Things and Thoughts. Central Europe and the Mediterranean in the 4th–1st centuries BC

ISBN 978–80–7671–005–4 THINGS AND THOUGHTSJan Kysela Jan Kysela THINGS AND THOUGHTS Central Europe and the Mediterranean in the 4th–1st centuries BC Since the beginnings of research into the Central European Iron Age, interactions of the region with the Mediterranean played an important role in the interpretation of various archaeological phenomena but also and importantly as narrative elements. However, they rarely became asubject of study in their own right. The present volume investigates the contacts between Central Europe and the Mediterranean in the 4th–1st centuries BC based on the complex analysis and contextualisation of all the available written and more importantly archaeological sources. Not only does it bring new information on the topic itself but it also sheds new light on various aspects of the Central European Late Iron Age archaeology. ISBN 978–80–7671–005–4 THINGS AND THOUGHTSJan Kysela Jan Kysela THINGS AND THOUGHTS Central Europe and the Mediterranean in the 4th–1st centuries BC Since the beginnings of research into the Central European Iron Age, interactions of the region with the Mediterranean played an important role in the interpretation of various archaeological phenomena but also and importantly as narrative elements. However, they rarely became asubject of study in their own right. The present volume investigates the contacts between Central Europe and the Mediterranean in the 4th–1st centuries BC based on the complex analysis and contextualisation of all the available written and more importantly archaeological sources. Not only does it bring new information on the topic itself but it also sheds new light on various aspects of the Central European Late Iron Age archaeology. Jan Kysela THINGS AND THOUGHTS Central Europe and the Mediterranean in the 4th–1st centuries BC Studia Hercynia, monographs 1 Studia Hercynia, monographs 1 Jan Kysela: Things and Thoughts. Central Europe and the Mediterranean in the 4th–1st centuries BC Reviewers Anne ‑Marie Adam Natalie Venclová Proofreading John Collis Judd Burden Front cover: strainer thumb ‑piece cat. n° S089, drawing J. Kysela, P.Kazakova This work was supported by the European Regional Development Fund project “Creativity and Adaptability as Conditions of the Success of Europe in an Interrelated World” (reg. no.: CZ.02.1.01/0.0/0.0/16_019/0000734) and by the Faculty of Arts, Charles University program Progress 9 “History, Key for Understanding the Globalised World”. Exchange: Institute of Classical Archaeology, Faculty of Arts, Charles University, Celetná 20, 110 00, Praha 1, Czech Republic Subscription: Charles University, Faculty of Arts, Publishing House, nám. Jana Palacha 2, 116 38, Praha 1, Czech Republic ([email protected]) Publisher: Charles University, Faculty of Arts, nám. Jana Palacha 2, 116 38 Praha 1, Czech Republic Printed by Togga Prague ©Charles University, Faculty of Arts, 2020 ISBN 978‑80‑7671‑005‑4 (print) ISBN 978‑80‑7671‑006‑1 (online : pdf) Contents Reasons for abook — 13 I. The settings 1. The historical setting — 23 Central Europe in the last centuries BC — 23 The places — 24 The actors and the events — 24 The readings — 27 Abottom line — 33 Romans drawing close — 34 2. The geographical and archaeological background — 39 Bohemia, ‘Narrower Central Europe’, WnCE, EnCE, the Mediterranean — 39 Bohemia and Central Europe in the La Tène period — 41 3. Chronological setting — 48 Historical chronology — 48 Mediterranean imports — 49 Chronologies of the Second Iron Age in temperate Europe — 51 Chronologies of the Middle and Late Iron Age in Bohemia — 55 Závist, Gate D and the chronology of the oppida period — 56 The pre ‑oppida period — 63 Numismatic chronology — 63 The chronological framework– some conclusions — 64 II. The Things and the Thoughts 1. Introduction — 69 The corpus of the imports — 70 Provenance of the artefacts — 71 Stradonice old collections — 71 Excavations — 75 Metal detector finds — 76 Imports or local products? — 76 2. The Things– Mediterranean imports in Late Iron Age Central Europe — 77 Bronze vessels — 77 Bronze vessels in Central Europe of the pre ‑oppida period — 78 Bronze vessels of the 2nd–1st centuries BC — 89 Buckets — 89 Pans — 98 Jugs — 102 Feet — 107 Simpula — 111 Strainers — 114 Mugs — 125 Amphorae — 127 Basins — 128 Basin foot? — 128 Vessel body fragments and undetermined pieces — 129 Aryballos handle? — 130 Shafted strainer — 130 Oenochoai — 131 Ashell ‑shaped attachment — 133 Basin attachment — 133 Figural boss — 134 Cordiform balsamaria — 134 Asilver duck ‑head pendant from Závist — 138 Glass vessels — 140 Mirrors — 144 Jewellery — 149 Beads — 149 Finger rings — 153 Amber ring — 166 Gold — 168 Agold filigree disc — 168 Gold amphora ‑shaped pendant — 169 Phallic pendants — 171 Medical and toilet instruments — 173 Writing instruments and (more or less) related topics — 176 Writing — 176 Writing instruments — 177 Styli — 177 Wax tablets (or not) — 181 Seal boxes — 183 Pottery — 184 Black gloss (‘Campanian’) pottery — 185 ‘Pre ‑sigillata’ and early terra sigillata — 192 Thin ‑walled pottery — 193 Unguentarium — 195 Common wares– jugs and jars, internal red ‑slip plates, mortaria — 196 Amphorae — 200 Pottery imitations — 209 The pre ‑oppida period — 209 The oppida period — 212 Beasts and plants: ecofacts and artefacts — 215 Other — 218 The question of brass and anote on bronze — 218 Oil lamps — 219 Figural attachment — 221 Strigil suspension rings — 222 Box fittings — 224 Omega ‑shaped handles and decorative nail heads — 224 Handles in the form of two opposing dolphins — 226 Anthropomorphic statuary — 226 Arms and armour — 227 From north to south — 230 Bohemica among the Italian Boii? — 230 Coins — 231 Amber, slaves, and other ghosts — 233 3. More Things– Mediterranean coins in Late Iron Age Central Europe — 236 Roman coins — 237 Greek coins in Iron Age Central Europe — 239 The Němčice mercenaries — 239 Central Europe and north ‑eastern Italy– comparison and some reflections — 241 The selection criteria and the comparison method — 241 The regional spectra — 243 Three chronological horizons — 245 Regional spectra– the cities and authorities — 245 The ‘oppida period’ — 249 An African digression — 250 Several certainties and intermediate reflections — 251 The function of Greek coins beyond the Alps: several hypotheses — 252 To sum up — 256 14 THINGS AND THOUGHTS ture; 2) local innovations based on Mediterranean models; and 3) the history of research in these matters. The last point is the right place to begin our discussion. The Mediterranean– the Greek and Roman world– has always been the natural point of reference for the archaeology of the Bronze and Iron Ages in temperate Europe. Mediterranean written sources provided a(no matter how vague and self ‑centred) narrative framework, in rare cases also suggesting names of peoples and places or hints at events. Finds of transalpine objects in the Mediterranean area or vice versa became pivotal points of relative chronologies and of considerations of cultural interaction between the two regions. According to the formulation repeated ad nauseam by many scholars, the Iron Age of Cen‑ tral Europe ‘entered into the light of written sources’. It was seldom added how dim this light was in actual fact and how little able it was to advance our knowledge. As far as the diverse categories of Mediterranean evidence are concerned, it was the information from the written sources that the archaeologists cared most about in the early stages of Bohemian research. It will be shown in detail in the chapter on the historical setting (chapter I.1) that their chief goal was to identify the historical peoples occupying the territory of the Czech lands in the Iron Age. The objects of Mediterranean origin excited them far less. Already Píč in his publication of the oppidum of Stradonice (Píč 1903– regardless of its flaws one of founding works of the Late La Tène archaeology of Central Europe) was aware of the foreign origin of most catego‑ ries of objects treated in this text. He, however, rarely referred to them explicitly as ‘Roman’ but rather ‘Roman provincial’, partly because at that time publications of material from Gaul and the eastern Alps (Gurina) were much more available than from actual Mediterranean contexts, partly due to his personal research agenda when he tried to prove by all possible means his argument that Stradonice was the seat of the Germanic ruler Maroboduus with close cultural links to the Roman world and (according to Tacitus Ann. II, 62.3) frequented by Roman traders. Mediterranean objects enjoyed less attention from the scholars of the next generation. Some ignored them completely (Niederle 1900; 1909; Šimek 1923; 19341), others (Schránil 1928; 1940; Böhm 1941; Filip 1948) duly listed the Mediterranean/‘Roman’2 finds from Stradonice without however exploiting them for more than at most astatement of the site’sinvolvement in long ‑distance trade. Discussions of interaction between the transalpine and Mediterranean world usually did not go beyond enumerating Celtic clashes with Greeks and Romans, and recognition of Greek and Roman inspiration in the Celtic coinage. In this atmosphere of general disregard for Mediterranean matters the remarks made by Jaroslav Böhm concerning the similarities between the Late La Tène oppida and Mediterranean towns (Böhm 1941, 426–449; Böhm 1946) stand out. Böhm pointed out that while the concept of urban settlements may have come to central Europe as aresult of Mediterranean inspiration, ‘this external impulse did not get lost because the indispensable [economic and social] premises for the creation of towns had already been put in place by the Celts themselves’ (Böhm 1941, 427). Böhm then went on to analyse these premises and never came back to the role of Med‑ iterranean influences which he apparently viewed as completely marginal. Later on (Böhm 1946, 29, 37–39) he even vehemently argued against the idea of Mediterranean inspiration for the oppida ascribing it to ‘insufficient knowledge of European archaeology, typical of the last century’ (unfortunately without referring to aspecific work or scholar). 1 Šimek’ssilence in this regard is somewhat stunning considering at what length in the same works he discusses bronze vessels of the 1st century AD (Šimek 1923, 59–73). 2 Their precise origin is not asubject of curiosity by these scholars and it often gets blurred by the commonly used unfortunate term ‘antique’. 15REASONS FOR ABOOK This state of affairs took anew turn with the work of Jan Filip (1900–1981, professor at Prague University in 1948–1980), almost the only archaeologist of the previous generation also active after WW2. Filip replaced the strictly material ‑based approach of the pre ‑war generation with avery particular and rather unfortunate method in which archaeology and aheavily reconstructed (and in some parts quite imaginary) grand historical narrative backed each other’sflimsy hypotheses. However, even though largely dependent on Greek and Roman written sources and not averse to stretching their evidence, Filip remained rather restrained– perhaps unimaginative or even disinterested– when it came to assessing the interaction between the two sides of the Alps. Focused primarily on the La Tène Culture itself and viewing pre ‑Roman Europe basically as an arena of apower struggle between the ethnic blocks of Celts, Germans, and Romans, Filip felt little urge to account in detail for the nature of Roman influence in Central Europe. In passing he mentioned the (in his opinion possible) Mediterranean inspiration of Late La Tène painted pottery (Filip 1956, 508); he claimed agrowing Roman influence on the oppida in their latest phases without specifying on which grounds other than historical conjecture (Filip 1956, 329–330). In dealing with the actual Roman imports in Stradonice (Filip 1956, 331) he was happy to come back to Píč’sideas that since the site in his opinion survived down to the period of Maroboduus, the finds testify to the presence of Roman traders at that time. In awork for the general public, Filip touched for the first time upon the idea of connec‑ tion between Bohemia and the migration of the Boii from northern Italy to central Europe (Filip 1960, 61), though not implying any historical or archaeological consequences of this. This narrative scheme, based on amention by Strabo (v, 1.6) was first introduced by German and Austrian numismatists in the 1930s (Paulsen 1933, 21–23; Pink 1936, 18–19; Pink 1960, 20–21) and soon after Filip’smention was embraced wholeheartedly in Slovakia by Eva Kol‑ níková (1963). In 1966, Libuše Jansová explicitly invoked the migration of the north Italian Boii to Bohemia as the cause of the foundation of oppida in Bohemia (Jansová 1970, 329, 335).3 Shortly before Wolfgang Dehn had published astudy comparing formal features of oppida with Mediterranean towns (Dehn 1961; see also Dehn 1977) but Jansová did not refer to it even implicitly and we have seen that Böhm disputed such hypotheses as early as 1941. The idea must have been around though never formulated in print. These voices remained rather isolated. After Filip’swork, major culture historical mat‑ ters were considered settled and Czechoslovak research of the 1950s–1980s focused more on fieldwork, material studies, and broadly economic issues. The Mediterranean was in no way atheme in the major synthesis of the whole of Bohemian pre‑ and protohistory from 1978 (Pleiner– Rybová eds. 1978), apart from an innovative but rather bland and vague statement of ‘experience gained by contact with the advanced regions of the Mediterranean’ contributing to the development of La Tène civilisation (Pleiner– Rybová eds. 1978, 590); the only element 3 ‘Les influences méditerranéennes à propos de l’origine des oppida celtiques et surtout à propos de leur fortifications avec un agger derrière les murs en pierres et avec les portes aux ailes et retour d’équerre avaient été envisagées déjà auparavant par quelques auteurs. Les influences romaines à propos des substructions en pierres sèches des habitats celtiques, surtout à la fin du 1ère siècle av.n.e. ont été supposées aussi par J. Filip. On ne doit aussi oublier que les Boii, après avoir été repoussés au 2ème siècle par les Romains de l’Italie du Nord, auraient pu– comme l’admet J. Filip– trouver une nouvelle patrie sur le territoire de la Bohême où ils sont historiquement attestés au premier siècle av.n.e. Ce fait aurait pu leur faciliter l’adoption d’une façon plus évoluée de bâtir les habitations qu’ils pourraient avoir appris à connaître dans les cités étrusques de leur ancienne patrie, p.ex. Marzabotto, […]’. Only one year earlier her attitude in this matter was very close to that of, for instance, Böhm (Jansová 1965, 11–12). 16 THINGS AND THOUGHTS of interest for our study is the already obligatory overview of southern imports from Stradon‑ ice brought there ‘by trade’ (Pleiner– Rybová eds. 1978, 614, 623). Asimilar sober tone was characteristic also of the first publications of research into Czech oppida– even those intended for the general public: e.g. Břeň 1966; Motyková– Drda– Rybová 1978a; Meduna 1980, 164. Only in the 1980s did some of the key categories of imports start to be published. Apart from the study of the only surviving Stradonice intaglio (Ondřejová 1981), glass imports (vessels and ring ornaments) were published by Natalie Venclová in 1990 in her overview of Prehistoric Glass in Bohemia (Venclová 1990). The most significant publication in this respect was the overview of ‘antique imports’ in the last two centuries BC in the MA thesis of Hele‑ na Svobovodá (1981) published later in two articles focused on bronze vessels (Svobodová 1983) and other import categories (Svobodová 1985). This study unfortunately suffered from several drawbacks outside the author’scontrol: the limited availability of the actual finds (some unpublished, some inaccessible to study, some kept in Vienna, beyond the reach of Czechoslovak students in the 1980s) and therefore the need to study them mostly on the basis of sometimes inadequate publications; the limited availability of appropriate publications (Svobodová’sstudies came quite early in the history of research and were to some extent pioneering, though on the flipside there was little bibliography to work with– all the more so in 1980s Czechoslovakia); and finally, under these circumstances, it was difficult if not impossible to take the important step of comparing the Czech facies with imports from other parts of Europe and trying to interpret them. This stage of research was summed up in an article by Jan Bouzek finally dedicated entirely to (the most significant part of) the topic of our study– Bohemian oppida and the Mediterranean (Bouzek 1989).4 The text is extremely dense and not always specific but still it is the most concise and clear overview of the rather restricted opinions expressed so far, not deviating from them in any particular way: the oppi‑ da do resemble Mediterranean towns (it is not specified whether and in what way they were inspired by them; p.129); the numerous imports indicating contacts with Italy, presumably via the north Italian Celts since many of these artefacts find analogies in sites like Ornavasso; fibulae on the other hand document shared fashions between both regions (p.130); there were technological innovations coming from the south including also e.g. rotary querns and the potter’swheel (p.131); ‘writing was probably not limited to Roman traders’ whose presence is thus implicitly assumed (p.131); the migration of the Boii from Italy headed exclusively to the Middle Danube area (p.132). The upcoming 1990s were to bring amajor interpretational upheaval to these sober attitudes. The idea that the north Italian Boii had arole in the cultural development of Iron Age Bo‑ hemia was in the meantime, while hibernating in Bohemia proper, being further developed in Italy and France by Venceslas Kruta (Kruta 1978, 174; 5 Kruta 1980b, 199–201; Kruta 1988, 288; 6 4 The first version of the paper (Bouzek 1982) focused specifically on imports abstaining almost completely from interpretations. 5 ‘On peut se demander dans quelle mesure la diffusion de la nouvelle mode [d’habitation en Europe centrale] n’apas eu pour principaux agents les groupes boïens qui quittèrent en 191 la Cispadane. Dans le cas où l’hypothèse d’une relation entre ce phénomène et le développement proto ‑urbain pourrait être confirmée par des arguments moins subtils que ceux dont nous disposons actuellement, les Boïens de Cispadane auraient pu avoir joué un rôle de premier ordre dans l’apparition des oppida transalpins.’ 6 ‘La disfatta dei Boi e il ritorno di una parte almeno dei loro effettivi Oltralpe, è documentata da tutta una serie di indizi di forti e dirette influenze peninsulari che rendono probabile una loro partecipazione attiva allo sviluppo degli oppida, le prime formazioni urbane dei Celti transalpini.’ 17REASONS FOR ABOOK Kruta In: Formazione CdM 1988, 315;7 Kruta– Manfredi 1999, 195–196; Kruta 2001, 341–343; Kruta In: Kruta– Lička eds. 2001, 68–69; 123; Kruta In: Kruta– Lička– Cession ‑Louppe eds. 2006, 205–207 and passim; Kruta 2018). Kruta based his hypothesis almost entirely on historical considerations and very confidently asserted, but very selectively documented, eth‑ nic affinity between the populations of southern Bohemia and northern Italy. As to the actual archaeological evidence, he relied mainly on Middle La Tène period artefacts to support his view of an almost constant flow of people and ideas between Italy and Central Europe. Evidence from Bohemian oppida, let alone the Mediterranean objects there, does not form part of his arguments. In this respect he relied on asynthesis of the Bohemian Iron Age published by Petr Drda and Alena Rybová (Drda– Rybová 1995; 1998; Drda– Rybová 1997, 109–113), whose main points were extremely complementary with Kruta’s. This synthesis is centred on the oppidum of Závist through which the authors, the exca‑ vators of the site, view all the major processes and even events of the Bohemian Iron Age. It is here that they bring the newcomers from northern Italy (the authors duly describe this migration as hypothetical but attribute it an indispensable role in the story) to establish here the first oppidum beyond the Alps. This romantic account, bearing excessive similarity e.g. to Tolkien’sneomythologies quoted at the head of the chapter, was supported with only some doubtful circumstantial evidence (an unexpectedly high chronology for the foundation of Závist corresponding roughly with the supposed arrival of the newcomers ‘in ca 175 BC’). Material proof of this narrative is limited exclusively to aseries of local vessels imitating, in the authors’ opinion, Mediterranean forms. Apart from afew marginal voices (Polišenský 2003; Jančo 2003), the idea was endorsed especially by Jan Bouzek (Bouzek 2007, 136, 143–145; Bouzek 2011a, 146; Bouzek 2011b, 69; Bouzek 2011c). Bold as it is and notwithstanding all the criticism it duly deserves (and which will be exposed below), this hypothesis has been so far the most coherent if not the only compre‑ hensive approach to the issue of Mediterranean connections in the later Iron Age of Bohemia and Central Europe. Over the last few decades, new evidence of Mediterranean contacts have started piling up and new statements and hypotheses in this respect have became more and more common: the discovery of the lowland agglomeration in Němčice nad Hanou with its numerous finds of Greek coins opened anew chapter in assessing contacts in the pre ‑oppida period (Čižmář– Kolníková– Noeske 2008; Kolníková 2012); the excavations on the Castle Hill of Bratislava (Musilová– Barta– Herucová eds. 2014) and more recently in Vienna (Adler ‑Wölfl– Mosser 2015; Mosser– Adler ‑Wölfl 2018) have shed acompletely new and unexpected light on the latest stages of the La Tène period about which we previously believed we had avery complete picture based on both archaeological and written sources– neither of them made us anticipate these new discoveries. The remarkably swift development of numis‑ matic research (cf. Militký 2015a; 2015b; 2018a; 2018b; Smělý 2017) has shown more vividly than ever before how dynamic, multifaceted and profound the Mediterranean connections were from very early on; new finds of both individual objects (e.g. Kyselaetal. 2017) and entire find assemblages (e.g. Kysela– Danielisová– Miltký 2014; Adler ‑Wölfl– Mosser 2015) from sites known and unknown before have broadened the available corpus… And yet, not even the finds collected in Stradonice in the 19th century have been fully published and properly analysed and contextualised. 7 ‘L’esperienza urbana che igruppi celtici acquisirono vivendo per quasi due secoli acontatto con le popolazioni italiche costituì al momento dell’occupazione romana della Cispadana e del conseguente ritorno oltr’Alpe di una parte dei Boi (Strabone V, 1.6), un fattore di primaria importanza per lo sviluppo e la diffusione degli oppida transalpini.’ 18 THINGS AND THOUGHTS While the Mediterranean fascination of late 19 th and early 20 th century scholars translated mostly into their reliance on written sources and their painstaking efforts to project these mentions into apassive archaeological reality, in the second half of the 20th and in the early 21st century, mentions of the Mediterranean role in the Middle La Tène period cultural transfor‑ mation (e.g. Venclová ed. 2008/2013; Danielisová 2011) and of long ‑distance trade between Mediterranean and Central Europe in the Late La Tène period (e.g. Salač 2004; Venclová ed. 2008/2013, 144; Danielisová 2011) have almost imperceptibly become commonplace (though not always asignificant factor) of archaeological discussion whether arguing in favour or against it. And yet the explanatory models available to Central European archaeologists who wish to go beyond amere statement of Mediterranean imports, influence, or simply ‘contacts’, are ultimately those present in their interpretational toolbox since the times of Filip if not Píč: returning Celtic mercenaries (but were the Celts from Central Europe ever involved in such enterprises? Did mercenaries ever return?); the Heimkehr of the Italian Boii (corroborated by nothing but wishful thinking); Roman traders in the oppida (amodel going back ultimately to the times when Stradonice was believed to be Maroboduus’ seat)… Most importantly, these topics have often been treated in isolation without sufficient concern about their context both in Central Europe and in the Mediterranean. At this point asynthesis of the available– written and archaeological– evidence on what we know about the mutual relations between Central Europe and the Mediterranean in the 4th–1st centuries BC may turn out to be useful. Throughout this quick overview of the history of research, we should realise that acommon feature of dealing with Mediterranean evidence in Central Europe has been overstatement and misrepresentation due to overrating narrative and individualised (évenementiel) aspects of apast reality, as if the Mediterranean evidence was (by virtue of it being Mediterranean?) more able to structure this than the local archaeological data. In the past, Ihave expressed critical and polemical opinions about some of these approach‑ es (Kysela 2009; 2011; 2015a; 2018/2019a). This is not the place to repeat my criticisms, not because Ihave grown more conciliatory in these respects but because the point of this study is not to engage in polemics with particular interpretations but rather to produce evidence, judge it, and advance hypotheses on the basis of this evidence itself (though it will not prevent me from occasional reference to previously expressed opinions). It brings us to the question of what this book will actually be about, what will be the main material of this study and how it will be treated. Its main part (chapters II.2 and II.3) will consist of arather old ‑fashioned typochronological and contextual study of the corpus of objects of Mediterranean origin discovered in Central European contexts of the 4 th –1 st centuries BC with an emphasis on trying to obtain afull and complete picture. The situation in Bohemia will be set in abroader picture by comparison with two neighbouring regions in Central Europe. The method may seem unsophisticated. Iwill gladly agree that such positivism should not be the ultimate step in the intellectual development of mankind, but Ifirmly believe it to be the most adequate approach for this material at this stage of research and somewhat superior to the previous impression ‑based storytelling. Some adaptations of the method will be indispensable for the subsequent chapter (II.4) in which Iwill try to judge, on the basis of several selected examples, the scope and impact of technological innovations of Mediterranean origin on the transalpine world. Both large scale phenomena and individual specific cases will be treated, once again in abroader geo‑ graphical context. Throughout this study Iwant to follow one principle– my ultimate aim is to study past communities (and their mutual relations) as archaeological units. In order to do so, Ido not need to know their names. Idon’t believe in Celts, Idon’t believe in Boii and Idon’t care about 19REASONS FOR ABOOK them; Icare about the inhabitants of Central Europe in the last few centuries BC and about their material remains and it is through their material remains that Iwant to read their stories. Ido not believe that other sources are even partially capable of helping me in this effort and Ihope to be able to demonstrate this later on (chapter I.1). Before finishing, it is important to make some quick comments about how Iactually use some of the terms mentioned above. Iabstain completely from the word ‘Celts’ unless its use in the context is justified by the ancient written sources (along the lines of Collis 2003/2010; cf. Kysela 2018/2019a); the term ‘La Tène Culture’ in my usage is not intended to be (as is sometimes the case) apolitically correct whitewash for the concept of ‘Celts’, i.e. aset of es‑ sential traits inherent to an individual or acommunity. No one was ever born aLatènian… In my view, it is ajust asimplistic tag for our own means of classification. Ido not understand a‘La Tène Culture’ or the Mediterranean as monolithic blocks of essential properties, nor as analytical units on their own, or as actors in our studies, but rather as very general settings, abackground against which we can project our material, agauge of what is normal and what is unusual. They are for me mere artificially circumscribed aggregates of material features within aspatial continuum in which two neighbouring cultures (e.g. the La Tène and Italic cultures in northern Italy) may seem closer to each other than the two extreme ends of this single culture (e.g. La Tène in northern Italy and in Britain). We may of course assume that the shared material traits reflect also some shared values and social or economic models but we must assume them to be as dynamic and changeable in time and space as the material fea‑ tures. Both material culture and social strategies are direct products of human societies and therefore they are naturally maintained as long as it is advantageous for these societies, and transformed or replaced when it stops being the case, no matter whether it is for economic reasons or of prestige. These quite basic principles of human agency should be of no surprise to anyone but Ifind it useful to lay them out. Their principal point is that unlike some previous approaches and in spite of the seemingly bipolar terminology Iuse (La Tène/Central Europe vs. Mediterranean) we are not going to approach the evidence as apart of alarge ‑scale nar‑ rative in which collective actors (‘Celts’, ‘Boii’, ‘Romans’) play out adrama of two civilisations clashing with each other. Europe of the last few centuries BC was, as Iwill try to demonstrate in what follows, aloose continuum of small self ‑contained worlds, each defined by, as well as co ‑defining, the underlying archaeological cultures and geographical areas, each in their own specific way, each with fluid and highly permeable borders; each of these small worlds has its own ways of dealing with elements of single archaeological cultures (that is of living culture of which little remains for us to judge), accepting, rejecting or passing them on to other regions. Our impression of what an entire archaeological culture is depends largely on these dialogues and intersections within the single small worlds and between them. I. THE SETTINGS 23I. THE SETTINGS 1. The historical setting CENTRAL EUROPE IN THE LAST CENTURIES BC Our knowledge about historical events taking place in Central Europe in the last centuries BC, let alone their actors, is extremely limited, based exclusively on asmall number of Greek and Roman written sources, as arule, of ambiguous interpretation. In what follows we will go through the relevant quotes followed by abrief comment. Particular attention will be devoted to the topic of the Boii who play akey role in the history of research. The available sources are few (all the relevant quotes are listed in their entirety in Appen‑ dix 1) and there have been innumerable attempts at making sense of them and synthesizing them which has resulted in dozens of slightly varying statements of wie es eigentlich gewesen ist. That is exactly what we will try to avoid doing here, aware of the specificities of the use of written sources in protohistory. The seeming closeness of the Greek and Roman writers to the described events and the comprehensibility of their writings make an apparence trompeuse of these text fragments, as solid pieces of information forming part of asingle story which we can simply reconstruct by lining up the fragments into chronological order and filling in the blank spaces with our imagination. Iwill argue that this is the incorrect way of dealing with written sources relating to aprotohistorical situation in which, from the very definition of protohistory, the source was written down in acultural environment different from that described in it. In our specific case, we have to keep in mind that our sources are seldom contemporary with the described events and not contemporary with each other either, thus potentially describing different realities; that these texts, true to the principles of Greek and Roman scholarly writing, are not necessarily independent creations based on aproper research for the facts but rather compilations of earlier texts, aconscious component of abroad Greco ‑Roman cultural discourse rather than adialogue with reality; that for the Greeks and Romans the barbarians were interesting primarily as props for their own histories and therefore their literary representations are necessarily selective, if not purposefully formulated from the start. In our case we are lucky enough to study the interaction between the Mediterranean and the transalpine world and therefore some information written down in the sources (i.e. information concerning these interactions mainly from the Greek/Roman side of this equa‑ tion) can be of simple and direct use. On the other hand, all the information concerning the transalpine world itself has to be subjected to strict contextualisation in order to understand the author’sown sources, his use of the information, as well as his motivations. Only correctly contextualised texts can be used for synthesis. Previous syntheses of the fragmentary evidence consisted often in combining the fragments into alogical order and in case of incongruity between them, choosing one and refuting the other as wrong. For most of these syntheses, the information preserved in the written sources formed the basic building blocks around which astory was constructed. The urge to turn the fragmentary sources into anarrative was so compelling that these constructed hypothetical events filling in the voids soon overshadowed the actual sources and some of the later syntheses do not distinguish the one from the other. This issue is partly due to the lack of adequate training with written sources on the part of 30 THINGS AND THOUGHTS Hallstatt cultures in southern and central Bohemia, and the La Tène flat graves represented two distinct populations, both of Celtic origin and partially overlapping chronologically, based on astill unclear chronology and the common association of Hallstatt and La Tène set‑ tlement finds. This was an important point; the idea, inherited from the earlier generation, of the co ‑existence of an earlier subdued, and an incoming dominant, population.13 One of these populations was always considered to be the Boii– usually it was the southern Bohe‑ mian barrow culture as suggested already by Niederle. The then assumed continuity of this culture until the 1st century BC made this idea still relatively plausible. The flat graves were in this case attributed either to the Volcae or anon ‑specified Celtic tribe. Exceptional in this respect was Josef Schránil (1928; 1929) who argued correctly that in 113 BC when the written sources mention the Boii for the first time, the archaeologically visible populations are the flat graves (thanks to their late dating) and the oppida. It is not to say that Schránil was ‘right’ (the search for truth is not our aim here); it is meant to demonstrate that his logical fact ‑based arguments, still perfectly within the prevailing paradigm of the time, stood no chance against the consensus of the majority. The issue of the end of Celtic presence in Bohemia took anew spin thanks to new knowledge on non ‑La Tène cultural groups in northern Bohemia, dated from as early as the Middle La Tène period (Reinecke 1915; Mähling 1944a; 1944b). ‘Gradual infiltration’ and ‘increasing pressure from the Germans’ beginning in the Middle La Tène pe‑ riod and continuing through the Cimbri, and Ariovistus down to Marobuduus became alocus communis in discussions of the reasons for the departure of Boii,14 which became universally viewed as along term process beginning in 80s or 60s BC and continuing through to the end of the millennium. Three points of this migration story were widely shared: 1) the Boii left, for Gaul to join the Helvetii and for the Middle Danube region to establish aBoian ‘empire’ there, later destroyed by the Dacians. It was only at this point, and primarily in Czech research that the Bohemian/Bavarian and the Pannonian Boii were presented not as two parallel entities but as the latter being alater offshoot of the former.15 The idea is primarily based on acom‑ bined and adapted reading of some of the written sources (Strabo: ‘the Boii earlier dwelt in the Hercynian Forest’; Caesar: ‘the Boii crossed Danube and besieged Noreia’; Tacitus: ‘the Marcomanni gained their territory in land from which the Boii had been expelled’) to which archaeological evidence was only secondarily added in order to back it up. 2) The notion of an ‘empire’ is consistently present in all the interpretations though not supported by any written or archaeological source other than perhaps Strabo’sstatement about their border on the Parisus, believed to be the Tisza. 3) Another important point on which almost all the inter ‑war scholars insist is the persistence of some of the Celts in Bohemia until the Roman Iron Age as can be assumed from the laténised nature of the earliest Germanic culture– the idea of an incoming dominant population co ‑existing with the indigenous people persists even though the Slavic card is played only exceptionally. Curiously enough, this universally agreed upon basic scheme was reached by individual scholars by completely different and sometimes radically differing chains of arguments and proofs: the Boii left for Pannonia no matter if they had occupied the south or the centre of 13 In this context it is worth noting that although the issue of Slavicity was explicitly rejected by some scholars (e.g. Schránil 1928), Jan Filip in particular continued to hint at it (Filip 1936–1937) and later even explicitly reintroduced it into the discussion (Filip 1946). 