elements and small finds place to the 5th–7th centuries. Standing out is Basilica A, which is over 40 m long and richly decorated, revealing the great significance of this site as an Early Christian centre. Investigations in the settlement interior have only just begun and have not yet provided more concrete evidence, though the com- plex of three large churches and the rich artefacts offer indirect evidence of the significance of this settlement complex in the 5th and 6th centuries. The partly exca- Fig. 3.171: Gradina Zecovi near Čarakovo. Plan of the fortifica-vated churches revealed evidence of their renovation, tion (Basler 1972, Fig. 26). Fig. 3.172: Gradina Zecovi near Čarakovo. fortification from Fig. 3.173: Gradina in Bakinci. Plan of the hil top site and the the east (1988). churches (Vujinović 2014, Fig. 24). 254 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.174: Gradina in Bakinci. Hil top site from the north Fig. 3.176: Gradina in Bakinci. Ruins of a smaller church (2014). (Basilica C) outside the wal s (2014). with earlier elements reused in the cathedra, the clergy bench added in Basilica A and the baptistery in Basilica B later partitioned. Particularly intriguing is the find of an impost inscribed with the names Constantius, Andreias and Constans, the first two of which correspond with the names of the bishops present at the synods of Salona in 530 and 533. The state of investigations does not allow a more precise dating of individual phases of the site, while coins and other small finds indicate an even earlier habitation in the 3rd and more intensely in the second half of the 4th and the early 5th century. The partly investigated fortification at Grad near Gornji Vrbljani ( Fis. 3.177) lies in a remote mountainous area close to the spring of the River Sana (Bojanovski Fig. 3175: Gradina in Bakinci. Excavated building inside of the 1976; id. 1980; Ciglenečki 1987a, 106–107; Špehar 2008, fortified settlement (2014). Fig. 3.177: Grad near Gornji Vrbljani. Plan of the fortification (Bojanovski 1980, Fig. 1). 255 Slavko CIGLENEČKI Gradina in Grkovci Gradina in Bastasi Gradina in Gubin Kesićeva gradina near Čelebić Gradina in Donji Rujani Gradina near Velika Gradina Žirović near Veliki Kablić Velika gradina Gradina in Vašarovine in Donji Rujani Gradina in Lištani Gradac above Potočani Gradina in Vidoši Fig. 3.178: Late Antique fortifications (red dots) of Livanjsko polje (from Benac 1985, map 3). 569–571). It measures roughly 110 × 65 m and extends the limited investigations do not provide evidence for a across most of the prehistoric hil fort above the canyon of more detailed discussion. I only present those where the the Lučica stream, which envelops the site from two sides. remains are sufficiently visible on the surface to enable It is additional y protected with defensive wal s, which are at least a brief description of the defensive elements and merely 0.65 m thick and reinforced from both sides with other architectural features, but have also witnessed some buttresses. A ditch was dug in front of the wal s in Late trial trenching. The density of Late Antique settlement in Antiquity. In the part of easier access, a 22 × 10 m large the area can best be il ustrated on the example of Livanjsko residential building leans against the defensive wal s in polje, where field surveys in the 1980s primarily aimed at the interior. It has a small anteroom presumably with a investigating prehistoric hillforts also detected and briefly staircase, a spacious main room, small room with a large described the Roman remains (Benac 1985). In addition fireplace, court, kitchen and small cistern. The ruins of to defensive wal s, some even revealed the remains of a two other buildings are visible along the western defensive proteichisma, which was lower and built in the drystone wal s. Ivo Bojanovski, who excavated the site, presumed technique (Ciglenečki 2003a, 271). that the fort was constructed in the first half of the 5th and Gradac above Potočani ( Fig. 3.179) holds clearly persisted to the late 6th century, though he also mentioned visible Late Antique remains (Benac 1985, 104–106). the possibility that it was constructed in the time of the This fort on a high karst ridge 100 m above the valley is Gothic Wars. The above-mentioned military housing protected by rocky faces in the south, while all others points to the presence of a small garrison, while the fort sides were protected already in prehistory by a thick also provided shelter for the population of a wider area. earthwork rampart. It is a small fortification, measuring A detailed research of the adjacent lines of Late Antique c. 70 × 30 m. Thick mortared wal s were built on top of communication would certainly shed light on the role the rampart, ten metres in front of it also a drystone of this fort. proteichisma. Traces of other wal s along the defensive Archaeological surveys and rare trial trenching walls suggest buildings with floors sunken into the campaigns in southern Bosnia indicate the existence of prehistoric rampart. Leading up to the fortification is a numerous Late Antique fortifications ( Fig. 3.178), but pathway still clearly visible today. 256 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.179: Gradac above Potočani. Remains of the wal s and a proteichisma from the north-east (1988). Fig. 3.181: Gradina in Vašarovine. Plan of the hilltop site (Ciglenečki 2003a, Fig. 13). Fig. 3.180: Gradina in Vidoši. Plan of the hil top site (Basler 1972, Fig. 33). An example of a Late Antique refuge frequently mentioned in literature is that on Gradina in Vidoši ( Fig. 3.180), where a mortared wall was constructed in Late Antiquity in the highest part of the prehistoric hillfort (Basler 1972, 60; Benac 1985, 97–99; Basler 1993, 38). Fig. 3.182: Gradina in Vašarovine from the south-east (1988). 257 Slavko CIGLENEČKI Fig. 3.183: Gradina in Bastasi from the north-west (1988). Gradina in Vašarovine ( Figs. 3.181, 3.182) holds 1985, 39–41). The prehistoric hillfort holds the remains readily observable defensive elements (Benac 1985, of Late Antique wal s that ful y integrated the prehistoric 114–116; Ciglenečki 2003a, 271). This fort lies high defensive constructions and the natural terrain. The fort above the valley and provided effective control of the measures c. 140 × 35 m, with additional artificial terraces surroundings. A very thick and up to 10 m high prehis- visible on the steep slope below. Its 0.8 m thick defensive toric rampart encloses the fort from three sides, while wal s were built on top of the prehistoric earthwork ram- a very steep slope protects it in the east. Substantial, part, in some places heavily reinforced. It is very likely 1.5–1.8 m thick wal s were built in Late Antiquity on that buildings (towers?) were sunken into the rampart. top of the rampart. In the most exposed part, the wal s A drystone proteichisma is visible in some places not widened into a bastion (tower?). At a distance of 20 m, far below the defensive wal s. the wal s were lined along the whole length by a drystone Of the many fortifications on the fringes of karst proteichisma. The 110 × 50 m large fortification has poljes, some may only have been used as refuges, par- differently-sized habitation terraces below. ticularly those located in less exposed areas. They may The fortification at Gradina in Bastasi ( Figs. 3.178, have been used for brief periods, by people otherwise 3.183) has a completely different location (Benac 1985, living in the lowland. Examples are known in Kupreško 121–123). It lies in the lowland, on the fringes of a karst polje, where Dimitrije Sergejevski unearthed two Late polje and close to a water spring. It rises barely 10 m Antique forts with the remains of contemporary settle- above the surrounding area and is protected with a ments at the foot of the respective elevations, most nota- deeply carved bed of a stream. The shorter sides of the bly the remains of churches (Sergejevski 1942, 126–131, plateau hold two stretches of a thick prehistoric rampart, 147–153; Basler 1972, 56–57). Of the two fortifications, on top of which are the remains of Late Antique wal s more is known on the one at Gradina in Podgradina with a tower and buttresses. A 0.8 m thick wall is also Kamenska ( Figs. 3.184, 3.185), where researchers ex- discernible along the longer sides of the settlement. amined the most characteristic elements before the site Similar observations have been made for the for- became heavily overgrown (Sergejevski 1942, 147–150; tifications on other karst poljes. The best preserved is Basler 1993, 36). The plateau above the precipitous face Gradina above Kovaći, in the west part of Duvanjsko of a torrential stream was already inhabited in prehis- polje that rises 100 m above the surroundings (Benac tory, when it was protected with substantial earthwork 258 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.186: Gradac in Lepenica (Homolj). Plan of the fortification (Basler 1972 Fig. 80). additional protection. A proteichisma was added on the two sides of easier access. The fort was additional y protected with a deep and wide ditch along the south- Fig. 3.184: Gradina in Podgradina Kamenska. Plan of the west. In the 90 × 65 m large fortified area, Sergejevski fortification (Sergejevski 1942, Fig. 18). observed the remains of a large building and a cistern and also presumed two towers on top of the ramparts. A smal er refuge came to light on the natural y well-protected hill of Gradac in Lepenica (Homolj) ( Fig. 3.186) (Skarić 1932; Basler 1972, 89–91; id. 1993, 57–59; Špehar 2008, 572–574). Its 70 × 50 m large flat summit was enclosed with defensive walls. Leaning on the wal s from the interior was a rectangular tower. A richly decorated small Early Christian church was found in the northeast corner, presumably already built in the 5th century, but renovated and enlarged in the 6th century. Just below the hill are abundant remains of a Roman settlement not known in detail, but one Fig. 3.185: Gradina in Podgradina Kamenska from the north. that likely continued into Late Antiquity and was thus contemporary with the fort on the hil . One of the last more extensively investigated Late ramparts that are particularly high along the shorter Antique settlements in Bosnia is located at Gradac on sides. In Late Antiquity, 1.5 m thick mortared wal s Ilinjača near Sarajevo ( Fig. 3.187) (Fekeža 1990; ead. were built on top of the rampart, associated in the south 1991). On this natural y well-protected hil , 0.8–1 m corner with a wall that traversed the slope and offered thick defensive wal s were constructed in Late Antiquity Fig. 3.187: Gradac on Ilinjača near Sarajevo. Plan of the hil top site (Fekeža 1991, Fig. 1). 259 Slavko CIGLENEČKI Fig. 3.188: Debelo Brdo above Sarajevo. Plan of the hil top site (Fekeža 1991, Fig. 9). on top of the rampart of a prehistoric hillfort, with the wal s enclosing a 115 × 40 m large area. Integrated into the wal s from the interior was a small Early Christian church. The church was presumably already built in the late 5th or early 6th century, whereas the defensive wal s were erected in the second quarter of the 6th century and incorporated the church into the system of defence. No other buildings are known in the interior, only a cultural layer is mentioned, with modest finds from the 3rd and 4th centuries, and richer finds from the 5th and 6th centuries. There is insufficient evidence as to the type of settlement (refuge?) that existed here from the 3rd to the 5th century when the defensive wal s had not yet been built. Did the prehistoric rampart provide sufficient protection? Fig. 3.189: Gradina in Biograci near Lištica. Plan of the forti-The fortification at Debelo Brdo above Sarajevo fication (Čremošnik 1989, Fig. 85). ( Fig. 3.188) is mainly known from early excavations (Fiala 1894) and small-scale investigations in recent times (Fekeža 1991, 176–178). The fort was constructed 90 × 40 m stands on a low elevation with an excellent on a hill well-protected already in prehistory. The 1.3 m view of the karst polje of Mostarsko blato. Its defensive thick defensive wal s are completely adapted to the ter- wal s were only 0.6–0.7 m thick, but associated with rain and enclose a 120 × 70 m large area. Part of the wal s buildings densely spaced along the interior and increas- survived, as did the south tower, while a report mentions ing the wal s’ defensive potential. A single rectangular another tower. The interior revealed a fairly thick Late protruding tower was found, while two towers presum- Antique layer that indicates more permanent habitation. ably stood in the interior near the west entrance. The site The remains of a Roman settlement are mentioned below was additional y protected with a proteichisma, part of the hil , near the River Miljacka, which also yielded coins which was unearthed several metres below the wal s, both and other finds from the 5th and 6th centuries. mortared. The function of the fortification can be inferred Gradina in Biograci near Lištica ( Figs. 3.189, 3.190) from its location near a road that connected the coast is among the rare well-researched sites in this area, but and the hinterland, but also from the concept of a fort has so far been published only briefly (Čremošnik 1989; with buildings lining the interior of the defensive wal s. Chevalier 1995, 413–414; Špehar 2008, 577–579). The There are no independent churches, while a small church densely built-up Late Antique fortification measuring is presumed in the protruding tower with an internal 260 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.190: Gradina in Biograci near Lištica from the north (1988). apse (Chevalier 1995, 413–414). The chronology is not completely reliable as it is based on architectural features, high-quality pottery and parallels with the then known fortifications; Irma Čremošnik dated its construction to the 3rd or 4th century and continued existence throughout the 5th and 6th centuries. Considering the similar sites, it is more likely that the defensive wal s were built in the second half of the 4th century, whereas the dating of the majority of the buildings in the interior remains open. The construction of the proteichisma can, as elsewhere, be dated to the Justinian period, when the church may also have been integrated. The scarce military equipment and a variety of tools indicate a pronounced civilian use, but also the presence of a small garrison to control the Fig. 3.191: Mogorjelo. Remains of the Early Christian church communications. in the fortification (2009). The fortified vil a in Mogorjelo ( Figs. 3.191−3.193, 3.309) was investigated already in the early 20th century, but its function remains a matter of discussion (Dyggve, in the 3rd century and subsequently incorporated in the Vetters 1966; Basler 1972, 38–42, 97–100; Zaninović new complex. The prevailing opinion is that this was a 2002; Paškvalin 2003; Busuladžić 2008; id. 2011, 149; fortified vil a of a large imperial estate supplying the Turković, Maraković 2012, 63). In the late 3rd or early needs of nearby Narona. Its regular rectangular layout 4th century, a large fortified complex with a highly sym- with strong defensive wal s, towers and reinforced en- metrical layout of rooms, measuring 103 × 86 m, was trances, as well as the series of evenly distributed build- built on the spot of an earlier Roman vil a. It was pro- ings, however, are more similar to a military fort that, in tected with four rectangular towers in each of the corners combination with its economic part, protected and also and one round tower, while pairs of protruding towers supplied agricultural produce down the River Neretva flanked three of the entrances and one tower protected (cf. Duval 1989−1990). The complex was abandoned by a smaller entrance. The up to 1.65 m thick defensive the mid-5th century. Afterwards, the stil wel -preserved wal s were constructed in the opus mixtum technique. architecture was turned into a fortified settlement in the Small living quarters and industrial facilities were evenly late 5th century; this last phase was dated indirectly, on distributed along three of the interior sides, while a large the basis of a double church in the centre, which was residential building, with mosaic floor in parts of the built into the ruins of the vil a and deviates completely upper storey, took up the entire south side. Only the from the orientation of earlier buildings, showing these economic part survived of the early vil a burnt down must already have been destroyed at that time. Research 261 Slavko CIGLENEČKI Fig. 3.192: Mogorjelo. Plan of the fortified vil a (Basler 1993, Fig. 4). Fig. 3.193: Mogorjelo. Wal s with residential buildings (2009). 262 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.194: Koštur near Dabrica. Plan of the fort (Ciglenečki 2011b, Fig. 5.22). established repairs to the defensive wal s, while the smal finds reach to the late 6th century and reveal that the local population sought shelter here. Slobodan Ćurčić presumes Mogorjelo was a bishopric see, similarly as Louloudies or Gamzigrad (Ćurčić 2010, 139). The large fort at Koštur near Dabrica near Stolac ( Figs. 3.194−3.197) lies in a mountainous and not readily accessible area above the canyon of the River Radimlja (Basler 1972, 50–51; id. 1988; id. 1993, 32; Ciglenečki 2011b, 678–681). The present-day name of the site originates from the word castrum, which underscores its military nature. Steep and precipitous slopes protect the fort from three sides and it is only accessible from the west, where it rises slightly above the lower part of the ridge. Investigations on the hill were limited to the field surveys of the clearly visible architectural remains. The fort is 180 m long and 65 m wide. It is skilful y adapted to the terrain, though the builders managed to maintain a fairly straight line of the defensive wal s, giving them a near pentagonal outline. The wal s are particularly well-fortified in the easily accessible west side, where they reach 2.3 m in thickness in contrast to 1.1–1.4 m in other parts. Three towers are visible in the interior of the wal s. Associated with them in the southwest is a short wall that guarded the access to the slope below the defensive wal s and thus prevented a potential siege from the south. The wal s are very regularly constructed, with faces of finely dressed rectangular stones. Traces of a small ditch are visible below the western defensive wal s, behind it also a small rampart (possibly proteichisma). Fig. 3.195: Koštur near Dabrica. Wal s on the northern side (1988). In the interior, a large building measuring roughly 42 × 263 Slavko CIGLENEČKI Fig. 3.196: Koštur near Dabrica. Fort from the north-west (1988). 18 m leans against the northern defensive wal s in the part natural y best protected. The building shows several rooms and a hexagonal cistern touching the defensive wal s. There are traces of another building, not identi- fied in detail, near the western defensive wal s. Without further research, it is difficult to date the fort that has been attributed to the 6th century in literature; this dating is supported by the surface finds of amphorae, though its construction may be of an earlier date. The small Late Antique fort at Gradac near Todorovići ( Fig. 3.198) was constructed on a steep slope high above the gorge of the River Bregava and is only known from the data gathered in the late 19th century Fig. 3.197: Koštur near Dabrica. Part of a large building (1988). (Truhelka 1893, 296–298). At that time, Ćiro Truhelka conducted trial trenching and drew the plan of the c. 40 × 30 m large triangular fort. One of its sides was natural y protected with rocky faces and the other two fortified with a 1.3 m thick defensive wall reinforced with a tower where access was easiest. Building remains and several burials were unearthed in the interior along the defensive wal . The smal , but heavily fortified post presumably protected the road that led from the valley to the Erguda Plateau. It may date as early as the 4th or early 5th century, though it may later have been used as a refuge. In size and shape, it is similar to the Justinian period fort at Bosman in Đerdap (see below). Fig. 3.198: Gradac near Todorovići. Plan of the fort (Truhelka 1893, Fig. 68). 264 EASTERN PART The fortress at Veliki Gradac near Donji Milano- vac (presumed Taliata) ( Fig. 3.199) was built in a strategi-Serbia cal y very important location and is consequently quite large, measuring 134 × 126 m (Popović, V. 1984). The Research in Serbia long focused on major Roman beginning of its Late Roman phase is dated to the second cities, particularly the Late Antique fate of Sirmium and half of the 3rd century, when the Roman fort was exten- its historical significance. Much scholarly attention was sively renovated. Following a fire and abandonment in also paid to Caričin grad, investigated now for a century. the first half of the 5th century, it was again renovated in In the 1960s, there were extensive rescue investigations the time of Justinianic fortification of the Danube limes. in the Đerdap area, which revealed many important sites The renovations involved repairing the defensive wal s (for example Lepenski Vir) that include a considerable and raising the rectangular towers. The north entrance number of Roman forts with several renovations in Late was retained and all others walled up. New and larger Antiquity. Large-scale systematic investigations of Late round towers were built in the corners, in the interior a Antique sites in the countryside began in the 1970s with horreum and a church. There are traces of smal housing the investigations of the fortification of Ras. When this units of non-durable materials across the interior. The site was found to be Late Antique rather than medieval, church, leaned against a tower of the defensive wal s, attention began to be more widely paid to the fortifications shows two construction phases, the second one dating to of Late Antiquity. In recent decades, this brought a great the Early Byzantine period. In this last phase, the fortress amount of new topographic data and several trial trench- was more like a fortified vil age than a military camp, ing campaigns, but the results are only briefly published. which is evidenced in the church being renovated and The numerous characteristic Roman and Late the horreum torn down and reused as a metal working Antique forts offer a reliable insight into the creation, facility. It was destroyed in a fire, which has been linked function, development and end of the Empire’s first line to the Avar raids in 595/596. of defence (overview in Petrović 1980; Kondić 1984; The medium-sized, but integral y excavated fort Vasić, Kondić 1986; Vasić 1995). Their architecture and at Boljetin ( Fig. 3.200) clearly shows the characteristic small finds faithful y reflect the increasing calamities that renovation phases in Late Antiquity (Zotović 1984). It befell the Empire in Late Antiquity along its exposed measures 80 × 50 m and was constructed on a partly border on the Danube. The forms of the Late Antique fortified river terrace. It was already built in the 1st, fortifications range from simple towers, fortified ports, extensively renovated in the second half of the 3rd, aban- small forts to fortresses at strategic locations. Sited on doned in the early 5th century and reused in the time of river terraces and on the spots of earlier Roman forts, the Early Byzantine restoration of the limes. This phase they share a fairly regular shape, which distinguishes involved reinforcing the defensive wal s, modifying the them from the multitude of contemporary fortifications entrances and rebuilding all exterior towers. A simple constructed on hil tops. single-aisled church with an apse and narthex was Fig. 3.199: Veliki Gradac near Donji Milanovac. Plan of the Fig. 3.200: Boljetin. Plan of the fort (Zotović 1984, Fig. 2). fortress (Popović, V. 1984, Fig. 4). 265 Slavko CIGLENEČKI 2 1 3 4 5 87 89 7 6 9 8 88 90 76 10 11 85 86 77 12 79 15 78 13 14 20 19 91 16 17 80 92 23 18 24 21 81 25 28 29 22 82 26 27 34 32 45 94 93 35 30 31 83 33 44 95 84 38 42 37 41 51 43 47 46 39 48 36 50 52 53 40 55 54 49 56 58 57 59 73 74 72 75 60 61 71 62 63 70 64 65 66 69 67 68 0 100 km Fig. 3.201a: Newly-founded fortifications in the countryside from the eastern part, discussed in the book. 266 3. SETTLEMENT IN THE COUNTRYSIDE 1 − Bosman 25 − Hum near Tutin 49 − Belgrad near Dvorište 72 − Palaiokastro 2 − Boljetin 26 − Đurđevica in Đerekare 50 − Gradište Stenče 73 − Drama 3 − Veliki Gradac 27 − Tupi krš 51 − Markovi kuli (Vodno) 74 − Ai Giannis 4 − Kraku Lu Jordan 28 − Čečan/Gornji Streoc 52 − Pelenica near Dračevo 75 − Khortokopi 5 − Vrelo - Šarkamen 29 − Kekola near Keqekol a 53 − Markovo Kale near Malčište 76 − Gradishteto (Riben) 6 − Liška Ćava 30 − Gegje 54 − Gradište near Pakoševo 77 − Golemanovo Kale 7 − Bedem near Maskare 31 − Harilaq 55 − Pešna near Devič 78 − Sadovsko Kale 8 − Ukosa in Stalać 32 − Veletin 56 − Kale near Debrešte 79 − Mezdra 9 − Timacum Minus 33 − Korishë/Koriša 57 − Venec near Miokazi 80 − Borovets 10 − Gradina in Radalica 34 − Martinička gradina near Spuž 58 − Gradok (Markovi Kuli) 81 − Chertigrad 11 − Kula in Kaludra 35 − Đuteza near Dinoša near Čanište 82 − Elenska Basilica 12 − Gradina in Vrsenice 36 − Stari grad, Ulcinj 59 − Čebren (Grad) near Zovik 83 − Krasen kale 13 − Samograd 37 − Pecës 60 − Qafa 84 − Markova Mehana 14 − Kulina in Rogatac 38 − Gradishta (Bardhoc) 61 − Golem Grad, Konjsko 85 − Dichin 15 − Gradina Ras 39 − Bushati 62 − Paleokastra near Gjirokastra 86 − Dobri Dyal 16 − Južac near Sopoćani 40 − Domaj 63 − Ҁuka e Ajtojt 87 − Kartal kale near Ruyno 17 − Gaj in Babrež 41 − Gradište (Sobri) near Oraše 64 − Kastritsa 88 − Madara 18 − Zlatni Kamen 42 − Kale near Gorno Svilare 65 − Kastro Rizovouni 89 − Odartsi 19 − Kale in Bregovina 43 − Gradište near Pčinja 66 − Kephalos 90 − Gradishteto (Debrene) 20 − Balajnac near Niš 44 − Žegligovski kamen 67 − Olympia 91 − Harmana 21 − Kale, Zlata 45 − Kula near Čelopek 68 − Monemvasia 92 − Gradishte (Gabrovo) 22 − Hisar above Leskovac 46 − Kalata near Kamenica 69 − Isthmia 93 − Dyadovo 23 − Trojan 48 − Gradište near Delisinci 70 − Velika 94 − Carassura 24 − Gradina Ramoševo 47 − Kula near Kalauzlija 71 − Louloudies 95 − Kaleto (Castra Rubra) Fig. 3.201b: Newly-founded fortifications in the countryside from the eastern part, discussed in the book. Site names correspond with the digits on Fig. 3.201a. built in the central part. The remains of three well-built masonry edifices (presumed workshops) and smaller rectangular features (presumed tent substructures) are mentioned in the fort interior. Small finds consist of coarse and imported wares, farming tools, arrowheads and others. The fort was destroyed in a fierce fire in the last years of the 6th century. A special example in the Đerdap Gorge (Iron Gate) is the fort at Bosman ( Fig. 3.202), which is the only Justinian period construction built completely anew. It was also not repaired or reused later, thus offering the most authentic image of the military architecture of the mid-6th century (Kondić 1984). It was ful y adapted to the terrain and triangular in plan, with 45 m long sides. Fig. 3.202: Bosman. Plan of the fort (Kondić 1984, Fig. 1). The substantial defensive wal s, built in the opus mixtum technique, were 2.3 m thick on average and reinforced with three round projecting towers. The only entrance was from the river, which provided additional protec- 160 × 80 m and is protected with steep slopes and thick tion. The interior held a masonry well and traces of wal s. The complex extends along the southern slope modest buildings with wattle-and-daub wal s and floors down to the river. Investigations thus far examined of beaten loam. The coin hoard from 596, found under seven buildings along the interior of the southern the ruins, very precisely dates the end of the fort and wal s and three square towers. The buildings revealed also of this section of the Danube limes. furnaces, smithies and other workshops that produced Partly investigated in the area of considerable mining tools. The fortification also served as a mining metal ore deposits, particularly the gold deposits in the administration centre, which was presumably located valley of the River Pek, is the fortified metal urgic com- in the western part of the complex. This is where the plex at Kraku Lu Jordan ( Fig. 3.203) (Bartel et al. 1979; production of gold and distribution of gold ingots was Tomović 2000). The fortification lies on a low elongated being controlled, while also providing a place to safely ridge at the confluence of the Rivers Brodička reka and store the gold. Coins and other small finds show that Pek, 70 km south of the Danube limes. It measures the fortified complex was constructed in the late 3rd 267 Slavko CIGLENEČKI Fig. 3.203: Kraku Lu Jordan. Plan of the fortification (from Tomović 2000, Fig. 4). Fig. 3.204: Vrelo - Šarkamen. Plan of the palace (Tomović, Vasić 1997, Fig. 4). century and later gradual y extended. It was destroyed interconnected structures. Most of these were never in a fire in the late 4th century and abandoned. finished as the construction was brought to a halt at one The group of magnificent fortified imperial resi- point. The foundations of the northern defensive wal s dences in Illyricum includes the presumed palace of were finished, though stratigraphic evidence shows no Maximinus Daia at Vrelo - Šarkamen ( Fig. 3.204) signs of habitation; only the mausoleum was finished (Tomović, Vasić 1997; Popović, I. 2005; Mladenović and put to use. It is smaller than the other two palaces 2009). It extends across almost 10 ha and incorporates (Gamzigrad and Diocletian’s palace in Split), discussed a fortification, mausoleum, burial mound and several among the newly-founded cities (see Chapter 2.5), and 268 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.205: Liška Ćava. Plan of the fortified settlement (Milinković 2010, Fig. 276). also less well-known, but enclosed with equal y impres- sive defensive wal s with towers. Future research might offer additional evidence on the fortified palaces that very ostentatiously heralded a new era. In addition to the above-mentioned exceptional forms of Late Antique fortifications, there are many other fortified posts in Serbia, but they were only rarely systematical y investigated. One located in northwest- ern Serbia is at Liška Ćava on a 667 m high and natu- ral y well-protected hill ( Fig. 3.205) (Radičević 2009; Milinković 2010, 210–216). It extends across two levels, one on the smaller summit plateau, partly protected with rocky terrain, and the other on the larger lower part on the northern slope. More is known on the remains from the 155 × 50 m large summit plateau. Its north side is protected with 1.36 m thick mortared defensive wal s. The interior held several residential buildings with clay- bonded wal s. A surprising discovery is a smal , only Fig. 3.206: Timacum Minus. Plan of the fort (Petrović 1986, 11 × 6.2 m large Early Christian church with an apse, Fig. 2). also of clay-bonded wal s. Another church, of a similarly modest construction and only very slightly larger, was excavated on the ridge next to the fortification. Small is a typical Roman cohort fort with the surface of 1.7 ha, finds, predominantly coarseware and farming tools, built in the lowland next to a river. Research established but also two precious brooches, date to the 6th century three phases, the last one dating to the 4th century and and reflect a heavily ruralised autarchic hil top settle- involving extensive renovations and the construction of ment that persisted to the late 6th century or possibly thicker (up to 3 m) defensive wal s with large protruding slightly longer. rectangular towers. Coins and other small finds indicate The fort of Timacum Minus ( Fig. 3.206) lies in the this well-protected fort was also in use in the 5th and 6th valley of the River Timok on an important communica- centuries, when there were no more repairs of the defen- tion that connected the Adriatic (Lissus) with the Danube sive features, only adaptations to the interior structures Basin (Ratiaria) (Petrović 1986; Petković, Ilijić 2012). It at the entrances and along the wal s. 269 Slavko CIGLENEČKI Bedem near Maskare ( Fig. 3.207), located at the can be seen as the consequence of destruction in a Hun confluence of the Rivers Južna Morava and Zahodna incursion. The fort is believed to have been renovated Morava, is the only lowland Early Byzantine fort from in the late 5th and abandoned in the late 6th or early 7th the 6th century in Serbia (Milinković 2015, 63; Rašković century. 2021, 279–281). It is protected with both rivers and up A large number of Late Antique fortifications came to 3 m thick defensive wal s, and played an important to light in southwestern Serbia, in the area of Novi Pazar strategic role, possibly even controlled the river traffic. and Raška. We will only present those that best il ustrate In the 130 × 130 m large fort, excavations revealed a the last posts before the decline of the Empire. large three-aisled church. In the central Balkans, a very telling example is Ukosa in Stalać holds an example of a significant the fortification at Gradina Ras ( Fig. 3.208) (Popović, Early Byzantine settlement built on a long ridge above a M. 1999, 73–103). It is one of the first systematical y gorge (Rašković 2016). The few trial trenches at the site investigated sites of this type in Serbia with reliably dated confirmed the existence of defensive wal s bound with phases of use. The Late Antique phases are only one seg- lime mortar. The wal s survive up to 1 m high, while ment of the settlement on this high rocky hill above the their thickness has not been established. A thick Early confluence of the Rivers Raška and Ibar, in an area rich Byzantine layer contained numerous coins, pottery, in metal ore deposits. The habitation remains from the glass, bone products, a variety of artisanal and farming middle and second half of the 3rd century are primarily tools and a single spearhead. Also mentioned are a few dated with coins. These remains were found on the east- storage pits. The coin hoard from the mid-5th century ern slope of the hil , at Podgrađe, where Marko Popović Fig. 3.207: Bedem near Maskare. Plan of the fort (Milinković 2015, Fig. 27). 270 3. SETTLEMENT IN THE COUNTRYSIDE presumes a masonry building representing a speculum; the coins from sealed contexts, as the steep terrain he also dates a timber house on the slope to the mid-3rd often makes the stratigraphic record unclear. For this century. Stray finds from this time, predominantly coins, reason, continued habitation does seem questionable, as also came to light on top of Gradina and show the first, the analysis does not take sufficiently into account the short-term habitation on the hill in the second half of possibility of later use. Analysis did provide a relative the 3rd century, when the natural defences of the rocky chronology of individual buildings and their renova- domed hil top sufficed. Popović observes that the hill tions. At the foot of the hil , investigations also revealed was fortified in the initial decades of the 4th century. the remains of a three-aisled church; this is the largest It incorporated the prehistoric hil fort and the slope church in the area of Novi Pazar and was presumably (Podgrađe) with the two above-mentioned buildings built contemporaneously with the renovations of the from the 3rd century. The defensive wal s are completely defensive wal s. The end of the settlement cannot be adapted to the terrain and enclose a surface in excess reliably dated, but presumably occurred at the transition of 5 ha. They are 1.2 m thick and erected in places of from the 6th to the 7th century. easier access, while elsewhere the site’s defences relied The fort at Gradina in Vrsenice ( Fig. 3.209) was on the steep rocky slopes. Only three towers were built. systematical y investigated and comprehensively pub- Two parallel wal s form a narrow entrance into the fort. lished (Popović, Bikić 2009; Milinković 2015, 259). It lies Even though the fortification is vast, the steep slopes on a hill with steep slopes that greatly hinder access, with only allow buildings to be constructed in the levelled the summit located at 1330 m asl. The first phase dates to parts along the defensive wal s. Twenty-two Late An- the 2nd and 3rd centuries, when a building of a military tique houses were investigated, which revealed several function was put up, serving primarily as a watch post. In construction phases. Most were timber buildings on the last third of the 4th century (360 or 370s), 1.2–1.5 m drystone foundations, only few were masonry houses. thick wal s were built on the hil top that enclosed a c. The remaining space was presumably used to shelter 0.5 ha large area. A tower was incorporated in the area people and livestock. Stratigraphic evidence shows the of easiest access. The earlier building was renovated and fort was continuously inhabited from the early 4th to the added a new, longer building next to the defensive wal s, late 6th century. Within this time, the first phase dates both serving to house the troops. Only some timber to the 4th and 5th, the second to the 6th century and is buildings were found in the fort interior. In Late Roman marked by the renovation of the defensive wal s in the times, the fort is believed to have protected the road used late 520s or the 530s. At Podgrađe, part of the defensive to transport valuable metals from the mining centres wall renovation dates to the first decades of Justinian’s along the border between the provinces Dalmatia and reign. Popović primarily based this interpretation on Upper Moesia to the valley of the Ibar. There are no reli- Fig. 3.208: Gradina Ras. Plan of the fortification (Popović, M. 1999, Fig. 8). 271 Slavko CIGLENEČKI able traces of the fort being continual y inhabited to the to suggest an increased military presence. The fortified 5th century. After the fort fel into ruin, which is believed settlement is believed to have primarily functioned as the to have occurred as the consequence of the earthquake shelter for the local population, though the possibility in 518, it was again renovated in the second quarter or it formed part of the regional defence system, as it did even the mid-6th century. The old defensive wal s were earlier in the Late Roman period, cannot be excluded. repaired and additional, 1.4 m thick wal s (proteichisma) We should also briefly mention several sites only constructed at a distance of 10−20 m from the earlier detected through field surveys and trial trenching, but wal s. Also built was a small and modest single-aisled which importantly contribute to our understanding of the church with an apse and narthex, in the fort interior also chronology and function of the sites. One is the hil top several timber buildings; the tower is believed to have fortified settlement at Zlatni Kamen near Novi Pazar been renovated at this time as wel . The cultural layer ( Fig. 3.210) (Ivanišević 1990). This high hill (953 m asl) shows greater intensity of use than in the earlier period with steep sides revealed a 176 × 66 m large settlement and the number of inhabitants increased; they mainly girded with defensive wal s and hosting a small church practised livestock farming, while there are no elements and several other buildings indicative of a more per- manent settlement. Investigations also yielded furnaces and slag, as well as numerous deposits of iron ore in the vicinity. It was first settled in the 4th century. It was also inhabited in the 6th century, when the settlement included the above-mentioned church, the smaller buildings and yielded a number of small finds. The mortared 0.9−1 m thick wal s were reinforced with two small towers in the part of easier access. The modest cultural layer shows that the occupation of the hil top in the 6th century was of a short duration (Ivanišević 1990, 11). The construction of the fortification at Južac near Sopoćani ( Fig. 3.211) was reliably dated to the 4th/5th centuries (Popović, M. 1987). At that time, 1.4 m thick and mortared defensive wal s were built on top of the prehistoric hillfort and reinforced with a protruding rec- tangular tower in the area of easier access. Several trial trenches explored the 150 × 65 m large settlement and unearthed the remains of a burnt-down drystone and timber construction next to the wal s. It was a building Fig. 3.209: Gradina in Vrsenice. Plan of the fortified settlement that small finds attribute to the second half of the 4th or (Popović, Bikić 2009, Fig. 32). first half of the 5th century. Later, a masonry building Fig. 3.210: Zlatni Kamen near Novi Pazar. Plan of the fortified settlement (from Ivanišević 1990, Fig. 1). 272 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.212: Hum near Tutin. Plan of the fortification(Ivanišević 1988, Fig. 1). Fig. 3.211: Južac near Sopoćani. Plan of the fortification (Popović, M. 1987, Fig. 1). was put up on the same spot, which together with small finds shows the fortification was renovated and used in the 6th century. A characteristic fortified refuge is located at Hum near Tutin ( Fig. 3.212), at an altitude of 1502 m asl (Ivanišević 1988). The smal , 75 × 50 m large fortified area protected with 0.9–1 m thick wal s only revealed a thin cultural layer from the 6th century. Of a similar nature is the site at Gradina Ramoševo Fig. 3.213: Gradina Ramoševo near Tutin. Plan of the fortifica-near Tutin ( Fig. 3.213) (Ivanišević 1987). The hil top tion (Ivanišević 1987, Fig. 1). revealed 1.4 m thick and mortared defensive walls and two protruding towers that Late Antique pottery roughly dates to the 4th−6th centuries. No buildings were observed in the interior. constructed on the hil top (1351 m asl) that encircled One of the earliest fortified sites at the edge of the a 210 × 85 m large area. In the 3rd century, the rampart Pešter plateau is that at Trojan ( Fig. 3.214) (Ivanišević (presumably also the palisade) was renovated, but 1989). Trial trenching explored the basic elements of without using mortar. Rare coins and other small finds the fortification that draws its origin from the prehis- date the modest habitation remains to the first half of toric period, when a substantial earthwork rampart was the 3rd century, while the settlement or visits to the site 273 Slavko CIGLENEČKI Fig. 3.215: Kulina in Rogatac. Plan of the fortification (Simić 1987, Fig. 1). Fig. 3.214: Trojan. Plan of the fortification (Milinković 2007, Fig. 7). presumably continued to the late 4th century. The interior revealed no building traces. Vujadin Ivanišević posits that it accommodated a small garrison controlling the nearby communications, though the size of the forti- fication suggests it may also have been in refugial use. The hil top site at Kulina in Rogatac ( Fig. 3.215) was likely inhabited only briefly (Simić 1987). The 1.4 m thick wal s protect the 100 × 60 m large site in the north, west and south, with steep slopes providing additional protec- tion. A rectangular tower was found in the southwest. We should also mention a thick earthwork rampart in the function of a proteichisma located in the part of easier access. The interior only revealed a thin cultural layer from the first half of the 6th century and no building remains. The fortification at Gaj in Babrež ( Fig. 3.216) is a a single-phase site from the 6th century, located on a Fig. 3.216: Gaj in Babrež. Plan of the fortification (Premović- hill dominating the valley of the River Jošanićka reka Aleksić 1989, Fig. 18). (Premović-Aleksić 1989). The 1.6–1.9 m thick defensive wal s girdle a 100 × 40 m large area of the hil top. At the highest point, the wal s are reinforced with a U-shaped The fortification at Gradina in Radalica ( Fig. tower that combined defensive and watch functions, 3.217) lies on a dominant hill affording control over the while its interior held the remains of a smelting furnace. surrounding areas (Kalić, Mrkobrad 1985). Remains of Two other masonry buildings were found at the wal s. buildings are visible in several places, particularly nu- The cultural layer dates to the 6th century. The location merously near the defensive wal s. The c. 60 × 40 m large and the thickness of the wal s suggest the fortification fortification is believed to have been established in the controled traffic and surrounding territory. 6th century, though it also revealed pottery from the 4th 274 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.218: Tupi krš. Plan of the fortified settlement (Milinković 1985, Fig. 1). Fig. 3.217: Gradina in Radalica. Plan of the fortified settlement (Kalić, Mrkobrad 1985, Pl. 1). century. The defensive wal s are 1.4 m thick and reinforced with a rectangular tower at the highest point (1313 m). A small fortification was built on the high hill of Tupi krš ( Fig. 3.218) (1286 m) that enabled control over the upper reaches of the River Ibar (Milinković 1985). The 47 × 33 m large fortified settlement was in one part enclosed with 1.05 m thick wal s and elsewhere protected by rocky faces. Several buildings are visible in the interior, one of which was trial trenched and found to have dry-stone foundations. The small finds indicate its construction in the 6th century. The site at Kula in Kaludra ( Fig. 3.219) is heavily fortified (Popović, M. 1984). The 120 × 60 m large fort was built on a natural y well-protected hill at 1041 m asl and additional y protected with substantial defensive wal s and presumably eight protruding towers. The origi- nal defensive wal s were 1.2 m thick and later reinforced to the thickness of 1.8 m. Investigations revealed two phases, both dating to the 6th century. In the first, the wal s and towers were built. The second phase followed soon after and involved reinforcing the wal s and add- Fig. 3.219: Kula in Kaludra. Plan of the fort (from Popović, ing a staircase leading to the top of the defensive wal s. M. 1984, Fig. 1). The east part, where access was easier, was additional y protected with a proteichisma at a distance of c. 10 m from the wal s. At a distance of 20–30 m, a substantial The fort at Đurđevica in Đerekare ( Fig. 3.220) is al- earthwork rampart was also built that enclosed the so heavily fortified and measures 110 × 60 m (Milinković fort from three sides. The modest pottery remains and 1983). It is encircled with 1.8–2.5 m thick wal s with particularly architectural features date the fort to the three towers along the part of easier access and elsewhere 6th century, with possible limited occupation already protected with rocky faces. The uniform cultural layer in the 4th century. is attributable to the 6th century. A damaged burial of a child was found at the wal s. Milinković presumes the 275 Slavko CIGLENEČKI Fig. 3.220: Đurđevica in Đerekare. Plan of the fort (Milinković Fig. 3.222: Balajnac near Niš. Plan of the cistern (Jeremić, G. 1983, Pl. 1). 1995, Fig. 10). Fig. 3.221: Balajnac near Niš. Plan of the fort (Ćurčić 2010, Fig. 3.223: Kale in Bregovina. Plan of the fortified settlement Fig. 183). (Milinković 2015, Fig. 5). fort formed part of a wider system of defence, but also slope. The 2 m thick defensive wal s are reinforced with one that could serve as a refuge, as it was defended by round and semicircular towers. The upper part measures civilians with their families. 113 × 70 m, the lower one 135 × 50 m. Trial trenching The same time frame, but greater quality of construc- campaigns in the fort only unearthed two large build- tion can be observed in the unusual fort at Balajnac near ings, presuming the interior was not densely inhabited. A Niš ( Figs. 3.221, 3.222) (Jeremić, G. 1995). It was built on three-aisled basilica with an apse and high-quality marble a low hill some 60 m above the flatland, in the vicinity of furnishings was partial y excavated. A special feature is the road that led from Naissus to Caričin grad. It has two the 17.2 × 15.5 m large cistern with up to 1.95 m thick parts. The upper part is enclosed with rectilinear defensive wal s, which was very well constructed. Other wal s were wal s and makes use of the natural y well-protected sum- unearthed in its proximity, possibly even the transverse mit of the hil . The lower part extends across the western wall of the fort. The elements found thus far reveal a fort 276 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.224: Kale, Zlata. Plan of the fortified settlement (Milinković 2015, Fig. 170). of a specific character, with wal s built in the opus mixtum part of the settlement uncovered and damaged part of technique, which is rare in this area; also rare are round a church built of brick and fitted with a multicoloured towers. Even more unusual is the presence of two large mosaic floor. Reports also mention a great amount of buildings and absence of other, particularly residential architectural decoration and fragments of wall paintings. buildings. The site also yielded an exceptional find, Trial trenching established a rich cultural layer and two namely the bronze head of a Byzantine empress (Srejović, habitation phases, the first of which dates to the 6th and Simović 1959). early 7th century. It also unearthed part of a building There are several fortified Early Byzantine settle- with wal s of clay-bonded rubble. ments in the vicinity of Caričin grad (cf. Ivanišević et al. The fortification at Hisar above Leskovac lies on 2019), which include the partial y excavated settlement a dominant and natural y well-protected spot at 341 m at Kale in Bregovina ( Fig. 3.223) (Jeremić, Milinković asl (Stamenković 2013, 143). Investigations revealed 1995; Milinković 2015, 228–236). It was built on a low part of the defensive wal s and a small brick wall in the and easily accessible hill. The well-constructed and interior. Small finds date the first phase of the fort to the 1.6–1.9 m thick defensive wal s in the opus mixtum tech- 4th century (most likely second half), persisting to the nique form two defensive belts. The smaller, 60 × 65 m Hun incursions. Numerous small finds show a renova- large inner part with at least five protruding semicircular tion under Justinian and a continuation to the late 6th or towers revealed a richly decorated three-aisled basilica. early 7th century. A coin hoard was also found, dated to Paral el to it is a larger and partitioned masonry building 573/574. An Early Byzantine settlement existed on the with pilasters of a prestigious nature. Several smaller, but slope below, where slag and other finds indicate smelting poorly known drystone edifices are also mentioned. The and iron working. small finds include many farming tools and weapons. The outer belt is poorly known, as is its interior. The large fortified settlement at Kale, Zlata ( Fig. Montenegro 3.224), located 13 km from Caričin grad, is only known in outlines. It was built on a large elongated and natu- The newly-founded Late Antique fortification in ral y protected plateau and its size has not yet been ful y Montenegro are poorly known, which should primar- established (the inner part measures 165 × 100 m). An ily be attributed to insufficient fieldwork. There are, aqueduct was found outside the settlement, also visible however, several examples where a Late Antique phase is a large dam below it (Milinković 2015, 236–248). The has been established on previously known archaeologi- dam with the surviving size of c. 100 × 4.4 × 6 m is a cal sites. One of these is the fortified hil top settlement feat of hydrotechnical engineering (Milinković 2015, at Martinička gradina near Spuž ( Fig. 3.225) (Mlakar 242, 247). Recent construction work in the highest 1961, 224; Korać 2001; Stevović 2014, 102–105). The 277 Slavko CIGLENEČKI Fig. 3.225: Martinička gradina near Spuž. Plan of the fortified Fig. 3.226: Đuteza near Dinoša. Plan of the fort (from Bugaj settlement (Stevović 2014, Fig. 89). et al. 2013, Fig. 3). high hill above the fertile valley of the River Zeta shows at a distance of 8−20 m from the wal s. Archaeological 1.2–1.5 m thick defensive wal s with several protruding prospection and small-scale investigations in the fort rectangular and semicircular towers. The wal s enclose interior have revealed a simple apsidal church and traces the top of the hill and the vast, c. 2.5 ha large southern of other masonry buildings. The small finds are poorly slope. In its design and the form of the wal s with towers, known, while the defensive architecture and the simple it is similar to other Late Antique fortifications, though church show similarities with those from other Early poor investigations do not allow us to distinguish be- Byzantine forts in the area. tween Late Antique and later, medieval elements. The Only preliminary reports speak of the unusual hil top holds a three-aisled church with three apses, fortified settlement at Samograd (Mrkobrad, Jovanović which was built in the 6th century and renovated in the 1989; Mrkobrad et al. 1990). It rises almost impercep- Middle Ages. The remains of another, single-aisled apsed tibly above the valley of the River Brzava, but its core Early Christian church lie at the foot of the hil , which is protected from three sides with imposing vertical was later added a clergy bench and side rooms. rocky faces. The defended area measures c. 100 × 80 m. The construction of the fort at Đuteza near Dinoša Substantial defensive wal s guard the access to the set- ( Fig. 3.226) skilful y used the hil ’s natural defences and tlement. The interior holds the remains of prehistoric the earlier prehistoric earthwork ramparts (Velimirović habitation and three Roman and Late Antique phases. Žižić 1986; Bugaj et al. 2013). The hill with steep slopes The first defensive wal s are believed to have been built rises high above the fertile plain and is only readily ac- in the mid-2nd century, but the settlement was already cessible from the north, where a low saddle separates it destroyed in the mid-3rd century. The next phase lasted from the surroundings hil s. A small Late Antique fort from the 4th to the mid-5th century, after which the ( c. 65 × 42 m) was built within the larger prehistoric settlement was again abandoned. Intense occupation hillfort with impressive ramparts. In the south, the fort followed in the 6th century, when one Early Christian leans against the rampart of the prehistoric acropolis. church was erected in the interior and another one Elsewhere it is enclosed with 1.5–1.6 m thick, mortared outside the wal s. The area was again fortified with even wal s reinforced with three roughly horseshoe-shaped thicker wal s, measuring 1.7 m in the lower and 1.3 m towers. Two of the towers are smal , the third one large in the upper part. The settlement has been associated and hosting an additional entrance into the fort. As also with the exploitation of the lead, zinc, copper and iron observed at other fortifications, the Late Antique wal s ore deposits in the area. were constructed so that they were additional y pro- The area of Stari grad in Ulcinj, in the citadel inside tected with the prehistoric ramparts in the parts of easier medieval wal s, revealed highly mixed cultural layers that access or the ramparts were reused as a proteichisma included a Late Antique cultural layer and architectural 278 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.227: Harilaq. Plan of the fortification (Rama 2020, Fig. 6). remains (Mijović 1984–85, 73–79; Marković, Stanković (Ivanišević, Špehar 2005) are also only known from field 1987). The citadel was constructed on a rocky promon- surveys. Both have a natural y well-protected location tory, which is natural y well-protected and suitable for and defensive wal s, and are important for the many a fortification, and yielded rich remains from different small finds that shed light on the intensity of habitation periods. Found above the Hellenistic wal s were modest on the hil top settlements of the former province of remains of Late Antique defensive wal s dating to the Dardania. The published finds show at least two habita- Justinian period, while different structures, burials and a tion phases. Coins and other finds date the first phase to characteristical y Early Christian tomb came to light in the 3rd and 4th centuries, while the multitude of decora- the interior. The same time frame is also posited for the tive and everyday items of metal, bone and glass, as well remains of a cistern and a church indicated by decora- as tools and weapons indicate more intense habitation tive elements of furnishing. Although barely anything in the 6th century. This multitude includes many Early is known of the Late Antique phase of the citadel, its Byzantine coins spanning from those of Anastasius to natural y protected location above a favourable harbour Justin II. Authors emphasise the finds of farming tools, represents an ideal location for a well-defended settle- for woodworking and processing leather and posit the ment. The decision to build a fortification here would existence of a workshops for bone and glass goods. The have been made easier by the already existing Hellenistic published selection of finds clearly shows an autarchic defensive wal s and Roman structures. Numerous smal nature of the fortified settlements, where we can only finds support habitation in the Roman and Late Antique expect a limited production of goods for export (the periods. The fortification most likely formed part of the multitude of keys indicates the activity of a locksmith). Justinianic system of protecting navigable routes along An outstanding hillfort fortification with a pro- the eastern coast of the Adriatic. nounced sacral component has been investigated at Harilaq ( Fig. 3.227) (Rama 2020, 117–118). The exposed hill above the flatland west of Ulpiana hosts a fortifica-Kosovo tion of a triangular plan enclosed with defensive wal s reinforced with four towers and with four entrances. In For a long time, not much was known on the Late the west, where access is easiest, a large rectangular inner Antique countryside in Kosovo. Recently, however, in- tower was constructed that has a triangular projection vestigations have been conducted here as well and the on the exterior side of the wal . In the highest part, a results have already been discussed in the first overviews church complex was built in the opus mixtum technique (Përzhita, Hoxha 2003; Rama 2020). Field surveys estab- and composed of a three-aisled basilica in the centre lished over a hundred Late Antique fortifications (Rama and a pair of sacral buildings symmetrical y in front of 2020, 121). The hil top sites at Čečan and Gornji Streoc it. The complex was visible from far and was certainly 279 Slavko CIGLENEČKI Fig. 3.229: Gegje in Đakovica. Plan of the fort (Përzhita, Hoxha 2003, Fig. 34). Fig. 3.228: Veletin. Plan of the fortification (Rama 2020, Fig. 20). of great significance for the population. The two small sacral buildings are seen as martyria honouring Florus and Laurus, two local saints, and the fortified complex is seen as a pilgrimage site. The geostrategic location and size of the fortifica- tion at Veletin ( Fig. 3.228) show this was an important site in Late Antiquity, which has seen only partial investigations (Jovanović 2004; Rama 2020, 129). It lies on an exposed hil , only 5 km from Ulpiana and in proximity to major mining areas. It offers a good view over a vast area of the Kosovo polje and its lines of communication. Its significance is mirrored in habitation traces from several prehistoric periods. In Late Antiquity, the site was heav- ily fortified. The c. 1.2 ha large surface extended over the summit plateau and several terraces on the slopes. Its wal s were well-adapted to the terrain. Access was easi- est from the saddle in the north. Publications mention the remains of a semicircular tower, in the interior also architectural remains. The fort at Kekola near Keqekolla lies in the moun- tainous area northeast of Pristina and has only recently Fig. 3.230: Korishë/Koriša. Plan of the fortified settlement been investigated (Rama 2020, 124–125). It is natural y (Përzhita, Hoxha 2003, Fig. 55). protected from three sides, while in the south, where access is easiest, it has defensive wal s with a rectan- gular tower. In its first phase, the tower was part of the remains of houses and a simple apsidal church. The fort defensive wal s. It was additional y reinforced from the is dated to the 6th century, though the first phase may interior in the second phase and a triangular projection be slightly earlier. was added on the exterior, giving the tower a pentagonal The Late Roman fort at Gegje in the south part of plan. The fort with a surface of 0.6 ha is enclosed with the city of Gjakovë/Đakovica ( Fig. 3.229) is well-known 1.8−2 m thick defensive wal s. The interior revealed the in plan (Përzhita, Hoxha 2003, 69–73). Field surveys 280 3. SETTLEMENT IN THE COUNTRYSIDE roughly established the line of the defensive wal s, as wel defensive wal s were renovated then and multi-room as rectangular, semicircular and U-shaped towers. The buildings erected in the interior. The thin wal s and 150 × 90 m large fort was constructed on a low hill to low-quality construction indicate their upper parts were protect the nearby communications. The plan, shape of made of sun-dried mudbrick. Investigations in the north the towers and small finds date it to the early 4th century. part of the defended area revealed that the settlement The steep hill near the vil age of Korishë/Koriša was partly renovated in the second half of the 4th century. ( Fig. 3.230) holds the remains of a Late Antique forti- The renovation was presumably the consequence of fied settlement only known from field surveys that an earthquake. A crisis occurred in the first half of the revealed an interesting plan (Përzhita, Hoxha 2003, 5th century, attributed to the combination of a natural 97–101, 149–150). It comprises the upper part on the disaster (earthquake) and a devastating Hun attack. The fortified summit plateau with remains of housing and a settlement ceased to exist for several decades. Only in church, as well as the lower part on the terraces of the the late 5th or initial decades of the 6th century came slope that revealed the remains of masonry buildings. large-scale reconstruction efforts, when new defensive The 1.4–1.8 m thick defensive wal s presumably also wal s were built on top of the destroyed earlier wal s. The included several towers. defended area was increased with the construction of a short proteichisma in the east, at a distance of 5–7 m from the defensive wal s. Earlier edifices in the interior Bulgaria were torn down, which raised the ground level. Buildings were constructed with thicker wal s compared to those Bulgaria has a long history of research of Late of the Late Roman phase, many of them had more than Antique fortifications, particularly if also considering one storey. Another attack presumably brought about the work of brothers Škorpil in the early 20th century another renovation of the defensive wal s and buildings (cf. Boshnakov 2007). The very important and widely in the second half of the 6th century, marked by a low known work was the investigations by a German-Bul- quality of construction. The end of the monetary circu- garian team on the fortified settlement at Golemanovo lation and other finds prove that life in the settlement Kale, which together with the adjacent fort at Sadovsko ended in the late 6th century, which Sergey Torbatov Kale became the reference point for the study of Late associates with Avaro-Slavic incursions. Antique fortifications in a wider area (Uenze 1992). The fortification at Borovets near Pravets is In recent times, systematic surveys of the countryside located on a difficult-to-access hill at 847 m asl (Gri- and excavations at Dichin have been conducted in gorov 2011b). It was irregular in plan, measured 88 × col aboration with a British team (Poulter 2007c). The 62 m and was enclosed with 1.35–1.6 m thick wal s. first publications aimed at systematical y presenting the The highest point of the hil , which also holds remains fortified posts appeared in the 1980s (Ovčarov 1982) and from earlier periods, revealed a two-room building have been growing more numerous (Dintchev 1997; id. that presumably served either as the living quarters of a 2007; id. 2021b). military commander or a sanctuary. Coins of emperors from Gordian III to Aurelian and other small finds date Very instructive is the multi-period fort at Mezdra the building and the fort to the third quarter of the 3rd ( Fig. 3.231, 3.232) (Torbatov 2015). It lies on a rocky century. Also found was a fortified entrance initial y plateau above the River Iskar, at the edge of the city of protected with a rectangular tower in the south and a Mezdra in northwestern Bulgaria, along the Roman thick wall in the north. Later, in the second phase, the trans-Balkan road that led from Oescus on the Danube south tower was torn down and across it another, 2 m to Serdica. Investigations established several prehis- thick wall erected. The small finds from the entrance area toric habitation phases, a Roman settlement and in date also to the second half of the 3rd century. the mid-2nd century the construction of a Roman fort. Two of the Late Antique fortifications with earliest The fortification and the associated civilian settlement investigations were those at Golemanovo Kale and Sa- were destroyed in 170–190. Renovations began in the dovsko Kale, near the vil age of Sadovec (Uenze 1992). late 2nd or early 3rd century, but were never finished. A They are located only 500 m apart as the crow flies, on sanctuary was soon constructed within the ruinous wal s the opposite sides of the River Vit and overseeing the and developed into a large religious centre. Different valley that opens up not far behind them. More extensive Graeco-Roman, but also Oriental and local deities were investigations were conducted at Golemanovo Kale ( Fig. worshipped here. The second phase of the sanctuary 3.233), which is an excellent example of a fortified hil top dates to the second half of the 3rd century when another settlement with several phases of development (Uenze building was constructed, presumably a second temple. 