Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The archaeological record of the Galatians in Anatolia, 278-63 B.C. Nixon, Lucia Frances 1977

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


UBC_1978_A8 N59.pdf [ 5.68MB ]
JSON: 1.0094439.json
JSON-LD: 1.0094439+ld.json
RDF/XML (Pretty): 1.0094439.xml
RDF/JSON: 1.0094439+rdf.json
Turtle: 1.0094439+rdf-turtle.txt
N-Triples: 1.0094439+rdf-ntriples.txt
Original Record: 1.0094439 +original-record.json
Full Text

Full Text

THE ARCHAEOLOGICAL. RECORD OF THE GALATIANS IN ANATOLIA, 278-63 B..C. by Lucia Frances Nixon A.B., Bryn Maw College, 1971 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR -THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Department of Classics s, We accept this thesis as;conforming to the required standard THE..UNIVERSITY OF BRITISH COLUMBIA August, 1977 © Lucia Frances Nixon, 1977 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of . Glassies The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date pphmary 3fi r 1 Q7ft ii ABSTRACT The Galatians were a group of Celts who arrived in Anatolia from the west in 278 B.C. According to the historical sources, they earned their livelihood by plundering and by serving as mercenaries in the eastern Medit erranean. Ancient authors state that the Galatians constituted a definite threat to the cities of western Asia Minor before they were settled in central Anatolia. Galatia became a Roman province in 25 B.C.; by that time, the Galatians had been thoroughly absorbed by the local population. The purpose of this paper is to see what archaeological evidence exists for the presence of the Galatians in Anatolia during the pre-provincial per iod, and how that evidence can be obtained. Three types of evidence are examined: pottery, burials and grave goods, and forts and settlements. Galatian pottery is still a controversial sub ject requiring more study and excavation. Only one burial site, Karalar, can definitely be identified by an inscription in Greek. The evidence from this site suggests that the Galatians adopted various types of Hellenistic tomb architecture and that they placed a fundamentally Hellenistic selection of grave goods within their tombs and graves. Galatian burials are there fore hard to distinguish from ordinary Hellenistic burials in Anatolia. Three tores and three fibulae from burials at Karalar, Bolu, and Bogazk5y are probably Celtic; that there are so few of them suggests that they had been imported from Europe, and that the Galatians were not themselves metal workers in the Celtic tradition. Such objects cannot be used as the sole means of identifying Galatian burials. The situation is little better for forts and settlements. Some have been identified because they were inhabited by literate people before or after the arrival of the Galatians; others have been suggested because of the likelihood of their location. Settlement seems to be more dense west of the Halys but more surveys and excavation are necessary to test this emerging pattern. So far, the pre-provincial period has yielded little in the way of archaeological evidence for the presence of the Galatians in Anatolia, despite the solid background provided by the historical sources. The Gala tians had little connection with the European Celts and adapted easily to local customs. This capacity for adaptation makes it difficult to say what is Galatian and what is Anatolian Hellenistic. Only further work in the field can remedy this state of affairs. iv TABLE OF CONTENTS Introduction 1 Historical Outline 8 Pottery 35 Burials and Grave Goods 44 Forts and Settlements 69 Conclusion 87 Figures and Illustrations 89 Bibliography 106 V LIST OF FIGURES AND ILLUSTRATIONS 1. "Galatian" pottery. Maier, Jdl 78 (1963) 235, fig. 14. 2. Shapes of "Galatian" pottery. Maier, Jdl 78 (1963), pp. 221, 222, figs. 1, 2. 3. Map of Galatian, "Galatian", and other tombs. 4. Kurtkale. Mansel, Belleten 38 (1974), fig. 22. Igdir. T6kg6z, TurkArkDerg 22 (1975), p. 157, fig. 3 Gemlik. Mellink, AJA 71 (1967), p. 173, fig. 1. 5. Mudanya. Mansel, Belleten 38 (1974), fig. 16. Kirkagac. Mellink, AJA 67 (1963), p. 189, fig. 1. 6. Map of Karalar. Arik, Turk Tarih Arkeoloji Etnografya Dergisi 2 (1934), pi. 2. 7. Karalar C. Mansel, Belleten 38 (1974), fig. 21. Gordibn I. Young AJA 59 (1955), pi. 81, fig. 4. 8. Karalar A. Arik, TTAED 2 (1934), pi. 8. 9. Karalar A. Arik, TTAED 2 (1934), pi. 14. 10. Kucucek/Aykazi. Firatli, Belleten 17 (1953), fig. 5. 11. Karalar B. Arik, TTAED 2 (1934), pi. 11. 12. Bolu East. Firatli, AJA 69 (1965), pi. 94, fig. 5. 13. Buckle from Bolu West. Firatli, AJA 69 (1965), pi. 95, fig. 7. 