14 So much so that, for example Böhm (1941), localising the Boii in southern Bohemia, yet still needing to fit in the narrative of migrant Boii, let the tribe leave their homeland under the pressure of the Volcae from central Bohemia who on their turn were pushed by the Germans… 15 Šimek (1923) assumed an earlier migration from Bohemia concurrently to Italy and to the Middle Danube region, followed by asupplementary migration to the latter region in the 60s–10s BC. 31I. THE SETTINGS Bohemia; the Celts were under Germanic pressure no matter if it can be demonstrated by agrowing or, on the contrary, decreasing occupation of southern Bohemia; the Marcomanni were laténised no matter if they are represented by the later stages of Stradonice or by the Dobřichovice stage of the Roman Iron Age, etc. This seeming consensus was therefore not the result of aground ‑breaking understanding of the archaeological picture based on areinter‑ pretation of the written sources; both kinds of sources remained basically the same and what changed was not abetter understanding but amutual agreement on the ways to read them. This is clear from the fact that radical departures from this general scheme were possible and remained unopposed or could not be argued. For example, in Slovakia, Ondrouch published shortly afterwards an alternative idea localising ‘Iron Age Boiohaemum’ in the Middle Danube area (Ondrouch 1959). While this study is far from impeccable in terms of logical discussion and analysis of the sources, it shows to what extent the sources remained devoid of any deeper understanding, and open to any interpretation. This stage of research was synthesised in the works of Jan Filip (1956; 1960) from an archae‑ ological point of view and Josef Dobiáš (1964) from an historical perspective. Both scholars– two of the last few remaining from their generation– took unexpectedly liberal approaches in the matter. Filip in his study (Filip 1956) contented himself with summing up (not always precisely) the opinions of other scholars (Boii either in the south or in the centre, migration before 58 BC, aBoian ‘empire’ on the Danube), while not committing himself excessively; in Filip’sapproach the Boii did not matter– his narrative was about the Celts, the individual peoples were secondary (Kysela 2015a). Dobiáš (1964, 24–25, 28–30) took the same approach while demonstrating aparticularly whateveristic perspective on details: the Boii were to be identified in the flat burials because they occupied the more attractive part of the land which inherited their name, they nevertheless extended up to Danube as demonstrated by Boian set‑ tlements like Třísov and (for some reason) Linz (it is not explained why any of these is ‘Boian’). In comparison with his evasive formulations in his academic monograph, Filip showed amore opinionated attitude in his work for the general public (Filip 1960, 58–62) though camouflaged in hardly intelligible rhetorical twists and turns. He made the reader understand that the Boii are for him the population of the flat burials though insisting on the Celticity of the preceding population and mainly on its survival under the rule of the relatively small group of Boii long into the La Tène period– the conservative south remained untouched by the Boian invasion. He subsequently introduced anew element, extremely important for our inquiry: the Boii ex‑ pelled by the Romans from northern Italy migrated to the middle Danube area teste Strabo and ‘it is not excluded also to Bohemia’. In his (from today’sperspective extremely low) chronology in which LT B2 and C1 date to the 2nd century he could corroborate this assertion by pointing out the similarity of the Middle La Tène culture in Italy and in central Bohemia (contrasting it, however, to the Late Hallstatt culture of southern Bohemia rather than to other Middle La Tène regions of Europe). It was in his opinion from this moment on that the Boii (=LT B2/C1) started to penetrate into southern Bohemia in search of raw materials. The low chronology helped him equally to find archaeological proof for the departure of the Boii from Bohemia to Pannonia in the disappearance of burials in LT C1 Bohemia and their sudden increase in LT B2 and C1 in southern Slovakia. Another proof is the use of shell staters of Bohemian origin in Bratislava (Filip 1960, 116) as well as the similarity of pottery produced in Bratislava with that made in the Bohemian oppida (Jansová 1964). Filip certainly can be commended for trying to interlink the so far exclusively historical interpretations with archaeological corroboration and cannot be reproached for introducing in his interpretation the written sources for the Boian migration from northern Italy refuted by all his colleagues; we should realize howev‑ er, that the way he worked with written sources here changed considerably relative to the 32 THINGS AND THOUGHTS previous research: rather than an analysis of the single sources and of the secondary studies thereof, we are presented with acomplete story in which the source, its interpretation, and archaeological material to corroborate them intertwine into asingle whole. This is the case of both the work for the general public and for the academic study. These ideas of Filip were the last statement in the matter for several decades. Bohemian archaeologists took them for the norm or rather ignored them focusing on research questions other than historical interpre‑ tations (only vague statements can be found e.g. in Pleiner– Rybová eds. 1978). Filip’sbook for the general public kept being republished until 1995 without modifications. It was only in the 1990s and 2000s that historical interpretations, and with them also the written sources, returned as asubject of study. The prime mover of this approach was Vence‑ slas Kruta (1978; 1980a; 2000; 2006; Kruta 2018; Kruta– Manfredi 1999). His views overlap largely with those published by Petr Drda and Alena Rybová (1995; 1998), and by Jan Bouzek (e.g. Bouzek 2007; 2011a; 2011b; 2011c). Although it was of course necessary to accommodate many modifications mainly in terms of chronology, their scheme remains extremely similar to that presented by Filip or other scholars of his generation. In the opinion of Kruta, Drda, Rybová and Bouzek, Boii were present in Bohemia from the Early Iron/Middle Bronze Age (if not from the Eneolithic: Kruta 2018, 301), and can be identified with the southern (≃ Niederle, Šimek) but also central Bohemian (≃ Böhm) Hallstatt cultures. Kruta, extremely critical of the idea of amixed Iron Age population (no more acceptable in today’schronology) has basi‑ cally the entire population of southern Bohemia leave the country in the early 4th century to be replaced by an incoming flat grave population identified by him (and Bouzek, unlike Drda and Rybová) with the Volcae who occupy lowland Bohemia and Moravia. After their defeat in Italy, the Boii retreat to the Middle Danube area while, ‘it remains an open question whether some […] did not proceed to Bohemia’ (≃ Filip; Drda– Rybová 1997). Archaeological proof of this idea is looked for in the foundation of oppida in the southern part of Bohemia reputedly reoccupied by the Boii; their spread to the north is not possible due to the pressure from the Volcae (≃ Böhm). Reoccupation of southern Bohemia in search for resources (≃ Filip) is also mentioned although much more difficult to bring into correspondence with the present chronology. Thinning down of the population in the 1st century BC (≃ Schránil, Böhm, Filip) is now demonstrated on the evidence from the oppida; at this point the Boii partially retreat to the Middle Danube area where they gain ahegemony centred around Bratislava. Once again, there is essentially nothing wrong about efforts to verify atext ‑based histor‑ ical reconstruction with archaeological evidence; in our case one has nevertheless clearly the impression, that it is the story which existed in the first place and to which archaeological proof had to accommodate. It is obvious from the fact that the narrative line is basically iden‑ tical with that outlined several decades earlier by Jan Filip, regardless of the fact that rather than asource analysis, it was itself only arelatively loose interpretation of the then available source analyses. It is remarkable how pervasive the story has remained; the majority of present day Czech archaeologists deliberately abstain from historical approaches either ignoring historical interpretations completely or recognizing explicitly the inadequacy of written sources (e.g. Venclová ed. 2008/2013, 9–10; Salač 2012, 338; Mangel– Danielisová– Jílek 2013, 15; cf. Kysela 2015a, 154–155, fig. 3), but still, some elements of the inherited Boii story may appear, surely subconsciously, in their writings (e.g. the entirely imaginary migration of Boii from Bohemia to Bratislava, albeit reduced to the migration of the elite rather that the entire pop‑ ulation: Venclová 2008/2013, 10; Danielisová 2011, 116, 121). We can see that over the history of research and in particular in the 20th century, the syn‑ theses of the written sources became to acertain extent aself ‑contained doctrine growing 33I. THE SETTINGS increasingly detached from the sources themselves. The interpretations have leapt from one scientific generation to another not with the advancement of knowledge but with the pro‑ gressive establishment of ashared and agreed upon narrative. No matter how much these interpretations were propped up with archaeological arguments or how much they were called upon to explain archaeological phenomena, they were not shaped by them. What remained from one generation to another were the single narrative elements; what changed were the archaeological proofs necessary to make the narrative function in the archaeological state of art of the period. Suggestions of one generation became hypotheses in the following one and certainties in the third. Unsubstantiated hypotheses formulated from historical sources and even simple rhetorical means gradually overshadowed the sources themselves up to apoint in which the pursuit of their narrative became an end in its own right. ABOTTOM LINE Ihave tried to show elsewhere (Kysela 2012b; 2015a; 2018/2019a) how these narratives were to acertain extent unconsciously shaped by the personal, cultural, not to say political agen‑ das of individual scholars, that they too obviously reflect the scholars’ intellectual heritage, their pre ‑existing positions and dispositions or that they, in other words, formed part of the broader cultural discourse of the period. It is necessary to insist that we are dealing in all cases not with deliberate manipulations but with adaptations of an unconscious nature, always made in optima fide on the basis of each researcher’spersonal and scholarly habitus. The fact remains nevertheless that instead of arranging facts to provide us with insights into events in Central Europe in the last centuries BC, the modern syntheses are rather palimpsests of thoughts from different times and places which entrap us in labyrinths of invented stories. This is not the place to undertake aprofound discourse analysis of these texts; suffice it to point out the gaping difference between the actual sources and the interpretations constructed on top of them. For all these reasons Iwill argue that it is necessary to step back from most historical con‑ siderations proposed so far for central European protohistory and to admit the inadequacy of the available sources for any such exercise. As far as the localisation of the protohistoric populations, namely of the Boii, is concerned, Ihave previously (Kysela 2014b; 2018/2019a) endorsed ‘maximalist’ approaches (Stöckli 1979a, 198; Waldhauser 2001, 13; Collis 2003/2010, 117; Fichtl 2006; Rieckhoff 2009) which do not explain the multiplicity of localisations with invented migrations between them but rather by redefinition of these ethnic groups and by considering how we perceive them. The ubiquity of the Boii in Central Europe could therefore be explained by the fact that rather than asingle tribe we may be dealing here with acollective ethnonym at the level of Caesar’sCelts, Belgae and Aquitani rather than Aedui or Remi. It is worth realizing moreover that the Boii in the sources constantly give the impression of marking the limits of the writers’ knowledge: at first they seem to be at the geographical margin of the world Rome had to deal with, while later they become the horizon of slowly waning memories. Under these circumstances it is not at all impossible that the term Boii was used in the Greco ‑Roman world only as ageneric catch ‑all label for the central European Celts. If it were so, it is entirely possible that the term acquired this status per extensionem from amore specific Selbstbezeichnung– after all, the civitas Boiorum in Pannonia obviously consisted of people calling themselves Boii; there is however no way of finding out what was the correct use of the term in the actual pre ‑Roman Central Europe. Whether it was so or not, it is clear that an ethnonym about which we cannot even be sure 34 THINGS AND THOUGHTS what kind of ethnic group it describes is hardly auseful basis for any historical considerations. In the following text, the use of ethnic terms will be systematically avoided; it contributes in no way to our knowledge and only obfuscates our ability to reason on purely archaeological grounds. The word Boii will only be applied to the ‘Boian coinage’ with full awareness of it being amodern technical term with no significance for apast reality. References in ancient sources to Boii will be understood as broadly relating to Central Europe unless the source explicitly states otherwise, e.g. the clash between them and the Dacians in which the Boii are clearly to be understood as inhabitants of the northwestern edge of the Carpathian Basin. As far as historical events are concerned, we will ignore the information from Livy about Boii arriving in Italy in the 4th (let alone 6th century BC) as amere ex post reconstruction. The same is true for Strabo’sidea of areverse migration from Italy to the Middle Danube area, clearly debunked by Dobesch. Likewise, Caesar’sreference to the Volcae occupying at an unspecified period an unspecified part of Central Europe is of no value as it is too vague and has no link to any other written or archaeological record. Under these circumstances, the earliest written reference to Central Europe is the campaign of the Cimbri sometime before 113 BC.Avictory of the Boii is the only thing we can be sure of; the exact location or the political and social consequences of this encounter are only mere guesswork. Sometime before 58 BC agroup of Boii besieged Noreia, probably contrary to Roman interests. Once again, we know nothing else and extending this event to broader political considerations is nothing but storytelling. In 58 BC agroup of 30,000 Boii joined the campaign of the Helvetii in Gaul. In the Carpathian Basin at one point the Boii controlled an area up to (or rather not extending beyond) the river Tisza but were defeated and annihilated by the Dacians led by Burebista. This war can be dated to sometime around 46–35 BC (Göbl 1994; Dobesch 1994/2001; 1995/2001). In the 1st century AD acommunity identifying itself with the Boii demonstrably lived in the then Roman Middle Danube area and it makes sense therefore that the most consistent information on the Boii comes from here, but the Boii were clearly remembered or considered to be the original inhabitants of aregion including present day Bohemia and probably parts of Bavaria. There is no way of supplementing this flat and static picture with dynamic elements like events and status (formation of the tribe, migration of its parts, mutual dependence be‑ tween the areas, alleged hegemony in the Middle Danube region); no source provides any help on this matter. No source is earlier than 113 BC and any considerations of events prior to this date, including ‘the formation of the Boii in Bohemia in the 5th–4th century BC’, are mere fantasy. Most importantly, Iwill insist that we do not need written sources at all. If their informa‑ tion is vague and ambiguous, let us not hesitate to dismiss it. Archaeological sources provide us with apicture which is rich and complex enough to more than make up for the lack of any written information. Those who miss in this case the kinds of information which only written sources can provide (ethnonyms, events, and persons) are advised to turn to other historical periods. The sources on Central Europe in the last centuries BC can only provide blurred im‑ pressions, never clear pictures. They can only be background music to the main role played by archaeology according to the rules of that discipline. ROMANS DRAWING CLOSE Compared with the dearth on information on Central Europe in the period under study, we are somewhat better off as far as the information on the Roman presence in the regions neighbouring Central Europe is concerned (overview e.g. in Vedaldi Isabez 1994). The Ro‑ 35I. THE SETTINGS mans conquered northern Italy in the aftermath of the 2nd Punic War. By the early 180s BC, they controlled in one way or another the whole of the Po plain; while the regions south of the river were under direct Roman rule with the annexation of large portions of land and the foundation of both Latin and Roman colonies and other centres; Transpadana and Veneto were tied to Rome by aseries of alliances (Brizzi 1987). Transpadane settlements only started acquiring colonial status after the Social War in 91–88 BC, and that only as so ‑called fictitious colonies, i.e. by the elevation of indigenous settlements to colonial status rather than by new foundation by colonists from central Italy (overview e.g. in Denti 1990, 31–53). In this perspective the case of Aquileia, founded as aLatin colony as early as in 181 BC in the farthest northeast corner of the Adriatic, is exceptional. The foundation was preceded by aremarkable series of events in which the site of the future colony was first occupied in 186 BC by acontingent of 12,000 Gauls ‘from beyond the Alps’. The Romans approached them in aresolute though stunningly peaceful way; first the Senate required an explanation from representatives of their nation directly in their homeland (Liv. xxxix, 22.7). Asubsequent ‘punitive’ expedition against the newcomers in 183 BC only contented itself with disarming them. The Gauls left after an unsuccessful appeal to the Senate; the transalpine authorities once again sided with Rome in this matter (Liv. xxxix, 54–55). Another failed attempt by Celts to settle in this region is recorded in 179 BC (Liv. xl, 53.5). The insight this series of events gives us into the Roman northeastern policy is extremely valuable and hardly finds an equal among the Roman written sources in this respect. Not only does it allow us to understand what detailed ideas the Romans had about even the most distant parts of the Cisalpine region and about their strategic interests in it, but also it makes clear that even at this extremely early date the Alpine foothills were not the horizon of Roman diplomacy. Both the Romans and the transalpine elites demonstrated an extraordinary degree of mutual understanding and ability to act in concert; what we are observing here is either the moment in which dip‑ lomatic relations and mutual policies were being established or possibly even the enactment of previously reached accords. The foundation of Aquileia followed in 181 BC (Liv. xl, 34). Further reinforcement of the colonial contingent occurred in 169 BC: Liv. xl, 34) and the colony was attached to the rest of the Roman world by the via Postumia in 148 BC and the via Annia in 153/131 BC.It clearly had the purpose of strategically securing the northeastern corner of Italy whether by protecting it from incursions or by laying foundations for further expansions, directed in the first dec‑ ades of the 2nd century mainly against Istria and Illyria (Šašel Kos 2013). Throughout this period, the Romans seem to have continued to cultivate good relationships with some local populations. Already in 178 BC we hear about 3,000 Celtic auxiliaries assisting the Romans in afailed campaign against the Istrians (Liv. xli, 1.8). Although no details are given and the Celts could surely come from northern Italy (e.g. Insubrians or Cenomani) it is legitimate to assume that they may be one of the Alpine populations, involved for the first time in the Roman alliance system (Šašel Kos 1997, 26). When in 171 BC the consul C. Cassius Longinus recklessly led his troops from Aquileia along the Adriatic coast, the Senate responded to the complaint by the Celtic ruler Cinibilius with ahigh profile investigation and compensation to the impaired (Liv. liii, 5.1–10). Polybius reported on discovery (‘during his lifetime’ = 203–120 BC) of gold mines by the Taurisci in whose exploitation Italians took part at least for aperiod of time (apud Strabo iv, 6.12). In 113 BC the Romans first clashed with Cimbri ‘in the territory of Norici who had friendly relations with the Romans’ (Strab. v, 1.8; App. Kelt. 13). Between 58 and 49 BC, Caesar spent his proconsular years as governor of Cisalpine Gaul and Illyricum and although he is now much better known today for his activity in Transalpine Gaul in the same period, he spent aconsiderable amount of time in northern Italy, repeatedly 36 THINGS AND THOUGHTS quartered in Aquileia, held courts and was involved in matters concerning the politics of the northern Adriatic and eastern Alps (Vadaldi Iasbez 2000; Šašel Kos 2000; Santangelo 2016). We have already seen Caesar alluding to the Boian attack on Noreia as athreat to Roman interests (BG i, 5.3) for which we lack the broader context certainly well known to his Ro‑ man readers. Adecade later, Caesar’stroops were assisted by 300 riders sent by the king of Noricum in the war against Pompey (BC i, 18.5); although the latter case may testify to relations on apersonal rather than at state level, it clearly makes the case for shared interests of both men and probably of arelationship cultivated between them. Fig. 1: East Alpine area in the Late Republican period– sites quoted in the text. At least over the last century of the Republican period, the Romans gradually gained afoot‑ hold in the part of the eastern Alps adjoining the territory of Aquileia (Fig. 1). Strabo (iv, 6.10) describes the itinerary beginning in the Ocra Pass and following to the river Nauportus (today’sLjubljanica) and asettlement of the same name. This territory, dominated later by the city of Emona (today’sLjubljana) was considered part of the ager of Aquileia and still in the Imperial period it belonged to Italy rather than Pannonia (Šašel Kos 2013, 195–199). Ar‑ chaeological sources clearly corroborate these reports; several sites have been excavated in the Ocra (today Razdrto) Pass, from which especially Mandrga has produced plentiful finds of Republican date (Horvat– Bavdek 2008). The site of Nauportus (today Vrhnika some 20 km upstream from Ljubljana/Emona) has been known for quite some time thanks to its Augustan warehouses (Horvat 1990; Mušič– Horvat 2007); avery recent rescue excavation however identified asignificant Late Republican horizon on the site, documenting aRoman presence from at least the 1st century BC (Vojaković– Bekljanov Zidanšek– Toškan 2019). It was the area of Emona which became the bridgehead for Octavian’swar in Illyricum in 35–33 BC conquering both the Illyrian and Dalmatian coast and the interior centered on Siscia 37I. THE SETTINGS (Šašel Kos 2005; Šašel Kos 2013, 187–195). This circumstance makes clear the significance of this route for communication of Italy with the Balkans and Pannonia. The situation is somewhat less clear in anortherly direction, through which the routes to central Europe most probably passed. In immediately pre ‑Augustan and Augustan periods we can reasonably assume aRoman presence in (and Roman commercial and political interests relying on) the sites forming acommunication axis Zuglio (Forum Iulii) / Moggio Udinese– Gu‑ rina– Magdalensberg. For earlier phases of the Late Iron Age evidence is unclear. In the whole of northern Friuli, Romanisation surely promoted by Aquileia becomes evident from the 2nd century BC (Vitri 2001). The examples located in the Pre ‑Alps are Zuglio in the valley of the Bût and the very recently published Moggio in the valley of Fella (Faleschini 2018); both the mentioned rivers are tributaries of the Tagliamento flowing from here to Aquileia. Excavations in the area of the forum of the later Roman town (of Caesarean or Augustan foundation) in Zuglio and in different spots of Moggio identified occupation going back to the the middle/ end of the 2nd century BC and including imports of Italian tableware. Continuing upstream of the Bût and having crossed the next Alpine ridge there is the site of Gurina; its recent publication (Gamper 2015) clearly demonstrates the significance of the site (also in terms of contacts with Roman Italy) but unfortunately some aspects of the author’sagenda make full understanding of the site (mainly its chronology) problematic. Other local hillforts feature some, though only relatively meagre, evidence of contacts with Italy prior to the Augustan period (overview in Gugl 2000, 129–139) The most important site in this part of the Alps is without doubt the Magdalensberg further downstream the Gail and Drau/Drava. The site is generally famous for what is plausibily presented as asettlement of Roman/Italian merchants and craftsmen on the slopes of the mountain; the nature of the native (?) occupation on the hilltop and its chronology as well as the date at which the Roman merchants first settled on the slopes has been subject of much debate. The traditional dating of the foundation after the middle of the 1st century is argued in Dolenzetal. 2007; 2009 often with explicit rejection of the low chronology (foundation only in 16/15 BC) proposed by P.Gamper (synthesis in Gamper 2015). D. Božič (2008) drew attention to the sporadic presence of finds from the first half of the 1st century BC.Gamper’slate dating of the site is out of the question and contradicts any available evidence. Iam more inclined to see the beginning of the occupation in the first half of the 1st century though admittedly the evidence is meagre. Whether in the Magdalensberg or elsewhere, there must have been sites in the eastern Alps which were in contact with the Roman world already in the first half of the 1st if not in the 2nd century– linking points between sites like Zuglio or Moggio in Italy and Central Europe (the intimacy of these early contacts is made clear by the site of Wien ‑Rochusmarkt, see below). The Alps as awhole came under Roman rule as aresult of acombined campaign led by Drusus and Tiberius in 15 BC (Strab. vii, 1.5; Horat. Carm. iv, 14.14). Among the regions dealt with in this study it is mainly the central Alps with the valleys of the rivers Adige and Inn. We do not have clear information as to how major the Roman impact on southern Germany was at this point. Although the province of Raetia was not established until the Claudian period, the Roman presence was most likely constant from the Augustan period, its exact manifesta‑ tions are, however, difficult to circumscribe archaeologically (Hüssen– Irlinger– Zanier eds. 2004; Zanier 2004). Similar uncertainties concern the neighbouring regions of Noricum. We do not hear about conquests in this region previously attached to Rome by presumed alliances or elite relationships. When in AD 9, Tiberius concentrated his forces in Carnuntum, ready to launch the offensive against Marobuduus, we cannot be sure whether he was admitted there by the Noricans (or the Boii? There are no hints that Norican rule extended so far north) or whether 38 THINGS AND THOUGHTS the territory had already been in one way or another under Roman control (as suggested e.g. by Kovács 2015). This issue is, nevertheless, unimportant for our needs: the mention lets us note that the Romans obviously had aperfect knowledge of this area and complete freedom of action there including large scale military operations; under these circumstances, the ex‑ act administrative status of the region has no significance for us. The more important point, when exactly the Romans gained such afree access to this region– was it in connection with the 15 BC Alpine campaign or earlier?– is not addressed by any source. 39I. THE SETTINGS 2. The geographical and archaeological background Bohemia in solo barbarico trans Danubium sita, Germanie porcio est […] Gelida prorsus regio […] Plebs toto regno bibula, et ventri dedita, supersticionum obsequax ac avida novitatum […] Qui paulo excellunt atque inter plebem nobilitatemque medii sunt, audaces, versuti, ingenio vario, lingua precipiti, rapinarum avidi sunt. Et quibus nihil satis esse possit. Enea Silvio Piccolomini, Historia Bohemica, 1457 BOHEMIA, ‘NARROWER CENTRAL EUROPE’, WNCE, ENCE, THE MEDITERRANEAN The working area of this study (Fig. 2) is in the first place Bohemia, conceived as the drainage basin of the Elbe and the Vltava and delimited by the mountains along the present Czech po‑ litical frontier. The precise definition of the frontier between Bohemia and Moravia need not trouble us here as the La Tène occupation in this frontier area, though extremely significant from many points of view, will not be part of our inquiry. Fig. 2: Narrower Central Europe– WnCE, Bohemia, EnCE. 46 THINGS AND THOUGHTS unclear: Parma 2014). Numerous hill forts in the northern‑ and easternmost part of Moravia belong to the (non ‑La Tène) Púchov Culture (for the Púchov Culture in Moravia see Geršl in print). Plavecké Podhradie is the only site which can be classified an oppidum in the Small Carpathians in the east. The most prominent hill top site in Lower Austria is the rather small Oberleiserberg (for the hillforts in the region in general Karwowski 2015). The research focus on the oppida in Bohemia has left us with relatively little information on the lowland open settlements of LT C2–D1. The agglomeration of Lovosice surely survived through this period but little is known about the settlement apart from its approximate extent and chronology (Salač 1990). The small hilltop site of Kolo near Týnec nad Labem remains little known (Sedláček 1981) but ariver harbour has recently been excavated at its foot (Beneš 2015; Militký– Beneš 2016) providing insight into the complexity of the settlement pattern of this period. As far as simple open settlements are concerned, little information has been systematically published and we are mostly relegated to scanty information from rescue excavations and metal detector finds. The available data lend themselves rather to macro‑ scopic territorial studies (Zavřel 1996; Venclováetal. 2001; 2008; Danielisová 2010). Only afew lowland settlements are entirely excavated and published in the region: e.g. Bořitov in Moravia (Čižmář 2003), Michelstetten in Lower Austria (Trebsche 2010), or in Bohemia Radovesice (Waldhauseretal. 1993). In some regions, settlements were extensively sampled by surveys or/and excavations and pottery typochronologies were established providing afair insight into settlement dynamics (Říčany and Loděnice regions: Venclová ed. 2000; 2008; NW Bohemia: Salač– Kubálek 2015). In Lower Austria with its numerous investigated sites, acomplex hierarchy has been convincingly outlined (the large central place of Roseldorf, the middle ‑sized settlements of Haselbach and Stripfing, small hamlets such as Michelstet‑ ten: Trebsche 2014; 2016). In spite of these advances, we still have difficulties to generalise this picture, to easily distinguish e.g. pre ‑oppida and oppida period settlements in most of the study territory and thus to understand the transition between these two phases, and to characterise their nature. Although the settlement structure in the north and south of 1st century Bohemia differed considerably due to the presence of oppida in the latter region and their absence in the former, the same coinage (Fig. 4) circulated in both of its halves– the so ‑called ‘shell staters’ and their fractions based on the earlier EnCE coinage of the Athena Alkidemos series (Militký 2015a; 2015b; 2018/2019). Local varieties of shell stater coinage were used in EnCE and in WnCE the so ‑called Vindelician coinage circulated in the south while specific local coinages are char‑ acteristic of northern Bavaria and Thuringia. The oppida period came to its end in LT D1b or the very beginning of D2 by which time the oppida were being or had been abandoned; some such as Manching, Staré Hradisko or Závist probably already came to an end in LT D1b while Stradonice and Třísov probably persisted some time into LT D2 (Danielisová– Miltký 2014). This period between LT D1b and D2 is characterised by the sudden rapid rise of the op‑ pidum of Bratislava which stands out in its Roman style masonry architecture (Musilová– Minaroviech 2014; Resutík– Minaroviech 2014; 2017). Another, probably slightly earlier site should be mentioned here, the recently excavated Wien Rochusmarkt/Kundmanngasse which yielded numerous unexpected finds of Italic origin (Adler ‑Wölfl– Mosser 2015; Mosser– Adler ‑Wölfl 2018). Little is clear about the end of the La Tène period in central Europe. Moravia shows no signs of being occupied between roughly the middle of the 1st century BC and the beginnings of the new era, only in the first decades AD the ‘Germanic’ finds start to appear consistently (Zeman 2017). In Bohemia, the circumstances of the survival/departure of the La Tène pop‑ 47I. THE SETTINGS ulation (at the transition of LT D1 and LT D2) and their mixing/replacement with incoming Germans (first of the Großromstedt, later of the Elbe ‑Germanic culture) are still subject to debate (Droberjar 2006; overview in Beneš 2018/2019). The discussions are equally incon‑ clusive but more heated in Bavaria (cf. Rieckhoff 1995; 2012 and Zanier 2004). The distinction between our working areas (Bohemia, EnCE, WnCE) was based exclusively on geographical criteria. Above, we have seen nevertheless, that the each area differs from the other e.g. in terms of settlement pattern or of coinage. In these two respects there appears to have been alarge scale cultural affinity if not unity in LT C period EnCE with its widespread Athena Alkidemos coinage and shared settlement model of large open agglomerations. Bo‑ hemia in the same period with its numerous small scale coinages (dependent largely on the EnCE model) and no clear central places gives amuch more disorderly and decentralised impression. In the oppida period on the other hand, Bohemia appears to have stuck to asin‑ gle monetary model of the shell type stater regardless of the apparent division of the land in terms of settlement organisation between the south based on the oppida model and the north where open settlement remained the norm. Jiří Waldhauser (1996) attempted to break LT C2–D1 Bohemia down into aseries of regions based on their characteristic pottery production, which, in spite of the author’sclaims, surely cannot reflect more than the natural variability of material culture. EnCE participated in the oppida period in the Bohemian monetary shell stater model though in its own characteristic way. The oppidum of Bratislava seems to have been aunit completely on its own– rather in‑ sular in time and space– which at the very end of the La Tène period further developed the earlier shell stater model combining it with large silver coinage, the hexadrachmas of the Biatec series (Militký – Torbágyi forthcoming). We have no clues as to whether these material ‑based groupings do actually reflect apast reality of some kind, nor to what social units these archaeological structures could correspond. In western Europe, in particular in Gaul, territorial issues of Late La Tène archaeology have often been discussed over recent decades based on the notion of civitates– i.e. the single peo‑ ples, tribes or states as recorded by Caesar (e.g. Fichtl 2004a). This approach goes beyond the antiquarian compulsion for squeezing the written sources in order to identify and localise ancient peoples for the sake of their identification. It tries instead to delimit actual social units which may help us to better understand material culture. It is clear, however, that such atask is impossible without the help of written sources which in our case are, as we have seen above, absolutely insufficient. The comparisons between these historically documented communities and territorial groupings based on material culture in Gaul have shown moreover how diffi‑ cult, if not impossible, it is to bring them into focus: in terms of pottery asingle civitas may host several different pottery productions but other ceramic categories may broadly cross the borders of the same civitas (Barral 2003; Bonaventure 2011, 253–263); the same is true of coins (Gruel 2002). 48 THINGS AND THOUGHTS 3. Chronological setting The present study unfolds on numerous stages, combining evidence of an historical and an archaeological nature, and elements from regions using different chronological systems. Syn‑ chronicity of specific events and archaeological phenomena was also akey argument of some past interpretations which we are dealing with. Therefore, establishing asolid chronological framework is an indispensable basis of our inquiry and method of analysis. HISTORICAL CHRONOLOGY As has been shown above, ancient sources have preserved little information concerning the history of pre ‑Roman Transalpine and especially Central Europe. Most of these mentions are moreover so vague that they leave alarge margin for interpretation, speculation and manip‑ ulation. In order for an historical interpretation to work, the only requirement is proving at least approximate synchronicity of the archaeological context in question with the event we are projecting on to it. In this way historical interpretations and chronology have often been interconnected in avicious circle in which one allegedly corroborated the other. The dearth of historical events recorded in pre ‑Roman Central Europe makes any of them very desirable and endlessly re ‑usable for historical interpretations, often in various regions. So the campaign of the Cimbri and Teutones has been assumed to have caused the aban‑ donment of Němčice (Čižmář– Kolníková– Noeske 2008), the fortification of Manching (Sievers 2003), the settlement shift in Basel from the Gasfabrik to the Münsterhügel (Rieck‑ hoff 1995 20 ), the appearance of oppida in the Rhineland (Leifeld 2007), and even in Gaul. Similarly, the Boii were believed to have arrived in Central Europe from northern Italy and to have brought the knowledge of building oppida to Bohemia (Drda– Rybová 1997), as well as importing Greek objects (Benadik 1981) and imitations of Campanian coins (Kolníková 2012) to western Slovakia. Enough time has been dedicated above to expressing reserves about these historical interpretations. Ihave naturally no objections to using artefacts from undoubted (!) historically dated contexts (say, Alesia) as benchmarks for protohistoric chronologies; the problems begin when chronologies are based on events supposed to have occurred in the re‑ gion in question (e.g. the destruction of Manching by the Romans in 15 BC; the abandonment of oppida in Bohemia after the arrival of Germans around AD 0; the genocide of the Boii at the hands of the Dacians after 44 BC). The most (in)famous of these events include: 191 BC– the defeat of the Boii in northern Italy and their alleged return to Central Europe (cf. chapter I.1). 