1992; Werner 1992). It lies on a rocky peak east of the The sanctuary existed to the initial decades of the 4th River Vit, with the southern slopes gently descending century. A new construction horizon from the late 300s towards the valley and a short, but steep north slope or early 310s marks the end of the sacral complex. The that separates it from the fertile plain below. In the Late 281 Slavko CIGLENEČKI Fig. 3.231: Mezdra. Plan of the fortified settlement (second fourth of the 4th century) (Torbatov 2015, Fig. 8). Fig. 3.232: Mezdra. Plan of the fortified settlement (second half of the 6th century) (Torbatov 2015, Fig. 10). Roman period, a 12 × 11 m large multi-storey tower rooms. The houses were predominantly two-storeyed with 2 m thick wal s was constructed in the highest with clay-bonded stone foundations, and wood or other part, connected with defensive wal s. The quality of non-durable materials used for the upper storeys. The construction suggests that the Roman army erected ‘Nestor house’ had two rooms and stands out in its size the wal s. Investigations unearthed no contemporary and rich interior furnishings; it was the only one with buildings in the interior. This fort is believed to have glass window panes and also revealed a hoard of gold been destroyed in the early 5th century. Under Justinian coins, indicating a high social status of the owner. The (presumably around 540), it was densely inhabited in its numerous finds of farming tools, woodworking and full extent, boasting 35 to 40 houses and 40 to 50 store- smithing point to an autarchic farming community. 282 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.233: Golemanovo Kale. Plan of the fortified settlement (Vetters 1950, Fig. 5). Leaned against the defensive wal s in the upper part of down. The eastern, higher part has not been investigated, the settlement is a two-storey church where mass was but seems also to hold building remains. Below the 1.7 m held in the upper storey, while the ground floor held a thick defensive wal s is a 1.8 m thick proteichisma, which baptistery. Another church from the Justinian period, protects a large area on the western slope. An additional not of a cemeterial nature, was found outside the settle- defensive feature is a strong, multi-storeyed tower with ment. Around 560, the settlement was burnt down but 2.7 m thick wal s. Leaning on the wal s and the tower resettled immediately afterwards. It remained inhabited are casemates. The wal s of the buildings in the interior at least to a time soon after 584. In this last phase, the were clay-bonded, similarly as those in the interior of exterior church had already been abandoned and the the settlement at Golemanovo Kale. The structure of the lower part of the interior church was filled with stones. coins and the similar construction of wal s with those of The settlement may be seen as a fortified vil age, where the adjacent fort suggest the fort at Sadovsko Kale was the Byzantine army led the construction of the defensive constructed under Justinian, most likely around 540. wal s, while the locals built-up the interior. The farmers The interior revealed no churches and shows a more living in the vil age supplied foodstuffs to the army and military character (Werner 1992, 415–417). The ground the hoards of gold coins indicate that surplus goods floor of the tower revealed traces of a smithery. The silver were sold to the border troops and civilian population. mounts of two belts and a pair of silver brooches prove The fort at Sadovsko Kale ( Fig. 3.234) shows a the presence of an important, most likely Gothic family different layout and a different nature (Uenze 1992). It as mercenaries in Byzantine service. All other finds point was established and abandoned in the 6th century, hence to the presence of the autochthonous population, living representing an important monument of the fortified in a settlement of farmers and artisans who defended architecture of its day. It is roughly 85 × 50 m large. Inves- themselves when the need arose. In this, it is in stark tigations were conducted in the western part of the ter- contrast with the adjacent fortification at Golemanovo race below the summit and on the proteichisma further Kale that otherwise shows several commonalities. The 283 Slavko CIGLENEČKI Fig. 3.234: Sadovsko Kale. Plan of the fort (Uenze 1992, Fig. 31). Fig. 3.235: Chertigrad near Brousen. Plan of the fortified settlement (Dintchev 2006, Fig. 91). smaller fort with casemate-type dwellings may indicate 3rd or early 4th century that controlled the traffic along the presence of irregular military units. Significant traces the road between Oescus and Philippopolis. Roughly of a devastating fire in combination with trilobate ar- in the mid-4th century, a strong fortification developed rowheads suggest that Avars destroyed the fort in the from the burgus. It witnessed a major cataclysmic event late 6th century. during the Second Gothic War under Valens (376–378) The fortified hil top settlement at Chertigrad near and was temporarily abandoned. It was soon reinhabited Brousen ( Fig. 3.235) was first used as a refuge and later (probably by the foederati) and continued to the mid-5th became permanently inhabited (Velkov, Gočeva 1972; century when it was burnt down and abandoned. Dintchev 2007, 509, 529). It lies hidden in the mountains Investigations at Dichin ( Figs. 3.236−3.238) pro- at 1284 m asl and occupies the site of a prehistoric hillfort. vide important results for understanding the chronology It is natural y protected with steep slopes on all sides and and character of several forts (Poulter 1998, 329–343; additional y protected with a 2.2 m thick wall on the side id. 2007c, 82–94). Only 11 km from Nicopolis, a low of easier access. In the 4th and 5th centuries, the hillfort was domed ridge rises 10 m above the floodplain on the reused as temporary refuge, followed in the 6th century west bank of the River Rositsa and hosts the remains of by the construction of several houses on this difficult- a fort with very well constructed defensive instal ations. to-access hill and constituting a permanent settlement. Its 2 m thick defensive wal s had the inner and outer The systematic investigations at Gradishteto near faces made of limestone blocks interrupted by bands of Riben revealed several Late Antique phases (Torbatov brickwork, while the core was of mortared rubble. The 2018). The fortified hil top settlement was built on the surviving stairs suggest the wal s reached 9 m high. The right bank of the River Vit, on a natural y well-protected wal s integrated rectangular towers, with round towers rocky elevation. The site of several prehistoric settle- in the corners. In front of them was a mortared stone ments and a Roman sanctuary hosted a burgus in the late proteichisma. The interior held a regular layout of large, 284 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.236: Dichin. Plan of the fort (Poulter 2007c, Fig. 13). Fig. 3.237: Dichin. Low domed ridge with the fort from the south-west (2012). mainly clay-bonded buildings on either side of the main the inhabitants performed military tasks and practised street. A small church was also found, with an apse, arable farming. The establishment of the fort has been narthex and stone foundations that supported partly dated to around 400. It was intentional y destroyed be- mortared and partly clay-bonded wal s. One part held a tween 474 and 520, renovated soon after and remained series of storerooms, while equal y-sized buildings (22 × in use to the late 6th century when it was destroyed 13 m) were found on both sides of the street that had in a fire. Poulter sees the settlement as an army fort, storage space on the ground floor and soldiers’ accom- presumed supply base, garrisoned by Gothic foederati. modation in the upper storey. In addition to weapons, Although the fort was planned by military engineers, it the debris layer revealed a number of tools, indicating seemed to have been garrisoned by irregular troops with 285 Slavko CIGLENEČKI Fig. 3.238: Dichin. Remains of the buildings by the defensive wall (2012). policing tasks who also farmed the surrounding land. The imported goods and amphorae indicate a supply system associated with the annona. The small Late Roman fort at Dobri Dyal ( Fig. 3.239) was found 15 km from Veliko Tarnovo (Poulter 2013). The 20 m high hill has 2.15–2.50 m thick defen- sive wal s. The interior held several buildings with thick clay-bonded wal s and some sunken-featured buildings. The hundreds of coins show that the circulation first increased in the third quarter of the 3rd century. Based on the absence of masonry structures, Poulter argues that the site served as refuge for the local population who possibly reused the earlier, Bronze Age ramparts. The coins reliably place the beginning of the strong fort to the late 4th and early 5th century, planned and built by Roman military engineers. It was presumably manned by a small garrison (fort size of only 0.4 ha) controlling the junction of two major roads. Small finds show the presence of both soldiers and civilians. The fort was abandoned in the 440s. At Kartal kale near Ruyno ( Fig. 3.240), a small refuge came to light in the east corner of the later medieval fortress (Atanasov 2015). The Late Antique refuge is located on a rocky and natural y well-protected hill surrounded by the dry riverbed of the Senebir. It is triangular in plan and protected with an imposing Fig. 3.239: Dobri Dyal. Plan of the fort (Poulter 2013, Fig. 4). entrance tower or two U-shaped towers in the part of easiest access. The 1.5 m thick wal s are bound by strong white mortar. There are two settlements from the 2nd−4th A small (0.37 ha), but impressive fort was construct- centuries in the vicinity. The small finds from the refuge ed on a high rocky plateau above the plain near Madara date to the late 4th and early 5th centuries and reveal ( Figs. 3.241−3.243) (Dintchev 2007, 522), mentioned it was abandoned following the Hun incursions that above as the site of a Late Roman vil a (see Chapter 3.2.2). devastated the area of Dobrudja after 422, particularly It is believed to have been established in the late 5th and between 434 and 447. inhabited throughout the 6th century. It was protected 286 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.240: Kartal kale near Ruyno. Plan of the fortification (Atanasov 2015, Fig. 7). with rocky faces and thick wal s with towers. The main entrance was reinforced with a pair of pentagonal tow- ers and a propugnaculum. A large building for the troops leaned against the wal s. Only three small single-room buildings are known in the interior. The water supply came from the spring in a nearby cave accessed via a rock- cut stairway. The plateau in the vicinity of the settlement holds the remains of a contemporary, but poorly known unfortified settlement. A good example to il ustrate the evolution of a Late Antique fortification is that at Gradishteto near De- brene ( Fig. 3.244) (Torbatov 2002a, 346–348; Dintchev Fig. 3.241: Madara. Plan of the fort (Dintchev 2006, Fig. 87). Fig. 3.242: Madara. Position of the fort on a high rocky plateau (2012). 287 Slavko CIGLENEČKI Fig. 3.243: Madara. View through the entrance to the interior of the fort (2012). 2007, 509, 530). In the second half of the 4th century, a this small and natural y well-protected plateau with the surface of 0.23 ha was partial y fortified. On two sides, it was protected by the steep slope of a deep natural depression, while clay-bonded wal s were constructed on the side of easier access; this part was additional y protected with a ditch. No buildings were observed in the interior. In the 6th century, the hill was refortified, this time with thick mortared wal s and the entrance reinforced with a tower. Leaning on the wal s from the interior was a three-aisled basilica with a baptistery. The defended area is too small for a permanent settlement, suggesting it served as a refuge with a church to an unprotected settlement in the vicinity. The fort at Odartsi ( Fig. 3.245) lies on a mountain ridge natural y protected from three sides (Torbatov 2002b; Kirilov 2007, 337–338; Dintchev 2007, 500). Cov- b ering a surface of 1.1 ha, it was inhabited in several peri- ods, with three phases attributable to Late Antiquity. In the first phase, 2–2.9 m thick wal s were built presumably in the second quarter of the 4th century. It was renovated on several occasions and had five towers in the part of easiest access, while a large tower terminated the fort on the south side. Traces of a proteichisma were found in front of the wal s. The small finds and the fort layout suggest this was a military post tasked with controlling the surrounding communications. This phase terminated in a thick layer of Fig. 3.244: Gradishteto near Debrene. Plan of the fortified hil top site from the second half of the 4th (a) and 6th centu- ries (b) (Dintchev 2006, Figs. 92, 93). 288 3. SETTLEMENT IN THE COUNTRYSIDE burnt debris, indicating destruction in the late 4th or the initial decades of the 5th century, presumably during the Hun incursions. The next phase followed soon after and the fort changed its nature to become a civilian settlement. Old buildings were repaired and there were also minor repairs on the wal s. Major renovation took place in the third phase, in the first half of the 6th century, when part of the earlier buildings was reused and numerous new ones built that were adapted to the earlier layout. The wal s and towers were also repaired. The newly fortified settlement persisted to the 600s or early 610s, when it was completely destroyed and abandoned. The comprehensively investigated and excellently preserved fort at Markova Mehana (Procopius’ Stenos?) ( Fisg. 3.246−3.249, 3.314) offers a remarkable insight into the appearance of Early Byzantine military posts (Mitova Džonova 1998; Dintchev 2007, 524–525). It was located on the border between the prefectures of Illyricum and Oriens, on the Succi Pass that became strategical y important in Antiquity as it hosted a major land route connecting East and West. The fort lies on the summit and slopes of a low elevation. It was only 70 × 36 m large, protected with two pentagonal and one triangular tower. Its interior revealed an open gallery along the whole perimeter of the 2.2 m thick wal s and, above the gallery, accommodations for the troops and their commander. Numerous dolia show that the southern part of the gal- lery was used as a granary. The wal s were constructed in the opus mixtum technique with five-line bands of brickwork. The centre of the fort was empty, surprisingly without a church. The deep niches in the rooms in the upper storeys of the pentagonal tower indicate the rooms were used as a praetorium and quaestorium. The fort was Fig. 3.245: Odartsi. Plan of the fort (Dintchev 2007, Fig. 23). already constructed under Anastasius, abandoned after Fig. 3.246: Markova Mehana. Plan of the fort (Mitova-Džonova 1998, Fig. 2). 289 Slavko CIGLENEČKI Fig. 3.247: Markova Mehana. Well-protected entrance (2012). Fig. 3.248: Markova Me- hana. Pentagonal towers and a saddle bellow the fort (2012). Fig. 3.249: Markova Mehana. Interior of the fort with re- mains of the buildings along the wall (2012). 290 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.250: Elenska Basilica near Pirdop. Plan of the fortified site (Dintchev 2007, sl 50). a presumed earthquake and renovated under Justinian; the two phases are separated by a layer of burnt debris. It is possible that some smaller fortified complexes represented early fortified monasteries. Dintchev (2007, Fig. 3.251: Krasen kale. Plan of the fortification (Grigorov 530–532) sees such examples at Orlandovtsi, Lozenets 2011a, Fig. 4). and Shkorpilovtsi, but a corroboration is hindered by the fact that these sites are only known from early excavations. A specific example is the site at Elenska Basilica near Pirdop ( Fig. 3.250) fortified with thick wal s and four towers (Dintchev 2007, 532–533; Borisov 2010a, 203). The defended area is small (0.16 ha) and revealed no other buildings with the exception of a church. This suggests the fort was a refuge for the inhabitants of an unfortified settlement in the vicinity. The largely systematical y investigated fortification at Krasen kale ( Fig. 3.251) forms a revealing strategic-residential unit together with a contemporary settlement in proximity (Grigorov 2011a). It was constructed on a steep hill and has two concentric wal s, with the outer wal s enclosing an area of 205 × 100 m and the inner wal s a 112 × 62 m large area. The distance between them narrows to 7 m on the most endangered north side. The outer wal s were interpreted as a proteichisma and include a pentagonal tower and a ditch in the area of easiest access. The outer wal s are 2.2 m thick in the north and 1.2–1.4 m elsewhere. The north part of the area between the two wal s also holds a two-storeyed prestigious building with several rooms, dated to the 6th century. The fortification interior revealed a cistern and simple houses with one, two or three rooms. A church from the 4th/5th century stands on the plain be- low the hil . With the pentagonal tower attributable to Fig. 3.252: Gradishte near Gabrovo. Plan of the fortified settlement (from Dintchev 2007, Fig. 8). 291 Slavko CIGLENEČKI the Justinian period, it is believed that the fortification buildings of different floor plans and without upper was constructed in the late 5th or the first half of the storeys. The houses were irregular in plan and had 6th century. Its end came in the final decades of the 6th two or three rooms, with areas for the livestock added century. The small-scale archaeological investigations at in the rear. The settlement was additional y fortified the foot of the hill detected a large settlement from the with a thick proteichisma and towers. A large Early 4th to the 6th century, which would certainly also shed Christian church was built in the interior and another light on the fortification above if further investigated. church, smaller and of a cemeterial function, outside An earthwork rampart was built on the natural y the wal s. The quality of construction and the number well-protected spot on top of the high hill of Gra- of buildings in this 2.4 ha large settlement indicate its dishte near Gabrovo ( Figs. 3.252−3.255) in the early central role within a wider area. 4th century and timber buildings erected in its interior The Early Byzantine fort at Kaleto (Castra Rubra) (Koycheva 1992; Dintchev 1997, 51; Koycheva 2002). near Izvorovo ( Fig. 3.256) lies on the Via Diagonalis that In the second half of the 4th century, the summit was traversed Thrace and connected the East and West of the protected with thick wal s and two-storeyed buildings Empire. The fort was well-protected with its location were built in the interior. The intensity of habitation on a slightly raised plateau enclosed on three sides by further increased in the 5th and 6th centuries, with new a river. It is elongated pentagonal in plan and has four Fig. 3.253: Gradishte near Gabrovo. Remains of the fortified settlement from the north (2012). Fig. 3.254: Gradishte near Gabrovo. Thick wal s on the eastern side of the fortified settlement (2012). 292 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.255: Gradishte near Gabrovo. Early Christian church in the interior (2012). Fig. 3.256: Kaleto (Castra Rubra). Plan of the fort (Borisov Fig. 3.257: Dyadovo. Plan of the fort (Borisov 2010b, Fig. 32). 2010a). rectangular towers with an entrance in one of the large the fort. Another contemporary settlement came to light towers. Its construction is dated to the first half of the a further kilometre away. 6th century (Borisov 2010a). The fortification at Carassura on the Via Diagonalis The tell of Dyadovo ( Fig. 3.257) holds an Early ( Fig. 3.258), which is marked as a mutatio in the itinerar-Byzantine fort of a sub-rectangular plan with thick ies, has an unusual layout (Dintchev 2000, 74–75; id. 2007, wal s with four pentagonal towers in the corners and a 494). Its wal s enclosed a 3.2 ha large area spanning two reinforced entrance (Borisov 2010b). The interior only small elevations and the flatland with a river between revealed small barracks for the troops. The fort presum- them. The wal s were probably constructed in the second ably accommodated a small garrison that controlled the half of the 4th century. The buildings in the interior have nearby road and also offered shelter for the inhabitants been attributed to two phases, the first one from the late of the unfortified settlement in the vicinity. Only 400 m 4th and early 5th century, the second one from the 6th away, investigations revealed a civilian settlements with and early 7th centuries. The buildings were largely made a large building (probably church) in the centre that of stone and mudbrick. One church came to light in the yielded small finds contemporary with those found in northern part of the fortification and another, larger 293 Slavko CIGLENEČKI Fig. 3.258: Carassura. Plan of the fortification (Dintchev 2007, Fig. 18). and more prestigious one was found extra muros. The construction of the latter has been dated to the 4th and renovation to the 6th century. The complex was a large central settlement functioning as a local administrative, possibly also economic centre. The remains of an unusual fort were investigated at Harmana ( Fig. 3.259) that formed part of the barrier system in the eastern Stara planina mountains or the Gates of Haemus (Dintchev 2012). The remains show an octagonal plan with a 1.2 ha large defended surface. The fortification wal s were associated in the north and south by contemporary barrier wal s. The passage through the fort was guarded on either side by a large pentagonal tower, while round protruding towers stood in all other corners. The 2.65 m thick wal s were constructed in the opus mixtum technique. A large drystone edifice occu- pied the centre of the fort. The modest cultural layer and the fortification features date the beginning of the fort to the late 4th or more likely the early 5th century. The Fig. 3.259: Harmana. Plan of the fort (Dintchev 2006, Fig. 82a). observed layer of burnt debris has been ascribed to the wars with the Ostrogoths in the 470s or 480s. After this time, minor rebuilding took place in the towers, which included narrowing the entrance, but the work was of a lesser workmanship compared with the original. The second phase presumably lasted from the late 5th to the late 6th or early 7th century. 294 3. SETTLEMENT IN THE COUNTRYSIDE Albania substantial, 1.2–2.8 m thick wal s and a large U-shaped tower. Several houses were excavated in the 1 ha large Research in Albania revealed a multitude of forti- interior, while habitation traces also came to light outside fied sites, but only rare ones have been more extensively the walled area. An Early Christian church was built in investigated. A 0.54 ha large fort that witnessed some the interior in the 6th century. A substantial cultural investigations lies on the high hill of Gradishta of Bard- layer from the 4th–6th century shows that the settlement hoc ( Fig. 3.260) (Përzhita, Hoxha 2003, 58–64, 142–144). served as an administrative and religious centre for a An intriguing feature is mentioned in connection with wider area. its defensive wal s, namely that both longer 2.6 m thick The small fort at Bushati ( Fig. 3.262) was built on sides are built in the drystone technique, whereas the a strategical y significant spot overlooking a major Ro- equal y thick short sides are mortared. The west short man road (Përzhita, Hoxha 2003, 37–42, 135–137). Its side was reinforced with a large rectangular tower. mortared wal s enclose a 0.7 ha large interior and are Several quality masonry buildings came to light near reinforced with as many as twelve towers of different the wal s that presumably accommodated the troops protecting the road from Lissus to Naissus. The small finds date the fort to the 6th century. Enjoying an important strategic location in the valleys of the Drini and Bardhë is the fortified settle- ment at Pecës ( Fig. 3.261) (Përzhita, Hoxha 2003, 50–58, 139–141). Systematic investigations showed that this site, hosting a fortified settlement already in prehistory, was again extensively fortified in the late 3rd or early 4th century. The river provides natural protection from three sides, while the fourth one was defended with Fig. 3.260: Gradishta of Bardhoc. Plan of the fort (Përzhita Fig. 3.261: Pecës. Plan of the fortified settlement (Përzhita, 1993, Pl. 1). Hoxha 2003, Fig. 21). 295 Slavko CIGLENEČKI shapes indicating construction in the early 4th century. Via Egnatia ranks among the large settlements (Cerova The entrance was flanked with a pair of U-shaped tow- 1987; Hoxha 2001; Ćurčić 2010, 183). Extensive investi- ers. The architecture shows a characteristic army fort gations were conducted on this hil , rising high above the from the early 4th century that could host one cohort. valley of the River Devoli and next to a road that crosses A smaller fort lies near the vil age of Domaj ( Fig. the knee of the river and the mountainous terrain. The 3.263) (Përzhita, Hoxha 2003, 43–49, 137–139). Wal s rocky plateau hosting the settlement is protected with of a high-quality and 1.1−2.1 m of thickness enclose a precipitous slopes on two sides and is heavily fortified 0.11 ha large interior. They were shaped as the prow of a on gentler slopes. The wal s survive high and are bound ship in the part of easier access, with a rectangular tower by strong mortar. They are 220 m long and 2.3–2.5 m protecting the only entrance. The cultural layer from thick, reinforced with three triangular, one U-shaped the 6th century and the characteristic shape of the wal s and one rectangular tower, the last one guarding the indicate a small Byzantine fort with a garrison control- entrance. The 0.8 ha large defended area was addition- ling the nearby road and signalling. The fort may also al y protected in areas of easier access with a 1.4 m have served as a refuge for the local population. thick proteichisma that lined the main wal s at a dis- The exceptional y well-preserved fortified settle- tance of 10–20 m. The interior shows traces of thirteen ment at Qafa ( Fig. 3.264) south of Elbasan and on the simple single-room buildings, only two of which were Fig. 3.262: Bushati. Plan of the fort (Përzhita, Hoxha 2003, Fig. 3.263: Domaj. Plan of the fort (Përzhita, Hoxha 2003, Fig. 10). Fig. 15). 296 3. SETTLEMENT IN THE COUNTRYSIDE partitioned. A surprising observation is that the wal s The site of Paleokastra near Gjirokastra ( Fig. of these buildings were clay-bonded, while their roofs 3.265−3.267, 3.310) holds one of the rare lowland forts were covered with brick or slate tiles. Small finds date (Baçe 1981; Bowden 2003, 180–181). It was built as a the cultural layer to the middle or second half of the 6th Roman cohort fort in the early 4th century. It is the shape century. The defensive architecture and the small finds of an isosceles trapezium in plan and covers a c. 0.915 ha indicate a double, i.e. military and agricultural nature of large surface, using its location above the confluence of the fortification presumably controlling the local popu- the Rivers Drinos in Kardhiq for defensive purposes. lation and the mountain pass. It shows a combination of Fortifications include fourteen towers and three rein- strong and well-constructed defensive instal ations with forced entrances. The up to 2.3 m thick wal s are built an irregular layout of buildings in the interior that point in the opus mixtum technique. The interior revealed to the presence of local inhabitants. Ćurčić even saw the traces of army barracks built in the opus spicatum tech- fortification as a small Late Antique urban settlement. nique. The fort was already abandoned in the late 4th century, but civilian population reused the remains in the 5th and 6th centuries. In the interior, they erected a three-aisled church with a narthex and an extension. A smaller single-aisled church came to light extra muros. Coins and other metal finds date the 4th-century phase, while the phase from the 5th and 6th centuries mainly yielded coarseware, fineware and amphorae. The small finds and church architecture are evidence of the fort’s transformation from a Roman military post to a fortified civilian settlement of Late Antiquity. Fig. 3.264: Qafa. Plan of the fortified settlement (Cerova 1987, Fig. 3.265: Paleokastra near Gjirokastra. Plan of the fort Fig. 3). (Bowden 2003, Fig. 7.18). 297 Slavko CIGLENEČKI Fig. 3.266: Paleokastra near Gjirokastra. Remains of the fort with a tower in the foreground (2009). Fig. 3.267: Paleokastra near Gjirokastra. Wal s of the lowland fort (2009). The fortification at Çuka e Ajtojt ( Fig. 3.268) un- North Macedonia derwent a slightly different transformation in Late An- tiquity (Bowden 2000, 21–22; id. 2003, 180). The steep The present territory of the Republic of North hill that oversaw the communications between northern Macedonia is most varied and rich in terms of settle- and southern Epirus witnessed a drastic decrease of the ment remains, but also most extensively researched. earlier, Hellenistic wal s so that they only enclosed a The numerous field surveys, trial trenching campaigns 1.3 ha large surface on the summit and part of the slope. and individual systematic excavations of fortified Late The rebuilt wal s were reinforced with three towers. Antique sites have provided us with a wealth of data that Small finds date the fortification to the 4th–6th century. greatly contribute to our knowledge of the contemporary settlement pattern. 298 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.268: Çuka e Ajtojt. Plan of the fortification (Bowden 2003, Fig. 7.15). Fig. 3.269: Gradište Stenče. Plan of the fortified settlement (Kostadinovski, Cvetanovski 2008, Fig. on p. 189). An important settlement with a major ecclesiasti- acropolis is a large cistern and smaller ones at the wal s. cal centre of Late Antiquity was established at Gradište A considerable part of the settlement remains uninves- Stenče ( Fig. 3.269), on a natural y well-protected hil top tigated, while the southern slopes revealed the remains (Kostadinovski, Cvetanovski 2008). The c. 150 × 70 m of a potential unfortified settlement. large settlement is enclosed with 1.9 m thick walls. The ridge of Gradište (Sobri) near Oraše ( Fig. Investigations thus far focused on the highest part of 3.270) holds traces of a complex fortification (Mikulčić the elevation, i.e. the acropolis that revealed an Early 2002, 463–466). It comprises the acropolis, separated Christian complex with two churches, a baptistery and from the rest of the ridge with two ditches, a residential as many as three piscinae. This complex is separated area of a rectangular plan, with buildings visible in from the rest of the settlement; Kostadinovski and outlines, and a large walled area below the acropolis Cvetanovski posit an episcopal centre here. Next to the that reached to the foot of the hil . The site revealed a 299 Slavko CIGLENEČKI Fig. 3.270: Gradište (Sobri) near Oraše. Plan of the fortified settlement (Mikulčić 2002, Fig. 382). myriad of defensive elements ranging from substantial, ment and a partial y excavated basilica, as well as traces 1.9 m thick wal s, towers, ditches and proteichismas, of ancient mining. with the recovered small finds supporting its dating The site Markovo Kale near Malčište ( Fig. 3.271) is to the Early Byzantine period. In proximity, there are large and well-fortified (Mikulčić 1982a, 92–95; id. 2002, traces of a 3 ha large unfortified contemporary settle- 173–174). The substantial and up to 2.2 m thick wal s 300 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.272: Pelenica near Dračevo. Plan of the fortified vil a (Mikulčić 2002, Fig. 55). Fig. 3.271: Markovo Kale near Malčište. Plan of the fortification (Mikulčić 2002, Fig. 64). protect a 160 × 80 m large area divided into the sepa- rately fortified acropolis and the lower walled area. The wal s are buttressed. The substantial fortifications of the acropolis in the south include several wal s (proteichis- ma?) added on different occasions, as well as four large towers, of which two are pentagonal, one triangular and one rectangular. The lower area hosts evenly arranged and equal y-sized houses that are particularly numer- ous along the northwestern wal s. The large building on the acropolis has been interpreted as an Early Christian church. The fortification has been dated from the 4th to Fig. 3.273: Pelenica near Dračevo from the south (2004). the 6th century, during which time its defences witnessed several renovations. It is believed to have guarded the road leading from the Skopje Basin towards the gold while the traces of a three-aisled basilica indicate a long mines further south. Found in the vicinity were traces use of the walled area. of a settlement and Late Antique burials, on the slope Standing out among the numerous fortifications below the hill also an Early Christian basilica. is that at Markovi Kuli on Mount Vodno ( Fig. 3.274, The complex of buildings at Pelenica near Dračevo 3.275), a site located high above the valley and marked ( Figs. 3.272, 3.273) has been interpreted in literature as in several publications as a ‘late city’ (Mikulčić 1982a, a Late Antique fortified vil a (Mikulćić 2002, 164–165). 48–53; id. 2002, 190–195). The exceptionally large Trial trenching of the complex revealed it was walled, but fortified settlement with 2.4 m thick wal s enclosing also largely protected with its location at the edge of a an area of 400 × 90 m boasts some thirty pentagonal, slope. Although protected both artificial y and natural y, triangular and rectangular towers. It encompasses its defences appear less convincing as in most other sites three separately fortified parts. The partly investigated discussed here. The array of buildings (rotunda with acropolis revealed two large cisterns and a large house, mosaic floors, traces of complex architecture) supports while the majority of housing is presumably located the hypothesis of a late countryside vil a, which the smal in the lower part. The third walled area is in the east, finds and the mosaic date to the 4th and 5th centuries, where rescue investigations established several rec- 301 Slavko CIGLENEČKI Fig. 3.274: Markovi Kuli on Mount Vodno. Plan of the fortification (from Mikulčić 2002, Fig. 85). Fig. 3.275: Markovi Kuli on Mount Vodno from the south-west (2004). tangular towers and two construction phases, one throughout the 6th century, albeit in a reduced extent, Early Byzantine and the other medieval. No significant thus disproving the above-mentioned supposition. It habitation traces came to light in the last part (Ristov is unusual that neither the acropolis nor the two lower 2018). Mikulčić set the first phase of the site as a whole walled areas revealed any traces of an Early Christian to the time of Justinian, the second to the last third of church. The large ( c. 6 ha) and heavily fortified site the 6th and the third to the early 7th century. The large certainly exhibits a primarily military character and bastion on the most easily accessible west side revealed one that could, at least in the two lower areas, welcome several construction phases and confirmed that the refugees from the surrounding area in times of danger. Late Antique fort was already established prior to the The site cannot be tied to any settlements mentioned mid-6th century and abandoned in the late 6th or early in the ancient texts. 7th century (Talevski 2018). The settlement is frequently The small fort at Gradište near Pakoševo ( Fig. referred to as a ‘city’, to which the inhabitants of Scupi 3.276) is located on top of a high crag above the River supposedly moved after the devastating earthquake of Vardar (Mikulčić 2002, 178–180). It is a 32 × 20 m large 518 (Mikulčić 1982a, 48 (143); Saradi 2006, 467). The rectangular structure with a triangular projection on one recent investigations in Scupi, however, revealed irrefu- side. Some habitation traces are visible inside the fort, table evidence of the city being continual y inhabited while better-preserved remains of roughly eight houses 302 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.277: Kula near Čelopek. Plan of the fortress (Mikulčić 2002, Fig. 12). Byzantine period, while the remains of a contemporary Fig. 3.276: Gradište near Pakoševo. Plan of the fort (Mikulčić settlement were unearthed in the plain nearby. 2002, Fig. 70). Visible from afar, the prominent basalt plateau of Žegligovski kamen near Mlado Nagoričane ( Figs. can be observed on the terraces just below it. Small finds 3.279, 3.280) holds the remains of another typical Late date the fort from the 4th to the 6th century. Antique settlement (Mikulčić 2002, 133–137). The 150 × The fortress at Kula near Čelopek ( Fig. 3.277), 90 m large fortification is protected with precipitous north of Kumanovo, has only been investigated with slopes on all sides and additional y fortified with wal s. field surveys (Mikulčić 2002, 124–125). The remains of Excavated in its interior was a three-aisled basilica and wal s clearly visible on the surface show a c. 175 × 53 m a two-part cistern. Investigations also indicated the ex- large fortress with 1.8 m thick wal s and several towers. istence of large houses in the centre of the plateau and The parts of easier access show a drystone proteichisma. small buildings along the edge. Small finds and burials Small buildings are visible in the interior, arranged along show that the fortification was already inhabited in the the interior of the wal s, there is also a larger building 4th, though more intensely in the 6th century. An Early with six rooms. These are elements characteristic of Christian hypogeum came to light on the slope below. an Early Byzantine fortress without a church, but with The prominent location and the vicinity of major roads large headquarters and accommodations for the troops. show the fortified settlement must also have accom- A very clear division between military and civil- modated at least a small army unit. ian parts can be observed at Kula near Kalauzlija The fortification at Gradište near Pčinja ( Fig. ( Fig. 3.278) (Mikulčić 2002, 390–391). This mountain 3.281), which controlled a major communication, was ridge holds two fortification, each located on its own built in a bend of the River Pčinja, in the narrowest part peak separated by a natural depression. The west part of the gorge (Georgiev 1989; Mikulčić 2002, 139–142). hosts a triangular, 160 × 120 m large fortress with 1.8 m It shows characteristic fortification features. It only has thick wal s and three towers, of which the one in the a thin wall on the three sides protected by the river, area of easiest access is pentagonal and the other two while the east side of easiest access was guarded with rectangular. The interior of this fortress has a large thick wal s additional y reinforced on two occasions. building next to the entrance, while habitation remains The length of the east wal s was lined with a wide ditch, are more prominent in the lower part of the site. The while the access across the saddle was additionally fortified area on the opposite peak measures 145 × protected with two shorter ditches. In the Late Roman 90 m and is protected merely with a 2 m thick drystone period, the wal s on the saddle were reinforced with a wal . Excavated in its interior was an unusual church rectangular tower, while two large pentagonal towers of a square plane. Both fortification date to the Early were built there in the mid-6th century. Towards the end 303 Slavko CIGLENEČKI Fig. 3.278: Kula near Kalauzlija. Plan of the hil top site (from Mikulčić 2002, Fig. 296). Fig. 3.280: Žegligovski kamen near Mlado Nagoričane from the west (1984). Fig. 3.279: Žegligovski kamen near Mlado Nagoričane. Plan of the fortified settlement (Mikulčić 2002, Fig. 27). Fig. 3.281: Gradište near Pčinja. Plan of the fortified settlement (Mikulčić 2002, Fig. 31). 304 3. SETTLEMENT IN THE COUNTRYSIDE of the 6th century, a new rectangular tower was built on the spot of the earlier one. The fortified area measured c. 3 ha. The acropolis revealed the remains of an Early Christian church, while the slopes below hold housing remains. Buildings were also found on the unprotected slope east of the fortification. Kalata near Kamenica ( Fig. 3.282) is a large ag- glomeration with the features of a small city (Mikulčić 1975, 122–131; id. 2002, 211–213; Lilčić 2013, 1023). It was constructed on a terrace above the confluence of the Rivers Kamenica and Bregalnica. As an economic and ecclesiastical centre on a major crossroads, Mikulčić belives it became a city in the 6th century. The c. 2 ha large acropolis holds the remains of multi-phase wal s. In the Justinian period, thick wal s were built on top of drystone foundations and a drystone wal (proteichisma) added along the exterior. The terraces in the interior hold the remains of differently-sized building that presum- ably include a basilica. In the southeastern part of the settlement below the acropolis, excavations unearthed a basilica with a large atrium from the 6th century and housing remains. The buildings in the interior date the settlement to the 3rd−6th century. A richly decorated three-aisled basilica from the 6th century was found at Begov Dab, a site west of the settlement. Another, cem- eterial church is located northeast of the acropolis. Two other Early Christian churches were identified slightly further away from the settlement. All this reveals an Early Christian centre, but the settlement itself has not been extensively investigated and cannot be reliably Fig. 3.282: Kalata near Kamenica. Plan of the fortified settle-interpreted. ment (Mikulčić 2002, Fig. 102). The trial trenching at Venec near Miokazi ( Fig. 3.283) revealed a fortification from the 6th century (Spasovska-Dimitroska 1989; Mikulčić 2002, 473–474). are evenly spaced small buildings with clay-bonded wal s The c. 1.5 ha large fortress was enclosed with 1.65 m thick and interiors with hearths and pithoi. A three-aisled wal s reinforced on the exterior side with evenly spaced church with a baptistery stood on the top. Publications rectangular towers and in areas of easier access also with also mention traces of larger edifices not identified more a drystone proteichisma. Lining the wal s in the interior precisely. The layout of the fortress with towers and army Fig. 3.283: Venec near Miokazi. Plan of the fortress (Mikulčić 2002, Fig. 391). 305 Slavko CIGLENEČKI barracks systematical y arranged along the wal s point to a strong army post next to a junction of major Roman roads. The size, thick wal s with towers and the location on the Via Egnatia just before its ascent towards the important Barbara Pass show that the fortified site at Kale near Debrešte ( Fig. 3.284) was a significant fortress charged with protecting the above-mentioned road (Hensel, Rauhutowa 1981; Mikulćić 2002, 348–350). It was built in the flatland, but used the surrounding wet- land for protection. A fortress measuring 175 × 157 m was built here in the Late Roman period, which had four reinforced entrances and wal s strengthened with round towers in the corners and square ones elsewhere. In the Early Byzantine period, the north and south entrances were wal ed up, the other two reinforced with pentagonal towers and a propugnaculum. The up to 2.4 m thick wal s with towers were also renovated. A three-aisled basilica with a baptistery was constructed at the east entrance. Fig. 3.284: Kale near Debrešte. Plan of the fortress (Mikulčić Next to it was a large rectangular Late Roman building 2002, Fig. 248). of an unknown function. Most of the interior has not been investigated, though it did reveal traces of other buildings. Also found were traces of a contemporary unfortified settlement in immediate proximity (cf. Mikulčić 2002, 350). The numerous finds show the site was abandoned in the late 6th century. In the southernmost part of North Macedonia, systematic excavations unearthed a small settlement on the island of Golem Grad, Konjsko in Lake Prespa, naturally protected with the waters of the lake that enabled a fairly peaceful existence in Late Antiquity (Bitrakova Grozdanova 2011). The island is 21 ha large and delimited with rocks that rise 40 m above the cur- rent surface of the lake. It was presumed to have hosted a walled refuge (Mikulčić 2002, 379–381), which the recent excavations did not confirm. Only the part of easiest access to the rocky plateau was walled. Excava- tions on the best protected part of the plateau, on the saddle between two elevations, revealed several simple buildings, on the very edge of the plateau also a well- constructed cistern that collected rainwater from the roofs of the adjacent buildings. Building remains also came to light in the western part of the island. The group of Late Antique houses included a small Early Christian church, while another one, slightly larger and in part decorated with mosaic floors came to light further away. Near the latter was a small cemetery from the 6th century. The excavated coins and other small finds, as well as grave goods date the settlement from the 4th to the early 7th century. The natural y protected settlement in the middle of the lake could certainly welcome a large number of refugees fleeing from danger. An interesting example of an Early Byzantine fortress was unearthed at Belgrad near Dvorište ( Fig. 3.285) (Mikulčić 2002, 423–425). The elongated, 170 × 64 m large fortress had up to 1.5 m thick wal s with Fig. 3.285: Belgrad near Dvorište. Plan of the fortress (Mikulčić equal y-sized casemates lining the interior. A special fea-2002, Fig. 332). 306 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.286: Pešna near Dević. Plan of the fort in the cave (Mikulčić 2002, Fig. 192). ture is a rock-hewn 19 × 8 m large cistern below the wal s on the exterior side, which was enclosed with a drystone wall and additional y protected with two towers of the fortress wal s. This cistern was certainly also available to exterior visitors, with the troops overseeing its use. A very special fort is that built in the large hori- zontal cave of Pešna near Devič ( Fig. 3.286) (Mikulčić 2002, 293–295). The cave entrance was guarded with impressive, 2.1−2.5 m thick wal s. The fort measured 105 × 35–50 m. In the interior, a cistern was built next to the wal s. The fort is dated to the 6th century and protected the nearby road. The hill of Gradište near Delisinci ( Fig. 3.287) revealed an example of a reused earlier fortification (Mikulčić 2002, 415–416). As at several other sites it occupied the summit and slopes already fortified in prehistory. The 2.2 m thick Late Hellenistic drystone Fig. 3.287: Gradište near Delisinci. Plan of the fortified hil top wal s were reused without reinforcing or in any way site (Mikulčić 2002, Fig. 325). improving them. The numerous remains of Late Roman pottery show that the site was either used as a refuge on several occasions or was inhabited over a long time. The slopes also yielded burials from the 3rd and 4th centuries. The hill Gradok (Markovi Kuli) near Čanište ( Figs. 3.288−3.290) holds the remains of a vast two- part fortified settlement (Mikulčić 2002, 344–347). It was built on the summit and its southern slope. The surface of 4.5 ha was protected with 1.5−1.7 m thick wal s with several protruding rectangular towers. The summit plateau of the hill has a wall separating it from the lower part; it represented the main habitation area with the entrance reinforced with towers. This acropolis holds the remains of two churches, the larger Fig. 3.288: Gradok (Markovi Kuli) near Čanište. Plan of the fortified settlement (Mikulčić 2002, Fig. 245). 307 Slavko CIGLENEČKI Fig. 3.289: Gradok (Markovi Kuli) near Čanište. Fortified settlement from the west (2004). Fig. 3.290: Gradok (Markovi Kuli) near Čanište. Remains of the west church (2004). of which is a three-aisled basilica in the east, while the west church is functional y associated with a large residential building with three rooms. There are also traces of other buildings on the acropolis, while such remains are not readily identifiable in the lower south area. The small finds and church architecture date the settlement to the 6th century. Its two-part layout shows a more permanently inhabited upper part and a larger lower area that was either in refugial use or served for economic purposes. The site at Kale near Gorno Svilare ( Fig. 3.291) hosts a small (80 × 60 m) fort (Mikulčić 2002, 166–168). It was constructed on a high hill that afforded control of Fig. 3.291: Kale near Gorno Svilare. Plan of the fort (Mikulčić 2002, Fig. 58). 308 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.292: Čebren (Grad) near Zovik Plan of the fortified settlement (from Lilčić Adams 2018, Fig. on p. 72). the communication at the point where the River Vardar Greece exits the Derven Gorge. Its wal s have five protruding towers and densely spaced buildings along the interior. Newly-founded fortifications of Late Antique are Also visible in the interior is a large cistern and a small not well-known in Greece. With the scholarly attention Early Christian church. A proteichisma in the form of focused on the Classical period, such sites were often a thick stonework rampart was constructed in areas overlooked or only briefly mentioned. The situation is of easier access. In the south, the fort was additional y changing gradual y in recent times and research indi- protected with two parallel wal s. The architecture and cates that such sites are fairly numerous, particularly defensive elements reveal an Early Byzantine fort, while in northern Greece, and possibly show a distribution the small finds indicate it may already have been estab- similar to that observed in areas to the north. lished in the Late Roman period. The steep hill of Čebren (Grad) near Zovik ( Fig. At least fourteen hil top sites (refuges?) of a Late 3.292) holds the remains of a c. 8 ha large fortified set-Antique date are mentioned in the area around the plain tlement established in a bend of the River Crna reka at Philippi, which were presumably associated with the (Mikulčić 2002, 287–288). The better-protected highest unfortified settlements in the plain. They may in part part holds the acropolis with a large rectangular ter- be strategic forts manned with soldiers (Dunn 2004, race below. It is defended with up to 2.2 m thick wal s 548–555). An example is the first-phase fort at Drama, reinforced with several towers. The lower part revealed which boasts impressive opus mixtum wal s. One or even a large Early Christian church. The habitation terraces several churches are mentioned in the interior, while and numerous small finds indicate intense habitation. small finds include Late Antique pottery. Mikulčić presumes a fortress here in the 4th−5th cen- A site that appears particularly important is that turies and a settlement of an urban character in the 6th at Ai Giannis, Adriani, where an Early Christian century. The fertile surroundings hold traces of intense basilica was found on the acropolis and beside the ba- Late Antique habitation and mining activities. In the silica densely arranged buildings that also yielded Late 6th century, the settlement can be seen as a large forti- Antique pottery. Given the continuity of the toponym, fied settlement with an economic hinterland and the the site has been linked with the Justinianic phrourion capacity to accommodate a large number of soldiers Adrianon, the successor of the previously sprawling city in its interior. of Hadrianopolis (Dunn 2004, 549). 309 Slavko CIGLENEČKI The fortified settlement at Palaiokastro near Palaiochori, at the foot of Mount Pangaion, revealed the remains of stone houses, a cistern and an Early Christian church, but also Late Antique pottery (Dunn 2004, 550). The small fort at Khortokopi II was enclosed with thin, but double wal s (proteichisma?). It yielded the foundations of stone buildings, a cistern and a smal church (Dunn 2004, 550). We should again mention the fortified church complex at Louloudies near Katerini, which is already discussed among newly-founded cities (see Chapter 2.5). Leaving aside the hypotheses of the episcopal see being transferred to this quadriburgium and limiting the discussion to the basic archaeological evidence, this site shows a familiar form of settlements frequently men- tioned in areas to the north. The 90 × 80 m, i.e. 0.9 ha large quadriburgium held a large Early Christian church, parallel to it a large residential building and smaller auxiliary edifices lining the wal s. Also characteristic is the proximity to a 4 ha large fort from the 6th century, located only 150 m away. In northwestern Greece, fortifications are also poorly known. Most were established on top of the earlier Hellenistic fortified settlements. An example of this can be found at Kastritsa near Ioánnina ( Figs. 3.293−3.295), a high hill at Lake of Ioánnina with re- mains of a multi-period fortified settlement (Blackman 2000, 46, 65–66; Bowden 2000, 40–42; id. 2003, 181). It revealed substantial remains from Late Antiquity. The summit is enclosed with Hellenistic wal s that in places show Late Antique repairs and additions. Excavations unearthed two earlier buildings that were later subdi- vided. One of the buildings revealed numerous coins from the 4th to the late 6th century, but also a hoard of coins that end with a follis of Mauricius Tiberius from 582/583. Another house is marked by numerous pithoi and Late Antique coarseware, which suggests it served as a storehouse. Fig. 3.293: Kastritsa near Ioánnina. Plan of the fortified settlement (Bowden 2003, 7.19). Fig. 3.294: Kastritsa near Ioánnina. Remains of the buildings on the western slope of the settlement (2013). 310 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.295: Kastritsa near Ioánnina. Settlement in the upper part of the site from the south (2013). Not much is known on the large fortification at Kastro Rizovouni ( Fig. 3.296) (Bowden 2000, 107–109; id. 2003, 180). It is located on a dominant steep hil above the Thesprotikon Plain and was inhabited in several periods. The extensive Hellenistic wal s were partial y renovated in Late Antiquity and a construction with reused stone and brick courses added on a top of the earier wal s. Underneath the later church are the remains of a three-aisled Early Christian basilica. The field survey in the interior revealed modest remains of Late Antique pottery, which led to the hypothesis that it only functioned as a refuge. It was presumably used by the people living in a nearby lowland settlement at present-day Rizovouni, where the remains of an Early Christian church and other buildings came to light. The research conducted in the last two decades in Thessaly revealed a mass of unfortified, but also several fortified hil top settlements. The most important and also best researched among these is undoubtedly at Velika, a high hill above a city of the same name and just above the coast (Sdrolia 2016, 127–129). In the Early Byzantine period, earlier defensive elements were Fig. 3.296: Kastro Rizovouni. Plan of the fortified hil top site reused to build a 2.1 ha large fortified settlement. The (Bowden 2003, 7.14). wal s are well-built, 2 m thick and reinforced with tow- ers. Standing out among the buildings in the interior is a three-aisled basilica and a large building with three indicate a major role of the Late Antique fortified set- wings enclosing a court. It is believed to house the office tlement in the Aegean trade network. for supervising produce trading. Found on the slopes outside the wal s was another church complex with an A special type of island settlements or ‘refuge oil press and storerooms. Coins, other small finds and islands’ has been observed along the Greek coast that architecture date the settlement to the Justinian period. were predominately established in the late 6th century The settlement was abandoned after the 7th century. The (Bowden 2003, 186–190; Veikou 2012, 177–188). A par- great quantity of imported goods and transport vessels ticularly characteristic and well-known example is the 311 Slavko CIGLENEČKI settlement on the island of Kephalos in the Ambracian of Greece. Bowden sees them as merely a maritime Gulf, near the city of Vonitsa in Epirus. The previously variation of hil top settlements (Bowden 2003, 189). The uninhabited island measuring 200 × 50 m witnessed archaeological evidence is il uminated by the accounts a hive of construction activity in the 6th century that in the Chronicle of Monemvasia and other historical included two Early Christian basilicas, but also numer- sources that speak of the population from urban centres ous remains of residential buildings. Coins, other small dispersing to the islands (Liebeschuetz 2001, 284–291; finds and mosaics date the settlement throughout the Veikou 2012, 182–188). 6th and part of the 7th century. The limited space avail- able on the island meant that the settlement could only Associated with the trans-isthmian wall of the function in communication with the mainland. Similar Hexamilion, which in Late Antiquity barred access to settlements were also found along the southern coast the Peloponnese, is the large fortress at Isthmia ( Figs. 3.297, 3.298) (Gregory 1993; Kardulias 1993). Its size and eminent strategic location reveal the main fortress protecting the barrier wal . The trapezium-shaped for- tress was the heart of the defensive system and took full advantage of the slightly raised terrain. It covers a sur- face of 2.71 ha and its wal s are 2.3 m thick. Insufficient research offers limited insight into the interior, where only the outlines of several buildings were recognized. More is known of both monumental entrances and the protruding rectangular tower, while the interior only revealed the outlines of several buildings. The fortress and the Hexamilion were constructed in the 410s and renovated under Justinian in the mid-6th century, indi- rectly corroborating the two fortification periods in the wider Balkan region. A small Early Byzantine fortified settlement was created in the plain holding the ruins of ancient Olympia ( Figs. 3.299, 3.300) (Curta 2014, 35; Baldini, Bazzechi 2016, 705–707; Völling et al. 2019). The records of the early excavations from the second half of the 19th cen- tury allow for a partial reconstruction of the settlement believed to have been established in the 5th century on Fig. 3.297: Isthmia. Plan of the fortress (Kardulias 1993, Fig. 1). Fig. 3.298: Isthmia. Well-preserved barrier wall of Hexamilion near the fortress (2007). 312 3. SETTLEMENT IN THE COUNTRYSIDE the ruins of Greek temples, persisting throughout the 6th To further il ustrate the late fortification pattern, at least to the early 7th century. It had substantial wal s, we should mention another one of the many natural y made of reused stone and reinforced with four towers. protected sites that offered shelter for those living in the It revealed two phases, one dating to the 5th century and southeastern Peloponnese, one reported in the Chroni- the other that began in the mid-6th and lasted to the cle of Monemvasia (Cameron 2012, 154; Veikou 2012, early 7th century. The numerous farming tools indicate 184– 188). The mighty plateau of Monemvasia on a a highly agrarian community. The finds from the associ- difficult-to-access peninsula is presented in the chronicle ated cemetery show this was a settlement of Romanised from the 10th/11th centuries as a natural y excellently inhabitants that was particularly densely inhabited in defended location where people sought refuge, fleeing the 6th and early 7th century. from the incursions of the Slavs. Fig. 3.299: Olympia. Plan of the fortified settlement (Baldini, Bazzechi 2016, Fig. 10). Fig. 3.300: Olympia. Remains of the Early Christian church outside the fortified settlement (2007). 313 Slavko CIGLENEČKI 3.4 INTERPRETATION 1974; Mulvin 2002). We should note the pioneering OF THE SETTLEMENT PATTERNS work of Edith Thomas, who gathered vast amounts of IN THE COUNTRYSIDE data and defined the types and functions of the vil as (Thomas 1964). A particular feature of the vil as in Hun- The overview of the different forms of countryside gary is the fortification of the most important residential settlement has revealed two major categories of sites. The buildings, though not to the degree of forming fortified first predominantly consists of unfortified settlements settlements proper. rooted in the tradition of the earlier Roman settlement The research of the vil as in Istria has revealed clear with vil as, farmsteads and vil ages rational y distributed signs of continuity. An example is the vil a at Dragonera across fertile lowland areas. Settlements of this category Jug , where excavations established a Late Antique phase are quite rare in the mountainous hinterland, where with high-quality renovation of the residential part and their subsistence was tied to specific economic activities. a continuation of the vil a to the early 7th century (Sta- Settlements of the second category are a novum rac, A. 2010). A continuous life has also been evidenced and comprise fortified sites of diverse forms that paint for the vil as on several islands in the Adriatic (e.g. a completely different picture with their great numbers Muline on Ugljan, Novo selo Bunje on Brač). and partly with locations in natural y well-protected Vil as in Bulgaria ceased to be inhabited after the mountainous or even high-altitude areas. This category invasion of the Huns in the 440s. Simple dwellings were also incorporates unfortified settlements, but their re- erected in their ruins that show a transformation similar mains are modest and poorly-investigated, and thus do to those observed in the western part of the Empire not allow for reliable conclusions as to their appearance (Dintchev 1997). and duration. Less is known of other forms of settlements such as vici and roadside stations, and it is not possible to draw conclusions regarding their continuity. The recent large- 3.4.1 SETTLEMENTS WITH CONTINUITY scale rescue investigations in Austria, Slovenia, Croatia, Serbia and Bulgaria have confirmed their existence only The overview above briefly presented a selection of in the 4th century. vil as and other sites that represent a continuation of the The Dalmatian hinterland has revealed several ex- habitations established in the first three centuries AD amples of small Roman settlements that witnessed mul- ( Figs. 3.1, 3.26). Some of these were already abandoned tiple renovations and persisted to the end of Antiquity; in the last third of the 3rd century, though for most the an example is the group of buildings at Bisko near Knin process of abandonment unfolded in the course of the (Milošević, A. 1986). Other such examples associated 4th century, few persisted to the first half of the 5th, while with contemporary Early Christian churches came to only some vil as around Lake Garda, along the Adriatic light along the coast (e.g. Muline on Ugljan, Lovrečina coast and in Greece survived into the 6th century, albeit on Brač, Proložac Donji, Ubli on Lastovo; Fig. 3.301). in a heavily altered or reduced form. An interesting example of an unfortified settlement Numerous investigations have been conducted in have been discovered at Cherven breg in Bulgaria, in Italy, particularly its northern part, that provide a wealth which an Early Christian church was erected in the of data on the fate of the vil as in Late Antiquity and mid-4th century and the settlement persisted as such to beyond (cf. Brogiolo, Chavarría Arnau 2014, 233). They the late 5th century (Petrova 2019). reveal a rather standardised evolution and end of vil as Several unfortified settlements in northern Greece across the Empire, but also that their transformation al- presumably also continued to the 6th century, but evi- ready began in the 3rd century. Initial y, the most apparent dence on them is scarce (Dunn 2004, 539–542). change was in the function of rooms, which became pro- duction units. From the 5th century onwards, life in these vil as continued on a modest scale and included minor 3.4.2 NEWLY-FOUNDED artisanal activities; it is not clear whether they were stil UNFORTIFIED SETTLEMENTS inhabited by Roman aristocracy. Brogiolo dates the final phase of the vil as around Lake Garda from the late 5th The unfortified settlements include many that were to the late 6th century and notes a degradation of earlier built soon after the crisis in the 3rd century and display buildings, subdivisioning with primitive constructions a continuation of the former settlement patterns ( Figs. and burials in the vil a interiors (Brogiolo 1997, 300). 3.33, 3.49). These vil as are fewer in number, but many Much research has been dedicated to the vil as show a more luxurious interior; this is a development in Hungary, which results in a good understanding of similar to that observed in the vil as with continuity. the different aspects of their existence, but even more In addition, a number of modest settlements were also importantly in a reliable typology and chronology of established in the lowland, mainly composed of timber construction and abandonment (Thomas 1964; Bíró buildings, that can be seen as late vil ages (primarily 314 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.301: Ubli on the island of Lastovo. Remains of the unfortified settlement with a church near the harbor (2017). from the 4th century). The majority of these end in the importance as they point the direction in which the miss- late 4th century and only rare ones persist into the 5th ing links of Late Antique settlement should be sought, century. Some very late vil as established in remote areas particularly for the poorly-known 5th century. They also (Ostrovica, Mirje, Rankovići near Travnik, Kruče near represent the settlements from which the inhabitants fled Ulcinj, Donje Nerodimlje) were either abandoned or to seek shelter in nearby refuges (see below). destroyed presumably as late as the 6th century. More at- tention should be paid to these intriguing settlements, as they are only partial y investigated and briefly published. 3.4.3 NEWLY-FOUNDED In addition to the strong continuity of vil as, Istria FORTIFICATIONS and the Adriatic coast boast several other forms of set- tlements that appeared in the Late Roman period and The phenomenon of fortified (hil top) settlement lasted to the end of Antiquity. Examples of these include forms of Late Antiquity ( Figs. 3.54, 3.201) is – as already the unfortified settlement at Rim near Roč in Istria, a outlined above – not well-known to the broader profes- small rural complex of simple buildings at the church of sional public and should therefore be more extensively St Chrysogonus near Milohnići, on the island of Krk, and discussed. This is in part due to the fact that these sites the complex of buildings at Vučipolje near Dugopolje were frequently seen as a local phenomenon and subject (Marušić 1986; Janeš 2015; Borzić, Jadrić 2007). to different interpretations and names within different A similar continuity can be established for the smal regional archaeological practices. The discussion below vil age with a church in North Macedonia, at Trpčeva will therefore open with a short outline of the history of Crkva near Dunja (Kepeska, Kepeski 2006). research into these sites and continue with the difficulties The intensive surveys in Greece show a countryside regarding their interpretation and the understanding of dotted with dispersed unfortified buildings or settle- their role within the broader settlement pattern. ments (Bintliff 2012, 352–359). This settlement pattern is primarily the result of the geographic location deep in the south that made it less exposed to barbarian incur- HISTORY OF RESEARCH sions, but also meant greater proximity to the capital of the Empire. Evidence on the forms of settlements To il uminate the gradual y developing awareness mainly comes only from intensive surveys, which cannot of the changes in the settlement pattern, we should provide details on the nature, size and duration of a site. begin the discussion with a brief presentation of the Most unfortified countryside settlements lie in re- main centres of research. The fortified hil top settle- mote areas, often also at higher altitudes, they are modest ments of Late Antiquity in the eastern Alps were first and unorganised. This, however, does not diminish their identified as an important settlement phenomenon 315 Slavko CIGLENEČKI Georgenberg Martinsbühel Ampass Burgkofel Lavant Burgbichl Kappele Sabiona Pittersberg Tesimo Cuel Budin San Pietro Invillino Appiano Castelfeder Gemona Osoppo Artegna Castelraimondo Nimis Doss Trento Attimis Ragogna Castelvint Udine San Martino di Lundo/Lomaso Saint'Andrea di Loppio Rocca di Garda Sirmione 0 50 km Fig. 3.302: Newly-founded fortifications 316 3. SETTLEMENT IN THE COUNTRYSIDE Pürgg Knallwand Gröbming Kugelstein Heiliger Berg Riegersburg Teurnia Grazerkogel Lug ins Land Frauenberg Lamprechtskogel Duel Ulrichsberg Tscheltschnigkogel Kathreinkogel Hoischhügel Hemmaberg Ajdna Ančnikovo gradišče Gračnica Brinjeva gora Bašelj Sv. Radegunda Sv. Jakob Donačka gora Tonovcov grad Rifnik Kranj Krvavica Zbelovska gora Vipota Puštal Sveta gora v Zasavju Tinje Mali Njivč Prapretno Lobor Polhov gradec Sv. Lambert Vranje Svete gore Lonja Sv. Katarina Veliki vrh Vrtovin Ad Pirum Velika Strmica Korinjski hrib Kuzelin Doberdo Limberk Velike Malence Planina Križna gora Zidani gaber Ulaka Kučar Rodik Šilentabor Črnomelj Ozalj Pasjak Crkvišće Bukovlje 0 50 km in the south-eastern Alps. 317 Slavko CIGLENEČKI Fig. 3.303: Ulrichsberg. Plateau of the settlement from the east (2005). Fig. 3.304: Ajdovski gradec above Vranje. The first sistematical y examined church complex in the fortified settlement (2023). in Austria in the early 20th century, when Rudolf Eg- of a systematic detection of fortified Late Antique sites ger investigated Early Christian churches on hil s and in the area under discussion and wider, and had a great observed they frequently lie in natural y and artificial y impact on their subsequent interpretation and research. well-protected areas (Egger 1916). The military for- Later research in Austria focused primarily on tification at Hoischügel was also investigated at that the systematic investigations at two major fortified time, which was followed between both world wars by hil top sites, namely Hemmaberg and Teurnia, while systematic excavations at Duel (Egger 1929). Using the minor investigations were also conducted elsewhere research results from several other fortified settlements (overview in Glaser 1997, 96–120, 131–140; Bauer (Ulrichsberg ( Fig. 3.303), Teurnia), Egger published the 1997; Ladstätter 2000; Glaser 2008; Winckler 2012; first historical-archaeological outline of the Late Antique Steinklauber 2013). settlement, entitled Die Spätantike in Ostalpen, as early as Playing an important role in the research of 1942 (Egger 1942). The book treated Late Antiquity as a Late Antique fortifications was the Kommission zur distinct chronological unit and also identified an altered archäologischen Erforschung des spätrömischen Raetien settlement pattern. Egger’s research marks the beginning in Munich. It was headed by Joachim Werner, who saw 318 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.305: Gradina on the island of Žirje. Large residential building along the wal s of the Early Byzantine fort (2004). the significance of such sites for understanding the 1960s, when they began with the excavations of the forts transition period between Antiquity and the Middle forming part of the Claustra Alpium Iuliarum barrier Ages. Alongside his investigations of the Late Roman system (Šašel, Petru 1971). The col aboration with Ger- forts in Germany (Lorenzberg, Kellmünz, Goldberg), man colleagues in the early 1970s brought large-scale he initiated and enabled systematic investigations in systematic investigations at Hrušica and Ajdovski gradec Italy (Invillino, Sabiona), Austria (Georgenberg near above Vranje (Petru, Ulbert 1975; Ulbert 1981). These Kuchl) and Slovenia (Hrušica, Ajdovski gradec above were joined by numerous other excavation campaigns, Vranje; Fig. 3.304) as early as the 1960s. He organised the for example at Rifnik, Sv. Pavel above Vrtovin, Ajdna, publication of important Bulgarian sites of Golemanovo Kranj, Korinjski hrib, Tinje above Loka pri Žusmu, Kale and Sadovsko Kale, which had been excavated Gradec near Prapretno, Brinjeva gora and Ajdovščina before World War II (Uenze 1992). He recognised the above Rodik (overview in Ciglenečki 1987a). The last great significance of investigating Late Antique sites in three decades also brought monographic publications North Macedonia and also encouraged and enabled of the investigations at Kučar (Dular et al. 1995), Tinje the publication of its Late Antique fortified cities and above Loka pri Žusmu (Ciglenečki 2000), Tonovcov grad fortifications, which is undoubtedly one of the most near Kobarid (Ciglenečki et al. 2011; Modrijan, Milavec important publications for understanding the settlement 2011) and Korinjski hrib (Ciglenečki et al. 2020). patterns in the Balkans (Mikulčić 2002). Investigations in Croatia focused on the forts along During his excavations at Invillino, Volker Bier- the Adriatic coast. The first synthetic publication and brauer published an overview of the fortified hil top assessment of sites (Gunjača 1986) was followed by a sites in South Tyrol and Friuli (Bierbrauer 1985; id. series of field surveys that detected innumerable forts 1987). It is a topic that received much attention in (cf. Tomičić 1996). Their good preservation and the northern Italy over the last decades. In addition to absence of vegetation enabled the first interpretation of the research at Invillino and Sabiona, there were also these predominantly Early Byzantine forts. There were systematic investigations at the vast fortified site at also large-scale excavations of the fort at Gradina on Monte Barro, which is located outside the area under Žirje ( Fig. 3.305) (Pedišić 2001; Karađole, Borzić 2020). discussion, and the first monographic overview of In continental Croatia, significant investigations were the fortified posts in northern Italy (Brogiolo, Geli- conducted at Kuzelin, Marija Gorska near Lobor and chi 1996). Investigations are currently conducted at Crkvišće Bukovlje (Sokol 1994; Filipec 2007; Azinović Sant’Andrea di Loppio (Maurina 2016; ead. 2020) and Bebek, Sekulić 2019). San Martino di Lundo/Lomaso (Zagermann, Cavada In Bosnia and Herzegovina, Đuro Basler published 2014; Cavada, Zagermann 2020b). an insight into the Late Antique settlement pattern In Slovenia, systematic investigations of the Late already in 1972 that included a discussion of the as- Antique fortifications have been conducted since the sociated Early Christian churches (Basler 1972). Many 319 Slavko CIGLENEČKI Fig. 3.306: Caričin grad, pressumed Iustiniana Prima. The location of a fortified settlement on a low, but natural y well-protected hil (2012). of the subsequent investigations were interrupted when Antique fortifications in a wider area (overview in Uenze the war broke out, though their large scale is reflected 1992). The first publications aimed at systematical y in the finds collected and synthetical y published in presenting fortified sites came out in the 1980s (Ovčarov Arheološki leksikon Bosne i Hercegovine (Čović 1988), 1982) and have since been growing in number (Dintchev but also in several later brief overviews (Čremošnik 1997; id. 2007; id. 2021b). Recently, there are also intensive 1990; Špehar 2008). surveys of the countryside and the excavations in the From 1912 onwards, investigations of the Late fort at Dichin, conducted in cooperation with a British Antique fortified sites in Serbia focused on excavating team (Poulter 2007c). The numerous investigations are Caričin grad ( Fig. 3.306), a typical fortified hil top settle-revealing an important share of fortified sites in the local ment and presumed Iustiniana Prima. The construction settlement pattern. of a hydropower plant in Đerdap Gorge in the 1960s The topographic surveys that Ivan Mikulčić headed brought large-scale rescue excavations that unearthed in North Macedonia revealed an exceptional y high important sites including many Roman forts renovated in number of Late Antique posts (Mikulčić 1982a; id. 1986; Late Antiquity (see the reports in Starinar 23−24 (1984)). id. 2002, 87–89). They include 290 reliably identified and Investigations of the fortified countryside sites in Serbia 95 presumed Early Byzantine forts, a large part of which began in the 1970s with the excavations at Gradina Ras, dates to the time of Justinian. Particularly important is which revealed a Late Antique habitation horizon under the distinction between military sites, on the one side, the medieval remains. A vast amount of new topographic and civilian settlements and refuges, on the other, that data has been gathered in recent decades, also several trial could be made for several sites (Mikulčić 2002, 87). The trenching campaigns, but their results are only partly field surveys and excavations conducted in recent times published (overview in Milinković 2008; id. 2015). add to the wealth of knowledge on the Late Antique Bulgaria enjoys a long history of research at fortified settlement (Lilčić 2013). sites in the countryside, particularly if also considering Recent investigations are beginning to shed light on the topographic efforts of brothers Hermann and Karel the Late Antique countryside of Kosovo (Përzhita, Hoxha Škorpil in the early 20th century (cf. Boshnakov 2007). Of 2003; Rama 2020). Field surveys have established over a great importance was the research of a German-Bulgarian hundred fortified Late Antique sites, but they only wit- team in the fortified settlement at Golemanovo Kale; nessed limited further investigations (Rama 2020, 121). together with the adjacent fort at Sadovsko Kale, this site The recent field surveys in Albania indicate a long served as the reference point for the study of Late site density similar to that in neighbouring North 320 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.307: Distribution of Early Byzantine fortified settlements in south-eastern Europe (showing largely just the state of research!) (Milinković 2014, Fig. 4). Macedonia, though the sites have only rarely been on the Clasical period, Late Antiquity has often been investigated further. The first overview of the fortified overlooked or only summarily outlined. More frequent sites was already published in 1976 (Bace 1976), while publications in recent times, however, suggest that forti- later topographic surveys offer additional insight into fied sites were fairly numerous, particularly in northern this segment of the settlement pattern (e.g. Përzhita, Greece, with their density possibly comparable with that Hoxha 2003). in the areas to the north (Dunn 2004; Veikou 2013, 131; Less is known on this segment of the Late Antique Sdrolia 2022). settlement in Greece. With scholarly attention focusing 321 Slavko CIGLENEČKI FUNCTIONAL IDENTIFICATION between castel a (i.e. small forts) and refuge enclosures OF THE NEWLY-FOUNDED FORTIFICATIONS in the Dalmatian hinterland, at least as far the location and size were concerned (Basler 1972, 61). The nature of the different types of Late Antique In Slovenia, the Late Antique settlement pattern fortified sites has frequently been discussed, but often appeared clear in the 1970s: people lived in lowland with contradictory results. Literature uses different settlements to the late 4th century, then largely moved terms for sites of the same form, a problem already to natural y well-protected hil top settlements in remote mentioned in the introduction. It is also a problem that areas and persisted there to the arrival of the Early Slavs hinders an appropriate understanding of the settle- in the late 6th century. Most researchers believed that ment patterns and we should therefore briefly present evidence indicated the existence of fortified hilltop and critical y assess some of the early interpretations of settlements from the second half of the 4th (beginning fortified sites that continue to be used in current publica- presumably after 378) to the late 6th century, but also tions. The myriad of substantial y different fortifications that they were of a ful y uniform type (Petru 1978, 362; in particular requires a very careful examination that Slabe 1978, 386; Šašel 1980, 14). Such interpretations must take into account the specifics of individual types, primarily rested on the site of Ajdovski gradec above as it is only possible to gain a comprehensive picture of Vranje, which was well-known at the time, but one the Late Antique settlement pattern by understanding whose church complex makes it an exception rather than its diversity. a rule among such sites in Slovenia. This relatively logical In the history of research, we already mentioned interpretation initial y corroborated with discoveries at the important role of Rudolf Egger, who considered the select sites soon proved to be oversimplified. fortified settlements with Early Christian buildings as In Italy as wel , Volker Bierbrauer used the inves- refuges for the Roman population, saw two of them as tigations at Invillino to interpret these fortifications as military posts guarding a major route to Italy and one as Roman settlements constructed in a specific historical a permanently inhabited settlement or vil age (Ulrichs- situation around 400 that remained inhabited to the end berg). At that time, however, the number of such sites of Antiquity (Bierbrauer 1985, 511). known in the eastern Alps was small and did not allow Field surveys in the eastern Alpine area have a systematic approach to their study. brought to light numerous new fortified sites, which Similar functional interpretations can be observed dotted this highly endangered area at the doorstep of in Basler’s overview of the Late Antique architectural Italy (Ciglenečki 1987a; Brogiolo, Gelichi 1996; Bauer remains in Bosnia and Herzegovina, which had a great 1997; Ciglenečki 2008; Glaser 2008). Being exposed to impact on later interpretations across former Yugoslavia mass migrations, the Roman population retreated to (Basler 1972, 60–61; id. 1993, 38–40). He identified large remote areas and to fortifications. In Slovenia, there are forts with towers as castra established on the orders no traces of habitation in unfortified lowland settlements of the central authorities and manned with soldiers in the 5th and 6th centuries, with only minor exceptions protecting major communications, those east of the (cf. Milavec 2020, 160), which is in contrast with other River Neretva also the Byzantine-Gothic border; these areas discussed here, where the settlement picture is not forts may have provided shelter for civilians in times of so uniform. Slovenian sites thus enabled the identifica- danger. He marked smaller forts as castel a, which were tion of different types of fortified sites, among which primarily built as part of the local population protection the permanently settled fortified settlements were the and saw the soldiers stationed in them as militia, i.e. most characteristic and most numerous. In addition to armed local inhabitants. The third category of fortified the early forts of the 4th century, we were able to identify sites he interpreted as refuge enclosures, merely serving several military fortifications from the 6th century, but to shelter civilians, though not excluding some form of also a variety of refuges. Excavations and the analysis armed self-protection. He correlated the size of these of the small finds from sealed contexts have enabled a refuges with the size of the local population; some also more reliable dating of individual settlement waves in- hosted churches. This suggests that he presumed refuges ferred from at least three pronounced habitation phases to be merely temporary shelters and not permanent set- (Ciglenečki 1987a; id. 2008). Both chronological y and tlements. He also treated the fortifications at Crkvina typological y, the previously uniform settlement picture near Halapić and Podgradina Kamenska as refuges, now disintegrated and revealed many variations in the already marked as such by Dimitrije Sergejevski, even types of fortified sites and their gradual occurrence. though their location on a major communication and In line with the early interpretations, the different their strong defences indicate a combination of a fort and fortifications were long only understood in the fort – refuge (Sergejevski 1942, 147–150; Basler 1972, 56–57). refuge dichotomy. Particularly those from the 5th and 6th What Basler did not recognise among these sites is per- centuries were marked as refuges, but this did not cor- manent settlements, which is the most common type respond with the actual finds on the newly-discovered in many regions. He found practical y no differences sites. In my article from 1979 with a slightly provocative 322 3. SETTLEMENT IN THE COUNTRYSIDE title of Kastell, befestigte Siedlung oder Refugium? , I high-the feature most underlined is their high altitude, lighted this problem and used examples from Slovenia which is predominantly above 500 m and commonly to show that we are dealing with all three categories of even above 1000 m, not so much their relative altitude Late Antique fortified sites, among which the fortified (Milinković 2007, 167). In the first detailed overview, permanent settlements actual y predominate (Ciglenečki Mihailo Milinković interpreted hil top fortifications as 1979). I later enhanced this division in many aspects and fortified vil ages also inhabited by women and children included sites from the whole eastern Alpine area in the and not as military forts (Milinković 2007, 169, 174, discussion (Ciglenečki 1987a). In this paper, I emphasised 180). He also noted a chronological hiatus in these set- that only rare sites could be seen as clear examples of a tlements in the second half of the 5th century. A good type and that the fortifications had different sets of func- example of the changes in the understanding of set- tions in different periods, with the function changing for tlement remains is Caričin grad, presumed Iustiniana instance from a refuge in the initial phase to a permanent Prima, which is discussed here as a newly-founded city. settlement later, when it may also have served as a fort The early research focused on the church architecture (Ciglenečki 1987a, 110). Marking some of the sites in the and only revealed a small portion of the actual extent western part as military was received with scepticism at of the city, but used it as a prime example of Procopius’ first, but was later corroborated by new investigations at exaggeration with regard to size and richness! Recent in- many sites. Recently, there are even views that go in the vestigations and the use of modern techniques (LiDAR, opposite direction, i.e. seeing merely military fortifica- geophysical survey, aerial photography and others), tions in all the sites, at least in the eastern part of the area however, have revealed a considerably larger city with under discussion (cf. Curta 2017). a densely built-up interior (see Chapter 2.5.2). In his first comprehensive study of the Late Antique In his study of fortified sites in the province Epirus forts in the area of Skopje, Ivan Mikulčić proposed a Vetus, William Bowden tackled the question of how to fairly detailed division of the fortified sites (1982, 58–64). categorise hil top sites (Bowden 2003, 180) that he saw His first group only comprised praesidia and burgi from as cities, fortified sites or episcopal sees. He found there the 1st to the 3rd century, which he paralleled with castra were no differences between cities and other kinds of and watchtowers. The second group consisted of castel a hil top forts as mentioned in Procopius or Hierocles, and Early Byzantine oppida that developed into proper where the use of the term ‘polis’ and the description small cities by the end of Antiquity. The third group of the attributes of a city were also anachronistic. He combined refuges located in remote areas, devoid of agrees that the primary motivation for the location of the defensive features (towers) characteristic of castel a these sites is the natural defensive features of the terrain and permanently settled at the end of Antiquity. The last or earlier fortifications, positing that some of these sites group also included fortified vil ages ( vici) with thinner only served as refuges. In his opinion, only excavations wal s without towers and interiors hosting Early Chris- can resolve the question of whether the sites were refuges tian churches and other buildings. As a specific example or permanent settlements and that not enough is known different from all others he mentioned the fortified vil a on the Late Antique settlement in Epirus to draw any at Pelenica near Dračevo. general conclusions (Bowden 2003, 182–184). His later research led him to emphasise that half The Early Byzantine fortifications in Bulgaria were of all fortifications were of a civilian character and initial y interpreted primarily as component parts of only one quarter proper military forts, while the last the defensive system protecting the heart of the Eastern quarter were oppida and oppidula of a mixed civilian Empire. Dimitar Ovčarov identified a defensive system and military nature representing the most significant with three lines of defence. He mentioned three types group (Mikulčić 1986, 275). The monograph on the of fortifications: large urban centres, small and middle- Late Antique cities in North Macedonia incorporates sized cities, and newly-built castel a (Ovčarov 1982). The castel a, most of which he believed lost their military increasingly intensive investigations in the last decades character in the 5th and 6th centuries and became forti- have brought a more in-depth knowledge of the Late fied settlements; only rare examples retained a military Antique fortifications. Discussing many of these sites role. He noted that these sites can only conditional y be in Bulgaria, Ventsislav Dintchev came to conclusions termed castel a (Mikulčić 1999, 363–388). He discussed similar to those indicated by the research in the eastern his division at greater length in the major work on Late Alps and North Macedonia (Dintchev 1997). Present- Antique and Early Byzantine forts in North Macedonia, ing the Early Byzantine fortifications in the dioceses of which presents a synthesis of all the findings mentioned Thrace and Dacia, he documented three basic categories above (Mikulčić 2002, 58–68). of newly-founded fortified sites. The first category is In Serbia, the Late Antique fortified sites were fortified settlements that comprise a large part of the initial y mostly marked as refugia, but it soon became sites from the 5th and 6th centuries. The second category clear that these sites were of a more complex nature is military forts. The third category comprises refuges, (Milinković 2007, 166). In defining hil top settlements, forts protecting churches, fortified monasteries and 323 Slavko CIGLENEČKI fortified vil as, the last ones already abandoned in the investigation that does not reveal the basic features. In first half of the 5th century (Dintchev 2006; id. 2007). addition, their nature may be subject to change in time; He also mentioned the different functions of fortified for instance, a refuge may develop into a permanent sites in different periods of Late Antiquity (for example, settlement, a military fort may become a civilian settle- Chertigrad was original y a refuge that transformed into ment and vice versa, and a myriad of other possibilities. a permanent settlement in the 6th century). He also noted The permanently inhabited fortified settlements that settlements with a militarised population predomi- form the most numerous and most diverse category. nated and proper military forts were rare. Most lie in natural y well-protected elevated locations We already mentioned the position of Florin Curta, and are additional y fortified with different defensive who presumes military garrisons stationed at most forti- elements (wal s, towers, entrance reinforcement). They fied sites in the eastern part of the Empire (Curta 2001a; are predominantly self-sufficient settlements of the civil- id. 2001b). He later altered this view somewhat and won- ian population that have, at least in the 6th century if not dered, even in the title of a paper, whether these sites were earlier, one or more churches in the interior. real y forts or rather refuges, concluding that the ‘hill-top The military fortifications have (at least in their sites in the Balkans may not all have been military, but early phase) a rather regular plan with substantial forti- none of them appears to have functioned as a fortified fications, with housing concentrated along the interior vil age’ (Curta 2013, 812, 837). In his following paper, he of the defensive wall and sometimes also in the towers. writes on a matter of current (?!) debate, namely whether The interior is predominantly empty, some have an the sites were ‘civilian (fortified vil ages) or military (forts Early Christian church built in the central part in the inhabited by soldiery)’ and concludes that the hil top sites 6th century. The military presence is corroborated by the were mainly forts. In his opinion, ‘the world of the Balkans prevailing finds of military equipment, weapons, also was not one of fortified vil ages, but one of strongholds coins. They are located in proximity to major commu- maintained and supported by the state’ (Curta 2017, 439, nications or other strategical y important spots. 449). In some measure, this opinion may hold true for the The refuges have very modest habitation traces eastern Balkans, but is certainly not valid for the western in the interior, with the merely occasional occupation part where fortified settlements and refuges predominate, confirmed by scarce smal finds. Prehistoric ramparts are with the exception of some Byzantine forts on the islands. often reused for defensive purposes and the fortification Summarising the different views that researchers wal s are improvised. have on the functional diversity of fortified sites, we can observe that most distinguish between three basic The main characteristics and examples of each of categories, with the most elusive being the distinction the three categories are presented below. Also discussed between military forts and civilian settlements, but also are the sites that display a hybrid function that either the identification of hybrid sites with both military and simultaneously combine the characteristics of different civilian components. We can propose that investigating categories or their function changed through time. the poorly known refuges and the associated unfortified settlements in their vicinity appears to be among the priorities of future research. Permanently inhabited fortified settlements Fortified settlements (vil ages) were the quintes- CATEGORIES OF FORTIFIED SITES sential Late Antique fortified sites, inhabited either permanently or over prolonged periods of time. Their The fortified sites of Late Antiquity ( Fig. 3.307) numerous, frequently masonry buildings in the inte- display a great diversity of form and function. There are rior make them most readily identifiable. They can in several elements that allow a distinction between differ- short be defined as fairly densely populated settlements ent types of such sites, namely plan and size, defensive on natural y well-protected spots that were often ad- wal s, towers, form and distribution of buildings in the ditionally fortified with man-made structures. They interior, characteristic array of small finds, density of constitute a vast group of settlements of many forms small finds in the cultural layer, location with regard to that mainly differ from one another in the extent and major communications, presence of foreign elements distribution of buildings in the interior. One or more and others. Early Christian churches most often dominated the These elements define three main categories: settlement from the highest point, while other build- permanently inhabited fortified settlements, forts and ings were arranged either evenly across the interior or refuges. We should emphasise that these are not clearly concentrated in certain parts. In addition to a great separated categories and the boundary between them relative altitude, these sites are also typical y located in is often uncertain, in some cases because of indistinct areas of considerable absolute altitude. The structure features, but more frequently because of a poor state of of the small finds recovered from the interior, largely 324 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.308: Ajdna above Potoki. Location of the settlement on a hard-to-reach and high rocky peak (1992). consisting of local y made pottery and tools, suggests In the eastern part, the numerous permanently that a great majority of them were autarchic. The sur- inhabited fortified settlements include Gradina Ras, rounding landscape with its natural characteristics may Liška Ćava, Zlatni Kamen near Novi Pazar, Samograd, have been used for small fields and pasturing. Some Korishë, Pecës, Ai Giannis − Adriani, Palaiokastro, sites even revealed traces of long-distance trading. As Kastritsa near Ioánnina, Velika and Olympia. They opposed to forts, the appearance and structure of the predominantly hold buildings of clay-bonded stone or fortified settlements from the 4th to the early 7th century sun-dried mudbrick (e.g. Gradina Ras, Tupi krš, Gole- do not show differences. manovo Kale, Dichin, Qafa), while wal s of mortared These settlements greatly differ from one another stone were only used for churches and other important in terms of protection features and display the complete buildings. At many sites, even the defensive wal s are array of the defensive elements used in Late Antiquity. drystone constructions (Dintchev 2007, 483). Some Some, for example Monte San Martino − Riva del Garda of the large settlements show a division into a better and Ajdna ( Fig. 3.308), were sufficiently protected by protected acropolis and a lower part; such settlements their location on a rocky crag. Most others have defen- have only been recorded in the eastern part, for example sive wal s ranging from thin to substantial, a limited at Balaj nac, Kale in Bregovina, Pecës, Korishë, Gradište number of differently shaped towers, ditches, ramparts, (Sobri) near Oraše, Markovo Kale near Malčište, Gradok buttresses, proteichismas, reinforced entrances and near Čanište and Markovi Kuli on Mount Vodno. special y fortified acropoleis (cf. Basler 1972, 60–61; The fortified settlements include many where Ciglenečki 1987a, 128–130; Mikulčić 2002, 91–104; nature and duration can only be surmised from a thick Dintchev 2007, 483–494; Milinković 2015, 49–108). cultural layer, while there are no known architectural re- A particularly il ustrative example in the western mains in the interior due either to a lack of investigations part is Gradec near Prapretno, located in a mountainous or to a poor state of preservation and we may presume area away from major Roman lines of communication the existence of buildings of perishable materials. Exam- (Ciglenečki 1981). Its 0.8 ha large interior holds more ples can be found at Debelo Brdo above Sarajevo, with than twenty buildings of different layouts and a church thick defensive wal s and a rich cultural layer, Gradac in the centre. Similar settlements are those at Ajdna on Ilinjača, which only revealed the remains of a church above Potoki, Kappele near Jadersdorf and Ulrichsberg. (Fekeža 1991), and Ukosa in Stalać that only revealed Buildings were mostly made of mortared stone (e.g. storage pits in addition to a thick Early Byzantine cul- San Martino – Riva del Garda, Ulrichsberg, Gradec tural layer (Rašković 2016). near Prapretno, Rifnik, Ajdovski gradec above Vranje, Many fortified settlements were created within Tonovcov grad, Ajdovščina above Rodik, Gradina the defensive wal s of Roman or Late Roman fortifi- Bakinci), rarely of clay-bonded drystone (Invillino) or cations. A typical example is Mogorjelo ( Fig. 3.309), timber (Tinje above Loka pri Žusmu). which transformed in the second half of the 5th century 325 Slavko CIGLENEČKI Fig. 3.309: Mogorjelo. In the strongly fortified Roman vil a, traces of the settlement from the 5th and 6th centuries are visible (2009). Fig. 3.310: Paleokastra near Gjirokastra. Lowland fort with traces of later church and settlement (2009). from a heavily fortified vil a into a fortified settlement Some of the forts in Đerdap were transformed in the (Ćurčić 2010, 139). Similar examples are the military second half of the 6th century to settlements of a rural forts of Timacum Minus and Paleokastra ( Fig. 3.310), population protected with strong defensive instal a- which were inhabited by a civilian population in the tions (cf. Popović, V. 1984, 280; Špehar 2010, 146–151, 5th and 6th centuries (Petković, Ilijić 2012, 168; Bowden 154). The opposite has been observed at Rifnik, where 2003, 180–181). The fort controlling a road at Gradina the fortified settlement received towers reinforcing the in Vrsenice changed into a permanently occupied set- circuit wal s and a small garrison in the Justinian period tlement; this also occurred in the fort at Odartsi that (Ciglenečki et al. 2020, 236). became a civilian settlement in the second half of the Large ecclesiastical centres (at least two churches, 5th century (Popović, Bikić 2009; Dintchev 2007, 500). baptistery and associated facilities) were located in some 326 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.311: Hemmaberg above Globasnitz. Remainns of the earliest church at the top of the settlement plateau (2014). of the fortified settlements. Standing out among these 157–160). It lies in the lowland, on a natural y excellently is Hemmaberg ( Figs. 3.78, 3.311) with as many as six protected rocky plateau above a river, it is heavily forti- churches which led to interpretations as a pilgrimage fied and densely inhabited. A similar combination can destination (Glaser 1997, 118–120). Similar sites in be observed at Đuteza near Dinoša, Kekola, Kale near the western part are also Ajdovski gradec above Vranje Debrešte, Žegligovski kamen, Gradište near Pčinja and (Petru, Ulbert 1975) and slightly larger Tonovcov grad elsewhere. near Kobarid (Ciglenečki et al. 2011). In the eastern part, The fortification at Golemanovo Kale is an exam- the acropolis of the settlement at Gradište Stenče holds ple of a hybrid fort, built on the incentive and with the two churches with three baptisteries (Kostadinovski, involvement of the army, but one that hosted civilians Cvetanovski 2008). A specific site is the memorial ec- who supplied the army deployed to the limes (Werner clesiastical centre at Harilaq, which has a church flanked 1992, 413). Dichin is a similar example, serving as by a pair of symmetrical y positioned chapels (Rama a supply base with a garrison of foederati in the 5th 2020, 117–118). century. Its inhabitants were presumably not only Some settlements were built on naturally less performing military tasks, but also engaged in farming protected spots, frequently on low elevations or at river the surrounding land (Poulter 2007b, 38). Both are sites confluences, but in proximity to communications or in the limes hinterland, which explains their specific other strategic points. They were heavily artificially hybrid function. fortified and presumably hosted army troops in addition The presence of both soldiers and civilians has to the civilian inhabitants. It is a group of settlements also been established at Dobri Dyal, a fort from the late with features similar to those of forts, but still clearly 4th and early 5th centuries (Poulter 2013). The site at displaying a double nature; the thick fortification wal s Gradishteto near Riben was a large foederati settlement in some cases combined with towers indicate an in- (Torbatov 2018). volvement of the army in their construction, whereas In the western part, a pronouncedly civilian nature the distribution of buildings in the interior points to a can be ascribed to Kastrum on Veliki Brijun, a fort and civilian settlement. The recovered small finds reveal a a port that served as the starting point for the shortest predominantly civilian population, but they also include crossing of the Adriatic en route for Ravenna (Marušić pieces of military equipment and weapons. A typical 1975, 338). A revealing fact here is that the church was example of settlements with a distribution of buildings only constructed outside the fort after it was added a in the interior pointing to civilian use and at the same strong civilian use. time having certain military tasks (controlling and pro- The problem already touched upon in the previ- tecting communications, signalling) is Mezdra, where a ous chapter regards the difficulties in distinguishing Late Antique fortified settlement grew on the spot of a between presumed urban and other large settlements. Roman fort and pre-Christian sanctuary (Torbatov 2015, There are some settlements that cannot be seen as cit- 327 Slavko CIGLENEČKI Fig. 3.312: Sv. Pavel above Vrtovin. Fortified settlement with habitation terraces from the north (2021). ies, but their size, distribution of buildings, extent of decorated with mosaics and an impressive dam on the church buildings and other features indicate a greater river, though the lack of investigations hinders a posi- significance within the multitude of smaller fortified tive identification of the nature of this settlement, at a settlements. The well-researched territory of Slovenia distance of only 13.5 km from Caričin grad (Milinković holds numerous hilltop fortified settlements that 2015, 236–248). include some of an outstanding size. The 10 ha large In Bulgaria, Ventzislav Dintchev used size to settlement in Kranj has been classified as a city. Smaller distinguish between several categories of sites. Settle- settlements, which are still large compared with most ments extending across more than a hectare were seen others, with a densely built-up interior, are those at as quasi-urban centres (Dintchev 2007, 482). Standing Sv. Pavel above Vrtovin (4 ha; Fig. 3.312; Ciglenečki out among these in size, thick fortification wal s and two 2021) and Ajdovščina above Rodik (3.8 ha; Slapšak large churches is Carassura. 1997). Both hold buildings with complex layouts, Ivan Mikulčić presumes a semi-urban status for two Sv. Pavel also three Early Christian churches. In the large fortified settlements in North Macedonia, namely heavily fortified settlement at Črnomelj (2.8 ha), the Kalata near Kamenica and Čebren (Grad) near Zovik architecture in the interior is only partial y known, but (Mikulčić 2002, 211–213, 287–288). its location at a confluence and its thick defensive wal s The settlement at Qafa has been defined as a small are similar to those in Kranj (Mason 1998). city in literature (Ćurčić 2010, 182–183). Bearing in An example of a large settlement with a central mind its relatively small size (0.8 ha), fairly simple build- function in Croatia is at Marija Gorska near Lobor (sur- ings and a single church, however, this definition seems face of c. 2.5 ha), where investigations unearthed part of questionable. It seems more reasonable to classify it as a the fortification features and a church with a baptistery large fortified settlement. Given its location on top of a (Filipec 2007). The Late Antique fortified settlement at mountain pass and its mighty wal s with evenly spaced Gradina in Bakinci in Bosnia is not known in detail, but towers along the side of easier access, it could also be its central role can be perceived from the mighty church partly military in nature. buildings on the nearby saddle; it has been tentatively identified as the city of Balkis, which the Avars report- Fortified monasteries have only rarely been identi- edly captured on their march towards Dalmatia in 597 fied in the areas under discussion and were presumably (Vujinović 2014, 179–182). not common (see Curta 2001c, 53–57; Ćurčić 2010, The site at Kale, Zlata in Serbia may be identified as 142–145; Uroda 2013). An example that has been reli- a newly-founded city given its size ( c. 7–10 ha), a basilica ably identified is located on the Majsan Island (Fisković 328 3. SETTLEMENT IN THE COUNTRYSIDE 1980), others are presumed at Orlandovtsi, Lozenets and Shkorpilovtsi (Dintchev 2007, 530–532). Military fortifications Ever since research began, attempts at defining military fortifications in the mass of different fortified sites have only seen limited success (cf. Basler 1972, 60–61; Ciglenečki 1987a, 111–114; Brogiolo, Gelichi 1996, 11–22; Mikulčić 2002, 63–68; Përzhita, Hoxha 2003, 119–123; Dintchev 2007, 516–528; Sarantis 2013, 9–18; Brogiolo 2014, 153–154). Having said that, most researchers agree that the forts from the late 3rd and the 4th century are considerably easier to identify in compar- ison with those built in the second half of the 5th and the 6th century. The former only relied on natural defences in a small measure and were primarily protected with thick wal s with evenly spaced towers. The layout was fairly regular as most were built on low elevations or even in the lowland, which did not require adaptations to the terrain. Most were also located in immediate proximity to main communications, crossroads, the limes, and many are mentioned in ancient texts. The group of forts from the early period of Late Antiquity certainly includes those of the Claustra Alpium Iuliarum system, particularly Ad Pirum – Hrušica and Ad Frigidus – Castra, but also many located along the limes and at strategic points in the interior (e.g. Doberdo, Fig. 3.313: Predjama. A small military crew was occasional y Tokod, Velike Malence, Pasjak, Iatrus, Harmana, Gegje, stationed in the cave in the middle of the rock wall (1992). Bushati, Palaiokastro, Bargala). They existed alongside smaller forts, primarily burgi, such as Castelraimondo, Gradishteto near Riben and Gradište near Pakoševo. cations protection?). Such an example from Slovenia is at Contemporaneously, there were other fortifica- Mali Njivč, which characteristic military equipment and tions, located at natural y better protected and higher weapons date to the last third of the 4th century (Istenič sites. Most had thinner wal s and less towers. The range 2015, 373). Another one is Jama below Predjama Castle of small finds is similar to that from the reliably identi- ( Fig. 3.313), where a cave passage in the middle of a fied military sites, but combined with traces of civilian rocky face revealed habitation traces from the Late Ro- presence. Examples of such forts are Georgenberg near man period that included military equipment (Korošec Kuchl from the second half of the 4th and the early 5th 1983). An example of an excellently protected site is the century, Ančnikovo gradišče from the second half of large fort on the Sirmione Peninsula at Lake Garda from the 4th and initial decades of the 5th century, Tonovcov the second half of the 4th and first half of the 5th century, grad from the second half of the 4th and initial decades which was enclosed with a thick defensive wall and could of the 5th, Kuzelin from the late 4th and early 5th centu- welcome large numbers of refugees. ries, Crkvišće Bukovlje from the second half of the 4th Common to all these forts dating from the last third century, Gradina in Vrsenice from the last third of the of the 3rd to the initial decades of the 5th century is the 4th, Gradina Ras from the 4th, Borovets near Pravets absence of Early Christian churches. from the third quarter of the 3rd and Odartsi from the second quarter of the 4th to the first half of the 5th cen- It is much more challenging to define the forts of tury. Such hybrid forts with a combination of soldiers the later group, from the second half of the 5th and the and civilians were primarily tasked with controlling 6th century, as these in a large measure completely lack and protecting the communications, but also provided the features defining the earlier military fortifications. shelter for refugees. Similarly as for the group of hybrid forts from the late The forts in natural y protected locations without 3rd to the first half of the 5th century, it is often very artificial defences were also occasional y used for mili- difficult to define their basic function and distinguish tary purposes (watch posts, signalling posts, communi- them from civilian settlements, because they frequently 329 Slavko CIGLENEČKI Fig. 3.314: Markova Mehana. Strong fort of a regular millitary unit guarded the important Succi Pass (2012). combine the features of both. They are mostly located on the most important route connecting East and West, and higher and natural y much better protected spots, were also on the border between the prefectures of Illyricum entirely adapted to the configuration of the terrain and and Oriens (Mitova Džonova 1998). With its location consequently had thinner defensive wal s. They were on a low hil , it boasts lofty defensive wal s enclosing the predominantly somewhat removed from the commu- interior that only holds army barracks arranged along nications they controlled. Living quarters were arranged the wal s and, very significantly, no church. in the towers or in the buildings abutting the defensive Of a similar nature is the fort on the small island of wal s. Many had an Early Christian church in the interior Veliki Sikavac (Gluščević, Grosman 2015) that control ed and only rarely another large masonry building. the access of the navigable route to the coastal belt. It has Ancient texts adequately explain the reasons be- thick wal s with evenly distributed towers and its interior hind the proliferation of hil top forts under Justinian, holds accommodations for a fairly large garrison that when numerous already existing and many newly-built depended entirely on provisions from the outside. forts were used to station troops (overview in Sarantis Another characteristic example of a strictly military 2016, 188–198). However, the texts do not (or not fort is Gradina on the island of Žirje, located on the navi- clearly) distinguish between the newly-built forts and the gable route along the outer side of the Adriatic islands earlier ones that were either modified or reused without (Pedišić 2001). It is protected with wal s, strong towers modifications. All this results in confusion and hinders and a proteichisma, and has a large well-constructed the identification of military posts. building abutting the wal s from the interior, presumably Many authors believed that it was not possible to the accommodation of a commander or the headquarters. identify the forts of the regular army in the last period of This group also comprises fortifications along the Antiquity, though the archaeological y better researched limes, in the Đerdap Gorge for example Veliki Gradac sites have provided indications to this very effect. Stand- near Donji Milanovac (Popović, V. 1984), Saldum ing out in the multitude of forts are some that reveal (Petrović 1984, 129–134), Bosman (Kondić 1984, 137– an entirely state-based incentive and construction, as 145) and Ravna (Kondić 1984, 233–251). These are Late well as an entirely military character. They were built Roman forts renovated in the 6th century according to for imperial purposes in neuralgic locations, primarily the same principles and with the emphasis on repairing along the limes and major lines of communication. Most or raising the defensive wal s, as well as modifying or were established on low elevations, with fairly rectilinear even constructing new towers that were larger than the wal s and are thus partly similar to the ‘classic’ Roman previous ones. Simple churches with a baptistery were camps. The undoubtedly most il ustrative example is the constructed in the interior of large forts such as Čezava, fort at Markova Mehana ( Fig. 3.314) from the early 6th Boljetin and Veliki Gradac, which were otherwise largely century, set on a strategical y vital location on the pass of occupied by poorly-built dwellings (Vasić 1984; Zotović 330 3. SETTLEMENT IN THE COUNTRYSIDE Sv. Marko Lopar Korintija Veliki Brijun Sv. Damjan Prizna Palacol Svetojanj Sv. Juraj Sv. Trojica Veliki Sikavac Modrič Pustograd Vrgada Toreta Gustjerna Gradin ŽIRJE a Svetac Osinj Ošlje Majsan Ston 0 100 km Fig. 3.315: Selected Early Byzantine forts in the eastearn Adriatic coastal area. 1984; Popović, V., 1984,). In the time of Justinianic reno- often proteichismas, as well as housing arranged along vations, these forts may have hosted regular army units, the wal s. They have no churches in the interior, which while the interiors became ruralised in the second half are otherwise present in most contemporary forts. This of the 6th century, which indicates an influx of the local may only be the consequence of a lack of space in smaller population (Špehar 2010, 146–151, 154). forts, but can it also be a sign that these were newly- In addition to these and the already mentioned built forts of the regular army? Final y, this group also fort at Markova Mehana, the Balkan interior holds includes substantial independent towers such as those at other forts at strategic locations, for instance at Koštur Toreta, Svetac and Ošlje – Gradac (Baraka Perica, Grbić near Dabrica from the 6th century (Basler 1972, 50–51), 2019). Evidence suggests that all these were constructed Madara ( Fig. 3.316) from the late 5th and the 6th century according to a pre-conceived plan and by the army and (Dintchev 2007, 522), Castra Rubra and Dyadovo from that they hosted units of the regular army. the first half of the 6th (Borisov 2010a; id. 2010b) and the Much more numerous is the group of fortified fortress at Isthmia from the mid-6th century (Gregory sites from the second half of the 5th and the 6th century 1993). They have thick defensive wal s with towers and with less pronounced military features. Defences in 331 Slavko CIGLENEČKI Fig. 3.316: Madara. Early Byzantine fort dominated the plain (2012). particular are less substantial for some fortifications, Grad near Gornji Vrbljani is an example of a hybrid with thinner wal s and less densely spaced towers. The fort constructed at a slight distance from the major com- modest housing is mainly arranged along the wal s and munications. The large building on the most exposed spot the interiors frequently hold a church. Small finds show of the defensive wal s, identified as the accommodations both a military and civilian character and some sites of a permanent garrison, marks this construction as a revealed inhumation burials of women and children. The military fort (Bojanovski 1980). The wal s are thinner and interpretation of these fortifications is not unambiguous; presumably include a single tower. The interior is large we may presume at least partial presence of army troops enough for hosting both refugees and livestock. stationed here together with their families, though they Less readily identifiable are the remains at Balajnac, may also have functioned as supply centres. They can be where the thick wal s with densely spaced towers on the defined – similarly as the less characteristic forts from acropolis only protect a large church and a very well- the last third of the 3rd and the 4th century – as hybrid constructed cistern (Jeremić, G. 1995). It is strategical y fortifications for which the ratio between military and located on a major road, while the acropolis indicates a civilian inhabitants cannot be reliable assessed. prestigious character of the fort with the lower part that These forts include Duel, which had a military could host additional troops and a civilian settlement. character combined with a civilian presence. Accom- The fortification at Čuker near Mokro polje ( Fig. modations were densely spaced along the wal s, while 3.317) shows an unusual layout (Delonga 1984, 277). Its the interior was empty with the exception of sacral location in the lowland, on a river terrace, and the evenly buildings. This led to assumptions on the interior hosting distributed rooms built into the prehistoric rampart re- shelters for the people from nearby unfortified settle- veal a fort that protected a line of communication and a ments (Egger 1929, 204–210; Petrikovits 1985). The fort river crossing. There are several similar examples in Bos- at Hoischhügel had a similar role (Glaser 1996, 69–72). nia and Herzegovina, for example Gradina above Kovaći In Italy, characteristic examples are Monte San and Gradac above Potočani, but they are only known Martino di Lundo/Lomaso and Sant’Andrea di Loppio. from surface surveys (Benac 1985, 39–41, 104–106). The former was presumably a strategic fort from the The group of hybrid fortifications may further in- second half of the 5th and the 6th century tasked with clude Krassen Kale (late 5th and the 6th century), Gradina providing supplies and controlling communications Zecovi near Čarakovo, Gaj in Babrež, Kula in Kaludra, (Zagermann, Cavada 2014). The latter is a fortification Sadovsko Kale, Gradishta of Bardhoc, Kula near Čelopek, from the late 5th and the 6th century on a natural y excel- Venec near Miokazi and Kale near Gorno Svilare (all 6th lently protected insular site with less pronounced man- century). Belgrad near Dvorište holds an unusual fortress made defences, while the small finds show a military from the 6th century that protected a large cistern abutting nature with a civilian presence, but also reflect several the defensive wal s from the exterior. A well-investigated changes of government (Maurina 2016, 751). example of a hybrid fortification is that at Gradina in Bi- 332 3. SETTLEMENT IN THE COUNTRYSIDE Fig. 3.317: Čuker near Mokro polje. Fort on the edge of the river terrace (2013). ograci, which is believed to have accommodated soldiers ficulties in determining whether a site is merely a refuge with their families (Čremošnik 1989). of civilian population or one that was at least in part used Some military fortifications, such as Duel, Zecovi for military purposes (watch post, signalling post). near Čarakovo and Dyadovo, were designed to have Refuges were occupied in times of danger and were a large empty space in the interior to welcome people not conceived for permanent habitation. Their natural y fleeing the nearby unfortified settlements. protected location, however, caused many of them to become inhabited later, which covered the traces of the Final y, there is a number of sites where the current original occupation. At several sites, the initial refugial state of research does not allow a reliable distinction nature can only be inferred from the well-preserved finds between military posts, permanent civilian settlements lost here in the first phase of occupation. The fortified and – the most likely option – a combination of the settlement at Kirchbichl in Lavant is an example where two (e.g. Zidani gaber, Gradina in Radalica, Tupi krš, this first phase can be clearly identified as a refuge or Đurđevica in Đerekare). They do, however, aptly il us- the beginning of a permanent settlement. The diagnostic trate the broad range of the different forms of military finds date as early as the second half of the 3rd century, and even more of the hybrid fortifications in the last whereas the buildings indicate a later time (Grabherr, period of Antiquity. Kainrath 2011, 188–191). The sites not occupied later (Veliki vrh near Osredek pri Podsredi), those where later occupation Refuges was modest (Trojan) and the sites where investigations were able to reliably date the early phase and establish its This is a group that joins the fortified sites with lesser elements (Kartal kale near Ruyno) are therefore all the defensive features, where the wal s were most frequently more valuable. The last site also confirms that only sys- improvised or the earlier prehistoric ramparts and natural tematic excavations will enable a more substantial num- steep slopes reused for protection. The interior only very ber of refuges to be identified. The initial investigations rarely revealed evidence of permanent housing and smal frequently did not pay enough attention to the modest finds are few in number. They show differing degrees early features, which are not adequately described and of occupation or habitation, which makes it difficult to interpreted in the publications, while the artefacts from distinguish them from the briefly inhabited settlements. the Early Roman period were – regardless of the state of With regard to small finds, the sites identified as refuges preservation – only identified as residual finds. in this book yielded only roughly 10% of the small finds We already mentioned above that refuges were often recovered at permanently inhabited settlements. The set up within prehistoric hillforts, of which the ramparts scarce (mainly metal) finds unearthed at the early refuges and remote location provided sufficient protection. An from the last third of the 3rd century frequently cause dif- example is Veliki vrh above Osredek pri Podsredi, where 333 Slavko CIGLENEČKI traces of makeshift dwellings and rare artefacts came to only became more intensely occupied in the 6th century light on the small terraces of the site. Coins and diagnostic (Popović M. 1987). metal finds set the brief occupation to the second half of A good example in Bulgaria is Chertigrad, where a the 260s or the early 270s (Ciglenečki 1990a, 147–154). refuge was established in the 4th/5th century at an altitude A similar dating can be proposed for the first use of Mali of 1284 m asl, within a well-protected prehistoric hillfort, Njivč above Novaki, a rocky crag with only natural de- which only became a permanent settlement in the 6th fences (Istenič 2015, 373). The systematic investigations century when housing was constructed (Dintchev 2007, at Tonovcov grad near Kobarid and Korinjski hrib above 529). A different example is Gradishteto near Debrene, Veliki Korinj also revealed the first peak in monetary in the lowland, which was strengthened with a wal circulation in the last third of the 3rd century, which is and a ditch in the second half of the 4th century. In this the time frame for several well-preserved metal finds as phase, but also later, in the 6th century when the wall was wel . Neither of the sites yielded remains of contempo- reinforced and a church built in the interior, it served as rary buildings, while the prehistoric rampart sufficed as a refuge (Dintchev 2007, 530). The natural y protected defence (Ciglenečki et al. 2011, 292; Modrijan, Milavec location was used at Kartal kale near Ruyno, which was 2011; Ciglenečki et al. 2020, 342). reinforced with defensive wal s and a fortified entrance. An increase in small finds at Marija Gorska above This site was briefly used as a refuge between the late 4th Lobor has been detected in the 3rd century, while defen- and the middle of the 5th century, when it was abandoned sive wal s were presumably built on top of the prehistoric in the face of the invading Huns (Atanasov 2015). rampart in the 3rd–4th century (Filipec 2007, 415–416). At Gradište near Delisinci in North Macedonia, Occupation traces from the last third of the 3rd century Late Roman pottery came to light inside the Hellenistic were reliably established at Kuzelin, when the summit defensive wal s that speaks of an occasional refugial use was presumably only enclosed with a palisade and the of the site (Mikulčić 2002, 416). The whole island of site served as a refuge (Sokol 1994, 201–202). Golem Grad, Konjsko in Lake Prespa has been positively The refuge at Gradina Ras in Serbia was sited on identified as a refuge in the 4th–6th century (Mikulčić a natural y protected rocky peak. It is coin-dated to the 2002, 379–381; Bitrakova Grozdanova 2011). It revealed middle and second half of the 3rd century (Popović, M. individual buildings inhabited over a long time, while 1999, 70–71). At Trojan, a palisade is believed to have the access to the vast rocky plateau above the lake was been erected on top of the prehistoric rampart in the additional y protected with a wall and thus provided 3rd century. Investigations revealed no buildings. The shelter for many people from the surrounding area. site may also have been occasional y inhabited in the 4th century. It presumably controlled the nearby lines of Not much is known on the refuges from the second communication, but was also used as a refuge (Ivanišević half of the 5th and the 6th century, when other forms of 1989). fortified sites predominated. In Italy, the fort at Castelrai- Monetary circulation shows that the refuge at mondo is believed to have been abandoned as a military Dobri Dyal in Bulgaria dates to the last third of the 3rd construction around 430, while a refuge was established century, when a Bronze Age rampart was reused for in the ruins of earlier buildings in the second half of the defence (Poulter 2013, 369). 