20 The author criticised harshly Furger ‑Gunti’sview (Furger ‑Gunti 1979) that the settlement shift can be connected with the migration of the Helvetii, dismissing it as methodically flawed and impermissibly contaminating archaeology with historical interpretations. Nevertheless, she herself integrated the migration of the Helvetii in her discussion, explaining with it the ultimate abandonment of the Münsterhügel. Rieckhoff did not make it clear in what respect her method differs from that proposed by Furger ‑Gunti. 49I. THE SETTINGS Before 113 BC the passage of the Cimbri through the territory of the Boii (Strab. vii, 2.2) linked, for example, with the abandonment of Němčice nad Hanou and the construction of oppida in Bohemia, etc. 113 BC– the defeat of the Romans by the Cimbri at Noreia (Liv. Per. 63; Appian Kelt. 13). 113–100 BC– the campaign of the Cimbri and Teutones (Plut. Marius xi–xxvii; Caesar BG i, 33.3). Before 58 BC– the Boii besiege Noreia (Caesar BG i, 5), sometimes considered the terminus ante quem for the end of the oppida in Bohemia (= the abandonment of Bohemia by the Boii). Before 58 BC– Ariovistus in Gaul (Caesar BG i, 31, 43–46). His passage from ‘Germany’ to Gaul is often believed to be reflected in La Tène Central Europe. 58–52 BC– Caesar’sGallic War. 44–35 BC– Burebista’sextermination of the Danubian Celts (Strab. vii, 3.2, 11; vii, 5.2), tradi‑ tionally considered the date of the end of the La Tène Culture in the Middle Danube area and the destruction of the oppidum of Bratislava. 35 BC– Octavian conquers Pannonia (Cassius Dio xlix, 37–38). 15 BC– the Alpine campaign (Cassius Dio liii ‑lix; Tac. Ann. i–ii), originally believed to be the cause of the abandonment of Manching. 11–9 BC– Drusus‘ campaign in Germany (Cassius Dio liv–lv, Tac. Ann. ii, 8). ca 6 BC?– Maroboduus and the Marcomanni occupy Bohemia (Tac. Germ. 28, 42). 6 AD– Tiberius‘ campaign against Maroboduus (Velleius Paterculus ii, 109)– associated in various ways with the oppidum of Bratislava. Though providing auseful elementary timeline, the utility of these events for chronological interpretation is extremely limited as their relation to any specific region, site or context is purely hypothetical. Therefore, the only archaeological contexts taken into consideration on historical grounds as chronological anchor points will be those directly related to the Romans. MEDITERRANEAN IMPORTS Mediterranean objects imported to Transalpine Europe will be dealt with in much more detail in the following chapter. At this point it is only necessary to make some remarks about their ability to provide information relevant to Transalpine chronology. In the period and area most likely to provide abasis for cross ‑dating with our area of study, that is in 3rd/2nd–1st century Italy; well ‑dated sites are relatively few. Settlements dated to this period were often also occupied throughout the Imperial and Medieval periods up to the present day and well preserved (and published!) Pre ‑Roman or Republican phases are as arule hard to find. It is still in this period (corresponding with the spread of the Roman way of life and death) that burials ceased being furnished with grave ‑goods from the find categories exported across the Alps. Only afew regions remain of relevance in this respect, mainly Celtic northern Italy (Veronese: Salzani ed. 1995; 1996; 1998; Lombardy: De Marinis 1986; canton Ticino: Martin ‑Kilcher 1998; Pernet– Carlevaroetal. 2006). Late Republican Roman military camps are best known from the Iberian Peninsula (Luik 1997). Out of these, however, only Numantia (pre‑133 BC; Luik 2002) and Cacéres el Viejo (pre‑79/72 BC; Ulbert 1984) have been published to any extent. In Gaul, only the siege ‑works of Alesia (Reddé– von Schnur‑ bein 2001), Gergovie, and Uxellodunum can be unequivocally linked with the Caesarean war whereas the military presence at La Chaussé Tirancourt, the Titelberg, Petriesberg or Her‑ meskeil are probably one generation more recent (Poux ed. 2008). Extraordinarily valuable for chronological discussion are famously the Augustan camps on the Rhine– Dangstetten (15–9 BC), Oberaden (12/11–9 BC), Haltern (7/5 BC–9/14 AD), etc. Shipwrecks, often loaded with 50 THINGS AND THOUGHTS hundreds of amphorae, as closed contexts par excellence have agreat theoretical potential for sorting out chronological questions; too often though they become in this respect rather the object of wishful simplifications (as denounced by Tchernia 1990 and Olmer 2012). Another factor to bear in mind is the considerable discrepancy between Italy and Transalpine Europe in terms of the research attention given to the various artefact categories. This is par‑ ticularly obvious in the case of brooches: the cornerstones of Transalpine chronology, still well studied in northern Italy, are basically ignored south of the Apennines. Yet, the few examples published or displayed in museums21 often correspond with or at least nicely complement La Tène typologies. Could these brooches, often too quickly and too easily labelled as ‘Celtic’, be in reality the standard element of Roman clothing of the period, lost in avicious circle of the small number of finds and alack of research interest? The publication of fibulae from northern Italy by Demetz (1999) is one of the most precious contributions to the topic of Late Iron Age brooches. Last, but not least, afactor negatively influencing our understanding is the relatively low artefact variability over the last two centuries BC.As in many other respects, the breaking point for Mediterranean chronologies is the Augustan period when basically all categories of objects, from pottery and amphorae, through glyptics and toreutics, to coinage, underwent radical and rather dynamic transformations. In the century or century and ahalf before that on the other hand the materials seem to have evolved almost imperceptibly. For metal vessels, the term ‘Late Republican forms’ is as all ‑encompassing as it is chronologically vague. In the case of amphorae, we may talk about developmental trends rather than strictly defined and chronologically delimited types (Olmer 2012). Following Lamboglia (1955), amphorae of type Dressel 1 attested by hundreds of thousands if not millions in Gaul, are traditionally sub‑ divided depending on the height of the lip into sub ‑types 1A (<55 mm), 1B (>55 mm), and (less relevant for our needs) 1C. The transition between the types Dressel 1A and 1B is usually placed in the 70s BC, based on the Mandrague de Giens wreck in which the majority of the published amphorae rims oscillate around 55 mm and which at the same time yielded coins minted in the timespan 155–75 BC.22 Things are however not so simple and clear ‑cut as they are sometimes pre‑ sented. Dressel 1 amphorae were in fact produced all along the Tyrrhenian coast in numerous independent workshops and rather than astrict distinction of types 1A, 1B we should envisage aconcurrent production of numerous varieties which may, but need not, follow the general trend, and may reflect deliberate choices linked to workshop specificities, marketing strategies and other factors. The general trend of gradual rim lengthening remains in principle valid, but its pace was probably not uniform and anchoring one stage or another of this development to an absolute date may be an illusory task as the typologically earlier forms (‘Dr 1A’) kept being produced albeit in smaller volumes alongside the more developed ones throughout the produc‑ tion period of the type (Olmer 2002; 2012). Even the opportune Mitfund of amphorae and coins in asingle closed context (such as the Mandrague de Giens wreck) is therefore not asufficient ground for dogmatic conclusions. Asimilarly gradual transition is characteristic of the Adriatic 21 E.g. Cosa, antiquarium: unpublished to my knowledge. Bolsena: Santrot– Santrot 1995, 28, fig. 8:45 (an Almgren 65); Gubbio: Malone– Stoddart eds. 1994, fig. 6:10‑2 (aMiddle La Tène scheme LT C2 brooch); Rome, the temple of Vesta: Piana Agostinetti 2006. The situation in southern Italy and Sicily has recently been synthesised by H. Baitinger (2012) whose conclusions go in adirection very similar to ours. 22 The date of the ship’ssinking proposed by the excavators of the wreck is in reality stated as ‘75–30 BC, most probably 60–50 BC’ (Tcherniaetal. 1978, 15–17; other reflections in Tchernia 1990). It is on acomparison of the Madrague rim heights that the Clemency tomb, one of pivotal points of the Rhineland chronology has been dated to the ‘70s–60s’ (Metzleretal. 1991); the date was later raised to around 80 BC (e.g. Metzler– Gaengetal. 2009, fig. 402). 51I. THE SETTINGS amphorae: the type Lamboglia 2 was replaced by the type Dressel 6A only by means of different transitional forms spanning two or three decades before the 30s BC. The black gloss (‘Campanian’) pottery adheres in its major classes to the same form rep‑ ertoire throughout the Late Republican period and attempts at defining chronologically sen‑ sitive types (e.g. Božič 2008) have proved unconvincing or of only limited regional validity. Decisive changes occurred in Mediterranean pottery only in the 40s BC with the appearance of the first red ‑slip wares with anew form repertoire which also partially influences the black gloss production. Chronologically the most promising imported objects are Roman coins, often dated with the precision of ayear. As we will see, however, very few Roman coin finds come from reliable contexts, with many deposited and even imported only in later periods (cf. Militký 2009a; and below). In chapter II.3 we will (in vain) try to gain from Roman coins some information on the nature of Transalpine contacts. At this point, we may nevertheless make adigression into one particular phenomenon significant for chronological considerations. The inventory of Roman coins discovered in Bohemia in credible La Tène contexts (cf. chapter II.3) includes only pieces minted in the 2 nd and the very beginning of the 1 st century BC.Imports basically come to an end in the 80s BC.Asimilar situation can be observed in Bavaria (Kellner 1990; Ziegaus 2004) and was already described in the cemetery of Ornavasso where the coin series comes to an end in 82/81 BC, the decades between 70s–30s BC are each represented solely by asingle coin and only in the Augustan period does the series regain its consistency (Graue 1974, 135–144). Such ageographically general phenomenon invites discussion concerning the region of the coins’ origins rather than in the reception zone. In other words, the absence of Roman coins in Transalpine Europe need not tell us anything about the local ability to acquire them but rather about the ability of these coins to cross the Alps. Between 90 and 35 BC Italy was being torn apart in aseries of civil wars culminating in the late 80s (Sulla’sdictatorship) and the 40s (the terror of the 2nd triumvirate and struggles between the triumvirs themselves). This tense situation is reflected by Italian coin hoards concentrating in 90–75 and 50–40 BC (Backendorf 1998, Abb. 10). Hoards from the 70s–60s are almost unheard of and coins from this period are as arule under ‑represented in later hoards (e.g. Bilić 2012). Apart from the atmosphere of anxiety in atime of crisis this situation reflects also the economic and financial crisis raging in Italy between the Social War and Augustan period with the outcome of alack of currency.23 CHRONOLOGIES OF THE SECOND IRON AGE IN TEMPERATE EUROPE « … se in questo stesso istante io fossi sulla linea del meridiano sarebbe la mezzanotte in punto, ma se guardassi aoccidente vedrei la mezzanotte di venerdì e se guardassi aoriente vedrei le mezzanotte di giovedì. » Umberto Eco, L’isola del giorno prima. There is no shortage of local or regional Late Iron Age chronologies in Europe; however, things get complicated when we try to establish asingle universal scheme valid for the entire territo‑ ry of the La Tène Culture (from western France to the Carpathian Basin, from northern Italy to southern Poland), let alone for the neighbouring regions relevant for our inquiry (from Portugal to the Crimea, from Tuscany to the Baltic Sea). 23 Instructive in this sense is the wording (or at least the literary representation) of Pompey’sletter to the Senate from the Sertorian War of 74 BC in Sallust, Hist. ii, 82. 52 THINGS AND THOUGHTS The actual definition of chronological phases of the Late La Tène was first undertaken in southern Germany and was based principally on metal small finds. In Gaul on the other hand, the local chronologies were developed mostly on the grounds of settlement pottery. Unlike the Middle La Tène period, Late La Tène burials are available only in some regions (Middle Rhine ‑Main ‑Mosel region; northern Italy; Eastern Alps); in some of them the seriation of grave ‑goods is based more on pottery, in others on metal objects. Yet in all these regions the chronological phases defined on such disparate material are labelled in the same way. The correct approach is, in my opinion, to utilise the universal phase labels exclusively for de‑ scribing phases defined by universally widespread types of metal artefacts, i.e. mainly (if not exclusively) brooches. Local ceramic chronologies naturally have their validity for the region in which they were constructed and should be even preferred as ameans of dating at alocal or regional level; they must, however, be distinguished (ideally also in terminology) from the overarching chronology based on metal artefacts, which is the only common denominator for these floating and not necessarily mutually corresponding local chronologies. It seems illogical and inacceptable to claim for local chronologies the universal phase labels or to ‘correct’ the universal metal chronology by using local ceramic ones. Current La Tène research has (seemingly) come to auniversal agreement on using the Reinecke chronological terminology of Stufen (LT B1a, b, B2, C1a, b, C2, D1a, b, D2a, b).24 This agreement, nevertheless, disguises aprofound incongruity in the actual meaning of these dangerously uniform terms. Apart from causes rooted in the history of research this problem is due to the extent of the La Tène world being composed in reality of numerous local (sub‑) groups with locally specific means of answering chronological questions. Moreover, the truly universal types of artefacts connecting all of these groups and allowing for cross ‑dating are in reality much less frequent than we would believe. The chronology of the Middle La Tène period, based principally on the seriation of grave goods (e.g. Hodson 1968; Waldhauseretal. 1987; Gebhard 1989a) has been relatively stable and uncontroversial over the last few decades. Some corrections may be necessary in the future to the absolute chronology due to corrections to the dating of Mediterranean imports (Kavur– Blečić Kavur 2018) but these are details to be discussed when dealing with these objects. On the contrary, the chronology of the Late La Tène period in Europe has been subject to major polemics. As is generally known, the two systems of Late La Tène chronology cur‑ rently in use in Europe are what we can call the Bavarian or Eastern and the Rhineland or Western model. Though sharing the terminology of Reinecke they diverge in the meaning of these terms both in arelative and in an absolute sense. In brief,25 the Bavarian model was developed between the 1960s and the early 1990s, based on evidence from Manching, Basel, and Altenburg ‑Rheinau, but also Czech oppida (Krämer 1962; Waldhauser 1983; Fischer 1988). At first, LT D1, characteristic of the oppida period and accompanied by Nauheim type brooches, was distinguished from LT D2 characterised by geschweifte Fibeln which are not found on the oppida of Central Europe. Subsequently, LT D1 was further subdivided into the 24 The validity of the very terms of phase/Stufe has been recently questioned by John Collis (2009), criticising its box ‑like nature and suggesting replacing these notions with the concept of ahorizon. There are many valid points in his critique. Nevertheless, we will stick here to the phase ‑based chronology, first and foremost because in this chapter we are going to deal with the historiography of phase ‑based chronologies, secondly because in the following ones chronological issues will be only of auxilliary value and not worth further complicating the discussion. 25 For adetailed discussion on the history of research in Late La Tène chronology with special focus on the East ‑West controversy cf. principally Leifeld 2007 and Božič 2008; the chronology down to the mid‑1970s is well synthesised in Stork 2007. 53I. THE SETTINGS Nauheim horizon LT D1a and alater stage LT D1b characterised by the A65 and spoon ‑bow brooches. In absolute terms, the end of the oppida and thus the transition between LT D1b and LT D2 was set at about 50 BC (Christlein 1964; Waldhauser 1983); the beginning of the period was dated to roughly 120 BC.The higher Rhineland chronology was developed in the 1990s. First, Miron (1991; 1998), based on the seriation of grave ‑goods in the Main ‑Mosel region, defined a‘pre ‑Nauheim’ horizon of Late La Tène brooches claiming for it the label of LT D1a and confusingly renaming the following stages previously established in Bavarian chronol‑ ogy.26 He also raised the absolute chronology by about ageneration by amere arithmetical recalculation (and without any good archaeological reason) simply in order to make it fit his new system better. His proposal was endorsed and further developed by Jeannot Metzler and Sabine Rieckhoff (Metzleretal. 1991; Metzler 1995; Metzler– Metzler ‑Zens– Méniel et. al. 1999; Metzler– Gaengetal. 2009; Rieckhoff 1995). Rieckhoff’swork in particular is of interest for us since she tried to combine evidence from Gaul, Bavaria, non ‑La Tène northern Germany but also Bohemia. In acomplex and not entirely logical discussion based on archae‑ ological arguments and historical/narrative constructions she claimed the geschweifte Fibeln had appeared and Manching was abandoned (and Bavaria basically depopulated) already by the 80s BC.Her main external argument for such ahigh date was the chronology of amphorae found in Manching which were re ‑studied shortly before (cf. Lyding Will 1987).27 Over the following quarter century the higher chronology became the norm in the Rhineland and in French ‑speaking research (Kaenel 2008; Barral– Fichtl eds. 2012), though its advocates never managed28 to defend the system against the numerous criticisms of it as awhole and of its individual components.29 There is not enough space for adetailed discussion but in conclusion, Ido not find convincing arguments for using the high western chronology which Ifind methodologically flawed in terms of relative chronology30 and insufficiently grounded as far as absolute dates are concerned. LT C2, akey phase in the social and economic transformation of Transalpine Europe, is no longer documented by grave finds in Bohemia and Moravia though graves are still relatively 26 In this system thus the Nauheim brooch horizon became LT D1b, A65 and spoon ‑bow brooch characterise LT D2a, while the geschweifte Fibeln appear with LT D2b. 27 Though as we have seen above, amphorae are not the kind of material fit for fine chronological divisions. 28 Rieckhoff 2012 is strong on rhetoric but less so on arguments. 29 Cf. e.g. criticism of the Rieckhoff brooch chronology: Peschel 1999; Brandt 2001, 91–92; Bockius– Luczkiewicz 2004; criticism of the ‘pre ‑Nauheim horizon’ in the Rhineland: Krausse 2006, 108–118, 130–131; criticism of the idea of depopulation of Bavaria around 80 BC: Hüssen– Irlinger– Zanier eds. 2004; Schaich 1998, 167 with evidence of major constructions in the Viereckschanzen of Pocking Hartkirchen in around 50 BC; complex deconstructions: Fischer 1999; Meller 1999; Leifeld 2007; Božič 2008. 30 With the criticism of the idea of apre ‑Nauheim phase Ido not suggest that the Late La Tène brooches appeared as the fully developed Nauheim type; the simple filiform Late La Tène brooches may indeed pre ‑date these. Irather question the possibility of defining this as achronological phase on its own, and even less so the need to allot it an entire quarter century of duration. The simple filiform Late La Tène brooches continue appearing alongside the Nauheim type in its phase if not beyond. The main problem of the ‘pre ‑Nauheim horizon’ is its negative definition by the absence of Nauheim brooches which will necessarily end up by dating any simple non ‑Nauheim Late La Tène brooch to the earliest part of LT D1a and thus with an artificial raising of the dating. For all these reasons Ibelieve that dividing LT D1a into two phases is counter ‑productive. The Nauheim brooch remains avalid indicator of this phase though not necessarily of its beginning which is marked by appearance of Late La Tène scheme alongside long ‑lived Middle La Tène brooches. 54 THINGS AND THOUGHTS numerous in Bavaria. This transitional character of LT C2 makes its artefactual definition somewhat problematic as we will see below. The ‘leading fossils’ of LT C2 are certain types of brooches (sharing several characteristics such as ashort foot and an external chord), prin‑ cipally the Mötschwil brooch with thickened bow (Hodson 1968; Meduna 1970b; Márton 2004). The definition of LT C2 is quite blurry also in absolute terms. Objects characteristic of LT C1 were still in circulation in 229 and 208 BC (cf. La Tène and Wederath: Haffner 1979) while the first more or less precisely dated context to produce objects of LT C2 type31 is Mon‑ te Bibele destroyed in ca 190/180 BC.The dating artefacts from these sites are however from disparate categories and none of them belongs to the Mötschwil group of brooches which are decisive for the chronology of Central Europe. The transition LT C1/C2 in ca 200 BC or in the 190s seems plausible, but the vagueness of this definition shows clearly how far we are from being sure what this transition is actually supposed to mean. We are even more clueless as to the inner chronological articulation of this extremely significant but archaeologically elusive phase even though it seems to have lasted for some 60–80 years. As already said, as far as the Late La Tène period is concerned, Ifavour the Bavarian chro‑ nology not only because in terms of relative chronology the distinction between the ‘oppida‑’ (LT D1a–b) and ‘post ‑oppida’ (LT D2) periods describes the situation in Bohemia handily and pertinently. As mentioned above (note 30), Iadmit the existence of apre ‑Nauheim horizon in LT D1, though Irefuse to recognize it as arelative ‑chronological phase on its own; Iwill be more inclined to believe that once we are able to distinguish alate phase of LT C2 correspond‑ ing in many traits to early LT D1 and beginning sometime after the mid‑2nd century BC.The actual Late La Tène period, i.e. LT D began in my opinion in ca 130/120 BC. 32 LT D1 can be further subdivided into two sub ‑phases: LT D1a accompanied by the Nauheim brooch which however does not necessarily mark its beginning (that is the job of an enclosed catch ‑piece, no matter on what type of brooch it is present) nor does it naturally disappear with its end. 33 LT D1b is defined by the Almgren 65 and spoon ‑bow brooches (which again may not have appeared simultaneously). The transition between LT D1a and LT D1b occurred in my opinion in the 70s BC based on the evidence from Treviglio in northern Italy (discussed by De Marinis 1986; Demetz 1999; Božič 2008). It is worth pointing out that the excavations of Caesar’ssiege ‑works from 52 BC at Alesia produced aseries of Nauheim and three spoon ‑bow brooches besides, naturally, the Alesia type, this latter lying already on the transition to the following phase (Brouquier ‑Reddé– Deyber– Sievers 2001, 297–298, pl. 90: 43–45, 90: 37 and 91: 48, 55–61). Apart from some isolated and meagre indices, LT D2, defined mainly with the geschweifte Fibel (as well as the Alesia, A238, and Jezerine brooches) is not attested in the oppida in Bohemia, Moravia or Bavaria, while it is very well represented in the oppidum of Bratislava where the geschweifte Fibel is the best attested type (Čambaletal. 2015). Outside our working area the earliest La Tène Culture find contexts of geschweifte Fibeln date to around the middle of the 1st century BC: At Besançon ‑Parking de la Mairie, they are dated to ‘pre‑40 BC’ and 40–30 BC (dates established by dendrochronology; Feugère 1992). The spread of the geschweifte Fibeln in Central Europe (leaving aside its evolution in the area of central Germany) therefore seems 31 Glass bracelets: Orfanelli– Vellani 1992, 5, n° 51 and 52. 32 The earliest Nauheim brooches date to post 120 BC (Fellbach ‑Schmiden: Planck 1980; Lombardy: De Marinis 1986, 139, note 150) and the allowance for the pre ‑Nauheim horizon need not be more than adecade or so, certainly not in Central Europe. 33 E.g. in Besançon (Feugère 1992), or in the Roman military camp of Lomba do Canho in Arganil, Portugal founded probably during the Caesarean campaigns in the region during his propraetorship (62/61 BC) with potential continuity till the battle of Munda in 45 BC.The finds include two Nauheim brooches (De Castro Nunes 1959) and amphorae classified as of Dr 1B type (Fabião 1989). 55I. THE SETTINGS clearly attested only in the third quarter of the first century; there are no definite finds earlier than the middle of the century. The beginning of LT D2 can therefore be dated around or only shortly before the middle of the 1st century BC.Its internal development and its end date are not relevant for our study. Fig. 6: Chronologies of LT C–LT D after various scholars. CHRONOLOGIES OF THE MIDDLE AND LATE IRON AGE IN BOHEMIA For apart of the 20th century, the chronological framework of the La Tène period in Bohemia was Jan Filip’smodel (Filip 1956), conceiving chronological phases more as historical epochs and expressing the chronology exclusively in absolute dates (very low ones from the current point of view). In this vision, the phases of flat burials and of the oppida were considered very close if not overlapping. This situation started changing in the 1970s mostly under the influence of scholars trained in the Brno University and adhering to the Reinecke system. Jiří Meduna (1970b) defined the phase LT C2 on the basis of finds from Staré Hradisko. Miloš Čižmář (1970; 1976) proposed arelative chronology for Moravian burials. Both these scholars assumed that the cemeteries came to an end before the beginning of LT C2 within which Staré Hradisko was founded. Another Brno alumnus, Jiří Waldhauser, refined the chronologies of both the flat graves (LT B1–C1b; Waldhauseretal. 1987) and the oppida periods (Waldhauser 1983). In the latter he distinguished earlier and later phases, LT D1a and D1b respectively, and argued for raising the date of the end of the Bohemian oppida from the time of Maroboduus to the mid‑1st century BC. 62 THINGS AND THOUGHTS Fig. 9: Chronology of occupation in major oppida in Central Europe based on brooch type ratios. Danielisová– Militký 2014, obr.7. (The authors use the Rhineland terminology, thus their LT D1 = our LT D1a, LT D2 = our LT D1b). The same pattern can also be discerned on other Czech oppida (Fig. 9; Danielisová– Militký 2014). LT D1a is always the best represented period; LT C2 is systematically present although in Stradonice and Třísov it is represented by only afew examples. Hrazany and especially Staré Hradisko are relatively richer in these early brooches. In all the studied sites we observe asteep drop in brooches in LT D1b, though occasional LT D2 pieces can be found in Třísov and Stra‑ donice. Manching with its extremely strong (pre ‑oppidum phase) LT C2 and extremely poor LT D1b corresponds therefore more closely with the occupation dynamics of Staré Hradisko and Závist rather than with Stradonice and Třísov. The sudden rise of the oppidum of Bratislava into ashort ‑lived regional centre occurred in the period of the spoon ‑bow brooch and lasted mainly through the period of Almgren 18/ Alesia/ Almgren 238 (Čambaletal. 2015). An outline of the absolute chronology is provided by the numismatics as the local silver Biatec coinage imitates Roman denarii from 70/69–46/45 BC (Göbl 1994; Militký – Torbágyi forthcoming). The bulk of the amphora assemblage from the Bratislava Castle seems to date to roughly the middle or the third quarter of the 1st century BC (Kysela– Olmer 2014). Imported tableware includes black gloss ware, (and single pieces of) pre ‑sigillata, thin ‑walled pottery and earliest sigillata, testifying to the settlement continuing into the 30s or early 20s BC.An important role in outlining the Bratislava chronology had always been assigned to the supposed destruction of the oppidum by the Dacians, believed to have oc‑ curred sometime between 45 (the latest Roman coin imitated in Bratislava coinage) and 35 BC (by 63I. THE SETTINGS which date Burebista who was accredited with wiping out Celtic power in the Middle Danube area had already died). Research of the last decade has significantly down ‑scaled the significance of this event (Vrtel 2015) and some form of continuity beyond the 40s or 30s BC is not excluded. The pre ‑oppida period The chronology of the Late La Tène period, although subject to lively discussion, is extremely rich in data. The situation is quite the opposite in the misty period leading to the foundation of the oppida. As already mentioned in the previous chapter, the phase LT C1 marks amajor change in the evidence base in Bohemia and in Central Europe. Archaeologically detectable burials gradually disappeared over LT C1 while it is only sometime in LT C2 that the oppida came to existence. In spite of our gradually increasing knowledge about the LT C1–C2 agglom‑ erations, no closed contexts enabling artefact seriation are known from this period, and an entire artefact phase consisting of objects no longer represented in the burials but present in the agglomerations or in the (earliest phase?) of the oppida seems to be floating in avoid. In this way a‘gap’ of unknown duration seems to be yawning between these two well studied and well systematized blocks of data of burials on the one hand and of the oppida on the other. Several agglomerations are known from our working area, but only from Němčice nad Hanou is there enough published evidence to allow adiscussion of chronology (Kolníková 2012; Venclová 2016; Čižmář– Čižmářová– Meduna 2018). The occupation of the site may have begun in LT B2 but the combined chronology of glass ornaments and brooches point to the earliest signs of production activities in LT C1a, asettlement peak in LT C1b and LT C2 with possibly some continuity of human presence, albeit on an extremely reduced scale in LT C2/D1. Including the floruit and abandonment of Němčice and at the same time the beginning of occupation at Závist as well as at Staré Hradisko, LT C2 becomes ahighly event ‑loaded phase. Archaeological evidence proved insufficient to help us understand the correct order of these events as the Mötschwil brooches characterising the occupation of late Němčice and early Staré Hradisko and Závist do not allow for any finer chronology.39 In order to try to better un‑ derstand these issues we will turn to another find category whose study has recently enjoyed aparticularly lively development, the coins. Numismatic chronology Coins have already been discussed to some extent but they cannot be omitted here for their upmost importance providing answers to chronological questions. Not only are they extremely numerous but especially, unlike any other artefacts, they create an independent self ‑contained system with its own inner logic (gradual loss of weight; in some cases debasement of metal purity over time; and the barbarisation of the original Mediterranean iconographic models). Recent advances in the seriation of the Bohemian and Moravian coinage (Militký 2015a; 2018a; Smělý 2017) help to describe in detailed their internal development; what we lack however is asynchronisation of numismatic and archaeological chronologies. So far, we are capable of attaching one specific coinage with abroad chronological phase, mainly thanks to key sites. In this way we can clearly distinguish the ‘Němčice’ and ‘Stradonice’ chronological phases (corresponding roughly to settlement peaks at these sites in LT C and LT D1) and even outline afairly detailed development within them. However, we are unable to tell for example which 39 We can nonetheless be sure that the repeatedly proposed idea of an abandonment of Němčice and a‘retreat’ to Staré Hradisko as areaction to the campaign of the Cimbri has no material justification. In the 110s BC Němčice must have already been abandoned for several decades with some residual population at the best. 64 THINGS AND THOUGHTS stage of the Stradonice coinage corresponds precisely with LT D1a and which with LT D1b. Of particular value therefore are moments of transition from one numismatic phase to another and acomparison of different sites which help us recognise absences. In this way, Smělý (2017, 69) argued that Staré Hradisko came into existence in the years of decline (but still existence) of Němčice nad Hanou though suggesting aconsiderable chronological gap between the end of Němčice and the beginnings of large ‑scale coin production in Stradonice (Smělý 2020). This would correspond with the delay with which Stradonice appeared on the scene according to the brooch chronology. Militký on the other hand considers the coins filling in this gap to have been minted in Stradonice (Militký 2015a, 168; Militký 2017, 117–118). What is sure is that there was no significant chronological overlap between Němčice and the Czech oppida but at the most ashort symbiosis if not agap. At the other end of the oppida period, coinage clearly corroborates the continuity of Stradonice into LT D2 (only vaguely suggested by brooches) and synchronicity of its latest phase (admittedly one of deep crisis) with the rise of Bratislava (Militký 2015a, 168 ; Milit‑ ký – Torbágyi forthcoming). THE CHRONOLOGICAL FRAMEWORK– SOME CONCLUSIONS The term ‘recent La Tène period’ denotes in our usage the period LT C1/C2–LT D2. The term ‘Late La Tène period’ traditionally describes LT D1–D2. In relative chronology Ifollow the Bavarian system assigned absolute dates as follows: LT C2 = 200/190–130/120; LT D1a = 130/120–ca 70s; LT D1b = ca 70s–ca 50; LT D2 = from ca 50. Fig. 10: Chronological overview. 