5th century and remained in use throughout the 6th and Ivan Mikulčić identified a number of refuges in part of the 7th century (Santoro Bianchi 1992, 195–204). North Macedonia (Mikulčić 2002, 61–63). It is often In the vicinity of Tonovcov grad, a c. 120 × 20 m difficult, however, to discern his division line between large post was found at Gradec near Logje that was the fortified sites with a thick cultural layer and those only protected with very steep slopes. The find of an with only sporadic smal finds. For the large fortified set- equal-arm brooch dates it to the last phase of Antique tlement at Kalata near Kamenica, for example, he men- occupation, while the modest cultural layer defines it as tions artefacts from the 3rd–6th century, while the survey a refuge (Osmuk 1985; ead. 2001). results are unclear as to whether the site was a refuge The group of late refuges may also include Kučar, from the 3rd century on the acropolis enclosed with a where the defensive wall encircled an ecclesiastical drystone wal , or the whole site has been permanently centre and a vast empty space. In times of danger, only inhabited from the beginning (Mikulčić 2002, 211). a great number of refugees could mount a sufficient Simultaneously with refuges, many earlier unfor- defence of the site (Dular et al. 1995). tified settlements in the lowland were still inhabited, The numerous fortified sites along the edges of which suggests that most refuges may be expected in the karst poljes in Bosnia and Herzegovina only saw the last third of the 3rd and the 4th century and less later. limited investigations, which makes it difficult to dis- tinguish refuges from settlements. The site at Biograci, Of the refuges from the second half of the 4th for example, was seen as a refuge in 1972 (Basler 1972, century in Serbia, we should mention the presumed 59–60), whereas later excavations unearthed a perma- first phase of the fort at Južac near Sopočani, which nently inhabited fort (Čremošnik 1989). A refuge can 334 3. SETTLEMENT IN THE COUNTRYSIDE reliably be identified at Gradina in Vidoši, where a (Grigorov 2011a), in North Macedonia near Gradište mortared defensive wall was constructed on top of pre- near Pčinja, Markovo Kale near Malčište, Gradište near historic ramparts (Basler 1972, 60; Benac 1985, 97–99). Pakoševo, Kale near Debrešte and Gradište (Sobri) near The scarce finds from the 3rd and 4th centuries found Oraše (Mikulčić 2002, 142, 174, 180, 350, 466), but at Gradac on Ilinjača indicate the possibility that the also a settlement with an ecclesiastical centre at Stenče prehistoric hillfort was reused as an occasional refuge (Kostadinovski, Cvetanovski 2008). In times of danger, in that period (Fekeža 1991, 189). the inhabitants of these settlements sought refuge in A thin cultural layer from the 6th century came to nearby fortifications. light in the interior of the prehistoric hillfort at Hum near Tutin, at 1502 m asl, which was reinforced with a thin wall (Ivanišević 1988). A similar refugial nature can On whose initiative were fortifications constructed? be observed at Gradina Ramoševo near Tutin, a slightly better fortified site with no traces of permanent habita- There are different opinions in literature as to who tion in the interior (Ivanišević 1987). initiated the construction of fortifications. In this book, The site at Elenska Basilica near Pirdop, protected the subject is indirectly tackled in the discussion of all with wal s and towers, served in the 6th century as a three categories of fortified sites. refuge for the inhabitants of a nearby unfortified set- There is extremely little evidence on private initia- tlement (Dintchev 2007, 532–533). The earlier refuge tive. For the eastern part, evidence might be seen in the at Gradishteto near Debrene was refortified in the 6th inscription from Izbičanj (se Chapter 2.5.2), dating to century, and also added a church. the mid-6th century, which probably indicates just such An interesting complex in North Macedonia is Kula private initiative if further research confirms a fortified near Kalauzlija, which holds a fort and a less fortified settlement mentioned in the inscription at the nearby refuge with a church. The interior of the last revealed no site of Kovingrad. other buildings, only sporadic pottery sherds (Mikulčić Several such examples are known outside the area 2002, 391). under discussion. One is the fort at Lake Como that is In Greece, the Late Antique activities at Kastro believed to have been constructed on the incentive of Rizovouni only involved the renovation of a part of Marcelliano, Subdeacon of Milano (Brogiolo, Gelichi the earlier, Hellenistic wal s and the construction of an 1996, 20; Marano 2016, 749). There is also an inscription Early Christian church in the interior. The absence of found in the French Alps that relates a private individual other buildings and only rare fragments of Late Antique by the name of Cl. Postumus Dardanus who set up a ref- pottery indicate a refuge that provided shelter to the uge named Theopolis in a high mountain valley (Johnson inhabitants of an unfortified settlement in the vicinity 1983, 242). This inscription, carved into a rock at the en- (Bowden 2000, 107–108). trance into the valley, is not supported by archaeological Across the area under discussion, there was also evidence of a settlement here and it would seem that the a number of caves and rock shelters inhabited or oc- location in a high-altitude and remote valley sufficed for casional y occupied in Late Antiquity, though not much a refuge. At the eastern border of the Empire, in Syria, a is known of these sites (cf. Ciglenečki 1999, 294). A private fort was built in the mid-6th century in the centre more detailed analysis and identification of functions in of the large settlement of El Andarin, at the edge of a the different phases of Late Antiquity will only be pos- desert, which is clearly recorded in an inscription above sible after cataloguing and systematical y investigating the entrance (Strube 2011, 216–219). such sites. As an example, we should mention Pećina Leaving aside the private initiative, which sources Laganiši in Istria, which revealed prehistoric and reveal to have played a minor role, discussion mainly Late Antique artefacts, stone structures and hearths, focuses on the relationship between local and state presumably left behind by a Romanised population initiative. Most authors see state initiative primarily in (Komšo 2008). connection with military forts and rarely with fortified settlements. All three categories of sites include examples In the time between the last third of the 3rd and where traces of contemporary unfortified settlements the middle of the 5th century, there is evidence of both were found in proximity to fortified sites. Such sites state (characteristic military forts, some even mentioned not explored in detail have been mentioned at Sirmione in ancient texts) and local initiative (refuges and in a and Garda in Italy (Brogiolo, Gelichi 1996, 171) and at small measure also fortified settlements). This is also Podgradina Kamenska in Bosnia (Basler 1972, 56). Many the time when several hybrid fortifications were con- examples also came to light during the field surveying in structed, with impressive defensive elements that point the eastern part of the area under discussion. In Bulgaria, to the initiative and construction by the army and with there are examples near the forts at Dyadovo (Borisov inhabitants composed of soldiery and civilians (e.g. 2010b), Madara (Dintchev 2007, 522) and Krasen Kale Ančnikovo gradišče). 335 Slavko CIGLENEČKI It is more difficult to differentiate between state locations were constructed on the initiative of the Em- and local initiative for the second half of the 5th and peror (e.g. in Liebeschuetz 2007, 107). Other authors the 6th century, though distinctions between the east- presume that the forts constructed after the 4th century ern and western parts come into sharper focus. In the were part of central y-implemented measures, but their western part, a single source refers to state initiative, form and construction depended on local organisation namely Cassiodorus’ report of the edict of Theoderic (cf. Wilkes 2003, 754; Brogiolo 2014, 153). An example that urges the inhabitants of Tridentum to move to the of this is Sant’Andrea di Loppio, a fort constructed on natural y-protected hill of Verucca (Doss Trento) (Bi- state intervention using military engineers with a par- erbrauer 1985, 497–498; Brogiolo, Gelichi 1996, 18). In ticipation of the local population (Maurina 2016, 751). the eastern part, Procopius mentions the construction It is also posited that local authorities initial y instigated of numerous military posts in the time of Justinian, the construction of fortifications, while the army gov- though they can only rarely be reliably identified in the erned them in the Gothic and Langobard eras (Brogiolo archaeological remains (cf. Sarantis 2016, 161–198). 2014, 153). The question raised in connection with the A large part of the fortified sites in the west was sites in Bulgaria is whether they represent new Gothic created on local initiative. This can primarily be inferred settlements or merely refuges of the local population from the choice of location at remote, high and natural y (Poulter 2007b, 15). Many believe that the construction protected sites. The defensive elements are modest and of fortified settlements merely formed part of Justinian’s poorly built, the buildings in the interior show a hap- military strategy (cf. Curta 2013, 837–839). hazard distribution. The small finds in them reveal a predominantly self-sufficient economy. The fortified set- We may presume that most fortified settlements tlement on Ulrichsberg was presumably built by farmers and refuges were initiated and also constructed by the (Egger 1942, 266). All the castra in South Tyrol and Friuli local communities, whereas the category of military forts are believed to have been constructed on local incentive reveals an imperial interest and therefore also incentive. (Bierbrauer 1985, 511, similarly in Brogiolo, Gelichi There are several hybrid options, where the defensive 1996, 35–36). Gradec near Prapretno presumably hosted wal s were presumably built with imperial help, whereas a local refugee community (Ciglenečki 1981, 430). In the buildings in the interior were the responsibility of the the eastern part as wel , many fortified hil top sites in local population (Golemanovo Kale, Dichin). We should Serbia are believed to have been vil ages (Milinković also mention the possibility of the same sites being sub- 2007, 172–176). Most fortified settlements of the 5th ject to different incentives in different periods. There is and 6th centuries in Bulgaria were presumably built to ample archaeological evidence of settlements fortified host a local population (Dintchev 2007, 482–483); this on local initiative, but additional y fortified by a minimal local incentive has been particularly emphasised for the participation of the state that used the sites to control refuge at Kartal kale (Atanasov 2015, 225). lines of communication or even accommodate smal Alongside the great local initiative, however, there army units. An example is the settlement on Rifnik, is a series of forts where we may suppose state interven- built at least in the late 5th century on local incentive tion in a variety of forms, ranging from reliably identifi- and receiving a new defensive wall with towers under able military forts to settlements constructed with the Justinian, which together with a varied array of military help of the state, but also vice versa, local initiatives equipment indicates a state-prompted fortification. It subsequently transformed to serve the needs of the would appear that numerous fortified settlements in the state. These are marked by a location in the vicinity of eastern part were transformed according to this model major routes or other strategic points, thick defensive in the 6th century, but also some along the major roads wal s with evenly spaced towers and a great amount of in the western part. Fortified settlements, constructed military finds. The opinion most frequently voiced in on local initiative, predominated in remote and high- literature is that the local population constructed some altitude areas, which the authorities merely controlled, fortified hil top settlements, while those on strategic but not helped with state interventions. 336 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS. GENERAL REMARKS ON THE CHRONOLOGY OF THE SETTLEMENT CHANGES In order to better comprehend the changes in the There is a large number of sites that have only been settlement patterns, we should begin with a diachronic broadly dated, either to the Late Roman, Late Antique, examination of the basic transformations that settlement Early Byzantine periods or to one or more centuries. witnessed from the last third of the 3rd century, when These sites are usual y only known from surface surveys the first major changes can be observed, to the first or small-scale trial trenching and do not provide more third of the 7th century, when Antiquity came to an end precise dating of the settlement shifts. Dating the duration across a large part of the area between Constantinople of settlements with coins alone is also not sufficiently relia- and Ravenna, and left behind an utterly different set- ble, as it has long been established that earlier, particularly tlement picture. Late Roman coins remained in circulation for extended The process of abandoning unfortified lowland periods of time and even came to light at sites from the habitations (primarily vil as) most likely already began 5th and 6th centuries (cf. Liebeschuetz 2001, 43–46; Guest in the second half of the 3rd century. Settlements were 2007, 298–299; Kos 2011, 229–237). For a precise dating, then numerously deserted in the second half of the 4th we primarily used the sites where investigations yielded century and only rare ones persisted beyond this date. reliable stratigraphic and chronological data. For most of the 5th and the 6th century scant evidence of Systematic investigations at many sites broadly dated their existence mainly comes from field surveys or indi- to the 4th–6th or even 3rd–6th centuries revealed several rectly from contemporary buildings (mainly churches) waves of settlement separated by gaps of differing dura- in their proximity. tions. The fortified hil top settlement at Tonovcov grad, for Concurrently with the decline of cities and the example, shows a brief occupation of the hill in the 270s, abandonment of unfortified lowland settlements, we followed by a long hiatus, a lengthy period of habitation see the appearance of fortified forms of settlement. In from the last third of the 4th century to the 430s, again a the western part, they have often been dated between hiatus, of half a century, and final y a permanent settle- the 4th and the 6th century, and seen as simply replacing ment from the late 5th to the opening decades of the 7th the earlier, unfortified settlements, while in the eastern century. Similar observations have been made for other part they were predominantly attributed to the time of sites, for instance the forts in the Đerdap Gorge (Vasić, the Emperor Justinian. A more detailed examination Kondić 1986, 549–558), at Iatrus-Krivina and Mezdra. of all the settlement forms reveals, however, that such The settlement changes are not uniform across generalised conclusions oversimplify a very complex the vast and geographical y heterogeneous area under process, which involves migrations from cities to the discussion, visible in the urban fabric and the changed countryside and vice versa, but also from unfortified to appearance of the countryside. Their administrative, fortified settlements. Studying this topic has to contend religious, economic and primarily military-political un- with a poor knowledge of the final phases of cities and derpinnings may be well-known, but only archaeological an unreliable dating of the different phases of repeated research can reveal the major changes at a particular site. settlement in unfortified and fortified sites. The number Here, we can observe the changes by tracing several key of well-researched sites has been growing in recent times, elements that include the destruction or abandonment but we still know very little (or cannot identify) settle- of a city or countryside settlement, shift of habitation ment remains in the periods when the unfortified sites location, construction or renovation of defensive instal- were already largely abandoned and the fortified posts lations, major changes in the housing and the construc- only served as refuges. This hiatus is particularly glaring tion or abandonment of Early Christian architecture. In in the middle third of the 5th century in the western part, the countryside, the main indication is the construction and throughout the second half of the 5th and partly in of new military fortifications and a mass use of fortified the first third of the 6th century in the eastern part. settlements and refuges. 337 Slavko CIGLENEČKI Fig. 4.1: Rome. Aurelian Wal , the most important and symptomatic example of the fortification of cities in the empire in the last third of the 3rd century (2007). 4.1 FIRST CHANGES OF THE SETTLEMENT Some cities show destruction and abandonment PATTERN IN THE SECOND HALF OF THE 3RD that were soon followed by renovation, as well as the AND FIRST HALF OF THE 4TH CENTURY spread and creation of new fortified urban areas. In Teur- nia, habitation terraces were abandoned in the second A sharp break in settlement, both in the cities half of the 3rd century. A large fort was constructed at and in the countryside, is archaeological y detectable Favianis in the last third of the 3rd century that subse- in the second half of the 3rd century, particularly in its quently developed into an urban settlement. The city last third (cf. Sodini 1984, 393–396; Ciglenečki 1990a, of Gorsium ( Fig. 4.2) was deserted in the 260s and only 154–156). rebuilt in the first third of the 4th century. The first Late In cities, this break is primarily visible in the Antique renovation of Ulpiana dates to the late 3rd or fortifications and in some cases also in the reduction early 4th century. Serdica was greatly enlarged under of urban areas. Many Roman cities in northern Italy the Tetrarchs or Constantine. The vicus in the city of become enclosed with city wal s at this time (Johnson Novae was destroyed in the Gothic raid in the second 1983, 117–121; Christie 2007, 563–565). In Verona, an half of the 3rd century, which led to an extension of the inscription dates their construction to 265. Very reveal- adjacent legionary fortress that accepted the refugees. An ing is the construction of the Aurelian Wall in Rome il ustrative example is the destruction and abandonment ( Fig. 4.1), the heart of the Empire, which indirectly tel s of the city of Styberra, where the last sealed contexts of the great endangerment of urban and non-urban are dated with the coins of Gallienus (Mikulčić 1999, settlements in the provinces closer to the limes. Under 58–59). Thick layers of burnt debris reveal that Heraclea Aurelian, fortification of cities and the countryside also Lyncestis and Stobi were destroyed at the same time. begins in Pannonia (Thomas 1964, 389–390). Histria, Settlements in the countryside also show ample located in the north-eastern part of the limes, was forti- signs of destruction and abandonment of some set- fied with a wall in the late 3rd or early 4th century that tlements in the second half of the 3rd century. Both only encircled the natural y best protected acropolis. In vil as on the Sirmione Peninsula were devastated in North Macedonia, city wal s began to be constructed the mid-3rd century, the villa at Ponte Lambro was after the third quarter of the 3rd century, while cities deserted. At Castelraimondo, a break in habitation was not thus protected ceased to be inhabited (Mikulčić noted around 275, when several buildings were dam- 2002, 78–79; Snively 2009, 39). The major defensive aged. There are also observable traces of renovations of wall renovation in Philippi has been dated to the late old and construction of new settlements. At Invillino, 3rd century and the heavily reduced post-Herulian wall the construction of both building complexes on Colle in Athens to a time soon after 267. Santino has been dated to the late 3rd century (Martin 338 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS Fig. 4.2: Gorsium. Roman city from the south-east (2011). 1992, 261). Reliable traces of Late Antique habitation at headquarters at Ajdovščina (Castra) were constructed Lavant belong to the second half of the 3rd century. In soon after 260. The fort at Kuzelin was used in the last the vil a at Tatárszál áson, the destruction of a timber third of the 3rd century, when the modest housing in the building in a fire has been dated to 260/270. The fort at interior was only protected with a palisade. Környe was built on a levelled layer of the previous set- The refortification phase of the forts along the limes tlement destroyed under Gallienus. In Gamzigrad, the in the Đerdap Gorge (e.g. Veliki Gradac and Boljetin) earthquake and destruction of the early vil a dates just has been well-researched and dated to the second half before the construction of the Palace of Galerius next to of the 3rd century. In North Macedonia, nine military it. In Bulgaria, many vil as were deserted in the mid-3rd forts are believed to have been built in the late 3rd century century (Dintchev 1997, 121–123). or during the battles between Constantine and Licinius In addition to abandonment, the second half of (Mikulčić 2002, 65). The forts at Čučer and Gegje belong the 3rd century also witnessed the creation of several to the second half of the 3rd century. new vil as and other forms of settlement, for instance at Prehistoric hillforts in the vicinity of the new low- Radvanje, Mali Mošunj, Lisičići, Obelija and Pešterica. It land settlements begin to be inhabited in the last third of is not always possible to precisely date their beginning, the 3rd century (most often refuges), some were also addi- though we may surmise they were predominantly built tional y fortified (Ciglenečki 1987a, 123–124; Milinković after the hiatus in the 260s and 270s. They are joined by 2008, 538–544; Brogiolo, Chavarría Arnau 2014). numerous others in the course of the first half of the 4th In Slovenia, short-lived habitation from the second century, for example the vil as at Löffelbach, Skelani, half of the 3rd century is most clearly visible at Veliki Mediana, Kostinbrod and the masonry building of the vrh near Osredek pri Podsredi, which was not used roadside station at Tatárszál áson. after this time and thus represents a time capsule from A contemporaneous occurrence is the creation of the 270s. Such habitation remains are much less read- new, already slightly differently sited and shaped mili- ily discernible at multi-layer settlements such as Rifnik tary fortifications. A characteristic example is the fort and Tonovcov grad, where few coins and wel -preserved at Pasjak, which was erected in the early 270s on the artefacts indicate a brief human presence in the last third major Roman road from Aquileia to Salona. The first of the 3rd century. concentration of coin finds at Hrušica, an important The appearance of fortified hil top sites in Serbia military post, also fal s in the 270s and points to the has been dated to around the middle of the 3rd century first significant presence of the army at the site, while (Milinković 2008, 538–544). Standing out among these the construction of the fortress has been attributed to is Trojan, which was presumably inhabited to the third the Constantinian period (Kos 1986, 199–200). The quarter of the 4th century. The forts at Ras and Kraku 339 Slavko CIGLENEČKI Fig. 4.3: Hissar. Formation of the mighty city wal s with towers is dated to the beginning of the 4th century (2012). Lu Jordan were created in the second half of the 3rd of which those of Diocletian and Galerius were later century. No major destruction has been documented inhabited by civilian populations, while not much is in the forts on the limes in the Đerdap, though we do known on the palace at Šarkamen. see additional reinforcement of the defensive elements Several military fortifications can also be dated (Vasić, Kondić 1986, 549–550). In Bulgaria, the fort at to the opening decades of the 4th century, in addition Dobri Dyal briefly served as a refuge in the last third of to the already-mentioned Hrušica also Abritus, Iatrus, the 3rd century and Borovets near Pravets revealed reli- Paleokastra, Scampis, Bargala and the forts in the able traces of habitation primarily in the 260s and 270s. Đerdap Gorge. In Hungary, ‘inner fortifications’ were Most cities in the area under discussion show the erected in the middle third of the 4th century. In the last peak of prosperity after the reforms of the Emperors second quarter of the 4th century, the fortified settle- Diocletian and Constantine. This is perceptible in the ments at Mezdra and Pecës were renovated and the fort construction of city wal s, for instance in Aquileia and at Odartsi constructed. Celeia, but even more so in the construction of private domus. The last peak is clearly visible in Aquileia, Flavia Solva, Sirmium, Athens, Corinth, Gortyn and elsewhere. 4.2 MARKED CHANGES IN CITIES, GRADUAL These cities now host the first large Early Christian build- ABANDONMENT OF LOWLAND SETTLEMENTS ings such as the basilica in Aquileia, dated to the 310s, AND THE APPEARANCE OF NUMEROUS and Basilica D or Octagon in Philippi, dated to around HILLTOP SITES (LAST THIRD OF THE 4TH AND 340 (Jäggi 1990, 173; Saradi 2006, 403). Diocletianopolis FIRST HALF OF THE 5TH CENTURY) (Hissar; Fig. 4.3), a city protected with a lofty wal , was constructed in the early 4th century. The inscription Major and long-term changes in the settlement pat- from 315/317 reliably dates the construction of the city terns occur in the second half of the 4th century, mostly Tropaeum Traiani. Dinogetia is created in roughly the towards its end. Ancient texts from this period tell of same period. numerous incursions and invasions, which coincide The safe conditions of this period enabled life to with profound political and socio-economic changes intensify in the lowland, where numerous renovations perceptible across a large part of the Empire (overview and new constructions were observed within the already in Wickham 2005; Poulter (ed.) 2007a; Cameron 2012). existing settlements, while lavish vil as also became com- The unsafe conditions are clearly reflected in the transfer mon. The great majority of the refuges set up in fortified of the western capital from Milano to the natural y better sites in the late 3rd century was abandoned, only some protected Ravenna, also accessible from the sea, in 402. showed modest habitation traces. Three large imperial A more detailed chronology of the changes depends on countryside palaces were built in the early 4th century, the geographic location in this vast and diverse area; 340 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS Fig. 4.4: Stobi, Domus Fullonica. Building shows significant changes at the end of the 4th century (2013). the changes were radical in many places (such as the The Roman cities in Hungary were mostly aban- abandonment of most cities in the western part) to the doned in the early 5th century, with only Scarbantia degree of representing one of the greatest breaks in the showing prolonged existence. Sopianae was destroyed in history of human settlement. a fire towards the end of the first third of the 5th century After the last construction works in the first half and abandoned. The second phase of the ‘inner forti- of the 4th century, cities gradual y began to decline and fications’ dates to the Valentinian period, though they some were already abandoned in the late 4th or early 5th were already abandoned in the 430s with the exception century. Many urban vil as were subdivided into smal er of Keszthely-Fenékpuszta. housing or economic units, for example in Brescia, Sir- The last reliably dated habitation layers in Poetovio mium, Stobi ( Fig. 4.4) and Tropaeum Traiani. Housing date to the first third of the 5th century, while the last shows the degradation and gradual ruralisation of the convincing Late Antique trace may be identified in the urban areas. As for public buildings, earlier ones such small cemetery with the burial of a warrior from the as theatres, baths and forums were being abandoned, time of the Hun incursions ( Fig. 4.5). In Emona, the whereas church architecture prospered. destruction and abandonment of the episcopal centre The peak of prosperity that Aquileia witnessed in soon after 423, most likely even before the Hun incur- the 4th century gradual y faded in the first half of the 5th sion, can be seen as an excellent indication of the decline century. The Hun incursion greatly affected the city, cor- and abandonment of the city. roborated by an extensive layer of burnt debris from the Many Roman cities witnessed the construction mid-5th century. Neither the unfortified Virunum nor the or renovation of city wal s that frequently also brought walled Celeia thus far revealed any buildings that could about the reduction of the urban area. Along the Adri- reliably be dated to a time after the first third of the 5th atic coast, Nesactium was walled in the late 4th century, century. Around 400, the defensive wall and the first- the defensive wall in Salona was extensively renovated phase diocesan church were constructed in Teurnia. The under Theodosius II. Deultum was enclosed with a influx of people to Favianis after 370/380 caused the fort wall slightly before 383, the wall of Stobi was renovated to be enlarged and the buildings in its interior renovated. in the Valentinian period. In the early 5th century, the Some unfortified cities on the fringes of the Pan- perilous conditions are also clearly reflected in the con- nonian Plain, such as Sal a, Flavia Solva, Neviodunum struction of the land wall to protect the eastern capital and Andautonia, were already largely deserted towards of Constantinople. the end of the 4th century, with the modest small finds The construction or renovation of city wal s has indicating an only occasional human presence (overview been dated in Ulpiana to the late 4th or early 5th cen- in Ciglenečki 2011a, 185). tury, in Scodra to the early 5th century, in Heracleia (Perinthus) and Thessaloniki roughly to the first half 341 Slavko CIGLENEČKI Fig. 4.5: Poetovio. Grave of a nomadic warrior from the middle of the 5th century is the last reliably dated grave in the city (2000). of the 5th century. The refortification of Sparta presum- 400, and a further addition of a three-apsed triclinium ably occurred in the time of Alaric’s Goths, when the around 420. acropolis was enclosed with a wall that incorporated At the same time in the countryside, we can trace a theatre (Lawrence 1983, 180). In Athens, the agora the gradual abandonment of vil as and other forms of was destroyed after the Gothic raid in 396, a time when unfortified lowland settlements (Thomas 1964; Henning most public buildings were also devastated in Corinth. 1987, 35; Dintchev 1997; Mulvin 2002, 42; Milinković New wal s constructed in Dion and Thespiae in the late 2008, 545). This process occurred similarly as for the 4th century greatly decreased the size of the respective cities, i.e. first on the continent and with a slight delay cities. Up to the mid-5th century, Serdica was reduced in the coastal areas. On the other hand, rare new vil as to the extent it had before Constantine. A large part of (e.g. Höflein) and modest unfortified settlements (e.g. Nicopolis ad Istrum was destroyed around 447. Vučipolje) were still being built in remote areas. The inscription from the fort at Bargala relates that The process of weakening cities and abandoning it became a city at least in 371 and that it hosted both unfortified settlements is paralleled in the countryside soldiery and civilians. The renovation of Romuliana by a mass construction of different new fortified sites. dates to the last quarter of the 4th century, being densely These are particularity numerous in the most endan- occupied by soldiers and civilians up to the mid-5th gered areas in the last third of the 4th century, though it century. The housing in Stobi and Heraclea Lyncestis should be noted that the current state of research rarely already underwent great changes towards the end of the allows us to distinguish between military and civilian 4th century. Transformation of housing has also been de- sites; the multitude of such sites and the small finds tected in Philippi, where some urban vil as were turned recovered from them certainly indicates civilian use as into workshops in the second half of the 4th century. wel . Some were inhabited only occasional y and others Priscus reports of the Huns destroying Naissus permanently (Ciglenečki 1997). in 441 or 447, which is corroborated by archaeological The emergence of fortifications is observable across evidence. The sharp break in settlement that occurred a large part of the area between Constantinople and in Stobi in the mid-5th century has also been tied to the Ravenna; they comprise renovated earlier settlements Hun incursion of 447. and newly-established military posts such as the large In contrast with the examples above, the south- fort on the Sirmione Peninsula, the fort at Tokod dated ern parts of Illyricum and even more the coast of the to the Valentinian period and the Late Roman phase of Mediterranean show a greater continuity of residential the fort at Golemanovo Kale. This is also the peak for architecture. In the coastal city of Buthroton, archae- the forts of the Claustra system, which is corroborated ologists documented the expansion of an existing vil a by the recorded intensification of the monetary circula- with the addition of a peristyle and portici around tion in 364–378. Some forts in the Đerdap Gorge show 342 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS Fig. 4.6: Gradishte near Gabrovo. In the second half of the 4th century a fortified settlement was built (2012). refortification under Valentinian and even more so tural remain of the settlement from the late 5th century. under Theodosius I. For the fort at Iatrus (Krivina) in The first defensive wall at Rifnik was presumably erected Bulgaria, greatly intensified habitation has been noted around 400, a dating provided by the imported pottery in the last third of the 4th and first half of the 5th century. recovered in particularly great quantities along this wal . The construction of the fort at Dobri Dyal has been dated At nearby Ajdovski gradec, modest building remains to around 400. The defensive system in the Stara planina from the second half of the 4th century came to light in Mountains with the fort at Harmana belongs to the late the area of the baptistery. 4th or early 5th century, while the construction of the Small finds show that Kuzelin was most intensely Hexamilion Wall and the associated fortress in Greece occupied in the last quarter of the 4th and first half of has been set to the 410s. the 5th century. A lofty defensive wall was constructed There are also rare fortified settlements that per- in this time with timber buildings along its interior. The sist to the second half of the 5th century, for example same phase at the complex site of Gradina in Bakovci is Ajdovščina above Rodik and Tinje above Loka pri only indicated by well-preserved metal artefacts. Žusmu with a hybrid civilian and military nature, or For Serbia, many elements suggest a partial shift Kučar of a purely sacral function. Their interiors hold of civilian population to fortified sites in the late 4th or masonry and timber buildings. Phase II at Invillino first half of the 5th century (Milinković 2008, 540–545). has been dated from the middle/second half of the 4th Numerous forts are only dated with coins, which are not to the first half of the 5th century or, according to an a consistently reliable indicator on sites of long duration alternative interpretation, to the decades around 400. (problem of residuality). This problem is particularly At Castelraimondo, the fort was presumably militarily apparent at Gradina Ras, where a coin of Licinius has active to around 430, when it witnessed a fire and the led to the construction of the defensive wall being dated destruction of most of its buildings. to the early 4th century, whereas a large part of the finds Part of the settlement at Kappele is dated to the first from the fort belongs to the second half of the 4th cen- half of the 5th century, which is supported by numerous tury (Popović, M. 1999, 79). Fortification of Gradina metal finds from the second half of the 4th and the early in Vrsenice has been attributed to the last third of the 5th century. The earliest church at Hemmaberg was con- 4th century. The fort at Ukosa revealed small finds from structed around 400, which is also the dating post quem the 4th century and a hoard from the time of the Hun non for the fortification wall and the densely distributed incursions, the latter suggesting that the hill was used timber buildings in its interior. at least as a refuge in that period. Tonovcov grad appears to have been intensely in- Among the examples from Bulgaria, we should habited from the 370s to the 420s or 430s. Building traces particularly mention the extensively investigated fort from this period have been unearted under the architec- at Iatrus (Krivina), where the second habitation phase 343 Slavko CIGLENEČKI Fig. 4.7: Dyrrachium. The strong wal s were built during the time of Emperor Anastasius (2009). presumably already began in the 360s and intensified 4.3 TRANSFORMATION OF THE SETTLEMENT in the 380s, lasting to the devastation in the mid-5th PATTERNS AFTER THE MID-5TH CENTURY: century. The establishment of the forts at Dichin and LANDSCAPE OF FORTIFICATIONS Dobri Dyal has been set to around 400. The small AND DECLINING CITIES refuge at Gradishteto near Debrene was already forti- IN THE 6TH AND THE END OF fied in the second half of the 4th century. The refuge at LATE ANTIQUE SETTLEMENT Kartal Kale was briefly used between the late 4th and the IN THE FIRST THIRD OF THE 7TH CENTURY mid-5th century, after which it is believed to have been abandoned due to the Hun invasions. Gradishte near The middle and second half of the 5th century is a Gabrovo ( Fig. 4.6) was fortified and its interior built-up time least clear with regard to the settlement pattern. in the second half of the 4th century, but in contrast with The great destruction brought on by the Hun incur- other fortifications went on to be even more intensely sions resulted in a fateful break in settlement. This was occupied in the 5th and 6th centuries. particularly apparent along the major lines of communi- Most forts in North Macedonia were already con- cation that the Huns used in their progression through structed in the 4th and first half of the 5th century, and the Balkans and towards Italy, which signified similar temporarily abandoned in the middle of the 5th century consequences in both parts of the area under discussion. similarly as in Bulgaria (Mikulčić 2002, 84–87). Better- Alongside political, economic and social changes, the known examples include Markovo Kale near Malčište, Hun incursions are doubtlessly the main contributing Gradište near Pakoševo, Žegligovski kamen, Gradište factor in the radical change of the settlement pattern, as near Pčinja, Kale near Debrešte and Čebren near Zovik. they led after the mid-5th century to a complete destruc- The end of the Late Roman phase of the forts in the tion and abandonment of numerous cities or to a greatly Đerdap has been dated to the mid-5th century (presum- reduced continuation of others. ably to 441–443) and linked primarily to Hun incursions Major changes in the west occurred towards the (Vasić 1995, 44). A destruction in the mid-5th century end of the 5th century and slightly later in the east. is also well-recorded at Iatrus (Krivina), where a thick Archaeological evidence does not allow for a precise dat- layer of burnt debris covered a large part of the fort. ing of individual sites, even though the small finds include The fortified settlement at Gradishteto near Riben was slightly more numerous coins. We should first mention abandoned in the mid-5th century. The destruction of the radical changes in cities, for example the abandon- the fort at Odartsi has been dated to the late 4th or the ment of Diocletianopolis in Thessaly and Pydna in the late opening decades of the 5th century, presumably also as 5th century, and the construction of fortified settlements the result of Hun incursions. The abandonment of the with an urban character, such as Golemo Gradište near fort at Dobri Dyal has been set to the 430s. Konjuh and Louloudies in Greece, the latter hosting the 344 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS Fig. 4.8: Carevec. At the beginning of the 6th century the small settlement was changed to a city (2012). bishopric transferred from Pydna. Some cities witnessed were renovated in the second half of the 5th century. The a shift to natural y protected locations in proximity, for thick city wal s in Diocletianopolis (Hissar) and Philippi example at Tridentum, the inhabitants of which sought were added a proteichisma in the sections of easier ac- refuge on the hill Doss Trento, and Iulium Carnicum, cess. Literary sources and brick stamps suggest that the with its inhabitants fleeing to the nearby hill San Pietro. wal s of Dyrrachium ( Fig. 4.7) were constructed under In Noricum Ripense, population was evacuated Anastasius. Recent research has dated the renovation of from the endangered areas along the Danube in the the Hellenistic wal s of Buthroton to around 525. 480s, under the leadership of St Severinus (Ubl 1982; It is also a time when numerous changes in the Liebeschuetz 2001, 370, 378). The last period at Favi- housing and church architecture are observable in the anis, which fal s in the time of Severinus, only revealed city interiors. Housing is becoming increasingly modest modest habitation traces and abandonment after 488. and poorly built, while there is a proliferation of large After the devastation around 447, Nicopolis ad Early Christian churches, both inside cities and outside. Istrum became a greatly reduced city of a completely Under the Ostrogothic rule, several houses in Brixia and different layout. The incursion of the Ostrogoths in 482 Verona were renovated, in Verona the fortified palace of led to the people from the large city of Demetrias to flee Theoderic was constructed on a nearby hil . to the nearby hill of Iolkos. Modest huts were constructed in the second half of Reductions of city wal s took place in the late 5th the 5th and the 6th century in Sirmium, inside formerly and early 6th centuries, observable for example in Pharia lavish buildings. An isolated settlement of huts was on the island of Hvar, in Deultum, on the east side of erected inside the hippodrome. The last phase of the Heraclea Lyncestis and in Nicopolis in Epirus, with the semi-urban settlement in Gamzigrad dates from the great reduction in the last city dated to the time of the late 5th to the early 7th century. Emperor Anastasius. After a devastation brought by the Hun incursion, The second major indication of settlement changes the fort at Iatrus (Krivina) was reinhabited in the late 5th is the construction or reparation of city wal s. A trans- or early 6th century, while the smal settlement at Carevec verse wall was presumably built in this time in Aquileia, ( Fig. 4.8) was enlarged and transformed into a city. The which considerably reduced the city area. At the foot fort at Dichin was destroyed between 474 and 520, but of the Alps, Forum Iulii was fortified with double wal s renovated soon afterwards and continued to the late 6th (proteichisma) and pentagonal towers. In Verona, the century. city wall was renovated and integrated the amphitheatre Important new constructions in Heraclea Lyncestis under the Ostrogothic rule. include a large episcopal basilica, next to it a smaller The reconstruction of the city wal s in Pola began church and an episcopal palace, all enclosed with defen- after the Hun incursions, the city wal s in Mesembria sive wal s. The construction of the episcopal basilica in 345 Slavko CIGLENEČKI Fig. 4.9: Byllis. Eastern part of Victorinos‘ wal s with the entrance (2009). Bargala has been attributed to the late 5th century and The Justinian period also witnessed the reinforce- its renovation already in the early 6th century. ment of defensive wal s in numerous cities, for example in Tergeste, Salona, Serdica, Philippopolis, Scodra, Byllis The final part of Late Antiquity also brought the last and Heracleia (Perinthus). The new wal s in many cities peak of urban prosperity under the Emperor Justinian. brought a great reduction of the urban area. A typical Along the Mediterranean coasts and the eastern part, example is Byllis ( Fig. 4.9), where the construction work we can observe a renewed prosperity in some cities al- is well-documented with inscriptions. ready in the early 6th century, which became particularly The reinforcement of the city wal s with towers, apparent in the second third of the 6th century. There buttresses and a proteichisma in Asseria ( Fig. 4.10) has are three inscription that mention public construction been reliably dated, as one of the buttresses was built on work on the initiative of bishops in the time of Justin- top of a tomb from the first half of the 6th century. Also ian, more precisely in the late part of his reign. One is dating to the Justinian period is the minor renovation an inscription from Heraclea Lyncestis dated to 561, of the wal s of Varvaria. In Salona, the period brought relating that Bishop John had a beautiful y decorated a reinforcement of the wal s with a ditch and the addi- fountain built. The second inscription, from Serdica, tion of triangular terminals to the rectangular towers. speaks of a renovation of an aqueduct and the third Ancient texts reveal that the construction was done in one from Izbičanj even describes the construction of a haste, which is archaeological y visible in the numerous fortified vil a or settlement. amphorae used to fill the core of the city wal s. In the eastern part, two cities were established The cities of the time witnessed a proliferation and under Justinian, which bear his name. Iustiniana Prima, peak of Early Christian architecture. Numerous new identified at Caričin grad, is first mentioned in ancient churches were constructed, for example in Tergeste, texts in 535, hence its formation can be dated to the 530s, Salona, Histria and Caričin grad, while others in most which is corroborated by archaeological evidence. It active cities were renovated and embellished. The great saw numerous additions and renovations, and was only significance of sacral architecture is visible in the Justini- abandoned in the opening decades of the 7th century. anic church in Iatrus (Krivina), which was constructed In Ulpiana (Iustiniana Secunda), a new fortified city inside the former fort marked in the late sources as a city was erected next to a partial y abandoned earlier one. and greatly stands out in size and quality of construc- Also under Justinian, the fort of Iatrus was destroyed in tion from the very modest contemporary housing that a fire in the first half of the 6th century and afterwards surrounds it. renovated and mentioned in ancient sources as a city. In Most cities cease to exist in the late 6th or first half Thessaly, Justinian had the inhabitants of the devastated of the 7th century. In a time of gradual decline, cities Diocletianopolis move to the island in Lake Castoria. witnessed the final blow with the Avaro-Slav incur- 346 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS Fig. 4.10: Asseria. During Justinian‘s restoration, a wall buttress was built over the tomb from the first half of the 6th century (2004). sions (Popović, V. 1982, 547–548; Liebeschuetz 2001, Vučipolje near Dugopolje, Bisko, Ograja in Putovići, 284–291; Sodini 2007, 331–332). Dating to this period Trpčeva crkva, as well as some large vil as that either is the transfer of the metropolitan see from Aquileia to continued in a slightly altered form or were newly built Grado. Safety and connection with the maritime routes at this time, for example at Barbariga, Dragonera Jug, being of paramount importance is also visible in the case Vižula near Medulin, Mirje near Postire ( Fig. 4.11), Novo of Pola, which continued on the same spot, whereas selo Bunje, Polače on Mljet, Rankovići near Travnik, Nesactium, located in proximity but slightly removed Kruče near Ulcinj, Mirište in Petrovac, Diaporit and from the coast, was completely abandoned in the late 6th Akra Sophia. century! Salona shows no signs of violent destruction, but stray finds indicate it was gradual y abandoned in This is a time when fortifications came to dominate the first half of the 7th century while at the same time the area between Constantinople and Ravenna, in the the smaller, but well-fortified palace of Diocletian was western part already in the late 5th century and in the inhabited. The Avars took Sirmium already in 582. The eastern part slightly later, and went on to persist to the modest housing in Stobi reveals that the city persisted in end of Late Antiquity. Most of these sites in the western the second half of the 6th century, but was deserted even part very likely appeared in the 490s. In addition to the before the arrival of the Slavs. Heraclea Lyncestis was diagnostic finds from the settlements, this dating is cor- inhabited longer and includes among the late buildings roborated by the associated cemeteries that revealed no in the city centre some that were presumably still oc- elements predating the late 5th century (e.g. cemeteries cupied in the Middle Ages. A similar fate befell cities in at Hemmaberg, Rifnik, Kranj). Greece, most of which were abandoned between the last Two forts were constructed on the Sirmione quarter of the 6th and first quarter of the 7th century, and Peninsula in the second half of the 5th century. The only rare ones exhibit reliable signs of continuity into creation of the forts at San Martino di Lundo/Lomaso the 7th century; an example of the latter is the excellently and Sant’Andrea di Loppio in the Alpine area has been fortified centre of Thessaloniki (Sodini 1984, 393–396). attributed to the second half or the late 5th century. At In the far south, Gortyn on Crete was renovated Hemmaberg, two large complexes of double churches around 627, when an aqueduct was also restored. It was were erected in the early 6th century. The densely built- only the earthquake around 670 that turned the city into up settlement at Kappele has been dated to the late 5th a small settlement. and the 6th century. In the last third of the 5th and the 6th century, there The most intensive habitation on the well-re- was merely a handful of unfortified settlements in the searched fortified settlements at Ajdovski gradec above countryside, for instance Castel Antico, Castelletto di Vranje, Rifnik ( Fig. 4.12), Ajdna, Tonovcov grad and Ko- Brenzone, Rim near Roč, St Chrysogonus in Glavotok, rinjski hrib already began in the late 5th century, which 347 Slavko CIGLENEČKI Fig. 4.11: Mirje near Postire. Remains of Late Antique vil a from the north-east (2021). is corroborated by numerous small finds. The Early believed to have been rebuilt after the Hun incursions, Christian complex at Marija Gorska near Lobor has probably in the second half of the 5th century, Krasen been broadly dated to the 5th and 6th centuries, though Kale in the late 5th or the first half of the 6th century. The the published finds do not allow us to more precisely second phase of the fort at Harmanovo, which formed define the beginning of this phase. part of barrier wal s, is dated from the late 5th to the late Small finds date the large and important fortified 6th/early 7th century. settlement at Gradina in Bakinci from the late 5th to the early 7th century. This dating is supported by the Similarly as in cities, the countryside also reveals church architecture next to the settlement with several the establishment or renovation of numerous fortified construction phases. The impressive vil a at Mogorjelo settlements under Justinian. In the western part, there transformed into a fortified vil age in the second half are many changes and additions in the already existing of the 5th century. fortifications. That at Duel, for example, was renovated. In Serbia, there are no known fortified sites The churches at Tonovcov grad show a major renovation from the middle and second half of the 5th century of the interior, which has been reliably dated with a coin (Milinković 2008, 545). Even along the limes in the of Justinian inside an amphora found under a newly- Đerdap, research shows a long habitation hiatus in the built altar. At the same time renovations took place at second half of the 5th and first two decades of the 6th Ajdovski gradec above Vranje. The Justinian period at century following the Hun devastation (Vasić 1995, Rifnik and Korinjski hrib brought the construction of 44). The three coin hoards from the mid-5th century, masonry towers and the renovation of both churches. found in the hil top sites at Sv. Petka on Veliki Jastrebac, There is a series of forts along the eastern coast of Ukosa and Jerinin grad in Trubarevo, are therefore al the Adriatic, both on the islands and on the mainland, the more important (Rašković 2021, 283–284); the the geographic distribution and fortification features fortifications may have been used as refuges at this time, of which testify to a major project of protecting the which would indicate an at least occasional occupation navigable routes in the time of Justinian’s reconquista. in the middle of the 5th century. This dating has been corroborated by the excavations at The major fort of the regular army at Markova Me- Gradina on Žirje, but also the numerous surface finds hana is dated to the reign of Anastasius. The extensive and small-scale excavations in several other forts, for renovation of the fortified settlement at Mezdra and the example at Veliki Sikavac, Korintija and Gradina near construction of the fort at Madara date to the late 5th or Modrić. early 6th century. The fortified settlement at Odartsi is 348 4. DIACHRONIC ASSESSMENT OF THE SETTLEMENT CHANGES AND OF THE CITY–COUNTRYSIDE DYNAMICS Fig. 4.12: Rifnik. Late Antique buildings in the most habitable part of the settlement (2023). Most of the newly-established fortifications in fortified settlement at Qafa to the middle and second Serbia have been attributed to the reign of Justinian and half of the 6th century. associated with his fortification programme described In the publications of surface surveys, several for- by Procopius (Milinković 2008, 545–557; id. 2015, 258– tification have been identified as either Early Byzantine 263). It is a period that saw the renovation of numerous (e.g. Belgrad, Velika), dated to the 6th century (e.g. Gra- forts along the limes, such as Veliki Gradac and Boljetin, dok near Čanište) or more precisely as Justinianic (e.g. or the construction of new ones (Bosman). Their reno- Gradina Žirje). Most of the small finds thus recovered, vation dates between 529 and 540 (Vasić 1995, 44). The however, are poorly diagnostic and therefore require fortified hil top settlements of this time, identified for caution. We should nevertheless note the fact that the example at Liška Ćava, Bedem near Maskare, Kula in reliably dated artefacts mainly belong to the Justinian Kaludra, Gaj in Babrež, Balajnac, Kale in Bregovina and period. Kale, Zlata, mainly had a single phase of occupation and In spite of the vastness and diversity of the area were established in the 530s. Gradina Ras and Gradina between Ravenna and Constantinople, the available in Vrsenice were renovated at this time. evidence shows a similarly dated end of the fortified The dense occupation of the interior of the forti- settlements in the countryside, which occurred in the fication at Golemanovo Kale has been attributed to the late 6th century (e.g. Rifnik, Ajdovski gradec above time around 540, when the nearby fort at Sadovsko Kale Vranje, Korinjski hrib, Kappele, Mezdra, Golemanovo was also presumably built. The first half of the 6th cen- Kale, Sadovsko Kale, Dichin, forts along the limes tury is the time when the forts at Dyadovo, Kaleto near in the Đerdap Gorge) or the early 7th century (e.g. Izvorovo, Đuteza, Kekola and Domaj were constructed, Tonovcov grad and Odartsi). A vast majority of these and when many earlier forts such as Markova Mehana, settlements were permanently abandoned and only a Odartsi, the fortress at the Hexamilion and Markovi handful reinhabited in the late 8th or early 9th century Kuli on Vodno were renovated. Small finds date the (cf. Milavec 2012; ead. 2020, 162). 349 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE This chapter outlines the settlement patterns in The process of abandonment in the westernmost other parts of the Empire and does that – for the pur- part of the Empire, i.e. Britain, in many areas along the poses of facilitating comparison – in an order similar Rhine limes or Aquitania, but also individual examples as above in terms of form (cities, countryside) and on the Iberian Peninsula began in the late 4th century geography (West, East). The outline provides a general (Liebeschuetz 2001, 88). The large Roman cities in Spain, description of the settlement patterns il ustrated with such as Taraco, Italica, Clunia, were in decline at this typical examples of sites. time, but still inhabited, albeit in a limited measure. In the 5th and 6th centuries, the southern provinces of Spain and Gaul exhibit a greater level of urbanisation 5. 1 WESTERN PART OF THE EMPIRE compared with the north (Liebeschuetz 2001, 90). Cities with continuity are much more numerous 5.1.1 CITIES in the West than in the area between Ravenna and Constantinople, and closely reflect the changes in the The main phases of development and decline of urban structure. Numbers notwithstanding, most of the cities in the West are clearly discernible despite the them underwent a transformation that is similar in its dense settlement of the majority of them also after Late basic traits to that presented in the previous chapters. Antiquity. There are marked regional differences, but Many cities in the Gallic provinces were walled in they show a general y similar transformation (overview the 3rd and 4th centuries, which often reduced their size. in Ward-Perkins 2000; Liebeschuetz 2001, 82–103; Concurrently outside the city wal s, suburbs gradual y Wickham 2005, 635–674; Quiroga 2016). Fig. 5.1: León. Well-preserved northwestern section of the Late Roman city wal s (2011). 351 Slavko CIGLENEČKI Fig. 5.2: Conimbriga. The late Antique city wal s were constructed on top of earlier buildings (2011). Fig. 5.3: Haïdra. A large fortification was constructed in the middle of the Roman city of Ammaedara, on the natural y best protected spot (2001). developed around cemeterial churches and monasteries capitolium were fortified at Dougga and Sbeitla. At (Johnson 1983, 82–135; Liebeschuetz 2001, 84; Brogiolo Haïdra (Ammaedara) ( Fig. 5.3), a mighty fortification 2014). Many cities in Spain were also fortified at this was built in the centre of the city that took advantage of time, with excellently preserved and extensive wal s a protected location on a gentle slope and held the most surviving at Lugo, León ( Fig. 5.1) and Astorga. The prominent buildings that included a church (Pringle wal s in Conimbriga in Portugal ( Fig. 5.2) cut through 1981, 179–181). the previously inhabited part of the city and radical y In general, people still lived in the cities and only reduced the populated area (Liebeschuetz 2001, 90). sought refuge behind defensive wal s protected by smal Interesting forms of urban fortifications can also garrisons in times of danger. be found in Byzantine Africa (Pringle 1981). Hastily The interior of western cities reveals buildings of built defensive wal s only encircled small parts of cit- lower quality, subdivisioning of earlier constructions ies that came to function more as forts or refuges than and encroachment on public spaces, leading to a spatial as a proper cities. There are also clever adaptations of fragmentation of the urban fabric and a reduction of earlier monumental buildings that were incorporated the population (Wickham 2005, 669; Kulikowski 2010). into new fortified complexes. Only the forum and the This is strikingly similar to the development in the ar- 352 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.4: Trier. Porta Nigra, the lofty north entrance (2022). Fig. 5.5: Merida. Remains of the Roman buildings and a theatre (2011). eas between Ravenna and Constantinople, sharing the incursions, when the Gallic prefecture was transfered dividing line of the mid-5th century that marks a severe to Arles. (In late 4th century, the imperial capital as well break in urban continuity. Also similar is the appearance was transfered to Milan). The Roman administration of numerous churches inside and outside the city wal s. of the city of few inhabitants lasted to the early second Only the cities that went on to become the capitals half of the 5th century, after which the city came under of Germanic kings recovered again towards the end of the Frankish rule. Antiquity (e.g. Metz) (Liebeschuetz 2001, 86). Trier Cities in Provence retained their Roman character ( Fig. 5.4) became the imperial residence in the late 3rd longer. For most of the 5th and the first half of the 6th century and reached the pinnacle of its development century, Arles remained a large city (Heijmans 2004). Its in the 4th century. It was a vast, 285 hectares large city cathedral was created in the 4th, the city wal s renovated enclosed with thick wal s built in the late 2nd century. The in the 5th century. Archaeological remains become scarce development of this major administrative and economic after 550, the number of inhabitants decreased, though centre was halted in the early 5th century with Germanic the city still hosted a bishop and administrative facilities. 353 Slavko CIGLENEČKI Simple buildings were constructed in the 5th century along the exterior of the circus, which may have still been in use, and were deserted in the mid-6th century. Other simple dwellings were constructed inside the abandoned Forum of Augustus and in the abandoned houses; the large two-storey palatial building was still in use. Life continued in the walled classical part of the city and outside it. Lyon was spatial y fragmented in the 6th century, with separate habitation cores inside the classical Roman city. One of these cores on the right bank of the River Saône was intensely inhabited until around 600 and then less intensely from the 7th to the 9th century (Wickham 2005, 665–666). Merida ( Fig. 5.5), as the most important city on the Iberian Peninsula, was densely populated in Late Antiq- uity, with the forum presumably still active in the first half of the 5th century. Very early on, a large Christian complex developed around the sanctuary of St Eulalia that became a popular pilgrimage destination (Kulikowski 2004, 290–293). In the second half of the 5th century, the large urban domus began to be subdivisioned. They also exhibit an intriguing transformation into collective hous- ing units with poorly built rooms surrounding a central court, which is a specific form of reusing lavish earlier buildings (Quiroga 2016, 81). Late Antiquity in Tarragona brought a noticeable shift of habitation to a slightly higher part of the city where the former monumental sacral and administrative complex was replaced by different newly-built dwellings. As in Merida, extramural churches Fig. 5.6: Saint-Bertrand-de-Comminges. Remains of the Ro- formed small habitation cores. man city in the foreground and the hill with a Late Antique A typical example of Late Antique urban transfor- settlement in the background (2011). mation can be found at Saint-Bertrand-de-Comminges Fig. 5.7: Saint-Bertrand-de-Comminges. Western defensive wall of the hil top settlement (2011). 354 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.8: Sbeitla (Sufetula). Well-preserved fortified house in the foreground and an oil press constructed on the street further along (2001). Fig. 5.9: Tipassa. House of the Frescoes (2018). ( Figs. 5.6, 5.7), where habitation area began shrinking in and Djemila (Cuicul) show several phases of renovation the 5th century, when people were moving from the lower (Lepelley 1992, 52). In Africa as wel , urban domus were city to the natural y well-protected and additional y forti- subdivided into smaller units; a good example is the fied elevation above (Wood 2002; Wickham 2005, 667). House of the Frescoes in Tipassa ( Fig. 5.9), which shows Different elements show a lesser degree of urban the characteristic subdivisioning of the previously grand ruralisation in different parts of the West, particularly urban building (Ellis 2000, 111). in Spain (Quiroga 2016, 83). Similarly as in the Balkans, Sbeitla shows an additional feature, namely the however, Carthage and Valencia did reveal modest construction of heavily fortified houses – towers in dwellings set up in the protected shel s of theatres. the city centre, which broadly date to the 5th and 6th Cities in North Africa displayed the classical urban centuries (Pringle 1981, 142, 284–285). In Dougga, a features longer. Great prosperity is observable to the small fort was constructed under Justinian on the spot early 5th century, though the numerous mosaics from of the forum and capitolium, and the city was enclosed this time show that the quality of urban living remained with a Late Antique wall of poorer quality. The city was high even in the Vandal period. The urban structure was additional y protected with several forts in immediate gradual y changing, with dwel ings and economic facili- vicinity (Pringle 1981, 244–246). ties encroaching on public spaces (Lepelley 1992; Potter Many large settlements, such as Albi, Orleans 1995; Liebeschuetz 2001, 74–76, 99). The disintegration and Auxerre, were given the status of civitates in Late of the basic urban elements occurred from the late 6th Antiquity and as such welcomed a bishop (Liebes- century onwards. In Sbeitla (Sufetula) ( Fig. 5.8), for ex- chuetz 2001, 88). These new cities or episcopal sees are ample, an oil press was set up in the middle of the main comparable with some of the newly-founded cities in street (Potter 1995, 67). Aristocratic houses in Carthage the eastern Alpine area and the Balkans. In Spain, the 355 Slavko CIGLENEČKI Fig. 5.10: Tolmo de Minateda. Hill with the remains of an episcopal centre on top (2011). Fig. 5.11: Reccopolis. Re- mains of a palace in the fortified Visigoth royal city (2011). fortified hil top settlement with an ecclesiastical centre presented in the chapters above (cf. Chavarría, Lewit at Tolmo de Minateda ( Fig. 5.10) has been identified 2004; Brogiolo, Chavarría Arnau 2014, 233). as Elo, an episcopal see mentioned in literary sources It is commonly believed that the Roman villas (Quiroga 2016, 76). in the West represented the most important and best A frequently mentioned example is Reccopolis ( Fig. known segment of the countryside settlement (Van Os- 5.11), which was a city newly-founded in the late 6th sel, Ouzoulias 2000; Liebeschuetz 2001, 96; Wickham century (Quiroga 2016, 85–88). It is primarily marked 2005, 465–495; Chavarría Arnau 2007; Teichner 2011). by its natural y protected location, fortifications wal , Between the late 3rd and the late 5th century, some were church on a prominent spot and numerous, but simple abandoned and others lost their original function with masonry dwellings, all of which are features similar to the the instal ation of agrarian facilities, storerooms and newly-founded cities in the Balkans; the only difference workshops, which gave them a ful y economic function. is a presumed royal palace at the edge of the settlement. Only rare ones continued as residential vil as beyond the 5th century (e.g. S. Giovanni di Ruoti in southern Italy). Modest dwellings were constructed in villas 5.1.2 THE COUNTRYSIDE in the 5th and 6th centuries, some were turned into churches or cemeteries. Churches next to vil as were Countryside settlement in the West shows similar predominantly built after the change of function. It is patterns regarding the unfortified settlements as the one frequently difficult to distinguish between continuity 356 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.12: Torre de Palma. Presented remains of a Ro- man vil a (2011). Fig. 5.13: Torre de Palma. Remains of a church in im- mediate proximity to the vil a (2011). Fig. 5.14: Sao Cucufate. Wel -preserved remains of a Late Antique vil a (2011). of a vil a or its reoccupation with a different function. The large vil a at Torre de Palma ( Figs. 5.12, 5.13) In spite of their transformation, the utilitarian or cult was inhabited up to the first half of the 5th century, use of the vil as suggests that people continued to farm when the complex transformed and was subdivided into the adjacent land. smaller residential units, persisting in this form into the 357 Slavko CIGLENEČKI Fig. 5.15: Timgad. Early Byzantine fort from the 6th century, located at the edge of the Roman city (2018). 7th century. A church was constructed next to it, presum- and work buildings (Chavarría, Lewit 2004, 35–36). In ably already in the 4th century, the area around which some places, they are already observable in the 3rd and was used as a cemetery from the 5th to the 7th century 4th centuries, but become more numerous in the 5th and (Maloney, Hale 1996; Chavarría Arnau 2007, 269). The 6th centuries. They are typical y of a shorter duration and excellently preserved vil as, such as those at Centcelles frequently shifting location. In a large measure, they are and Sao Cucufate ( Fig. 5.14), were used as residential similar to the settlements of simple wooden buildings in vil as to the mid-5th century, later subdivided on several the discussed part of the Balkans and the eastern Alps, occasions and used as production units to the 6th cen- with the difference that the latter do not reach beyond tury. On the Mediterranean coast in Tunisia, the vil a of the late 4th century (see Chapter 3). Nador was already subdivided to create accommodation The reasons behind the transformation of the clas- for several families and economic facilities in the early sical rural vil as and the location the inhabitants from 5th century (Potter 1995, 78–79). the unfortified settlements moved to are not entirely It is not possible to be certain that villas were clear. We can exclude the possibility of moving to cities, walled for defensive purposes (Van Ossel 1992, 164–165; which in the 5th and 6th centuries show no signs of major Chavarría Arnau 2007, 104–108). In Africa, fortified vil- influx of people. To a limited extent, they may have cho- las have been identified in mosaics depictions, but they sen fortified hil top locations, as indicated by examples served more as status symbols, used by the owners to from Italy, south-eastern France and partly also Spain control the works in the surrounding area and defend (Randsborg 1991, 56–64, 71–72; Liebeschuetz 2001, 96). them against local raids and roberrers.(McKay 1998, Until recently, fortified hil top settlements in the 176; Hirschfeld 2001, 264–266). Mosaic representations West were poorly known and only rarely considered in also show simple farms in the vicinity of vil as. the overviews of countryside settlement patterns. They Archaeological research is providing increasing were presented as a local phenomenon and predomi- evidence that the lavish vil as mentioned, for example, nantly interpreted as refuges or military posts (cf. Uslar by Sidonius Apollinaris and Venantius Fortunatus in 1964, 16–24; Petrikovits 1971, 187–189, 192–193; Senn- the 5th and 6th centuries are more likely a form of poetic hauser 1979, 152–156; Overbeck 1982; Johnson 1983, exaggeration. This discrepancy has been noticed early 226–244). The scholarly problems they pose, however, on, for example by Rudolf Egger who observed that the are similar to those in the area discussed in the previ- palace at Mosel (episcopal vil a) as described in a poem ous chapters, particularly with regard to distinguishing by Venantius Fortunatus actual y had traits recognisable between military and civilian sites. in the modest fortified ecclesiastical centres of the 6th The typical fortifications are the military posts century in the eastern Alps (Egger 1940, 87–94). dating from the 3rd to the early 5th century, which are From the 5th century onwards, there is another often also better investigated (overview in Petrikovits form of settlements newly-established in the lowland, 1971; Johnson 1983, 136–225). Later ones, particularly which persisted into the Early Middle Ages. These are Justinian period forts, came to light in Algeria and Tu- small compact groups or hamlets of very simple wooden nisia (cf. Pringle 1981). An excellent example of a large dwellings or sunken huts with associated agricultural Early Byzantine fort has been investigated at Timgad 358 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.16: Limysa. Excellently preserved quadriburgium (2001). Fig. 5.17: Roc de Pampelune. Plan of the fortified set- tlement (Schneider 2007, Fig. 7). ( Fig. 5.15), where strong perimeter defences were cou- from the last third of the 3rd to the early 5th century. pled with army barracks, a water cistern and a church in Investigations succeeded in distinguishing refuges, forti- the interior (Lassus 1981). A smal , but better preserved fied settlements and military forts (Gilles 1985). Such is the quadriburgium at Lymisa ( Fig. 5.16), for which sites, albeit in smal er numbers, have also been identified size and number of soldiers correspond with that of the in Switzerland and westernmost Italy (overview in Steuer smaller forts in the Đerdap and on the Adriatic islands. et al. 2008; Cavada, Zagermann 2020a). Civilian and civilian-military hil top sites are well- In France, only few fortified settlements with smal investigated in Germany, in the areas of Eifel and Hun- finds from the 5th and 6th centuries were known for some srück, which were created and used in different phases time. New research, particularly in the south-eastern 359 Slavko CIGLENEČKI Fig. 5.18: Roc de Pampelune. Hilltop settlement, view from the south (2011). Fig. 5.19: Larina. Late An- tique buildings in the hil top settlement (2011). Fig. 5.20: St. Blaise. Remains of the partly excavated de- fensive wal s (2011). 360 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.21: St. Blaise. Late Antique buildings on the hil top (2011). part, has revealed many new sites (overview in Schneider 2001; id. 2007; Constant et al. 2015; Kasprzyk, Monteil 2017). One is the relatively extensively investigated site at Le Roc de Pampelune near Argelliers (dep. Hérault) ( Figs. 5.17, 5.18), which displays features characteristic of fortified settlements (Schneider 2003). It was con- structed on a natural y well-protected hil , enclosed with a defensive wall reinforced with towers. A church stood on the highest spot, while simple two- or three-room buildings are visible elsewhere, particularly densely in the central part of the plateau around a large empty space. The numerous farming and other tools indicate a largely self-sufficient settlement of civilian population. Fortified sites further include large Late Antique settlements that have the characteristics of centres, such as Larina ( Fig. 5.19) and Saint Blaise ( Figs. 5.20, 5.21) (Schneider 2001, 436–439; Porte 2011). In many cases, these were set up in earlier hillforts. Standing out is the oppidum at Constantine in Provence, with dwellings and a church inside a hillfort, where the prehistoric rampart was mortared in Late Antiquity (Duperron 2013). An exceptional example is the natural y protected refuge of Theopolis near Sisteron ( Fig. 5.22). An in- scription reveals that a private individual by the name of Thomas offered the high mountain valley to general use, though archaeological evidence of this only consists of this inscription, which is carved into a rock near the Fig. 5.22: Sisteron. Entrance to the mountain valley with an entrance to the valley (Johnson 1983, 242; Ripoll Arce inscription cut into the rock (2011). 2000, 97). Recent investigations have also provided more information on fortified sites in Spain (Catalán et al. 2014; Sánchez-Pardo, Galbán Malagón 2015; Fernán- dez Pereiro, Sánchez-Pardo 2022). Two such sites have been partial y investigated in the area conquered by the Byzantines, namely Cerro de Montroy ( Fig. 5.23) and El Monastil ( Fig. 5.24) (Sánchez 2009, 192−194, 246−248, 421, 463−465). 361 Slavko CIGLENEČKI Fig. 5.23: Cerro de Montroy. Defensive tower of the Late Antique fort (2011). Fig. 5.24: El Monastil. Ves- tiges of the Early Byzantine fortified settlement (2011). 5.2 EASTERN PART OF THE EMPIRE ter known than in the West (overview in Liebeschuetz 2001, 30–74; Morisson, Sodini 2002, 184–193; Butcher The settlement patterns in the East differ more 2003; Bagnal , Rathbone 2004; Wickham 2005, 609–635; substantial y from those observed in the wider area of Saradi 2006; Quiroga 2016; Zanini 2016; Rizos 2017b, the Balkan Peninsula. To name only two major differ- 20–24, 32–36). There was a continuous development ences, cities display more prominent continuity that is across a large part in the 6th century and many survived clearly mirrored in the archaeological record, while in through the fateful 7th century, as the arrival of Islam the countryside we can see more numerous and widely did not signify a total urban break. Cities prospered in distributed unfortified settlements that peak in the 5th Late Antiquity and the homes of the urban aristocracy and 6th centuries. frequently continued to develop. The presence of com- mercial buildings points to a lively trading activity. There are also a number of literary sources that tell of a busy 5.2.1 CITIES urban life; they mention the use of baths throughout the 6th century (Kennedy 1985, 8–10). Many eastern cities ceased to be inhabited after the The disintegration of cities in the East began in the end of Late Antiquity, hence their development is bet- 6th century. Similarly as in other parts of the Empire, 362 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.25: Ephesus. Remains of Roman buildings in the upper part of the city (2002). Fig. 5.26: Ephesus. The lower part of the city with the Ar- cadiane, a thoroughfare that led to the harbour (2002). this is observable in the abandonment of large public Ephesus ( Figs. 5.25–5.29) was among the key cities buildings and homes of the elite, as well as the appear- in the East, a metropolis, emporium and major Chris- ance of modest dwellings and workshops. Architecture tian centre (Liebeschuetz 2001, 32–36, 49; Ladstätter of an inferior quality encroached on streets and roads, 2011; Pülz 2011; Ladstätter 2017). It transformed after groups of burials appeared in cities, the sewage and water the political changes and earthquakes in the 3rd and 4th supply systems were neglected, refuse was piling up in centuries. Many buildings altered their function and the streets (Morisson, Sodini 2002, 189). Shops and dif- whole insulae were either abandoned or transformed. ferent workshops were set up along the main porticated Dwellings were set up inside public buildings and the streets, forming a linear marketplace as the beginning of city centre shifted from the upper part to the harbour the Oriental souq. Dense habitation and urban life can area. The city only recovered in the mid-4th and wit- be traced to the end of the 6th century. nessed great prosperity in the late 4th century. Different Concurrently with strong continuity and qualita- administrative buildings and housing complexes devel- tive transformation of housing, there was a large-scale oped next to the harbour, with shops nearby. The city construction of church architecture, which reached its wal s, although not yet precisely dated, only protected peak in the 6th century, producing a series of splendidly the main, greatly reduced lower part of the city that was decorated churches (cf. Krautheimer 1986, 201–282). densely inhabited in the 6th century. An array of shops 363 Slavko CIGLENEČKI Fig. 5.27: Ephesus. Remains of an important Early Byzan- tine building (2002). Fig. 5.28: Ephesus. The church of Mary (2002). Fig. 5.29: Ephesus. A large pilgrimage centre formed in the 6th century next to the church of St John (2002). 364 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.30: Sardis. Densely spaced Late Antique shops in the portici of the gymnasium (2002). Fig. 5.31: Palmyra. View of the city from the southeast. The Diocletians’s fortress on the right side (2004). and workshops lined the main streets. The suburbs were above. Numerous small finds from these shops indicate less densely inhabited, though the church buildings in lively trading that continued to the early 7th century. them became an obligatory stop on the Early Christian Palmyra ( Fig. 5.31), a desert city on a major caravan pilgrimage itinerary. In addition to important public route, witnessed great changes in Late Antiquity (Rizos and private buildings, there is a recent discovery of a 2017b, 20–21; Intagliata 2017). Under Diocletian, this ‘Byzantine palace’ with reception rooms, a bath complex, lowland city in an oasis was added a legionary fortress on oil production facilities and a large garden. Life in the the natural y better protected slope of the adjacent hil , city partly continued even in the 7th century. which exhibits both the characteristic change in the loca- Sardis ( Fig. 5.30) was a major provincial centre tion of Late Antique military forts and the first example (Liebeschuetz 2001, 43–44, 48–49; Rautman 2017). It of the marriage between a city and a legionary fortress. only saw few investigations, but these include a well- The provincial capital of Apamea ( Fig. 5.32), with researched bath complex incorporating a gymnasium its monumental colonnades, sumptuous houses and and synagogue. The covered portici of the gymnasium churches, reflects the splendour of the Late Antique city, received numerous shops and workshops in the 5th and with numerous domus forming its centre. It suffered two 6th centuries, known as the ‘Byzantine Shops’ (Stephens earthquakes in the first half of the 6th century and was Crawford 1990). They were built in two storeys with both times renovated. It only began to change radical y shops on the ground floor and dwellings of the owners in the second quarter of the 7th century with the formerly 365 Slavko CIGLENEČKI Fig. 5.32: Apamea. Impres- sive colonnades along the cardo maximus (2004). Fig. 5.33: Bosra. The excel- lently preserved Roman theatre was later added de- fensive towers (2004). lavish buildings subdivisioned into ruralised dwellings The defensive architecture shows considerable (Liebeschuetz 2001, 56). regional differences. Cities in western and central Asia Inscriptions and well-preserved architecture of Minor were largely unfortified. Some were walled in Bosra and Gerasa confirm intense construction activi- the 3rd and 4th centuries, presumably more as a display ties even in the first half of the 6th century (Liebeschuetz of urban confidence than necessity (Niewöhner 2007, 2001, 59–62). Bosra ( Fig. 5.33) was the capital of the 122–135; Jacobs 2012, 117–125). In the 5th and 6th cen- province Arabia and a major military and trade centre turies, wal s were only built exceptional y. The mighty in the area on the border with Persia. Investigations city wall of Nicaea ( Fig. 5.35) was already built under here mainly unearthed important military and church the Emperor Claudius II in 268/269 and later reinforced buildings. A special feature is a theatre subsequently with numerous towers (Rizos 2017b, 32–33). The city reinforced with towers and turned into a fortress. In of Sagalassos ( Fig. 5.36) was walled soon after 400, the excellently preserved Gerasa ( Fig. 5.34), changes in though the wal s only enclosed one third of the whole the urban fabric only become visible in the second half urban area and soon fell into disrepair (Waelkens 2005; of the 6th century. At this time, public buildings were id. 2019, 17–18); its construction is believed to be the abandoned, modest housing erected and ruralisation consequence of Isaurian raids, but also the expression was increasingly present. In spite of these marks of de- of civic pride. cline, there was intensive construction of churches even Because of a permanent Persian threat, cities in the in the second half of the 6th century (Brenk 2003, 86–91). Levant were well fortified and militarised, which is vis- 366 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.34: Gerasa. Elaborate entrance to the cathedral complex (2008). Fig. 5.35: Nicaea. Thick city wal s from the early 4th century (2003). Fig. 5.36: Sagalassos. View of the city from the southeast (2011). 367 Slavko CIGLENEČKI Fig. 5.37: Amida (Diyarba- kir). Mighty wal s of the city, view from the south (2003). Fig. 5.38: Amida (Diyarba- kir). A section of the well- preserved city wal s with an entrance (2003). ible at Antioch, Palmyra, Sergiopolis (Resafa), Zenobia Ammianus Marcellinus wrote that it permanently hosted (Halebiye), Dara and elsewhere. Procopius reports on one legion and had the capacity to accommodate four many renovations under Justinian, for example the resto- additional legions in the case of military operations. Its ration of Antioch following the Persian raid in 540, when lofty defensive wall could also provide shelter for the all the urban features of the city were re-established, but inhabitants from the surrounding areas. also the renovation of the pilgrimage city of Sergiopolis. Justinian had the fortress of the Queen Zenobia on The Late Antique wal s of Miletus, Sardis, Ephesus and the Euphrates transformed into a heavily fortified border Pergamon are not completely reliably dated, though city of Zenobia (Halebiye) ( Fig. 5.39; Blétry 2017). It has most likely belong to the 6th century, and show consider- a characteristic location on a natural y protected slope able reductions of the urban areas (Liebeschuetz 2001, between the river and the top of the hill and is marked 49–52). In Miletus, for example, the ‘Justinianic’ wall by a strong defensive wall with densely spaced towers, a only encircled one quarter of the Roman city. separately fortified citadel on the hil top, an excellently Many cities were established in Late Antiquity preserved praetorium and the interior only hosting two primarily for military purposes (Rizos 2017b, 20–24, churches, a bath complex and housing. 32–36). Amida (Diyarbakir) ( Figs. 5.37, 5.38) became a Sergiopolis (Resafa) ( Figs. 5.40−5.42) is an unusual city-fortress at least in the mid-4th century, though pos- example combining a pilgrimage centre, city, military sibly already under the Tetrarchs (Assénat, Pérez 2017). fortress and trading post on a caravan route, boasting 368 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.39: Zenobia. Heavily fortified Early Byzantine city on the slope above the River Euphrates (2004). Fig. 5.40: Sergiopolis (Re- safa). Vast city with strong wal s and a pilgrimage centre (2004). Fig. 5.41: Sergiopolis (Re- safa). Ornate facade of the north entrance to the city (2004). 369 Slavko CIGLENEČKI Fig. 5.42: Sergiopolis (Re- safa). Interior of the city’s largest church (Basilica A) (2004). Fig. 5.43: Arif. Remains of buildings along the defen- sive wall (2011). Fig. 5.44: Arif. Early Chris- tian church (2011). 370 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.45: Ören Tepe. Hill with the ruins of a small Late Antique city (2011). Fig. 5.46: Umm el-Jimal. Panoramic view of the vast settlement (2008). a thick defensive wal , splendid church complex, large newly-established cities in the Balkans, with its defensive cistern and residential buildings (Ulbert 1993). The wall with towers and the interior only hosting simple 1.8 km long city wall was already built under Anasta- dwellings and three churches, but completely devoid of sius and subsequently reinforced, particularly in the the urban features of a classical city and particularly of Justinian period (Hof 2016; Gussone, Sack 2017). It public buildings. It would appear that a small portion of undoubtedly had an ostentatious character, but also a the population still lived in the old city, from where they great defensive significance in the exposed area near the could flee to the natural y protected rocky acropolis, high border with Persia. above the city, in times of danger. Some cities in Asia Minor display quite radical re- A similar shift occurred at Panemoteichos in structuring that actual y resulted in the construction of Pisidia, where the city was transferred to the natural y new, smaller, but natural y better protected cities. An out- protected hill of Ören Tepe ( Fig. 5.45) in the vicinity standing example is Arykanda in Lycia, where the classical (Saradi 2006, 390). The ruins on this hill reveal a com- city was deserted in Late Antiquity and the inhabitants pact series of buildings lining the defensive wal , while moved to a smaller ( c. 250 x 100 m) and natural y much the interior only holds a single large and partitioned better protected location at Arif ( Figs. 5.43, 5.44), some residential building with a church on the summit. 2 km away (Knoblauch, Witschel 1993, 229–262; Saradi An example of a newly-founded small unforti- 2006, 469). The latter is in many aspects similar to the fied city from the 6th century was found at Viran ş ehir, 371 Slavko CIGLENEČKI Fig. 5.47: Umm el-Jimal. Remains of Late Antique architecture (2008). which has been identified as Mokisos, established under wal s, eleven churches, a military fort and public baths. Justinian. It was built in a remote area of Asia Minor, The dating of the inscriptions indicates a peak of the below Mt Hasan Dağı at 1500 m asl (Berger 2017). It settlement in the 5th and 6th centuries. One inscription holds dispersed simple dwellings, while only churches relates that a wealthy inhabitant named Thomas had a are mentioned as public buildings. walled kastron constructed in its centre in 558–559, and There is also a group of large settlements marked in that it also served as a refuge. epigraphic and literary sources as kóme or metrokomía (see Dagron 1979). These are mostly unfortified, but display great similarities with some of the newly- 5.2.2 THE COUNTRYSIDE established cities. A particularly characteristic example is the large Our knowledge of the countryside in the East settlement at Umm el-Jimal in Jordan ( Figs. 5.46, 5.47), is extensive, particularly because of the innumerable with more than a hundred multi-storey houses and as unfortified settlements surviving in an excellent con- many as fifteen churches, which peaked in prosperity dition (overview in e.g. Hirschfeld 2001; Morisson, in the 5th and 6th centuries (Graf 2001, 234; Kennedy Sodini 2002; Bagnal , Rathbone 2004; Wickham 2005, 2004, 86–91; Quiroga 2016, 82). The distribution of the 442–465). buildings speaks against a uniform design and public The countryside in the East holds a number of buildings are missing with the exception of the churches. fortified sites. In addition to the forts inside cities, which A small fort stood at the edge. The settlement was re- are mentioned above, there was a greater number of duced with the arrival of Islam, but only abandoned after quadriburgia established along the endangered eastern a plague that struck in the mid-8th century. border ( Strata Diocletiana), as well as a network of forts Large examples of such settlements include El Bara in Egypt (Kennedy 2004; Bagnal, Rathbone 2004). Most in the Limestone Massif of northern Syria, as well as of these persisted throughout Late Antiquity, when they Israeli metrokomía at Shivta, Nessana and Rehovot in were renovated and hosted different army units and the Negev . They are sometimes referred to as ‘secondary foederati with their families. A characteristic example is cities’ (Morisson, Sodini 2002, 179–183). the quadriburgium at Qasr Bshir ( Fig. 5.51), built under Standing out in Syria is a large and densely in-Diocletian and also used later (Kennedy 2004, 48–151). habited settlement of Androna (Al-Anderin) ( Figs. Its central court was surrounded by multi-storey build- 5.48−5.50; Strube 2011; Mundell Mango 2011; ead. ings that accommodated soldiers in the upper storey and 2017). It covered an area of 160 hectares, had double provided rooms for animals on the ground floor, which 372 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.48: Al-Andarin. Ruins of a church in the foreground and a large settlement with the remains of a fort in the background (2004). Fig. 5.49: Al-Andarin. Par- tially uncovered entrances to the fort (2004). Fig. 5.50: Al-Andarin. En- trance to the fort with the inscription of the founder (2004). 373 Slavko CIGLENEČKI Fig. 5.51: Qasr Bshir. Quad- riburgium, view from the south (2008). Fig. 5.52: Luxor. Defensive wal s and towers of the le- gionary fortress on the west side of the Temple of Amun (2006). suggests a cavalry unit. In Egypt, forts were constructed survey revealed a church and a cistern, while the best along the Nile and the approaches leading to its valley preserved tower is believed to have served defensive, under the Tetrarchs. Well-known is the legionary for- watch and signalling functions. The authors of its pub- tress in Luxor ( Fig. 5.52), which enclosed the Temple of lication interpret the site as a post tasked with control- Amun and cleverly made use of its architecture (Bagnal , ling the major road in the valley below. They believe a Rathbone 2004, 191–192). garrison of limitanei was permanently stationed in this A special example is the fortification at Kale’i fort that also served as a refuge. They found parallels in Zerzevan ( Fig. 5.53), not far from Amida in northern different military forts in Africa and the Đerdap, and Mesopotamia, with the location on a hundred-metre- the closest parallel in the fortified settlement at Ajdovski high and naturally excellently protected rocky peak gradec above Vranje (see Chapter 3.3). that is in many ways similar to the hil top settlements presented in previous chapters (Deichmann, Peschlow The hallmark of the countryside in the East is the 1977). It is a fortification from the Justinian era that has hundreds of still clearly visible settlements. Their excel- a thick defensive wall with evenly distributed towers and lent state of preservation, wealth of architectural remains interior showing the characteristics of a settlement. A and epigraphic monuments make them an invaluable 374 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.53: Kale‘i Zerzevan. Plan of the fort (Deichmann, Peschlow 1977, Fig. 1). Fig. 5.54: Baude. Remains of a small Late Antique set- tlement in Syria’s Limestone Massif (2004). source of information for understanding the Late Antique to the 9th century. Recent research has shed light on many countryside (Morisson, Sodini 2002, 178–179; Butcher details of life in these rural settlements, their chronology 2003, 145–179; Wickham 2005, 443–465). The vil ages and economic base (Sodini et al. 1980; Tate 1992; Butcher are particularly characteristic of Syria and some areas of 2003, 145–179). They still exhibit most achievements of Turkey. They have been known in literature ever since the the Roman civilisation. Baths survive in several of them, seminal work by Georges Tchalenko on the excellently buildings show very high quality of construction, as preserved settlements in the Limestone Massif of north- well as great diversity of architecture and architectural western Syria (an example in Baude; Fig. 5.54), between decoration. The baths in Serjil a ( Figs. 5.55, 5.56), one the major centres of Antiquity in Antioch and Beroe of the best surviving vil ages, have been dated to 473 on (Tchalenko 1953−1958). They form a network of some the basis of an inscription. These vil ages of independent 700 settlements, termed ‘Dead Cities’ in early literature farmers were not built according to a preconceived plan. due to the extensive remains. Some date as early as the 3rd Inscriptions allow us to distinguish between two main and 4th centuries, though most were built in the 5th and 6th phases, one spanning from the late 3rd or early 4th to the centuries. They continued to develop in their full extent early 5th century, the second one from the first half of the in the first half of the 7th and were less intensely inhabited 6th to the 7th century. They were autonomous vil ages with 375 Slavko CIGLENEČKI Fig. 5.55: Serjil a. A medium- sized and well-preserved settlement, with baths from the 5th century in the fore- ground (2004). Fig. 5.56: Serjilla. Produc- tion room with an oil press (2004). specific functions that bring them close to cities in nature Unfortified vil ages are also typical of Egypt, but (Butcher 2003, 159). were inhabited also after the end of Late Antiquity and In addition to the most comprehensively known are therefore less well-known. An excellent example is settlements in the Limestone Massif of north-western Karanis in the Fayum, with continuity from the Hel- Syria and Hauran, there are similar settlements in the lenistic period to the 6th century, but also Jeme near Negev in Israel, i.e. the already mentioned metrokomía at Luxor ( Fig. 5.60), where a vast settlement was set up in Shivta, Nessana and Rehovot (Hirschfeld 2001, 270). Also the ruins of the Temple of Ramses III, with continuity well-known are those in Cilicia, a remote part of Turkey into the Early Middle Ages (Bagnal , Rathbone 2004, in the hinterland of Seleukeia, that show great similarity 131–134, 193). with the vil ages in Syria (Hild, Hellenkemper 1990). They Also in Egypt, in the Western Desert, a large settle- also host well-preserved multi-storey buildings, though ment at Ain Labakha and a small Late Antique fort were with less architectural decoration. The settlement at established on an important crossroads in the Kharga Karakabakli ( Figs. 5.58, 5.59), for example, has numerous Oasis ( Fig. 5.61) (Bagnal , Rathbone 2004, 255–257). buildings, mostly multi-storey, and two churches, at the This is a combination of a military fort and a settlement edge also a tetrapylon at the intersection of vil age streets in its vicinity similar to the numerous fortified sites in (Hild, Hellenkemper 1990, 290). the Balkans. 376 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.57: Jerade. Ruins of a Late Antique settlement and its tower (2004). Fig. 5.58: Karakabakli. Ru- ins of residential buildings (2010). Fig. 5.59: Karakabakli. Re- mains of a tetrapylon and ancient road (2010). 377 Slavko CIGLENEČKI Fig. 5.60: Jeme, Medinet Habu. Late Antique and later buildings abutting the earlier wall of a temple pre- cinct (2006). Fig. 5.61: Kharga. The fort at Ain Labakha with an associ- ated settlement in a small oasis of Egypt’s Western Desert (2006). Fig. 5.62: Akkale. Late An- tique villa of a high state official, view from the west (2010). 378 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.63: Qasr Ibn Wardan. Palace with a church and army barracks at the edge of the desert (2004). Fig. 5.64: El Breij. Monastery on the edge of a fertile plain, in the densely inhabited Limestone Massif in Syria (2004). Widespread in the semi-arid belt of Lybia is a type A luxury complex associated with a harbour lies at of fortified tower-like farmhouses with an inner court Akkale ( Fig. 5.62), near the city of Elaiousa-Sebaste on known as gusr, which is characteristic of the 4th–6th the Cilician coast (Hild, Hellenkemper 1990, 165–166; century. Similar farmhouses are also mentioned in the Sodini 2003, 35). It has a multi-storey palatial build- Negev in Southern Palestine (Hirschfeld 2001, 266; ing in its centre and around it buildings that include Morisson, Sodini 2002, 179). baths, cistern, oil press and a small domed edifice. The In contrast with the western part of the Empire, complex is seen as an aristocratic vil a from the late 5th typical Roman vil as are rare in the East. Many in the century, an inscription from which mentions the owner Levant were associated with compact settlements and Illous, who was an important person at the court of the cities or sited among farmhouses in the suburbs. More Emperor Zeno. is known on the luxury vil as in the Daphne suburbs A unique complex has been unearthed at Qasr Ibn of Antioch, renowned for beautiful and high-quality Wardan ( Fig. 5.63), at the edge of the Syrian Desert, mosaics (Hirschfeld 2001, 262). An example of a forti- and comprises a two-storey palace, church and army fied vil a has been identified at Ramat Hanadiv in Israel barracks. It is presumably the home of a military com- (Hirschfeld 2001, 266). mander, dated in 564 (Krautheimer 1986, 247–249, 259). 379 Slavko CIGLENEČKI Fig. 5.65: Mt Sinai. Fortified monastery of St Catherine, view from the north (2019). Fig. 5.66: Alahan monas- tery. Excellently preserved church, view from the west (2003). Archaeological research in the East shows differ- an abrupt end of the tradition of Classical Antiquity. ent settlement patterns, but also different densities of Many cities and vil ages lived on into the Umayyad Early Byzantine settlement (Graf 2001, 220–223). The period and only show decline in the late 8th or the early favourable climate conditions between 300 and 600 are 9th century (Graf 2001, 234). believed to have enabled cities to prosper and the coun- tryside to be maximal y inhabited (Decker 2017, 4–5). Monasteries and hermitages are a specific settle- The existence of settlements in remote and previously ment phenomenon, occurring in such a number that uninhabited areas is proof of a growing population and suggests they represented a special type of settlement indicates a shift from cities to rural areas. We see many (Hirschfeld 2001, 270–271; Liebeschuetz 2001, 67–70). examples of modest construction in cities and, at the Of the two, monasteries are better known and also same time, numerous well-built and beautiful y deco- recorded in contemporary texts. Many were erected rated constructions in countryside settlements. at the edge of settlements in the Limestone Massif of Recent investigations have shown that the Islamic north-western Syria ( Fig. 5.64) and in Egypt (Butcher conquest in the East did not have a fatal impact, causing 2003; Bagnal , Rathbone 2004). Egypt also holds several 380 5. BRIEF OUTLINE OF THE LATE ANTIQUE SETTLEMENT PATTERNS IN OTHER PARTS OF THE ROMAN EMPIRE Fig. 5.67: Qalaat Samaan. Octagon with the remains of the pillar of St Simeon Stylites (2004). Fig. 5.68: Qalaat Samaan. Monastery in a pilgrimage centre, preserved to a sub- stantial height (2004). proper monastic settlements with an ecclesiastical centre remains of the pil ar of St Simeon Stylites, and beside it a surrounded by numerous monk cel s, for example at large multi-storey monastery. Its renown drew masses of Kellia, Nitria and Wadi Natrun, while there are larger believers. Below the hill was Telemos, a large settlement fortified complexes in more exposed areas, an example with several monasteries and pilgrims hospices along of which is the monastery of St Catherine on Mt Sinai the edge. The complex is very well preserved today and ( Fig. 5.65) (Bagnal , Rathbone 2004, 107–115, 123–125). offers an insight into the significance and functioning Another feature particular to the East is the nu- of pilgrimage centres in these areas. merous newly-formed pilgrimage centres such as Abu The settlement of the East is highly diverse and Mina, St John in Ephesus, Resafa, Alahan ( Fig. 5.66), documents many aspects that are archaeological y less monastery of St Thecla near Seleukeia and others. A cen- readily visible and less preserved in the areas between tre important in the 5th and 6th centuries was at Qalaat Ravenna and Constantinople, and in the West in general. Samaan. ( Figs. 5.67, 5.68) with nearby monasteries in the As such, they provide invaluable evidence for a better Syrian Limestone Massif. On top of this hil , a large and understanding of the dynamics between different forms beautiful y decorated church was built c. 480 around the of Late Antique settlement patterns. 381 6. CONCLUDING REMARKS AND AN ATTEMPT TO INTERPRET THE SETTLEMENT PATTERNS In Late Antiquity, the area between Ravenna and The overview of the characteristic Late Antique cit- Constantinople experienced a number of political, ies has revealed great differences between them, but also military and socio-economic crises that, together with the different degrees of transformation in different areas a relentless barbarian pressure from beyond the limes, and phases of Late Antiquity. In some cities, changes resulted in a fundamental y transformed landscape. The were taking place within the original design, others prosperity, gradual disintegration and abandonment reduced their urban area and still others were partial y of cities, the changes in the typical forms of Roman or completely transferred to better protected locations. countryside settlement and the creation of numerous The cities that saw a more intensive development in this new, predominantly fortified settlements reveal very period are those that became provincial centres and dynamic settlement processes that unfolded in the area those sited on major land and, even more prominently, under discussion. maritime routes. Because of their strategic significance, The intensification of research from the 1970s on- all enjoyed imperial support. This is particularly ap- wards has fundamental y changed our perception of the parent in the eastern half of the area and is reflected in settlement patterns. Despite this, the perceived changes numerous important constructions. in the settlement are only very briefly addressed in the There are three categories of cities that differ in major syntheses on Late Antiquity, in which much of the time of their beginning and abandonment, but the area is discussed only exceptional y and often er- also in specific functions. The first one comprises cities roneously interpreted. abandoned before the end of Late Antiquity. They can The time frame of this book is delimited by changes be found in the continental part of the western half. in the settlement, both its beginning and its end. The They largely retained the tradition of the Roman urban beginning in the last third of the 3rd century is primar- design and their degradation only began just before the ily marked by the first intense use of hil top sites, which abandonment. become the most important aspect of countryside set- The second, largest group consists of cities that tlement in the course of the next three centuries. The continued to the end of Late Antiquity, some even end comes with the major settlement changes of the late longer. They retained some achievements of the classical 6th and early 7th century, when most of the remaining urban design, but also witnessed changes reflected in the cities are abandoned and fortified hil top settlements disintegration and ruralisation of the urban fabric that predominantly cease to be inhabited, leading to a com- intensified after the mid-5th century. These cities under- plete change of the settlement paradigm. went a reduction, sometimes drastic, of the urban area. There are terminological difficulties associated with Cities with continuity in the western half are lo- this topic, as already discussed in the Introduction, that cated in the well-protected Italy and along the eastern stem from the different names used for the same types of coast of the Adriatic. From Ravenna, Brescia, Verona, settlements in the area. The different names range from the drastical y reduced Aquileia to the major cities on ancient terms, used fairly inconsistently, to completely the Adriatic such as Parentium, Pola, Iader and Salona, neutral descriptive names that were developing within they show a widespread degradation of housing, but also different national archaeological traditions. In addition, well-constructed churches and housing of the elite. The there are different functional interpretations of hil top cities in the eastern half survived with several attributes sites from a time when the boundary between civilian of Roman cities, for example street grid, aqueducts and and military sites became interceptible in the archaeo- baths, and changes are observable in the renovation of logical record. All this greatly hinders our understanding city wal s, as well as monumental constructions and of the otherwise very similar settlement phenomena renovations of Early Christian churches in many inte- across the area under discussion. riors. In these cities as wel , the most apparent changes 383 Slavko CIGLENEČKI pertain to housing, with subdivisioning into new and the unfortified lowland settlements that have their last smaller units, but also to the civic infrastructure being peak of prosperity in the 4th century. Similarly as in other left in disrepair, the introduction of rural elements and parts of the Empire, we can detect an accumulation of the increasing numbers of intra muros burials. Also char- estates here as wel , and with it the creation of a small acteristic is the concentration of modest housing in cores number of large vil a estates (Bruckneudorf/Parndorf, particularly in the shel s of theatres and hippodromes. Balácapuszta, Panik). Some shaped into a different type The third group comprises newly-founded settle- of settlements and into economic or even religious build- ments that may, in accordance with the rather modest ings (Manerba del Garda, Mirište in Petrovac). In the standards of Late Antique cities, be seen as possessing central part of the area under discussion, understanding some urban characteristics. We can distinguish between the vil a transformation is hindered by a poor state of two chronological y and functional y distinct subgroups. investigations that do not reveal the extent and much less The first joins cities that were newly established on the structure and function of later habitations. Roman imperial initiative in the late 3rd and early 4th century, settlements only very rarely continued to be inhabited to primarily in the limes area and its hinterland, some the late 6th century – most frequently in the coastal belt also at strategical y important locations in the interior and on the islands (Rim near Roč, Dragonera Jug, Ubli). (Hissar, Iatrus, Tropeaum Traiani, Dinogetia). Their As stated, more attention is paid to the better- emphasis was on the military function and massive wal s investigated fortified hil top sites, where it was frequently that signalled the Roman presence in an endangered possible to identify several phases of occupation, albeit area. Some were located in the lowland, but no longer not always chronological y clearly definable. The great- included monumental buildings. est density of such sites has been noted in the eastern The second subgroup consists of large settlements part, where several hundred were identified in North established after the mid-5th century on well-protected Macedonia alone, some with associated unfortified elevations that can also be interpreted as small cities. settlements in proximity. We expect a similar density They differ from contemporary fortified settlements in in southern Serbia, as well as in Bulgaria, Montenegro, their greater size and the fact they retained rudimentary Albania, Kosovo and northern Greece. Slightly fewer residential, administrative, military and increasingly im- sites have been observed in the western part, which is portant ecclesiastical functions. They were less numer- in part the result of the (lower) degree of investigation ous in the western part (Kranj, Blagaj), but already began and less favourable conditions for surface surveys. This after the mid-5th century, whereas those in the eastern is confirmed by a more detailed survey of the prehistoric part appeared in greater numbers, but only under Jus- hillforts in the area of the karst poljes in Bosnia and tinian (Caričin grad, Vinica, Shoumen, Louloudies). Herzegovina, which showed that many were reinhabited They represented the centres of smal territorial units. In in Late Antiquity (Livanjsko polje). In the eastern Alpine them, the classical city was transformed into a large and area, i.e. Slovenia, Austria and the fringes of northern well-fortified settlement dominated by an ecclesiastical Italy, concentrations of Late Antique fortifications in- centre and eventual y permanently accommodating a dicate that most people living in the endangered areas small garrison, though a large part of the interior held along the main routes to Italy fled there in the face of numerous, but modest dwellings. barbarian incursions. In terms of the settlement pattern, In comparison with cities, the countryside settle- we see the countryside settlement in Late Antiquity as bi- ment of Late Antiquity displays an even more dynamic polar, exhibiting an interdependence and intermingling transformation and intermixing of patterns; we can of unfortified and fortified settlements, but also a shift observe a tendency towards a gradual abandonment of from the lowland to the hil s and vice versa. unfortified lowland settlements in favour of new fortified For a better insight into the transformation of forms. This also includes a gradual transformation of cities and the countryside, we provide a chronological military posts with the boundary between military and overview of the milestones and phases of settlement civilian sites becoming increasingly blurred. changes. The first changes in cities are perceptible in the The transformation of the countryside is also the last third of the 3rd and early 4th century and take the part of the Late Antique settlement that receives greater form of abandonment of a city (Styberra), reduction of attention in this book compared with earlier major city wal s (Athens) or destruction of buildings (Stobi). studies that rather focused on lowland settlements and Many old city centres were revived in the first half of provided ample knowledge on the Roman vil as and the 4th century, following the reforms of Diocletian and their transformation, whereas the multitude of fortifi- Constantine. This is a time of great prosperity in cities, cations was barely considered and poorly understood. which is reflected in an accelerated renovation of private To correct the skewed view of settlement patterns this luxury buildings. Both emperors also prompted the focus created, we attempted to provide a more balanced creation of several new cities of a mixed civilian-military picture and present the known fortified forms of settle- nature (Hissar, Tropaeum Traiani). ment in particular detail, while only briefly outlining 384 6. CONCLUDING REMARKS AND AN ATTEMPT TO INTERPRET THE SETTLEMENT PATTERNS In the countryside of the last third of the 3rd cen- Carnicum, Aguntum), where in many cases they per- tury, numerous unprotected lowland settlements were sisted to the end of Late Antiquity. Wherever this was not destroyed or abandoned. This was soon followed by an option, people moved to safer areas, largely deeper the creation of new, fortified habitation forms, which into the interior of the Empire or to the Mediterranean had a primarily refugial character and part of which coastline. shows a short-lived presence of army units. These Contemporaneously in the countryside, we can display a slightly broader span; reliable evidence in trace a gradual abandonment of vil as and other forms some cases points to their beginnings as early as the of lowland settlement, with only rare examples of unpro- time of Gallienus or Aurelian (Pasjak, Veliki vrh near tected settlements that were newly established. On the Osredek), while elsewhere changes only become clearly other hand, we can observe a great rise in different forms perceptible under Diocletian and Constantine (Bushati, of fortified sites. They are particularly numerous in areas Paleokastra). Military forts were established along the greatly endangered in the last third of the 4th century main roads in the interior of the area under discussion (Ančnikovo gradišče, Odartsi). Some were inhabited that show an increasingly irregular layout, as they were only occasional y, others already permanently (Rodik, constructed in natural y better protected locations that Gabrovo). Most, including those of a pronounced mili- required adaptations to the terrain (Hrušica). tary character, were abandoned in the first half of the The first occupation of hil top sites in the last third 5th century, only rare ones persisted. of the 3rd century was brief and followed by a hiatus that The paucity of settlements in the middle and cannot always be dated more precisely. The small finds second half of the 5th century, both in the lowland and (primarily coins) from some sites show roughly equal on elevations, mirrors the state of insecurity caused by numbers also in the first half of the 4th century, though barbarian incursions. People no longer renovated the it should be noted that these finds were not unearthed in old cities and settlements, but were also not prepared reliably dated sealed contexts. It would appear that they for a definite resettlement. In terms of Late Antique are likely residual finds, but we cannot exclude the possi- settlement, this is a time of least available evidence. bility that some of the hil tops were occasional y visited, A fatal break came with the devastation caused by the potential y even hosting sanctuaries. These first shifts Hun raids, particularly along the major routes, which from the lowland to fortified hil tops are harbingers the Huns took to progress to the Balkans and towards of the changes that become the new normality in Late Italy. Archaeological y, evidence of this devastation can Antiquity and went on to become the settlement norm. in some cases be seen in the Hun artefacts found in The next major and longer-lasting changes occurred layers of burnt debris, though we also have evidences in the second half of the 4th century, most commonly to- in contemporary texts. In many places, cities and Ro- ward its end when ancient texts relate numerous military man vil as were being gradual y abandoned even earlier. incursions and invasions in the area. Cities display signs of Alongside socio-economic changes, the Hun invasions gradual decline and many also abandonment (Virunum, were certainly the most powerful motivation for a radical Sopianae) in the late 4th and the opening decades of the change of habitation. Many cities were either completely 5th century. Several small and exposed cities were already destroyed or lived on in a reduced form after the middle abandoned towards the end of the 4th century, primar- of the 5th century. Some reveal that their inhabitants ily those along the Danube limes and in the Pannonian moved to nearby elevations (Demetrias), others show a hinterland. Others persisted to the early and some even substantial reduction of the urban area (Philippopolis, the middle of the 5th century. A reliable indication of their Dion). The housing of this time is archaeological y least existence after the early 5th century are Early Christian detectable and we may presume that buildings partly in churches, but these were abandoned or destroyed soon ruins were reused for habitation. afterwards, which consequently proves the abandon- The cities that either did not suffer Hun attacks or ment of cities (Emona). There are few such examples in soon recovered show no major changes. This is primarily the eastern part in spite of the fact that barbarian raids the case with numerous cities on the coast, but these are brought great devastation here as wel , causing variously less well-known due to the continuity of the urban use long interruptions in occupation. into later periods. It would appear that partial y altered Many urban vil as were modified into small habita- houses in them mainly continued to be inhabited. tion or economic units, there is gradual ruralisation of The Hun incursions posed physical danger, but the urban area, many cities were walled at this time or also had a tremendous psychological effect, which in their wal s were reinforced (Brescia, Sirmium, Stobi). the countryside generated a strong impulse toward Only those along the Mediterranean coast display establishing new fortified settlements. We see indica- stronger housing continuity. tions of small-scale and improvised constructions in the The population of the abandoned cities initial y middle third of the 5th century, but also occasional use sought refuge in the nearby military forts (Carnuntum, of refuges on fortified hil s (Ukosa). Only in the second Aquincum) and natural y protected locations (Iulium half of the 5th century do we see an intense occupation 385 Slavko CIGLENEČKI of fortified hil top sites in the western part, which lasts show signs of urban continuity in the classical tradition, in most cases to the late 6th or even early 7th century with an emphasis on church architecture (Iader, Thes- (Rifnik, Tonovcov grad, Lobor, Bakinci). It seems that saloniki). In the time of the Avaro-Slav incursions in the the people who had previously hoped to return to their late 6th and early 7th century, we also see several large former homes now decided to permanently settle in semi-urban units that could be associated with the influx better protected locations. Coins and other small finds of the Roman population from the interior (Ulcinj). indicate this settlement took place even before the ar- The settlement picture again changed in the late 6th rival of the Ostrogoths (Rifnik, Ajdovski gradec). Most or first half of the 7th century, when most Roman cities settlements established at this time show local initiative. ceased to exist and the countryside witnessed a shift In the western part, the time of renewed peace under from the fortified hil top settlements back down to the Theoderic consolidated the resettlement and the forti- lowland. This is also a time when housing underwent fied settlements were included in the administrative and a complete transformation, losing all connection with economic system of the new state. For several strategi- the classical tradition and largely mirroring the liv- cal y significant cities settlement was a state initiative. In ing standard of newcomers. Continuity has only been the mass of newly-founded fortified hil top settlements, observed in some cities in northern Italy and along the some even seem to have been minor regional centres Mediterranean coastline. with modest urban features (Carnium). The comparison with parts of the Roman Empire In the eastern part of the area under discussion, a outside the area under discussion has revealed many complete shift of the settlement paradigm mainly oc- commonalities, but also great differences that under- curred from the 530s onwards, when a great majority of score the specific nature of the area between Ravenna new fortifications was established as part of Justinian’s and Constantinople. Cities in the West show all the altered strategy of state defence. Compared with the features of transformation and abandonment observed western part, the creation of large central settlements, in the area under discussion. The great crisis of the sort of semi-urban units no longer akin to classical 5th century is fol owed in many places by a renewed cities, is even more clearly visible here (Caričin grad, intensification of urban life in the 6th century. We can Jelica, Vinica, Shoumen). Throughout this geographi- also observe a greater prosperity of cities in the Medi- cal y, political y and economical y diverse area, we can terranean coastline contrasted with a modest existence observe a common trend of settlement shifts, but their in the interior, in the 5th century also migrations from speed and intensity differ. the lowland to the nearby elevations (Saint-Bernard- The creation of fortified hil top settlements was de-Comminges). largely a spontaneous reaction of the local inhabitants, The settlement patterns in the countryside of the as inferred from the numerous pre-Justinian period West are very similar, but