65I. THE SETTINGS The term ‘pre ‑oppida period’ naturally covers the time ‑span preceding the establishment of the oppida, both the period of flat graves in LT B–C1 and of agglomerations in LT C1b–LT C2. The term ‘oppida period’ describes in Bohemia (but also in EnCE) the advanced stages of LT C2 and LT D1. The latest phase consistently attested in the oppida throughout our working area is LT D1b. The extremely meagre evidence of LT D2 shows that Stradonice and Třísov may have survived until the beginning of this stage, probably on amuch reduced scale. In Bratislava on the contrary LT D2 is the peak period of occupation. In terms of historical chronology, the oppida in Bohemia may have lasted till the time of the Gallic War and it is in these same years that Bratislava rose, lasting to the time of the Triumvirate or perhaps the earliest Augustan period. By that time the oppida in Bohemia and Moravia had long been abandoned and the traces of La Tène Culture were gradually disappearing in the beginning (‘Germanic’) Roman Iron Age (RIA). II. THE THINGS AND THE THOUGHTS 69II. THE THINGS AND THE THOUGHTS 1. Introduction In view of the almost complete absence of written sources, all information on contacts be‑ tween the Mediterranean and Central Europe is provided by the ‘contact indicators’ as defined by Gilles Pierrevelcin (2009; 2012, 59–66); the same author also presents asolid theoretical framework for the study of long ‑distance contacts. At this point it is therefore only necessary to point out the features in which the present study will– by virtue of different material– of necessity depart from this framework. First, the present study is intended as asymmetrical: in most cases, the contacts will be studied in only one direction through the southern indicators in the Transalpine Central Eu‑ ropean milieu. This asymmetry is largely justified by the aims of the study (we are interested in the local effects of the contacts); at the same time, identification of specifically Bohemian finds in the Mediterranean, that is distinguishing them from the bulk of La Tène finds may prove extremely difficult, as will be demonstrated in ashort chapter dedicated to that issue at the very end of this section. The ‘contact indicators’ are divided into two large categories: we are first dealing with the material ‘imports’, i.e. artefacts (chapter II.2) including the somewhat special category of coins (chapter II.3) and subsequently (chapter II.4) with the much less clearly defined spiritual or immaterial imports as defined by Natalie Venclová (2002). As already explained, Bohemia will be systematically studied in comparison with two neighbouring regions– WnCE and EnCE– in order to properly contextualise the Bohemian results. Although the core of the study will be the Late Iron Age, i.e. oppida period, the much scarcer contacts in the 4th–3rd century will be studied with the due diligence and the two pe‑ riods will already be differentiated in this analytical chapter. The basis of the study are the contact indicators, single objects (or fragments of objects) which Istrove to collect and document as fully as possible. The analysis will proceed by find categories: metal vessels, ornaments, pottery, etc. The core of the study is formed by already published objects, including those published as contact indicators, regardless if this interpre‑ tation is tenable or not. Particular attention will be paid to cases in which the artefacts were published as imports (both material and immaterial) without substantiation and we will be ready at any point to dismiss such dubious artefacts. Unpublished objects can be included thanks to numerous colleagues who have provided me with valuable information on such pieces, granted me access to various public or private collections where Icould identify so far unrecognized objects, and who have given me per‑ mission to publish them. Isuppose that some published pieces have escaped my attention, in particular those from the ‘bibliographically more distant’ regions like southern Germany or Austria. Ihope nevertheless that the corpus presented here, though not necessarily complete, is at least representative of the phenomena Iintend to investigate. The value of the single finds and find categories may vary, and the depth with which they will be dealt will therefore vary accordingly. The length to which the finds or categories will be discussed need not always directly reflect their significance within the mass of imports… Ido not mind running this risk. 70 THINGS AND THOUGHTS THE CORPUS OF THE IMPORTS The word ‘import’ is intended here as an artefact deposited in adifferent geographical and cultural environment than the one in which it had been created, in our case therefore objects made in the Mediterranean and found in Central Europe (or more rarely vice versa). Moreover, in order for the object to be taken into consideration, its production, transfer, and deposition (or at least two of these) had to occur within the chronological timespan studied here (acon‑ dition which is not always met and not always easy to verify). The word ‘import’ is therefore used here in apurely neutral way, without any cultural or economic significance, e.g. that implied by its present ‑day commercial value (cf. Pierrevelcin 2012, 35–37 for discussion on modalities of import transfer). The Mediterranean imports found in the study area are presented in the Catalogue in Appendix II in geographical order (region > site > functional category). Throughout the study, each object is referred to by alphanumerical codes in square brackets consisting of site ab‑ breviation and object number. In order to facilitate the orientation in the Catalogue and to leave space for possible later entries, Itried to avoid continuous numerical lines in the richer assemblages: therefore, in the richest sites like Manching and Stradonice the first number refers to find category (000–199 = bronze vessels; 200–299 = glass vessels; 300–399 = mirrors; 400–499 = jewellery and finger rings; 500–599 = medical, toilet and writing instruments; 600– 699 = pottery; 700–799 = amphorae; 800–899 = ecofacts; 900–999 = other).40 Thus e.g. acode [Zá12] corresponds to artefact number 12 in Závist whereas [S412] refers to artefact number 12 in the category of jewellery and finger rings (400–499) at Stradonice. All abbreviations are explained in detail in the introduction to the Catalogue, AppendixII. Coins (Chapter II:3) are catalogued separately in AppendixIII. As mentioned above, the corpus also includes objects not considered to be actual imports for one reason or another, often those which were incorrectly published as imports in the past. These objects and the reasons for their exclusion from our consideration are discussed within the single import categories in which they (in reality do not) belong. These rejected non ‑imports are catalogued in Appendix II too, and identified by an ‘x’ within their individual code (e.g. [Sx02] = second disproved object in Stradonice). Bibliographical references to the individual objects are as arule relegated to the Catalogue in order not to overload the text itself. For the same reason, authorship of the drawings of the catalogued objects are presented in Credits at the end of the book rather than in the individual captions. For all the used Institution abbreviations cf. the list in pages 331–332. In the final evaluation of the corpus, in the majority of categories each entry, albeit only afragment, will be counted as one individual object. In other words, Iwill not employ any method of quantification. One exception to this rule will be pottery and amphorae on the one hand and mirrors on the other. In both cases the nature of these categories runs the risk of excessively overestimating their importance if counted by asimple number of fragments; therefore pottery, amphorae and mirrors will be quantified using the minimum number of individuals (MNI). 40 Such an approach is adopted (albeit not so rigorously) also in other sites, such as e.g. in Třísov, where new categories sometime start from the next ten up. 71II. THE THINGS AND THE THOUGHTS PROVENANCE OF THE ARTEFACTS The sites which have yielded artefacts of Mediterranean origin are relatively few. In the Cat‑ alogue in Appendix II the overview of artefacts is listed according to site, each of which is introduced by ashort paragraph describing the most important aspects of the site itself. At this point, as apreamble to the actual artefact study, it is necessary to present several important facts concerning the ways and circumstances by and under which the collections of these objects were constituted which will help us better understand the sometimes very particular nature of some of these assemblages. Stradonice old collections The greater part of the objects with which we will deal in Bohemia are finds and ‘finds’ from Stradonice. The problems connected with the extremely rich Stradonice collection are gen‑ erally known (Sklenář 2015; Hlava 2015b): the collections were accumulated mostly during the ‘gold rush’ in 1877 and 1878 when dozens if not hundreds of local inhabitants pillaged the site in order to sell the finds to collectors (Fig. 11). The Stradonice collections consist of (tens of?) thousands of artefacts often absolutely extraordinary or unique in aregional if not aEu‑ ropean context. However, basically none of these objects has documented find circumstances (controlled excavations on the site– cf. below– have been few and generally unproductive). Even the provenance ‘from Stradonice’ is often questionable as the earliest Stradonice col‑ lections were already contaminated with fakes, and ‘Stradonice’ became in its time simply abrand name used freely by dealers with antiquities to embellish their merchandise. Each confrontation with artefacts allegedly from Stradonice therefore requires aparticularly subtle critical approach. The Stradonice finds are currently dispersed all over Central Europe; there is however ahandful of collections which constitute the vast majority of them. The Berger collection. Štěpán Berger (1844–1897) was alawyer, landowner, and politician who assembled during his lifetime an enormous collection of antiquities from Bohemia and abroad. The largest portion of Berger’scollection consisted of finds from Stradonice (less than 10 km as the crow flies from his residence in Svatý Jan pod Skalou). Berger (or his agents) was present on the spot in Stradonice during the ‘gold rush’, buying antiquities directly from the excavators. He knowingly did not distinguish actual finds from fakes, partly from ignorance, but at least partly as an act of beneficence towards their precarious creators. It is not known to what extent Berger purchased Stradonice antiquities outside the actual excavations from antiquities dealers. It is very probable that should the opportunity have arisen, he would surely have done so. He was certainly an avid buyer of antiquities from many sources as confirmed by other parts of his collection, including e.g. finds from Hallstatt. After Berger’sdeath, the collection was after some negotiations transferred to the National Museum in Prague in 1898; apreliminary catalogue (‘taxation list’) was created at this occasion though with numerous flaws and imprecisions (Hlava 2015b). Serious cataloguing of the collection started only in 1913, numbering at that time over 22,000 objects (Hlava 2015b, 151). It may never have been systematically carried out since some Stradonice finds from the Berger collection were only inventoried in the 2000s. The Grosse collection. In the 1870s, Wilhelm Grosse was amanager of asmelting plant in Nová Huť uNižbora, and therefore an immediate neighbour of the Stradonice pillagers with the easiest access to them. This is well reflected in his collection. Although much smaller than Berger’s(the early 20 th century inventory lists some 680 items: Collection Grosse n.d.), it is composed of obviously selected quality pieces. The collection came to the Vienna Hofmuseum, today the Naturhistorisches Museum Wien. 78 THINGS AND THOUGHTS afew finds come from primary contexts which could suggest to what use the vessels were put and although obsequence to Mediterranean wine consumption habits cannot be reasonably assumed in Transalpine Europe in the first place, these few finds hint that the vessels were employed there for practices different from those for which they were originally intended. BRONZE VESSELS IN CENTRAL EUROPE OF THE PRE ‑OPPIDA PERIOD Before delving into the voluminous topic of Late Republican bronze vessels, an overview of the much less numerous but more disparate 4th–3rd century material is necessary. In Bohemia, this period is represented by asingle vessel, abronze basin from the hoard of Stebno ‑Nouze [Stb1]. The basin has arounded bottom, out ‑turned rim decorated with an egg ‑and ‑dart frieze and asingle horizontal handle with an attachment depicting agrinning human or monstrous face (Fig. 13). Fig. 13: Stebno ‑Nouze, Etruscan bronze basin [Stb1]. 79II. THE THINGS AND THE THOUGHTS Basins of this type (misleadingly called teglie = baking pans) are relatively common in Hellenistic central Italy (Liepmann 1981; Cianferoni 1992; Bini– Caramella– Buccioli 1995, 163–185; Esposito 2007; 2010). As far as the evidence goes, their relatively large typological variability does not reflect any chronological or regional differences (Esposito 2010, 15–17).42 Anumber of such basins have been discovered in closed contexts– most often central Italian elite graves– dated to the last decades of the 4th and early 3rd century BC.As to their function, the basins most probably originally served for ablutions during feasts. The decoration on the attachment from Stebno is somewhat unusual representing ap‑ parently the face of Gorgo or of one of the unidentified Etruscan demons whose depiction is inspired by that of Gorgo (Paschiger 1992, 180–191). In Etruria, representations of Gorgo or gorgoneion are relatively common (Krauskopf 1988) and follow the same development as in Greece– from the monstrous gorgoniea of the Archaic period to the Hellenistic type of ‘beautiful Gorgo’, i.e. the face of ayoung woman occasionally recognisable at the most for who she is by her snake hair and with apathetic Hellenistic expression rather than monstrous. The being on the Stebno attachment belongs to the transitional type– lacking any monstrous attributes (e.g. claws) and depicted in anaturalistic way but still grinning with wide open eyes and tongue stuck out. This transitional type is ashort ‑lived phenom‑ enon dated to the late 4th and early 3rd century BC (Krauskopf 1988, 343–344). No matter of what type, agorgoneion is an unusual motif on basin attachments which are (regardless of the basin’stypological variety) most often decorated with apalmette or in some cases with entire animal figures or very rare mythological scenes (for acomplete overview cf. Kyse‑ laetal. 2017, 82). The only two other instances of gorgoneion on ateglia attachment known to me come from Tarquinia ‑Monterozzi, t. 842/1593 (used between the late 4th century and the Augustan period: Cavagnano Vanoni 1996, 82, fig. 18:163, tav. lviii: a) and from an un‑ known site (presumably in southern Etruria), currently kept in the ex ‑Falcioni collection in the Vatican (Caliò 2000, 219–220, n° 393).43 On the Tarquinia piece the gorgoneion is of the Archaic (in this case ‘archaising’) variety and of rather poor workmanship. Its connection with the Stebno piece seems rather accidental. The vessel in the Vatican seems to be amuch closer parallel but the depiction is not well preserved. The distribution of these vessels (the lists presented by Esposito 2007; 2010 need to be complemented by finds listed by Liepmann 1981 and Cianferoni 1992) includes Etruria (where they were certainly produced) as well as the Marche and Emilia where they are curiously found in La Tène Culture graves.44 The northernmost find (?) so far has been an 42 For completeness sake, the teglia from Stebno corresponds to the var. B1 after Caramella (Bini– Caramella– Buccioli 1995, 179–181) or IIb1 after Esposito (2010; or type 20 212b after Esposito 2007). 43 Also in the imitations of this vessel shape in pottery apalmette is the prevalent decoration. Cf. black ‑gloss pottery: Morel 1981, 396, pl. 195, genre 6300; ‘ceramica argentata’(tin ‑coated); the teglie produced in Orvieto are only decorated with palmettes while on smaller strainers of the same shape, the attachments bear figural scenes (Michetti 2003, 200–202, fig. 24–25, tav. lxxiii–lxxiv); lack of decoration is characteristic of the same shapes in the Volterra production (Michetti 2003, 209–211, fig. 26); it is only on Faliscan examples that faces occasionally appear, on the attachments of both teglie and strainers beside the customary vegetal motifs: aSilenus face on ateglia from Civita Castellana (Michetti 2003, 238, n° 510, tav. cxx), and afemale head on astrainer in Rome (Michetti 2003, 239, n° 517, fig. 38). 44 Montefortino: Brizio 1899, 688, tav. x: 11; Filottrano: Baumgärtel 1937, 256–259, pl. xxiv: 4–5; Bologna ‑Benacci, gr. 138 and 953: Vitali 1992, 138, tav. 15: 2, 291, tav. 34: 10; Monte Bibele ‑Monte Tamburino, gr. 132: Vitali ed. 2003, 421–422, tav. 228: 10. 80 THINGS AND THOUGHTS attachment kept in Trento (Marzatico 1997, 358–359, tav. 82). 45 No certain Transalpine finds are known to me.46 The basin from Stebno ‑Nouze is remarkable also in its find circumstances: it formed part of asmall hoard consisting (apart from the basin) of aset of bronze ornaments and aceramic bowl, the whole deposited probably sometime in LT B2 or LT C1 in apart of Bohemia which was not inhabited at that time. During anon ‑systematic study of the collection of Němčice nad Hanou kept in the Moravian Museum in Brno47 two objects of Mediterranean origin have been identified, both apparently fragments of asitula: ahandle termination with abaluster ‑shaped finial [NH01] and acast bronze arched foot with lateral appendices [NH02]. Both these fragments can be convincingly linked with Hellenistic situlae (Fig. 14) There are two basic forms of bronze situla in the Late Classical and Hellenistic periods relevant to our inquiry (Fig. 15A) 48 : the simple bell ‑shaped buckets with slightly rounded profiles, and spouted situlae in which the handle attachments are decorated, one with ahuman 45 Marzatico argues for an authentic local discovery although the find circumstances are not recorded. 46 Three dubious handles are recorded: one from the Rodenbach mound in Hessen, one said to be from the Rhineland in the Kestner Museum in Hannover, and one documented only by adrawing from the Hrdina collection in Prague. In the latter case adetailed analysis of the archive documentation leads us to believe that none of the Mediterranean imports from the collection was necessarily agenuine find (Kysela– Hlava 2014). The Hannover handle is avery similar case– acollection sold en bloc by the collector’sheirs with presumable loss and confusion of information. The Rodenbach handle is the only one from a(partially documented) excavation though its find context dates clearly to the 5th century with no hints at another later burial at the same place (Engels 1972). Liepmann (1981, 17) correctly realised that the style of the palmette on the Rodenbach handle clearly differs from that on other basins and stands closer to the palmettes on 5th century jugs. Also, the handle from Trento (cf. above) and one attachment kept in the Falcioni collection (Caliò 2000, 221, n° 394) are similar in style. This point certainly needs further research. The Rodenbach handle in any case is not avalid parallel for the Stebno piece. 47 Iam deeply grateful to Dr.Jana Čižmářová who granted me access to the unpublished collection and who allowed me to study the finds and to make them public. 48 For more detail cf. Kysela 2020. Fig. 14: Němčice nad Hanou, fragments of Hellenistic situlae. 81II. THE THINGS AND THE THOUGHTS Fig. 15: Hellenistic situlae. A– principal types. 1– ovoid situla, 2– spouted situla, 3– kalathos. B– spouted situlae. Examples of the decoration of the backside attachments. Drawing, J. Kysela. 82 THINGS AND THOUGHTS or divine head or bust and the other with aspout shaped like aleonine head.49 Less important for us are the so ‑called kalathoi, or ‘baskets’ with concave profiles which continue to be pro‑ duced in the Hellenistic period. All these types share the same arrangement of handle– double swinging handles held in apair of spectacle ‑shaped attachments. On the other hand, only the spouted situlae and the kalathoi appear to have been provided with three arched feet soldered on to their base (Barr Sharrar 2000). These are usually in the shape of an ox ‑hide with small appendices on the inner pair of stubs. What both principal situla types have in common is aconsiderable uncertainty about their production area, Macedonia, Etruria, and Tarentum being the usual contenders. In both cases the most probable scenario is that of Macedonian models subsequently partly imitated in (among others) Etruscan workshops (cf. mainly Blečić Kavur 2010, 285–307; Blečić Kavur 2012 and the discussion below). The handle end [NH01] with carefully formed flattened sides of the shaft and quite strongly profiled baluster ‑shaped ends corresponds particularly well with the terminals on situlae of the late Classical and Hellenistic period (discussed in detail below). In the handle terminals of the Late Republican period situlae, the balusters are less strongly profiled and the shafts are rounded in section. Numerous parallels can be cited among Hellenistic situlae of all types, both from Greece and Italy.50 In the case of the foot from Němčice nad Hanou [NH02] it is worth noting that its shape with small appendices on the stubs is dissimilar from Late Republican shapes (cf. below51) and identical with feet on spouted situlae in bronze and silver (cf. Zimi 2011, 196, fig. 28)52 as well as to those present in the remarkable assemblage from Nesactium in Istria discussed below (Mihovilić 2017, pl. 4: 12, 14). In conclusion it seems that at least one Hellenistic situla, most probably of the spouted type, was present at Němčice nad Hanou. In order to correctly understand or contextualise the finds from Němčice (but also the Steb‑ no basin discussed above), we are obliged to step back, out of the limits of our working area, and have alook at the situation in the whole of broader Central Europe and its neighbouring areas. Only afew vessels can be cited for the 4th–3rd century from the area between the Rhine and the Carpathian Basin, the majority of which are well known and only ashort mention will suffice (Figs. 16, 17). 49 Bell ‑shaped situlae = type F after Giuliani Pomes 1957, type Aafter Zahlhaas 1971a; spouted situlae = type C after Zahlhaas 1971a. 50 As acompletely random selection of parallels: e.g. from Populonia (Cianferoni 1992, spouted situla 15–17, fig. 4–5, bell ‑shaped situla fig. 8–9), Tuscania, tomba Curunaś II/23 (Moretti– Sgubini Moretti 1983, 101, tav. ci– deposited perhaps in the 2nd/1st century BC), Karlsruhe Museum (Jurgeit 1999, spouted situla 327–328, Nr. 537, bell ‑shaped situla 332–333, Nr. 543), Cetamura del Chianti (de Grummond ed. 2017, 118, fig. 77), Thrace (Sideris 2016, 198–223, nos 78, 81, 84, etc.). 51 One similar foot is present in the Stradonice collection [S051] and one is known from Manching [M027]. Though formally identical with the piece from Němčice, the chronology of Stradonice is naturally much too late for the circulation of this vessel type. If the Stradonice foot actually comes from aHellenistic situla, it must have reached the site as an antiquity or mere scrap. In the case of the Manching piece, the odds are much higher that it could actually come from avessel present in the site already in the 3rd century BC.Considering these incertitudes, both feet will be considered to be of oppida period. 52 No bell ‑shaped situlae with feet are known to me (their base is either simply flat, or it is shaped into aprofiled ring foot; in some cases, it is stabilised with alayer of lead poured into it). However, this may only be due to insufficient evidence (the situlae in general are few and only the upper parts of vessels are sometimes preserved) rather than the actual state of things; Iprefer not to be too dogmatic on this point. 83II. THE THINGS AND THE THOUGHTS Fig. 16: Hellenistic bronze vessels in Central Europe. Hu– Hurbanovo (Bujna 2007); Bg ‑K– Beograd ‑Karaburma (Blečić Kavur– Kavur 2010); Sz– Szob (Tankó 2014); Mnd– Mannersdorf (Ramsl 2011); Wdg– Waldalgesheim (Joachim 1995). Not to scale. 84 THINGS AND THOUGHTS Fig. 17: Hellenistic bronze vessels in Central Europe and other finds mentioned in the text. The only region in which 4th–3rd century Mediterranean metal vessels appear with some con‑ sistency is the Carpathian Basin (cf. arecent overview in Repka 2015, 110–117). One bronze lekythos was discovered in tomb 18/64 in Hurbanovo, okr. Komárno, SK (Benadík 1981; Bou‑ zek 2002; Bujna 2007). It belongs to the Talcott type common in Greek contexts from the late 4 th –first half of the 3 rd century BC (for an overview of the type in various materials cf. Zimi 2011, 46–47). The grave itself contains artefacts of LT B2/C1a date, so there is no serious discrepancy between them. Two bronze drinking vessels were discovered in tombs at Szabolcs (askyphos with annular handles; Szabó 2000/2008, 80, fig. 2) and Szob (akantharos; Szabó 2000/2008, 80, fig. 3; Tankó 2014). Parallels to the Szob piece date to third quarter of the 4th century (e.g. Sideris 2016, 266–267). The Szabolcs cup is somewhat uncommon in metal but vessels of the same form in pottery date from the second quarter to the end of the 4th century BC with some continuation into the 3rd century (Sparkes– Talcott 1970, 123, fig. 7, nos. 720–721; Rotroff 1997, 90–91, fig. 10, nos. 102–109, and esp. 114). In all these cases, the origin of these vessels is unmistakably Greek and they have been commonly connected with the movement of Celtic armies during (and in the aftermath of) the campaign in Thrace and Greece in 280–278 BC (e.g. Szabó 2008; Repka 2015) although in reality they could have arrived and probably did arrive in the Carpathian Basin well before 85II. THE THINGS AND THE THOUGHTS this date, already at the end of the 4th or very beginning of the 3rd century BC (Kavur– Blečić Kavur 2018). The single category of bronze vessels documented with some consistency not only in the Carpathian Basin but also further west are situlae, as further confirmed by the finds from Němčice nad Hanou. On the westernmost fringe of the Carpathian Basin in the necropolis of Mannersdorf am Leithagebirge, abronze kalathos was found in aLT B1b gr. 13 (Ramsl 2011, 155, 199, 249–251 passim, Abb. 124/A, Taf. 63). The necropolis spanning the period from LT Ato LT B2/C1 with apeak in LT B1 has produced evidence of far ‑flung long ‑distance contacts (Ramsl 2011 passim; Ramsl 2014) from eastern France (the openwork decoration of the weaponry: Gi‑ noux– Ramsl 2014) to the Carpathian Basin and the Balkans (an omega ‑shaped double pin of Illyrian origin; an amphoriskos ‑shaped glass pendant cf. below). Peter Ramsl correctly classified the situla as type D after Giuliani Pomes (1957; cf also Pellegris 2004; Montanaro 2015, 74–77) and pertinently observed the chronological dissonance between find contexts of this type in Italy (5th/early 4th century BC) and the Mannersdorf grave (late 4th century) in which it was deposited as an Altstück (‘heirloom’). It does not, however, answer the question in which of the two time horizons the bucket crossed the Alps– was it one of many bronze vessels traded between the Felsinean and LT Acultural areas or was it brought back in the turbulent period following the Celtic invasion of Italy? The easternmost regions of the Early La Tène Culture seem disinterested in the importation of Etruscan vessels (though the picture may be skewed: cf. Goláňová in print). At the same time, we should keep in mind the high degree of connectivity manifested by the community burying their dead at the Mannersdorf cemetery including contacts with Italy (azoomorphic brooch and pottery both originating in the Fritzens ‑Sanzeno Culture of the Trentino ‑Alto Adige region: Ramsl 2014; Roncador 2016, 176–181). For all these reasons, the idea that the Mannersdorf bucket was not akeimelion treasured for three generations by Transalpine hoarders but that it arrived in Central Europe only in the 4th century BC, does not seem to be pure fantasy. Two complete bell ‑shaped Hellenistic situlae are known in the Carpathian Basin, one from grave 22 in the cemetery of Belgrade ‑Karaburma and one dredged from the Danube in Budapest.53 Both correspond to Macedonian types of the late 4th–early 3rd century BC (Blečić Kavur– Kavur 2010; Blečić Kavur 2010 with further references). The Karaburma tomb moreover contained abronze ribbed phiale; its origin can also be looked for in northern Greece. Based on the analysis of the grave goods the tomb can be dated to the late 4th century (Blečić Kavur– Kavur 2010) thus predating the Celtic invasion of Greece (and the retreat) with which it has traditionally been connected (e.g. Szabó 2008). At the opposite end of broader Central Europe there is the famous Waldalgesheim situla. Discovered in 1869 in alate LT B1 female princely grave, this bell ‑shaped bucket has been the subject of numerous studies resulting in somewhat conflicting conclusions about the object’sproduction date and origin (Zahlhaas 1971b; Schiering 1975; Shefton 1985; Rol‑ ley 1987, overview in Joachim 1995, 27; for bell ‑shaped situlae in general cf. Zimi 2011, 53–55). Zahlhaas (1971b, 124) believed it to have been made in Tarentum as early as the first half of the 4th century BC (unlike all the other scholars who prefer the 340s–320s). Schiering, apart from lowering the date, questioned the idea of Tarentine production– in his opinion some stylistic imperfections betray anon ‑Greek, i.e. Italic craftsman. An even more radical depar‑ ture from Zahlhaas’ classification came from Brian Shefton (1985) who pointed out the almost 53 It was wrongly believed to date to the 2nd century BC by Zahlhaas (1971a), followed by others. The correct chronology was instituted by Blečić Kavur (2010). 86 THINGS AND THOUGHTS total lack of evidence in Italy of both finds of this variety of bell ‑shaped situlae and of some of their characteristic technical traits. Situlae of the same variant (A) are commonly found in the Aegean and Pontic areas; situlae of the related type B were certainly produced there. Also Rolley (1987) pleaded in favour of aMacedonian origin for the Waldalgesheim situla on the basis of comparison with the situlae from Derveni and Vergina. The decisive argument for abroadly eastern (Greek, Macedonian) production of the entire family of ovoid situlae to which the Waldalgesheim piece belongs was finally put forward on the basis of the method of manufacture by B. Barr ‑Sharrar (2000). An eastern rather than awestern origin for the Waldalgesheim bucket seems indisputable; it remains an open question whether it reached Central Europe by way of Italy (cf. avery similar situla in Montefortino: Brizio 1899, 774, tav. xi: 8) or through the Balkans and the Carpathian Basin (cf. Fig. 17). Far less known and yet extremely important for our understanding of the phenomenon are the situlae from the eastern Adriatic coast, recently synthesised mainly thanks to Martina Blečić Kavur (2010, 285–307). There are five fragments of bell ‑shaped situlae corresponding to at least two vessels from agrave in Rijeka (Blečić ‑Kavur 2010, 285, 446–447, tab.18), two rim fragments, one base, eight handles or their fragments and twelve attachments and aseries of feet (about which later) from various contexts in Nesactium (Mihovilić 2017); finally, an entire vessel missing the handles in which ahoard of jeweller’smoulds and instruments was found in Ošanići in Herzegovina (Marić 1978; 1995). Based on afresh look at the entire class of bell ‑shaped situlae, Blečić Kavur established on good grounds the distinction between At‑ tic, Macedonian, and Etruscan types; she considers the Waldalgesheim situla aMacedonian product in accordance with Rolley, Shefton, and Barr ‑Sharrar (there are other Macedonian pieces further south on the eastern Adriatic coast) while the pieces from Rijeka, Nesactium, and Ošanići are in her opinion Etruscan. They are securely dated to the late 4th–early 3rd cen‑ tury BC though the piece from Ošanići was buried much later, perhaps sometime in the first half of the 2nd century BC. Asmall series of situla attachments from the Adriatic area is unusual in that in these dou‑ ble handle attachments of classical form, i.e. with two flat eyelets flanking acentral vertical element, the vertical post terminates with depictions of human heads (Fig. 18). Three such anthropomorphic attachments were present in the Nesactium assemblage along with nine plain ones (Mihovilić 207, 265–266, sl.8, t. 2: 1–2, 7); another example was found in acoin hoard from Vrankamen in Bosnia (Truhelka 1893, 188, fig. 5). Relief decoration of the central vertical post is relatively common in all the situlae no matter if discovered in the Balkans or in Italy; this decoration is however, mostly vegetal rather than anthropomorphic. Entire human or mythological figures first appear in this role in the 5th century BC (Jurgeit 1999; Benedettini ed. 2012, 454, no 1417) and continue occasionally to re ‑appear (reduced to mere heads as in the Nesactium and Vrankamen pieces) in the 4 th /early 3 rd century BC, though not necessarily on bronze situlae: cf. an unusual two ‑handled kettle from atomb in Orvieto/ Bolsena territory in the Castellani collection (Moretti Sgubini ed. 2000, 152–153, no 104.3) or in ceramic imitations of situlae.54 Although formally close, the eastern Adriatic pieces do not correspond to these in their execution and perhaps we may be dealing here with agroup on its own, inspired by the Etruscan pieces but produced somewhere between Etruria and 54 Human heads in this position appear regularly in ceramic imitations of both ‑bell shaped and spouted situlae, in particular in Orvieto products (Michetti 2003, 169–170, nos. 142–144, fig. 9, tav. iii–v/bell ‑shaped situlae with female heads/ and 170–176, figs. 12–14 /spouted situlae with Dioscuri heads/); such figural decoration is not on record in Volterran products while in the Faliscan area situlae do not seem to have been produced (Michetti 2003). 87II. THE THINGS AND THE THOUGHTS the eastern Adriatic coast. It is not fully excluded that only replacement attachments were produced rather than entire vessels. The elegant and problematic spouted situlae have been the subject of numerous studies (overview in Jurgeit 1999, 328; Bini– Caramella– Buccioli 1995, 127–131, Zimi 2011, 49–53 and most recently in Blečić Kavur 2012); most scholars agree on their chronology (second half of the 4 th century BC) but no unanimity has been reached as to their production area. The suggestions include Macedonia (e.g. Barr ‑Sharrar 1982), Thrace (Venedikov 1977), Etruria (Boucher 1973; Adam 1984, n° 19), and southern Italy, that is Tarentum (Zahlhaas 1971a; Candela 1985; Jurgeit 1999, 328). Their wide distribution and common occurrence in both Italy and the Balkans is in my opinion most probably ahint at polyfocal production (at least in Greece/Macedonia and in Etruria) as proposed already by Pfrommer 1990; the same conclusions were reached by M. Blečić Kavur (2012, 161) who considers the majority to have been made in Macedonia though admitting local production of the late examples (down to the beginning of the 3rd century?) and relatively low quality pieces found in Etruria. No more clear than these vessels’ origin is their role in Transalpine contacts. Stephanie Boucher (1973, 85–96) in her first overview of these vessels’ distribution in Europe took an excessively opti‑ mistic approach considering authentic Transalpine finds to include many kept in European museums without certain provenance. In reality, local provenance cannot be proven for any of the finds listed by Boucher (Waal, Cologne, Lyon). Slightly more, but still not fully, convincing seems the case of avessel kept in the museum of Bonn (Menzel 1986, 203–204, Nr. 561). Its provenance from Efferen near Cologne was claimed at its acquisition (Lehner 1903, 354) 55 without any definite conclusion as to its origin (find? collection?). In the Adriatic, spouted situlae are represented by finds of handles with their attachments from Novi Vindolski on the northern Liburnian coast (Blečić Kavur 2012) and two lion‑ ‑headed spouts from Nesactium in Istria (Mihovilić 2017, 266–267). Moreover, Nesactium produced six ox ‑hide shaped situla feet. To my knowledge, no bell ‑shaped situlae with feet are documented. In contrast the trio of ox ‑hide shaped feet is repeatedly documented on spouted situlae and kalathoi in both the Balkans and in Etruria. In summary, the Adriatic coast provides us with an insight into an otherwise unprece‑ dented circulation of bronze vessels in the 4 th –3 rd century BC, complementing our view which sometimes gets too restricted by focusing on the La Tène Culture area and interactions exclu‑ sively between Celts and Greeks or Etruscans. Macedonian and Etruscan situlae circulated concurrently around the coasts of Picenum (Montefortino), Istria (Nesactium), Liburnia (Rijeka, Novi Vindolski), and Dalmatia (Ošanići). The vessels we find here were produced in the same workshops (sometimes of the same types) and at the same time as those from La Tène contexts in Waldalgesheim or Karaburma. We should certainly keep this in mind also when pondering the ways in which such vessels could have arrived in Central Europe. It is not to say that we should automatically discard Celtic migrations as an explanation, but neither should we consider it the only explanation available. 55 ‘Die römischen Bronzen wurden vermehrt durch eine Statuette der Venus, welche sich das Brustband anlegt, aus Gohr bei Neuss, einen Schlüsselgriff in Gestalt eines Pferdekopfes, mehrere Fibeln und einen Armring aus Bonn, eine sehr gut erhaltene Bronzeapplike mit dem Vorderkörper eines Pegasus aus Lannesdorf bei Muffendorf, den Doppelhenkel eines Bronzeeimers mit schön gearbeiteten Mascarons aus Effern, eine Bronzescheibe mit Minervakopf aus Blankenheim und das Ortband eines Schwertes aus Remagen.’ (underlined by j.k.). 94 THINGS AND THOUGHTS Fig. 23: Situlae Eggers 18–23, the fragments discovered in WnCE. 95II. THE THINGS AND THE THOUGHTS Fig. 24: Situlae Eggers 18–23, the fragments discovered in EnCE. It is also not improbable that the situla in which the Podmokly hoard was deposited [Pdm1]62 spent apart of its life in the nearby oppidum of Stradonice. The bucket of which only the rim and the handle are preserved is of type E20 with quite rudimentary ornithomorph handle 62 The NM inventories list one greater part of asitula handle including one of its ornitomorph finials (inv. n° 65133). Although very similar to the handle associated today with the Podmokly situla, it differs slightly from it in its description and considerably in its dimensions. With its 135 mm overall length, the NM handle is shorter than the piece [Pdm1] (187 mm) but at the same time too long to be its missing other half. The object was inventoried already by Vocel († 1871) and therefore its entry in the Museum collections predate the Stradonice pillaging. The object is documented only by acursory sketch and it has not been located in the Museum reserves, therefore any statement in its regard is even more difficult to make. It is to be considered astray object without clear provenance and not discussed further. 96 THINGS AND THOUGHTS Fig. 25: Situlae Eggers 18–23, the fragments discovered in Bohemia. 97II. THE THINGS AND THE THOUGHTS finials. Apart from Stradonice, situla fragments are also present in Třísov (one E18 and one E19 attachment eyelets [Tř05, Tř06], one certain (and one possible) baluster ‑shaped handle finial [Tř02, Tř0363] and two ornithomorph ones [Tř04, Tř34], all from the recent metal de‑ tector surveys); and from Závist (one baluster ‑shaped and one fragment of aornithomorph handle finial from Balda, both in aprivate collection [Zá01, Zá02]). The last find to mention is the E18 attachment fragment from Kolo near Týnec nad Labem [Ko01]– only the figural part is preserved. Fig. 26: The situla from the Pod‑ mokly hoard and the engraving of the hoard after Voigt 1771. On the other hand, three bronze vessel feet [S044, S048, S054] are inventoried in the NM with the provenance from Stradonice but also Podmokly. All are of different forms and dimensions and surely do not come from asingle vessel (unless it had been mended with haphazardly chosen spare parts). In this case, Iconsider the double attribution to be amistake and classify all three as from Stradonice. 