the changes do not follow the fortifications across the western and less numerously same pace. The transformation of Roman vil as occurred also in the eastern part. The Emperor Justinian made slightly later; many persisted in their original function to clever use of this self-preservation reflex of the popula- the first half of the 5th century or even beyond, and we tion and deployed the army as his architects to support know much more of the different types of transformation primarily those sites that played major strategic and in the 5th and 6th centuries. New research suggests that administrative roles (Markova Mehana, Hexamilion). A fortified hil top sites in some parts (particularly south- large number of late fortifications with defensive instal- eastern France) were settled much more extensively than lations, however, were less well-constructed and show thought until recently, which brings the West close to the an absence of the imperial interest. The process also ran area under discussion in terms of settlement patterns. in the opposite direction, with the local population ar- Greater differences are visible in the East. The large ranging modest dwellings in the safety of earlier military and flourishing cities of Asia Minor, the Levant and forts (Veliki Gradac near Donji Milanovac, Golemanovo Egypt are hardly comparable with those in the eastern Kale, Odartsi). Alps, Pannonia and the northern Balkans, as the latter The last major step in the transformation of cities show a much more pronounced degradation of the can be detected in the final decades of the 6th century, urban area, decline of monumental buildings and other when there is a final migration wave of people from the urban achievements than their counterparts in the East. interior or from other abandoned cities to the natural y The border with the Sassanian Empire did constitute a well-protected locations along the Adriatic coast. The threat, but it also allowed for a continued existence of clearest examples include the formerly powerful me- cities and a fairly intense settlement of the countryside. tropolis of Aquileia, from which the ecclesiastical centre This development cannot be interpreted in any other way moved to the well protected, but much smaller castrum than by a lesser pressure, in both scale and frequency, of Grado on an island in the nearby lagoon, but also the on the eastern border of the Empire as the one noted in last inhabitants of Salona moving to the former palace the Balkans and the Middle Danube Basin. of Diocletian in Split. Only few formerly major cities 386 6. CONCLUDING REMARKS AND AN ATTEMPT TO INTERPRET THE SETTLEMENT PATTERNS In the cities of the East, we can trace important into the opening decades of the 7th century. With the public buildings, as well as still existing water supply end of these last nuclei of the ancient civilisation, we and sewage systems in some places even towards the can speak of the end of Antiquity in a large part of the end of the 6th century, whereas such examples between continental area between Ravenna and Constantinople. Ravenna and Constantinople are mere exceptions. What For the area under discussion, we still lack evi- is similar is the proliferation of church buildings, par- dence for a reliable distinction between a military and ticularly in the 6th century. Fortifying cities with wal s civilian nature of a large part of the late fortified sites. was an existential necessity in the area under discussion, Archaeological evidence shows that those in strategical y in the East often only a display of civic pride. It is true, significant locations had at least some military assign- however, that city wal s were important factors in the ments and hosted garrisons for that purpose (soldiers urban design of cities in endangered areas, for example with their families?, foederati?, local militia?), while Syria (Sergiopolis, Zenobia, Amida). those at less exposed sites were largely the result of the Radical changes in the East largely occurred in the local initiative. What is surprising is the low-quality 6th, while they already become perceptible in the area construction of housing in the Late Antique cities both under discussion in the 5th century. Furthermore, the in the western and eastern halves compared with the decline of cities in the East came later, with the classical mortared masonry constructions of high quality in achievements, architectural in particular, continuing in numerous contemporary hil top fortified settlements the early Islamic cities. To summarise the comparisons, that persisted to the late 6th century. This shows that we can state that cities everywhere witnessed extensive only the government administration with the Church at construction of church architecture in the 5th and 6th his head remained in the ruralised cities, whereas part centuries, but there are differences in the continuation of the elites – particularly in the western part under the of other public buildings and urban amenities, as well Ostrogoth rule – moved to smaller, but better protected as the size of the urban population; throughout Antiq- settlements. Even in difficult living conditions, well- uity, in fact, cities in the East were particularly densely built housing was constructed here by employing the populated. A commonality pertains to some cities be- knowledge of construction brought with them from the ing established in easily defensible places both in the cities. Masonry architecture predominates in the forti- East (Arif) and in the eastern Alpine and Balkan areas fied settlements in the western half. In the eastern half (Caričin grad). masonry architecture is only used for the most promi- The greatest differences are observable in the coun- nent buildings and in a small portion of such settlements tryside. In a large part of the East, in the 5th and 6th (Jelica). Further investigations will show whether this centuries, unfortified settlements were the basic form only mirrors the social status of the refuges or was the of living. Fortified settlements are extremely rare and consequence of the later construction (in the 530s). The Roman vil as only slightly more common. The dense set- outlined characteristics of the housing construction in tlement of the countryside in the eastern Mediterranean, cities and fortified settlements indicate ever smaller, in especial y with the intensely inhabited Limestone Massif some cases non-observable differences between them. in northern Syria, is without parallels in the area under How does the knowledge of the Late Antique set- discussion; closest are some areas in the central and tlement patterns between Ravenna and Constantinople eastern Balkans, in the well-known North Macedonia contribute to the scholarly discourse on the decline in particular. or transformation of the classical world? The urban The countryside in the West reveals settlement pat- culture clearly shows a decline, whereas the diversity terns similar to those discussed in the previous chapters, of the countryside settlement rather points to a radical with differences pertaining to fortified settlements being transformation of human habitation with the funda- less represented in the West and the transformation mental achievements of Antiquity still visible in the of unfortified countryside settlements being less well- newly-established settlements. The newly-founded known in the eastern Alps and the Balkans. cities as wel , which are predominantly large central The settlement patterns in the eastern and west- settlements, continue the traditions of the ancient ern parts of the Empire show the processes in cities urban lifestyle, albeit in a considerably more modest to be similar and thus universal in their significance. form. The boundary between cities and late, mostly The countryside, on the other hand, shows substantial fortified settlements is blurred in many parts and only differences. The exposed position in the hinterland of the administratively significant cities in the interior the limes influenced an earlier transformation in com- and the major trade centres along the coasts of the parison with most other parts of the Empire and also Mediterranean stand out. caused migrations on the largest scale. The numerous Archaeological research and Procopius’ writings hil top settlements are the last manifestation of the clas- are unveiling a cleverly devised defence system of forts, sical settlement tradition, preserving many civilisational but also cities and fortified settlements operational in elements of Antiquity to the late 6th and in many places the time of Justinian. The system functioned by com- 387 Slavko CIGLENEČKI bining the different elements, all of which could serve the continuation of the classical civilisational achieve- military purposes (surveil ance, signalisation, occasional ments and a link between the old Roman populations presence of army troops) in addition to their civilian and the newcomers. functions. We clearly see a variety of contemporary The political, economic, military, ecclesiastical, interventions that additional y fortified cities and earlier and cultural history enables general conclusions, while settlements, thereby allowing an at least short-term pres- systematic investigations of the different types of coun- ence of army units in many of them. These interventions tryside settlement forms are required for a comprehen- were most often aimed at strengthening city wal s in sive understanding of the changes in the settlement key sections, while those in the interior only focused patterns within smaller areas. The discussion so far on modifying and enlarging churches and, even more mainly focused on the sites with well-preserved archi- prominently, adding baptisteries. It would thus appear tectural remains and valuable artefacts, but these only that a great majority of the fortifications noted in Pro- show the most readily visible aspect of the settlement copius merited to be mentioned as such due to limited, patterns, while all others are poorly known. Research but effective measures. At the same time, some new posts in the future should focus on a detailed analysis of the were built, both along the limes and in its hinterland. vast quantities of small finds, but also on the economic Particularly significant is the discovery of a series of and environmental factors. Only comprehensive re- forts along the eastern coast of the Adriatic, which are search of individual settlements and their economic not mentioned in The Buildings by Procopius, but show hinterland can offer a reliable answer with regard to clear Early Byzantine fortification features, with the the different settlement patterns in different areas. If archaeological finds recovered from them confirming this answer is certainly more complex for cities, the such dating (Gradina Žirje). The choice of natural y creation of fortified settlements is largely an indication protected locations and the predominantly thinner wal s of the dangerous conditions in the period of civil wars testify to the haste in construction and the necessity to and barbarian incursions when these settlements were protect the conquered territories, especial y the land and inhabited precisely because of their natural y protected maritime routes connecting East and West. In relation location. The mass of countryside settlements, particu- to Early Byzantine fortifications, we should mention the larly the more readily perceptible fortified, but also the proteichisma, a feature unknown in the western part unfortified ones persisted to the end of Late Antiquity of the area under discussion prior to the Gothic War. and attest to a much greater population density that Exclusively military forts were only constructed at key previously posited. strategic points (Veliki Sikavac). The settlement patterns in a vital, but also the most The example of Victorinus in Illyricum shows that endangered area of the Empire, between Ravenna and the state only provided help to some cities and large Constantinople, in many ways complement the trans- newly-founded fortifications in the shape of experienced formation of cities and even more so of the countryside architects. For the strategical y significant forts, on the across the whole of the former Roman Empire. The other hand, we may presume the army was involved in previously oversimplified patterns, represented in the their construction, but also that imperial funding was East primarily by the excellently preserved unforti- secured for the purpose. fied settlements in Syria’s Limestone Massif and in the The settlement picture of Late Antiquity shows a West by the characteristic Roman villas, have been gradual decline of cities in favour of smaller fortified substantial y enhanced in a large part of the Empire settlements that could be more easily supplied and were with different forms of fortified sites. It is no longer often even completely autarchic. The latter are the most acceptable to discuss and reiterate the simplified settle- characteristic expression of the changed political and ment patterns, as the settlement in Late Antiquity has socio-economic conditions. Their existence was short- been shown to be much more complex and nuanced lived in some places, but longer or recurrent in more in general, not only in the area between Ravenna and exposed areas. In a limited and modest extent, they are Constantinople. 388 BIBLIOGRAPHY ABRAMIĆ, M., A. COLNAGO 1909, Untersuchungen skata krepost prez rimskata epoha (II-IV v.). – in Norddalmatien. – Jahreshefte des Österreichisches Археология ( Arheologija) 20, 13–24. Archäologisches Institut 12, Beiblatt, 13–112. ANTONOVA, V. 1981 = АНТОНОВА, B., Шуменската ACHIM, I. A. 2012, Paysage urbain tardo-antique à крепост − твŭрдина през вековете (Shoumen Histria: les églises paléochrétiennes entre le cadre fortress, a powerful stronghold through the ages). – architectural et la liturgie. – Dacia 56, 125–167. Sofia. Acta XIII Congressus 1998 = Radovi XIII. međunarodnog ASSÉNAT, M., A. PÉREZ 2017, La Topographie antique kongresa za starokršćansku arheologiju / Acta XIII d’Amida (III Siecle apres J. C.-VIe Siecle apres J. Congressus internationalis archaeologiae christi­ C.) d’apres les sources litteraires. – In: Rizos (ed.) anae 3 (Split, Poreč, 25.9 - 1.10.1994), Vjesnik za 2017a, 57–70. arheologiju i historiju dalmatinsku, 87/89. – Split, ATANASOV, G. 2013, The Roman-Byzantine castel- Roma. lum of Durostorum – Dorostol on the bank of the ADAM, R. 1764, Ruins of the Palace of the Emperor Dio­ Danube in Silistra. – Arheologia 1, 52–70. cletian at Spalatro in Dalmatia. – London. ATANASOV, G. 2015, Le refugium de l’Antiquité tardive AIROLDI et al. 2016 = Airoldi, F., A. Bona, C. Cattaneo, du ІV–V s. pres du vil age de Ruyno, région de D. Cesana, M. Fortunati, A. Fossati, E. Grassi, M. Durostorum (Silistra – Bulgarie). – In: L. Vagalin- Mattia, C. Perassi, F. Sacchi, M. Sannazaro, M. ski, N. Sharankov (eds.), Limes XXII. Proceedings Uboldi 2016, Ponte Lambro tra età romana e tardo- of the 22nd International Congress of Roman antica: la pars rustica del a vil a in località Schieppo. Frontier Studies Ruse, Bulgaria, September 2012, – Rivista archeologica del ’antica provincia e diocesi Sofia, 217–226. di Como, 172–249. AZINOVIĆ BEBEK, A., P. SEKULIĆ 2014, Kasnoantičko ALAJBEG et al. 2019 = Alajbeg, A., P. Dresler, M. Vágner nalazište Crkvišće Bukovlje kod Generalskog Stola 2019, Archaeological investigations at Otres in (The late ancient site of Crkvišće Bukovlje near 2017 and 2018 / Arheološka istraživanja na Otresu Generalski Stol). – Starohrvatska prosvjeta 3, 41, 2017. i 2018. Godine. – Starohrvatska prosvjeta 165–175. 46,189–204. AZINOVIĆ BEBEK, A., P. SEKULIĆ 2019, Late Antiq- ALEKSOVA, B. 1997, Loca sanctorum Macedoniae. The uity Hil top Fortress Crkvišće Bukovlje. – In: For­ Cult of Martyrs in Macedonia from the 4th to the 9th tifications, defence systems, structures and features centuries. – Skopje. in the past. Proceedings of the 4th International AMORE et al. 2005 = Amore, M. G., L. Bejko, Y. Cerova, Scientific Conference on Mediaeval Archaeology I. Gjipali 2005, Via Egnatia (Albania). Project: of the Institute of archaeology Zagreb, 7th – 9th results of fieldwork 2002. – Journal of Roman June 2017, Zbornik instituta za arheologiju / Serta Archaeology 18, 336–360. Instituti Archaeologici 13. ANAMALI, S. 1993, Architettura e decorazione tardoan- BAÇE, A. 1976, Fortifications de la basse antiquite en tica in Albania. – Corso di cultura sull arte Ravenate Albanie. – Monumentet 11, 69–75. e bizantina XL, 447–474. BAÇE, A. 1981, Kështjel a e Paleokastrës. – Iliria 11/2, ANGELESCU et al. 2017 = Angelescu, M. B., V. Bottez, 165–235. I. Achim 2017, New research on the Early Christian BADURINA, A. 1982, Bizantska utvrda na otočiču basilicas in the southern part of the acropolis . – Palacol. – Izdanja Hrvatskog arheološkog društva 7, Dacia 61, 103–153. 171–177. ANTONOVA, V. 1978 = АНТОНОВА, B., Шуменската BAGNALL, R. S., D. W. RATHBONE (eds.) 2004, Egipt. крепост през римската епоха (ІІ-ІV в.) (Šumen- From Alexander to the Copts. – London. 389 Slavko CIGLENEČKI BALABANOV, K. 2011 = БАЛАБАНОВ, K., Виничко BAVANT, B. 1984, La ville dans le nord de l’Illyricum кале: митологија, религија и историја пишувани (Pannonie, Mésie I, Dacie et Dardanie). – In: Vil es со глина / Vinica Fortress: Mythology, Religion and 1984, 245–288. History written in clay. − Skopje. BAVANT, B. 2007, Caričin grad and the Changes in the BALDINI LIPPOLIS, I. 2001, La domus tardoantica. Nature of Urbanism in the Central Balkans in the Forme e rappresentazioni del o spazio domestico Sixth Century. – In: Poulter (ed.) 2007a, 337–374. nel e città del Mediterraneo. – Col ana di studi e BAVANT, B., V. IVANIŠEVIĆ 2003, Iustiniana Prima scavi del Dipartimento di Archeologia, Università – Caričin Grad. – Catalog de l’exposition organi- degli studi di Bologna . sée par le Centre culturel français et l’Institut ar- BALDINI, I., E. BAZZECHI 2016, About the Meaning chéologique de Belgrade, Belgrade. of Fortifications in Late Antique Cities: the Case of BEAUDRY et al. 2018 = Beaudry, N., D. Moreau, G. Athens in Context. – In: Frederiksen et al. (eds.) Atanasov, V. Yotov, A. Milanova 2018, Zaldapa 2016, 696–711. in Late Antiquity: Archaeological Perspectives on a BARAKA PERICA, J., N. GRBIĆ 2019, Rezultati fortified Christian City of the Lower Danube’s Hin­ istraživanja utvrde Ošlje-Gradac i selo Ošlje u terland. – XVII International Congress of Christian općem povijesno-arheološkom kontekstu. – Anali archaeology. Zavoda za povijesne znanosti Hrvatske akademije BEGOVIĆ DVORŽAK, V. 2001, Fortificationi sklop znanosti i umjetnosti u Dubrovniku 57, 145–174. Kastrum - Petrovac na Brijunima. – Histria Antiqua BARNEA, A. 1986, La fortresse de Dinogetia a la lumiere 7, 177–190. de dernieres fouilles archéologiques. – In: Studien BEGOVIĆ, V., I. SCHRUNK 2001, Preobrazbe rimskih zu den Militärgrenzen Roms 3. 13. Internationaler vila na istočnom Jadranu u kasnoj antici i ranom Limeskongreß, Aalen 1983, 447–450. srednjem vijeku. – Prilozi Instituta za arheologiju BARTEL et al. 1979 = Bartel, B., V. Kondić, M. R. Werner u Zagrebu 18, 157–172. 1979, Excavations at Kraku lu Jordan, Preliminary BEGOVIĆ et al. 2012 = Begović, V., I. Schrunk, M. Report, 1973–1976 Seasons. – Journal of Field Ugarković 2012, Rimska vila u uvali Soline na Archaeology 6, 127–149. otoku Sv. Klement kod Hvara. Preliminarna BASLER, Ð. 1972, Arhitektura kasnoantičkog doba u analiza arhitekture prema geofizičkom istraživanju Bosni i Hercegovini. – Sarajevo. i sondiranju / Roman vil a in the Soline cove on BASLER, Đ. 1977, Rimski metalurški pogon i naselje u the Island of St. Clement near Hvar. Preliminary dolini Japre (Römisches Eisenwer und Ansiedlung analysis of the architecture according to geophysi- im Japra-Tal). – Glasnik Zemaljskog muzeja Bosne i cal investigations and sondages. – Prilozi Instituta Hercegovine u Sarajevu 30/31, 1975–1976, 121–216. za arheologiju u Zagrebu 29, 143–166. BASLER, Đ. 1988, Koštur, Dabrica, Stolac. – In: Arheološki BELAMARIĆ, J. 2003, The Date of Foundation and leksikon Bosne i Hercegovine 3, Sarajevo, 184. Original Function of Diocletian’s Palace at Split. – BASLER, Đ. 1993, Spätantike und Frühchristliche Ar­ Hortus Artium Medievalium 9, 173–185. chitektur in Bosnien und der Herzegowina. – Schrif- BELDEDOVSKI, Z., T. NACEV 2006 = БЕЛДЕДОВСКИ, ten der Balkankommission, Antike Abteilung 19. З., T. НАЦЕВ, Новооткриен термален објект во BASSI, C., E. CAVADA 2013, Dal a Trento romana al a Баргала (Newlу discovered Late Roman Therma Trento cristiana. – In: La citta e l’archeologia dell in Bargala). – Macedoniae acta archeologica 17, sacro, Trento, 13–25. 289–302. BAUER, I. 1997, Römerzeitliche Siedlungen in der BLÉTRY, S. 2017, L’urbanisme et l’habitat de la ville de Steiermark mit besonderer Berücksichtigung des Zénobia-Halabiya: résultats de la mission franco- archäologischen Fundmaterials. – Fundberichte aus syrienne (2006-10). − In: E. Rizos 2017a 137−152. Österreich 36, 71–192. BENAC, A. 1985, Utvrđena ilirska naselja (I). Delmat­ BAUSOVAC, M. 2003, Poznoantični kovinski predmeti ske gradine na Duvanjskom polju, Buškom blatu, z Gradca pri Prapretnem (Spätantike Metallgegen- Livanjskom i Glamočkom polju ( Agglomérations stände von Gradec bei Prapretno). – Arheološki fortifiées il yriennes (I). Les “gradinas” delmates de vestnik 54, 315–324. la plaine di Duvno, de Buško blato et des plaines de BAUSOVAC, M., J. KRAJŠEK 2020, Municipium Livno et de Glamoč). – Djela ANUBiH 60, Centar za Claudium Celeia. – Celeia Antiqua. balkanološka ispitivanja 4 / Monographies, Acad- BAUSOVAC, M., R. KREMPUŠ 2009, Celje – Knežji emie des sciences et des arts de Bosnie-Herzegovine dvorec. – Varstvo spomenikov. Poročila 45 (2008), 60, Centre d’études balk. 4. 44–45. BERG, H. 1985, Bischöfe und Bischofssitze im Ostalpen- BAUSOVAC, M., D. PIRKMAJER 2012, Late Roman und Donauraum vom 4. bis zum 8. Jahrhundert. – Glazed Pottery from Rifnik near Celje. – Rei Cre­ In: H. Wolfram, A. Schwarz, Die Bayern und ihre tariae Romanae Fautorum Acta 42, 33–39. 390 BIBLIOGRAPHY Nachbarn 1, Denkschriften der Österreichischen BITRAKOVA GROZDANOVA, V. 2009, Lychnidus a Akademie der Wissenschaften 179, 61–108. L’epoque paleochretienne et son noyau urbain. – BERGER, A. 2017, Mokisos – Eine Kappadokische Zbornik radova Niš i Vizantija VII, 23–36. Fluchtsiedlung des sechsten jahrhunderts. – In: BITRAKOVA GROZDANOVA, V. 2011, Golem Grad, Rizos (ed.) 2017a, 177–188. Prespa. – Skopje. BIERBRAUER, V. 1985, Frühmittelalterliche Castra im BLACKMAN, D. 2000, Archaeology in Greece 1999– östlichen und mittleren Alpengebiet: germanische 2000. – Archaeological reports 46, 1–125. Wehranlagen oder romanische Siedlungen? Ein BOJANOVSKI, I. 1970, Nova epigrafska potvrda Del- Beitrag zur Kontinuitätsforschung. – Archäologi­ miniuma sa Duvanjskog polja / New epigraphic sches Korrespondenzblatt 15, 497–513. evidence of Delminium from Duvno polje. – BIERBRAUER, V. 1987, Invil ino­Ibligo in Friaul 1, Glasnik Zemaljskog muzeja Bosne i Hercegovine u Die römische Siedlung und das spätantik­frühmit­ Sarajevu, N.s. 25, 5–18. telalterliche Castrum. – Münchner Beiträge zur BOJANOVSKI, I. 1974, Dolabelin sistem cesta u rim­ Vor- und Frühgeschichte 33. skoj provinciji Dalmaciji. – Djela 47, Centar za BIERBRAUER, V. 1988, Invil ino­Ibligo in Friaul 2, Die balkanološka ispitivanja. spätantiken und frühmittelalterlichen Kirchen. – BOJANOVSKI, I. 1976, Ranosrednjovjekovno utvrđenje Münchner Beiträge zur Vor- und Frühgeschichte 34. u Vrbljanima na Sani. – Glasnik Zemaljskog muzeja BIERBRAUER, V. 2003, Rifnik. – In: Real exikon der Bosne i Hercegovine u Sarajevu, N.s. 29, 245–258. Germanischen Altertumskunde 24, Berlin, New BOJANOVSKI, I. 1980, Kasnoantički kaštel u Gornjim York, 613–619. Vrbljanima na Sani. – Glasnik Zemaljskog Muzeja BIERBRAER, V. 2004, Säben. – In: Real exikon der Sarajevo, N.s. 34, 105–126. Germanischen Altertumskunde 26, Berlin, New BOLTA, L. 1981, Rifnik pri Šentjurju. Poznoantična York, 69–73. naselbina in grobišče. – Katalogi in monografije 19. BIERBRAUER, V., H. NOTHDURFER 2015, Die BOLTIN-TOME, E. 1986, Predloka – antična in Ausgrabungen im spätantik­frühmittelalterlichen zgodnjesrednjeveška lokaliteta (Predloka – Antike Bischofssitz Sabiona–Säben in Südtirol I . Frühchrist­ und Frühmittelalterliche Lokalität). – Izdanja Hrliche Kirche und Gräberfeld. – Münchner Beiträge vatskog arheolokog društva 11/2 (1987), 189–207. zur Vor- und Frühgeschichte 58. BONETTO, J. 2004, Difendere Aquileia, città di fron- BIERNACKI, A. B. 2005, A City of Christians: Novae tiera. – Antichità Altoadriatiche 49, 151–196. in the 5th and 6th C AD. – Archaeologia Bulgarica BONETTO, J., L. VILLA 2003, Nuove considerazioni IX/1, 53–74. sulle cinta fortificate di Forum Iulii al a luce dello BIERNACKI, A. B. 2013, The Bishopric of Novae (Moe- scavo di casa Canussio. – Forum Iulii 27, 15–67. sia Secunda, 4th-6th Cent.): History, Architecture, BORISOV, B. 2010a = БОРИСОВ, Б., Кастра Рубра Daily Life. – In: Acta XV Congressus internationalis през Античността и Средновековието ( Castra archaeologiae christianae (Toleti 8.-12.9.2008), Rubra during Antiquity and the Middle Ages). – Ve- Episcopus, Civitas, Territorium, Pars I, Citta de liko Trnovo, 185–198. Vaticano, 895–914. BORISOV, B. 2010b, Dyadovo. Vol. 2. The Sanctuary of BINTLIFF, J. 2007, The Contribution of Regional Survey the Thracian Horseman and The Early Byzantine to the Late Antiquity Debate: Greece in its Mediter- fortress. – Varna. ranean Context. – In: Poulter (ed.) 2007a, 649–678. BORISOV, B. D. 2015 = БОРИСОВ, Б., Новите BINTLIFF, J. 2012, The Complet Archaeology of Greece. археологически проучвания на ранновизан- From Hunter­gathers to the 20th Century A.D. – тийската и средновековна крепост Туида в Chichester. Сливен – един провален проект (The latest re- BÍRÓ, M. 1974, Roman Villas in Pannonia. – Acta search of the Early Byzantine and medieval fortress Archaeologica Hungarica 26, 23–57. of Tuida in Sliven – a ruined project). – Епохи / BITENC, P., T. KNIFIC, 2008, Oggetti di origine ostro- Epochs 23/1, 19–29. gota e ceramiche domestiche di epoca gota rinve- BORZIĆ, I., I. JADRIĆ 2007, Novi prilozi arheološkoj nute in Slovenia. – In: M. Buora, L. Vil a (eds.), topografiji Dugopoljskog kraja. – Archaeologia Goti dal ’Oriente al e Alpi, Archeologia di frontiera Adriatica 1, 145–173. 7, 98–108. BOSHNAKOV, K. 2007 (ed.) = БОШНАКОВ, К., BITRAKOVA GROZDANOVA, V. 1975, = БИТРАКОВА Сборник в памет на Карел и Херменегилд ГРОЗДАНОВА, B., Староxристижански Шкорпил / Proceedings of a conference in memory спомениcи во Оxридско (Monuments paléochréof the brothers Hermenegild (Hermann) and Karel tiens de la région d’Ohrid). – Ohrid. Škorpil (22 May, 2004, Sofia). – Jubilaeus 6. BOWDEN, W. 2000, Gazetter of sites in Late Antique Epirus Vetus (online publication; last access: 391 Slavko CIGLENEČKI 23rd March 2023). [https://www.academia. BREŠČAK, D. 1990, Gradec nad Mihovim, Novo edu/10282608/Gazetteer_of_Late_Antique_Sites_ mesto. – In: Arheološka najdišča Dolenjske, Arheo. in_Epirus_Vetus] Dossier Dolenjska, 105–106, Ljubljana. BOWDEN, W. 2003 , Epirus Vetus: The Archaeology of a BROGIOLO, G. P. 1988, Lo scavo di via Alberto Mario. – Late Antique Province. – London. In: G. Panazza, G. P. Brogiolo (eds.), Ricerche su BOWDEN, W. 2007, Butrint and Nicopolis: Urban Plan- Brescia altomedievale. Vol. 1, Gli studi fino al 1978 ning and the “Romanization” of Greece Epirus. – In: lo scavo di Via Alberto Mario, Brescia, 37–218. I. L. Hansen, R. Hodges (eds.), Roman Butrint. An BROGIOLO, G. P. 1997, Continuita tra tarda antichita Assessement, Butrint Archaeological Monographs 1. e altomedioevo attraverso le vicende delle ville. – BOWDEN, W., J. MITCHELL 2007, The Triconch Palace In: E. Roffia (ed.), Vil e Romane sul Lago di Garda, at Butrint: The Life and Death of a Late Roman Desenzano del Garda e del a Riviera, 299–313. Domus. – In: Lavan et al. (eds.) 2007, 455–474. BROGIOLO, G. P. (ed.) 1999, S. Giulia di Brescia. Gli BOWDEN, W., L. PËRZHITA 2014, The Roman vil a scavi dal 1980 al 1992. Reperti preromani, romani and Early Christian complex at Diaporit. – In: L. e alto medievali. – Firenze. Përzhita et al. (eds.), Proceedings of the International BROGIOLO, G. P. 2010, Città e suburbio tra tardoan- Congress of Albanian Archaeological Studies (Tirana tico e altomedioevo in Italia settentrionale. – In: 2013), 469–484. D. Vaquerizo, Las áreas suburbanas en la ciutad BOWDEN et al. 2004 = Bowden, W., L. Lavan, C. Macha- histórica. Topografía, usos, función, 67–82. do (eds.) 2004, Recent Research in the Late Antique BROGIOLO, G. P. 2011a, Le origini del a città medievale. Countryside. – Late Antique Archaeology 2. – Post-classical archeologies Studies 1. BRAJKOVIĆ et al. 2013 = T. Brajković, Ž. Krnčević, E. BROGIOLO, G. P. 2011b, From ancient to early medieval Podrug 2013, Arheološki vodič po Danilu. – Šibenik. town: reflections starting from the case of Verona. – BRANDS, G., H. G. SEVERIN (eds.) 2003, Die spätantike In: Heinrich-Tamáska (ed.) 2011a, 195–210. Stadt und ihre Christianisierung. Symposion vom 14. BROGIOLO, G. P. 2014, Costruire castelli nel ’arco al- bis 16. Februar 2000 in Hal e/Sal e. – Spätantike - pino tra V e VI secolo . – Archeologia Medievale 40, Frühes Christentum - Byzanz. Kunst im ersten 143–156. Jahrtausend. Reihe B. Studien und Perspektiven BROGIOLO, G. P. 2018, Nuove storie di Sirmione 11, 1–26. raccon tate dal ’archeologia. – In: E. Roffia (ed.), BRATOŽ, R. 1982, Evgipij, Življenje svetega Severina. Sirmione in eta antica. Il territorio del comune dal a Uvod, prevod in komentar. – Ljubljana. preistoria al medioevo, 9–12. BRATOŽ, R. 1992, Nekatera nerešena in nerešljiva (?) BROGIOLO, G. P., A. CAGNANA 2005, Nuove ricerche vprašanja iz zgodovine severnojadranskih dežel v 6. sul ’origine di Grado. – In: G. P. Brogiolo, P. De- in 7. stoletju. − Zgodovinski časopis 46/3, 297−307. logu (eds.), L’Adriatico dalla tarda antichità al ’età BRATOŽ, R. 2001, Koprska škofija od prve omembe carolingia, Atti convegno (Brescia 2001), 79–108. (599) do srede 8. stoletja. – Acta Histriae 9/1, 37–64 . BROGIOLO, G. P., A. CHAVARRÍA ARNAU 2014, BRATOŽ, R. 2000 Das Patriarchat Grado im mono- Vil ae, pretoria e aedes publiche tardoantichi in theletischen Streit. – In: R. Bratož (ed.), Slovenija Italia settentrionale: riflessioni a partire di alcune in sosednje dežele med antiko in karolinško dobo 2 / ricerche recenti. – In: P. Pensabene, C. Sfameni Slowenien und die Nachbarländern zwischen Antike (eds.) 2014, La Vil a restaurata e i nuovi studi und karolingischer Epoche 2, Situla 39, Razprave sul ’edilizia residenziale tardoantica, 227–239. 18, 609–658. BROGIOLO, G. P., S. GELICHI 1996, Nuove ricerche BRATOŽ, R. 2009, Cristianesimo in Istria. Una sintesi e sui castel i altomedievali in Italia settentrionale. – alcune riflessioni. – In: E. Marin, D. Mazzoleni (eds.), Quaderni del Dipartimento di archeologia e storia Il cristianesimo in Istria fra tarda antichità e alto me­ delle arti. Sezione archeologica, Siena. dioevo. Novità e riflessioni, Città del Vaticano, 9–46. BROGIOLO et al. 2000 = Brogiolo, G. P., N. Gauthier, BRATOŽ, R. 2011a, Die kirchliche Organissation in N. Christie (eds.) 2000, Towns and their Territories Westil yricum (vom späten 4. Jh. Bis um 600) – between Late Antiquity and the Early Middle Ages. Ausgewälte Fragen. – In: Heinrich-Tamáska (ed.) – Leiden. 2011a, 211–248. BUGAJ et al. 2013 = Bugaj, U., P. Lutovac, Z. Polak, M. BRATOŽ, R. 2011b, Die Auswanderung der Bevolkerung Trzeciecki 2013, Relics of masonry structures on aus den pannonischen Provinzen während des 5. Đuteza Hil . – Materialy i sprawozdania Rzeszow­ und 6. Jahrhunderts. – In: Konrad, Witschel (eds.) skiego Osrodka Archeologicznego 34, 149–165. 2011, 589–614. BUGARSKI et al. 2016 = Bugarski I., O. Heinrich- BRENK, B. 2003, Zur Christianisierung der spätrö- Tamáska, V. Ivanišević, D. Sybe (eds.) 2016, Grenz­ mischen Stadt im östlichen Mittelmeerraum. – In: Übergänge. Spätrömisch, frühchristlich, frühbyzan­ Brands, Severin (eds.) 2003, 85–95. tinisch als Kategorien der historisch­archäologischen 392 BIBLIOGRAPHY Forschung an der mittleren Donau. – Forschungen dioevo. – In: La Fin de la cité antique et le début de la zu Spätantike und Mittelalter 4. cité medievale: de la fin du IIIe siècle à l’avènement de BUNGURI, A. 2007, Gjetje në Kalanë e Grazhdanit (Finds Charlemagne. Actes du colloque tenu à l’Université at Grazhdan’s castle). – Monumentet, 100–121. de Paris X Nanterre (1993) , Bari, 239–271. BÜLOW, von G. 1995a, Die Entwicklung des Siedlungs- CANTINO WATAGHIN, G. 2004, La città tardoantica: bildes von Iatrus in der Periode B/C. – In: Iatrus- il caso di Aquileia. – Antichità Altoadriatiche 59, Krivina 1995, 29–54. 101–119. BÜLOW, von G. 1995b, Die Siedlungsperiode D 1 in CARRIE, J.-P., D. MORREAU 2015, The Archaeology of Iatrus. – In: Iatrus-Krivina 1995, 55–60. the Roman Town of Abritus: The Status Quaestionis BÜLOW, von G. 1995c, Die Siedlungsperiode D 2 in in 2012. – In: L. Vagalinski, N. Sharankov (eds.), Iatrus. – In: Iatrus-Krivina 1995, 61–66. Limes XXII. Proceedings of the 22nd international BÜLOW, von G. 2011, Romuliana – Gamzigrad. Ort Congress of Roman Frontier Studies (Ruse, Bulgaria, der Errinerung oder Herrschaftsort? – In: Bülow, September 2012), Bulletin of the National Archaeo- Zabehlicky (eds . ) 2011, 153–163. logical Institute 42, Sofia, 601–610. BÜLOW, von G., K. WACHTEL 2015, Zur Geschichte CATALÁN et al. 2014 = Catalán, R., P. Fuentes Melgar, der Grabung Iatrus-Krivina. – Bul etin of the Na­ J. C. Sastre Blanco (eds.) 2014, Las Fortificaciones tional Institute of Archaeology 17, 203–216. en la Tardoantigüedad. Élites y articulatión del ter­ BÜLOW, von G., H. ZABEHLICKY (eds.) 2011, Bruckritorio (siglos V­VIII D. C.). – Colleción simposia 5, neudorf und Gamzigrad. Spätantike Paläste und Madrid. Grossvil en im Donau­Balkan­Raum. – Akten des CAVADA, E. 2019, Trento in eta tardoantica e altomedi- Internationalen Kolloquiums in Bruckneudorf vom evale (IV – Vi secolo), il dato archeologico. Status 15. bis 18. Oktober 2008, London. questionis. – Antichità Altoadriatiche 90, 97–119. BÜLOW et al. 2007 = Bülow, von G., B. Böttger, S. Conrad, CAVADA, E., M. ZAGERMANN (eds.) 2020a, Alpine B. Döhle, G. Gomolka-Fuchs, E. Schönert-Geiss, Festungen 400–1000. Chronologie, Räume und Funk­ D. Stančev, K. Wachtel 2007, IATRUS­KRIVINA. tionen, Netzwerke, Interpretationen / Fortezze Alpine Spätantike Befestigung und frühmittelalterliche Sied­ (secoli V–X). Cronologia, spazi e funzioni, sistemi, lung an der unteren Donau. Bd. VI: Ergebnisse der interpretazioni. Akten des Kol oquiums in München Ausgrabungen 1992–2000. − Limesforschungen 28. am 13. und 14. September 2018. – Münchner Be- BUORA, M. 1992, I dati archeologici sul popolamento iträge zur Vor- und Frühgeschichte 68. del settore alpino in epoca romana. – In: Castelrai­ CAVADA E., M: ZAGERMANN 2020b, Die spätantike mondo. Scavi 1988–1990, I. Lo scavo, Cataloghi e Festung auf dem Monte San Martino (Lomaso, Tren- monografie dei Civici musei di Udine 2 , 99–109. tino). Ein italienisch-deutsches Forschungsprojekt BUORA, M. 2010, Attimis. – Peuce 8, 185–202. seit 2008. – In: Cavada, Zagermann 2020a, 451–480. BURKHARDT, N., R. H. W. STICHEL (eds.) 2010, Die CAVALIERI MANASSE, G., B. BRUNO 2003, L’edilizia antike Stadt im Umbruch. Kolloquium in Darm- abitativa a Verona. – In: J. Ortalli, M. Heinzelmann stadt (19. bis 20. Mai 2006), Wiesbaden. (eds.), Abitare in città. La Cisalpina tra impero e BUSULADŽIĆ, A. 2008, Zbirka fibula iz Mogorjela (The medioevo / Leben in der Stadt. Oberitalien zwischen Fibulae Collection from Mogorjelo). – Opuscula römischer Kaiserzeit und Mittelalter. Convegno / archaeologica 32, 2008 (2009), 21–54. Kolloquium (Roma, 4-5 novembre 1999), 47–64. BUSULADŽIĆ, A. 2011, Rimske vile u Bosni i Hercego­ CEDILNIK, A. 2004, Ilirik med Konstantinom Velikim in vini / Roman Vil as in Bosnia and Herzegovina. – Teodozijem Velikim: Balkansko­podonavski prostor Zemaljski muzej Bosne i Hercegovine, Sarajevo. v poročilih Atanazija, Hilarija, Sokrata Sholastika, BUTCHER, K. 2003, Roman Syria and the Near East. – Sozomena, Teodoreta in Filostorgija. – Thesaurus London. memoriae, Dissertationes 3. BUZOV, M. 2010, Antički gradovi u kontinentalnoj CEROVA, Y. 1987, La forteresse de Qafa dans la region Hrvatskoj. – Izdanja Hrvatskog arheološkog društva de Sulova. – Iliria 17/2, 155–185. 25, 137–171. CHAVARRÍA, A., T. LEWIT 2004, Recent research CAILLET, J. P. 1993, L’évergétisme monumental chrétien on the late antique countryside: a bibliographical en Italie et à ses Marges. – Collection de l’école essay. – In: Bowden et al. (eds.) 2004, 3–51. française de Rome 175. CHAVARRÍA ARNAU, A. 2007, El final de las vil as CAMBI, N. 2002, Antika. Povijest umjetnosti u Hrvaten Hispania (siglos IV–VIII). – Bibliothèque de skoj. – Zagreb. l’Antiquité Tardive 7. CAMERON, A. 2012, The Mediterranean World in Late CHEVALIER, P. 1995, Salona II: Ecclesiae Dalmatiae. – Antiquity, AD 395­600. – London, New York. Rome. CANTINO WATAGHIN, G. 1996, Quadri urbani CHEVALIER, P. 2015, Les villes de l’actuelle Albanie et nel ’Italia settentrionale: tarda Antichità e alto Me- leur évolution aux ve-viie siècles –quelques réflex- 393 Slavko CIGLENEČKI ions. – In: Y. Marion, F. Tassaux (eds.), AdriAtlas Die Völker Südosteuropas im 6. bis 8. Jahrhundert, et l’histoire de l’espace adriatique du VIe s. a.C. au Südosteuropa-Jahrbuch 17, 265–286. VIIIe s. p.C. , Actes du colloque international de CIGLENEČKI, S. 1990a, K problemu datacije nastanka Rome (4-6 novembre 2013), 227–245. rimskodobnih višinskih utrdb v jugovzhodnoalp- CHEVALIER, P., N. BEAUDRY 2018, Une ville du VIe skem prostoru (Zum Datierungsproblem der Entste- siecle retournant àla ruralité: desurbaniaiation et hung der römerzeitlichen Höhenbefestigungen im abandon du siège épiscopal de Byllis. – In: J-L. Lam- Südostalpenraum). – Arheološki vestnik 41, 147–176. boley, L. Përzhita, A. Skenderaj, L’Il yrie Méridionale CIGLENEČKI, S. 1990b, Zidani gaber, Novo mesto. – et l’Épire dans l’antiquité VI. Actes du VIe colloque In: Arheološka najdišča Dolenjske, Arheo. Dossier international de Tirana (20 – 23 mai 2015), 435–448. Dolenjska, 113–116. CHEVALIER, P., J. MARDEŠIĆ 2006, La ville de Salone CIGLENEČKI, S. 1993, Arheološki sledovi zatona dans l’Antiquité tardive: déprise spatiale, mutations antične Petovione. – In: Ptujski arheološki zbornik, et renouveau de la parure monumentale. – Hortus Ptuj, 505–520. Artium Medievalium 12, 55–68. CIGLENEČKI, S. 1994, Höhenbefestigungen als Sied- CHRISTIE, N. 2000, Towns, land and power: German- lungsgrundeinheit der Spätantike in Slovenien. – Roman survivals and interactions in fifth- and Arheološki vestnik 45, 239–266. sixth-century Pannonia. – In: Brogiolo et al. 2000, CIGLENEČKI, S. 1997, Strukturiranost poznorimske 275–297. poselitve Slovenije / Strukturierung spätantiker CHRISTIE, N. (ed.) 2004, Landscapes of Change. Rural Besiedlung Sloweniens. – Arheološki vestnik 48, Evolutions in Late Antiquity and the Early Middle 191–202. Ages. – London. CIGLENEČKI, S. 1999, Results and problems in the CHRISTIE, N. 2007, From the Danube to the Po: Late archaeology of the late antiquity in Slovenia / Roman defence in Pannonia and Italy. – In: Poulter Izsledki in problemi poznoantične arheologije v (ed.) 2007a, 547–577. Sloveniji. – Arheološki vestnik 50, 287–309. CHRISTIE, N., A. AUGENTI 2012, Vrbes Extinctae: CIGLENEČKI, S. 2000, Tinje nad Loko pri Žusmu: Archaeologies of Abandoned Classical Towns. – poznoantična in zgodnjesrednjeveška naselbina / Ashgate. Tinje oberhalb von Loka pri Žusmu: Spätantike CIGLENEČKI, S. 1976, Kasnoantično utrjeno naselje und frühmittelalterliche Siedlung. – Opera Instituti Gradec pri Prapretnem (Eine spätantike befestigte Archaeologici Sloveniae 4. Siedlung am Hügel Gradec beim Ort Prapretno). CIGLENEČKI, S. 2003a, The basic characteristics of the – Areološki vestnik 26 , 259–267. late antique settlement pattern within the eastern CIGLENEČKI, S. 1979, Kastel, utrjeno naselje ali refugij? Alpine region and Dalmatia. – Histria antiqua 11, (Kastel , befestigte Siedlung oder Refugium?) – 263–281. Arheološki vestnik 30, 459–472. CIGLENEČKI, S. 2003b, Frühchristliche Kirchenanlagen CIGLENEČKI, S. 1981, Rezultati prvih raziskovanj in Slowenien. – In: H. R. Sennhauser (ed.), Frühe na Gradcu pri Prapretnem (Ergebnisse der er- Kirchen im östlichen Alpengebiet. Von der Spätantike sten Forschungen auf Gradec bei Prapretno). – bis in ottonische Zeit, Abhandlungen der Bayerischen Arheološki vestnik 32, 417–453. Akademie der Wissenschaften, Philosophisch- CIGLENEČKI, S. 1985, Potek alternativne ceste Siscija- historische Klasse N. F., H 123, 581–595. Akvileja na prostoru zahodne Dolenjske in Notranj- CIGLENEČKI, S. 2005, Langobardische Präsenz im ske v času 4. do 6. stoletja. Preliminarno poročilo o Südostalpen im Lichte neuer Forschungen. – In: W. raziskovanjih Korinjskega hriba in rekognosciranjih Pohl, P. Erhart (eds.), Die Langobarden: Herrschaft zahodne Dolenjske (Der Verlauf der Alternativs- und Identität. Forschungen zur Geschichte des trasse Siscia-Aquileia im Raum von Westdolenjsko Mittelalters 9, 265–280. und Notranjsko in der Zeitspanne vom 4. Bis zum CIGLENEČKI, S. 2008, Castra und Höhensiedlungen 6. Jh. Präliminarbericht über die Erforschung des vom 3. bis 6. Jahrhundert in Slowenien. – In: Steuer Korinjski hrib und die Rekognoszierungen von et al. (eds.) 2008, 481–532. Westdolenjsko). – Arheološki vestnik 36, 255–284. CIGLENEČKI., S. 2011a, Von römischen Städten zu CIGLENEČKI, S. 1986, Zgodnjekrščansko središče spätantiken Höhensiedlungen zentralörtlichen na Kučarju v Beli Krajini. − Izdanja Hrvatskog Charakters. – In: J. Macháček, Š. Ungerman (eds.), arheološkog društva 10, 137−150. Frühgeschichtliche Zentralorte in Mitteleuropa, Stu- CIGLENEČKI, S. 1987a, Höhenbefestigungen aus der dien zur Archäologie Europas 14, 186–192. Zeit vom 3. bis 6. Jh. im Ostalpenraum. – Dela 1. CIGLENEČKI, S. 2011b, Problem odsotnosti cerkva razreda SAZU 31. v nekaterih poznoantičnih kastelih Ilirika (The CIGLENEČKI, S. 1987b, Das Weiterleben der Spätantike Problem of Absence of Churches in some Late bis zum Auftauchen der Slawen in Slowenien. – In: Antique Fortifications in Illyricum). – In: Zbornik 394 BIBLIOGRAPHY u čast Emilija Marina za 60. rođendan / Miscel anea Stadt. Oberitalien zwischen römischer Kaiserzeit Emilio Marin sexagenario dicata, Zbornik Kačić und Mittelalter. Convegno / Kolloquium (Roma, 41–43, 673–688. 4-5 novembre 1999), 37– 56. CIGLENEČKI, S. 2012, Spremenjena podoba pozno- CLAUDE, D. 1969, Die Byzantinische Stadt im 6. Jahr­ antičnih urbanih središč – prispevek k transformaciji hundert. – München. poselitvene slike v jugovzhodnoalpskem prostoru. – CONSTANT et al. 2015 = Constant A., J.-A. Ségura, M. In: I. Lazar, B. Županek (eds.), Emona med Akvilejo Valenciano 2015, Hil top settlement dynamics in in Panonijo ( Emona between Aquileia and Panno­ Provence between the 5th-9th centuries. Results and nia), Annales Mediterranei, Koper, 459–479. research prospects. – In: M. Vicelja-Matijašić (ed.), CIGLENEČKI, S. 2014, The Changing Relations Swords, Crowns, Censers and Books. Francia Media ­ Between City and Countryside in Late Antique Cradles of European Culture, Rijeka, 374–403. Illyricum (Mijene u relacijama između gradova i CORAZZA et al. 2001 = Corazza, S., P. Donat, F. Oriolo unutrašnjosti u kasnoantičkom Iliriku). – Hortus 2001, Iulium Carnicum. Centro alpino tra Italia e Artium Medievalium 20/1, 232–250. Norico dal a protostoria al ’età imperiale. Atti del CIGLENEČKI, S. 2015, Late Roman army, Claustra Convegno (Arta Terme – Cividale, 29–30 settembre Alpium Iuliarum and the fortifications in the south- 1995). – Studi e ricerche sul a Gallia Cisalpina 13. eastern Alps. – In: Istenič et al. (eds.) 2015, 385–430. CRAWFORD, S. 1990, The Byzantine Shops at Sardis. – CIGLENEČKI, S. 2016, Claustra Alpium Iuliarum, trac­ Archaeological Exploration of Sardis, Monograph 9. tus Italiae circa Alpes and the defence of Italy in the CROCE DA VILLA, P. 2003, Concordia. – In: Abitare in final part of the Late Roman period. – Arheološki città: la Cisalpina tra impero e medioevo / Leben in vestnik 67, 409–424. der Stadt: Oberitalien zwischen römischer Kaiserzeit CIGLENEČKI, S. 2020, Namen utrdbe Korinjski hrib und Mittelalter. Convegno / Kolloquium (Roma, in pregled zadnjih antičnih utrdb z vojaškimi 4-5 novembre 1999), 65–72. nalogami / The function of the Korinjski hrib CROW, J. 2001, Fortifications and urbanism in late fort and an overview of the last antique forts with antiquity: Thessaloniki and other eastern cities. military tasks. – In: Ciglenečki et al. 2020, 259–271. – In: L. Lavan, Recent research in Late­antique CIGLENEČKI, S. 2021, Poznoantična semiurbana Urbanism, Journal of Roman Archaeology, Suppl. utrjena naselbina Sv. Pavel nad Vrtovinom – 55 ser. 42, 89–105. let pozneje (Late Antique Fortified Semi-Urban CROW, J. 2002, Late antique cities in eastern Thrace Settlement Sv. Pavel above Vrtovin – 55 Years (Europa). – In: Slokoska et al. 2002, 342–351. Later). − Goriški letnik 45, 193−214. CUNJA, R. 1996, Poznorimski in zgodnjesrednjeveški CIGLENEČKI, S., Z. MODRIJAN 2020, Bronze chande- Koper. Arheološko izkopavanje na bivšem Ka­ liers from Vipota (Slovenia) and the decline of the pucinskem vrtu v letih 1986–1987 v luči drobnih churches of Celeia . – Hortus Artium Medievalium najdb 5. do 9. stoletja / Capodistria tardoromana 26, 141–150. e altomedievale. Lo scavo archeologico nelléx orto CIGLENEČKI et al. 2011 = Ciglenečki, S., Z. Modrijan, dei cappuccini negli anni 1986­al a luce dei repertivi T. Milavec 2011, Poznoantična utrjena naselbina dal v al IX secolo D.C. – Koper. Tonovcov grad pri Kobaridu. Naselbinski ostanki CURTA, F. 2001a, Peasants as ‘makeshift soldiers for the in interpretacija / Late antique fortified settlement occasion’: sixth-century settlement patterns in the Tonovcov grad near Kobarid. Settlement remains Balkans. – In: Th. S. Burns, J. W. Eadie, Urban Centand interpretation. – Opera Instituti Archaeologici ers and Rural Contexts in Late Antiquity, 199–217. Sloveniae 23. CURTA, F. 2001b, The Making of the Slavs. History CIGLENEČKI et al. 2020 = Ciglenečki, S., Z. Modrijan, and Archaeology of the Lower Danube Region c. T. Milavec 2020, Korinjski hrib in poznoantične 500–700. – Cambridge. vojaške utrdbe v Iliriku / Korinjski hrib and late CURTA, F. 2001c, Limes and cross: the religious dimen- antique military forts in Il yricum. − Opera Instituti sion of the sixth-century Danube frontier of the Archaeologici Sloveniae 39. early Byzantine empire. – Starinar 51, 45–70. CIRELLI, A. 2008, Ravenna: archeologia di una città. − CURTA, F. 2013, Horsemen in forts or peasants in vil- Sesto Fiorentino. lages? Remarks on the archaeology of warfare in the CIRELLI, A. 2010, Ravenna: Rise of a Late Antique Capi- 6th to 7th century Balkans. – In: Sarantis, Christie tal. − In: S. Dani, G. Speed (eds.), Debating Urbanism. (eds.) 2013, 809–850. Within and Beyond the Wal s A.D. 300−700, Leicester CURTA, F. 2014, The Edinburgh History of the Greeks, C. Archaeology Monographs 17, 239−263. 500 to 1050. The Early Middle Ages. – Edinbourgh. CIURLETTI, G. 2003, Il caso Tridentum. – In: J. Or- CURTA, F. 2017, Coins, forts and commercial exchanges talli, M. Heinzelmann (eds.), Abitare in città. La in the sixth- and early seventh-century Balkans. – Cisalpina tra impero e medioevo / Leben in der Oxford Journal of Archaeology 36(4), 439–454. 395 Slavko CIGLENEČKI CUSCITO, G. 2001, “In Castra gradensi ac plebe sua”: muzeja Bosne i Hercegovine u Sarajevu, n.s. 42/43 Lo sviluppo del castrum di Grado dalla tarda (1987−1988), 83–128. antichità al ’alto medioevo. – In: Paolo Diacono ČREMOŠNIK, I. 1990, Rimska utvrđenja u BiH s 2021, 387–406. osobitim osvrtom na utvrđenja kasne antike (Der CVETKOVIĆ-TOMAŠEVIĆ, G. 2008 = ЦВЕТКОВИЋ- römischen Befestigungen in Bosnien und Her- ТОМАШЕВИЋ, Г., Епископски двор и манастир zegowina mit besonderer berücksichtigung der у ранохришчанској Хераклеји (Episkopal Anlagen der Spätantike). – Arheološki vestnik 41, court and the monastery in the Early Christian 355–364. Heraclea). – In: Археолошките откритија ĆURČIĆ, S. 2010, Architecture in the Balkans from Dio­ на почвата на Македонија ( Arheološkite otcletian to Süleyman the Magnificent. – Yale. kritija na počvata na Makedonija), Прилози за DAGRON, G. 1979, Entre vil age et cite’ la bourgade истражувањето на историјата на културата на rurale des IV-VIIe siècles en Orient. – Koinônia 3, почвата на Македонија (Prilozi za istražuvanjeto 29–52. na istorijata na kulturata na počvata na Make- DAGRON, G. 1984, Les villes dans l’Illyricum proto- donija) 19, 823–834. byzantin. – In: Villes 1984, 1–20. ČAUŠEVIĆ-BULLY, M., S. BULLY 2020, Najnovija DAIM, F., J. DRAUSCHKE (eds.) 2011, Byzanz – das istraživanja ranokršćanskih ‘crkvenih sklopova’ Römerreich im Mittelalter 2 (1, 2). – Mainz. Kvarnera: na razmeđi arhitekture i teritorija. – In: DECKER, M. 2017, Approaches to the environmental M. Sanader, D. Tomčinić, I. Kaić, V. Maitjević (eds.), history of Late Antiquity II: Climate Change and Zbornik I. skupa hrvatske ranokršćanske arheologije the End of the Roman Empire. − History Compass (HRRANA), Zagreb, 271–287. 15/10, 1−8. ČIČIKOVA, M. 1994, Novae a l’epocque du Bas-Em- DE FRANCESCHINI, M. 1998, Vil e Romane del a X pire. – In: G. Susini (ed.), Limes, Studia di storia Regio Venetia et Histria. – Studia Archaeologica 93. 5, 127–138. DEICHMANN, F. W., U. PESCHLOW 1977, Zwei ČOVIĆ, B. (ed.) 1988, Arheološki leksikon Bosne i Herspätantike Ruinenstätten in Nordmesopotamien. – cegovine 1–4. – Sarajevo. Bayerische Akademie der Wissenschaften, Philos- ČREMOŠNIK, I. 1955a, Nova antička istraživanja kod ophisch-Historische Klasse 2. Konjica i Travnika. – Glasnik Zemaljskog muzeja DELOGU, P., S. GASPARINI (eds.) 2010, Le trasformazi- Bosne i Hercegovine u Sarajevu, n.s. 10, 122–127. oni del V secolo. L’Italia, i barbari e l’Occidente ČREMOŠNIK, I. 1955b, Novi srednjevjekovni nalazi romano. – Atti del Seminario di Poggibonsi, 18-20 kod Prijedora. – Glasnik Zemaljskog muzeja Bosne ottobre 2007, Venezzia. i Hercegovine u Sarajevu, n.s. 10, 137–147. DELONGA, V. 1984, Prilog arheološkoj topografiji ČREMOŠNIK, I. 1956, Rimski ostaci na Gradini Zecovi. Mokrog polja kod Knina. – Starohrvatska prosvjeta – Glasnik Zemaljskog muzeja Bosne i Hercegovine 14, 259–283. u Sarajevu, n.s. 11, 137–146. DEMANDT, A. 1989, Die Spätantike. Römische Geschichte ČREMOŠNIK, I. 1957, Dalja istraživanja na rimskom von Diocletian bis Justinian 284 bis 565 n.Chr. – naselju u Lisičićima. – Glasnik Zemaljskog muzeja Handbuch der Altertumswissenschaft III/6. Bosne i Hercegovine u Sarajevu, n.s. 12, 143–162. DEMO, Ž. 1994, Ostrogothic coinage from col ections ČREMOŠNIK, I. 1958, Arheološka istraživanja u okolici in Croatia, Slovenia and Bosnia & Herzegovina. – Bihaća. – Glasnik Zemaljskog muzeja Bosne i Her­ Situla 32. cegovine u Sarajevu, n.s. 13 , 117–136. DI VITA, A. 2010, Gortina di Creta. Quindici secoli di ČREMOŠNIK, I. 1959, Izkopavanja u Ljusini 1957 vita urbana. – Roma. godine. – Glasnik Zemaljskog muzeja Bosne i Her­ DIMITROVA, E. 2012, = ДИМИТРОВА, E., Bини- cegovine u Sarajevu, n.s. 14, 137–147. ч ката мистерија. Керамичката ризница од ČREMOŠNIK, I. 1965, Rimska vila u Višićima (Die доцноантичкото кале / The Vinica Mystery. The Ce­ römische Vil a in Višići). – Glasnik Zemaljskog ramics Treasures of a Late Antique Fortress. − Vinica. muzeja Bosne i Hercegovine u Sarajevu, n.s. 20, DINTCHEV, V. 1997 = ДИНЧХЕВ, В., Римските вили в 147–260. днешната българска територия (Roman Roman ČREMOŠNIK, I. 1974, Prvi tragovi kršćanstva na vil as in the present­day Bulgarian lands). – Sofia. nalazima vile u Paniku (Die erste Spuren des DINTCHEV, V. 1999, Classification of the Late Antique Christentums in den Funden von Panik). – Situla Cities in the Dioceses of Thracia and Dacia. – 14/15, 243−248. Archaeologia Bulgarica 3/3, 39–73. ČREMOŠNIK, I. 1976, Panik. – Glasnik Zemaljskog muze­ DINTCHEV, V. 2000, The Limit of Urban Life in the ja Bosne i Hercegovine u Sarajevu, n.s. 29, 41−164. Late Antique Dioceses of Thracia and Dacia: the ČREMOŠNIK, I. 1989, Rimsko utvrđenje na Gradini Overestimated Centres. – Archaeologia Bulgarica u Biogracima kod Lištice. – Glasnik Zemaljskog 4/2, 65–84. 396 BIBLIOGRAPHY DINTCHEV (= DINCHEV), V. = ДИНЧХЕВ, В., DJURIĆ, B. 2005, I primi mosaici paleocristiani di Ранновизантийските крепости в България и Emona (Venetia e Histria) e la loro officina. – In: H. съседните земи (в диоцезите Thracia и Dacia) Morlier, C. Bail y, M. Tahri, D. Janneteau (eds.), La (Early Byzantine Fortresses in Bulgaria and its mosaique gréco­romaine IX, I, Collection de l’École Neighbouring Territories (In the Dioceses of française de Rome 352, 663–676. Thracia and Dacia). – Разкопки и проучвания DJURIĆ, S. 1994, Mosaic of philosophers in an early (Razkopki i proučivanja) 35. Byzantine vil a at Nerodimlje. – In: VI Coloquio DINTCHEV (= DINCHEV), V. 2007, The Fortresses of Int. sobre Mosaico Antiguo (Palencia-Merida, Oct. Thrace and Dacia in the Early Byzantine Period. – 1990), Guadalajara, 123–134. In: Poulter (ed.) 2007a, 479–546. DÖHLE, B. 1995, Die Siedlungsperiode A in Iatrus. – In: DINTCHEV, V. 2008, Habitations représentatives de Iatrus-Krivina 1995, 9–28. l’antiquité tardive, découvertes dans la région du Bas DOLENZ, H., W. GÖRLICH 1935, Eine spätantike Danube. – In: P. Scherrer (ed.), Domus: das Haus in Fliehburg in Warmbad Vil ach. – Carinthia I, 125, den Städten der römischen Donauprovinzen. Akten 133–140. des 3. Internationalen Symposiums über römische DOLENZ et al. 2016 = Dolenz, H., J. Bonetto, D. Ebner- Städte in Noricum und Pannonien [vom 19 bis 22. Baur, A. Fiamozzi, C. Flügel, K. Strobel 2016, Der April 2005 in St. Pölten], Wien, 405–416. Bischofssitz und die spätantike Stadt Virunum. – DINTCHEV, V. 2009 = ДИНЧХЕВ, В., Новите Römisches Österreich 39, 47–172. късноримски центрове на Scythia и Moesia DOMIJAN, M. 1983, Crkva Sv. Andrije na Vrgadi nakon secunda (Novitе kăsnorimski tsеntrovе na Scythia posljednjih istraživačkih i konzervatorskih radova. i Moesia secunda). – In: Сборник в памет на – Starohrvatska prosvjeta 13, 123–138. професор Велизар Велков (Sbornik v pamеt na DÖRFLER et al. (eds.) 2017 = I. Dörfler, P. Gleirscher, profеsor Vеlizar Vеlkov), Sofia, 414–448. S. Ladstätter, I. Pucker (eds.) 2017, AD AMUSSIM. DINTCHEV, V. 2012, The East-Balkan barrier wall and Festschrift zum 65. Geburtstag von Franz Glaser. – the “Gates of Haemus” complex. – In: L. Vagalinski, Kärntner Museumsschriften 85. N. Sharankov, I. Torbatov (eds.), The Lower Dan- DUBOLNIĆ, M. 2007, Argyruntum i njegov teritorij ube Roman Limes (1st – 6th c. AD), Sofia, 495–526. u antici. – Zavod za povijesne znanosti HAZU u DINTCHEV, V. 2016, The town and the church during Zadru 49. the late antiquity period. Architectural and urban DULAR, J. 1978, Early christian church on Kučar near dimensions. – In: Zbornik za Boris Borisov, Sofia, Podzemelj. – Arheološki vestnik 29, 528–532. 309–324. DULAR et al. = Dular, J., S. Ciglenečki, A. Dular 1995, DINTCHEV, V. 2018, Town and Church in Late Antiquity. Kučar. Železnodobno naselje in zgodnjekrščanski Architectural and Urbanistic Dimensions. – Pro- stavbni kompleks na Kučarju pri Podzemlju / ceedings of the first International Roman and Late Eisenzeitliche Siedlung und frühchristlicher Gebäu­ Antiqe Thrace Confrerence “Cities, territories and dekomplex auf dem Kučar bei Podzemelj. – Opera Identities” (Plovdiv 3rd – 7th October 2016 ). Bulletin Instituti Archaeologici Sloveniae 1. of the National Archaeological Institute 44, 357–370. DUNN, A. 1994, The transition from polis to kastron DINTCHEV, V. 2021a, The Provincial Capitals of the in the Balkans (III-VII vv): General and regional Late Antique Dioceses of Thracia and Dacia. – In: perspectives. – Byzantine and Modern Greek Studies M. Raycheva, M. Steskal (eds.), Roman Provincial 18, 60–80. Capitals Under Transition. Proceedings of the DUNN, A. 2002, Was there a militarization of the International Conference (Held in Plovdiv 04.-07. southern Balkans during late antiquity? – In: Limes November 2019), Sonderschriften des Österreichis- XVIII. Proceedings of the XVIIIth International chen Archäologischen Institutes 61. Holzhausen – Congress of Roman Frontier Studies held in Am- Wien, 233–254. man, Jordan (September 2000), BAR International DINTCHEV, V. 2021b = ДИНЧХЕВ, В., Днешната Series 1084, Vol. 2, 705–712. българска територия през Късната античност DUNN, A. 2004, Continuity and Change in the macedo- (The present-day Bulgarian territory in Late Antiqui- nian Countryside from Galienus to Justinian. – In: ty). – In memoriam Theophili Ivanov. Bul etin of the Bowden et al. (eds.) 2004, 535–586. National Institute of Archaeology XLVIII, 273–312. DUNN, A. 2005, The Problem of the Early Byzantine DJURA JELENKO, S. 2020, Colatio – Stari trg pri Slovenj Vil age in Eastern and Northern Macedonia. − In: J. Gradcu. – In: Horvat et al. (eds.) 2020a, 305–324. Lefort, C. Morrisson, J-P. Sodini (eds.), Les Vil ages DJURIĆ, B. 1976, Antični mozaiki na ozemlju SR dans l’Empire byzantin (IVe-XVe siècle), Réalités Slovenije (Les mosaïques antiques sur le teritoire de byzantines 11, 267−278. la R. S. Slovénie). – Arheološki vestnik 28, 537–626. 397 Slavko CIGLENEČKI DUPERRON, G. 2013, Le mobilier céramique tardo­ EITLER, J. 2017, Neues zu bekannten Befunden - Über- antique (Ve et VIe s.) de l’oppidum de Constantine. − legungen zum Forum von Teurnia. – In: Dörfler et Lançon-Provence, Bouches-du-Rhône. al. (eds.) 2017, 87–103. DUVAL, N. 1989–1990, Mogorjelo: camp ou palais? – EITLER, J., Y. SEIDEL 2022, Das Gipfelplateau des Starinar 40, 253–259. Hemma bergs. Bruch ­ Wandel ­ Kontinuität. DUVAL, N. 1996, L’urbanisme de Caričin Grad. Une ville Ergebnisse des FWF­Projekts Cult Continuity at artificielle et ses bâtiments d’apparat: une spécificité the Summit of Hemmaberg (P29452­G25). (Online locale ou une étape décisive dans la typologie des publication; last access: 23rd March 2023) [https:// principia militaires? – Antiquité tardive 4, 325–339. landesmuseum.ktn.gv.at/aktuelles?nid=241]. DVORŽAK SCHRUNK, I. 1989, Dioklecijanova palača ELLIS, S. P. 2000, Roman Housing. – London. od 4. do 7. st. – Vjesnik Arheološkog muzeja u Za­ ELLIS, S. 2007, Late Antique Housing and the Uses of grebu 22, 91–105. Residential Buildings: An Overview. – In: Lavan et DYGGVE, E., H. VETTERS 1966, Mogorjelo. Ein al. (eds.) 2007, 1–22. spätantiker Herrensitz im römischen Dalmatien. – FABER, A. 1988, Osvrt na neka utvrđenja otoka Krka Österreichische Akademie der Wissenschaften, od vremena prethistorije do antike i srednjeg Schriften der Balkankommision, Antiquarische vijeka. – Prilozi Instituta za arheologiju u Zagrebu Abteilung 13. 3/4, 113–140. DŽIN, K. 2006, Stancija Peličeti. Roman Vil a Rustica – FABIJANIĆ et al. 2012 = Fabijanić, T., M. Međušić, new find at the Istrian “Y” highway corridor. – I. Radić Rossi 2012, Dokumentiranje postoječeg Izložba / Exhibition. Arheološki muzej Istre, Pula. stanja i istraživanja utrvrde Turete, luke i sakralnog DŽIN, K., V. GIRARDI JURKIĆ 2005, Rimska gospodar­ kompleksa na prostoru crkve Gospe od Tarca na ska vila u Červar Portu kod Poreča / Roman vil a otoku Kornatu. – Histria antiqua 22, 401–412. rustica in Červar near Poreč. – Izložba / Exhibition. FADIĆ, I. 1995, Starigrad u antici / Starigrad in ancient Arheološki muzej Istre, Pula. times. – In: Paklenički zbornik, 229–236. DŽIN, K., V. GIRARDI JURKIĆ 2008, Vižula i Burle FADIČ, I. 2001, Bedemi Aserije. – Histria Antiqua 7, u antici. – Izložba / Exhibition. Arheološki muzej 69–89. Istre, Pula. FEISSEL, D. 1988, L’architecte Viktorinos et les fortifica- DŽIN, K., I. MIHOLJEK 2019, Medulin Bay in Late tions de Justinien dans les provinces balkaniques. – Antiquity: The Antique and Late Antique Site of Bulletin de la Société Nationale des Antiquaires de Vižula near Medulin, Croatia. – In: E. Cirelli, E. France, 136–146. Giorgi, G. Lepore (eds.), L’Adriatico centrale tra FEISSEL, D. 2000, Les edifices de Justinien au temo- tarda Antichità e alto Medioevo, BAR International ignage de Procope et de l’epigraphie. – Antiquité Series 2926, 505–510. Tardive 8, 81–104. ĐIDROVA et al. 2011 = Đidrova, L., E. Nasuh, L. FEKEŽA, L. 1990, Kasnoantičko utvrđenje i crkva na Lozanovski 2011= Ђидрова, Л., Е. Насух, Л. lokalitetu Gradac na Ilinjači u Gornjem Kotorcu Лозановски 2011, Водоснабдувањето на Heraclea kod Sarajeva. – Glasnik Zemaljskog muzeja Bosne Lincestis (The Water Supply of Heraclea Lyncestis). i Hercegovine u Sarajevu n.s. 45, 155–170. – Macedoniae acta archaeologica 20, 381–404. FEKEŽA, L. 1991, Kasnoantičko utvrđenje na lokalitetu EGGER, R. 1914, Ausgrabungen in Norikum 1912/13. – Gradac na Ilinjači u Gornjem Kotorcu kod Sarajeva Jahreshefte des Österreichischen Archäologischen (nastavak), Glasnik Zemaljskog muzeja Bosne i Instituts in Wien 17, Beiblatt, 59–86. Hercegovine u Sarajevu n.s. 46, 175–202. EGGER, R. 1916, Frühchristliche Kirchenbauten im FEKEŽA, L. 1998, Ranokršćanske crkve u refugijima südlichen Noricum. – Sonderschriften des Öster- i u nizinskim naseljima na području centralne reichischen Archäologischen Institutes in Wien 9. Bosne. – In: Acta XIII Congressus 1998, 251–260. EGGER, R. 1929, Ausgrabungen in Feistritz a.d. Drau, FELGENHAUER-SCHMIEDT, S. 1993, Das Kappele Oberkärnten. Der Hügel bei Duel. – Jahreshefte des (“die Kåpile”) ob Jadersdorf. Eine spätantik­früh­ Österreichischen Archäologischen Institutes in Wien mittelalterliche Höhensiedlung in Oberkärnten. − 25, Beiblatt, 190−216. Aus Forschung und Kunst 27. EGGER, R. 1940, Vom Ursprung der romanischen FERNÁNDEZ PEREIRO, M., J. C. SÁNCHEZ-PARDO Chorturmkirche. – Wiener Jahreshefte 32, 85–99. 2022, Buscando un castillo altomedieval entre la EGGER, R. 1942, Die Ostalpen in der Spätantike, Das vegetación. El caso de Castro Valente (Padrón, neue Bild der Antike 2, 395ss. (Cited by Römische Galicia). – Cuadernos de Arqueología de la Univer­ Antike und frühes Christentum. Ausgewählte Schrifsidad de Navarra 30/2, 55–72. ten von Rudolf Egger 1, 257–271). FIALA, F. 1894, Jedna prehistorička naseobina na De- EGGER, R. 1950, Der Ulrichsberg. Ein heiliger Berg belom Brdu kraj Sarajeva. – Glasnik Zemaljskog Kärntens. – Carinthia I, 140, 9–78. muzeja Bosne i Hercegovine u Sarajevu 6, 107–140. 398 BIBLIOGRAPHY FILIPEC, K. 2007, 10 Jahre archäologischer Grabung in GÁBOR et al. 2004 = Gábor, O., G. Karpati, I. Lengvari, Lobor (1998 - 2007). – Hortus Artium Medievalium S. Pozsarko 2004, Sopianae. – In: Šašel Kos, Scher- 13/2, 411– 422. rer (eds.) 2004, 269–294. FILIPEC et al. 2020 = Filipec, K., O. Heinrich-Tamáska, GASSNER et al. 2002 = Gassner, V., S. Jilek, S. Ladsttätter M. Šimek 2020, New results from the eastern mar- 2002, Am Rande des Reiches. Die Römer in Öster­ gins of the Alps – Hil top settlements in Pannonia reich. − Österreichische Geschichte 15. in Late Antiquity. – In: Cavada, Zagermann (eds.) GASPARI, A. 2020, Ulaka. – In: Horvat et al. (eds.) 2020a, 199–224. 2020, 141−172. FINGAROVA, G. 2015, Das christliche Erscheinungs- GAVRILOVIĆ VITAS, N., B. POPOVIĆ 2015, Kasno- bild Serdicas. – In: R. J. Pillinger (ed.), Neue antički domus u Skelanima ( Municipium Malve­ Forschungen zum frühen Christentum in den siatium) (Late Antique Domus in Skelani ( Mu­ Balkanländern, Archäologische Forschungen 26, nicipium Malvesiatium)). – Starinar 65, 197–220. Denkschriften der philosophisch-historischen GELICHI et al. 2012 = Gelichi, S., C. Negrelli, S. Leardi, Klasse 484, 109–122. L. Sabbionesi, R. Belcari 2012, Doclea alla fine FISKOVIĆ, I. 1980, O ranokršćanskim spomenicima del ’Antichità. Studi e ricerche per la storia di una città naronitanskog primorja. – In: I. Marović, Ž. abbandonata del a Prevalitania / Duklja na kraju Rapanić (eds.), Dolina rijeke Neretve od prethistorije antičkog doba. Studije i istrazivanja istorije jednog do ranog srednjeg vijeka (Val ée du fleuve Neretva napuštenog grada Prevalitanije. – Nova antička Duk­ depuis la préhistoire jusqu’au début du Moyen Age) , lja III / New Antique Doclea III, Podgorica. Izdanja Hrvatskog arheološkog društva 5, 213–266. GEORGIEV, Z. 1989 = ГЕОРГИЕВ, З., Градиште, с. FISKOVIĆ, I. 1982, O ranokrščanskoj arhitekturi na Пчиња – Доцноантички и рановизантиски otocima Braču i Šolti. – Arheološki radovi i rasprave кастел (“Gradište”. Une forteresse de la basse 8–9, 159–216. antiquite et de la haute periode Byzantine pres du FISKOVIĆ, I. 1998, Late antique buildings in Polače vil age de Pčinja). – Macedoniae Acta Archaeologica on the island of Mljet. – In: Acta XIII Congressus 10 (1985–1986), 199–214. 