63 In its primary publication we questioned its possible identification as ahandle finial due to the extreme length of its shaft (Kysela– Danielisová– Militký 2014, 575). Still, no interpretation seems more fitting and the interpretation is possible though with this caveat. 98 THINGS AND THOUGHTS Some doubtful finds moreover have not been taken into account, such as the iron attach‑ ment from Stradonice (Píč 1903, tab xiv: 53) which though strongly resembling its counter‑ parts on situlae of type E20 could also be from alocally made wooden bucket. The issue of bronze vessel feet, some of which may be from situlae, will be discussed separately below. Pans The origins of the bronze pans of the so ‑called Aylesford type (Fig. 27) are in central Italy, mainly in Etruria (Castoldi 1991; Bolla 1994, 16; Bini– Caramella– Buccioli 1995, 187–188, 190; Sueur 2018, 30–60). The earlier varieties (types Montefortino and Povegliano Scaldasole) which are identical in shape and differ principally only in their proportions (adeeper basin with ashorter handle) date to the 3 rd –2 nd century BC, while the Aylesford type appeared from the late 2nd and lasted till the end of the 1st century BC (Feugère– De Marinis 1991, 100, 104–108). Only the Aylesford type is widespread in Central Europe, while the Montefortino and Povegliano pans are present in northern Italy (Feugère– De Marinis 1991), appearing also in the Balkans. 64 The pans seem to have remained in production until at least the Augustan period (Castoldi 2003, 209–210). Fig. 27: Aylesford type pan. Feugère– De Marinis 1991, fig. 11. The pans were cast in one piece and finished by hammering. The basin (diameter 14–31 cm) has around bottom and acharacteristic thickened T ‑shaped rim often decorated by simple incised lines, chevrons and Xs. The handle terminates in ahook embellished with the head of awater bird (no other animals are attested) of various degrees of naturalism or stylisation. Afew vessels have their original feet preserved; these are made of lead in the shape of ashell and are soldered on to the bottom, three per vessel (cf. e.g. Furger ‑Gunti, 1979, 68, Abb. 39: 8). There has been intense discussion about the function of these vessels. Willers (1907) characterised them vaguely as ritual implements; Werner (1954, 66) believed them to have 64 Represented with at least one example from the enigmatic collective burial in Krajčinovići in western Serbia (Popović 1992, 64; Zotović 2007, fig. 5: 1–3– the proposed date of the tumulus to the mid‑2nd century AD seems too late). 99II. THE THINGS AND THE THOUGHTS been used for heating wine mixed with water and herbs as did Petrowski (1993, 187) while Caramella (Bini– Caramella– Buccioli 1995, 187) regarded them as very down ‑to ‑earth cooking vessels. Werner (1978, 8) later reconsidered his original idea as the lead feet exclude the use of the vessel in direct contact with the fire and the vessel’sbroad open form is little suited for heating liquids. He proposed instead to interpret the pans as parts of aset (basin and jug) for washing hands during feasts, based on aNuber’s (1972) study on the utilisation of such sets (Kanne und Griffschale) throughout Antiquity. This hypothesis (with validity for both Italy and the Transalpine world) was generally embraced also by the participants of the Lattes conference (Feugère– De Marinis 1991, 108; Bolla 1991b; Bolla 1994, 16–17; cf also Sedlmayer 1999, 76). Sabine Rieckhoff on the other hand rejected any practical function for these vessels (Rieckhoff 1998, 511–512) and argued for their use only in aritual sphere, be it for ablutions or for libations. Without specifying which features make them so suited for libations with their single handle, inward sloping profile and acapacity of three litres…). To finish, Bienert (2007, 71–72) considers the pans to have been used for food preparation (not necessarily over afire). Fig. 28: Distribution of Aylesford type pans in northern Italy, Central Europe and eastern Gaul. The finds outside the working area are not quantified. The considerations on the vessel’ssuitability for one activity or another, nevertheless, only concern their Mediterranean homeland (including northern Italy). The only evidence of use of these vessels in the Transalpine world come from tomb B in Goeblange ‑Nospelt in which apan was discovered containing pork ribs (Metzler– Gaeng 2009, 82, 276–277, fig. 70). The publishers observed on the vessel damage caused by exposure to heat for which they found parallels from elsewhere in the Transalpine world concluding that the pans were commonly used here in this function. The fact that the original lead feet of the Goeblange ‑Nospelt pan 100 THINGS AND THOUGHTS were replaced by more sturdy bronze pieces cannibalized from at least two different bronze jugs clearly shows that the pans were not originally intended for this purpose. In general the evidence from northern Gaul reviewed recently by Quentin Sueur (2018, 45–49) pleads more in favour of aculinary than hygienic use. The idea of alocal i.e. Transalpine production of Aylesford pans has occasionally been suggested (Břeň 1975a, 13; Hamm 1999, 43, pl. xx: 5 and 19) on the grounds of some seemingly half ‑finished examples (i.e. the duck ‑head terminals). It is, however, implausible: the duck heads in all sorts or metalwork differ strongly in their degree of stylisation and carefulness of execution; pieces lacking any engraved details are not unknown also in Italy (e.g. Feugère– De Marinis 1991, fig. 13; Galliazzo 1979, 196). Unlike their typological predecessors, Aylesford type pans are widespread (Fig. 28) in northern rather than central Italy and very often form part of La Tène grave assemblages (Feugère– De Marinis 1991). In the East Alpine area, finds are known from the Magdalensberg and Gurina (Sedlmayer 1999, 75–76, Taf. 29: 16). Beyond the Alps, they concentrate in particu‑ Fig. 29: Aylesford type pans, fragments discovered in WnCE and EnCE. * = not to scale. 101II. THE THINGS AND THE THOUGHTS lar in the Rhine ‑Mosel ‑Main area and along the course of the Danube (Kappel: Feugère– De Marinis 1991; Ulm: Wehrberger– Wieland 1999; Manching [M010–M012], Langensteinach in Mittelfranken: Dannheimer– Gebhard eds. 1993, 292, Nr. 141; Wamser– Flügel– Ziegaus eds. 2000, 312, Nr. 3a1). East of Bohemia there is one handle fragment from the Oberleiserberg [Ob04]. Moreover, one fragment of apan handle comes from Staré Hradisko [SH03] and one (now lost) from Hostýn [Hos1]. Jana Čižmářová (1996a, 120–121) classified them as fragments of horizontal simpula; in my opinion their interpretation as pans is more likely. Also, two duck heads also from Staré Hradisko (one finely engraved, one crude and undecorated) can be identified as pan handle finials [SH04, SH05] (Fig. 29). Fig. 30: Aylesford type pans, fragments discovered in Bohemia. * = not to scale. 102 THINGS AND THOUGHTS Among the numerous duck heads from Stradonice, nine can be attributed to the pans [S018– S026]; one fragment of an Aylesford pan handle from Stradonice is moreover documented from the Lehmann collection [S027]. One almost complete handle was discovered during the excavations in the northern acropolis at Třísov [Tř07] and another duck ‑head fragment comes from the Třísov detector surveys [Tř08]. One fragment of the body of abronze vessel from Závist [Zá03] can be associated with apan due to the characteristic engraved decoration on the thickened rim; in the case of another fragment from the same area this identification is less certain [Zá08] (Fig. 30). During the metal detector surveys of the Třísov oppidum an object was discovered in the form of awater ‑bird head [Tř33] of extraordinary size (length 67 mm!). The head itself is flattened from the sides and the beak from the top; the neck turns up. The head surface is finished by detailed engraving with extraordinary care and attention. The form of the object with the flattened head and upturned neck is closest to the hook termination on Aylesford pans. The object is nevertheless three times as big as the usual duck heads. Interpreting it as the termination of abucket handle seems improbable– the neck is turned in the opposite direction; the large beak would require auselessly large eyelet. In the first publication of the object (Kysela– Danielisová– Militký 2014, 585–587) we expressed doubts about it and hes‑ itated to class it among pan handle terminations. This hesitancy turned out to be unnecessary since aduck ‑head of identical dimensions and execution can be found on acomplete (large) pan from grave 92 in the cemetery of Belgrade ‑Karaburma (Todorović 1972, 30, tab.28). Jugs The category of bronze jugs is typologically very varied. The function which most immedi‑ ately comes to mind is that of serving wine though in the Mediterranean they may have been employed also for ablutions in association with pans or basins. Over the decades of research, the bronze jugs of the last two centuries BC have been divid‑ ed into aseries of types (Fig. 31; comprehensively Boube 1991; Sueur 2018, 68–69).65 On the grounds of two main typological criteria (the body profile and the form of attachment) the following types have been defined: sharply biconical jugs with heart ‑shaped (Gallarate type) or male bust (Piatra Neamţ type) attachments; S ‑shaped jugs with astraight rim (Ornavasso type) with the attachment either heart ‑shaped (the Ornavasso ‑Montefiascone type) or in the form of amale bust (Ornavasso ‑Ruvo type); S ‑shaped jugs with overhanging rim and heart‑ ‑shaped attachment (Kjaerumgaard type); piriform jugs with amassive handle terminating in an attachment of agrinning male face (Silenus or aNew Comedy character) with two pointed protrusions (Kappel ‑Kelheim type). This typology need not be fully exhaustive66 but it is more than sufficient for our needs. What all these types have in common are three bronze feet soldered on to their base (they will be dealt with in detail below) and the general shape of the handle with arched branches attached to the rim and athumb ‑piece at their crossing with the actual handle. 65 Werner (1954; 1978) differentiated the basic types of Kelheim and Ornavasso, followed by Ulbert (1984) who recognized the Kjaerumgaard type as alater variety of the latter. De Marinis (1975) and Tizzoni (1981) focused on the biconical jugs labelled the Gallarate type. Boube (1991) finally distinguished aPiatra Neamţ type from aGallarate type and with Montafiascone and Ruvo varieties within the Ornavasso type. 66 These typologies omit e.g. the S ‑shaped jug with ahandle attachment in the form of afemale (Medusa?) head between two vertical dolphins, which is widespread in the Mediterranean (e.g. Uroz Rodríguez 2015, 192, fig. 13; Luik 2002, Abb. 174: 80; Bolla– Castoldi 2016, 152, n° 7 etc.). The only ‘Celtic’ area they reach is southern Gaul (Girard 2010). 103II. THE THINGS AND THE THOUGHTS Fig. 31: Late Republican bronze jugs– types present in Central Europe. What reality does the complex world of jugs reflect? The Kjaerumgaard type (not attested in Late Iron Age Central Europe and only present in Augustan period contexts) is apparently alate ‑comer which is not relevant for our study. The Gallarate type jugs seem to have been around from the early 2nd century BC, developing in Etruria from earlier local forms (Boube 1991; Castoldi 1991). It is from the 120s BC that jugs of the Gallarate and Piatra Neamţ types appear in La Tène graves in northern Italy (Boube 1991) and the Kelheim type in various La Tène contexts (Barbau 2019, 31–32) to be followed by the Ornavasso type probably also of central Italian origin and attested seemingly only in 1st century BC contexts (Boube 1991, 35– ‘from the 70s BC’; Bolla 1994, 23– ‘first century’). In the case of the Piatra Neamţ type (rare in Italy and with ‘très fruste’ execution of the Jupiter bust on the attachment), Boube suggested ‘local imitation of exported models’. The argument is not very clear (the models would be produced in Italy anyway) and the equation ‘ugly ≠ Italian’ is misleading (many Italian vessels of this period show an equally poor and / or sloppy execution). Boube herself listed (‘only’) eight Italian finds,67 Bolla and Castoldi (2016, 132) mention three more new finds from northern Italy, and at least three more attachments 67 We may add to them the only recently published piece from the Bonifacio Falconi collection in the Vatican Museum: Caliò 2000, 209, n° 376. 110 THINGS AND THOUGHTS Fig. 37: Bronze vessel feet discovered in EnCE and WnCE. 111II. THE THINGS AND THE THOUGHTS One already mentioned foot from Třísov [Tř11] merits adigression– the ox ‑hide ‑shaped foot was visibly deformed by fire. Although accidental damage cannot be excluded, we may reasonably hypothesise that it had been the subject of intentional recycling (Kysela– Da‑ nielisová– Militký 2014, 581). Isolated vessel feet were discovered also in the Bibracte work ‑ shops (Hamm 2005, 67–68), ready for recycling or for reuse as spare parts, similarly to those secondarily attached to the previously mentioned pan from Goeblange ‑Nospelt tomb B. One of these was kidney‑shaped, two were ox ‑hide shaped. The situation in Bohemia is worth comparing with both neighbouring regions: in WnCE feet are represented mostly in Manching [M016–M029] while the Karlstein [Kst2], Leonberg [Lnb1], and Karlstadt [Ktd1] count one piece each. Out of these, the kidney shape prevails over the ox ‑hide shape in aratio of 11:6. In EnCE the situation is not so dramatic: five kidney ‑shaped feet, two of them from Staré Hradisko [SH07, SH08], two from the Oberleiserberg (both of the same form and size, potentially from the same vessel) [Ob01, Ob02], and one from Plavecké Podhradie ‑Pohanská [PPP1] as opposed to another five in the ox ‑hide form with variously configured stubs. The updated distribution of bronze vessel feet in central Europe (Fig. 35 below) thus does not seem to support my earlier assumptions (Kysela 2014a) about jugs as asupplementary indicators of distribution zones of southern imports. Simpula In Republican Italy, there are two basic forms of simpula or ladles on along shaft used for pour‑ ing wine from the mixing to the drinking vessels– those with ahorizontal shaft and aseparate removable bowl (type Pescate; Fig. 38A) and those with vertical handle integral with the bowl (Fig. 38B). In both types the handle terminates in asuspension hook decorated with abird or canine head.75 The duck heads from simpula can be told apart from those of pans thanks to their flattened curved beaks and the more three ‑dimensional approach to the depiction. The execution is as arule much more careful and elegant than in the often quite rushed pans. The horizontal Pescate type simpula (Castoldi in Castoldi– Feugère 1991; Sueur 2018, 174–184) are documented by numerous finds in the most diverse contexts and cultural areas, well reflecting the different forms of Roman cultural impact in the 2nd–1st century BC: from Roman camps (and native settlements) in Hispania (Mansel 2004, 20; Erice Lacabe 2007, 198; Uroz Rodriguez 2015, 180–186), through Mediterranean ship ‑wrecks (present as mer‑ chandise or as equipment? Genovesietal. 2013) up to settlements and sanctuaries in central Italy (Talamone: Ciampoltrini 1994, 375, fig. 4: 1–2; Volterra: Rosselli in Cateni ed. 2007, 219; Cetamura di Chianti: De Grummond ed. 2017, fig. 81), and burials in central and northern Italy. Arecent addition to this list comes from a1st century BC casa retica in Vadena/Pfatten‑ ‑Laimburg, prov. Bolzano (Marzoli 2012–2013). These various find contexts help establish their chronology to the uselessly long time span of ca 120 BC to the Augusto ‑Tiberian period. Finds are much less common in inland Europe. In Gaul they rarely leave the confines of the Roman Province (for which cf. Girard 2010, 333–334; Demierre 2015, 173, fig. 19: 5–6), though 75 Both animal species may appear on both ladle types. Their size and style vary considerably independent of the type. Benjamin Girard (2010, 333–334) proposed aclassification of the canine heads on vertical simpula according to the degree of their naturalism/stylisation into three groups, there is however no hint that this distinction reflects any past reality (e.g. chronological or chorological variety) other than aworkshop or even individual idiosyncracies. 112 THINGS AND THOUGHTS one detached bowl was found at the bottom of the Rhine in Mainz (Sueur 2018, 176).76 Two finds are attested in southern Germany (adetached bowl in the Kappel hoard: Wieland 1996, 209, Taf. 7: 1; one shaft fragment at the Karlstein [Kst3]), and one damaged but complete piece comes from the northern periphery of the La Tène Culture, from aLT D1 cremation grave in Seebergen, Lkr. Gotha in Thüringen (Huck 1999, 277–278, Abb. 3). Fig. 38: Simpula. A– type Pescate (after Popović 1992) and fragments discovered in Central Europe. B– vertical type (after Girard 2010) and fragments discovered in Central Europe. A and B not to scale. 76 Afind from Villeneuve ‑St.‑Germain was originally published by Debord (1998, fig. 1: 8) as apan hook and reinterpreted by Sueur (2018, 176) as probably aPescate type simpulum hook. Ifavour the original interpretation in view of the stylisation of the head which is characteristic of pans. 113II. THE THINGS AND THE THOUGHTS In the eastern part of the La Tène world, Pescate type simpula are quite abundant from the East Alpine regions (Gurina: Jablonka 2001, 167, Taf. 121: 21; Idria pri Bači: Guštin 1991, 66–67), through the Carpathian Basin (15 pieces, eight of them from the Karaburma cemetery alone are listed by Popović 1992, 64–66; cf. Todorović 1972) up to the territory of present day Romania (Rustoiu 2005). Fig. 39: Distribution of simpula in northern Italy and Transalpine Europe. The finds outside the working area are not quantified. In the immediate vicinity of Bohemia there are three handle fragments from Staré Hradisko: two flat bars [SH14, SH15] and avery accurately modelled and engraved canine head from afinial hook [SH16].77 In Bohemia itself, horizontal simpula are represented by ahandle frag‑ ment [Tř18] and atiny broken ‑off canine head [Tř19] from Třísov and one hook with afinely 77 The head by itself cannot be unequivocally ascribed to one simpulum type or another. The Pescate type seems the more likely candidate in view of the shape of the shaft which, though bending axially, has asimple circular section and mainly in view of the distribution of both types in central Europe (cf. below). 114 THINGS AND THOUGHTS sculpted duck head from Stradonice [S062]. 78 The piece referred to by Werner (1954, E:20, Abb. 4: 8) as Pescate type (here [S064]) belongs in reality to avertical simpulum (see below). Vertical simpula (Feugère in Castoldi– Feugère 1991; Sueur 2018, 174–184) develop from earlier Etruscan ladles (e.g. Jurgeit 1999, 439–447, Nr. 743–747, 750–753) in comparison with which their bowls grew deeper and the hook (previously at right angles) bent towards the handle with which the animal head on it is parallel. This makes them easy to distinguish from Aylesford pans. Ladles of this form are attested both in bronze and silver79 with achronology spanning from the late 2nd or beginning of the 1st century BC to the Augustan period.80 Vertical simpula are relatively widespread in the Mediterranean world and in specific parts of its neighbouring territories: in Italy they concentrate in its northern part (they are reportedly very numerous in the region of Aquileia; Bolla 1996) mirrored in the East Alpine region (an overview of Slovenian finds: Guštin 1991, 16, 71, Abb. 14: 11; supplemented by new finds from Ljubljana: Gaspari 2014, 99–100, 106, fig. 103b, 110, and one broken off head from the Magdalensberg: Deimel 1987, Taf. 17: 5). In Gaul they are relatively numerous in the south (Girard 2010, 333–335 lists 72 fragments translating to 46 MNI), but further north they appear only singly in sites of aspecial status (Barbau 2019, nos 193–212;81 Sueur 2018, 177, fig. 127): Gondole, Bibracte (5 fragments), Fossé des Pandours (2 fragments), Titelberg (3 fragments). No vertical simpula or their fragments are known to me from southern Germany or elsewhere in the Middle Danube area. In surprising contrast to this dearth of finds in the whole of Central Europe, there are as many as three fragments of these in Stradonice: one almost complete handle with acanine head on the hook [S064], one bowl [S063], and one hook terminal in the shape of aduck head [S065]. No other simpula types (e.g. types 5 and 6 with astrainer attached to the handle) have been documented in Central Europe. After all, these are as arule characteristic of the last few decades of the 1st century BC. Strainers The elements of what is conventionally called bronze strainers (Guillaumet 1991) are the most common bronze vessel type in Transalpine Europe. As traditionally reconstructed (by Reinecke and Déchelette and passing through Břeň, Christlein, and Guillaumet), the bowl itself 78 The duck head in this piece is unusually large corresponding in its size and shape with apiece from Pompeii (today kept in Berlin) and dated by astamp F HEL CAT to the Augustan period (Castoldi in Castoldi– Feugère 1991, 66–67, fig. 5, n° 7 with further bibliography; Petrovszky 1993, 27, 154, Tf. 17: H.02.01). This of course does not mean that we should consider an Augustan date for the Stradonice piece as well (it has no stamp). 79 Silver examples– Arcisate (now in the British Museum: Piana Agostinetti– Priuli 1985); Boscoreale (now in Antiquario Comunale, Rome: Talamo 1993; Guzzo ed. 2006, 165, n° 216), Tivoli (now in the Metropolitan Museum of Art, New York: Oliver 1981, fig. 2 in p.59); Palmi (Guzzo 1977–1979); Menghíbar (Raddatz 1969, 94–95, 227, Taf. 24). 80 The statement by Eduard Droberjar (2006, 14) that vertical simpula are the latest imported objects datable to the Augustan period is imprecise. This date concerns only some varieties not attested in Central Europe (Castoldi– Feugère 1991) while for those present in Late Iron Age contexts, the Augustan period is only the latest phase of their existence, when probably they were no longer in production. 81 Finds from Feurs and Paule, listed by Clémentine Barbau as simpula hooks should in my opinion be classed rather as terminals of abucket handle and of apan respectively. 115II. THE THINGS AND THE THOUGHTS was hemispherical (diameter 7–13 cm) with perforations in geometric or vegetal patterns while the grip is composed of two characteristic parts: atwo ‑arch handle with its ox ‑hide shaped attachment plate and aY ‑shaped thumb piece with two stylized ornithomorph protomes on the front and aswallow ‑tail shape on the back (Fig. 40). The thumb piece was soldered on to the vessel rim above the handle and both elements may also have been joined together as hinted by traces of solder often observed on the downside of the thumb pieces. This configuration, providing amore secure grip and easier handling of the vessel in all directions, is derived from Hellenistic skyphoi (Fig. 40 right; cf. silver pieces Segall 1938, 51–52, Taf. 15: 39; Piana Agostinetti– Priuli 1985, 187–191, fig. 2; Baratte 1989; Colivicchi 2002, 286–287, fig. 46.5; pottery: Morel 1981, èspeces 3150–3170, 3310–3330) on which the first reconstruction of these handles by Reinecke (1911) was based. Fig. 40: Bronze strainer reconstruction. Guillaumet 1991, fig. 1: 1, modified after Uroz Rodríguez 2015, fig. 12. Skyphos held in ahand– fragment of astatuette from Nauportus (after Horvat 1990). Metzler (et al. 2016, 242) and Sueur (2018, 130–132) recently pointed out that all three compo‑ nents are never present in any of the (rare) complete pieces: the bowl carries either the handle (Wallendorf, Valeggio sul Mincio, Orbetello) or the thumb piece (Nogara: Bolla– Castoldi 2016, 150; Floheim: Guillaumet 1991; Arcisate: Piana Agostinetti– Priuli 1985, 192–193, fig. 3: 1) and in none of these cases does the missing piece seem to have ever been present). Already Giulio Ciampoltrni (1994) suggested that apart of the strainers (in his opinion from the later phase of production) lacked the thumb pieces whose function was fulfilled by their large horizontal rim such as those in the strainers from Orbetello (only asingle detached thumb piece was preserved along with nine strainers) (Ciampoltrini 1994) and from Idria pri Bači (Guštin 1991, Taf. 20: 5). Metzler went so far as to claim that the two ‑piece grip was never actually used, that the bevelled face of the thumb ‑piece branches exclude its soldering to the universally vertical profiles of the preserved strainers and that the prevalence of thumb pieces over the handles suggests that they were actually used independently, the thumb pieces even not necessarily on strainers. This radical statement can be disproven by at least one actual find of astrainer in which all the components are preserved82 in Libisosa in southern Spain (Uroz Rodríguer 2015, 186–187, fig. 12, n° 9). As to the much higher number of thumb pieces than handles, it is worth realising that the attachment area of the handles (asquare plate) is much more solid than that of thumb pieces (a1–2 mm thin horizontal line under constant strain) and the thumb pieces are therefore 82 The bevelled face of the thumb piece sits very comfortably on itsdiagonally inclined rim. 116 THINGS AND THOUGHTS much more likely to break off. The essence of Metzler’sargument is nevertheless fully valid and it is entirely possible that some strainers may have been made with only one of the elements. As to Metzler’sidea that these elements need not necessarily come from strainers, Ifind it quite reasonable, having already argued in the same vein (Kysela– Danielisová– Militký 2014, 582). It is true that none is documented attached to the body other than one that is perfo‑ rated; the number of these elements found in any functional connection is however very small (half adozen pieces). In Mediterranean metal ware and pottery, the same arrangement is doc‑ umented also in drinking cups (skyphoi, kantharoi) and it is truly remarkable that not asingle category of drinking vessels (perhaps with the exception of Istria type mugs) is documented among the metal vessels imported to Transalpine areas. The wide distribution of thumb pieces and handles in Transalpine Europe would make this option ever more suggestive. Throughout this text Irefer to these vessels systematically with their conventional label ‘strainers’; Ithink however that, although there is no proof of this, ascribing some of the ‘strainer’ elements to drinking cups is not an improbable hypothesis. After all, Reinecke (1911) as the first scholar to link both parts together, considered them as parts of cups (kantharoi) rather than strainers (cf. also Břeň 1966, 107–109, obr.27). The interpretation as astrainer was later put forward by Christlein (1963–1964) and the topic was developed in particular by Guillaumet (1977; 1991). The large formal variability, low production quality and alleged presence of half ‑finished examples at Bibracte as well as the (then) seeming absence of finds in the Mediterranean and quick disappearance after the Roman conquest of Gaul led Guillaumet (1977, 245) to postulate alocal Transalpine production if not origin for these artefacts. This radical statement was later moderated (Guillaumet 1991, 93); the idea of Transalpine production of these pieces however lingers on (Poux 2004, 244; Karwowski 2017, 266–267). The seeming absence of these vessels in the Mediterranean can be accounted for. First, this absence is more apparent than real: the strainers or their parts are very numerous in Spain (Christlein 1963–1964; Ulbert 1984, 87–89; Fabião 1999, 179–183; Mansel 2004, 25; Erice Lacabe 2007, 199–200; Uroz Rodríguez 2015, 186–188). In Italy, the artefacts were not unknown in the first place (eight finds listed by Guillaumet 199183) and more finds have appeared (or were previously overlooked) both in north Italian Celtic tombs (three pieces from the Veronese alone: Bolla– Castoldi 2016, 150), but more importantly also in peninsular Italy (nine vessels or their fragments from awreck probably off ‑shore from Orbetello (Ciampoltrini 1994); ahandle at Blanda, Calabria (La Torre– Mollo 2006, 412, tav. cxix: 15); aminuscule silver thumb piece (possibly from acup rather than astrainer) in Ancona ‑Colle Gardetto, t. 1885/II (Colivicchi 2002, 252, n° 38.2); an angular thumb piece from an Imperial context in Macerata (Sorda 1971, 407, fig. 13: 1); and especially, one silver strainer is preserved in the Arcisate treasure (Piana Agostinetti– Priuli 1985) and two complete cups of silver in the Palmi treasure (Guzzo 1977–1979). More importantly, our distribution maps may be largely skewed by various circumstances. The rarity of finds in north Italian graves may be due to ritual reasons dictating which vessels should be buried with the dead (among which the strainer need not have been apreferred object).84 Moreover, research into Republican metal vessels is centred in northern Italy (M. Bolla, M. Castoldi, R. C. De Marinis). In the regions south of the Po where bronze vessels did 83 Apart from finds from La Tène or other peripheral contexts (Valeggio sul Mincio, Ornavasso; Castelvenere near Trieste), these included also finds from purely Roman sites: such as the Latin colony of Luni (Mercando 1971, 407, fig. 13; Frova 1973, tav. 135: 5CM2700), Herculaneum, the port of Mondragone in Campania, and Rome itself (all unpublished, quoted by Guillaumet 1991). 84 This is after all the case in Gaul as well where strainers are almost systematically present in settlements rather than in tombs (Barbau 2015, 569–577, ill. 372–373; Sueur 2018, 128). 117II. THE THINGS AND THE THOUGHTS not form part of burial assemblages, the researchers may therefore be less likely to recognize their fragments in settlement excavations (e.g. the Blanda handle mentioned above was not identified for what it is in the publication). Last, but not least, as with all kinds of Republican small finds we must not forget how fragmentary our knowledge of pre ‑Augustan settlements is and how notoriously rare metal finds are in them. For all these reasons Iprefer considering the strainers as mainly Mediterranean products. This does not mean that none of them could have been produced in Transalpine Europe. We are however not able to tell apart the local products from the imported ones; their function has in any case strong Mediterranean connotations, and to recognize their ‘mediterraneity’ seems more correct than refuting it. The strainers’ chronology extends, as with most bronze vessels, over the entire LT D period down to the Augustan age (for which cf. Fingerlin 1986, 78, Taf. 207: 2; Deschler‑ ‑Erb ed. 1996, 32, note 161). Sedlmayer (1999, 88) even mentions some pieces from full 1st and even 2nd century AD contexts. Represented by dozens of examples, the thumb pieces feature aconsiderable variety of form. Three proposals for their classification have appeared recently independently from each other, all agreeing in the basic trend of development and differing only in details (Fig. 41). Božič85 recognised atendency from earlier (LT D1) slender sinuous forms towards later (LT D2b–early Imperial) massive and angular ones. Metzler (et al. 2016, 238–242) distinguished two types within Božič’slater form (2: massive but still sinuous, 3: mas‑ sive and rigid), dating them both to the Augustan period. Also, according to Sueur (2018, 137, 145–146) the development went from the most common type Iwith its elegant curved shapes, to types II (partially contemporary with the former) and III progressively simplifying it to various degrees of rectilinear schematisation. He further defined atype IV with extremely 85 Lecture in the 2014 conference on the La Tène period in Central Europe in Oberleis ‑Klement. Cf. also http://iza.zrc ‑sazu.si/En/Raz_Lt.html#d8. Fig. 41: Strainer thumb pieces. Typology after Božič, Metzler, and Sueur. 118 THINGS AND THOUGHTS fine branches still preserving the curves of the first type, in which the bird heads on the fron‑ tal element become (almost) completely indistinct. The situation in Central Europe (with La Tène occupation mostly earlier than LT D2a) does not contradict this scheme. Thumb pieces of Sueur’stypes II/III are absent here; somewhat angular items even closer to type Iare doc‑ umented at Manching [M042], as well as at Stradonice [S093, S100, S101] where the presence of LT D2 artefacts is not fully excluded, and in Bratislava [Ba02] where it is expected. Thumb pieces of type IV are evenly represented in small numbers: Manching [M041], Stradonice [S102, S103], Třísov [Tř26], Staré Hradisko [SH27]. It is worth pointing out that while Early Imperial chronology of e.g. Metzler’stype II is quite probable, an angular shape by itself need not be achronologically decisive element: for example angular thumb pieces with straight squared of tail parts were present in two silver skyphoi discovered in 1930 in Pradela de Guiães, prov. Tras ‑os ‑Montes, Portugal and containg ahoard of 237 Roman Republican coins, the latest of which was minted in 64 BC (Raddatz 1969, 281, Abb. 35). The strainers probably served for filtering wine from ingredients or impurities (pitch conserving the wine and permeating the amphora body, or herbs added to it intentionally before consumption).86 The diffusion of strainers in Transalpine Europe largely exceeds that of the amphorae (both were believed to overlap by Christlein 1963–1964, 18–19, Abb. 3) and the strainer could thus theoretically be taken as alternative indicators of the wine trade (in areas where wine was transported in other containers?). In reality, it seems quite probable that in the areas where amphorae are missing the strainers were used for filtrating less noble (though no less intoxicating) beverages (cf. also Poux 2004, 244). Strainers are the most widespread metal vessels in the La Tène world (Fig. 42). After all, among the six artefacts with which Déchelette (1914, fig. 404) demonstrated the unity of the oppida period material culture, there are two strainer components. They are represented throughout inland Gaul (Barbau 2019, 31, 170–171; Sueur 2018, 148–162) as well as in the Carpathian Basin and the Balkans (Rustoiu 2005, 68–68, fig. 8). In Central Europe they are present both in the east and west. ‘Along the Amber route’ (Fig. 43): Celje, the Magdalensberg (Sedlmayer 1999, 87), Velem, Bratislava (thumb piece [Ba02]), Devín (handle [De01]), Zohor (thumb piece [Zoh1]), the Oberleiserberg (handle [Ob03]), the lowland settlement of Thunau am Kamp (handle [Thu2]), Staré Hradisko (14 fragments– body fragments [SH27, SH28]; thumb pieces [SH17–SH24b], handles [SH25, SH26]), Bořitov (thumb piece [Bř01], body frag‑ ment [Bř02]?). In the west (Figs. 44 and 45) there is asignificant concentration in the area of the Rhine ‑Main confluence, as well as in WnCE along the course of the Danube in Manching with 36 finds [M030–M063, M070, M071], Kelheim (thumb piece [Kel3]), Wallersdorf (body fragment [Wld1]), Egglfing (thumb piece fragment [Eg01]), and Passau (thumb piece [Pas1]); in southern Bavaria and the Alpine foothills in Steinebach am Wörthsee (thumb piece [SaW1]), the Karlstein (athumb piece and ahandle [Kst3, Kst4]); Salzburg (thumb piece [Slz1]); Bruck‑ ‑Fischhorn (thumb piece [BrF1]); and in northern Bavaria in Altendorf [Ad01, Ad02] and Weißenburg [Wei1]. This large presence is mirrored also in Bohemia where there are 39 thumb pieces [S066a– S103], 15 handles [S104–S113, S120–124], and three body fragments [S114–S116] from Stradonice itself, eight or nine thumb pieces [Tř 23–Tř30, questionable Třx1] and three handles [Tř20– 86 Margherita Bolla (1996, 189) observed complementarity between the distribution areas of simpula on the one hand (northern Italy) and strainers on the other (Gaul and Central Europe), interpreting them as areas of consumption of aromatised and pure wine respectively. But would not the consumers still need aladle or another instrument in order to transfer the wine into the strainer? 119II. THE THINGS AND THE THOUGHTS Tř22] from Třísov, one thumb piece and one handle from Závist [Zá05, Zá06]. Besides these well ‑known oppida, there is also one thumb piece from the hillfort of Sedlo uAlbrechtic [Sed1], another one from asettlement pit in the lowland settlement of Strakonice in southern Bohemia [Stk1] and ahandle from an open settlement in Lipec87 in the Elbe lowlands [Lip1]. 87 Iam grateful to Zdeněk Beneš (ÚAPPSČ) for information about this unpublished find. Fig. 42: Distribution of strainers in Central Europe (above) and Italy and Transalpine Europe (be‑ low). Finds outside of the working area are not quantified. 126 THINGS AND THOUGHTS Fig. 49: Mug fragments discovered in Central Europe. 127II. THE THINGS AND THE THOUGHTS Apart from the usual north Italian graves (e.g. eight pieces mentioned in the Veronese by Bolla– Castoldi 2016, 150–151), Idria mugs are present also in settlements and graves in Etruria (e.g. Cristofani 1975, 26, fig. 19, n° 67); as many as nine pieces were recovered from awreck in the estuary of Cecina (Genovesietal. 2013, 82–83 with references to other finds from Etruria); they are numerous in Spain (Erice Lacabe 2007, 203; Mansel 2004, 23; Uroz Rodríguez, 2015 194–197) and Portugal (Fabião 1999) and present also in the already mentioned Delos. Finds come also from the Eastern Alps (the eponymous Idria pri Bači cemetery to begin with, as well as the Magdalensberg) and from Gaul though most finds seem to concentrate there in the south– Barbau (2019, 180) lists four examples of mugs from inland Gaul. In Central Europe (Fig. 49), there is one entire vessel [M064] and three or four handle frag‑ ments [M065–M067, M068?] from Manching and another one– extremely fragmentary– from Liptovská Mara, outside our working area (Pieta 1996; 2008/2010, obr.119: 13). In Moravia there is one handle fragment from Staré Hradisko [SH29]; finally, in Bohemia one attachment and one handle from Stradonice [S117, S118] and one handle from Třísov [Tř32] can be mentioned. Amphorae Acordiform attachment with acentral rib from Stradonice [S119] originates from an amphora rather than ajug or asitula (Fig. 50). Such closed vessels with two symmetrical handles (Feu‑ gère 1991b) had previously been known only from the Mediterranean, apart from Italy (the list by Feugère 1991b can be supplemented by at least three examples from awreck off ‑shore of northern Etruria: Genovesietal. 