1998, 273–286. GERASIMOVSKA, D. 2010 = ГЕРАСИMОВСКА, Д., FISKOVIĆ, I. 1999, Jesu li Polače na Mljetu bile sijelo Контрасти меѓу станбените објекти во доцната vladara Dalmacije? / Were Polače on the Isle of антика во Република Македонија (Kontrasti među Mljet the Seat of the Rulers of Dalmatia? – Prilozi stambenite objekti vo docnata antika vo Republika Instituta za arheologiju u Zagrebu 13–14 (1996– Makedonija). – Patrimonium. MK 3/7–8, 119–126. 1997), 61−82. GERASIMOVSKA, D. 2012, House Development in the FITZ, J. 1976, Gorsium – Herculia. – Szekesfehervar. Cities of the Republic of Macedonia throughout FITZ, J. 2004, Gorsium. – In: Šašel Kos, Scherrer (eds.) Antiquity. – Haemus Journal 1, 1–119. 2004, 197–208. GERASIMOVSKA, D. 2015, = ГЕРАСИMОВСКА, FORTIS, A. 1774, Viaggio in Dalmazia (cited after: Put Д., Aрхитектурата во доцната антика во po Dalmaciji, priredio Josip Bratulić, Zagreb - Lju- Mакедонските градови / Architecture in Late An- bljana 1984, [283–284]). tiquity in Macedonian cities. − Folia Archaeologica FREDERIKSEN et al. (eds.) 2016 = Frederiksen, R., Balkanica 3, 253–265. S. Müth, P. I. Schneider, M. Schnelle (eds.) 2016, GEUENICH, D., Th. ZOTZ 2008, Castra und Höhen- Focus on Fortifications. New Research on Fortifica­ siedlunhen in der schriftlichen Überlieferung von tions in the Ancient Mediterranean and the Near der Spätantike bis zur frühen Karolingerzeit. – In: East. – Oxford. Steuer et al. (eds.) 2008, 795–820. FÜLEP, F. 1984, Sopianae. The history of Pecs during the GHICA et al. 2019 = Ghica, V., A. Milošević, N. Uroda, Roman era, and the problem of continuity of the Late D. Džino 2019, Varvaria / Breberium / Bribir Roman population. – Archaeologia Hungarica 50. Archaeological Project. The 2016 excavation sea- FURLANI, U. 1969, Una stazione militare romana sul son. – Hortus Artium Medievalium 25/1, 143–163. Castel azzo di Doberdo del Lago. – Aquileia nostra GHILARDI et al. 2006 = Ghilardi, M., C. J. Goddart, 40, 57−70. P. Porena 2006, Les cités de l’Italie tardo­antique, GÁBOR, O. 2008, Buildings, Building Elements in the IVe­VIe siècle: institutions, économie, société, cul­ Town of Sopianae and its Late Antique Cemetery. ture et religion. – Collection de l’École française – In: P. Scherrer (ed.), Domus: das Haus in den de Rome 369. Städten der römischen Donauprovinzen. Akten GILLES, K. J. 1985, Spätrömische Höhensiedlungen in des 3. Internationalen Symposiums über römische Eifel und Hunsrück. – Trierer Zeitschrift, Beiheft 7, Städte in Noricum und Pannonien [vom 19 bis 22. Trier. April 2005 in St. Pölten], 333–348. GLASER, F. 1991, Das frühchristliche Pilgerheiligtum auf dem Hemmaberg. – Aus Forschung und Kunst 26. 399 Slavko CIGLENEČKI GLASER, F. 1992, Teurnia. Römerstadt und Bischofs­ GRBIĆ, M. 1936, Arhitektura u Basijani I. – Glasnik sitz. – Klagenfurt. istorijskog društva u Novom Sadu 9, 19–31. GLASER, F. 1996, Frühchristliche Denkmäler in Kärnten. GRBIĆ, M. 1937, Arhitektura u Basijani kod Donjih Ein Führer. – Klagenfurt. Petrovaca u Sremu. – Glasnik istorijskog društva u GLASER, F. 1997, Frühes Christentum im Alpenraum: Novom Sadu 10, 1–7. Eine archäologische Entdeckungsreise. – Graz, GREGORY, T. E. 1985, An early Byzantine complex at Wien, Köln. Akra Sophia near Corinth. – Hesperia 54, 411–428. GLASER, F. 2002, Teurnia. – In: Šašel Kos, Scherrer GREGORY, T. E. 1993, Isthmia 5: The Hexamilion and (eds.) 2002, 135−147. the Fortress. – Princeton, New Jersey. GLASER, F. 2008, Castra und Höhensiedlungen in GREENSLADE, S. 2019, Butrint 6: Excavations on the Kärnten und Nordtirol. – In: Steuer et al. (eds.) Vrina Plain. Vol. 1: The Lost Roman and Byzantine 2008, 595–642. Suburb. – Butrint Archaeological Monographs 6/1. GLASER, F. 2014, Provinzialrömische Archäologie GRIGOROV, V. 2011a, The Byzantine Fortress “Krasen” und Antike Numismatik mit den Außenstellen near Panagyurishte. – In: Daim, Drauschke (eds.) Römermuseum Teurnia und Archäologischer 2011, 779–808. Park Magdalensberg. – Rudolfinum, Jahrbuch des GRIGOROV, V. 2011b = ГРИГОРОВ, B., Късноримска Landesmuseums für Kärnten 2013, 56–78. крепост на връх Боровец до град Правец GLASER, F. 2015, Teurnia, civitas Tiburnia. – In: (разкопки през 2008-2009 г.) (Late Roman Fortress Heinrich-Tamáska et al. (eds.) 2015, 11–26. on Mt. Borovets near Pravets (Excavations in 2008– GLAVIČIĆ, A. 1984, Arheološki nalazi iz Senja i okolice 2009)). – In: S. Stanev et al. (eds.), Сборник в чест (6). – Senjski zbornik 10–11, 7–28. на 90-годишнинaта на арх. Стефан Бояджиев / GLUŠČEVIĆ, S., D. GROSMAN 2015, Nova bizantska Studies in Honor of Stefan Boyadzhiev, Sofia, 57–76. utvrda na otočiču Veliki Sikavac kot otoka Paga / GROH, H. 2021, Ager Solvensis (Noricum) oppidum – A new Byzantine fortification on Veliki Sikavac municipium – sepulcra – territorium – opes natu­ islet off the island of Pag. – Diadora 29, 121–150. rales. – Forschungen zur geschichtlichen Landes- GÖMÖRI, J. 2001, Von Scarbantia zu Sopron. – In: kunde der Steiermark 92. Zwischen Römersiedlung und mittelalterlicher GROH, S. 2012, Fortificazioni Aquileia. Forschungen Stadt. Archäologische Aspekte zur Kontinuitätsfrage, zur Urbanistik und spätantik-byzantinischen Beitrage zur Mittelalterarchäologie in Österreich Fortifikation von Aquileia (Italien). – Jahreshefte 17, 223–231. des Österreichischen Archäologischen Institutes in GÖMÖRI, J. 2003, Scrabantia. – In: Šašel Kos, Scherrer Wien 81, 67–96. (eds.) 2003, 81–92. GROH, S., V. LINDINGER 2008, Neue Forschungen ǴORǴIEVSKA, A., E. NASUH 1997–1999 = ЃОР- in Immurium-Voidersdorf/St. Margarethen in ЃИ ЕВСКА, А., E. НАСУХ, Археолошки Salzburg. Die geophysikalische Prospektion 2007. – истражувања во Хераклеја – 1998 г. (Archaeo- Jahreshefte des Österreichischen Archäologischen logical investigations at Heraclea, 1998). – Mac­ Institutes in Wien 77, 77–89. edoniae Acta Archaeologica 16 (2005), 225–240. GROH, S., H. SEDLMAYER 2002, Forschungen im Kastel GOSTENCNIK, K. 2000, Die Ausgrabungen auf dem Mautern­Favianis. Die Grabungen der Jahre 1996 “Lug ins Land” bei Molzbichl von 1992–1995. Ein und 1997. – Der römische Limes in Österreich 42. Vorbericht. – Carinthia I, 190, 101–112. GRUBER, Ch. 2017, Zur baulichen Entwicklung der GRABHERR, G., B. KAINRATH (eds.) 2011, Die Palastvil a Loig bei Salzburg. – Archaeo Plus, Schrifspät antike Höhensiedlung auf dem Kirchbichl von ten zur Archäologie und Archäometrie der Paris Lavant. Eine archäologische und archivarische Lodron­Universität Salzburg 8, 75–91. Spurensuche. − Ikarus 5. GUEST, P. 2007, Coin Circulation in the Balkans in GRAČANIN, H., G. BILOGRIVIĆ 2016, Siscia in Late Late Antiquity. – In: Poulter (ed.) 2007a, 295–308. Antiquity. – Acta archaeologica Carpathica 51, GUDELJ, L. 1998, Proložac Donji – groblje kot svetog 103−143. Mihovila, rezultati istraživanja godine 1997. – GRAČANIN, H., A. RAPAN PAPEŠA 2011, Postrim- Obavijesti 30/98, 68–70. ski grad u južnoj Panoniji: primjer Cibala. – In: GUDELJ, L. 2000, Proložac Donji. Izvješče o istra- Scrinia Slavonica. Godišnjak Podružnice za povijest živanjima lokaliteta kod crkve Sv. Mihovila u Slavonije, Srijema i Baranje Hrvatskog instituta za Postranju. – Starohrvatska prosvjeta 27, 129–145. povijest, 7– 30. GUGL, C. 2011, Carnuntum und sein Legionarslager GRAF, D. F. 2001, Town and Countryside in Roman in Spätantike und Frühmittelalter. Von der Car­ Arabia during Late Antiquity. – In: T. S. Burns, J. nutensis scutaria zur frühmittelalterlichen Siedlung W. Eadie, Urban Centers and Rural Contexts in Late 9./10. Jahrhunderts. – In: Konrad, Witschel (eds.) Antiquity, East Lansing, 219–240. 2011, 505–532. 400 BIBLIOGRAPHY GUNJAČA, Z. 1986, Kasnoantička fortifikacijska HENNING, J. 1987, Südosteuropa zwischen Antike arhitektura na istočnojadranskom priobalju i oto- und Mittelalter. Archäologische Beiträge zur Land­ cima. – Materijali 22, 124–136. wirschaft des I. Jahrtausends u. Z. – Schriften zur GUSSONE, M., D. SACK 2017, Resafa / Syrien: Städte- Ur- und Frühgeschichte 42. bauliche Entwicklung zwischen Kultort und HENSEL, W., J. RAUHUTOWA 1981, Archaeological Herrschaftssitz. − In: Rizos (ed.) 2017a, 117−136. research at Debrešte (Macedonia) 1974–1978. – HARIZANOV, A. 2020, Socioeconomic Aspects of Ce­ Archaeologia Polona 20, 191–225. ramic Production in Roman Thrace. – Journal of HILD, F., H. HELLENKEMPER 1990, Kilikien und Isau­ Ancient History and Archaeology 7/2. rien. − Tabula Imperii Byzantini 5 (= Österreichi- HÉBRARD, E., J. ZEILLER 1912, Spalato. Le palais de sche Akademie der Wissenschaften. Denkschriften Dioclétien. – Paris. der philosophisch-historischen Klasse 215). HEIJMANS, M. 2004, Arles Durant L’Antiquité Tardive: HIRSCHFELD, Y. 2001, Habitat. – In: W. Bowersock, P. de la Duplex Arelas à L’urbs Genesii. – Collection Brown, O. Grabar (eds.), Interpreting Late Antiq­ de l’École française de Rome 324. uity: Essays on the Postclassical World, Cambridge, HEINRICH-TAMÁSKA, O. (ed.) 2011a, Keszthely­ 258−271. Fenékpuszta im Kontext spätantiker kontinuitäts­ HOBDARI, E., Y. CEROVA 2014, Scampis dans forschung zwischen Noricum und Moesia. – Castel- l’Antiquité tardive: la ville intra et extra-muros. – lum Pannonicum Pelsonense 2. Proceedings of the International Congress of Alba­ HEINRICH-TAMÁSKA, O. 2011b, Die spätromische nian Archaeological Studies (65th Anniversary of Innenbefestigung von Keszthely­Fenékpuszta: In­ Albanian Archaeology), 21-22 November 2013, nere Chronologie und funktionl er Wandel. – In: Tirana, 501–510. Heinrich-Tamáska (ed.) 2011a, 653–702. HOF, C. 2016, The Late Roman City Wall of Resafa/ HEINRICH-TAMÁSKA, O. 2012, Fortleben, Ab- Sergiupolis (Syria). Its Evolution and Functional bruch und Neuanfang: Spuren des Christentums Transition from Representative over Protective to in Pannonien in 4.-9. Jahrhundert. – In: O. Concealing. – In: Frederiksen et al. (eds . ) 2016 , Heinrich-Tamáska, O. Krohn, S. Ristow (eds.), 377–392. Christianisierung Europas: Entstehung, Entwicklung HOFMAN, B., M. ŽUPANČIČ 2008, Koper. – Varstvo und Konsolidierung im archäologischen Befund, spomenikov. Poročila 44, 107. Regensburg, 213–237. HORVAT, J. 2020, Nauportus­ Vrhnika. – In: Horvat et HEINRICH-TAMÁSKA, O. 2015, Civitates et castra im al. (eds.) 2020a, 93–112. Lichte der Kontinuitätsforschunge: zwei Beispiele HORVAT, J., A. DOLENC VIČIČ 2010, Arheološka aus Pannonien (5.–7. Jh. N. Chr.). – In: Heinrich- najdišča Ptuja / Archaeological Sites of Ptuj. Rabelčja Tamáska et al. (eds.) 2015, 45–69. vas. – Opera Instituti Archaeologici Sloveniae 20. HEINRICH-TAMÁSKA, O. 2016, Churches, Early HORVAT et al. 2003 = Horvat, J., M. Lovenjak, A. Dolenc Christianity and Roman Continuty in Pannonia and Vičič, M. Lubšina Tušek, M. Tomanič Jevremov, Z. Noricum (Fifth to seventh century). – In: E. Tóth, T. Šubic 2003, Poetovio. Development and Topogra- Vida, I. Takács, Saint Martin and Pannonia. Chris­ phy. – In: Šašel Kos, Scherrer (eds.) 2003, 153–190. tianity on the frontiers of Roman world. Exhibition HORVAT et al. (eds.) 2020a = Horvat, J., I. Lazar, A. Gas- catalogue, Pannonhalma, Szombathely, 123−136. pari (eds.), Manjša rimska naselja na slovenskem HEINRICH-TAMÁSKA, O. 2017, Castel um – Castrum prostoru / Minor Roman Settlements in Slovenia. – – Civitas? L’ évolution fonctionnelle des nouveaux Opera Instituti Arcaeologici Sloveniae 40. étabblissements de l’antiquité tardive en Pannonie HORVAT et al. 2020b = Horvat, J., B. Mušič, A. Dolenc et en Mésie Seconde: étude comparative. – In: Rizos Vičič, A. Ragolič 2020, Arheološka najdišča Ptuja / (ed.) 2017a, 39–55. Archaeological Sites of Ptuj. Panorama. – Opera HEINRICH-TAMÁSKA, O., R. MÜLLER 2009, Instituti Archaeologici Sloveniae 41. Keszthely-Fénekpuszta: spätrömische Festung HOTI, A. 1996 , Të dhëna arkeologjike për krishterimin und frühchristliches Zentrum? Einblicke in die e hershëm në Dyrrah (shek. IV­VII) / Des données Forschungsgeschichte. – In: O. Heinrich-Tamáska, archéologiques sur le christianisme ancien à Dyr­ P. Straub (eds.), Keszthely­Fenékpuszta im Spiegel rachium (IVe­VIIe siècles). − Il iria 26, 173–181. der Jahrtausende, Leipzig, Zalaegerszeg, 13−18. HOTI et al. 2008 = Hoti, A., J. Wilkes, E. Metal a, B. HEINRICH-TAMÁSKA et al. (eds.) 2015 = Heinrich- Shkodra 2008, The Early Byzantine Circular Fo- Tamáska, O., H. Herold, P. Straub, T. Vida (eds.) rum in Dyrrachium (Durrës, Albania) in 2002 and 2015, “Castel um, civitas, urbs”. Zentren und Eliten 2004–2005: Recent Recording and Excavation. – The im frühmittelalterlichen Ostmitteleuropa. – Castel- Annual of the British School at Athens 103, 367–397. lum Pannonicum Pelsonense 6. 401 Slavko CIGLENEČKI HOXHA, G. 2001, Phylon von Byzanz und die spät- Новопазарски зборник ( Novopazarski zbornik) antiken Befestigungen in Albanien. − Archäologis­ 14, 7–16. ches Korrespondenzblatt 31, 601−616. IVANIŠEVIĆ, V. 2011, Caričin Grad – the Fortifications HOXHA, G. 2003, Scodra dhe Praevalis në antikitetin e and the Intramural Housing in the Lower Town. – vonë / Scodra and Praevalis in the Late Antiquity. In: Daim, Drauschke (eds.) 2011, 747–778. – Skhoder. IVANIŠEVIĆ, V. 2016, Late antique cities and their HOXHA, G. 2014, Die Transformationsprozesse in der environments in Northerni Illyricum. – In: Daim, Stadt Lissus während der Spätantike. – Proceedings Drauschke (eds.), Hinter den Mauern und auf dem of the International Congress of Albanian Archaeooffenen Land ­ Leben im Byzantinischen Reich, logical Studies, 511–526. Byzanz zwischen Orient und Okzident 3, 89−99 HUDECZEK, E. 2002, Flavia Solva. – In: Šašel Kos, IVANIŠEVIĆ, V. 2017, Une capitale revisitée: Caričin Scherrer (eds.) 2002, 203–212. Grad (Justiniana Prima). – Comptes rendus des IATRUS-KRIVINA 1995 = Iatrus­Krivina. Spätantike séances de l’Académie des Inscriptions et Bel es­ Befestigung und frühmittelalterliche Siedlung an der Lettres I (janvier-mars), 93–114. unteren Donau, Bd 5. – Studien zur Geschichte des IVANIŠEVIĆ, V., M. KAZANSKI 2002, La nécropole de Kastel s Iatrus, Berlin. l’époque des Grandes Migrations à Singidunum. − IGL, R. 2011, Überlegungen zur Siedlungsgenese der Singidunum 3, 101−157. hoch mittelalterlichen Stadt Enns aus der spätan- IVANIŠEVIĆ, V., S. STAMENKOVIĆ 2013, = ИВА- tiken Siedlung im Legionslager Lauriacum. Ein НИШЕВИЋ, В., C. СТАМЕНКОВИЋ, Pазградња Beitrag zur Kontinuitätsdiskussion – In: Konrad, фортификације акропоља Царичиног града (De- Witschel (eds.) 2011, 461–474. construction of the Caričin Grad fortifications). – INGLEBERT, H. 2006, Conclusions. Periodise L’ Antiq- Лесковачки зборник ( Leskovački zbornik) 53, 22–31. uité tardive. – In: Les cités de l’Italie tardo­antique IVANIŠEVIĆ, V., P. ŠPEHAR 2005, Early Byzantine (IVe ­ VIe siècle). – Collection de l’École française Finds from Čečan and Gornji Streoc (Kosovo). – de Rome 369, 359–366. Starinar 55, 133–159. INTAGLIATA, E. E. 2017, Palmyra and its Ramparts IVANIŠEVIĆ et al. 2016a = Ivanišević, V., S. Stamenković, during the Tetrarchy. − In: Rizos 2017a, 71−83. S. Jović 2016a = Иванишевић, B., С. Стаменковић, ISKRA-JANOŠIĆ, I. 2004, Colonia Aurelia Cibalae. С. Јовић, Римско насеље и занатски центар у Entwicklung der Stadt. – In: Šašel Kos, Scherrer Mалој Kопашници / A Roman settlement and (eds.) 2004, 169–195. workshop center at Mala Kopašnica. – In: S. Perić, ISTENIČ, J. 2015, Mali Njivč nad Novaki / Mali Njivč A. Bulatović (eds.), Археолошка истраживања на above Novaki. – In: Istenič et al. (eds.) 2015, автопуту / Archaeological investigations along the 365–384. highway Route E75 (2011–2014), Beograd, 17–45. ISTENIČ et al. (eds.) 2015 = Istenič, J., B. Laharnar, J. IVANIŠEVIĆ et al. 2016b = Ivanišević, V. Bugarski, S. Horvat (eds.) 2015, Evidence of the Roman army in Stamenković 2016b, Nova saznanja o urbanizmu Slovenia / Sledovi rimske vojske na Slovenskem. – Caričinog grada. Primena savremenih metoda Katalogi in monografije 41. prospekcije i detekcije (New Insights into Urban ISZTIN et al. 2014 = Isztin, G., T. Tárczy, E. Tóth 2014, Planning of Caričin Grad. The application of Mod- Die Aula Palatina aus dem 4. Jahrhundert in Savaria ern Sensing and Detecting Methods). – Starinar (Revision und Aufgaben). – Archaeologiai Értesítő 66, 143–160. 139, 137–155. IVANIŠEVIĆ et al. 2019 = Ivanišević, V., B. Bavant, I. Bu- IVANIŠEVIĆ, V. 1987 = ИВАНИШЕВИЋ, В., Касно- garski 2019, Caričin grad – arheološka istraživanja антићко утврђење у Рамошеву (La forteresse de u 2017. godini. – In: Arheologija u Srbiji. Projekti Ramoševo). – Новопазарски зборник ( Novopazar­ Arheološkog instituta u 2017. godini, 151–158. ski zbornik) 11, 5–12. IVANOV, R. (ed.) 2012, Roman cities in Bulgaria. − Cor- IVANIŠEVIĆ, V. 1988 = ИВАНИШЕВИЋ, В., Рано- pus of ancient and medieval settlements in modern византијско утврђење на Хуму код Тутина (La Bulgaria, Vol. I., Sofia. forteresse protobyzantine de Hum). – Новопазарски IVANOV, T. 1980 = ИВАНОВ, Т., Абритус­римски зборник ( Novopazarski zbornik) 12, 5–12. кастел и ранновизантийски град в Долна Мизия IVANIŠEVIĆ, V. 1989 = ИВАНИШЕВИЋ, В., Античко ( Abritus. Rimski kastel i rannovizantijski grad v утврђење на Тројану (La fortresse antique de Dolna Mizija) . – Sofia. Trojan). – Новопазарски зборник ( Novopazarski IVEKOVIĆ, Ć. M. 1927, Otok Žirje. – Starohrvatska zbornik) 13, 7–15. prosvjeta 1, 45–59. IVANIŠEVIĆ, V. 1990, = ИВАНИШЕВИЋ, В., Касно- JACOBS, I. 2012, The creation of the Late Antique античко утврђење на Златном Камену код city: Constantinople and Asia minor during the Новог Пазара (La forteresse de Zlatni Kamen). – ‘Theodosian Renaissance’. – Byzantion 82, 113–164. 402 BIBLIOGRAPHY JÄGGI, C. 1990, Aspekte der städtebaulichen Entwick- JEREMIĆ, M., M. MILINKOVIĆ1995, Die byzantini- lung Aquileias in frühchristlicher Zeit. – Jahrbuch sche Festung von Bregovina. – Antiquité Tardive fur Antike und Christentum 33, 158−196. 3, 209–225. JANAKIEVSKI, T. 2001 = ЈАНАКИЕВСКИ, Т., JOHNSON, S. 1983, Late Roman Fortifications. – London. Доцноантичка микростанбена целина над JOVANOVA, L. 2015a , Scupi ­ Colonia Flavia Scupinorum. римскиот театар во Херацлеа Лyнцестис ( The Macedonian cultural and natural heritage. – Skopje. late­antique micro­residential ensemble above the JOVANOVA, L. 2015b, Scupi. Transition from Impe- Roman theatre in Heraclea Lyncestis). – Bitola. rial Colony to Late Antiquity Metropolis. – Folia JANEŠ, A. 2015, Late Antique Rural Complex by the archaeologica Balcanica III, 219–252. Church of St. Chrysogonus in Glavotok (Island of JOVANOVA, L., M. ONČEVSKA TODOROVSKA 2018 = Krk). – Portal. Godišnjak Hrvatskog restauratorskog ЈОВАНОВА, Л., М. ОНЧЕВСКА ТОДОРОВСКА, zavoda 6, 9–22. Cкупи­сектор jугоисточен бедем и комплекс JANKOVIĆ, Đ. 1981, La partie danubienne de la région терма ­ базилика со атриум / Scupi. Sector southdes’ Acquis au VIe et au débutdu VIIe siècle. – Beo-eastern defensive wall and the thermae ­ Atrium grad. basilica complex. – Skopje. JELIČIĆ-RADONIĆ, J. 1994, Ranokrščanske dvojne JOVANOVIĆ, V. 2004, Utvrđeno naselje Veletin crkve u Starom Gradu na Hvaru. – Regionalni zavod (Fortified Settlement Veletin). – Starinar 53–54 za zaštitu spomenika kulture Split, Split. (2003–2004), 139–161. JELIČIĆ-RADONIĆ, J. 1998, Nova istraživanja grad- JURKIĆ, V. 1981, Građevinski kontinuitet rimskih skih zidina Salone. – Prilozi povijesti umjetnosti u gospodarskih vila u zapadnoj Istri od antike do Dalmaciji 37, 5–36. bizantinskog doba. – Histria Historica 4, 77–105. JELIČIĆ-RADONIĆ, J. 2001, Rimsko naselje u Ublima KABAKCHIEVA, G. 2011, Die antike Vil a Armira – na Lastovu. – In: Arheološka istraživanja na po­ eine neue Touristattraktion in Bulgarien. – In: dručju otoka Korčule i Lastova, Znanstsveni skup, Bülow, Zabehlicky (eds.) 2011, 219–231. Vela Luka i Korčula (1991), Zagreb, 197–220. KALIĆ, J., D. MRKOBRAD 1985 = КАЛИЋ, Ј., Д. JELIČIĆ-RADONIĆ, J. 2006, Salona, The Urbs orientalis. МРКОБРАД, Градина у Радалици (Gradina u – Hortus Artium Medievalium 12, 43–54. Radalici). – Новопазарски зборник (Novopazarski JELIČIĆ-RADONIĆ, J. 2015, Salona – Metropolis of zbornik) 9, 39–46. the Roman Province of Dalmatia and its Cultural KANDLER, M. 2004, Carnuntum. – In: Šašel Kos, Scher- Environment in the Light of Recent Research. – rer (eds.) 2004, 11–66. Neue Forschungen zum frühen Christentum in den KARAĐOLE, A., I. BORZIĆ 2020, Keramički nalazi Balkanländern, 73–81. iz Bizantske utvrde Gradina na otoku Žirju. – JELINČIĆ VUČKOVIĆ, K. 2015, Rimsko selo u provinci Archaeologia Adriatica 14, 45–109. Gornjoj Panoniji: Virovitica Kiškorija Jug / Roman KARAGIORGOU, O. 2001a, Demetrias and Thebes: the Vil age in the Province of Upper Pannonia: Viro­ fortunes and misfortunes of two Thessalian port vitica Kiškorija South. – Monographiae Instituti cities. – In: L. Lavan, Recent research in Late­antique Archaeologici 7. Urbanism, Journal of Roman Archaeology, Suppl. JELINČIĆ VUČKOVIĆ et al. 2019 = Jelinčić Vučković, ser. 42, 182–216. K., E. Botte, A. Bertrand 2019, Arheološko KARAGIORGOU, O. 2001b, Urbanism and economy in istraživanje na lokalitetu Novo Selo Bunje na otoku late antique Thessaly (3rd–7th century A.D.): the Braču 2018. godine. – Annales Instituti Archaeo­ Archaeological evidence. – Oxford. logici 15, 201–216. KARAGIORGOU, O. 2013, “Christian” or “Thessalian” JEREMIĆ, G., T. ČERŠKOV 2021= ЈЕРЕМИЋ, Г, Т. Thebes: the Port City of Late Antique Thessaly. – In: ЧЕРШКОВ, Naissus. Топографија античког и J. Albani, E. Chalkia, Heaven and Earth: Cities and касноантичког Ниша. Археолошки трагови Countryside in Byzantine Greece, Athens, 156–167. на простору центра савременог града ( Topog­ KARAISKAJ, G. 1972, Të dhëna të reja për datimin e raphy of Roman and Late Roman Niš (Naissus). kalasë së Elbasanit. – Monumentet 3, 147–157. Аrchaeological traces in the area of the modern city KARAISKAJ, G. 1998, Die militärische Architektur vom center). – Niš. 6. Jh.bis zum Anfang des 7. Jh. in Albanien. Die JEREMIĆ, M. 1995, Balajnac, aglomeration proto- Neuerungen des 6. Jh. – In: Acta XIII Congressus byzantine, fortifiee (Region de Niš, Serbie du Sud). – 1998, 867–874. Antiquité Tardive 3, 193–207. KARAISKAJ, G., A. BAÇE 1975, La forteresse de Durres JEREMIČ, M. 2006, The Relationship Between the Ur- et les autres fortifications environnantes pendant la ban Physical Structures of Medieval Mitrovica and basse antiquite. – Monumentet 9, 29–33. Roman Sirmium. – Hortus Artium Medievalium KARDULIAS, P. N. 1993, Anthropology and population 12, 137–161. estimaes for the Byzantine fortress at Isthmia. – In: 403 Slavko CIGLENEČKI The Corinthia in the Roman Period, Journal of Ro- KNIFIC, T. 1979, Vranje pri Sevnici. Drobne najdbe z man Archaeology, Suplemetary Ser. 8, 139–148. Ajdovskega gradca (leto 1974) (Vranje bei Sevni- KASPRZYK, M., M. MONTEIL (eds.) 2017, Aggloméra­ ca. Kleinfunde aus Ajdovski gradec (J. 1974)). – tions, vici et castra du Nord de la Gaule entre Antiqui­ Arheološki vestnik 30, 732–785. té tardive et début du haut Moyen Âge. – Gallia 74/1. KNIFIC, T. 1993, Hunski sledovi v Sloveniji? – In: Ptujski KASTLER et al. 2017 = Kastler, R., F. Lang, S. Traxler arheološki zbornik, 521–542. 2017, Neue Forschungen zur ländlichen Besied- KNIFIC, T. 1994, Vranje near Sevnica. A Late Roman lung in Nordwest-Noricum - eine Einführung. – Settlement in the Light of Certain Pottery Finds. – Archaeo Plus 8, 7–50. Arheološki vestnik 45, 211–237. KATIĆ, M. 2003, The Late Antique Town on the Eastern KNIFIC, T. 2005, Gospe iz mesta Karnija. – In: Kranjski Adriatic Coast. – Histria Antiqua 11, 523–528. zbornik 2005, Kranj, 331–343. KATSAROVA et al. 2018 = Katsarova, V., H. Popov, N. KNIFIC, T., T. NABERGOJ 2017, Medieval stories from Tonkov 2018 = В. Кацарова, Х. Попов, Н.Тонков, the crossroads. – Ljubljana. Mежду Aвгуста Tраяна, Tуида и Kабиле. Cело KNIFIC, T., M. TOMANIČ JEVREMOV 1996, Prva zna- от римската епоха в землището на с. Cкобелево, menja velikega preseljevanja ljudstev v Sloveniji. – община Cливен ( Between Augusta Traiana, Thu- Ptujski zbornik 6/1, Ptuj, 369–393. idai, and Kabyle. A vil age from the Roman period KNOBLAUCH, P., Ch. WITSCHEL 1993, Arykanda in the territory of Skobelevo, Sliven municipal- in Lykien. Eine topographische Aufnahme. – ity). – Studia archaeologica Universitatis Serdicensis, Arzhäologischer Anzeiger, 229–262. Suppl. 6, 343–361. KOCSIS, A. 2020, GPR survey in the Late Roman fort KELEMEN, M. H. 2014, Késő Római vendégfogadó az of Tokod. – In: G. I. Farkas, R. Neményi, M. Szabó Aquincum–Brigetio közötti diagonális út mentén, (eds.), The Danube Limes in Hungary. Archaeological Kesztölc-Tatárszál áson. – Archaeologiai Értesítő Research conducted in 2015–2020, Pécs, 187–200. 139, 35–70. KOLŠEK, V. 1984, Nekaj podatkov o zgodnjem KENNEDY, D. L. 1982, Archaeological Explorations on krščanstvu v Celeji (Einige Angaben über das the Roman Frontier in North­East Jordan. – BAR Frühchristentum in Celeia). – Arheološki vestnik Internationa Series 134. 35, 342–345. KEPESKA, L., K. KEPESKI 2006 = Кепеска, Л., К. KOMŠO, D. 2008, Pečina Laganiši – mjesto života i Кепески 2006, Доцноантичка населба Трпчева smrti. – In: I. Porečki susret arheologa. Rezultati Црква (Late Antique Settlement Trpčeva Crkva). – arheoloških istraživanja na području Istre, Zbornik Folia Archaeologica Balcanica 1, 425–438. radova s međunarodnoga znanstvenog skupa KEPESKI, K. 2008, Vil a rustica во Пештериcа кау (Poreč 2006), Poreč, 9–16. Прилеп (Vil a rustica in Pešterica near Prilep). – KONDIĆ, V. 1984, Les formes des fortifications proto- In: Археолошките откритија на почвата на byzantines dans la région des Portes de Fer. – In: Македонија ( Arheološkite otkritija na počvata Villes 1984, 131–161. na Makedonija), Прилози за истражувањето KONDIĆ, V., V. POPOVIĆ 1977, = КОНДИЋ, В., В. на историјата на културата на почвата на ПОПОВИЋ, Царичин град. Утврђено насеље у Македонија (Prilozi za istražuvanjeto na istori- византијском Илирику ( Caričin grad. Utvrđeno jata na kulturata na počvata na Makedonija) 18, naselje u vizantijskom Iliriku). – Galerija Srpske 619–632. akademije nauka i umetnosti 33. KIRIGIN, B., A. MILOŠEVIĆ 1981, Odkritje. Svetac. – KONRAD, M., C. WITSCHEL (eds.) 2011, Römische Arheo 2, 45–51. Legionslager in den Rhein­ und Donauprovinzen – KIRILOV, Ch. 2006a, Die Stadt des Frühmittelalters in Nuclei spätantik­frühmittelalterlichen Lebens?. Ost und West. Archäologische Befunde Mitteleuropas – Bayerische Akademie der Wissenschaften. Phi- im Vergleich zur östlichen Balkanhalbinsel. – Stu- losophisch-historische Klasse. Abhandlungen 138. dien zur Archäologie Europas 3. KORAĆ, V. 2001, = КОРАЋ, В., Мартинићи – остаци KIRILOV, Ch. 2006b, Die Reduktion der ummauerten раносредњовековног града (Martinići – ostaci Städtefläche und die Frage nach dem Verfall der an- ranosrednjovekovnog grada). – Beograd. tiken Stadt. – Archaeologia Bulgarica 10/2, 55–104. KOROŠEC, P. 1983, Predjama konec 4. do sredine 5. KIRILOV, Ch. 2007, Der rissig gewordene Limes. st. (Predjama vom Ende des 4. Bis zur Mitte des Höhensiedlungen im östlichen Balkan als Zeugnis 5. Jahrhunderts). – Arheološki vestnik 33, 84–114. für die Schwäche des oströmischen Reichs in der KOS, P. 1983, Antični novci. – In: L. Plesničar Gec, Spätantike. – In: L. Vagalinski (ed.), The Lower Dan­ Starokrščanski center v Emoni, Katalogi in mono- ube in Antiquity (VI C BC ­ VI C AD). International grafije 21. Archaeological Conference (Bulgaria-Tutrakan, KOS, P. 1986, The monetary circulation in the southeastern 6-7. 10. 2005), 329–352. Alpine Region ca. 300 B.C.­ A.D. 1000 / Denarni 404 BIBLIOGRAPHY obtok na prostoru jugovzhodnih Alp 300 pr.n. š.­ KULIKOWSKI, M. 2004, Late Roman Spain and its cities. 1000. – Situla 24. Ancient society and history. – Baltimore, London. KOS, P. 2012, The construction and abandonment of the KULIKOWSKI, M. 2010, The Urban Landscape in His- Claustra Alpium Iuliarum defence system in light pania in the Fifth Century. – In: El tiempo de los of the numismatic material / Gradnja in opustitev “Bárbaros.” Pervivencia y transformación en Gal ia obrambnega sistema Claustra Alpium Iuliarum v e Hispania (ss V­VI D. C), Zona Arqueológica 11, luči numizmatičnega gradiva. – Arheološki vestnik 311−318. 63, 265–300. KUŠAN ŠPALJ, D. 2020, Aquae Iasae (Varaždinske KOS, P. 2014, Ad Pirum (Hrušica) in Claustra Alpium Toplice) – ranokrščanski horizont. – In: Zbornik Iuliarum. – Vestnik 26 (Zavod za varstvo kulturne I. skupa hrvatske ranokrščanske arheologije (HR­ dediščine Slovenije). RANA), Zagreb, 303 – 319. KOS, P. 2020, Ad Pirum – Hrušica. – In: Horvat et al. KUSETIČ et al. 2014 = Kusetič, J., P. Kos, A. Breznik, (eds.) 2020a, 61–76. M. Stokin 2014, Claustra Alpium Iuliarum – med KOSTADINOVSKI, S., V. CVETANOVSKI 2008 = raziskovanjem in upravljanjem / Between Research КОСТАДИНОВСКИ, С., В. ЦВЕТАНОВСКИ, and Management. – Ljubljana. Градисхте, виллаге Стенцхе (Gradishte, vil- LADSTÄTTER, S. 2000, Die Materielle Kultur der lage Stenche). – Makedonski arheološki pregled 1, Spätantike in den Ostalpen. Eine Fal studie am 184–189. Beispiel der westlichen Doppelkirchenanlage auf KOVACSOVICS, W. K. 2002, Iuvavum. – In: Šašel Kos, Hemmaberg. – Mitteilungen der Prähistorischen Scherrer (eds.) 2002, 165–202. Kommission 35. KOVAČIĆ, V. 2006a, Lovrečina. – Hrvatski arheološki LADSTÄTTER, S. 2002, Die Spätantike. – In: Gassner godišnjak 2 (2005), 362–363. et al. (eds.) 2002, 285 – 368. KOVAČIĆ, V. 2006b, Mirje. – Hrvatski arheološki LADSTÄTTER, S. 2011, Ephesus in Byzantinischer Zeit. godišnjak 2 (2005), 371–372. Das letzte Kapitel der Geschichte einer antiken KOVAČIĆ, V. 2010, Kasnoantička vila s portikom na Grossstadt. – In: Daim, Drauschke (eds.) 2011, Mirju kod Postira. – Klesarstvo i graditeljstvo 493–519. 21/3–4, Pučišća. LADSTÄTTER, S. 2017, Ephesus. – In: P. Niewöhner KOVALJEV, G. 1990, Donje Nerodimlje. Kasnoantička (ed.), Archaeology of Byzantine Anatolia: From the palata / Late Roman Palace. – Arheološki pregled end of Late Antiquity until the coming of Turks, (1988), 168–170. Oxford, 238–248. KOYCHEVA, K. 1992 = КОЙЧЕВА, К., Зхилисхта от IV LANG, F. 2019 , Der Georgenberg bei Kuchl in römi­ на VI в. на Градисхте краy Габрово (Zhilishta ot IV scher Zeit. – Castel um Cucullis 1, Archäologie in na VI v. na Gradishte kray Gabrovo). – Годисхник Salzburg 11. на моузеите от Северна Булгариа (Godishnik na LASSUS, J. 1981, La Forteresse byzantine de Thamugadi mouzeite ot Severna Bulgaria) 18, 41–54. 1. Fouil es à Timgad 1938–1956. – Paris. KOYCHEVA, K. 2002, = КОЙЧЕВА, K., Черква и LASZLO KLEMAR, K., M. ZEMAN 2010, Naselja i or- некропол при укрепеното селище на м. Градище ganizacija prostora na srednjem istočnom Jadranu край Габрово (Chеrkva i nеkropol pri ukrеpеnoto od antike do ranoga srednjeg vijeka – pitanje sеlishtе na m. Gradishtе kray Gabrovo). – In: Sloko- kontinuiteta (Bribir, Donje šibensko polje). – In: T. ska et al. 2002, 178–183. Šeparović (ed.), Zbornik Stjepan Gunjača i hrvatska KRAJŠEK, J., D. ŽERIČ 2010, Celeia: mesto pod mestom. srednjovjekovna povijesno­arheološka baština 1, Razstavišče Knežji dvorec. – Celje. Split, 147–165. KRAUSE, J.-U., Ch. WITSCHEL (eds.) 2006, Die Stadt LAVAN. L. 2003, Late Antique Archaeology: An Introducin der Spätantike – Niedergang oder Wandel? – His-tion. – In: Lavan, Bowden (eds.) 2003, VII–XVI. toria Einzelschriften 190. LAVAN, L., W. BOWDEN (eds.) 2003, Theory and Prac­ KRAUTHEIMER, R. 1986, Early Christian and Byztice in Late Antique Archaeology. – Late Antique antine Architecture. – New Haven, London (4th Archaeology 1. edition). LAVAN et al. (eds.) 2007 = Lavan, L., L. Özgenel, A. KRIŽ, B. 2021, Gorjanci med Rimom in Bizancem / The Sarantis (eds.) 2007, Housing in Late Antiquity. Gorjanci Hil s between Rome and Byzantium. – From Palaces to Shops. – Late Antique Archaeol- Novo mesto. ogy 3/2 (2005). KUHNEN, H.-P. 2020, Mehr als ein Bischofssitz: LAWRENCE, A. W. 1983, A skeletal history of Byzantine Sabiona-Säben, Gem. Klausen (Südtirol) und sein Fortification. – The Annual of the British School at siedlungsarchäologisches Umfeld. – In: Cavada, Athens 78, 171–227. Zagermann (eds.) 2020a, 419–450. LAZAR, I. 2002, Celeia. – In: Šašel Kos, Scherrer (eds.) 2002, 71–101. 405 Slavko CIGLENEČKI LEBEN, F., A. VALIČ 1978, Ajdna. – Arheološki vestnik archaeological finds in Ohrid from Antiquity). 29, 534–545. – In: Археолошките откритија на почвата LEHNER, M. 2004, Lissus Excavation Report 2004. – на Македонија ( Arheološkite otkritija na počvata Karl Franz Universität Graz. na Makedonija), Прилози за истражувањето LELEKOVIĆ, T., A. RENDIĆ MIOČEVIĆ 2012, Rural на историјата на културата на почвата на settlements. – In: B. Migotti (ed.), The Archaeology Македонија (Prilozi za istražuvanjeto na istori- of Roman Southern Pannonia. The state of research jata na kulturata na počvata na Makedonija) 18, and selected problems in the Croatian part of the 475–486. Roman province of Pannonia, BAR, International MALONEY, S. J., J. R. HALE 1996, The vil a of Torre de Series 2393, 279–311. Palma (Alto Alentejo). – Journal of Roman Archae­ LEPELLEY, C. 1992, The survival and fall of the clas- ology 9, 275–294. sical city in Late Roman Africa. – In: Rich (ed.) MANASKOVA, M., B. ANGELOVSKI 2017, Late an­ 1992, 50–76. tique residential object from the archaeological site LIEBESCHUETZ, W. 1998, Light on the Dark Ages. – Kale­Vinica. – Review of the Institute of History American Journal of Archaeology 102, 817–819. and Archaeology RIHA 4/1. LIEBESCHUETZ, J. H. W. G. 2001, The Decline and Fal MANEVA, E. 1989 = МАНЕВА, Е., Базиликата Де од of the Roman City. – Oxford. Хераклеја Линкестис (La basilique D d’Heraclea LIEBESCHUETZ, J. H. W. G. 2007, The Lower Danube Lyncestis). – Лихнид / Lychnid 7, 51–63. Region under Pressure: from Valens to Heraclius. − MANEVA, E. 2008, = МАНЕВА, Е., Базиликата Де од In: Poulter (ed.) 2007a, 101−134. Хераклеја Линкестис (The Basilica D from Hera- LIEBL, H., W. WILBERG 1908, Ausgrabungen in As- clea Lyncestis). – In: Археолошките откритија seria. – Jahreshefte des Österreichisches Archäo­ на почвата на Македонија ( Arheološkite otlogisches Institut 11, 17–88. kritija na počvata na Makedonija), Прилози за LILČIĆ, V. 2004, = ЛИЛЧИЋ, В., Рановизантијски истражувањето на историјата на културата на кастели и мали градови на северу скопско- почвата на Македонија (Prilozi za istražuvanjeto кумановског региона (Early Byzantine Castels na istorijata na kulturata na počvata na Makedonija) and small Fortresses in the Northern Part of the 19, 835–850. Skopje-Kumanovo Region). – In: Niš and Byzan­ MARANO, Y. 2011, The towns of central and eastern tium 2, Second Symposium (Niš 2003), 177–184. Venetia in the Ostrogothic period. – In: Heinrich- LILČIĆ, V. 2013 = ЛИЛЧИЋ, В., Македонија во Tamáska (ed.) 2011a, 173–194. доцната антика (археологија). (Makedonija vo MARANO, Y. 2012, Doppo Attila. Urbanesimo e storia docnata antika (arheologija)). – In: E. Dimitrova, ad Aquileia tra V e VI secolo d.C. – In: A. Didone, J. Donev (eds.), Македонија. Милениумски C. Previato (eds.), L’architettura privata ad Aquileia културно историски факти (Makedonija. Milein età romana, Padova, 517–590. niumski kulturno­istoriski fakti), Skopje, 963–1032. MARASOVIĆ, T. 1994, Diocletian’s Palace. The World LILČIĆ ADAMS, V. 2018, Fanula Papazoglu’s Research Cultural Heritage. – Zagreb. and the City of Pelagonia in the Light of Recent Dis- MARENSI, A. 2001, Contributo allo studio del ’edilizia coveries. – In: Učenici i kolege za profesora Fanulu privata in Norico: le ville. – Notizie Archeologiche Papazoglu / Students and col eagues for Professor Bergomens i 9, 157–272. Fanoula Papazoglou. Međunarodna naučna kon- MARIN, E. 1994, Civitas Splendida Salona. Geneza, ferencija (Beograd, 17−18 oktobar 2017), 53–78. profil i transformacija starokršćanske Salone. – In: LOLIĆ, T. 2003, Colonia Flavia Siscia. – In: Šašel Kos, Salona Cristiana, Split, 11–96. Scherrer (eds.) 2003, 131–152. MARIN, E. 2006, The urbanism of Salona and Narona LORGER, F. 1936, Poročilo o izkopavanju rimske inside Roman Dalmatia. Dalmatia – In: D. Davison, naselbine v Grobelcah pri Šmarju pri Jelšah. – V. Gaffney, E. Marin (eds.), Dalmatia. Research in Časopis za zgodovino in narodopisje 31, 77–82. the Roman Province 1970 – 2001. Papers in honour LOVENJAK, M. 2003, Municipium Flavium Latobico- of J. J. Wilkes, BAR Int. Ser.1576, 73–80. rum Neviodunum. – In: Šašel Kos, Scherrer (eds.) MARIN, E. 2018, Mursa. Hadrijanova kolonija uz limes 2003, 93–105. Rimskog carstva / Colonia Aelia at the limes of Ro­ LUBŠINA TUŠEK, M. 2004, Izročilo preteklosti med man Empire. – Zagreb, Osijek. Potrčevo in Volkmajerjevo cesto na Ptuju. – Zbornik MARKO, P. 2011, Die villa Löfelbach – Polygonale Bau- splošne bolnišnice dr. Jožeta Potrča Ptuj, Ptuj, 73–91. formen in spätantiken Villen und Palästen. – In: MADZHAROV, M. 2012, Diocletianopolis. – In: Ivanov Bülow, Zabehlicky (eds.) 2011, 285–291. (ed.) 2012, 439–466. MARKOVIĆ, Č., S. STANKOVIĆ 1987, Ulcinj/Stari MALENKO, V. 2008, = MАЛЕНКО, В., Нови ан- grad. Citadela/Citadel. – Arheološki pregled 27 tи ч ки наоди во Oхрид и Oхридско (Recent (1986), 158–160. 406 BIBLIOGRAPHY MAROVIĆ, I. 1984, Reflexions about the Year of the MAURINA, B. 2020, New military evidence from the Destruction of Salona. – Vjesnik za arheologiju i site of Sant’Andrea at Loppio (Trentino, Italy). – In: historiju dalmatinsku 77. (Disputationes Saloni- Cavada, Zagermann (eds.) 2020a, 595–603. tanae II), 293–314. MCKAY, A. 1998, Houses, Vil as and Palaces in the Ro­ MARTIN, M. 1992, Volker Bierbrauer, Invillino – Ibligo man World. – London. in Friaul II. Die spätitantiken und frühmittelalter- MENTZOS, A. 2002, A Sketch of the History of Late lichen Kirchen. – Jahrbuch der schweizerische Roman Dion. – In: Slokoska et al. 2002, 333–341. Gesel schaft für Ur­ und Frühgeschichte, 260–263. MIGOTTI, B. 1994, Arheološka građa iz ranokršćanskog MARUŠIĆ, B. 1967, Kasnoantička i bizantinska Pula. – razdoblja u kontinentalnoj Hrvatskoj. – In: Od ne­ Kulturno - povijesni spomenici Istre 6, Pula. pobjedivog sunca do sunca pravde. Rano kršćanstvo MARUŠIĆ, B. 1975, Neki problemi kasnoantičke i u kontinetnalnoj Hrvatskoj, Zagreb, 41–70. bi zantske Istre u svjetlu arheoloških izvora. – MIGOTTI, B. 1997, Evidence for Christianity in Roman Jadranski zbornik 9 (1973–1975), Prilozi za povijest Southern Pannonia (Northern Croatia). – BAR, Istre, Rijeke i Hrvatskog primorja i Gorskog kotara, International Series 684. 337–350. MIGOTTI, B. 2015, Early Christian Archaeology in MARUŠIĆ, B. 1986, Istraživanje arheološkog nalazišta Northern Croatia: The State of Research. – In: Neue Rim kod Roča. – Izdanja Hrvatskog arheološkog Forschungen zum frühen Christentum in den Balkandruštva 11, 235 – 243. landern, Archäologische Forschungen 26, 61–68. MARUŠIĆ, B. 1987, Rim kod Roča. – Izdanja Hrvatskog MIHAJLOVSKI, R., J. ROSPENDOVSKI 2011 = arheološkog društva 11/2, 235–243. МИХАЈЛОВСКИ, Р., J. РОЗПЕНДОВСКИ, MASELLI SCOTTI, F. 1990, Tergeste. – Antichità Alto­ Јустинијанова епископска фонтана во Хераклеја adriatiche 36, 333–345. Линкестис (Justinianic episcopal fountain from MASELLI SCOTTI, F., E. MONTAGNARI KOKELJ Heraclea Lyncestis). – Macedoniae Acta Archaeo­ 1989, Castel azzo di Doberdò. − In: E. Montagnari logica 20 (2006–2008), 405–420. Kokelj (ed.), Il Carso Gorizziano tra Protostoria e MIJOVIĆ, P. 1984−1985, Was Olcinium included in the Storia. Da Castel azzo a San Polo, Gorizia, 15−96. Justinian’s limes? − Balcanoslavica 11−12, 73−86. MASON, Ph. 1998, Late Roman Črnomelj and Bela MIJOVIĆ, P., M. KOVAČEVIĆ 1975 = МИЈОВИЋ, П., krajina (Poznoantični Črnomelj in Bela krajina). – М. КОВАЧЕВИЋ, Градови и утврђења у Црној Arheološki vestnik 49, 285–313. Гори ( Gradovi i utvrđenja u Crnoj Gori) . – Beograd. MATEJČIĆ, I. 2007, Assetto urbano dell’antica Pa­ MIKL CURK, I. 1978, Poetovio v pozni antiki (Poetovio rentium e ubicazioni degli edifici sacri. – In: A. in der Spätantike). – Arheološki vestnik 29, 405−411. Augenti, C. Bertelli (eds.), Felix Ravenna. La croce, MIKL CURK, I. 2005, Verkehrsrichtungen, Wehran- la spada, la vela: l’alto Adriatico fra V e VI secolo, lagen und ihr Einfluß auf die Raumentwicklung Milano, 73–80. einer römischen Stadt - Beispiel Poetovio. − MATEJČIĆ, I. 2014, Euphrasiana – Cathedral in Poreč. Godišnjak XXXIV, Centar za balkanološka ispiti­ – Pula. vanja ANUBiH 32, 239−251. MATEJČIĆ, I., P. CHEVALIER 2012, L’episcopium de MIKULČIĆ, I. 1982a, = МИКУЛЧИЋ, И., Старо Poreč / Des domus ecclesiae aux palais épiscopaux . Скопје со околните тврдини ( Skopje und umge­ – In: S. Balcon-Berry, F. Baratte, J.-P. Cail et, P. bende Festungen in der Antike und dem Mittelalter). Sandron (eds.), Actes du col oque tenu à Autun du − Skopje. 26 au 28 novembre 2009, Bibliothèque de l’Antiquité MIKULČIĆ, I. 1982b, Der Untergang der Paläste im tardive 23, 163–172. spätantiken Stobi, Nordmakedonien. – In: Papen- MATIJAŠIĆ, R. 1982, Roman Rural Architecture in the fuss, Strocka (eds.) 1982, 535−544. Territory of Colonia Iulia Pola. – American Journal MIKULČIĆ, I. 1986, Spätantike Fortifikationen in der of Archaeology 86/1, 53–64. S.R. Makedonien. – In: Corso di cultura sul ’ arte MATIJAŠIĆ, R. 1988, Kasiodorova pisma kao izvor za ravennate e bizantina 33, 253–277. poznavanje kasnoantićke povijesti Istre (Cass. Var. MIKULČIĆ, I. 1999 = МИКУЛЧИЋ, И., Антички XII, 22, 23, 24). – Zgodovinski časopis 42, 363–371. градови во Македонија ( Antike Städte in der Re­ MATIJAŠIĆ, R. 1998, La produzione agricola in Istria publik Makedonien). – Skopje. nel VI-VII secolo. – In: Acta XIII Congressus 1998, MIKULČIĆ, I. 2002. Spätantike und frühbyzantinische 1107–1120. Befestigungen in Nordmakedonien. Städte, Vici, MATIJAŠIĆ, R., K. BURSIĆ MATIJAŠIĆ 1996, Antička Refugien, Kastel e. − Münchner Beiträge zur Vor- Pula s okolicom. – Pula. und Frühgeschichte 54. MAURINA, B. 2016, Ricerche Archeologiche a Sant’Andrea MIKULČIĆ, I. 2003, Stobi. An ancient city. − Skopje. di Loppio (Trento, Italia). Il Castrum Tardoantico­ MILAVEC, T. 2012, Late Antique settlements in Slovenia Altomedievale. – Archaeopress Archae ology 2016. after the year 600. – In: V. Ivanišević, M. Kazanski 407 Slavko CIGLENEČKI (eds.), The Pontic­Danubian Realm in the Period MILOŠEVIĆ, A. 2017, Arheologija na Bribiru: jučer - of the Great Migration, Posebna izdanja 51, CNRS danas - sutra. – In: Održivi razvoj arheoloških Monographies 36, Paris, Beograd, 71−88. lokaliteta Araća 2017, Novi Bečej, 186–205. MILAVEC, T. 2020, Late Antique fortified hil top settle- MILOŠEVIĆ, G. 2002, New data on the topography of ments in Slovenia: Fifteen years later. – In: Cavada, Viminacium. – In: The Roman and Late Roman city. Zagermann (eds.) 2020a, 159–174. The international conference at Veliko Turnovo MILIN, M. 2004, Bassianae. – In: Šašel Kos, Scherrer (26–30 July 2020), 151–158. (eds.) 2004, 253–268. MILOŠEVIĆ, G. 2011, A residental complex at Me- MILINKOVIĆ, M. 1983 = МИЛИНКОВИЋ, М., Rano- diana: the architectural perspective. – In: Bülow, vizantijsko utvrđenje na Đurđevcu u Đerekarima. – Zabehlicky (eds.) 2011, 167–176. Новопазарски зборник (Novopazarski zbornik) MILOŠEVIĆ ZAKIĆ, B. 2019, Sipar, arheološka istra- 7, 29–37. živanja i zaštita lokaliteta od 2013 do 2015. godine MILINKOVIĆ, M. 1985 = МИЛИНКОВИЋ, М., (Sipar, archaeological excavation and protection oft Рановизантијско утврђење на Тупом кршу и he site from 2013–2015). – Starohrvatska prosvjeta, околна утврђења у тутинској области (Ranovi- ser. 3, 46, 205–221. zantijsko utvrđenje na Tupom kršu i okolna MIRABELLA ROBERTI, M. 1974–1975, Il castrum di utvrđenja u tutinskoj oblasti). – Новопазарски Grado. – Aquileia Nostra 45–46, 565–574. зборник ( Novopazarski zbornik 9), 47–54. MIRABELLA ROBERTI, M. 1976, Iulium Carnicum MILINKOVIĆ, M. 1995, Die Gradina auf dem Jelica-Ge- centro romano alpino. – In: Aquileia e l’arco al­ birge und die frühbyzantinischen Befestigungen in pino orientale, Antichità Altoadriatiche IX (1976), der Umgebung von Čačak, Westserbien. – Antiquité 91–101. Tardive 3, 227−250. MIRKOVIĆ, M. 2011, Kontinuität und Diskontinuität MILINKOVIĆ, M. 2002, Die Byzantinische Höhen- bei der Entwicklung der Stadt Sirmium. – In: anlage auf der Jelica in Serbien - Ein Beispiel aus Heinrich-Tamáska (ed.) 2011a, 87–96. dem nördlichen Illyricum des 6. Jh. – Starinar 51 MITOVA-DŽONOVA, D. 1998, Das frühbizantinisches (2001), 71−133. castrum Stenos an der “via a Singiduno per Hadri- MILINKOVIĆ, M. 2007, Stadt oder “Stadt”: Frühbyzan- anopolim usque ad Constantinopolim” am Succi- tinische Siedlungsstrukturen im nördlichen Il­ Pass (Claustro; Stenos). – In: Acta XIII Congressus lyricum. – In: J . Henning (ed.), Post­Roman Towns, 1998, 531–540. Trade and Settlement in Europe and Byzantium 2. MLADENOVIĆ, D. 2009, Astral Path to Soul Salvation Byzantium, Pliska, and the Balkans, 159–191. in Late Antiquity? The Orientation of Two Late MILINKOVIĆ, M. 2008, Die spätantik-frühbyzanti- Roman Imperial Mausolea from Eastern Serbia. – nische befestigten Höhenanlagen in Serbien. – In: American Journal of Archaeology 113, 81–97. Steuer et al. 2008, 533–557. MLAKAR, Š. 1961, Terenska istraživanja u okolini MILINKOVIĆ, M. 2010, = МИЛИНКОВИЋ, М., Duklje (Einige Terrain-forschungen in der Umge- Градина на Јелици. Рановизантијски град и bung von Duklja in Montenegro). – Starinar n.s. средњовековно насељe ( Die Gradina auf der Jelica. 11 (1960), 215–224. Frühbyzantinische Stadt und mittelalterliche Sied­ MLAKAR, Š. 1975−1976, Fortifikacijska arhitektura lung). – Beograd. na otoku Brioni – “Bizantski Kastrum”. – Histria MILINKOVIĆ, M. 2015 = МИЛИНКОВИЋ, М., Archaeologica 6–7, 5–49. Рановизантијска насеља у Србији и њеном MÓCSY, A. 1974, Pannonia and Upper Moesia. A His­ окружењу (Frühbyzantinische Siedlungen in Serbien tory of the Middle Danube Provinces of the Roman und Umgebung). – Beograd. Empire. – London, Boston. MILINKOVIĆ, M. 2020, Ein Gebäude mit Speicherraum MÓCSY, A. (ed.) 1981, Die spätrömische Festung und auf der Jelica–Gradina (nördliches Illyricum des 6. das Gräberfeld von Tokod. – Akadémiae Kiadó, Jahrhunderts. – In: I. Popović, S. Petković (eds.), Budapest. ILLYRICVM ROMANVM. Studiola in honorem MODRIJAN, Z. 2017, New military finds from Miloje Vasić, Posebna izdanja. Arheološki institut Ančnikovo gradišče. – In: Dörfler et al. (eds.) 2017, Beograd 73, 268–286. 159–174. MILOŠEVIĆ, A. 1986, Bisko kod Sinja. – Arheološki MODRIJAN, Z. 2020, Ančnikovo gradišče. – In: Horvat pregled (1985), 60–61. et al. (eds.) 2020a, 349–362. MILOŠEVIĆ, A. 1989, Bisko kod Sinja. – Arheološki MODRIJAN, Z., T. MILAVEC 2011, Poznoantična pregled 28 (1987), 94–95. utrjena naselbina Tonovcov grad pri Kobaridu. MILOŠEVIĆ, A. 1998, Arheološka topografija Cetinje. – Najdbe / Late Antique fortified settlement Tonovcov Muzej hrvatskih arkeoloških spomenika. Katalozi grad near Kobarid. Finds. – Opera Instituti Archeo- i monografije 3. logici Sloveniae 23. 408 BIBLIOGRAPHY MOREAU, D., N. BEAUDRY 2021, La fortresse tardo- through interdisciplinary analysis. – Journal of antique de Zaldapa (Bulgarie): bilan des deux Cultural Heritage 28, 27–36. premières saisons de la Mission archéologique NIEMANN, G. 1910, Der Palast Diokletians in Spalato. internationale. – Bulletin de la Société française – Wien. d’archéologie classique, 170–175. NIEWHÖNER, P. 2007, Archaologie und die “Dunklen MORISSON, C., J.-P. SODINI 2002, The Sixth-Century Jahrhunderte” im byzantinischen Anatolien. – In: Economy. – In: A. E. Laiou (ed.), The Economic Henning (ed .), Post Roman Towns, Trade and Set­ History of Byzantium: From the Seventh through tlement in Europe and Byzantium 2. Byzatnium, the Fifteenth Century, Dumberton Oaks, 171–220. Pliska and the Balkans, 119–157. MOSSER, M. 2011, Das Legionslager Vindobona – Wien NIKOLOSKA, A. 2015, The House of Parthenius Re- zwischen Spätantike und Frühmittelalter. – In: visited. – Folia Archaeologica Balkanica 3, 83–104. Konrad, Witschel (eds.) 2011, 425−460. NOVŠAK, M. 2007, Celje – arheološko najdišče Celje. MRKOBRAD, D., A. JOVANOVIĆ 1989 = МРКОБРАД, – Varstvo spomenikov. Poročila 43, 28–30. Д., А. ЈОВАНОВИЋ, Самоград − археолошка ONČEVSKA TODOROVSKA, M. 2017, Late Antiquity истраживања (Samograd − archaeological ex- Settlement in the Area of the Roman Theatre at plorations in 1988). – Новопазарски зборник Scupi. – In: A. Jakimovski (ed.), Rimski teatar ( Novopazarski zbornik) 13, 31−46. Skupi / Roman Theatre Scupi, Skopje, 245–317. MRKOBRAD et al. 1990 = Mrkobrad, D., L. Glušac, ONČEVSKA TODOROVSKA, M. 2018, The Christian P. Lutovac 1990, Samograd kod Bijelog Polja. – religion in Scupi represented through the elements Arheološki pregled 29 (1988), 185–187. of the Basilica with atrium. – In: Giving gift to God. MÜLLER, R. 1988, Die spätrömische Festung Valcum Evidences of votive offerings in the sanctuaries, am Plattensee. – In: G. Bott, W. Meier-Arendt temples and churches. Proceedings of the 1st & 2nd (eds.), Germanen, Hunnen und Awaren. Schätze der International archaeological conference (Skopje, Völkenvanderungszeit, Nürnberg, 270–272. Kumanovo, 2016-2017), Kumanovo, 183–200. MÜLLER, R. 2000, Der Untergang der Antike und ihr OSMUK, N. 1985, Logje, Tolmin. – Gradec. – Varstvo Nachleben im nördlichen Pannonien (Transdanu- spomenikov 27, 294–295. bien). – In: R. Bratož (ed.), Slovenija in sosednje OSMUK, N. 1997, Ajdovščina – Castra. Stanje arheoloških dežele med antiko in karolinško dobo / Slowenien raziskav (1994) (Ajdovščina – Castra. Forschungs- und die Nachbarländer zwischen Antike und stand (1994)). – Arheološki vestnik 48, 119–130. karolingischer Epoche 1, Situla 39, 241–254. OSMUK, N. 2001, Fibula. – In: P. Bitenc, T. Knific (eds . ), MULVIN, L. 2002, Late Roman vil as in the Danube­ Od Rimljanov do Slovanov. Predmeti. – Ljubljana, 47. Balkan region. – BAR, International Series 1064. OŠTARIĆ, I., A. KURILIĆ 2013, Arheološka karta otoka MUNDELL MANGO, M. 2011, Androna in Syria. – In: Paga. – Novalja. Daim, Drauschke (eds.) 2011, 245–290. OVČAROV, D. 1973 = ОВЧАРОВ, Д., Протеихизмата MUNDELL MANGO, M. 2017, Androna and the late в системата на рановизантиските укрепљениа antique cities of Oriens. – In: Rizos (ed.) 2017a, по нашите земи (Proteihizmata v sistemata na 189–204. ranovizantiskite ukrepljenia po našite zemi). – MUNTEANU, F. 2011, Contribuţii la topografia Histriei Arheologia 15, 11–23. romane târzii (I) (Contributions to the topography OVČAROV, D. 1982 = ОВЧАРОВ, Д., Византијски и of Histria in the Late Roman period (I)). – Istorie булгарски крепости V­X век (Vizantijski i bulgarveche si arheologie 62, 233–249. ski kreposti V­X vek). – Sofia. MUŠIČ, B. 1999, Geophysical prospecting in Slovenia: OVERBECK, B. 1982, Geschichte des Alpenrheintals in an overview with some observatons related to the römischer Zeit I. − Münchner Beiträge zur Vor- und natural environment / Geofizikalna prospekcija v Frühgeschichte 20. Sloveniji: pregled raziskav z nekaterimi ugotovit- OŽANIĆ ROGULJIĆ et al. 2018 = Ožanić Roguljić, I., I. vami glede naravnega okolja. – Arheološki vestnik Miloglav, D. Tončinić 2018, Arheološka istraživanja 50, 349–405. na lokalitetu Banjače / Archaeological excavations NACEV, T. 2016, Bargala. – Skopje. of the Banjače site. − Opuscula Archaeologica 39/40, NACEV, T., I. ATANASOVA 2017, New Findings Re- 135−149. garding the Economic Part of Bargala. – Review of Paolo Diacono 2001 = Paolo Diacono e il Friuli altomethe Institut of History and Archaeology 17, 15–26. dievale (secc. VI.­X.). Atti del XIV. Congresso inter- NEMETH-EHRLICH, D., D. KUŠAN ŠPALJ 2003, nazionale di studi sul ’alto Medioevo (Cividale del Municipium Andautonia. – In: Šašel Kos, Scherrer Friuli, Bottenicco di Moimacco, 24-29 settembre (eds.) 2003, 107–129. 1999), Spoleto. NERI et al. 2017 = Neri, E., B. Gratuze, N. Schibille 2017, PAPENFUSS, D., V. M. STROCKA (eds.) 1982, Palast Dating the mosaics of the Durres amphitheatre und Hütte. Bauen und wohnen im Altertum. – Sym- 409 Slavko CIGLENEČKI posium der Alexander von Humboldt- Stiftung PETRIKOVITS, H. 1985, Duel. − In: Real exikon der (Berlin 25.-30. Nov. 1979), Mainz am Rhein. Germanischen Altertumskunde 6, Berlin, New PAROVIĆ-PEŠIKAN, M. 1981, Antička Ulpijana prema York, 226−238. dosadašnjim istraživanjima (Ulpiana antique à la PETROVA, S. 2019, The ancient and medieval settlement lumière de l’état actuel desrecherches). – Starinar at the vil age of Cherven breg, municipality of Dup- 32, 56–74. nitsa in Bulgaria (Preliminary report). – In: Niš and PAROVIĆ-PEŠIKAN, M. 1989, Neka zapažanja o ur- Byzantium. Seventeen international symposium banom razvoju Ulpijane – ispitivanje ulica / Some (Niš, 3–5 June 2018), The collection of scientific observations of the development of Ulpiana. Re- works 17, 157–186. search of the streets. – Lihnid 7, 117–132. PETROVIĆ, P., 1980, Les fortresses du bas-Empire sur PAŠKVALIN, V. 1980, Antički nimfej u Putovićima le limes Danubien en Serbie. – In: W. Hanson, J. kod Zenice. – Glasnik Zemaljskog muzeja Bosne i Keppie (eds.), Roman frontier studies 1979, BAR. Hercegovine u Sarajevu 34 (1979), 55–84. International Series 71, 757–773. PAŠKVALIN, V. 1990, Arheološko nalazište Ograja u PETROVIĆ, P. 1984, Saldum, rimsko i ranovizantijsko Putovićima kod Zenice / The archaeological site utrvrđenje na ušću potoka Kožica (Saldum, fortifi- of Ograja in Putovići near Zenica. – Glasnik Ze­ cation romain et paléobyzantine sur l’embouchure maljskog muzeja Bosne i Hercegovine u Sarajevu de la rivière Kožica). – Starinar 23–24 (1982–1983), 45, 35–92. 129–134. PAŠKVALIN V. 2003, Arheološki nalazi iz Mogorjela PETROVIĆ, P. 1986, Timacum Minus und die Kastele kod Čapljine (Archaeological finds from Mogorjelo im Timok-Tal. – In: Studien zu der Militärgrenzen near Čapljina). – Izdanja Hrvatskog arheološkog Roms III, 514–518. društva 22, 253–266. PETRU, P. 1982, Arheološki oris poznoantične poselitve PEDIŠIĆ, I. 2001, Ostaci stambene arhitekture u bi- Slovenije. – Zgodovinski časopis 36/4, 295–310. zantskoj utvrdi na otoku Žirju. – Histria Antiqua PETRU, P. 1978, Poznoantična poselitev Slovenije (Late 7, 123–130. Roman settlement in Slovenia). – Arheološki vestnik PERIŠA, D. 2009, Rimski Delminij kao sjedište staro- 29, 359–367. kršćanske biskupije (Roman Delminium as an early PETRU. P., S. PETRU 1978, Neviodunim (Drnovo pri Christian bishopric). – Arheološki radovi i rasprave Krškem). Katalog najdb. – Katalogi in monografije 16, 225–246. 15. PËRZHITA, L. 1993, Gradishta e Bardhocit / Gradishta PETRU, P., Th. ULBERT 1975, Vranje pri Sevnici. de Bardhoc. – Illiria 23, 219−239. Starokrščanske cerkve na Ajdovskem gradcu / Vranje PËRZHITA, L., G. HOXHA 2003, Fortifikime të shekujve bei Sevnica. Frühchristliche Kirchenanlagen auf dem IV ­ VI në Dardaninë perëndimore ( Late antiquity Ajdovski gradec. – Katalogi in monografije 12. castles in Western Dardania). – Tirane. PICCOTTINI, G. 1976, Das spätantike Gräberfeld von PETKOVIĆ, S. 2011, Gamzigrad – Romuliana in der Teurnia. – Klagenfurt. Zeit nach dem kaiserlichen Palast. – In: Bülow, PIPLOVIĆ, S. 2016, Dioklecijanova palača u Splitu na- Zabehlicky (eds.) 2011, 113–128. kon Careve smrti. – Vjesnik za arheologiju i historiju PETKOVIĆ, S. 2016 = Петковић, C., Заштитна архео- dalmatinsku 109/1, 259−290. лошка истраживања на локалитету Давидовац PIRKMAJER, D. 1994, Rifnik. Arheološko najdišče. – Градиште. Прелиминарни резултати /Archa- Vodnik. – Celje. eological rescue excavations at the site Davidovac PISU, N. 2019, Lo scavo 2013–2015 nel ’ area di Piedi- – Gradište. Preliminary results. – In: S. Perić, A. castello. – Antichità Altoadriatiche 90, 197–215. Bulatović (eds.), Археолошка истраживања на PISU, N. POSSENTI, E. 2020, Monte San Martino (Riva автопуту / Archaeological investigations along the Del Garda, TN). Alle soglie del medioevo: cosa highway Route E75 (2011–2014), Beograd, 301–349. cambia? – ADA– Archeologia delle Alpi, 49−59. PETKOVIĆ, S., B. ILIJIĆ 2012, Prilog poznavanju rim- PLESNIČAR GEC, L. 1983, Starokrščanski center v skog naselja na lokalitetu Timacium Minus kod Emoni (Old Christian Center in Emona). – Katalogi Knjaževca. – Glasinik SAD 28, 153−178. in monografije 21. PETRICIOLI, I. 1970, “Toreta” na otoku Kornatu. – PLESNIČAR GEC, L. 1997, Emona v pozni antiki v Adriatica Praehistorica et Antiqua, 1970, 717–725. luči arhitekture (Die Architektur Emonas in der PETRIKOVITS, H. 1971, Fortifications in the north- Spätantike). – Arheološki vestnik 48, 359–370. western Roman Empire from the third to the fifth PLESNIČAR GEC, L. 2005, Emona in Late An- centuries A. D. – Journal of Roman Studies 61, tiquity Based on New Discoveries. – In: Il­ 178−218. lyrica antiqua II. In honorem Duje Rendić­ Miočević. Proceedings of the international 410 BIBLIOGRAPHY conference / Radovi s međunarodnog skupa o POTTER, T. W. 1995, Towns in Late Antiquity: Iol Cae­ problemima antičke arheologije, 401–409. sarea and its Context. – Sheffield. PLESNIČAR GEC, L., I. SIVEC 1978, Emona at the PÜLZ, A. 2011, Das Stadtbild von Ephesos in byzanti- beginning of great migration period. – In: Problemi nischer Zeit. – In: Daim, Drauschke 2011, 541–571. seobe naroda u Karpatskoj kotlini / Probleme der POULTER, A. 1992, The use and abuse of urbanism in Völkerwanderungszeit im Karpatenbecken, 59–64. the Danubian provinces during the later Roman PLESTENJAK, A. 2020, Blagovica. – In: Horvat et al. empire. – In: Rich (ed.) 1992, 99–135. (eds.) 2020a, 241–248. POULTER, A. 1994, Novae in the 4th Century AD: city or POPOVIĆ, I. (ed.) 2005, Šarkamen (Eastern Serbia): fortress? A problem with a British perspective. – In: A Tetrarchic Imperial Palace. The Memorial Com­ G. Susini (ed.), Limes, Studia di Storia 5, 139–148. plex. − Belgrade. POULTER, A. 1995, Nicopolis ad Istrum: A Roman, Late POPOVIĆ, I. 2008, Figuralno zidno slikarstvo Sirmi­ Roman and Early Byzantine City. Excavations 1985– juma / Figural wal painting in Sirmium. – Beograd. 1992. – Journal of Roman Studies Monographs. POPOVIĆ, I. 2017, Sirmium au V et VI siècle. – In: I. POULTER, A. 1998, L’avenir du passé, Recherches sur Popović, M. Kazanski, V. Ivanišević (eds.), Sirmium la transition entre la péeiode Romaine et le monde à l’époque des Grandes Migrations, Leuven, Paris, Protobyzantin dans la région du Bas-Danube. – Bristol, 7–24. Antiquité Tardive 6, 329–343. POPOVIĆ, M. 1984 = ПОПОВИЋ, М., Утврђење на POULTER A. G. (ed.) 2007a, The Transition to Late Кули у Калудри (Utvrđenje na Kuli u Kaludri). – Antiquity. On the Danube and Beyond. – Proceed- Новопазарски зборник ( Novopazarski zbornik) ings of the British Academy 141. 8, 11–18. POULTER, A. G. 2007b, The Transition to Late Antiq- POPOVIĆ, M. 1987, Južac kod Sopočana. − Arheološki uity. – In: Poulter (ed.) 2007a, 1–50. pregled 27 (1986), 115–117. POULTER, A. G. 2007c, The Transition to Late Antiq- POPOVIĆ, M. 1988 = ПОПОВИЋ, М., Светиња, нови uity on the Lower Danube: the City, a Fort and подаци о рановизантијском Виминацијуму the Countryside. – In: Poulter (ed.) 2007a, 51–97. (Svetinja – Contribution to the Study of the Early POULTER, A. G. 2010, The Lower Danubian frontier in Byzantine Viminacium). – Starinar 38 (1987), 1–37. Late Antiquity: evolution and the dramatic change POPOVIĆ, M. 1999, Tvrđava Ras ( The Fortress of Ras). – in the frontier zone c. 296–600. – In: P. Herz, P. Beograd. Schmid, O. Stol , Zwischen Region und Reich: Das POPOVIĆ, M., V. BIKIĆ 2009, Vrsenice. Kasnoantičko i Gebiet der oberen Donau im Imperium Romanum, srpsko ranosrednjovekovno utvrđenje ( Late Roman Berlin, 11–42. and Serbian Early Medieval Fortress). – Beograd. POULTER, A. G. 2013, Dobry Dyal: a Late Roman POPOVIĆ, V. 1982, Desintegration und Ruralisation der fortress on the lower Danube. – Journal of Roman Stadt im Ost-Illyricum vom 5. bis 7. Jahrhundert n. Archaeology 26, 362–382. Chr. – In: Papenfuss, Strocka (eds.) 1982 , 545–566. POULTER, A. G., E. MARKI 1998, Field survey at Lou- POPOVIĆ, V. 1984, Donji Milanovac – Veliki Gradac loudies: the discovery of a new Late Roman city in ( Taliata), rimsko i ranovizantinsko utvrđenje (Donji the Pieria. – Το Αρχαιολογικό Έργο στη Μακεδονία Milanovac – Veliki Gradac ( Taliata), fortresse roma- και Θράκη 9, 179–193. no-byzantine). – Starinar 23–24, 265–282. PREMOVIĆ-ALEKSIĆ, D. 1989 = ПРЕМОВИЋ- POPOVIĆ, V. 1987, Die süddanubischen Provinzen in АЛЕКСИЋ, Д., Касноантичко утврђење у der Spätantike vom Ende des 4. bis zur Mitte des Бабрежу (The Late-Antiquity Fort in Babrez near 5. Jahrhunderts. – In: Die Völker Südosteuropas Novi Pazar). – Новопазарски зборник (Novopazarim 6. bis 8. Jahrhundert, Südosteuropa-Jahrbuch ski zbornik) 13, 17–29. 17, 95–140. PRESHLENOV, H. 2012, Mesambria Pontica in orbis POPOVIĆ, V. 2004, Sirmium. Grad careva i mučenika. – Romanus. – In: Ivanov (ed.), 2012, 493–536. Sremska Mitrovica. PRESHLENOV, H. 2018, Late Antique Mesembria: (Re) PORTE, P. 2011, Larina et son territoire de l’Antiquité au Shaping of Public Spaces. – In: Proceedings of the Moyen Âge 1. Études archéologiques et historiques. first Internatonal Roman and Late Antique Thrace – Biarritz. Conference, Sofia, 393–407. POSSENTI, E. 2008, Altinum, la città e la Chiesa di Eli- PRINGLE, D. 1981, The Defence of Byzantine Africa odoro. – In: S. Piussi (ed.), Cromazio di Aquileia: from Justinian to the Arab Conquest. An account of al crocevia di genti e religioni, Milano, 416–419. the military history and archaeology of the African POSSENTI, E. 2013, Castelli tra tardo antico e alto provinces in the sixth and seventh centuries. – BAR, medioevo nel ’arco alpino centrale. – In: Castra, International Series 99. castel i e domus murate. Corpus dei siti fortificati PROVOST, S. 2001, City wall and urban area in Macedo- trentini tra tardo antico e basso medioevo, APSAT 6. nia: the case of Philippi. – In: Recent research in Late 411 Slavko CIGLENEČKI Antique Urbanism, Journal of Roman Archaeology. REGAN, K., B. NADILO 2009, Stare crkve u jugoistoč- Suppl. 43, 123–135. nom dijelu otoka Krka. – Građevinar 61, 1075–1087. QUAST, D. 2011, Eine völkerwanderungszeitliche REKA, V. 2015, Post-Roman Landscape in Pannonia: Höhensiedlung in Velem Szentvid? – In: Heinrich- 5th- Century Archaeological Sites. – In: Romania Tamáska (ed.) 2011a, 97–101. Gothica II, 669–677. QUIROGA, J. L. 2016, Early Byzantine Urban Land- RICH, J. (ed.) 1992, The City in Late Antiquity. – Studies scapes in the Southwest and Southeast Mediter- in Ancient Society 3. ranean. – In: Proceedings of the 23rd International RIEDL, E., O. CUNTZ 1909, Uranje in Steiermark, Congress of Byzantine Studies, Belgrade, 69–105. römische Bauten und Grabmäler. – Jahrbuch für RADIČEVIĆ, D. 2009 = РАДИЧЕВИЋ, Д., Прилог Altertumskunde 3, 1–34. проучавању рановизантијских фибула на тлу RIPOLL, G., J. ARCE 2000, The transformation and End Србије са посебним освртом на налаз са Лишке of Roman Vil ae in the West (fourth-seventh cen- ћаве код Гуче (Contribution to the Investigations turies): Problems and Perspectives. – In: Brogiolo of the Early Byzantine Fibulae in the Territory of et al. (eds.) 2000, 63–114. Serbia Considering Particularly the Find from RISTOV, K. 2012, Tauresium. Монети од археолошки Liška Ćava near Guča). – Zbornik Narodnog muzeja ископувања ( Tauresium. Coins from archaeologi­ 19/1, 403–421. cala excavations). – Skopje. RADOSLAVOVA, G. 2011, Abritus – eine spätrömisch- RISTOV, K. 2018, Маркови Кули Скопје – Источно byzantinische Stadt in Moesia Secunda. – In: Hein- подградие (Markovi Kuli Skopje – Eastern Sub- rich Tamáska (ed.) 2011a, 249–256. urb). – Folia Archaeologica Balkanica 4, 483–512. RAMA, Z. 2020, Dardania in Late Antiquity: An RISTOV, K. 2019, Taor. Early Byzantine Fortification Overview of 4th-6th Century Fortifications in the and Settlement from Prehistory to Middle Ages / Territory of Kosovo. – Archaeologia Adriatica 14, ТАОР Рановизантиско утврдување и населба 111–137. од праисторијата до средниот век. – Skopje. RAMSTETTER, K. 2019, Sieben Jahrhunderte in der RIZOS, E. 2011, The Late Antique wal s of Thessalonica Römerstadt Teurnia – Die Bebauung am Nor- and their place in the development of eastern mili- dostrand des Holzerberges. – Rudolfinum. Jahrbuch tary architecture. – Journal of Roman Studies 24, des Landesmuseums für Kärnten 2018, 30–38. 