2013, 83, fig. 12–15, and three from the Veronese: Bolla– Castoldi 2016, 143, 151, fig. 23a, tav. 4: xviii/3); there is one in aburial in Agde in southern France (discussed by Feugère 1991b) and asmall series of finds from Spain and Portugal (Erice Lacabe 2007, 197–198; Uroz Rodríguez 2015, 178–180, fig. 7; and aquestionable example from Cabeça de Vaimonte: Fabião 1999, 185). The only other possible bronze amphora fragment from Central Europe is arecent find from Manching [M069].88 The handle rightfully caused some perplexity: though doubtlessly coming from an oppidum period sealed context, and being aMediterranean bronze vessel 88 Also some handles interpreted here as those of jugs could in reality originate from amphorae, e.g. M015. Fig. 50: Bronze amphorae, fragments discovered in Central Europe. 128 THINGS AND THOUGHTS component, its only parallels can be found in the later 1st century AD Pompei (Sievers 2013, 228). We may explain this by the rarity of bronze amphorae in Republican Italy– with few published pieces, much can still be unknown to us.89 Basins Avery rare vessel form is the basin with two handles with wine ‑leaf attachments, type Eggers 94 (Bolla 1991c; Sedlmayer 1999, 57; Bolla– Castoldi 2016, 134). These artefacts, dated at latest to the second quarter of the 1 st century BC (though probably earlier), seem to concentrate in the southwestern Alpine piedmont; two were present in the Rhône ‑Saône valley, and one handle was discovered at the Karlstein in Bavaria [Kst6] (Fig. 51). Their rarity and exquisite workmanship suggest their foreign provenance even in northern Italy, with the eastern Med‑ iterranean as the possible area of origin. Basins Eggers 91/92 are known from Gaul (the Titelberg, Rouen, Nijmegen, and the camps of Dangstetten and Haltern: Sueur 2018, 109, fig. 74), from the Magdalensberg (Deimel 1987, 38, Taf. 20: 3–5; Sedlmayer 1999, 58–59), and from Dacia (Rustoiu 2005, 79–81). Three frag‑ ments were discovered in Bratislava [Ba0390, Ba05a, Ba05b] and one in Devín [De02] (Fig. 51). Dating from the early Augustan period on, these vessels actually exceed our timeframe (for their importance in the Early Roman Iron Age cf. Werner 1954; Karasová 2004, 24–26). In Bratislava they obviously mark the site’slatest phase. It is worth noting that their find contexts are either not described in detail ([Ba05a, Ba05b]) or are stratigraphically unclear ([Ba03]) and it is therefore not sure if they are to be connected with the oppidum period or with the subsequent occupation of the area. Basin foot? Aunique object from Staré Hradisko is kept in the Boskovice collection [SH30]. It is aslightly arched sheet (rounded at one end, pointed at the other) with adisc perpendicular to its con‑ vex face (Fig. 51). Idetical objects (without any information about their provenance) are kept in Museo Guarnacci in Volterra. The elementary structure of the object recalls strongly the feet attached (in their case by means of an ox‑hide shaped plate) in threes on the bottoms of basins of Tassinari type S1100, attested by anumber of examples in Pompeii (Tassinari 1993, 93, nos 3483, 7274, 18736, 8184, 1958, 41435). Dragan Božič (2002, 419–421) drew attention to the fact that in the territories of present day Slovenia, Croatia, Serbia, and Romania the feet of the S1100 basins appear with regularity as early as LT D1 (cf. also Dizdar– Tonc 2014, 589–590). On the other hand, Rusotiu (2005, 80–81, fig. 24) argued that none of the contexts quoted by Božič provides sure dates earlier than the last third of the 1st century BC, which would be of course in conflict with the chronology of Staré Hradisko. The Staré Hradisko (and Volterra) pieces differ from all the Tassinari S1100 feet known to me in that the attachment plate lacks the characteristic ox ‑hide shape; also, the actual disc ‑shaped foot seems less massive and fully circular rather than flattened on the bottom. 89 At the same time we may point out the similarity with the handles of the Talcott type lekythoi of the pre‑oppida period (cf. Hurbanovo above). The date would not be impossible in the case of Manching. 90 The bronze fragment representing agrapevine leaf found in alatest La Tène period context in Bratislava ‑Vydrica was treated in detail by Kvetánová– Kovár 2010. Various possible interpretations were proposed including abronze lamp, mirror, and astatue fragment and abronze vessel. The interpretation as afragment of E91/92 seems most probable. An exact parallel is the attachment of abasin from the Guştariţa ‑Sibiu hoard in Romania Rusotiu 2005, 79, fig. 13: 2. 129II. THE THINGS AND THE THOUGHTS Regardless of these differences. Their identification as feet ofbronze basins seems to me areasonable proposition. Vessel body fragments and undetermined pieces Late La Tène settlements regularly produce large quantities of bronze sheet (impressively demonstrated by Karwowski 2017, fig. 6), some of which may be body fragments of metal vessels. Ionly include rims and bases with asufficiently characteristic shape that they can be ascribed to specific vessel shapes. This is basically impossible in the case of body fragments, and only some strainer body fragments and some very characteristic pan rims can be considered. Fig. 51: Bronze basins, fragments discovered in Central Europe. * not to scale. 130 THINGS AND THOUGHTS Aryballos handle? What is apparently abronze handle [Dčx1] found in Dobročkovice was published by Jana Čižmářová among bronze vessels from Moravia (Fig. 52). The Y ‑shaped bar with out ‑turned upper branches corresponds with those of late aryballoi based on the Talcott type (Fig. 57: 2ab; Vasale 1979; Feugère 1991a). Very close parallels come e.g. from the wreck sunk near Comacchio in ca 30 BC (Berti 1990, 97) or from Milan (Feugère 1991a, fig. 8: 3). The aryballoi are only exceptionally attested in Pre ‑Roman (Trans)Alpine Europe (Ptuj in Slovenia; Antran, dpt. Vienne, F; the Magdalensberg: Deimel 1987, Taf. 106: 8–9). More importantly they are not dated before the Augustan period and for this reason the object has to be excluded from our corpus. It is nevertheless worth further attention, if for no other reason than because Augustan period occupation is currently not attested in Moravia. Fig. 52: Dobročkovice, bronze aryballos handle. Shafted strainer One of the Lehmann photographic plates (Fig. 53) unmistakeably depicts ashafted strainer [Sx05]. Similar objects are relatively common in Italy from the 6th/5th to the 4th /3rd centu‑ ries BC.In the Lehmann piece, the actual strainer is asmall cup clearly distinguished from the larger bowl; the handle is made from the same piece of metal as the bowl. This is atrait characteristic of the earlier, 5th century pieces. In the central Italian products of the 4th–3rd centuries, the cup is made as asingle piece with acontinuous profile– either rounded or carinated– and with perforations in its bottom (Esposito 2007, 137–148). In contrast, many close parallels can be found in southern Italy. Strainers with adistinct filter, handle integral with the cup, and with aornithomorph hook at its end (either axial or turned sideways as in the Lehmann piece) are documented by two dozen examples in Apulia and several others in Sicily and Basilicata and dated mostly to (the second half of) the 5th–(first half of the) 4th century BC (Tarditi 1996, 48–53, 142–143). 131II. THE THINGS AND THE THOUGHTS Fig. 53: Stradonice (?), the Lehmann collection. Bronze shafted strainer [Sx05]. Unknown scale. This (provenance and) chronology obviously contrasts strongly with its possible origin from Stradonice. Moreover, the photographic plate on which the object is depicted, is preserved in only one of the three extant sets of the Lehmann collection and differs from the others for instance in the lack of the pre ‑printed heading (instead, ‘Stradonitz’ is written by hand on the top). All this suggests that the plate depicts later entries in the Lehmann collection bought on the antiquities market rather than the crème de la crème collected by Mrs. Lehmann directly on the spot in the first weeks of the excavations. It is not even sure that the plate was supposed to be dedicated to Stradonice only, even though the majority of the depicted objects clearly date from Late La Tène period. At any event, the strainer will not be taken into consideration in this study. Oenochoai Among the Stradonice finds in RGZM Mainz originating from the Robert Forrer collection there is also the rim of abronze jug with concave neck and rolled rim decorated with astragalus [Sx07b]. On the neck there are two holes for attaching the handle, in one of which there is arivet. The collections of the National Museum in Prague include an unpublished slender loop‑ ‑shaped handle with an elongated oval attachment decorated in shallow relief with the com‑ bined motif of apalmette on the attachment itself and alotus bud at its root [Sx07a] (Fig.54). 132 THINGS AND THOUGHTS Fig. 54: Stradonice (?), bronze oinochoe rim (Forrer collection) and handle (Berger collection). Both objects differ significantly from what is common and normal at Stradonice or in the oppida in general. Karasová and Schönfelder (2004, 233–235) rightly liken the Mainz piece with the 5th–4th century Etruscan olpai (Weber 1983, iiiBEtr b) and asimilar classification can be proposed for the handle in Prague. Specifically it corresponds to the handles of bronze oenochoe type ‘ivEtr a’ after Weber (1983, 175–206) dated variously from the mid‑5th century BC (Weber 1983, 170–180) or to the third quarter or second half of 5th century (Jurgeit 1999, 393–396, Nr. 646–650). These handles were attached to the neck by two rivets similar to the configuration of the Mainz rim (like e.g. Jurgeit 1999, 395–396, Nr. 650). Both fragments might theoretically belong to asingle vessel. There is no way of finding out whether these two fragments were actually found in Central Europe (it is unlikely that they were from Stradonice itself) or whether they were excavated in Italy and only came to Central Europe (to be sold as Stradonice finds) as items of the antiq‑ uities trade.91 Karasová and Schönfelder (2004, 235) observe that vessels of this type are not common in Central Europe; aparallel to the handle nevertheless comes from Súlovo, Slovakia (Bouzek 2002, Abb. 3:2). In any case, the artefacts cannot be associated with the oppida period occupation of Stradonice. 91 The origin of the Forrer collection is unknown; as far as the fragment in the National Museum is concerned, no information is available as to which collection it comes from (Berger?). 133II. THE THINGS AND THE THOUGHTS Ashell ‑shaped attachment Among the bronze artefacts from Stradonice Píč (1903, tab.xiv: 63) published abronze shell with an eyelet in the top and probably arivet in its centre [Sx01] (Fig. 55). The object has unfor‑ tunately not been traced in the National Museum collections. Both the style and iconography of the object are of clearly Mediterranean heritage. Shell ‑shaped elements form part of Greek and Etruscan bronze instrumentum from the 5 th century BC (e.g. Bini– Caramella– Buccioli 1995, passim; Feugère 2011) and linger on into the Imperial period (Sedlmayer 1999, 47, Taf.19: 4). Adiscrete group of shell ‑shaped attachments somewhat similar to the Stradonice piece are concentrated in the East Alpine region (Bustia in Fogolari– Gambacurta 2001, 261, n° 248), the Magdalensberg (Deimel 1987, 145, Taf. 21: 6), Grad (Božič in Turketal. 2009). Božič dates the find from Grad (though basically without areliable context) to the Early Iron Age and all the others with it, although the finds from Lagole and certainly that from the Magdalensberg suggest rather a1st century BC date. An artefact very similar in shape and size was discovered in the Hellenistic bath in Musarna, southern Etruria (Broise– Jolivet 2004, 226, fig. 184, n° 674) and likened to lids of bronze lamps (an improbable interpretation for the Stradonice shell with its flat rather than transverse perforation). In the Musarna shell the top part is not preserved. Attachments based on acentral motif (human head or bust, leaf, etc.) with atransversal bar and asimple eyelet above and fixed to the vessel body with arivet are com‑ mon in the mid ‑Imperial period (2nd–3rd century AD) in the E35–36 buckets and in cauldrons (Sedlmayer 1999, 103–105, Taf. 46: 1–5; Bienert 2007, 154–156). Although the shell is to my knowledge not attested as one of the depicted motifs, the transversal bar is very suggestive. To conclude, no exact parallel to the object has been identified, and although adate during the Stradonice occupation cannot be ruled out, other possibilities are equally if not more probable. For all these reasons, the object will not be included in the corpus of Mediterranean imports. Basin attachment The bronze plate shaped as awine leaf with amassive square hook in the top from the Stra‑ donice oppidum [Sx03] (Fig. 55) was an attachment soldered on to an Eggers 83 basin dating from the late 2nd till the 3rd/4th century AD transition (Radnóti 1938, 126; Taf. xxxviii: 3; Eggers 1951; Sedlmayer 1999, 62–63, Taf. 25: 9; Bienert 2007, 124–125). Whether it actually comes from Stradonice or not, it has nothing to do with the oppidum occupation. Fig. 55: Stradonice (?), artefacts excluded from consideration. 134 THINGS AND THOUGHTS Figural boss The artefact which caught the scholars’ attention from the first publications of Stradonice (Osborne 1880) is adomed oval boss made of bronze sheet with avertical band broken off at the top. The oval carries arelief depiction of ahuman bust [Sx04]. The oval face (of afemale or ayouth?) is framed with fluffy but smooth hair, and the clothes are stylized into asymmetrical curl with two folds on the chest (Figs. 11: 5 and 55). The style is based on soft forms with very little detail. The object is covered with avery regular smooth dark patina. Svobodová (1983, 660, obr.2: 14) considered the object ajug handle attachment belonging to the Imperial period Eggers 124 jugs (cf. Kropotkin 1970, tab 64: 1; cf. also e.g. Radnóti 1938, tab.xxxix–xliv; Sedlmayer 1999, 13–33, Tf. 2–7). However, no credible parallel can be found to the Stradonice piece in these vessels. More importantly (adetail Svobodová could not verify on the artefact which was kept in then inaccessible Vienna) the ‘attachment’ is in reality only ahollow shell which could hardly be attached to anything; the supposed handle is only aU ‑bent bronze sheet which would make carrying the vessel very uncomfortable. To sum up, the style of the depiction does not fit with either Roman Republican, or Late La Tène Art; typologically the fragment does not correspond with any common vessel of the period and functionally it would be totally useless. Amodern fake (with an enormous investment of labour) would seem to be the most likely explanation. The possibility of alocal product of un‑ known function (and quite possibly of Mediterranean inspiration) cannot be ruled out either. Cordiform balsamaria This is asmall and very problematic group of bronze containers (?) (Fig. 56). One such artefact was discovered in the Hrazany oppidum in 1962 [Hrx1]; another (so ‑far unpublished) comes from Devínské Jazero in Záhorie in Slovakia [DJx1].92 An object (now lost) of similar form was discovered in the eponymous site of the Púchov Culture in Púchovská skala (Hoening– Halaša 1903, obr.25; Hoening– Halaša 1905, 137). All these three objects have more or less the same shape with abroadly heart ‑shaped profile, very slightly conical neck and thickened rim. In the Hrazany piece there is acylindrical bronze sheet insert in the neck and three holes pass through both it and the neck– two aligned circular ones and asquare one at aslightly lower point. The artefact from Devínske jazero has no obvious holes in it. In the piece from Púchovská skala aloop seems to be attached to the rim; however, the rather clumsy drawing does not allow for very conclusive statements. Although the Mediterranean origin of these objects was not always explicitly stated (e.g. Jansová 1965, 69) it is implicitly assumed in its function of a perfume bottle which various schol‑ ars have unanimously suggested (Hrazany: Jansová 1965, 69, obr.24; Jansová 1992, 50, 149, Taf. 254: 17; Drda– Rybová 1998, 163; Bouzek 2011a, 155; Púchovská skala: Hoening– Halaša 1905, 137). This supposition is quite logical; nevertheless, several points need to be mentioned: almost no other perfume or oil vessels are attested among the Mediterranean imports in Central Europe or even in Gaul (Barbau 2019, 182–183) until the Augustan period and more importantly, these cordiform balsamaria have no direct parallels among Mediterranean bronze vessels. At the same time, the very idea of these objects being vessels can be questioned. Ihave discussed all of these points in detail previously (cf. Kysela 2014c), and therefore only ashort summary is needed here. Their practical use as balsamaria would be hindered by the design of their orifices facili‑ tating their suspension but not the insertion of astopper (holes passing through the neck in the Hrazany piece; avery unfortunately placed loop in the find from Púchovská skála). 92 Unpublished information by K. Elschek presented in the 2017 La Tène period conference in Trenčianské Teplice. Iam grateful to K. Elschek for the permission to publish the find. 135II. THE THINGS AND THE THOUGHTS As already mentioned there is afrustrating lack of parallels to the object among contemporary Mediterranean oil vessels (cf. Fig. 57)93 with two significant exceptions: an iron aryballos from Tarentum (Fig. 57: 6; Colivicchi 2001, E.15; Hemper 2001, 189) and asimilar object from Tha‑ sos (Ghali ‑Kahil 1954, 244–246, fig. 30). In spite of the closeness of the general shape with apointed bottom, there are some significant dissimilarities: these aryballoi are low and broad which makes them tilted but extremely stable unlike the Central European pieces; at least in the Tarentum example, the neck is straight which resolves all the closure problems (unlike the Central European pieces, cf. above). Last but not least, the Central European ‘balsamaria’ differ from the Tarentum and Thasos aryballoi, and from all the Mediterranean oil vessels, in one highly significant criterion: their volume. The volume 94 of the Hrazany vessel is 17 ml (Fig. 58) while the Tarentum example can hold an almost sevenfold larger quantity of liquid: 110 ml. The volumes of other Classical, Hellenistic, Republican, and Early Imperial examples range from ca 44 to 500 ml with most values clustering between 60 and 150 ml (cf. Kysela 2014c for more details). The point of this seemingly pedantic discussion is that the volume of only ca 20 ml prevents the Central Eu‑ ropean cordiform ‘balsamaria’ from reasonably fulfilling the function of aryballoi. Volumes of ca 20 ml are only characteristic of the smallest unguentaria95 made as arule out of pottery 93 Globular aryballoi with afunnel ‑shaped orifice derived from Greek models of Talcott type, are widespread in Italy only from the Augustan period on (Vassalle 1979; Feugère 1991a; Antran : Pautreau ed. 1999, 43). In 3rd–2nd century BC Etruria oil bottles are represented by piriform bottles made of relief bronze sheet of the types Cianferoni 1 and 2 (Cianferoni in: Maggiani ed. 1985, 148–152; Bini– Caramella– Buccioli 1995, 211–218, tav. lxxix). New oil vessel types appeared only in the Augustan period and include globular aryballoi with trochilus ‑shaped orifice (Fig. 57: 5; Tassinari 1993, type F 1100), and aryballoi with troncoconic body (Fig. 57: 4; Tassinari 1993, type F2120; e.g. asilver aryballos in the Berlin Antikensammlung: Heilmeyeretal. 1988, 267, Nr. 6). 94 The vessel volumes were calculated by means of the application ‘Calcul de capacité d‘un récipient à partir de son profil’ run by CreA, Université Libre de Bruxelles http://lisaserver.ulb.ac.be/capacity/ (cf. Engels– Bavay– Tsingarida, 2009). The volumes were always measured up to the transition between the shoulder and the neck. 95 According to V. Anderson Stojanović (1987, 117–119) the volumes of ceramic unguentaria from various parts of the Greek world range from 6 to 190 ml, in most cases from 20 to 40 ml. Camilli (1999, 11, fig. 2) observed concentrations of glass unguentaria volumes at 4–6, 8–10, 16, 20–23, 40, 55 and 120 ml, i.e. divisions and multiplications of kyathos (45 ml). Fig. 56: Cordiform balsamaria from Hrazany, Devínské jazero, and Púchovská skala. The latter not to scale. 142 THINGS AND THOUGHTS Závist vessels (cf. Nenna 1999– max. thickness of 2 mm: e.g. C90, C101, C103–106, C120, C130, C163 (opaque blue); C184, C186, C191, C219, C238 (violet); C251, C256, C258; max. thickness of 1 mm: C209; C255). There does not seem to be any reason to exclude these fragments from the counts of monochrome moulded vessels. The La Tène date of these fragments is moreover secured by asolid find context of one of them (the other being found in asecondary deposit: Drda– Rybová 2001, 317). An almost identical (albeit slightly larger) base, this time of greenish transparent glass, was discovered in ‘Roman building I’ on the Bratislava Castle Hill. Preserved fragments of the rim make it possible to reconstruct the vessel as abeaker [Ba07]. Overall, glass vessels are one of the less well represented import categories. After all, the finds do not seem to be extremely common in Italy either (though much more numerous e.g. in Delos: Nenna 1999). The overview of mosaic glass finds in northern Italy by Giuliana Facchini (2011) lists only 108 vessels or their fragments; the majority is however of Imperial date and for none of them does the author suggest adate earlier than ‘late 1st century BC’ (including vessels of reputed south Italian origin such as Facchini 2011, cat. nos. 33–34). The reasons for Fig. 61: Mosaic glass vessels discovered in Central Europe. © Museum of the Boskovice region, Moravian Museum and Museum of the Nový Jičín region. 143II. THE THINGS AND THE THOUGHTS this surprising absence may be– apart from the exclusive nature of glass vessels in this period even in Italy– again the relatively limited knowledge of Republican contexts in Italy and the residuality of fragments of Republican vessels (not always easy to tell apart from the Imperial ones) in contexts of the Imperial period. As amatter of fact, their finds in burials do appear in Hellenistic/Republican Italy (e.g. Ancona: Colivicchi 2002, 18.5, 22.9–11, 23.5, 24.3, 26.5, 27.1, 31.6, 51.2; Todi, Duomo, late 3rd– early 2nd century: Giorgi– manconi 2009). This does not in any way diminish their status as exceptional luxurious objects as is clear from the wealth of the burials in which they appear. In Transalpine Europe, they are represented by relatively few fragments (more often mosaic than monochrome) only in the most prominent sites (for mosaic glass cf. Venclová 2016, fig. 75; for an overview of monochrome vessels in Gaul cf. Foyetal. 2008; it is also worth pointing out arecent discovery of as many as 19 glass vessel fragments in the centre of Toulouse ‑ZAC Niel: Demierre 2015, 173, fig. 17: 1–6). In the East Alpine area we can mention only one fragment of amonochrome bowl from Razdrto (Lazar 2003, 34; Horvat– Bavdek 2009, 56 fig. 35). Other finds of moulded glass from the territory of Slovenia all seem to date only to the Early Imperial period (Lazar 2003, 33–45). Fig. 62: Monochrome glass vessels discovered in Central Europe. © Museum of the Boskovice region and Moravian Museum. 144 THINGS AND THOUGHTS In Central Europe (Figs. 61–63) (apart from Bohemia discussed above) find spots of glass vessels include Manching (14 pieces [M200–M211]) and Bratislava (Ba06, Ba07). In this compar‑ ison, the three fragments from the otherwise always exceptionally rich Stradonice are almost surprisingly few. More importantly, Moravia stands out in this situation in two respects: the oppidum of Staré Hradisko with its 13 mosaic ([SH31–SH42]: 11 millefiori, 1 areticella, 1 ribbon fragments) and 9 monochrome [SH44–SH52] fragments (Venclováetal. 2015, 215–218; Ven‑ clová 2016, 88–91; Venclová– Jonášová– Vaculovič 2017; Venclováetal. 2018) is by far the richest site in Transalpine Europe. In her detailed analysis of the vessels, Natalie Venclová moreover argues that at least some of the local fragments– closer to the Canosa horizon than that of Antikythera ‑Délos– may belong to arelatively early date, i.e. second century BC rather than the usual transition 2nd–1st century BC (Venclováetal. 2015, 223; Venclová 2016, 90). Another noteworthy aspect of Moravia is the discovery of one fragment of an areticella bowl [Pož1] in the Púchov Culture hillfort of Požaha in the Moravian Gate on the western edge of the Moravian ‑Silesian Beskydy Mountains (Čižmářová 2004, 201),102 i.e. asettlement which shows no signs whatsoever of status comparable with any of the above ‑mentioned sites. Fig. 63: Glass vessels, distribution in Central Europe. MIRRORS Fragments of simple disc ‑shaped mirrors, ca 10 cm in diameter, cast from the characteristic ‘heavy’ alloy of silvery colour are rather common finds in the oppida (Fig. 64). In the finds from the Třísov surveys (Kysela– Danielisová– Militký 2014, 593, tab.3, obr.8), the mirror alloys are characterised by ahigh proportion of tin (45–80%) with correspondingly low representation of copper (52–10%) and asmall admixture of lead (0.3–3%, exceptionally 9%). Similar values have been obtained for the mirrors from the oppidum of Altenburg (Penz 2012, 806) and for the Roman mirrors from Augst (W. Stern in Riha 1986, 16–19, tab.1). All of these 102 Strictly speaking, Požáha lies ca 1200 m outside our working area, that is in the Oder rather than Morava drainage basin. The areticella sherd is asufficient reason for breaking our rules. 145II. THE THINGS AND THE THOUGHTS characteristics not only evoke the mention by Pliny (NH xxiii, 45) on mirrors being originally (= in the Republican period) made in Brundisium stamno et aere mixtis (indicating in this word order the importance of the individual metals), but also correspond fully with those of Roman period mirrors analysed in detail by Nigel Meeks (1995), and we may conclude that from atechnological point of view, all the mirror fragments from our corpus fit the category of high ‑tin (or less prob‑ ably tinned low ‑tin) mirrors of Roman type. Silver plating of the surface (sometimes described in the publications) is never documented by hard evidence and it is most probably to be considered only as atrope, based on amisleading visual impression and reiterated without any evidence. In Etruria disc mirrors appeared in the 3 rd century BC, having replaced the previous mirrors cast in one piece with their handle, and they remained in use until the Imperial period (Serra Ridgeway 1996, 289; cf. also Lloyd Morgan 1981; Riha 1986; many examples are documented in the Hellenistic and Republican tombs of e.g. Tarentum: Hemper 2001; Ancona: Colivicchi 2002; Volterra: Cristofani 1975, 28, fig. 19, nos. 76–77). Production of Roman mirrors was an unexpectedly challenging task requiring high precision casting, lathe ‑work and polishing of the very hard and brittle discs. In principle, it need not have been out of the reach of Transal‑ pine craftsmen, but still– as in the case of metal vessels– production in Mediterranean regions seem more likely in view of the absence of La Tène objects whose production would require the same technical skills. Moreover, in the Transalpine area they are very clearly concen‑ trated on sites of special status such as oppida. Last, but not least, similar to the case of the strainers, there is no obvious clue to distinguish local products (should there be some) from the imported ones. For all these reasons, mirrors are counted among southern imports in the present study. This statement does not solve all the issues concerning the utility of mirrors for us: these brittle artefacts easily recognizable from the tiniest fragment (into which they easily shatter) can get wastly over ‑represented. For all these reasons, mirrors will not be counted as one unit for each fragment but quantified according to the principle of the minimum number of individuals (MNI): each diagnostic part (in the case of mirrors the edge) is counted as one individual; fragments without the original rim preserved are not taken into account unless no diagnostic fragment can be connected with acontext (in our case with asite); in such acase all non‑diagnostic fragments are counted as one individual regardless of their actual quantity. Fig. 64: Bronze mirrors, distribution in Central Europe. 146 THINGS AND THOUGHTS Fig. 65: Mirrors discovered in Bohemia. 147II. THE THINGS AND THE THOUGHTS The finds from Stradonice counting all the public and private collections combined include at least103 64 mirror fragments [S301–S312, S300a–S300n] corresponding to 12 individuals. In two cases from the National Museum in Prague from the Berger collection ([S301–306] and [S307–S310]) the fragments are recomposed as ajigsaw puzzle into discs glued on to acard‑ board base. The single fractures however rarely fit between two neighbouring fragments; the fragment thickness varies as does the nature of the patina on the various fragments; some are bevelled others are plain… Both discs are clearly composed from fragments which have nothing in common apart from being mirrors from Stradonice. Two large mirror fragments (both rims) come from Závist [Zá11, Zá12], one from Hrazany [Hr02], 20 (= 7 MNI) from Třísov [Tř40a–Tř40n, Tř44–Tř46] (Fig. 65). In narrower Central Europe, mirrors are present in equally large numbers in Manching (20 pieces corresponding to 5 MNI [M300–M305, M310a–M310n]) (Fig. 66). No other find is known to me from WnCE. This may however be only due to the state of publication as mirrors are, in contrast, extremely numerous in EnCE (Fig. 67), present not only in the oppidum of Staré Hradisko, but also in anumber of small hillforts of both the La Tène and Púchov Cul‑ tures: Ohrozim [Ohr1], Loučka ‑Obírka [Ok01–Ok04], and (afragmented but complete piece) Rýsov [Rýs1]. The collection from Staré Hradisko itself alone consists of as many as 110 frag‑ ments– large small, and minuscule– albeit corresponding to only 17 MNI (diagnostic pieces: [SH53–SH69]; non ‑diagnostic fragments [SH69a–SH69x]). 103 In some cases (e.g. [S300h]) one inventory number corresponds to anon ‑specified number of fragments. Iwas not able to verify their real number in the collections of the NM. Fig. 66: Mirrors discovered in WnCE. 148 THINGS AND THOUGHTS Fig. 67: Mirrors discovered in EnCE. 149II. THE THINGS AND THE THOUGHTS Mirrors are not unknown at our usual reference point of the Magdalensberg (Deimel 1987, 78–81, Taf. 58–62) and while otherwise seemingly not common, many finds may remain unrecognized or unpublished as explained above. All the mirrors in our corpus whose form can be determined are simple circular plates; rectangular mirrors attested in Italy and e.g. at the Magdalensberg (Deimel 1987, Taf. 58–60) are not documented in Central Europe though in some cases straight breaks may be difficult to distinguish from astraight rim. The rims may be bevelled, rounded or squared off. The only decoration is an occasional groove around the rim. This corresponds to the common Mediter‑ ranean mirror types (Lloyd Morgan 1981, 3–20, 30–35; Riha 1986, 15). There are numerous fibres of wood preserved in the patina of one of the Třísov fragments [Tř14] which may be remains of its wooden frame (cf. Lloyd Morgan 1981, 3) or even of acoffer lid (Colivicchi 2002, 240, 34: 11) (cf. Fig. 87 below). On the back of one of the fragments from Loučka [Okx1] and on aunique mirror fragment from Stradonice [Sx31] aparticular arrangement of concentric circles in relief with connecting rays can be observed. In the Stradonice piece, there is moreover astraight central rib pierced by atransverse hole and agroove in the corresponding place in the circular rib. No parallel to this configuration of the mirror back is known to me from the last centuries BC; it is most probable that these mirrors date to the Migration Period (as identifed already by Čižmář– Salaš 2009; for the so ‑called ‘Hunnic mirrors’ cf. Werner 1956, 19–24, Taf. 13: 1, 2, 4, 5, 45: 1–9, 47: 9). Rare artefacts of this date are documented in the Stradonice collections (Jiřík 2012). The mirror handle preserved in the Naturhistorisches Museum in Vienna [Sx30] (Fig. 11: 2) raises some questions. It does not correspond in its shape with the mirror handles of either the Hellenistic/Republican or Imperial Periods (contra e.g. Bouzek 1982; cf. Serra Ridgeway 1996, tav. cv: 133; Lloyd Morgan 1981, 38, fig. 5, 46, fig. 9, 48, fig. 10, 55, fig. 11; Deimel 1987, Taf. 64: 2–3; Riha 1986, 14, Taf. 1: 2, 2: 13, 3: 14–20), nor is any parallel to it available from areas of the La Tène Culture. The handle is in my opinion most likely alocal ad hoc product. Certainly not of Mediterranean origin is aseries of mirror handles identified by Karol Pieta (1996) and discussed in detail by Dulęba (2018). There is no Mediterranean parallel available for the type and the find distribution in the Middle Danube area clearly disproves the putative ‘Norican’ origin (cf. Danielisováetal. 2018, 148; Dulęba 2018). JEWELLERY BEADS With the significant but rare exception of coral, the beads discussed here are predominantly made of glass. Glass ornaments are among the few find categories of Mediterranean imports also attested regularly in the pre ‑oppida period. The fundamental basis for any inquiry concerning glass beads (and other glass ornaments) in our study area are the works of Natalie Venclová (1990; 2016), and Maciej Karwowski (2004). This also includes the question of Mediterranean imports, though this is minor compared to local production. It is often impossible to demonstrate an undisputed Mediterranean origin of some bead types. They may be well represented in both the Mediterranean and Central Europe but it is exactly their omnipresence which makes any clear statement problematic. This is for instance the case of the eye ‑beads widespread in Central Europe from the Early Iron Age. Some beads appear to have had quite eventful lives and were being reutilised and worn long after their 150 THINGS AND THOUGHTS manufacture as illustrated for instance by the mask bead fragments from Staré Hradisko or the amphoriskos ‑shaped bead from Gélert. For this reason Iwill exclude from further dis‑ cussion beads found in obviously secondary contexts (e.g. the Hellenistic mosaic bead from aRoman Iron Age pit in Libkovice [Lbx1]) or typological groups with avery long date ‑range and circulation area which do not provide any clear information, and Iwill focus on clearly defined bead types common in the Mediterranean but rare in Central Europe (Fig. 68, 69). Fig. 68: Mediterranean glass bead types represented in Central Europe. After Venclová 2016. 151II. THE THINGS AND THE THOUGHTS Emblematic for the early phases of Mediterranean contacts are opaque or translucid amphoriskos ‑shaped glass beads/pendants, appearing often in necklaces along with other ornaments. These beads have been the subject of several syntheses or overviews (Popović 1997; 2008; Schönfelder 2008, 307–309, Fundliste 1; Rustoiu 2015; Blečić Kavur– Kavur 2017) and only ashort summary is therefore needed. Amphoriskos ‑shaped glass beads (ashape pres‑ ent in the Balkans and northern Greece in various materials from the Early Iron Age) appear in the late 5th century, and are widespread mainly in the Balkans in the 4th and occasionally in the early 3rd century BC.Aurel Rustoiu (2015) analysed their distribution concluding that two main areas can be identified, an eastern one in present day Romania (linked to production in Greek cities on the Black Sea coast) and awestern group including subgroups in the East‑ ern Alps, Middle Danube area, and EnCE as well as– less relevant for us– in the south of the Carpathian Basin (for further observations including also the presence of these beads in Italy, e.g. Monte Bibele cf. Blečić Kavur– Kavur 2017). In Rustoiu’sopinion, the western group was entirely supplied by producers in northern Greece, Macedonia, and the eastern Adriatic coast (the latter suggested by the regular combination of amphoriskos ‑shaped pendants with branches of coral). The East Alpine group then played akey role in distributing these objects to the Middle Danube area (and presumably also to northern Italy). Equally significant is the observation (Rustoiu 2015, 370; Blečić Kavur– Kavur 2017, 104–105) that rather than large migrations with historical ramifications or trade (contra Schönfelder 2007; Bouzek 2017, 143), these beads are exemplary witnesses of small ‑scale individual movements, gifts or exogamy. The current distribution map may result from their long ‑time circulation in the receiving region. Their southern origin need not have got lost but was only the departure point of along relay. This is ahighly plausible model also for other bead types. Branches of raw coral, very often present in necklaces with amphoriskos ‑shaped pendants leave no doubt as to their origin in the Mediterranean (in this case literally). In the necklace from Brno ‑Horní Heršpice the amphoriskos ‑shaped beads and coral branches were accompanied by small biconical beads of colourless glass. Such biconical beads in differ‑ ent colours (types 301–305) are considered of definite Mediterranean origin produced from the 6th century and certainly continuing until the 2nd (Venclová 1990, 58–59; Venclová 2015, 144; Venclová 2016, 28). Several examples are attested in Bohemia and Moravia mostly from LT B1 graves (Libenice [Lbn1], Libčeves [Lbč1], Brno ‑Horní Heršpice [BHH1])104; two tiny blue beads were collected in Němčice [NH05, NH06] and alate example is also present at Stradonice [S401].105 Ribbed beads of types 306–310 are somewhat tricky. Ubiquitous in the Mediterranean, most widespread in Transalpine Europe in Ha D2–LT A(Venclová 1990, 59–62; Venclová 2015, 144–145; Venclová 2016, 28) they correspond precisely to the biconical beads discussed above, so they can be regarded as possible imports. For coherence sake, they will not be dealt with as imports in the synthesis, but their presence is nevertheless worth keeping in mind. Finds are attested from LT B graves (Jenišův Újezd), LT C agglomerations (Němčice, Nowa Cerekwia), and from LT C2–D oppida (Třísov, Stradonice) and settlements (Bezdědovice; Venclová 1990, 59–60 with references). The so ‑called mask beads, type Venclová 901 (Venclová 1974; Venclová 1990, 96, 120, fig. 12, 17; Karwowski 2005), are of certain Mediterranean origin (with aparticularly strong Punic connection) and are widespread in the Pontic region (Alekseeva 1975, 1978, 1982). In the 4th–3rd century they are particularly present in the Carpathian Basin and northern Balkans 104 Others from uncertain find contexts have not been taken in consideration due to their potentially Ha D/LT Adate: Tvršice and Žatec area cf. Venclová 1990, 58. 105 This may still be an authentic local find considering the production of the type until the 2nd cen‑ tury BC. 254 THINGS AND THOUGHTS The Němčice horizon does not seem to manifest itself in central Europe (unlike in Italy) by finds of gold and silver pieces, and the bulk of the imported coinage consists of bronze pieces, whose monetary value only applied in their home territory outside which (in northern Italy, let alone in central Europe) they became nothing but somewhat funny pieces of metal. Still they need not have been completely worthless (cf. e.g. Picard 1998). Bronze coins are explicitly mentioned e.g. within the ‘humanitarian aid’ sent in 224 BC by Ptolemy III to the Rhodians affected by an earthquake (Polyb. v, 89). The example of the Mazin type hoards mentioned above (Mirnik 1981; Bertol– Farac 2012) clearly shows that (albeit in distant and culturally different areas), bronze coins were systematically collected and hoarded (along with scrap raw bronze). It is still unclear whether the Mazin type hoards should be interpreted as caches of raw material for further production or whether the bronze was hoarded because of its value. We must not forget, in fact, that although bronze is used for production of ‘worthless’ fiduciary coins, it is still metal with acertain intrinsic value. Raw bronze (aes rude) is believed to have been the earliest form of expressing value by the Italic nations, and e.g. by Celts in the Po valley as Charon’sobol down to the Late La Tène period (Bergonzi– Piana ‑Agostinetti 1987). The monetary system of some Italic nations (including Rome herself) was based exclusively on bronze (aes signatum) down to the 3rd century BC (Cattani 1987; Neri 2004). In practical terms: by simply adding up the weight of all the Mediterranean bronze coins found in Němčice itself we arrive at 810 g of bronze. Most of this weight is made up by the huge Ptolemaic bronzes weighing as much as 50 g each; even two normal coins would, however, produce enough material for achain‑ ‑belt link or any other of the locally produced bronze trinkets (Čižmář– Kolníková 2006; Čižmář– Kolníková– Noeske 2008). The question, whether bronze was for the Cis‑ and Transalpine Celts only material or whether it had also some value is rather academic: bronze must have been to acertain extent valuable material. The idea that Greek bronze coins may have been imported as asource of bronze (cf. already Militký 2013, 43) does not rely on actual evidence; it is clear, nevertheless, that the cultural and economic transformations occurring in the Transalpine world in LT B2–C1 required asignificant increase of raw materials of all kinds, bronze most of all. In the case of glass, we may be sure that the material came exclusively from the Mediterranean (Venclováetal. 2009) and we may hypothesise that bronze ingots or scrap bronze may have travelled along the same route. This hypothesis is naturally contradicted by the virtual absence of other imported bronze ob‑ jects of clearly Mediterranean origin in central Europe in the 3 rd –early 2 nd century BC. Apossible explanation would be these coins arriving via the NW Balkans rather than NE Italy– virtually the same facies is found in both zones. In that case one would be surprised not to find in central Europe hoards similar to those of the Mazin type, but we need not stretch that analogy too far. The idea of coins as asource of raw material can be developed also in another direction. We need not doubt that gold and silver coinage imported to central Europe was re ‑cast into local coins (Motyková– Drda– Rybová 1984a; Fröhlich 2012). We might also expect that bronze coins crossed the Alps as intruders in assemblages of gold or silver coins and started their Transalpine life when recognised for what they were and sorted out while the coins of precious metals were recast under strict control. We cannot leave without comment the idea of mercenaries as the main bearers of coins to central Europe (leaving aside for now the not unimportant question of whether any merce‑ naries actually went to the Mediterranean from Central Europe). The soldiers of Hellenistic armies (Launey 1950/1987, 735–764, 753–763)– mercenaries or not– were paid two sorts of wages: misthos, the actual salary paid in silver (4–8 obols aday = 200–400 drachmas per year) 255II. THE THINGS AND THE THOUGHTS or directly in gold (Polyaenus iv, 6). The bronze coins found in central Europe can therefore have nothing to do with misthos. Apart from misthos paid at the end of service, the merce‑ naries were also given the so ‑called sitos, i.e. means necessary for their everyday existence. Though preferably given to them directly as vital goods, the commanders could also (in order to get rid of the complicated logistics) pass the responsibility for these supplies to private hands and give the mercenaries money (roughly three obols or their equivalent in bronze aday; Tagliamonte 1994, 160–163). In this way barbarian mercenaries obviously could have become familiarised with the concept of fiduciary coinage. Adifferent question is, why they would take hundreds of these coins (and only the coins, given the dearth of other imports!) with them back to central Europe. Though intelligence is not aquality most usually associated with mercenaries, they must have understood that their home societies had hardly developed economies based on fiduciary coinage while they were away in Mediterranean lands and that their coins were going to be useless in their homeland. Noeske (Čižmář– Kolníková– Noeske 2008, 687) explains this paradox by alegitimate suggestion that the Greek coins were actually brought to Němčice by the returning merce‑ naries in order to be dedicated in ahypothetical sanctuary. Although the excavations of the sanctuaries of nearby Roseldorf did not produce any obvious examples of coin dedications (at least none are mentioned in the reports: Holzer 2009; Dembski 2009) this may be caused by the diversity of cults exercised in Němčice and in Roseldorf. Ahint at the presence of asanc‑ tuary in Němčice lies, in Noeske’sopinion, in the numerous small bronze animal and human statuettes. The presence of statuary in the Němčice horizon settlements is far from clearly explained and acultic function is not excluded (Goláňová– Kysela 2019); association of coins and small figural bronzes would liken Němčice clearly with the contemporary Venetic sanctuaries such as Este Baratella, Villa di Villa (Maioli– Mastrocinque 1992), San Pietro Montagnon (Dammer 1986) or Lagole (Fogolari– Gambacurta 2001). Significantly for our topic, the few known find contexts of Greek coins in northern Italy are often sanctuaries (e.g. Gorini 1995, 94). It also makes sense– dedicating to adivinity was the best thing one could do with these nice little and otherwise quite useless objects. Should we admit, nevertheless, that the Greek coins found in Němčice were votive of‑ ferings dedicated in ahypothetical local sanctuary, we recognise their actual function, no matter how symbolic and we may presuppose ademand for them in order to satisfy the need to fulfil this function. The mercenaries become unnecessary at this point or they may have been just one of many means of coin transmission. The important thing is not the means but rather the reasons and in my opinion these have to be looked for in the function of the coins in central Europe. Acoin is one of two things: agiven amount of metal and an image confirming it. We have spoken above about metal but abrief mention must be made also about the images. The peak of the site of Němčice and also of the ‘Němčice horizon’ of Greek coin influx to central Eu‑ rope dates to avery particular moment in the history of Transalpine Europe– amoment of transformation of society, transformation of the economy requiring raw materials but also atransformation of La Tène Art, which is discovering at this moment asudden interest in imagery. It is the moment of transition between the enigmatic and polyvalent abstract and elitist art of the Early and Middle La Tène periods on the one hand and of the straightfor‑ ward, iconic and to acertain extent democratic art of the oppida period on the other (Kysela 2018/2019b, 151–153). The Greek coins bear very evocative and strong images which must have made the objects of extraordinary interest for their Transalpine observers. No matter in what way these coins crossed the Alps, the local society was at its most receptive in its cultural development to 256 THINGS AND THOUGHTS appreciate them as bearers of iconographic message. This may be just aguess without much material support (few reliable contexts, mostly not telling much; total absence of Greek coins in graves; with asingle exception210 none of the studied coins was pierced to be worn on the neck as was common later in the Roman Iron Age).211 One object proves clearly, nevertheless, that the inhabitants of the Middle La Tène Transalpine area were not indifferent to Greek coins as an iconographic medium– it is the faïence disc from Jenišův Újezd, gr. 138 (Castelin 1978; Venclová 1978; Venclová 1991, 110, 263) imitating probably mid‑3rd century Romano‑ ‑Campanian silver coins. The object makes clear that coins were objects of interest, that they were possessed and, if not available, imitated in materials other than metal. At the same time, its material– faience– is not at all characteristic of transalpine workshops and it is very likely an import from outside Central Europe (Venclová 1978; 1991). TO SUM UP Based on acorpus which is certainly not complete and assembled through some admittedly courageous premises, we can suggest some conclusions. 1) The influx of Greek coins into central Europe can be divided into three chronological horizons the first of which (in the 4th and early 3rd centuries BC) may have been to agreat extent due to adirect contact with the eastern Mediterranean and Greece proper while in the middle (‘Němčice’) and late (‘oppida period’) the mediatory role of Italy comes to the fore. 2) The coins of the first and third horizon penetrated in acapillary way the entire central Europe, in both cases probably from several directions. In the Němčice horizon on the oth‑ er hand the ‘Amber route corridor’ is extraordinarily well represented while in southern Germany the coins of this horizon are much rarer and in Bohemia almost absent. 3) The majority of Greek coins in central Europe (that is mostly the coins of the second and third horizon) reflect the contacts of central Europe with northern Italy or the NW Balkans and there is no need to imagine contacts between central Europe and the areas where these coins were issued. The contacts through which the coins crossed the Alps need not have started at an extremely early date– the foundation of Aquileia in 181 BC is not out of place as the starting point of the second horizon, even though many of these coins predate it by many decades. 4) The comparison and remarkable mutual correspondence of the individual regional spec‑ tra make any explanation based on specific historical events (e.g. Punic, Syrian or other wars) unlikely, let alone individual activities (such as the return of Celtic mercenaries); the idea of migrations of the Boii after their defeat by the Romans is completely out of place. As usual, event ‑based explanations fail to account for the full complexity of the material record when analysed in detail while aprocessual approach proves much more suitable. 210 Afourré of adrachma of Apollonia in Illyria from Stradonice, today in the Fürstenberg collection in Křivoklát castle (Militký 2015, n° 2441). 211 This practice was in general very uncommon before the beginning of the Christian era and only became common in the middle Imperial period. The few perforated coins from La Tène areas universally feature ahole in their centre rather than near the edge and therefore were intended for fixing to asolid surface rather than for suspension (Doyen 2013). 257II. THE THINGS AND THE THOUGHTS 5) The diffusion of Mediterranean coins into central Europe suggests afunction for them there though we cannot be sure what that function was. The various ideas presented here– raw bronze, intruders in hoards of silver or gold coins, local interest in the coins’ iconography– need not be mutually exclusive; they remain nevertheless only suggestions awaiting further confirmation or refutation. 6) There is ahuge potential in the study of Greek coins in temperate Europe. However, in order to fully exploit it, we need to see as broad apicture as possible, combining transalpine and Mediterranean evidence and searching for long term processes rather than trying to pin them to historical events. There is not enough space here for such adaring exercise: this chapter must be understood only as apreliminary sketch for astudy to be undertaken in the future (hopefully by ascholar more competent for the task). 258 THINGS AND THOUGHTS 4. The Thoughts The categories of visible and invisible, and material and immaterial imports (after Venclová 2002, 74–75) partly overlapped in the previous text. We have touched upon the immaterial as‑ pects of our topic e.g. in the case of coinage, technological changes like the presence of brass, or individual details like one ‑off distant imitations of Mediterranean vessels. In this section we will focus on other cases of spiritual or immaterial imports, albeit stud‑ ied through their material manifestations. The approach will naturally differ substantially from that employed in the study of the imports. As in the former case, we will of course try to understand to what extent the assumption of Mediterranean influence is justified and what kind of contacts it actually refers to. There is on the other hand no correct quantification method to measure the Mediterranean impact from this evidence and no way to assess it in an, even seemingly, objective way. Although the adoption and spread of foreign technologies, behaviour, and approaches is obviously evidence of much deeper external influence than just the acquisition of aforeign object, it is at the same time true that through the process of their adoption, these foreign behaviours stop being foreign. And the more advantageous they are, the more indispensable they become, the more they become embedded into the everyday reality of the receiving society the less relevant their foreign origin grows. The study of invisible immaterial imports therefore must not, or at least ought not, content itself with stating the foreign origin of astudied element but should try to follow the process and dynamics of its adoption, however out of our reach this objective may often be. We must naturally focus on the initial moments of the innovation, including the actual circumstances of its adoption (When? Where? Through which mediation? In asingle place or contempora‑ neously at several places within the receiving society?) and on the beginnings of its spread. Once an innovation becomes commonplace, we need not worry any more. In the following, the emphasis will be on cases pointed out in the past as examples of Mediterranean influence. The point is not that of exhaustively covering all such suggestions; Iwill rather concentrate on the most obvious or most frequently cited examples, believing that athorough analysis of them may provide more reliable grounds for further reflexion rather than alaborious verification (or rather falsification) of each and every marginal and extravagant proposal. TECHNOLOGIES BRONZE TECHNOLOGY According to asuggestion by Pavel Sankot (in Fránaetal. 1997, 93; Sankot 2002, 96) southern inspiration transformed bronze casting technology. In his opinion, beginning in LT B2 (once again apre ‑oppida period phenomenon), bronze alloys became characterised by ahigher content of lead, intentionally added to them for technological reasons, apractice which has its clear antecedents in the Mediterranean. Both of these facts are true: alloying bronze with lead in order to facilitate casting is acommon Mediterranean practice, and asudden surge in leaded bronze during LT B2 and C1 has been confirmed by new large ‑scale analyses currently 259II. THE THINGS AND THE THOUGHTS under way (Bohemia: personal communication A. Danielisová; Moravia: Kmošek 2020). What is missing is aclear causal link between these two phenomena. Although the assumption of asouthern derivation of the technology sounds fully logical, 212 for the moment there is no obvious way to prove it. ROTARY QUERNS The introduction of rotary querns was another 3rd century innovation with far ‑reaching so‑ cial consequences (for Bohemia cf. Waldhauser 1981, 211) which is sometimes accredited to Mediterranean influence (Bouzek 1990, 131; Venclová ed. 2008/2013, 78). We should point out however, that this statement is true only if ‘Mediterranean’ is understood more broadly than we may have got used to in the previous pages. The earliest rotary querns are attested in the western Phoenician areas in the late 6th century BC and appear with some regularity in northwestern Spain and in Britain in the 5 th century (Wefers 2012, 87–90); in southern France they were apparently present in the 4th century (Reille 2000). There is some evidence of the possible adoption of rotary querns in the Early La Tène in southern Germany as early as the 5 th century BC (Wefers 2012, 90–91), while they were indisputably present in some areas from the 4th/3rd century, only to become commonplace from the 2nd century on. However, the dynamics and mechanisms of their spread in the Transalpine area between the 4th and the 2nd century BC are still insufficiently understood– the querns appear in various regions and at various moments without an obvious connection between them which can be translated into aclear process of diffusion. As aresult, it seems that the idea of rotary grinding was known in the Transalpine area at least from the Early La Tène period, it is unclear, however, to what extent its spread beginning in LT C period may be due to these Early La Tène foci and to what extent it was reintroduced from the south during the Celtic migrations (Wefers 2012, 90–93). It is important to stress that if the latter was the case, the ‘south’ would have to mean ‘southwestern France and northwestern Spain’, not Greece or Italy with which we usually associate this term. As amatter of fact, in both Italy and Greece, which have so far been the main, if not the only regions representing ‘the Mediterranean’ for us, rotary querns do not appear until rel‑ atively late. In Italy they are vaguely discussed for ‘the 2nd to 1st centuries BC’ though without much direct evidence (Dal Ri 1996; Donner– Carzoli 1996), and even as late as the 1st century in Greece (Runnels 1990). This may perhaps partly be due to the state of knowledge; it is nevertheless interesting to note that in the settlement of Monte Bibele, only hand and saddle querns are documented while rotary querns are completely unknown (Morrone 1999). In this respect the Cispadan Boii surely lagged behind their Transalpine cousins (Boii or not). POTTERY PRODUCTION The appearance and spread of advanced types of pottery kilns, namely the two ‑chamber vertical kiln, is another phenomenon often linked with Mediterranean influence. The topic has recently been exhaustively treated in an exemplary way in the detailed study by Tomáš Mangel and Richard Thér (2018), including the history of research, and acareful evaluation of all available evidence on aEuropean scale. The two ‑chamber vertical kiln appears in LT B– C1 in the Carpathian Basin whence they apparently spread within the same period to more northwestern areas including Moravia, Bohemia, southern Poland, and southern Saxony. Only 212 Though it is worth pointing out that on the Atlantic coast the addition of lead starts in the Late Bronze Age (Brown– Blin ‑Stoyle 1959). Iam grateful to John Collis for this reference. 260 THINGS AND THOUGHTS in asecond stage in LT C2–D1 do the vertical two ‑chamber kilns spread westwards to south‑ ern Germany and the Main ‑Rhine region where they intersect with the distribution of kilns with double entrance spreading in the same period from the west. As to the actual origin of the two ‑chamber vertical kiln, the authors convincingly link this technical innovation with inspiration either from mainland Greece or from the Black Sea littoral; on the other hand, this kiln type seems unknown in northern Italy at this period (Mangel– Thér 2018, 176–186). The adoption of the potter’swheel seemingly forms part of the same technological package leading to standardisation and to acertain degree to the industrialisation of pottery manu‑ facturing in the later La Tène period after aphase of purely household production during the Middle La Tène period. However, the exact nature of this adoption is much more complex to follow than in the previous case. The first appearance of the potter’swheel in Central Europe is suggested for La Tène fine wares, followed by its much broader use in LT C2–D. Thér, Mangel, and Gregor (2015) focused on the dynamics of the use of the potter’swheel in acase study from the Chrudim district in eastern Bohemia, concluding that what changed between LT Aand LT C2 was not only the product line (fine pottery only, replaced with all ‑round production), the increase of production volume, and of clientele (the entire society instead of anarrow elite), but also some basic working procedures such as achange from aclockwise to acounter clockwise direction of throwing. This suggests asubstantial change in technological tradition. Unfortunately, our knowledge about the situation in the 4th and 3rd centuries is insufficient due to the lack of settlement contexts (let alone production facilities) of this period. We do not know to what degree the findings from the Chrudim district can be generalised and applied to the whole of Bohemia or central Europe; in any case there is no clear evidence that this technological change occurred due to influence from the Mediterranean. GLASS WORKING La Tène glass ‑working, thriving in central Europe from the mid‑3rd century onwards (Ven‑ clová 1990; Gebhard 1989a; Karwowski 2004; Venclová 2016; Rolland 2017), was largely indebted to Mediterranean influences. The Mediterranean influence seems rather limited from aformal point of view since the most characteristic La Tène glass ornaments– glass bracelets and ring beads– have no counterparts in the Mediterranean, other than those which can be directly connected with manifestations of the La Tène Culture there. Glass bracelets are widespread in Italy mainly in the northern parts occupied by the Celts and in the neighbouring regions (Vellani 1995). The vast majority of glass ornament types present in northern Italy correspond chronologically with the development of their coun‑ terparts beyond the Alps. In the history of research there has been one significant exception to this statement: the glass bracelets of type Montefortino / Haevernick 1 have often been considered the earliest type of glass bracelets, produced in Italy as early as in the late 4th/ early 3rd century BC and thus potentially the early model for the entire Transalpine produc‑ tion (Venclová 1990, 131–132).213 This topic was nevertheless complicated by the unreliable find contexts of these bracelets (including and beginning with the eponymous grave 30bis in the necropolis of Montefortino whose dating proposals oscillated between 4th/3rd and 3rd/2nd century BC). Areinvestigation of this group by Roberto Tarpini (2007) confirmed the high chronology at least in some cases (five specimens from the necropolis of Spina), pointing out 213 On the other hand, the seemingly very high dates of some of the finds of Italian bracelets of Haevernick’sgroup 8 (Venclová 1990, 132) can be dismissed based on areconsideration of the find contexts (Vellani 1995, 17). 261II. THE THINGS AND THE THOUGHTS at the same time the poor typological definition of this group and difficulties of attributing to it the Transalpine finds (Dürrnberg, Mistřín, Maňa– all probably dated only to the mid‑3rd century). The origin of La Tène glass bracelets would thus be assigned to the 4th/3rd century in the northern Adriatic in which (on both shores) all the reliable finds seem to concentrate, but interestingly enough rarely in La Tène Culture contexts (cf. list in Tarpini 2007). The issue of the problematic classification of type 1 bracelets was recently pointed out also by Joëlle Rolland (2017, 173–179) who insisted on distinguishing the transparent bangles (the actual type 1) from those with ayellow insert. The latter are in her opinion aspecific technological group characteristic of southern Germany, the Upper Rhine, and possibly the Eastern Alps with asecondary distribution in other regions including northern Italy (Rolland 2017, 344, fig. 129). The actual bracelets or rings of type 1/Montefortino on the other hand derive from smaller rings of transparent glass, appearing as early as LT Ain northeastern France and the Rhineland and used mostly as torc pendants though arare bracelet is present in the LT Agrave in Reinheim. At least some of the Italian finds (e.g. those from Spina) maintained the original small size while others (e.g. that from Montefortino) grew to adiameter permitting their use as abangle. Therefore, although the earliest types of glass bracelets appear sporadically in northern Italy in LT B, it need not have been there that this idea first occurred and it was certainly not there that the necessary technology developed (Rolland 2017, 173–179, 382–383). While the technology for glass bracelet production could have been known in Transalpine Europe from the late 5th century and occasionally used through the 4th–3rd century, at the very beginnings of LT C1 or mid‑3rd century it expanded into large ‑scale production (Venclová 1990, 131; Gebhard 1989a, 128; Karwowski 2012; Rolland 2017). Akey region in this adoption seems to have been the Middle Danube area (it may not be an accident that two out of three central European finds of Montefortino type bracelets were unearthed in precisely this area) where the most intense and most coherent activity seems to have occurred in the earliest period of glass working. Originally (LT C1a) confined to the territory of EnCE, and northwest of the Carpathian Basin, in asecond step (LT C1b) the glass production spread further west to Bohemia, and in particular to southern Germany and the upper Rhine (Venclová– Militký 2104; Venclová 2016; Rolland 2017, 365, fig. 131). According to this scheme, the merit of the Middle Danube region would not be that of simply hosting atechnological innovation but rather that of providing favourable conditions for its development; these may have included an appreciation of the producers (translated in practical terms into ademand for their products giving them the liberty for innovation) as well as the availability of the raw materials. This is an important point at which the Mediterranean comes back into play. The Transalpine glassworkers were, in fact, fully dependent on the Mediterranean in terms of raw glass acquisition. The composition analyses of La Tène glass throughout its production from LT C1 to LT D have so far attested its very uniform character, universally corresponding to the natron glass produced exclusively in the eastern Mediterranean. Natron, used as the indispensable flux in primary smelting of raw glass, was possibly only available in Egypt and distributed among its neighbouring regions (Venclováetal. 2009; Venclová 2016, 107–109; Rolland 2017; contra Karwowski 2004, 81; Karwowski 2012, 244–245). The analyses also revealed aclear difference in the chemical composition of two chronologically defined groups of glass (reversing the ratio of the oxides of Zirconium and Strontium cf. already Karwowski 2012, 247–249) suggesting ashift of sources from Egypt to Syro ‑Palestine sometime on the transition between LT C1 and LT C2 (Venclová– Jonášová– Vaculovič 2017, 77; Rolland 2017, 50–51, 77–109, 117–118). Joëlle Rolland (2017, 154–155) reasonably searched the cause of this transition in the political and economic disarray occurring in Ptolemaic Egypt during the re‑ gency for the child king Ptolemy V, including the defeat in the Fifth Syrian War (202–195 BC) 262 THINGS AND THOUGHTS and the Egyptian Revolt. Whatever the actual cause, we may be sure that it was an internal issue of the supply side and beyond the reach of the receiving La Tène communities. Direct evidence of the Mediterranean origin of raw glass is provided also by aseries of wrecks carrying from dozens to hundreds of kilograms of raw glass: Su Pallosu (Oristano, Sar‑ dinia, 3rd century BC, ca 20 kg of glass: Salvi 2006), Sanguinaires A(Ajaccio, Corsica; late 3rd to early 2 nd century BC, the cargo was estimated to be about one ton of raw glass: Cibecchinietal. 2012), Lequin 2 (Porquerolles, Var, F; general overview in Rolland 2017, 31, 62; Venclová 2016, 118). The location of all the wrecks providing this random but highly impressive sample shows that the ships were obviously heading for Gaul. There must have been numerous other ships with the same cargo before and after that. Finds of raw glass have been reported from two dozen La Tène sites though Natalie Ven‑ clová in her last discussion of the issue (Venclová 2016, 109–110) argues that in only four of these sites has the protohistoric date of these finds confirmed by composition analysis– these sites include Hengistbury Head (irrelevant for our purposes), Manching [M950], Stradonice [S950], Staré Hradisko [SH90], and Němčice nad Hanou [NH90].214 Other finds (also from sites such as Mšec or Třísov) are doubtful or have been identified as of post ‑Iron Age date. We do not have any hints as to how the eastern part of the La Tène world was supplied with this commodity. It was arguably not through Gaul considering the delay in glass production in the necessarily intermediary regions of Bavaria and Bohemia, if not in Gaul itself. The volume of early production in the Middle Danube region and the Carpathian Basin may not have been fully comparable with that of late LT C1 or LT C2 Gaul, to which the Sanguinaires Aship was bound when it sank with its ton of raw glass aboard (cf. Rolland 2017, 365–369, fig. 131–132), but still, tens if not hundreds of kilograms of raw glass must have reached the Middle Danube region (obviously inaccessible to any seaborne trade) in one way or another by land routes; we may never know whether it was through Italy or the Balkans. COINAGE As early as the first half of the 3 rd century BC, the central European La Tène Culture communi‑ ties obviously understood the concept and functions of coinage and were able to successfully emulate them in their own production of asmall series of high value gold coins serving prob‑ ably alimited range of purposes including large payments, the setting of values and possibly some prestige activities. By the second half of the 3rd century, the same communities had adopted acomplex seven ‑denomination bimetallic monetary system of the so ‑called Athena Alkidemos (AA) coinage and launched large ‑scale coin production, achieving probably over an extremely short period of time acomplete monetarisation of their society (Militký 2015b; 2018a; Smělý 2017). This holds true of the entire area of the so ‑called ‘Boii’ coinage in Bohemia, Moravia, Lower Austria, and southern Poland. This coinage is based on Macedonian models both in terms of weight standards and iconography (Militký 2015c; 2018a; Smělý 2017). Mace‑ donian models are also the basis of the neighbouring ‘Vindelican’ coinage in Bavaria (Ziegaus 2010) which was, however, supplemented with aquinar, adenomination originating in Rome and shared also with some coinages of eastern Gaul. The Macedonian character of the Boii coinage was clearly set already in its earliest phase with the Nike staters imitating the issues of Alexander the Great, and this continued through 214 Due to the specific nature of raw glass, it is listed in the Catalogue simply in terms of presence (i.e. asingle Catalogue entry) for the needs of the final statistics without any attempts at quantification. 263II. THE THINGS AND THE THOUGHTS the rest of the pre ‑oppida period in the Athena Alkidemos coinage of the Amber route corridor and, at least from the point of view of weight standard, also in Bohemian local issues; the coin‑ age of the oppida period in this entire territory still consistently built on the same principles; the sole foreign element in it was the localised introduction of the Biatec type tetradrachms in its very latest phase. The adoption of, and consistent adherence to, the ‘world standard’ set by the staters of, or in the name of, Alexander the Great makes perfect sense in view of the chronology of this adoption, reasonably dated to the first half of the 3rd century BC (Militký 2018a, 22–23). It is also consistent with the general south ‑easterly orientation of EnCE and eastern central Europe in general in this period as suggested by numerous other indices. And yet, in spite of these very clear logical facts, an almost religious conviction stubbornly persisted among some central European scholars that the origins of the Boii coinage are some‑ how to be linked with the north Italian Boii (Pink 1936, 16–19; Kolníková 1963; Hildebrandt 2001, 18; Kolníková in Čižmář– Kolníková– Noeske 2008, 668; Pieta 2008/2010, 251; Kol‑ níková 2009, 12). This idea following apurely historical narrative hypothesis could have been legitimate at the time it was first formulated by Pink, at apoint in the history of research when nothing was known about coin circulation in northern Italy and when the chronology of the Boii coinage was believed by today’sstandards to have been extremely late. Most present‑day researchers never adopted this idea (Militký 2015a, 42; Rudnicki 2013; 2015; Fröhlich 2016; Smělý 2017) while others abandoned it (Kolníková 2012215). Still afew points can be added to this unified rejection. The universal coinage of pre ‑Roman northern Italy is arather simple system consisting basically of two silver denominations: obols, and the so ‑called Po valley drachmas (dramme padane); there are also extremely rare half ‑obols and didrachmas, documented by only single specimens (Pautasso 1966; Gorini 2014). These imitate the early Massilian heavy drachmas with avariously barbarized head of Artemis on the obverse and astanding lion on the reverse, sometimes completed with an inscription or pseudo ‑inscription. Their weights shift from 3.4/3.3 g in the 4th–3rd century, through 2.6–2 g in the late 3rd–2nd century, down to ca 1–0.5 g in the 1st century. Pautasso’searly attempts to link the issues with each and every Celtic (and Ligurian and Venetic) tribe mentioned by the written sources have been recently debunked by Gorini (2014) who explains their large typological variety with anatural development of coin dies within arelatively limited number of coin series, most of which are in his opinion to be localized north of the Po– it was only the Insubri, Cenomani, the Veneti, and some Ligurians who struck them, but not e.g. the Boii. Finds of dramme padane are rare north of the Alps and only documented as far as Iam aware from WnCE and from the Alps (e.g. Kelheim: Pauli 1993, 56, Taf. 137: 5; Dölsach, Wattens, the Karlstein, Schwarzach: quoted by Schachinger 2017, 60; Kellner 1990, Nr. 2243–2248) while there are no finds in Bohemia nor further east, i.e. in the area of the Boii coinage. The key element for arguing aCisalpine influence on central European coinage are silver drachmas, believed to be the models for the Athena Alkidemos coinage by the proponents of anorth Italian origin. These coins combine the obverse motif of afemale head with, on the reverse, astanding figure with araised shield and spear, clearly derived from the archaising depictions of Athena Alkidemos (or rather Promachos) common in Hellenistic coinage. Accord‑ ing to Pink (1936, 16–19), the precise model for this reverse motif were the coins of PhilipVof Macedon (238–179 BC) while the obverse was inspired by the head of Minerva/Bellona from 215 Eva Kolníková– along ‑time supporter of this theory (cf. Kolníková 1963)– abandoned it in her latest works (Kolníková 2012, 15; Kolníková 2015, 258–260) in favour of aMacedonian origin for the AA coinage. 270 THINGS AND THOUGHTS Fig. 123: The Bratislava castle– overall plan of the 2008–2014 MÚOP and MÚOP & Via Magna exca‑ vation. Musilová– Minaroviech 2014. The largest structure discovered in the Bratislava Castle is ‘Structure I’ in the castle court ‑ yard (Fig. 123: B). The walls, again made of opus incertum, delimited aspace of approximately 23.5 × at least 9 meters (further north, the walls are damaged); the wall thickness was double that of Structures Iand II, that is 180 cm. The floor was once again made in opus signinum. The structure was badly damaged; the walls were mostly traced only as robber trenches and only afew patches of the floor are preserved in the interior. Still, we can argue that the entire area was actually asingle large hall as is clear from the decoration in the opus signinum– aborder of interlaced meander ‑swastikas delimiting aband scattered with rosettes (Vrteletal. 2014). 