450–468. RANDSBORG, K. 1991, The First Mil ennium A. D. in RIZOS, E. 2012, The Making of Christian Society in the Europe and the Mediterranean: An Archaeological Late Antique Civil Diocese of Macedonia – Achaeo- Essay. – Cambridge. logical Evidence on Christianisation from Modern RAPAN PAPEŠA, A., D. ROKSANDIĆ 2016, Cibalae/ Greece. − In: O. Heinrich-Tamáska, O. Krohn, S. Vinkovci during Late Antiquity (fifth to sixth cen- Ristow (eds.), Christianisierung Europas. Entstehung, tury AD) – new insights about old assumptions. – Entwicklung und Konsolidierung im archäologischen In: Bugarski et al. (eds.) 2016, 145–161. Befund / Christianisation of Europe. Archaeological RAPANIĆ, Ž. 1984, Kasnoantička “palača” u Ostrvici evidence for it’s creation, development and consolida­ kod Gata (Poljica). – In: Cetinska krajina od prettion, Internationale Tagung im Dezember 2010 in historije do dolaska Turaka, Izdanja Hrvatskog Bergisch-Gladbach, Regensburg, 319−342. arheološkog društva 8, 149–162. RIZOS, E. 2013, Centers of the Late Roman military RAŠKOVIĆ, D. 2016, Ukosa u gradu Stalać. – Glasnik supply network on the Balcans: a survey of Hor- Srpskog arheološkog društva 32, 285–303. rea. – Jahrbuch des Römisch­Germanische Zentral­ RAŠKOVIĆ, D. 2021, Arheološka topografija i istorijska museums Mainz 60, 659–696. geografija centralnog Pomoravlja u ranovizan- RIZOS, E. (ed.) 2017a, New Cities in Late Antiquity. tijskom periodu. – Niš i Vizantija 19, 269–285. Documents and Archaeology. – Bibliothèque de RAUTMAN, M. 2017, Sardis. – In: P. Niewöhner (ed.), l’Antiquité Tardive 35. Archaeology of Byzantine Anatolia: From the end of RIZOS, E. 2017b, New cities and new urban ideals, AD Late Antiquity until the coming of Turks, Oxford, 250 – 350. – In: Rizos (ed.) 2017a, 19–38. 231–237. ROBERTSON BROWN, A. 2018, Corinth in Late Antiq­ RAYNAUD et al. 2003 = Raynaud M.-P., P. Chevalier, uity: A Greek, Roman and Christian City. – Library J.-P. Sodini, C. Vanderheyde, M. Wurch-Koželj, of Classical Studies 17. N. Beaudry, S. Muçaj 2003, Trois basiliques et un ROFFIA, E. 1997, Sirmione, le “grotte di Catullo”. – In: groupe episcopal des Ve-VIe siècles réétudiés à Byllis E. Roffia (ed.), Vil e romane sul Lago di Garda, De- (Albanie). – Hortus Artium Medievalium 9, 115–165. seznzano del Garda e del a Riviera, Brescia, 17–52. REDŐ, F. 2003, Municipium Aelium Sal a – In: Šašel ROSADA, G. (ed.) 1999, Oppidum Nesactium. Una città Kos, Scherrer (eds.) 2003, 191–235. istro­romana. – Canova, Traviso. 412 BIBLIOGRAPHY ROSSI, F. 2003, Transformazioni nel ’ edilizia abitativa race at Golemo Gradište, Konjuh, 2007–2010. – urbana a Brescia. – In: Abitare in città: la Cisalpina Macedoniae Acta Archaeologica 20, 347–364. tra impero e medioevo; convegno tenuto a Roma SANTORO BIANCHI, S. 1992, I risultati dello scavo. – il quattro e cinque novembre 1999 / Leben in der In: Castelraimondo. Scavi 1988–1990, I. Lo scavo, Stadt: Oberitalien zwischen römischer Kaiserzeit Cataloghi e monografie dei Civici musei di Udine und Mittelalter. Convegno / Kolloquium (Roma, 2, 125–227. 4-5 novembre 1999), 27–35. SARADI, H. G. 2006, The Byzantine City in the Sixth RUPEL, L. 1997, Rinvenimenti di superficie di materiale Century. Literary Images and Historical Reality. – archeologico nel comune di Raveo. – Forum Iulii Athens. 20 (1996), 35–41. SARANTIS, A. 2013, Waging War in Late Antiquity. – In: RUTAR, S. 1886, Ruine St. Paul im Wippacher Thale. – Sarantis, Christie (eds.) 2013 , 1–100. Mittheilungen der K. K. Central­Commission 12, SARANTIS, A., N. CHRISTIE (eds.) 2013, War and 81–82. Warfare in Late Antiquity. Current Perspectives. − SAGADIN, M. 1987, Ajdna nad Potoki. Poznoantično Late Antique Archaeology 8. naselje. − Arheološki pregled 28, 123−124. SARANTIS, A. 2016, Justinian’s Balkan Wars: Cam­ SAGADIN, M. 1988, Kranj – Križišče Iskra / Kranj – Iskra paigning, Diplomacy and Development in Il yricum, crossroads. – Katalogi in monografije 24. Thrace and the Northern World A.D. 527­65. – SAGADIN, M. 1991, Najstarejša cerkvena stavba v Kra- ARCA: Classical and Medieval Texts, Papers and nju. – In: Pod zvonom Sv. Kancijana, Kranj, 31–44. Monographs 53. SAGADIN, M. 1997, Ajdna nad Potoki. – Kulturni in SARIA, B. 1929, Začasno poročilo o izkopavanjih na naravni spomeniki Slovenije. Zbirka vodnikov 190. Gradišču pri Vel. Malenci. – Glasnik Muzejskega SAGADIN, M. 1998, Carniola u svjetlu novih nalaza. – društva za Slovenijo 10, 11–17. In: Acta XIII Congressus 1998, 71–724. SARIA, B. 1930, Drugo začasno poročilo o izkopavanjih SAGADIN, M. 2004, Poznoantična steklarska delavnica na Gradišču pri Vel. Malenci. – Glasnik Muzejskega (?) v Kranju (A Late Roman glass workshop (?) in društva za Slovenijo 11, 5–12. Kranj). – In: Lazar I. (ed.), Drobci antičnega stekla / SARIA, B. 1939, Doneski k vojaški zgodovini naših Fragments of Ancient glass, Annales Mediterranea, krajev v rimski dobi. – Glasnik Muzejskega društva 107–114. za Slovenijo 20, 115–151. SAGADIN, M. 2006, Potoki pri Žirovnici. – Varstvo SCHERRER. P. 2003, Savaria. – In: Šašel Kos, Scherrer spomenikov 39–41, 155–156. (eds.) 2003, 53–80. SAGADIN, M. 2008, Od Karnija do Kranja. – Unpub- SCHERRER, P. 2011, Noricum in der Spätantike – Zu lished doctoral dissertation, Univerza v Ljubljani, den Forschungen des vergangenen Jahrzehnts. Ljubljana. Mit einem Beitrag von Bernhard Schrettle, Neue SAGADIN, M. 2020, Carnium – Kranj. – In: Horvat et Forschungen in Rannensdorf und auf dem Frauen- al. (eds.) 2020a, 201–212. berg. Zur spätantike im Umfels von Flavia Solva. – SÁNCHEZ, J. V. 2009, La presencia bizantina en Hispa­ In: Heinrich-Tamáska (ed.) 2011a, 103–128. nia (siglos VI­VII). La documentación arqueológica SCHMID, W. 1943, Die Fortschritte der vorgeschichtli- (2009). Antigüedad y Cristianismo. − Monografias chen Forschung in Südsteiermark zwischen den sobre la Antiguedad tardia 24 (2007), Murcia. beiden Weltkriegen. – Zeitschrift des Historischen SÁNCHEZ-PARDO, J. C., C. J. GALBÁN MALAGÓN Vereins für Steiermark 36, 134–151. 2015, Fortificaciones de altura en el entorno de SCHNEIDER, L. 2001, Oppida et castra tardo-antiques. Santiago de Compostela. Hacia un primer análisis A propos des établissements de hauteur de Gaule arqueológico comparativo. Universidade de San- méditerranéenne. – In: Les campagnes de la Gaule tiago de Compostela. – NAILOS: Estudios Interdis­ à la fin de l’Antiquité, Actes du colloque AGER IV ciplinares de Arqueologia 2, 125–161. (Montpellier, 11-14 mars 1998), Editions APDCA, SANDERS, G. 2004, Problems in Interpreting Rural and 433–448. Urban Settlement in southern Greece, AD. 365- SCHNEIDER, L. 2003, Nouvelles recherches sur les 700. – In: N. Christie (ed.), Landscapes of Change habitats de hauteur de l’Antiquité tardive et du Rural Evolutions in Late Antiquity and the Early haut Moyen Âge en Gaule du Sud- Est : le cas du Middle Ages. – London. Roc de Pampelune (Hérault). – Les Nouvelles de SÁNDOR, M. 2001, Sopianae und Quiqueecclesiae. Die l’archéologie 92, 9–16. Frage der Stadtkontinuität zwischen römischer SCHNEIDER, L. 2007, Structures du peuplement et und mittelalterlicher Siedlung. – Beiträge zur Mit­ formes de l’habitat dans les campagnes du Sud-Est telalterarchäologie in Österreich 17, 233–239. de la France de l’Antiquité au Moyen Âge (IVe-VIIIe SANEV et al. 2012 = Sanev, G., C. S. Snively, M. Sto- siècles). Essai de synthèse. – Gallia 64, 11–56. janoski 2012, Excavations on the Northern Ter- 413 Slavko CIGLENEČKI SCHNURBEIN, S. 2001, Abriβ der Entwicklung der SNIVELY, C. S. 2002, Archaeological Investigation at Römisch-Germanischen Kommission unter den Konjuh, Republic of Macedonia in 2000. – Dum­ einzelnen Direktoren von 1911 bis 2002. – Bericht barton Oaks Papers 56, 297–306. der Römisch­ Germanischen Komm ission 82, SNIVELY, C. S. 2008, Golemo Gradishte, vil age Konjuh. 137−289. – Makedonski arheološk pregled 1, 179–181. SCHRETTLE, B. 2007, Römische Villen in Umland von SNIVELY, C. 2009, The Fates of Balkan Cities in Late An- Flavia Solva. Überlegungen zur Verbreitung der tiquity: Recent Views. – In: Niš i Vizantija 7, 37–44. bautypen in Südostnoricum. – Schild von Steier SNIVELY, C. S. 2011, The new basilica at Golemo Gra- 20, 253–267. dište, Konjuh: a sixth century church in the prov- SCHRETTLE, B. 2014, Der römische Tempelbezirk ince of Dardania. – In: Niš i Vizantija 9,187–201. auf dem Frauenberg bei Leibnitz Architektonische SNIVELY, C. S. 2017, Golemo Gradište at Konjuh: Funde der Grabungen 1951 bis 1953 und 1974 sowie A New City or Relocated One? – In: Rizos (ed.) Ergebnisse der archäologischen Untersuchungen 2017a, 205–220. 2008 bis 2013. – Fundberichte aus Österreich. SNIVELY, C. S., G. SANEV 2013, Late antique residences Materialhefte A 21. at Golemo Gradište, Konjuh, R. Macedonia. – In: SDROLIA, S. 2016, Habitation in the region of Mount Niš i Vizantija 11, 157–171. Ossa during the early Byzantine period. – Archaeol­ SODINI, J.-P. 1984, L’habitat urbain en Grèce à la veille ogy in Greece 2015–2016, 125–132. des invasions. – In: Villes 1984, 314–397. SDROLIA, S. 2022, Late antique hil top sites in Greece. SODINI, J.-P. 1995, Habitat de l’antiquité tardive 1. – – In: D. Pavlovič, O. Heinrich-Tamáska (eds.), Book Topoi 5 151–218. of Abstracts. International Workshop “Intentions SODINI, J.-P. 1997, Habitat de l’antiquité tardive 2. – and Meaning: A Comparative View of Late Antique Topoi 7/2, 435–577. Hil top Sites in Europe. ” (National Museum of Slo- SODINI, J. P. 2003, Archaeology and Late Antique social venia, Ljubljana, 22nd to 24th September 2021), Practices. – In: Lavan, Bowden (eds.) 2003, 25−56. Ljubljana, 30. SODINI, J.-P. 2007, The Transformation of Cities in Late SENNHAUSER, H. R. 1979, Der Profanbau. – In: Ur­ Antiquity within the Provinces of Macedonia and und frühgeschischtliche Archäologie der Schweiz 6, Epirus. – In: Poulter (ed.) 2007a, 311–336. Das Frühmittelalter, 149–164. SODINI et al. 1980 = Sodini, J. P., G. Tate, J-L. Biscop, SERGEJEVSKI, D. 1942, Putne bilježke iz Glamoča. – B. Bavant, D. Orssaud 1980, Déhès (Syrie du Nord). Glasnik Zemaljskog muzeja Bosne i Hercegovine u Campagnes I à III ( 1976­1978). – Recherches sur Sarajevu 54, 147–148, Fig. 20. l’habitat rural 15 (Syria). SIMIĆ, Z. 1987 = СИМИЋ, З., Утврђење на Кулини SOKOL, V. 1981, Najnovija arheološka istraživanja u у Рогатцу (Fortess on Kulina in Rogatac). – Prigorju. – Arheološka istraživanja u Zagrebu i Новопазарски зборник (Novopazarski zbornik) njegovoj okolici. Znanstveni skup o 100. obljetnici 11, 13–20. društva (Zagreb, od 14. do 16. studenog 1978. SKARIĆ, V. 1932, Altertümer von Gradac in der godine), Izdanja Hrvatskog arheološkog društva Lepenica (Bosnien). – Glasnik Zemaljskog muzeja 6, 169–185. u Bosni i Hercegovini 44, 1–20. SOKOL, V. 1994, Das spätantike Kastrum auf dem SIMONI, K. 1989, Funde aus der Völkerwanderungszeit Kuzelin bei Donja Glavnica. – Arheološki vest nik in den Sammlungen des Archäologischen Muse- 45, 199–209. ums in Zagreb. – Vjesnik Arheološkog muzeja u SOKOL, V. 1998, Rimski metal s Kuzelina / Roman Metal­ Zagrebu, 3. s., 22, 107–134. work from Kuzelin. – Sesvete, Zagreb. SLABE, M. 1975, Dravlje. Grobišče iz časov preseljevanja SOPRONI, S. 1978, Der spätrömische Limes zwischen ljudstev. – Situla 16. Esztergom und Szentendre. Das Verteidigungssystem SLABE, M. 1978, Govorica arheoloških ostalin o času der Provinz Valeria in 4. Jahrhundert. – Akadémiai selitve ljudstev na Slovenskem (Report on archaeo- Kiadó, Budapest. logical remains from the period of the migration of SOPRONI, S. 1985, Die letzten Jahrzehnte des pannoni­ peoples on Slovene territory). – Arheološki vestnk schen Limes. – Münchner Beiträge zur Vor- und 29, 379–391. Frühgeschichte 38. SLAPŠAK, B. 1978, Rodik-Ajdovščina. – Arheološki SOSSAU, V. 2018, Das Municipium Claudium Aguntum vestnik 29, 546–547. in der Spätantike. Eine erste Bestandsaufnahme. – SLAPŠAK, B. 1997, Starejša zgodovina Rodika. – In: In: M. Auer, H. Stadler (eds.), Ager Aguntinus, Rodik med Brkini in Krasom, Koper, 19–64. Historisch-archäologische Forschungen 1. SLOKOSKA et al. (eds.) 2002 = Slokoska, L., R. Ivanov, SOTINEL, C. 2005, Identité civique et christianisme. V. Dinchev (eds.) 2002, The Roman and Late Ro­ Aquilée du IIIe au VIe siècle. – Bibliothèque des man City. – Sofia. Écoles françaises d’Athènes et de Rome 324. 414 BIBLIOGRAPHY SPASOVSKA-DIMITROSKA, G. 1989 = СПА СОВ - STEINKLAUBER, U. 2020, Die Spätantike auf der Ka- СКА-ДИМИТРОСКА, Г., Остатоци на dischen. – Neues aus Alt­Vil ach 57, 55–99. доцноантичката архитектура на градиштето STEUER et al. (eds.) 2008 = Steuer, H., V. Bierbrauer, M. ‘Венец’ во Миокази­Кичевско ( Remains of lateanti Hoeper (eds.) 2008, Höhensiedlungen zwischen An­ que architecture on the site “Venec” at Miokazi near tike und Mittelalter von den Ardennen bis zur Adria. Kichevo). – Лихнид / Lychnid 7, 43–49. – Ergänzungsbände zum Reallexikon der Germani- SPIESER, J.-M. 1984, La ville en Grèce du IIIe au VIIe schen Altertumskunde 58, Berlin, New York . siècle. – In: Vil es et peuplement dans l’Il yricum STEVOVIĆ, I. 2014, Praevalis. Obrazovanje kulturnog protobyzantin. Actes du colloque de Rome (12-14 prostora kasnoantičke provincije / Praevalis / The mai 1982), 315–40. Making of the cultural Space of the Late Antique SREJOVIĆ, D., A. SIMOVIĆ 1959, Portret vizantijske Province. − Društvo arheologa Crne gore, Posebna carice iz Balajnca (Portrait d’une imprératrice byz- izdanja, Knjiga 1, Podgorica. antine de Balajnac). – Starinar 9–10 (1958−1959), STEPHENS CRAWFORD, J. 1990, The Byzantine Shops 77–85. at Sardis. – Archaeological Exploration of Sardis, SREJOVIĆ, D., Č. VASIĆ 1994, Imperial mausolea Monograph 9, Cambridge, Mass., London. and consecration memorials in Felix Romuliana. – STICOTTI, P. 1913, Die römische Stadt Doclea in Beograd. Montenegro. – Schriften der Balkankommission, STAMENKOVIĆ, S. 2013, Rimsko nasleđe u Leskovačkoj Antiquarische Abteilung 6. kotlini ( Roman Legacy in the Leskovac Val ey). – STRIČEVIĆ, D. 1961, Uvod u ispitivanje unutrašnjosti Arheološki institut, Posebna izdanja 53. romejskog limesa u Iliriku. – Limes u Jugoslaviji 1, STARAC, A. 2001, Neke spoznaje o bedemima Pule. – Beograd, 177–184. Histria archaeologica 7, 61–68. STRMČNIK, M. 1997, Mariborsko - bistriško območje STARAC, A. 2009, Quarter of St. Theodor in Pula. – v poznorimski dobi (Der Raum Maribor und Bis- Annales 19, 271–290. trica in spätrömischer Zeit). – Arheološki vestnik STARAC, A. 2010, Dragonera. Dva bisera / Two pearls. – 48, 269–288. Arheološki muzej Istre Pula. Monografije i kata- STRMČNIK GULIČ, M. 1990, Podoba antične poselitve lozi 19. med vzhodnim Pohorjem in Dravo (Abriß der an- STARAC, R. 1993, Prikaz rezultata pokusnih arheoloških tiken Besiedlung zwischen dem Ostteil des Pohorje istraživanja na Gradini kod Pasjaka. – Umjetnost na und der Drava). – Arheološki vestnik 41, 135–146. istočnoj obali Jadrana u kontekstu europske tradicije, STRMČNIK GULIČ, M. 2020, Slovenska Bistrica. – In: Rijeka, 27–36. Horvat et al. (eds.) 2020a, 339–348. STARAC, R. 2005, Cickini, Sršići. – Hrvatski arheološki STRMČNIK GULIČ, M., S. CIGLENEČKI 2003, godišnjak 2, 255–256. Ančnikovo gradišče pri Jurišni vasi. Poznoantična STARAC, R. 2009, Liburnijski limes – arheološko- in zgodnjesrednjeveška naselbina na Pohorju. – konservatorski radovi na Lokalitetima Vranjeno Slovenska Bistrica. i Za Presiku / Limes Liburnico – Ricerche ar- STRUBE, C. 2011, Al Andarin, das antike Androna. – In: cheologiche e lavori di recupero nelle località di Daim, Drauschke (eds.) 2011, 211–244. Vranjeno e Za Presikom. – In: L. Bekić, N. Radić STUPPNER, A. 2011, Zur Kontinuität in der Spätantike Štivić (eds.) , Tarsatički principij. Kasnoantičko vojno am norisch-pannonischen Limes in Niederöster- zapovjedništvo / Principia di Tarsatica. Quartiere reich. – In: Heinrich Tamáska (ed.) 2011a, 129–156. generale d’epoca tardoantica, Rijeka, 275–287. SUIĆ, M. 1960, Arheološka istraživanja u Mulinama na STARE, V. 1980, Kranj. Nekropola iz časa preseljevanja o. Ugljanu (Scavi archeologici a Muline sul ’isola di ljudstev. – Katalogi in monografije 18. Ugljan). – Ljetopis Jugoslavenske akademije znanosti STENKLAUBER, U. 1990, Der Duel und seine Klein- i umjetnosti 64/1957, 230–249. funde. – Carinthia I, 180, 109–136. SUIĆ, M. 1976a, Antički grad na istočnom Jadranu. – STEINKLAUBER, U. 2002, Das Spätantike Gräberfeld Zagreb. auf dem Frauenberg bei Leibnitz, Steiermark. – SUIĆ, M. 1976b, Obrambeni sustav i prospekt antičkog Fundberichte aus Österreich. Materialhefte A 10. Zadra s kopnene strane. – Radovi Filozofskog STEINKLAUBER, U. 2005, Inneralpine spätantike fakulteta u Zadru 6, 533–554. Höhensiedlungen im steirischen Ennstal. – Schild SUIĆ, M. 1980, Faze izgradnje bedema stare Varvarije. von Steier 18, 135−198. Uz jedan novi natpis iz Bribira. – Gunjačin zbornik. STEINKLAUBER, U. 2013, Fundmaterial spätan­ U povodu sedamdesete godine života i četrdeset i pete tiker Höhensiedlungen in Steiermark und Kärnten. godine znanstvenog rada, 31–42. Frauen berg im Vergleich mit Hoischhügel und Duel. SUIĆ, M. 1981, Zadar u starom vijeku. – In: Prošlost – Forschungen zur geschichtlichen Landeskunde Zadra 1, Posebna izdanja / Sveučilište u Splitu, der Steiermark 61. Filozofski fakultet Zadar, 347−366. 415 Slavko CIGLENEČKI SUIĆ, M. 1996, Odabrani radovi iz stare povijesti hrvat­ TALEVSKI, H. 2018, The Problem of the Early Byzantine ske. Opera selecta. – Zadar, 689–724. Settlement and the Questions for Stobi During SUIĆ, M. 2003, Antički grad na istočnom Jadranu (2nd the Middle Ages. – Folia Archaeologica Balkanica updated edition). – Zagreb. 4, 427–465. SVOLJŠAK, D. 1985, Sv. Pavel nad Vrtovinom. Rezultati TASSAUX, F. 2003, Orsera / Vrsar, une vil a maritime sondiranj leta 1966 (Sv. Pavel über Vrtovin. Ergeb- de l’Antiquité tardive en Istrie. – Histria antiqua nisse der Sondierungen im J. 1966). – Arheološki 11, 383–390. vestnik 36, 195 – 236. TATE, G. 1992, Les campagnes de la Syrie du Nord au SVOLJŠAK, D. 2015, Roman fort at Hrušica: Findings IIe au 7e siécle 1. – Institute Français d’arhéologie of investigations after 1979 / Rimska utrdba na du Porche - Orient, Bibliothèque archéologique et Hrušici, izsledki izkopavanja po letu 1979. – In: historique 134. Istenič et al. (eds.) 2015, 341–364. TAVANO, S. 1976, Grado. Guida storica e artistica. – SZABÓ, A., O. HEINRICH-TAMÁSKA 2011, Eine Udine. spätromische Innenbefestigung in Környe. – In: TCHALENKO, G. 1953–1958, Vil ages antiques de la Heinrich-Tamáska (ed.) 2011a, 47–60. Syrie du Nord: le massif du Bélus à l’époque romaine SZŐNYI, E. 2004, Mursel a. – In: Šašel Kos, Scherrer 3. – Bibliothèque archéologique et historique 50. (eds.) 2004, 85–98. TEICHNER, F. 2011, “Nan primum tibi mater Hispania ŠAŠEL, J. 1974, Koper (Capodistria (Koper)) – est, terris omnibus terra felicior”. Spätantike Groß- Arheološki vestnik 25 (1976), 446–461. villen und Residenzen auf der Iberischen Halbinsel. ŠAŠEL, J. 1980, Zur Historischen ethnographie des mit- – In: Bülow, Zabehlicky (eds.) 2011, 293–310. tleren Donauraums. – In: Die Völker an der mittleren TEICHNER, F. 2015a, Ulpiana – Iustiniana Secunda und unteren Donau im fünften und sechsten Jh., (Kosovo): das urbane Zentrum des dardanischen Wien, 13–17. Bergbaubezirks. – Ephemeris Napocensis 25, 81–93. ŠAŠEL, J., P. PETRU 1971, Claustra Alpium Iuliarum I, TEICHNER, F. 2015b, Neue Forschungen zur dardani- Fontes. – Katalogi in monografije 5. schen Doppelstadt Ulpiana – Iustiniana Secunda ŠAŠEL KOS, M. 1994, The embassy of Romulus to At- (Kosovo). – Bericht der Römisch­Germanischen tila: one of the last citations of Poetovio in classical Kommission 96, 281–338. literature. – Tyche 9, 99–111. TEICHNER, F. 2018, On the ancient twin-city of Ulpi­ ŠAŠEL KOS, M., P. SCHERRER (eds.) 2002, The Auana­Iustiniana Secunda (Kosovo): Capital of the tonomous Towns of Noricum and Pannonia / Die Metalla Dardanica. – In: J. M. Álvarez, T. Nogales, autonomen Städte in Noricum und Pannonien. I. Rodà (eds.), CIAC. Proceedings XVIIIth Interna­ Noricum. – Situla 40. tional Congress of Classical Archaeology 1, 271–275. ŠAŠEL KOS, M., P. SCHERRER (eds.) 2003, The Au­ TEICHNER, F., M. UGARKOVIĆ 2012, Zeugnisse einer tonomous Towns of Noricum and Pannonia / Die römischen Vil a maritima auf der Insel des Heiligen autonomen Städte in Noricum und Pannonien. Clemens (Dalmatien). – Germania 90, 97–125. Pannonia I. – Situla 41. THOMAS, E. 1964, Römische Vil en in Pannonien. – ŠAŠEL KOS, M., P. SCHERRER (eds.) 2004, The Au­ Budapest. tonomous Towns of Noricum and Pannonia / Die THOMAS. E. 1980, Vil a settlement. – In: A. Lengyel, autonomen Städte in Noricum und Pannonien. G. T. B. Radon (eds.), The Archaeology of Roman Pannonia II. – Situla 42. Pannonia, Budapest, 275–321. ŠILJEG, B. 2008, Arheološko istraživanje lokaliteta Mala TODOROVSKI, S. 2010 = ТОДОРОВСКИ, С., Т.н. luka i Baška na Krku. – Annales Instituti Archaeo­ “микростанбена целина врз театарот” во logici 4, 84–87. античкиот град Хераклеја Линкестис (The so- ŠKEGRO, A. 2008, The Diocese of Muccurum ( Eccle­ called micro dwelling unit at the ancient town of sia Muccuritana) in Dalmatia (Mukurska škofija Heraclea Lynkestis). – Patrimonium.MK 3/7–8, ( Ecclesia Muccuritana) v Dalmaciji). – Arheološki 111–118. vestnik 59, 291–303. TOMIČIĆ, Ž. 1988, Novija ranosrednjovjekovna ŠPEHAR, P. 2008, Late Antique and Early Byzantine istra živanja odjela za arheologiju. – Prilozi 3/4 fortifications in Bosnia and Herzegovina (hinter- (1986–1987), 141–173. land of the province of Dalmatia). – In: Steuer et TOMIČIĆ, Ž. 1990, Materijalni tragovi ranobizantskog al. (eds.) 2008, 559–594. vojnog graditeljstva u Velebitskom podgorju. – ŠPEHAR, P. 2010, Materijalna kultura iz ranovizantijskih Vjesnik arheološkog muzeja Zagreb, 3.s. 23, 135–162. utvrđenja u Đerdapu ( Material Culture from Early TOMIČIĆ, Ž. 1996, Auf der Spur der Reconquista Byzantine Fortresses in Djerdap). – Cahiers des Iustiniana: spätantike Befestigungsanlagen an der Portes de Fer 7. Nordküste Kroatiens. – Prilozi Odjela za arheologiju 10, 103–116. 416 BIBLIOGRAPHY TOMKA, P. 2015, Eine römische Stadt und ihre barba- (online publication; last access: 20th April 2023) rische Peripherie: Scrabantia. – In: T. Vida (ed.), [https://be-ja.org/index.php/journal/article/view/ Romania Gothica II. The Frontier World: Romans, be-ja-8-1-2018-17-65] Barbarians and Military Culture, Proceedings of the TORBATOV, S. 2021 = ТОРБАТОВ, С., Непознатият International Conference held at the Eötvös Loránd Римски лимес в България. Част 1. Между реките University (Budapest, 1-2 October 2010), 587–616. Вит и Осъм / The Unknown Roman Limes in TOMOVIĆ, M. 2000, Kraku Lu Jordan and Gold Min- Bulgaria. Part 1 . Between the Vit and the Osam ing and Metal urgy in Antiquity. – Starinar 50, Rivers. – Veliko Tarnovo. 155– 185. TOŠIĆ, G., D. RAŠKOVIĆ 2007 = ТОШИЋ, Г., Д. TOMOVIĆ, M., Č. VASIĆ 1997, Vrelo­Šarkamen, Carska РАШКОВИЋ, Ранохришчански споменици на palata. Zlatni nakit iz mauzoleja u Šarkamenu / источним падинама Копаоника (Early Chris- Vrelo­Šarkamen, The imperial palace. The golden tian Monuments on the Eastern Slopes of the jewel ery from the mausoleum at Šarkamen. – Beo- Kopaonik). − Зборник радова Византолoшког grad, Negotin. института ( Zbornik radova Vizantološkog insti­ TOPALILOV, I. 2012, Philippopolis. The city from the tuta), 44/1, 27−47. 1st to the beginning of the 7th c. – In: Ivanov (ed.) TÓTH, E. 1974, Vigilius episcopus Scaravaciensis. – Acta 2012, 2– 76. Archaeologiae Hungaricae 26, 269−275. TORBATOV, S. 1999, A note of Dinogetia. – In: G. von TÓTH, E. 1987, Bemerkungen zur Kontinuität der Bülow, A. Dimitrova-Milčeva (eds.), Der Limes an römischen Provinzialbevölkerung in Transdanu- der unteren Donau von Diokletian bis Heraklios, bien (Nordpannonien). − In: B. Hänsel (ed.), Die Vorträge der internationalen Konferenz Svištov Völker Südosteuropas im 6. bis 8. Jahrhundert, (1.-5. September 1998). – Sofia, 271–280. Südosteuropa-Jahrbuch 17, 251−264. TORBATOV, S. 2000 = ТОРБАТОВ, С., Касно­ TÓTH, E. 2000, Römische Binnenkastelle in den античният град Залдапа ( The Late Antique City Pro vinzen Pannonia Prima und Valeria. − In: of Zaldapa). – Sofia. Von Augustus bis Attila. Leben am Ungarischen TORBATOV, S. 2002a = ТОРБАТОВ, С., Укрепи­ Donau limes, Die Limesreihe - Schriften des Limes- телната система на провинция Скития (края museum Aalen, 33−36. на III – VII в.) ( The Defence System of the Late TÓTH, E. 2009, Die Spätrömische Militärarchitektur Roman Province of Scythia (the end of the 3rd – the in Transdanubien. − Archaeologiae Értesítő 134, 7th century A.D.) – Veliko Tarnovo. 31−61. TORBATOV, S. 2002b = ТОРБАТОВ, С., Монетната TÓTH, E. 2011, Die spätrömische palastanlage von Sa- циркулация в градището край Одърци (310 г. varia. – In: Bülow, Zabehlicky (eds.) 2011, 275−284. пр. Хр. – 610 г. сл. Хр.) ( Coin Circulation in the TRUHELKA, Ć. 1893, Zenica und Stolac. – Wissen­ Hil ­Top Settlement near Odartsi (310 B.C. – A.D. schaftliche Mitteilungen aus Bosnien und der Her610)). − Veliko Tarnovo. zegowina 1, 273−302. TORBATOV, S. 2015, The Transformations of a Roman TSONIOTIS, N. 2016, The Benizeli Mansion Excavation. Military Site: The Case with Mezdra (North-West- Latest Evidence on the Post-Herulian Fortification ern Bulgaria). – In: L. Vagalinski, N. Sharankov Wall in Athens. – In: Frederiksen et al. (eds.) 2016, (eds.), Limes XXII. Proceedings of the 22nd In­ 712−724. ternational Congress of Roman Frontier Studies TUFI, S. R. 2012, Doclea, citta romana del Montenegro. (Ruse, Bulgaria, September 2012) (= Bulletin of − In: BAR, International Series 2419, 477−489. the National Archaeological Institute XLII, 2015), TURKOVIĆ, T. 2011, The Late Antique “Palace” in Sofia, 151−161. Polače Bay (Mlijet) – Tetrarchic “Palace”? − Hortus TORBATOV, S. 2016 = ТОРБАТОВ, С., Anasamus/ Artium Medievalium, 211–233. Ansamus/Ἀσημοῦς/ Ἀσήμος (античните селищни TURKOVIĆ, T., N. MARAKOVIĆ 2012, Transforma- и фортификационни останки край устието на tion of Rural Landscape in the Southern Part р. Осъм) (Anasamus/Ansamus/Ἀσημοῦς/ Ἀσήμος of the Province of Dalmatia: Polače, Ostrvica, (the Roman, Late Roman and Early Byzantine set- Mogorjelo. – In: P. Diarte Blasco, A. Martín López tlement and fortification remains near the Osam (eds.), Arqueología de un Paisaje en Transición. River’s mouth)). – Bulgarian e­Journal of Archae­ Antigüedad Tardía y Alta Edad Media, 61–63. ology 6/1, 21−79. (Online publication; last access: TURKOVIĆ, T., M. ZEMAN, 2011, Prilog poznavanju 17th April 2023) [https://be-ja.org/index.php/ kasnoantičkih vila s područja provincije Dal- journal/article/view/be-ja-6-1-2016-21-79/122] macije - slučaj vile u Strupniću kraj Livna. – Ars TORBATOV, S. 2018 = ТОРБАТОВ, С., Градището Adriatica 1, 9–26. край Рибен (The Gradishteto hill site near Riben. − UBL, H. 1982, Die archäologische Erforschung der Bulgarian e­Journal of Archaeology 8/1, 17−65. Severin sorte und das Ende der Römerzeit im 417 Slavko CIGLENEČKI Donau-Alpen-Raum. – In: Severin. Z wischen of Thrace (4th-7th centuries)). – In: S. Stanev, V. Römerzeit und Völkerwanderung, Linz, 71–97. Grigorov, V. Dimitrov (eds.), Studies in honour of UBL, H. 2002, Lauriacum. Die zivilen Siedlungsräume. – Stephan Boyadzhiev, Sofia, 17–56. In: Šašel Kos, Scherrer (eds.) 2002, 257–276. VAN OSSEL, P. 1992, Établissements ruraux de l’Antiquité UBL, H. 2011, Die Legionslager und Hilfstruppenkas- tardive dans le Nord de la Gaule. – Gallia , Suppl. 51. tel e von Noricum seit dem 2. Jahrhundert bis zum VAN OSSEL, P., P. OUZOULIAS 2000, Rural settlement Abzug der Romanen aus Noricum Ripense und ihr economy in Northern Gaul in the Late Empire: Wiedererstehen als Städte des frühen Mittelalters. – an overview and assessment. – Journal of Roman In: Konrad, Witschel (eds.) 2011, 425–460. Archaeology 13, 133–160. UENZE, S. 1992, Die spätantiken Befestigungen von VANNACCI LUNAZZI, G. 2001, L’ insediamento di Sado vec (Bulgarien). Ergebnisse der deutsch­bulga­ Colle Mazéit a Verzegnis. – In: S. Vitri, F. Oriolo risch­österreichischen Ausgrabungen 1934−1937. – (eds.), I Celti in Carnia e nel ’ arco alpino centro Münchner Beiträge zur Vor- und Frühgeschichte 43. orientale, Atti del Convegno di Studi (Tolmezzo, UGLEŠIĆ, A. 2002, Ranokršćanska arhitektura na 30 aprile 1999), Trieste, 168–171. području današnje zadarske nadbiskupije. – Zadar. VASIĆ, M. 1970, Römische Villen vom Typus der Vil a ULBERT, T. 1981, Das spätrömische Kastell AD PIRVM rustica auf jugoslawischen Boden. – Archaeologia – Hrušica. – In: Ad Pirum (Hrušica). Spätrömische Iugoslavica 11, 46–81. Passbefestigung in den Julischen Alpen, Münchner VASIĆ, M. 1984 = Bасић, M., Чезава – Castrum No- Beiträge zur Vor- und Frühgeschichte 31, 3–50. vae (Čezava – Castrum Novae). – Starinar 33–34 ULBERT, T. 1993, Resafa - Sergiupolis. Archäologische (1982–1983), 91–122. Forschungen in der nordsyrischen Pilgerstadt. – VASIĆ, M. 1995, Le limes protobyzantin dans la In: E. M. Ruprechtsberger (ed.), Syrien. Von den province de Mésie Premiére. – Starinar 45–46 Aposteln zu den Kalifen, Linzer archäologische (1994–1995), 41–53. Forschungen 21, 112–127. VASIĆ, M., V. KONDIĆ 1986, Le limés romain et paleo- URANKAR, R. (ed.) 2021, Kranj, Kieselstein – severno byzantin des Portes de Fer. – In: Studien zu den Milidvorišče. Raziskave 2008–2009. – Monografije tärgrenzen Roms 3, Vorträge des 13 Internationalen CPA 13. Limeskongresses (Stuttgart 1983), 542–560. UREK, M., A. KOVAČIČ 2020, Fluvio Frigido, Castra – VEIKOU, M. 2012, Byzantine Histories, Settlement Ajdovščina. Raziskave / Investigations 2017−2019. Stories: Kastra, “Isles of Refuge”, and “Unspeci- – In: Horvat et al. (eds.) 2020a, 47–59. fied Settlements” as In-between or Third Spaces. URODA, N. 2013, Beginnings of Monasticism on Cen- Preliminary Remarks on Aspects of Byzantine tral Dalmatian Islands - Problems and perspec- Settlement in Greece (6th–10th c.). – In: T. Kious- tives. − Hortus Artium Medievalium 19, 113−120. sopoulou (ed.), The Byzantine Cities, 8th– 15th USLAR, R. von 1964, Studien zu frühgeschichtlichen Bef­ Centuries, Rethymno, 159–206. estigungen zwischen Nordsee und Alpen. – Bonner VEIKOU, M. 2013, Settlements in the Greek Coun- Jahrbücher 11. tryside from the 4th to 9th Centuries: Forms and UYTTERHOEVEN, I. 2007, Housing in Late Antiquity. Patterns. – Antichité Tardive 21, 125–133. Regional Perspectives. – In: Lavan et al. (eds.) VELIMIROVIĆ-ŽIŽIĆ, O. 1986, Ostaci fortifikacione 2007, 67–93. arhitekture na gradini Đuteza u Dinošima kod VAGALINSKI, L. 2008, Barbarian presences at Colonia Titograda. – Materijali 22, 80–87. Flavia Pacis Deultensiam (Southeastern Bulgaria). – VELKOV, V. 1977, Cities in Thrace and Dacia in Late In: B. Niezabitowska-Wiśniewska, M. Juścinski, Antiquity: (Studies and Materials). – Amsterdam. P. Łuczkiewicz, S. Sadowski (eds.), The turbulent (First edition 19671). epoch. New materials from the Late Roman Period VELKOV, V. 1988, Die Bedeutung von Nesebar in der and the Migration Period. 1 (= Monumenta Studia Übergangsperiode von der Antike zum Mittel- Gothica V), Lublin, 345–353. alter. – In: V. Velkov, Geschichte und Kultur Thrak­ VAGALINSKI, L. 2018, Chronology of the Late Antique iens und Moesiens, Amsterdam, 223–226. Fortifications of Deultum (Archaeological Data). – VELKOV, V., Z. GOČEVA 1972, Die thrakische Festung In: Proceedings of the First International Roman and Čertigrad im Balkan. – Thracia 1, 121–145. Late Antique Thrace Conference: Cities, Territories VETTERS, H. 1950, Ein befestigtes Dorf am Vit. – In: and Identities, Plovdiv, 3rd—7th October 2016 Dacia Ripensis, Österreichische Akademie der (Bulletin of the National Archaeological Institute, Wissenschaft, Schrifen der Balkan-Kommission, 44) 85–91. Antiquarische Abteilung 11, 1, 49–57. VALEVA, J. 2011 = ВЪЛЕВА, Ю, Елитна жилищна VIDA, T. 2007, Late Roman territorial organisation and архитектура в диозеца Тракия (ІV-VІІ в.) (Elite the Settlement of the Barbarian Gentes in Panno- house architecture and decoration in the diocese nia. – Hortus Artium Medievalium 13/2, 319–331. 418 BIBLIOGRAPHY VIDA, T. 2011, Die Zeit zwischen dem 4. und dem 6. al medioevo: spunti di riflessione. – Hortus Artium Jahhundert im mittleren Donauraum aus archäolo- Medievalium 12, 101–122. gischer Sicht. – In: Konrad, Witschel (eds.) 2011, VODIČKA MIHOLJEK, K. 2015, Arheološki lokalitet 615–648. Poje – antička vila u Njivicama na otoku Krku. VIDRIH PERKO, V. 1997, Rimskodobna keramika – In: Istraživanja na otocima. Znanstveni skup z Ajdovščine pri Rodiku (Roman Pottery from (Lošinj, 1.-4. listopada 2012), Izdanja Hrvatskog Ajdovščina near Rodik). – Arheološki vestnik 48, arheološkog društva 30, 57–66. 341–358. VOGRIN, A. 1991, Arheološko najdišče Kreuh. – In: VIDRIH PERKO, V., M. SAGADIN 2004, Gorenjska v Celeia antiqua. Exhibition catalogue, Celje, 17–20. antiki. – Kamniški zbornik 17, 207–224. VÖLLING et al. 2019, = Völling, Th., H. Baitinger, S. VIDRIH PERKO, V., M. ŽUPANČIČ 2003, Il popola- Ladstätter, A. Rettner, M. A. Miller 2019, Olym­ mento del a Slovenia sudoccidentale e del ’Istria pia in frühbyzantinischer Zeit. Siedlung – Land­ settentrionale nel periodo tardoromano e nel ’alto wirtschaftliches Gerät – Grabfunde – Spolienmauer. medio evo al a luce delle ceramiche di importazi- – Olympische Forschungen 34. one. – Histria antiqua 11, 457–476. VUJINOVIĆ, B. 2013 = ВУЈИНОВИЋ, Б., Накит VIKIĆ-BELANČIĆ, B. 1978, Elementi ranog kršćanstva рановизантијског града Балкис (Jewelry from the u sjevernoj Hrvatskoj (Les elements du chris- early byzantine city Balkis). – Gradiška. tianisme primitif en Croatie septentrionale). – VUJINOVIĆ, B. 2014, Ranovizantijski grad na Gradini Arheološki vestnik 29, 588–605. u Bakincima (Early Byzantine town at Gradina VILLA, L. 2001, Nuovi dati archeologici sui centri for- in Bakinci). – Godišnjak. Centar za balkanološka tificati tardoantichi – altomedievali del Friuli. – In: ispitivanja ANUBiH 43, 167–184. Paolo Diacono 2001, 825–861. VULIĆ, H. 2016, Eine frühchristliche Anlage in Ka- VILLA, L. 2002, Iulium Carnicum e Iulia Con- menica bei Cibalae/Vinkovci: Vorbericht zu den cordia: Il destino di due centri urbani minori archäologischen Untersuchungen in den Jahren nel ’altomedioevo. – Aquileia Nostra 73, 342–443. 2012 bis 2015. – In: GrenzÜbergänge. Spätrömisch, VILLA, L. 2003, San Giorgio (Attimis - Ud.). Campagna frühchristlich, frühbyzantinisch als Kategorien der di Scavo 2003. – Quaderni Friuliani di Archeologia historisch­archäologischen Forschung an der mit­ 13, 297–309. tleren Donau, Forschungen zu Spätantike und VILLA L. 2004, Aquileia tra Goti, Bizantini e Longobardi: Mittelalter 4, 133–144. spunti per un’analisi delle trasformazioni urbane VULIĆ et al. 2017 = Vulić, H., D. Doračić, R. Hobbs, J. nel a transizione fra Tarda Antichità e Alto Medio- Lang 2017, The Vinkovci treasure of Late Roman evo. – In: Antichità Altoadriatiche 59, 561–632. silver plate: preliminary report. – Journal of Roman VILLA, L. 2006, Le tracce della presenza Gota nell’ Archaeology 30, 127–150. Italia nord-orientale e il caso dell insediamento di WAELKENS, M. 2005, The Fifth to Tenth Century S. Giorgio di Attimis (UD). – In: M. Buora, L. Vil a AD In Southwest Anatolia – Transformation or (eds.), Goti nel ’arco alpino orientale, Archeologia Decline: A Case Study – Sagalassos and its Chora. di frontiera 5, 147–173. – In: Byzantine Symposium, Urban and Rural Set­ VILLA, L. 2020, Questioni di cronologia e di interpre- tlement in Anatolia and the Levant, 500­1000 AD: tazione storica circa lo sviluppo degli insediamenti New Evidence from Archaeology (Vienna, Austria fortificati in Italia settentrionale tra la tarda an- 2005), Symposium Abstracts. tichita e l’altomdioevo (V–VI secolo). – In: Cavada, WAELKENS, M. 2019, Sagalassos, Archaeology of. – In: Zagermann (eds.) 2020a, 23–57. Encyclopedia of Global Archaeology, 1–32. Springer, Villes 1984 = Vil es et peuplement dans l’ Il yricum Cham [https://doi.org/10.1007/978-3-319-51726- protobyzantin. – Actes du colloque organisé par 1_1121-3] l’École Française de Rome (Rome, 12-14 mai 1982), WALDE, E. 2002, Aguntum. – In: Šašel Kos, Scherrer Publications de l’École française de Rom e 77, Rome. (eds.) 2002, 149–163. VINSKI, Z. 1980, Ovrednotenje grobnih pridatkov. – In: WARD-PERKINS, B. 2000, Land, labour and settlement. Stare 1980, 17–32. – In: A. Cameron, B. Ward-Perkins, M. Whitby VISY, Z. 2003, The Ripa Pannonica in Hungary. – Bu- (eds.), The Cambridge Ancient History XIV. Late dapest. Antiquity. Empire and Successors, A.D. 425– 600, VISY, Z. 2018, A pannoniai késő római “belső erődök” 315–345 (3rd Edition, 2006). (Late Roman ‘inner forts’ of Pannonia) . – Archaeo­ WARNER SLANE, K., G. D. R. SANDERS 2005, Corinth: logiai Értesitö 143, 233–246. Late Roman Horizons. – Hesperia 74, 243–297. VITRI et al. 2006 = Vitri, S., L. Vil a, A. Borzacconi 2006, WERNER, J. 1992, Golemanovo Kale und Sadovsko Transformazioni urbani a Cividale dal tardoantico Kale. Kritische Zusammenfassung der Grabungs- ergebnisse. – In: Uenze 1992, 391–417. 419 Slavko CIGLENEČKI WHITBY, M. 2000, The Balkans and the Grece, 420–602. karischen Tälern (Trentino, Italien). – Bayerische – In: A. Cameron, B. Ward-Perkins, M. Whitby Vorgeschichtsblätter 79, 195–218. (eds.), The Cambridge ancient history XIV, Late ZANINI, E. 2009, Città, microterritorio e macroterri- Antiquity: Empire and Successors, A.D. 425–600, torio (e mobilità degli uomini) nel Mediterraneo 701–730 (3rd Edition, 2006). proto-bizantino: il caso di Gortina di Creta. – In: WHITTOW, M. 2007, Nicopolis ad Istrum: Backward G. Macchi (ed . ), Geografie del popolamento: casi di and Balkan? – In: Poulter (ed.) 2007a, 375–389. studio, metodi, teorie, Siena, 111–122. WICKHAM, C. 2005, Framing the Early Middle Age. ZANINI, E. 2016, Coming to the End: Early Byzantine Europe and the Mediterranean, 400–800. – Oxford. Cities after the mid-6th Century. − In: Proceedings WILKES, J. J. 1993, Diocletian’s Palace, Split: residence of the 23rd International Congress of Byzantine of a retired Roman emperor. – Sheffield 1986 (re- Studies (Belgrade, 22–27 August 2016), Belgrade, printed Oxford 1993). 127−140. WILKES, J. 2005, Provinces and frontiers. – In: A.K. ZANINOVIĆ, M. 1978, Gradina u Danilu i Tor nad Bowman, P. Garnsey, A. Cameron, The Cambridge Jelsom. Dva gradinska naselja u srednjoj Dalmaciji. Ancient History XII, The Crisis of Empire, A.D. – Materijali 15, 19–23. 193–337, 212–268 (second Edition). ZANINOVIĆ, M. 2002, Mogorjelo od vile do kastruma. WILKES, J. 2013, The Archaeology of War: Homeland – In: Godišnjak XXXII. Centar za balkanološka Security in the South-West Balkans (3rd–6th c. ispitivanja ANUBiH 30, 447–456. A.D.). − In: Sarantis, Christie (eds.) 2013, 735−757. ZORNIJA, M. 2016, A sign of faith in the open sea: the WINCKLER, K. 2012, Die Alpen im Frühmittelalter. early christian Church of St Mary on the Island of Die Geschichte eines Raumes in den Jahren 500 bis Kornat (Croatia). – Antiquité Tardive 25, 52–67. 800. – Wien, Köln, Weimar. ZOTOVIĆ, L. 1984, Boljetin (Smorna), rimski i rano- WISEMAN, J. R. 1984, The City in Macedonia Secunda. vizantijski logor. (Boljetin (Smorna), camp romain – In: Villes 1984, 289–314. et paléobyzantin). – Starinar 33–34 (1982–1983), WOLFF, H. 1991, Die Kontinuität städtischen Lebens 211–225. in den nördlichen Grenzprovinzen des römischen ZSIDI, P. 2004, Aquincum. Ergebnisse der topographi- Reiches und das Ende der Antike. – In: W. Eck, H. schen und siedlunghistorischen Forschungen in Galsterer (eds.), Die Stadt in Oberitalien und in den den Jahren 1969–1999. – In: Šašel Kos, Scherrer nordwestlichen Provinzen des Römischen Reiches, (eds.) 2004, 209–230. Kölner Forschungen 4, 287–318. ZSIDI, P. 2011, Vom spätantiken Aquincum zum mittel- WOOD, J. 2002, The wal top of the Late-Roman defences alterlichen Vetus Buda. – In: Konrad, Witschel at Saint-Bertrand-de-Cominges: interim report. – (eds.) 2011, 549–570. Journal of Roman Archaeology 15, 297–309. ŽELEZNIKAR, J., J. VISOČNIK 2020, Atrans – Trojane. ZABEHLICKY, H. 2011, Die Vil a von Bruckneudorf – In: Horvat et al. (eds.) 2020a, 249–294. – Palast oder Grossvil a? – In: Bülow, Zabehlicky ŽERJAL, T., M. NOVŠAK 2020, Školarice pri Spodnjih (eds.) 2011, 89–99. Škofijah. – Arheologija na avtocestah Slovenije 88. ZAGARČANIN, M. 2014, Nova istraživanja rimske i (Online publication; last access: 23rd March 2023) ranovizantijske vile na lokalitetu Mirišta u Petrovcu / [https://www.zvkds.si/sl/knjiznica/skolarice-pri- New reasearching of Roman and Early Byzantine spodnjih-skofijah] vil a on the site Mirišta in Petrovac. − In: Nova ŽERJAL, T., V. TRATNIK 2020, Fluvio Frigido, Castra – antička Duklja 5, Podgorica, 19−58. Ajdovščina. – In: Horvat et al. (eds.) 2020a, 9–46. ZAGERMANN, M., E. CAVADA 2014, … in locis ŽUPANČIČ et al. 2007 = Župančič, M., A. Trenz, D. firmissimis … Ausgrabungen auf dem Monte San Tomšič 2007, Koper. – Varstvo spomenikov. Poročila Martino di Lundo/Lomaso in den äußeren judi- 43, 88–91. 420 INDEX The digits in bold indicate pages on which figures appear, underlined digits indicate pages on which maps appear. Site Text on page, Fig. on page, Map on page Abritus (Razgrad) 23, 131, 131, 151, 340 Abu Mina 381 Acrae (Cape Kaliakra) 23, 134, 134 Aelium Cetium (St. Pölten) 22, 27 Aguntum (Lienz) 22, 24, 27, 28, 28, 105, 107, 150, 156, 385 Ai Giannis, Adriani 266/267, 309, 325 Ain Labakha 376, 378 Ajdna above Potoki 201, 219, 219, 220, 316, 319, 325, 325, 347 Ajdovščina (Castra) 201, 222, 223, 329, 339 Ajdovščina above Rodik 15, 201, 225, 225, 317, 319, 325, 328, 343, 385 Ajdovski gradec above Vranje 201, 232, 233, 233, 317, 318, 319, 322, 325, 327, 343, 347, 348, 349, 374, 386 Akkale 378, 379 Akra Sophia 196, 199, 199, 347 Alahan 380, 381 Albi 355 Alsóhetény 22, 107, 109, 109 Amida (Diyarbakir) 368, 368, 374, 387 Amphipolis 23, 92, 92, 158 Anasamus or Asemus 23, 127, 127 Ančnikovo gradišče near Jurišna vas 201, 226, 226, 227, 317, 329, 335, 385 Andautonia 22, 151, 156, 240, 341 Androna (Al-Anderin) 372, 373 Antioch 368, 375, 379 Apamea 365, 366 Aquae Iasae (Varaždinske Toplice) 22, 36, 36 Aquileia 22, 42, 43, 44, 44, 45, 70, 105, 149, 150, 153, 155, 158, 172, 224, 226, 339, 340, 341, 345, 347, 383, 386 Aquincum 22, 31, 31, 149, 171, 172, 385 Argyruntum (Starigrad na moru) 22, 37, 150 Arif 370, 371, 387 Arles 353 Armira near Ivailovgrad 182, 184, 185 Artegna 201, 205, 206, 206, 316 Arykanda in Lycia 371 Asseria (Podgrađe) 22, 52, 53, 53, 151, 153, 153, 346, 347 Astorga 352 Athens 12, 19, 23, 99, 99, 100, 101, 154, 158, 158, 159, 338, 340, 342, 384 421 Slavko CIGLENEČKI Site Text on page, Fig. on page, Map on page Auxerre 355 Balácapuszta 168, 171, 172, 384 Balajnac near Niš 266/267, 276, 276, 325, 332, 349 Banjače 191 Barbariga 168, 175, 347 Bargala (Kozjak) 23, 144, 144, 152, 153, 154, 160, 161, 329, 340, 342, 346 Bassianae (Sremski Petrovci) 23, 38, 38, 157 Bedem near Maskare 64, 266/267, 270, 270, 349 Bela voda near Pernik 196, 197 Belgrad near Dvorište 266/267, 306, 306, 332, 349 Bihovo near Trebinje 168, 181, 181 Bisko near Knin 168, 177, 314, 347 Blagaj 22, 118, 118, 119, 151, 164, 195, 384 Blagovica 168, 173, 173 Boljetin 265, 265, 266/267, 330, 339, 349 Borgo Regio in Desenzano 168, 169 Borovets near Pravets 266/267, 281, 329, 340 Bosar near Baška 22, 114, 115, 115, 151, 164, 164, 243 Bosman 264, 266/267, 267, 267, 330, 349 Bosra 366, 366 Brinjeva gora 317, 319 Brixia (Brescia) 22, 24, 40, 41, 149, 153, 159, 341, 344, 383, 385 Brodac near Bijeljina 188, 194 Burgbichl near Irschen 316 Bushati 266/267, 295, 296, 329, 385 Buthroton (Butrint) 19, 23, 78, 80, 81, 154, 186, 342, 345 Byllis 19, 23, 78, 78, 79, 154, 346, 346 Cabyle near Yambol 23, 136, 137, 137 Capris / insula Capritana (Koper) 22, 112, 112, 113, 165 Carassura 266/267, 293, 294, 328 Carevec (Veliko Tarnovo), presumed 23, 70, 129, 129, 130, 131, 151, 154, 286, 345, 345 Zikideva Caričin grad /Justiniana Prima/ 19, 23, 76, 104, 122, 123, 124, 125, 126, 126, 152, 153, 154, 161, 162, 163, 265, 276, 277, 320, 320, 323, 328, 346, 384, 386, 387 Carnium (Kranj) 22, 111, 111, 112, 163, 163, 164, 201, 317, 319, 328, 347, 384, 386 Carnuntum 22, 27, 149, 171, 385 Carthage 160, 355 Castel Antico 168, 170, 170, 347 Castelazzo above Doberdo del Lago 201, 207, 208, 317, 329 Castelletto di Brenzone 168, 169, 169, 347 Castelraimondo above Forgaria 201, 206, 207, 316, 329, 334, 338, 343 Castra Nicea (Dolenci) 23, 147, 147, 160 Castrum Gradense (Grado) 22, 44, 105, 105, 150, 347, 386 Čebren (Grad) near Zovik 266/267, 309, 309, 328, 344 Čečan 266/267, 279 Celeia (Celje) 22, 32, 33, 34, 34, 151, 157, 173, 227, 229, 230, 340, 341 Centcelles 358 Cerro de Montroy 361, 362 Červar, Porat 168, 175 Čezava 330 422 INDEX Site Text on page, Fig. on page, Map on page Chatalka near Stara Zagora 182, 184 Chertigrad near Brousen 266/267, 284, 284, 324, 334 Cherven breg 182, 183, 314 Cibalae (Vinkovci) 22, 36, 37, 37 Cickini near Sršiči 168, 176 Clunia 351 Colle di Zuca 205 Colle Santino above Invillino 201, 204, 205, 206, 316, 319, 322, 325, 338, 343 Concordia 22, 24 Conimbriga 352, 352 Constantine in Provence 361 Constantinople (İstambul) 9, 10, 11, 23, 62, 103, 127, 149, 152, 165, 337, 341, 342, 347, 349, 351, 352, 381, 383, 386, 387, 388 Corinth 19, 23, 100, 101, 103, 158, 186, 199, 340, 342 Crkvina near Halapić 118, 168, 180, 322 Crkvina near Otres 168, 177 Crkvišće Bukovlje (Generalski Stol) 201, 240, 241, 241, 317, 319, 329 Črnomelj 317, 328 Csúcshegy 168, 171 Cuel Budin 205, 316 Čuker near Mokro polje 201, 252, 252, 332, 333 Çuka e Ajtojt 266/267, 298, 299 Dara 368 Davidovac – Gradište 196, 196 Davina, Kula, Čučer 21, 23, 139, 139, 140, 161, 162, 163 Debelo Brdo in Sarajevo (fort) 201, 260, 260, 325 Debelo Brdo in Sarajevo (vil a) 168, 179 Demetrias 23, 24, 38, 39, 150, 345, 385 Deultum (Debelt) 23, 71, 71, 158, 341, 345 Diaporit 182, 186, 186, 347 Dichin 266/267, 281, 284, 285, 286, 320, 325, 327, 336, 344, 345, 349 Dinogetia 23, 138, 138, 152, 340, 384 Diocletianopolis (Argos Orestiko) 23, 38, 38, 147, 150, 344, 346 Dion 23, 92, 94, 158, 342, 385 Djemila (Cuicul) 355 Dobri Dyal 266/267, 286, 286, 327, 334, 340, 343, 344 Doclea (Podgorica) 23, 63, 63 Domaj 266/267, 296, 296, 349 Donja Glavnica 168, 174 Donje Nerodimlje 196, 197, 315 Doss Trento 42, 42, 150, 151, 316, 336, 345 Dougga 352, 355 Dračeva strana near Bileća 168, 181 Dragonera Jug 168, 175, 175, 314, 347, 384 Drama 266/267, 309 Drenje near Zaprešič 168, 174 Duel 47, 201, 210, 210, 211, 212, 317, 318, 332, 333, 348 Đurđevica in Đerekare 266/267, 275, 276, 333 Durostorum (Silistra) 23, 44, 70, 158 Đuteza near Dinoša 266/267, 278, 278, 327, 349 423 Slavko CIGLENEČKI Site Text on page, Fig. on page, Map on page Dyadovo 266/267, 293, 293, 331, 333, 335, 349 Dyrrachium (Durrës) 23, 76, 76, 77, 151, 158, 344, 345 El Bara 372 El Monastil 361, 362 Elenska Basilica near Pirdop 266/267, 291, 291, 335 Emona (Ljubljana) 22, 32, 32, 33, 33, 151, 151, 157, 157, 172, 173, 205, 220, 224, 225, 226, 238, 341, 385 Ephesus 363, 363, 364, 368, 381 Favianis (Mautern) 22, 26, 26, 27, 149, 338, 341, 345 Flavia Solva (Leibnitz) 22, 29, 29, 156, 340, 341 Forum Iulii (Cividale) 22, 45, 45, 153, 205, 206, 216, 220, 345 Frangoklisia 182, 186 Frauenberg 29, 317 Gaj in Babrež 266/267, 274, 274, 332, 349, Gata 191 Gegje (Gjakovë/Đakovica ) 266/267, 280, 280, 329, 339 Gemona 205, 316 Georgenberg near Kuchl 201, 208, 209, 316, 319, 329 Gerasa 366, 367 Golem Grad, Konjsko 266/267, 306, 334 Golemanovo Kale 164, 266/267, 281, 283, 283, 319, 320, 325, 327, 336, 342, 349, 386 Golemo Gradište near Konjuh 23, 142, 143, 154, 344 Gornji Streoc 266/267, 279 Gorsium 22, 107, 109, 109, 110, 151, 338, 339 Gortyn (Gortyna) 102, 103, 151, 156, 158, 340, 347 Grad near Gornji Vrbljani 201, 255, 255, 332, Gradac above Potočani 201, 256, 256, 257, 332 Gradac in Lepenica (Homolj) 201, 259, 259 Gradac near Ošlje 249, 331, 331 Gradac near Todorovići 201, 264, 264 Gradac on Ilinjača 201, 259, 259, 325, 335 Gradec near Logje 334 Gradec near Prapretno 15, 201, 229, 229, 317, 319, 325, 336 Gradina (Vrgada) 201, 249, 249, 250, 331 Gradina (Žirje) 201, 249, 250, 251, 319, 319, 330, 331, 348, 349, 388 Gradina above Kovaći 201, 258, 332 Gradina in Bakinci 201, 254, 254, 255, 328, 348, 386 Gradina in Bastasi 256, 258, 258 Gradina in Biograci near Lištica 201, 260, 260, 261, 332/333, 334 Gradina in Podgradina Kamenska 201, 258, 259, 322, 335 Gradina in Radalica 266/267, 274, 275, 333 Gradina in Vašarovine 201, 256, 257, 258 Gradina in Vidoši 201, 256, 257, 257, 335 Gradina in Vrsenice 266/267, 271, 272, 326, 329, 343, 349 Gradina near Modrić 201, 247, 247, 248, 331, 348 Gradina on Jelica 21, 23, 119, 120, 163, 386, 387 Gradina Ramoševo 266/267, 273, 273, 335 Gradina Ras 266/267, 270, 271, 320, 325, 329, 334, 343, 349 Gradina Zecovi near Čarakovo 201, 253, 254, 332, 333 Gradišče near Velike Malence 201, 238, 238, 317, 329 Gradishta of Bardhoc 266/267, 295, 295, 332 424 INDEX Site Text on page, Fig. on page, Map on page Gradishte Gabrovo 266/267, 291, 292, 292, 293, 343, 344, 385 Gradishteto near Debrene 266/267, 287, 288, 334, 335, 344 Gradishteto near Riben 266/267, 284, 327, 329, 344 Gradište (Sobri) near Oraše 266/267, 299, 300, 325, 335 Gradište near Delisinci 266/267, 307, 307, 334 Gradište near Pakoševo 266/267, 302, 303, 329, 335, 344 Gradište near Pčinja 266/267, 303, 304, 327, 335, 344 Gradište Stenče 266/267, 299, 299, 327 Gradok (Markovi Kuli) near Čanište 266/267, 307, 307, 308, 325, 349 Grazerkogel 149, 149, 150, 317 Grobelce near Šmarje pri Jelšah 168, 173, 173 Grušine 168, 177 H. Thomas/Agios Thomas 182, 186 Haïdra (Ammaedara) 352, 352 Harilaq 18, 266/267, 279, 279, 327 Harmana 266/267, 294, 294, 329, 343 Hemmaberg above Globasnitz 119, 150, 201, 213, 214, 214, 317, 318, 327, 327, 343, 347 Heracleia (Perinthus) 23, 103, 151, 157, 158, 341, 346 Heracleia Lyncestis (Bitola) 19, 23, 86, 87, 87, 88, 154, 158, 159 Herculia 109, 109, 110 Hexamilion 312, 312, 343, 349, 386 Hisar above Leskovac 266/267, 277 Hissar (Hisarya) presumably Diocle- 23, 70, 134, 135, 135, 137, 151, 152, 152, 153, 340, 340, 344, 345, 384, 385 tianopolis Histria (city) 23, 73, 73, 154, 158, 346 Histria (land) 11, 24, 32, 42, 187, 338 Höflein 188, 188, 342 Hoischhügel near Maglern 201, 211, 211, 317, 332 Hosszúhetény 188, 189, 189 Hrušica (Ad Pirum) 201, 224, 224, 225, 226, 317, 319, 329, 339, 340, 385 Hum near Tutin 266/267, 273, 273, 335 Iader (Zadar) 22, 52, 52, 53, 149, 151, 153, 383, 386 Iatrus (Krivina) 19, 23, 127, 128, 129, 152, 160, 329, 337, 340, 343, 344, 345, 346, 384 Ilidža in Sarajevo 168, 180 Immurium (Moosham) 168, 171 Isthmia 266/267, 312, 312, 331 Italica 351 Iulium Carnicum (Zuglio) 22, 24, 25, 150, 156, 345, 385 Iuvavum (Salzburg) 22, 25, 150, 208 Ivandvor 168, 174 Izbičanj near Prijepolje 23, 119, 164, 165, 335, 346 Jeme, Medinet Habu 376, 378 Jerinin grad in Trubarevo 348 Južac near Sopoćani 266/267, 272, 273, 334 Kalaja near Grazhdani 23, 138, 138 Kalata near Kamenica 162, 266/267, 305, 305, 328, 334 Kale in Bregovina 266/267, 276, 277, 325, 349 Kale near Debrešte 266/267, 306, 306, 327, 335, 344 Kale near Gorno Svilare 266/267, 308, 308, 332 Kale, Gradište (Vinica) 18, 23, 145, 145, 146, 151, 152, 153, 154, 161, 163, 384, 386 Kale, Zlata 266/267, 277, 277, 328, 349 425 Slavko CIGLENEČKI Site Text on page, Fig. on page, Map on page Kale’i Zerzevan 374 Kaleto (Castra Rubra) 266/267, 292, 293, 331, 349 Kappele near Jadersdorf 201, 209, 209, 316, 325 , 343, 347, 349 Karakabakli 376, 377 Karanis in the Fayum 376 Kartal kale near Ruyno 266/267, 286, 287, 333, 334, 336, 344 Kastritsa near Ioánnina 266/267, 310, 310, 311, 325 Kastro Rizovouni 266/267, 311, 311, 335 Kékkút 188, 190 Kekola near Keqekol a 266/267, 280, 327, 349 Kellia 381 Kephalos 266/267, 312 Keszthely-Fenékpuszta 22, 107, 107, 108, 109, 152, 341 Kharga Oasis 376, 378 Khortokopi 266/267, 310 Kirchbichl near Lavant 22, 24, 27, 105, 106, 107, 150, 165, 316, 333, 339 Knal wand near Ramsau 201, 208, 209, 317 Komló-Mecsekjánosi 188, 189 Korinjski hrib above Veliki Korinj 15, 201, 234, 234, 235, 235, 317, 319, 334, 347, 348, 349 Korintija near Baška 114, 115, 201, 243, 243, 244, 331, 348 Korishë/Koriša 266/267, 280, 281, 325 Környe 22, 107, 109, 339 Kostinbrod 196, 197, 339 Koštur near Dabrica 201, 263, 263, 264, 331 Kövágószőlős 168, 172 Kraku Lu Jordan 266/267, 267, 268, 339/340 Krasen kale 266/267, 291, 291, 335, 348 Kruče near Ulcinj 196, 197, 315, 347 Kučar above Podzemelj 119, 201, 236, 237, 238, 317, 319, 334, 343 Kula in Kaludra 266/267, 275, 275, 332, 349 Kula near Čelopek 266/267, 303, 303, 332 Kula near Kalauzlija 266/267, 303, 304, 335 Kulina in Rogatac 266/267, 274, 274 Kuzelin near Donja Glavica 201, 240, 240, 317, 319, 329, 334, 339, 343 Lambata 182, 184 Larina 360, 361 Lauriacum (Lorch) 22, 26, 149 Le Roc de Pampelune near Argelliers 359, 360, 361 (dep. Hérault) León 351, 352 Lisičići near Konjic 188, 194, 195, 339 Lisina (Hvar) 22, 116, 165 Liška Ćava 266/267, 269, 269, 325, 349 Lissus (Lezhë) 23, 75, 75, 151, 158, 269, 295 Ljusina near Bosanska Krupa 188, 194 Löffelbach 188, 188, 189, 339 Loig near Salzburg 168, 170 Louloudies near Katerini 23, 39, 119, 121, 147, 147, 148, 150, 150, 161, 165, 263, 266/267, 310, 344, 384 Lovrečina Bay (Brač) 168, 177, 178, 314 Lozenets 291, 329 426 INDEX Site Text on page, Fig. on page, Map on page Lug ins Land 201, 210, 210, 317 Lugo 352 Luxor 374, 374 Lychnidos (Ohrid) 23, 85, 86, 87, 158, 160 Lymisa 359 Madara (fort) 184, 266/267, 286, 287, 288, 331, 332, 335, 348 Madara (vil a) 182, 183, 183, 184, Majsan 201, 249, 252, 253, 253, 328, 331 Mala Kopašnica 182, 182 Mali Mošunj near Vitez 188, 194, 339 Mali Njivč above Novaki 201, 220, 317, 329, 334 Manerba del Garda 168, 169, 384 Marija Gorska near Lobor 18, 201, 239, 239, 240, 317, 319, 328, 334, 348, 386 Markova Mehana 266/267, 289, 289, 290, 330, 330, 331, 348, 349, 386 Markovi kuli (Vodno) 21, 162, 266/267, 301, 302, 325, 349 Markovo Kale near Malčište 266/267, 300, 301, 325, 335, 344 Martinička gradina near Spuž 185, 266/267, 277, 278 Martinščica 168, 176 Mediana 62, 196, 196, 339 Merida 353, 354 Mesembria (Nesebar) 23, 71, 71, 72, 151, 154, 155, 345 Metz 353 Mezdra 266/267, 281, 282, 327, 337, 340, 348, 349 Miletus 368 Miri near Ostrvica 188, 191, 191 Mirište in Petrovac 182, 185, 185, 347, 384 Mirje (Mirine) near Postire (Brač) 188, 191, 192, 315, 347, 348 Mogorjelo 21, 201, 261, 261, 262, 263, 325, 326, 348 Monemvasia 165, 266/267, 313 Montana 182, 183, 183 Monte San Martino di Lundo/Lomaso 201, 203, 203, 316, 319, 325, 332, 347 Montorio 188, 188 Mosel 358 Mt Sinai 380, 381 Muline (Ugljan) 168, 176, 177, 341 Municipium Riditarum (Danilo) 22, 24, 54, 55 Mursa (Osijek) 22, 36 Mušići near Višegrad 168, 180, 180 Nador 358 Naissus (Niš) 23, 39, 62, 157, 196, 276, 295, 342 Nauportus (Vrhnika) 168, 172 Nesactium 22, 51, 51, 52, 151, 341, 347 Nessana 372, 376 Neviodunum 22, 35, 151, 156, 341 Nicaea 366, 367 Nicopolis (Nikopolis) 23, 93, 95, 95, 96, 158, 159, 164 Nicopolis ad Istrum (Nikiup) 19, 23, 68, 68, 69, 131, 151, 164, 182, 183, 342, 345 Nitria 381 Novae (Svištov) 19, 23, 66, 67, 67, 68, 68, 127, 159, 338, Novo Selo Bunje 168, 177, 314, 347 Nuvolento near Pieve 168, 169 427 Slavko CIGLENEČKI Site Text on page, Fig. on page, Map on page Obelija 196, 197, 198, 339 Odartsi 266/267, 288, 289, 326, 329, 340, 344, 348, 349, 385, 386 Ograja in Putovići 168, 179, 180, 347 Olympia 266/267, 312, 313, 325 Orlandovtsi 291, 329 Orleans 355 Osoppo 205, 316 Ovilavis (Wels) 22, 26 Palace of Diocletian in Split 20, 21, 22, 115, 116, 117, 152, 152, 160, 201, 268, 347, 386 Palacol 249, 331 Palaiokastro near Palaiochori 266/267, 310, 325, 329 Paleokastra near Gjirokastra 266/267, 297, 297, 298, 326, 326, 340, 385 Palmyra 365, 365, 368 Pančićev vrh (Nebeske stolice) 196, 197 Panemoteichos in Pisidia 371 Panik near Bileća 168, 181, 181, 384 Parentium (Poreč) 22, 49, 49, 149, 154, 155, 383 Parndorf, Bruckneudorf 168, 170, 171, 171, 384 Pasjak near Klana 201, 242, 242, 317, 329, 339, 385 Pautalia (Kjustendil) 23, 70, 70, 158 Pecës 266/267, 295, 295, 325, 340 Pelenica near Dračevo 266/267, 301, 301, 323 Pergamon 368 Pešna near Devič 266/267, 307, 307 Pešterica near Prilep 196, 198, 339 Pharia (Stari grad) 22, 58, 58, 158, 345 Philippi 19, 23, 92, 92, 93, 153, 186, 199, 309, 338, 340, 342, 345 Philippopolis (Plovdiv) 23, 70, 70, 154, 158, 284, 346, 385 Piedicastello 42 150, 159 Pleven 196, 197, 197 Podvrh-Crkvine 182, 185 Poetovio (Ptuj) 22, 34, 35, 173, 227, 240, 341, 342 Poje in Njivice 168, 176 Pola (Pula) 22, 50, 50, 149, 151, 154, 345, 347, 383 Polače (Mljet) 188, 189, 192, 193, 347 Ponte Lambro 170, 338 Predloka 168, 174 Proboj near Ljubuški 168, 180 Proložac Donji 168, 178, 314 Pydna 23, 39, 148, 150, 344, 345 Qafa 266/267, 296, 297, 325, 328, 349 Qalaat Samaan 381, 381 Qasr Bshir 372, 374 Qasr Ibn Wardan 379, 379 Radvanje 188, 190, 190, 339 Ramat Hanadiv 379 Rankovići near Travnik 188, 194, 195, 315, 347 Ravenna 9, 10, 11, 20, 22, 38, 40, 40, 50, 112, 113, 116, 149, 150, 165, 242, 327, 337, 340, 342, 347, 349, 351, 353, 381, 383, 386, 387, 388 Reccopolis 356, 356 428 INDEX Site Text on page, Fig. on page, Map on page Rehovot 372, 376 Rifnik near Šentjur 201, 227, 227, 228, 317, 319, 325, 326, 336, 339, 343, 347, 348, 349, 349, 386 Rim near Roč 188, 190, 315, 347, 384 Riva del Garda (San Martino di Campi) 201, 202 Rocca di Garda 201, 202, 202, 316, 325 Romuliana (Gamzigrad) 20, 21, 23, 38, 89, 121, 121, 122, 122, 160, 161, 263, 268, 339, 342, 345 S. Giovanni di Ruoti 356 S. Pietro in Cariano 168, 170 Sabiona (Chiusa/Klausen) 22, 104, 104, 162, 165, 316, 319 Sadovsko Kale 266/267, 281, 283, 284, 319, 320, 332, 349 Sagalassos 366, 367 Ságvár 22, 107, 109, 160 Saint Blaise 360, 361, 361 Saint-Bernard-de-Comminges 354, 354, 386 Salla 22, 156, 341 Salemydessos-Medeia (Midye or Kujikoy) 23, 103 Salona (Solin) 19, 22, 54, 56, 56, 57, 57, 58, 58, 63, 116, 149, 151, 154, 162, 163, 191, 254, 255, 339, 341, 346, 347, 383, 386 Samograd 266/267, 278, 325 San Giorgio near Attimis 201, 205, 316 San Pietro di Carnia 24, 25, 316 Sant’Andrea di Loppio 201, 203, 204, 204, 316, 319, 332, 336, 347 Sao Cucufate 357, 358 Sardis 365, 365, 368 Savaria (Szombathely) 22, 29, 30, 30, 31, 31 Sbeitla (Sufetula) 352, 355, 355 Scampis (Elbasan) 23, 139, 139, 340 Scarbantia (Sopron) 22, 29, 30, 341 Scodra (Shkodër) 23, 73, 74, 151, 158, 341, 346 Scupi (Skopje) 21, 23, 24, 80, 81, 81, 82, 86, 139, 154, 156, 158, 160, 196, 302 Serdica (Sofia) 23, 24, 39, 65, 65, 66, 151, 157, 159, 197, 281, 338, 342, 346 Sergiopolis (Resafa) 368, 369, 370, 381, 387 Serjil a 375, 376 Shivta 372, 376 Shkorpilovtsi 291, 329 Shoumen 23, 132, 132, 164, 165, 384, 386 Sibrig domb 201, 215, 215 Singidunum (Beograd) 23, 39, 61, 61, 62, 157 Sirmione 200, 201, 202, 202, 316, 329, 335, 338, 342, 347 Sirmium (Sremska Mitrovica) 19, 23, 23, 37, 39, 59, 59, 60, 61, 61, 89, 154, 158, 160, 265, 340, 341, 345, 347, 385 Siparis 22, 113, 113 Siscia (Sisak) 22, 48, 48, 149, 156, 238, 254 Skelani 188, 195, 339 Skobelevo 182, 184 Školarice 168, 174, 174 Slovenska Bistrica 168, 173 Sopianae (Pécs) 22, 31, 31, 151, 172, 189, 341, 385 Sovizzo 168, 170 Sparta 23, 158, 159, 342 429 Slavko CIGLENEČKI Site Text on page, Fig. on page, Map on page St Chrysogonus in Glavotok (Krk) 188, 191, 315, 347 St Eulalia (Merida) 354 St Thecla near Seleukeia (monastery) 381 Stancija Peličeti 168, 176 Stari grad in Ulcinj 266/267, 278 Stari trg near Slovenj Gradec (Colatio) 168, 173 Stari Ulcinj (island) 197, 386 Stobi 19, 23, 23, 24, 82, 83, 84, 85, 85, 86, 154, 154, 156, 158, 160, 338, 341, 341, 342, 347, 384, 385 Storgosia 197 Strongyli 182, 186 Strupnić near Livno 168, 180 Stup near Sarajevo 188, 194 Styberra 338, 384 Sümeg 201, 215 Sv. Jakob above Potoče 112, 163, 164, 317 Sv. Klement near Hvar 168, 178 Sv. Pavel above Vrtovin 201, 221, 221, 222, 317, 319, 328, 328 Sv. Petka on Veliki Jastrebac 348 Sv. Trojica, Tribanj, Šibuljina 37, 150, 201, 246, 246, 331 Svetac 249, 331, 331 Szentkirályszabadja-Romkat 188, 190 Tanagra 23, 98, 151 Taor 21, 23, 140, 140, 141, 142, 151, 152, 161, 163 Taraco 351 Tarragona 354 Tatárszál áson 168, 172, 172, 339 Tergeste (Trieste) 22, 46, 46, 158, 225, 242, 346 Teurnia 22, 24, 25, 46, 46, 47, 48, 149, 151, 153, 154, 156, 208, 210, 317, 318, 338, 341 Theopolis near Sisteron 335, 361, 361 Thespiae 23, 98, 158, 196, 199, 342 Thessalian Thebes (Nea Anchialos) 23, 97, 97, 98, 151, 154, 158 Thessaloniki 19 23, 89, 89, 90, 91, 126, 144, 154, 165, 341, 347, 386 Timacum Minus 266/267, 269, 269, 326 Timgad 358, 359 Tinje above Loka pri Žusmu 15, 201, 230, 230, 231, 317, 319, 325, 343 Tipassa 355, 355 Tišina near Zenica 164, 188, 194, 194 Tokod 201, 216, 216, 329, 342 Tolmo de Minateda 356, 355 Tonovcov grad near Kobarid 15, 18, 201, 216, 217, 218, 218, 317, 319, 325, 327, 329, 334, 337, 339, 343, 347, 348, 349, 386 Toreta 201, 247, 248, 331, 331 Torre de Palma 357, 357 Tridentum (Trento) 22, 24, 42, 42, 150, 159, 336, 345 Trier 353, 353 Troesmis (Turcoaia, Iglita) 23, 138 Trojan 266/267, 273, 274, 333, 334, 339 Trojane (Atrans) 168, 173 430 INDEX Site Text on page, Fig. on page, Map on page Tropaeum Traiani (Adamclisi) 23, 137, 137, 160, 340, 341, 388 Trpčeva Crkva near Dunja 196, 198, 198, 315, 347 Tscheltschnigkogel 201, 212, 317 Tupi krš 266/267, 275, 275, 325, 333 Tzoides (modern-day city of Sliven) 23, 135, 136, 152 Ubli (Lastovo) 168, 178, 178, 179, 314, 315, 384 Ukosa in Stalać 266/267, 270, 325, 343, 348, 386 Ulaka 168, 174, 317 Ulpiana / Iustiniana Secunda (Gračanica) 23, 64, 64, 139, 158, 279, 280, 338, 341, 346 Ulrichsberg 150, 201, 212, 213, 317, 318, 318, 322, 325, 336 Umm el-Jimal 371, 372, 372 Valencia 160, 355 Varvaria (Bribirska glavica) 22, 52, 53, 54, 54, 346 Velem – Szent Vid 201, 214 Veletin 266/267, 280, 280 Veli grad (Krk) 201, 242, 242 Velika 266/267, 311, 325, 349 Veliki Brijun (Kastrum) 201, 241, 241, 327, 331 Veliki Brijun (Madona Bay) 168, 175, 241 Veliki Brijun (Verige Bay) 168, 175, 176 Veliki Gradac near Donji Milanovac 265, 265, 266/267, 330, 339, 349, 386 Veliki Sikavac 18, 201, 245, 245, 330, 331, 348, 388 Veliki vrh above Osredek pri Podsredi 15, 201, 233, 233, 234, 317, 333, 339, 385 Venec near Miokazi 266/267, 305, 305, 332 Verona 22, 24, 41, 41, 42, 149, 151, 159, 338, 345, 383 Venetia (land) 11, 24, 32, 42, 187 Viminacium (Kostolac) 23, 62, 62, 158 Vindobona (Vienna) 22, 27 Vipota above Pečovnik 34, 157, 317 Viran ş ehir (Mokisos) 371, 172 Virovitica Kiškorija South 168, 174, 174 Virunum (Zollfeld) 22, 28, 29, 29, 149, 149, 150, 151, 173, 341, 385 Višići near Čapljina 168, 180, 181 Vižula 168, 175, 347 Volpago 188, 188 Vrelo - Šarkamen 266/267, 268, 268, 340 Vrsar 168, 175 Vučipolje near Dugopolje 188, 191, 315, 342, 347 Wadi Natrun 381 Zaldapa 23, 133, 133, 152, 156 Založje near Bihać 188, 194, 194 Žegligovski kamen near Mlado 266/267, 303, 304, 327, 344 Nagoričane Zenobia (Halebiye) 368, 369, 387 Zidani gaber above Mihovo 201, 235, 236, 236, 317, 333 Zlatni Kamen near Novi Pazar 266/267, 272, 272, 325 431 Document Outline Between Ravenna and Constantinople Contents Preface 1 Introduction 2 Cities 2.1 Altered Appearances of LA cities 2.2 How to identify a LA city? 2.3 Cities abandoned before the end of LA 2.3.1 Western part 2.3.2 Eastern part 2.4 Cities with continuity 2.4.1 Western part 2.4.2 Eastern part 2.5 Cities newly-founded in LA 2.5.1 Western part 2.5.2 Eastern part 2.6. Concluding observations on LA cities 2.6.1 Transformation of LA cities 2.6.2 Characteristics of LA cities 2.6.3 Characteristics of LA cities and differences between them 3. Settlement in the countryside 3.1 Major settlement changes in the countryside 3.2 Settlements with continuity 3.2.1 Western part 3.2.2 Eastern part 3.3 Newly-founded settlements in the countryside 3.3.1 Newly-founded unfortified settlements 3.3.2 Newly-founded fortifications 3.4 Interpretation of the settlement patterns in the countryside 3.4.1 Settlements with continuity 3.4.2 Newly-founded unfortified settlements 3.4.3 Newly-founded fortifications 4. Diachronic assessment of the settlement changes and of the city–countryside dynamics 4.1 First changes of the settlement patter 4.2 Marked changes in cities, gradual abandonment of lowland settlements and the appearance of numerous 4.3 Transformation of the settlement patterns 5. Brief outline of the LA settlement patterns in other parts of the Roman Empire 5. 1 Western part of the Empire 5.1.1 Cities 5.1.2 The countryside 5.2 Eastern part of the Empire 5.2.1 Cities 5.2.2 The countryside 6. Concluding remarks and an attempt to interpret the settlement patterns Bibliography Index