271II. THE THINGS AND THE THOUGHTS The Bratislava structures can be studied from anumber of different points of view, each of which should be treated separately: construction techniques, planimetry, and function. As far as construction techniques are concerned, there is no doubt as to their entirely Ro‑ man character. The walls of all the structures are built in opus caementicium with the caementa being stone rubble of relatively large dimensions. Opus incertum is obviously astandard Roman Republican construction technique used mainly in the 3 rd –early 1 st centuries BC but appearing occasionally down to the 1st century AD (Adam 1999, 127–129). Acertain crude character of its execution need not be asign of incompetence or barbarised technical skills, but rather of concession to technical and logistic constraints in the foreign environment. The same holds true for the most striking feature of the Bratislava constructions; in spite of their highly representative character, they do not seem to have been roofed with terracotta tiles even though the three and six feet thick walls would surely be sufficiently solid to support their weight. The grand architectural project necessarily involved some extremely demanding collateral logistic tasks unprecedented in central Europe, such as, for example, the burning of an enormous quantity of lime. The production of roof tiles would have been an operation whose complexity would by far surpass any other task in the construction, and churning out enough of them to cover all the buildings on the Castle Hill could have literally taken years. It may have been for this reason that shingles or even thatch were probably the solution even in the most prestigious of the Bratislava buildings. This is in strong contrast with the situation in Gaul in which fired clay tiling is among the first Mediterranean elements adopted in local architecture. The earliest proofs of tile roofing in Gaul (overview in Clement 2013, 101–116, fig. 97) comes from the Rhône valley and dates as early as the middle of the 2nd century (e.g. Lyon‑ ‑Rue du Souvenir), in the late 2nd and early 1st centuries; clay roof tiles slowly spread to central Gaul (e.g. Corent, Gondole) and from the middle of the 1st century (i.e. in the Caesar’stime) they became commonplace (e.g. in the basilica ‑like building in Bibracte). It is interesting to note that until that point, terracotta tiles only covered buildings made of wood and mudbricks, sometimes on rubble footings (indispensable in order to support the weight) and sometimes with plastered and painted walls; floors in opus signinum, though present, are relatively rare. Already from the earliest phases the tiles were produced locally, though perhaps initially (mid‑2nd– early 1st century) in asingle workshop (Clement 2013, 101–102). This makes sense in aregion in which there is relatively widespread knowledge and demand for such prestigious merchandise, no matter that they were obviously limited to aspecific group of customers. In the case of Bratislava this last statement can be turned upside down: apart from the logistical problems, the absence of terracotta tiles in Bratislava can be explained also by the insular and self ‑contained nature of the site lacking acompetitor in prestige. One feature of the Bratislava buildings dispels any doubts as to the degree to which persons of Mediterranean origin were involved in their construction, the basically universal use of opus signinum floors. Not only does their technology faithfully correspond to later Vitruvian instructions (Vitruvius vii, 1) and compare well with their Italian counterparts, but also the decoration exactly matches the patterns documented in Italy, Gaul, and elsewhere in the Mediterranean (for opus signinum in general cf. Vasale 2006). The different decoration motifs present in Bratislava include simple diamond ‑shaped tesserae (Vasale 2006, 48–49: 4 th century BC to the Augustan period); rosettes (Vasale 2006, 49: examples from the 3rd to the late 1st cen‑ turies BC); and ameander ‑swastika border (Vasale 2006, 50: from the 3 rd /2 nd century onwards). Some comparable examples include Aquileia (white swastika meander border, late 2nd–1st century BC: Medri 2000, 270, fig. 5); Forlimpopoli/Forum Popilii (4+1 tesserae rosettes, end of the 2nd century BC; Ceccaglia 2010, 316–317, nos. 34:2–34:3); Este (swastika ‑meander border with rosettes in acentral field, second half of the 1st century BC: Donderer 1986, 137, Nr. 1, 272 THINGS AND THOUGHTS Taf. 46); Padua (alinear rosette laid out in acentral field delimited by aswastika meander: Donderer 1986, 168–169, Nr. 7, Taf. 54); Zuglio (4+1 tesserae rosettes filling afield delimited by achessboard pattern frame); Calvatone/Bedriacum– Casa dei signini (swastika ‑meanders as aborder but amore complex net pattern in the main field, Augustan period: Grassi 2001, 416, fig. 4); Brescia (an identical motif, 1st century AD: Donderer 1986, 106, Nr. 3, Brescia 19, Taf. 34); Cornebarrieu, Haute Garonne, F (swastikas/meanders as borders in abath dated before the mid‑1st century BC: Viers– Veyssière 2012, 118–122). The Mediterranean impression is completed in Bratislava by fragments of painted plaster including Egyptian blue (Vitruvius vii, 7.11) which in spite of its exotic name was already pro‑ duced in Italy in the 1st century BC and perhaps even in the eastern Alps (Heck 2004). Little can be said from astylistic point of view given the extremely fragmentary nature of the remains but from atechnical point of view the execution of the Bratislava paintings is of excellent quality, comparable with that of the floors.219 In fairness, plastered walls would be more than expected in buildings in whose walls and floors so much effort was invested as those on the Bratislava Castle Hill; already in the Republican period wall paintings can be considered asort of ‘standard luxury’ in northern Italy (for aquick overview of wall paintings mainly in the first Pompeian style in northern Italy cf. e.g. Murgia 2017, 4–7) with some appearing even in places in Gaul, showing early signs of the adoption of Mediterranean architectural models (e.g. Lyon, Rue de souvenir: Le Bot ‑Helly 2003). If there was no reason to question the Mediterranean nature of the Bratislava structures from the point of view of their construction techniques, the story is somewhat more compli‑ cated when we turn our attention to their layout. The ground plans of structures II and VII (and, should the reconstruction be correct, also structure I) with their large halls articulated by rows of pillars or columns are distant from any canonical plan of aRoman or Italian house and hardly find any parallel in central or northern Italy (cf. overview in Gros 2001, 30–92; Verzár ‑Bass ed. 2001); they resemble more asort of petrified version of Late La Tène timber structures (Fichtl 2018; Danielisová– Kysela– Křivánek 2018). Margaréta Musilová rightly drew attention to formal similarities between the buildings in Bratislava and those on the Mag‑ dalensberg, it should be realised however, that the (also somewhat similar) earliest buildings in e.g. Zuglio and Moggio) are characterised by much less sophisticated building techniques and less generous ground plans. More examples can be added from the north ‑eastern periphery of modern day Italy, principally from the Alpine area (Castelraimondo). The proposed inter‑ pretation of structures Iand II as elite habitations and potentially for storage of traded goods (the two are naturally not mutually exclusive– quite the opposite) seems more than justified. The large hall in the castle courtyard is different in many respects. The ground plan of the building shares many similarities with awell ‑established Roman building type, the basilica. Calling the structure ‘abasilica’ would be too provocative, and therefore let us do so (with only asmall caveat). By using the term ‘basilica’ Ido not suggest the building’sfunction which in aMediterranean setting is usually understood as amulti ‑functional public structure dedi‑ cated to trade, civic activities, and the administration of justice (for an overview of basilicas cf. Nünnerich ‑Asmus 1994; Lackner 2008, 266–270); the term is for me merely ashorthand for a‘basilica ‑like structure’. Ido not apologise for such agross over ‑simplification since the exact meanings and applications of the term ‘basilica’ have been subject to some controversy both in the Mediterranean and in the Transalpine world. In fact, the ‘basilica ‑like structure’ from Bratislava is far from the only such building in pre‑/proto ‑Roman temperate Europe. 219 Iam grateful to Nicolas Delferriere (Université de Bourgogne) for kindly sharing his opinion with me. 273II. THE THINGS AND THE THOUGHTS Two buildings, very similar to the Bratislava structure in dimensions, design and (in the case of one of them) controversial interpretation, can be found at the Magdalensberg and at Bibracte (Fig. 124). Fig. 124: The basilika ‑like buildings of Magdalensberg, Bibracte, and Bratislava in comprarison. The Magdalensberg basilica is the least problematic of them. It is a30 × 17 m structure with a21 × 9 m stylobate (i.e. an inner foundation for acolumned gallery). After the building’sde‑ mise, the walls and floor were almost entirely removed but in the few places where their remains were preserved, they show that the walls were built of stone rubble in acement matrix supported by adense grid of wooden posts anchored in the ground; the outer walls were 1.8 m thick, the stylobate wall about 1 m thick. Only patches of the lowermost levels survived consisting mostly of layers of mortar or clay conglomerate (Mörtel‑ or Lehmestrich) on apebble base. The basilica formed part of alarge ‑scale project together with aforum and tabernae, the first urban form of the settlement of Roman/ Italian trades‑ and crafts‑ ‑people (Piccottini 1986; Alzinger 1985). Its construction dates to the initial phases of the Roman settlement on the Magdalensberg in the 50s/40s BC and it remained in use until the mid ‑Augustan period when it was replaced by anew basilica (for the chronology cf. Za‑ behlicky‑Scheffenegger 1986). The reconstruction of the Bratislava building as abasilica ‑like structure is of course to some extent conjectural– it is based on the dimensions and thickness of the perimeter wall as well as the assumption of the hall being asingle open space as suggested by the continuity of the decorative pattern in its opus signinum floor; on the other hand, we have no traces of columns or of astylobate on which the columns would have to be placed. The need to stretch the overall width to more than 9 meters would, however, make some kind of inner support useful. It is also worth pointing out how close the dimensions of the Bratislava ‘basilica’ (23.5 × >9 m) are to those of the one in Bibracte (24 × 14 m). By this Ido not suggest any genetic link between the two buildings but rather the possibility of acommon solution to asimilar problem. 274 THINGS AND THOUGHTS The Bibracte basilica has been the subject of considerable controversy. The structure ex‑ cavated between the 1990s and the early 2000s in the Pâture du Couvent is a24 × 14 m large rectangle with astylobate for agallery of 8 × 4 columns, of which some fragments have sur‑ vived, and presumably facing alarge court (Szabo– Timár– Szabo 2007). In contrast to the credible reconstruction, its interpretation is more questionable. The building dated to the period of the very initial Romanisation in ca 50–20 BC has been interpreted by its excavators as acivic basilica in the full sense of the term. The numerous weak points of this very straight‑ forward hypothesis were criticised in detail by J.‑Y. Marc (2011). According to him, the Bibracte building is too small to be effectively used as abasilica220 but more importantly, the presence of abasilica would in Marc’sopinion necessarily presuppose the adoption of aspecifically Roman or civic form of social organisation by the local community.221 For these reasons Marc opted for amuch more likely interpretation of the structure as aprivate building (obviously belonging to atopmost elite inhabitant), as also suggested by its replacement with aprivate atrium house after its destruction. Such large private halls are not unknown even in Roman architectural terminology and are classifiable as an oecus (Aegyptius) or a‘private basilica’ as mentioned by Vitruvius (vi, 5.2, vi, 3.9). Asearch for such private basilicas in Italy itself has been mostly unfruitful, the potential candidates being mostly doubtful, peripheral, or much later than the Vitruvian remark, let alone our structures of Caesarean or triumviral date (Gros 2004). This is however not areason to doubt their existence; Pierre Gros correctly stressed the inadequacy of available evidence. What is useful to realise is the status of the term basilica, apparently shifting between the categories of public and private (perceived in avery different way by the Romans than by ourselves: Russel 2015). After all, the very origins of the building type (as well as obviously its name) are in aprivate building (most probably the private dining halls of Ptolemaic palaces) which only on Italian soil acquired its final shape and an eminently public function (Wilson 2005). The public basilicae played acrucial role in perpetuating in their names the private glory of the builder and his family (Russel 2015) while private basilicae had important functions in the public activities of their owner, whether it was in feasting with his peers, in trade, or for political purposes such as receiving his clients. At this point we have to come back to Bibracte and Bratislava. J.‑Y. Marc was surely right to reject the idea of acivic basilica in immediately post ‑conquest Gaul; he may, however, have fallen into the same trap of applying Roman (architectural and social) categories to non ‑Roman areas. The owner could have known what the terms ‘private basilica’ or ‘Egyptian oikos’ meant and could have demanded abuilding corresponding to this definition. They could, however, also have simply demanded to build the kind of building that traditionally personified the status of the Gallic aristocracy using prestigious high ‑tech materials, as if the Bibracte ba‑ silica were only apetrification of the large houses of e.g. Batilly (cf. Fichtl 2018). It was only 220 Of Italian examples, it is closest in its dimensions to the basilica in the small mountain centre of Iulium Carnicum dated to the second half of the 1st century BC (Vitri 2001) and it exceeds the very early (3rd century) small 18.4 × 11.7 m probable basilica structure in Ostia (Lackner 2008, 136, fig. in page 267). 221 This opinion may seem somewhat extreme. In my view, there is no particular reason to consider the basilica functional only in aRoman civic milieu. Basilicae were multi ‑functional spaces, asort of ‘bad weather fora’, hosting probably first and foremost various trade activities; government was on the other hand administered in dedicated structures (comitia and curiae) and it was probably not until the Imperial period that the basilica became aregular venue of tribunals; the fact that the tribunes of the plebs seemingly had their office in one particular basilica need not be extended to other magistracies (cf. discussion in Lackner 2008, 269–270). 275II. THE THINGS AND THE THOUGHTS in the second stage that showing off Roman building techniques was followed also by the adoption of aRoman style house layout and thus space management. The interpretation of this ‘non ‑basilica’ would thus depend entirely on the interpretation of large Gaulish houses. This function need not have been just one and the same for all of the relatively numerous buildings; Stephan Fichtl proposes arange of possible interpretations encompassing agricul‑ tural, residential and public/representational purposes (for prestige aspects of Gaulish elite architecture cf. also Adam 2018, 6–8). Their regular presence in the elite enclosures and the remarkable resources and effort invested in their construction plead strongly in favour of the last two which are surely not mutually exclusive; in Gaul, as in Rome, large elite houses must have been from their definition semi ‑public rather than private. This public aspect of the structures would naturally have nothing to do with the unlikely ‘citizen’ character of local communities– it could play out completely according to local rather than Roman social rules. Bratislava and Bibracte tend to merge into asingle explanation in the preceding lines; it is not necessarily correct. While it is much more likely that an Aeduan aristocrat could have formulated his order as a‘private basilica’, in Bratislava the odds are surely higher in favour of the scenario of alocal building type in Roman style. It is also suggested by the non ‑canonical plans of other local houses. In the case of the Magdalensberg basilica we need not doubt its basilica functions, and we may conjecture (without the possibility of ever finding out) whether it was the model for the structure on the Bratislava Castle Hill. THE ORIGIN OF OPPIDA The question, to what extent the emergence of oppida in central Europe was inspired from the Mediterranean is the elephant in the room of research in Transalpine relations in the Late Iron Age. As already hinted in the Introduction, the idea that the emergence and floruit of oppida222 in Transalpine Europe was due to an impulse from the Mediterranean has along tradition in the history of research and has been reflected in one way or another by numerous scholars.223 The idea in itself is not illogical considering the pre ‑eminence of cities in the settlement landscape of the ancient Mediterranean which inspired Transalpine peoples in other respects. Apart from such considerations, however, there have always been very few arguments to support these hypotheses. The evidence invoked by V. Kruta is at the best circumstantial, and based on hypothetical narrative constructions whose various aspects have been tested in several places in the previous pages with little success. There is extremely little evidence on the nature of the settlements of the north Italian Celts and what there is does not suggest any significant degree of urbanisation (Vitali 1996; 2004; Kysela 2009; 2010); as Kruta’sarguments are broadly culture‑ ‑historical rather than strictly archaeological, it is unclear what other Mediterranean models we should be looking for and what features they should share with the Transalpine agglomerations. O.‑H. Frey (1984; 1988) argued in favour of agenetic link between the oppida and Italian towns mainly on the grounds of 1) the agger type of rampart, and 2) the absence of towers which are 222 Over recent decades it has become clear that the emergence of the oppida was preceded throughout Transalpine Europe by agglomerations, concentrating various specialised activities. Surprisingly, aMediterranean origin was never claimed for these settlements but only for the hilltop fortified oppida. 223 Werner 1938/1979; Böhm 1946, 37–38; Dehn 1961; Kruta 1978, 174; 1980, 199–200; 1988, 288; Kru‑ ta– Lička eds. 2001, 123; Kruta– Lička– Cession ‑Louppe eds. 2006, 205–206; Frey 1984; 1988; Bouzek 1989, 129; 2007, 136, 143–145; Čižmář 1993a; Drda– Rybová 1997, 109–114; 1998, 127–132; Peroni 2000; Rieckhoff– Biel 2001, 247–249; Polišenský 2003; Sievers 2003; Gnesotto 2006. 276 THINGS AND THOUGHTS characteristic of contemporary Greek fortifications but absent in Etruria;224 he also realized that in oppida just like in the Etruscan cities the ramparts encircle much larger areas than was actually occupied. On the other hand, the hypothesis has been criticised by numerous scholars either refuting aMediterranean role in the creation of oppida, or pointing out the absence of any reasonable evidence for it, and with no need for such amissionary intervention on the part of Transalpine societies (e.g. Collis 1995, 78; Vitali 1996, 340; Buchsenschutz 2000; Kaenel 2006; Fichtl 2004b; 2005; Kysela 2012a; Salač 2012, 338). To focus on the formal aspects pointed out by Frey: just like those of the oppida, the walls of Etruscan and Italic cities (Città murata 2008; Miller 1998; Qiulici– Quilici ‑Gigli eds. 2000) are indeed simple. They mostly encircle the entire plateau on which the city is built running along contour lines and following the configuration of the terrain. The rampart as arule consists of astone wall reinforced on the inside with amassive earth embankment (agger: Quilici 1994). The stone wall itself is made of simple rubble in the earlier periods (Veii: Boitani 2008) while the majority of 4th–3rd century walls are made up of two stone faces filled in with rubble (e.g. Tarquinia: Baratti– Cataldi– Mordeglia 2008; Vulci: Sgubini– Moretti 2008). Neither in Italy nor in central Europe is the agger an architectural choice; first and foremost, it is an indispensable structural element backing and strenghtening the actual wall. Its value is in being the simplest, no matter how laborious, way of stabilising astone wall of sufficient height while providing no other architectural or strategic advantage (Fontaine 1990, 374). In Italy, the agger is mostly asimple rubble and earth structure leaning against the rear of the wall without any inner structure 225 (Fontaine 1990; 2008; Miller 1998; Oakley 1995). In contrast, timber ‑framed banks were already present in Late Hallstatt and Early La Tène fortifications throughout Europe (for Bohemia cf. Motyková– Drda– Rybová 1984b, 368–371). There is no reason why the Late La Tène builders should have reached out to the Mediterranean for technological knowledge which had clear local roots. Although there is no sign of ramparts being employed between the early 4th and mid‑2nd century, the old fortifications built in this way were still standing, and– e.g. in Závist– some of the oppida timber ‑framed ramparts were built directly on top of them. Asimilar case can be made for the pincer gate, very common both in Italy and in the oppida. This gate type was widespread in early Italian fortifications (Brands 1988, 16–29), though from the late 4th century BC on it was being replaced or developed by other, more sophisticated types based on the same principle (gates with an inner courtyard; projecting bastions, etc.). As in the case of the agger, the pincer gate is alogical solution widespread in numerous areas and periods (including Early La Tène Europe) and thus hardly requiring an outside inspiration. In Bohemia alone, pincer gates can be recognized in the Early La Tène hillforts of Sedlo uAlbrechtic, Podšely, Starý Plzenec or Svržno (Chytráček– Metlička 2004, 136–139, 236, 242, 247–253); even the pincer ‑type principal gate (D) in the oppidum of Závist was built on top of an Early La Tène gate with bastions working on the same principle (Motyková– Drda– Rybová 1984b, 371–372). Just like in the oppida, the ramparts of Etruscan and Italic towns are quite uninventive and unsophisticated (Fontaine 2008, 212–217). They are simple linear walls of considerable length 224 The question is therefore not, whether the oppida were inspired by Mediterranean or not but only where this inspiration came from. 225 Neither internal wood nor stone constructions are usually attested though they are discussed in written sources (Varro, Lingua Latina v, 24); their absence may be due to the state of research. Iam aware of only two cases of embankments with an inner wooden framework in pre ‑Roman Italy, in Spina (Malnati– Sassatelli 2008) and Forcello di Bagnolo San Vito (Hummler– Carver 1988), both in the 5th century Po valley. In both sites the embankment or levee lacks the stone facing and protected the sites from floods rather than from sieges. 277II. THE THINGS AND THE THOUGHTS in which the only elements of active defence are the gates. There are almost no towers or bas‑ tions, not even wall indentations; postern and sally ports are rare. Such features appear only in the latest Italian pre ‑Roman fortifications (beginning in the 3rd century BC), as aresponse to the advanced siege warfare emerging in Italy at this period (Pulicnelli 2010). It is in this con‑ cept of an adequate reaction to specific ways of waging war that we have to read fortifications, rather than in terms of their simple formal similarity. The mutual resemblance between Italian ramparts on the one hand and those of Late La Tène Europe on the other is not caused by the derivation of the latter from the former but by the simple fact that both responded to the same needs, i.e. to similarly unsophisticated archaic ways of warfare and siege (cf. Kysela 2015c). Though the oppida and their defences provide convenient grounds for formal compari‑ son, such an approach is too obviously divorced from reality. Oppida were only one of many components of complex settlement systems consisting also of open agglomerations, farms, enclosed households, and small hilltop settlements; more importantly, they were an artefact produced and utilised by aparticular human society and it is from this point of view that their origin should be studied. Nonetheless, when discussing settlements as asocial product rather than amaterial fact, Mediterranean archaeology becomes awelcome inspiration. Over recent decades, the topics of urbanisation, and of the origin, emergence and development of towns have been asubject repeatedly and vigorously discussed in both Greek and Italian archaeology (e.g. Müller ‑Karpe 1962; Formazione Lazio 1980; de Polignac 1984; Morris 1987; Rossi ed. 1988; Peroni 1996; Daamgaard Andersenetal. eds. 1997; Pacciarelli 2000; Reddéetal. eds. 2003; Lafon– Marc– Sartre 2003; Dinamiche di sviluppo 2005; Osborne– Cunliffe eds. 2005; Riva 2010). Significantly, what most of these studies understand as a‘city’ is not the actual agglomeration, but rather the political and social unit and rather than about cities we are therefore talking here about (city) states, poleis. After all, the notion of acity being ‘people rather than houses’ is acommon theme of ancient written sources (e.g. Thuc. ii, 77.7; Cicero Ad Atticum viii, 11.3; Paus. x, 4. 1–2). Consequently, Mediterranean archaeologists (curiously) often study the issue of city formation on the grounds of burial evidence (e.g. Müller ‑Karpe 1962, 61–62; Morris 1987; Peroni 1996; Pacciarelli 2000; 2010) while settlement evidence plays rather an auxiliary role (Peroni– Di Gennaro 1986). Though the creation of the actual settlement agglomeration surely has its significance (sometimes it is considered the ‘proto‑ ‑urban threshold’, e.g. in 9th–8th century BC Etruria), the actual ‘urban’ stage in the opinion of most scholars was reached in the 7 th or 6 th century when the cities were equipped with amenities such as road paving and sewers testifying to acommunity if not public projects, and most significantly with the appearance of central urban temples which is aclear sign of the emergence of community identities (cf. for the Greek world de Polignac 1984). Apart from polis communities centred on acity (‘city states’), another form of social inte‑ gration is well documented in the Italic world (Cristofani 1978, 88–102), namely ‘territorial states’ e.g. in Samnium (La Regina 1989; Lloyd 1995; Tagliamonte 1996; Rainini 2000; Giampaola 2000), Umbria (Malone– Stoddard eds. 1994; Dench 1995; Bradley 2000), but probably also in the territories of the north Italian Celts, mainly the Boii (Cato apud Plin. NH iii, 116; Polyb. ii, 17. 9–12; Livy xxxiv, 22; Peyre 1992; Vitali 1996; 2004; Schulze ‑Forster 1998; Kysela 2010a). These societies were organised into clans (pagi) and tribes (populi) forming anation (nomen) with dispersed hierarchically undistinguished settlements (vicatim or kata komas) and central functions often exercised by sanctuaries. True urban centres appeared only after integration of these regions into the Roman world (La Regina 1989; Faustoferri– Lloyd 1998). Traditionally looked down upon as underdeveloped (Aristotle Politics ii, 2.1261) these societies have received fuller appreciation in the last few decades with abetter understanding of their adaptation to specific territorial conditions. 278 THINGS AND THOUGHTS The shaping of this particular settlement structure along with aprofound transformation of local society in Samnium was analysed in detail by Gianluca Tagliamonte (1996, 128–136). According to him, the motor for these transformations was war which reinforced social co‑ hesion and self ‑definition vis ‑à‑vis others, and prompted the integration of warriors into fel‑ lowships, active locally or as mercenaries abroad, and constituted local and regional identities. Rural sanctuaries were natural meeting points for representatives of the various dispersed communities. As already pointed out by Collis (1995; 2012), settlement systems of Late Iron Age Transal‑ pine Europe are in many respects akin to the settlement models of pre ‑Roman Samnium or Umbria. The social model described by Caesar in Late La Tène period Gaul is that of ‘tribal’ states– civitates– based on aterritorial rather than acentralising principle and led by an oligarchic nobility controlling production and trade, and pursuing their political goals by interacting within acomplex net of kinship, client, and ‘feudal’ relations (Fichtl 2004a; Dobesch 2004, 27–28; Collis 2012). The roots of these political formations most probably go back to the Middle La Tène period (Fichtl 2004a). The most comprehensive model of social development in this crucial period was put forward by Stéphane Marion (2007; 2014) based on evidence from north ‑eastern France. Between LT B2 and C1 Marion observed atransfor‑ mation of the burial rite with acertain democratisation of grave goods. The original emphasis on warrior burials diminished from LT C1 along with the appearance of rural sanctuaries like Ribemont ‑sur ‑Ancre and Gournay ‑sur ‑Arronde suggesting ashift of certain aspects of self ‑representation from the private to the public sphere. In Marion’sopinion, war was the prime mover of social integration, the catalyst of economic rationalisation, the impetus for more efficient territory management. These processes going hand in hand with demographic growth and social differentiation ultimately forced the appearance of agglomerations (e.g. Bobigny in the case of the Île ‑de ‑France). This scheme not only accounts for the emergence of the equites whose polycephalic authority over their civitates was described afew generations later by Caesar but also corresponds with the social dynamics reconstructed in Samnium (Tagliamonte 1996, 128–136). Marion’sscheme though developed for France can, with the necessary modifications and caveats, be applied also to other regions of the La Tène world including Central Europe. The transformations of material culture, economic strategies, and settlement pattern there are roughly the same as in the Île ‑de ‑France (cf. industrial zones, ‘colonisation’ of previously vacant territories like southern Bohemia, etc.: Venclová 2001; 2009; Venclováetal. 2008; Schäfer 2002; 2010; Salač 2011). The establishment of polit‑ ical authority can be unequivocally deduced from the strictly controlled coin production: Smělý (2017) argues on numismatic grounds in favour of the existence of full ‑fledged state formation in the entire territory of EnCE in the 3rd century BC.The evidence from Gaul (and from Samnium) demonstrates the crucial role of sanctuaries in this process (Marion 2014; Barral– Nouvel 2012; Fernández ‑Götz 2012) and it is worth noting that also in the ear‑ liest agglomerations in central Europe sanctuaries seem to have been of utmost importance (Manching: Wendling 2013; Wendling– Winger 2014; Roseldorf: Holzer 2014). In Central Europe unlike in Gaul we are not able to reconstruct the extent of these political entities, to draw political boundaries, and thus to understand the relations between the specific sites, but this is not atopic which should matter at this point. What is important, the phenomena we observe in Central Europe are essentially identical in their material manifestations with those which in western Europe led to the appearance of early states. The transformation of the settlement pattern with the appearance of new settlement categories– agglomerations first and oppida afterwards– is aconstant symptom of these social developments and can be understood as their natural product with no need 279II. THE THINGS AND THE THOUGHTS for external inspiration, no matter if their establishment was dictated by practical needs or if there were other elements at play, e.g. prestige as recently convincingly argued by Moore and Ponroy (2014). Prestige and immaterial, seemingly irrational, reasons behind the foundation of oppida merit aspecial place in this discussion. As realised already by Woolf (1993), one of the functions which communities building them endow in their fortifications (i.e. one of the few forms of monumental architecture present and recognisable in prehistoric Europe) is to display their potency and their very existence. Apart from practical reasons which need not be doubted, asignificant point is that of the legitimisation of power through reference to earlier (in our case Early Iron Age) monumental constructions. Instances in which an oppidum comes to existence through fortification of apreviously open agglomeration are rare (Manching, Besançon?, Chateaumeillant?). More commonly we observe asparse ill ‑defined human presence in the future oppida (Závist, Stradonice, Hraza‑ ny?, České Lhotice, Plavecké Podhradie, Velem, the Donnersberg, Bibracte?). Very often– and frequently in the earliest oppida in central Europe– the oppida were founded in the location of an Early Iron Age settlement: Ha/LT Ahillforts: Závist, the Steinsburg (Peschel 2005), the Staffelberg (Abels 1987/1988), Corent (Milcentetal. 2014); Bourges (Augier– Krausz 2012); Ha /LT Aoccupation: Hrazany (Jansová 1982; Čtverák 2002); Staré Hradisko (Danielisová 2014), Bibracte (Guillaumet 2003); tumuli: Manching (Wendling 2013, 463–466), České Lhotice (Danielisová 2010). In some cases, acontinuous human presence can be hypothesised in these hillforts from the Early to the Late La Tène period, with areduction and change in character in LT B and C: Vladař, Závist?, the Steinsburg, Staffelberg, Donnersberg, Martberg, Dünsberg, Milsenburg, etc. To focus on Bohemia, its southern half is aregion in which the hillforts always played ahighly important role (Chytráček– Metlička 2004; John 2003). The appearance of oppida here is only another, not the first and not the last, chapter of this long tradition. The oppida may differ from Early La Tène hillforts, but so did 2nd–1st century society from that of the 5th century BC.Therefore, in my opinion, oppida were afully organic and func‑ tional component of the Late La Tène settlement landscape though operating more within the rationality of that period rather than ours. 286 THINGS AND THOUGHTS Overall, 770 artefacts have been classified as Mediterranean imports in our working area of narrower Central Europe (Fig. 125). Out of these, 25 artefacts date to the pre ‑oppida period, i.e. 4th–mid 2nd century BC (three in Bohemia, three in WnCE, and as many as 19 in EnCE). The rest, i.e. 745 objects reached Central Europe in the century between the middle/second half of the 2nd and the middle/third quarter of the 1st century BC.These consist of 301 objects in Bohemia, 245 in EnCE and 199 in WnCE. Within the oppida period we may further distinguish avery late horizon of Roman impact manifested in the oppida of Bratislava and Devín which are in all probability basically mutually exclusive with the other sites from achronological point of view. The site of Wien ‑Rochusmarkt/Kundmanngasse on the other hand dates, in my opinion, at least partially to aslightly earlier horizon and is contemporary with the late phases of sites like Stradonice, Staré Hradisko, and Manching). Should we set Bratislava and Devín with their 66 imported objects apart, the number of oppida period imports in EnCE drops to 179, becoming closer to the total of WnCE. THE PRE ‑OPPIDA PERIOD The majority of imports of the pre ‑oppida period are glass beads. Their under ‑representation in WnCE may be partly due to the fact that unlike this region for which the only available find overviews are those of Manching (Gebhard 1989) and later Egglfing (Uenze 2000; 2005), the glass ornaments in the eastern parts of our study region have been under the constant watchful and proactive control of Natalie Venclová, Maciej Karwowski, and most recently Hana Čižmářová. These inconspicuous beads are moreover less likely to be presented in summaries or preliminary excavation publications, unlike bronze vessels or finger rings. Still, it is interesting to note that even the types of glass beads identified as Mediterranean imports in EnCE and Bohemia hardly ever appear in WnCE. The explanation may indeed lie in the distinction between east and west, though not necessarily on apurely material level. The identifications of Mediterranean bead types by Natalie Venclová and Maciej Karwowski largely depend on the classification of beads in the northern Black Sea region by E. M. Alek‑ seeva (1975/1978/1982). No comparable comprehensive study is as yet available for Italy and the western Mediterranean (the collection of data by Th. E. Haevernick focuses mainly on the Early Iron Age). If there are specifically Italian types of glass bead (theoretically more likely to appear in WnCE than in EnCE) we are simply unable to identify them. Even so the contrast is enormous even between EnCE with 15 or 17226 glass beads and Bohemia with amere two, by themselves, pretty uninspiring, items. When dealing with more obvious imports (Figs. 126, 127), we struggle to find at least some coherence in the very motley contact indicators of the pre ‑oppida period: the two seeds from Roseldorf are unique curiosities, not only in the context of our study. The bronze vessels (or fragments thereof) from Stebno ‑Nouze and Němčice nad Hanou, the surgical instruments from Munich ‑Obermenzing, and the helmet cheek ‑piece from Manching all find some coun‑ terparts elsewhere in the Transalpine world (cf. Mannersdorf, Waldalgesheim, Hurbanovo, etc. for bronze vessels; Slatina nad Bebravou for, remarkably, both the medical instruments and the armour fragment). In reality, however, these objects have little in common in terms of their chronology and provenance: an Etruscan Hellenistic vessel from Stebno, and an Etruscan Late Archaic one from Mannersdorf, Greek/Macedonian vessels in Waldalgesheim, Němčice, Hurbanovo, etc.; an Italian cheek ‑piece from Manching as opposed to possibly aTarentine armour strap fitting from Slatina… 226 Two of them are residual finds from the oppidum of Staré Hradisko. [Document text truncated for crawler view.]