14. Gordion II. Edwards, Expedition 5 (1963), fig. 34. Besevler. Hoepfner, AthMitt 86 (1971), fig. 4. 15. Map of possible Galatian hill-forts and settlements, after Mitchell, D.Phil, thesis, Oxford, 1974. 16. Peium. Mitchell, AnatSt 24 (1974), p. 67, fig. 8. 17. Karalar. Arik, TTAED 2 (1934), pi. 19. vi ACKNOWLE DGEMENT I would like to thank Professor E.H. Williams for suggesting the topic of this paper, and Professors J.A.S. Evans and J. Russell for acting as thesis supervisors. All three have made helpful and appropriate comments, and have provided much-needed encouragement. I am particularly grateful to Stephen Mitchell for helping me with ref erences and for making his thesis on Galatia available to me. Additional thanks are due to Professors Machteld J. Mellink and Fred Winter, Peter Kuniholm, and David French for help with bibliography. David French also let me see the draught of the unpublished guide to the vilayet of Ankara. The University of British Columbia provided generous financial aid in 1971-72 and 1972-73 through the H.R. MacMillan Foundation, which made it possible for me to visit Galatian and "Galatian" sites when I was in Turkey in 1973. I would also like to thank my family and friends for all their help and support. 1. INTRODUCTION This paper is concerned with certain aspects of the archaeological record of the Galatians, a group of originally European Celts, from their arrival in Asia Minor in 278 B.C., until Pompey's eastern settlement of 63 B.C. The Galatians themselves were an illiterate people, but their activities in western Asia Minor and central Anatolia are fairly well documented in Greek and Roman historical sources. Thus it is possible to use the historical evidence as a basis for archaeological investigation. The task, then, is to combine both historical and archaeological infor mation, in order to obtain some notion of the Galatian cultural identity in Anatolia. The closest analogy for this kind of problem is probably that of the Kimmerians, who preceded the Galatians in Anatolia by a little over four centuries. They are mentioned in Greek, Assyrian, and Biblical sources, although they themselves were illiterate. The Kimmerians may have been .. driven from their homeland in the south Russian steppes by the Scythians; in the early seventh century they acted as mercenaries for Urartian kings, and later they constituted one of the nuisances which led to the downfall of the Assyrians. They posed a threat in Lydia and in Phrygia, where they made a raid on Gordion. It is thought that ultimately they settled in Cappa-docia."'' Archaeologically, the Kimmerians are very difficult to detect; in fact, without the evidence of the historical sources, their presence in Anatolia might well have gone unnoticed. The Kimmerians brought with them no distinctive style in art or weaponry, and were not apparently the build ers of substantial settlements. 2. The case of the Galatians is similar. They were described by the hist orians of other cultures; they had been on the move for long enough to have shed most of the characteristic cultural traits which might have linked them with their European counterparts. Thus they could easily have adapted to local customs. It is the evidence from historical sources, with the confirm ation from inscriptions and sculptural dedications, that makes the Galatian presence in Anatolia during the two centuries in question a fact rather than archaeological supposition. Nonetheless, an attempt must be made to piece together the available archaeological evidence. There are two main difficulties in doing this. In the first place, a great deal of work remains to be done on the archaeology of the Hellenistic period, especially in areas such as Anatolia which were not part of the cultural mainstream. The relationship between general hel-lenisihg influence and residual Phrygian elements needs to be firmly estab lished. At present it is hard to say how the Galatians reacted to the third century B.C. culture of central Anatolia in terms of what habits of their own were discarded or modified, since the material existence of the local people of the area is itself ill-defined. In the second place, it is difficult to be precise about the culture the Galatians brought with them to Asia Minor, although much is known about the European Celts and their relationships with the classical world in earlier periods. The Galatians were among the Celts who had penetrated into the Balkans by the mid-fourth century B.C., and who later invaded 2 Macedonia and attacked Delphi. This knowledge does not help us much, as there is no detailed study of the cultural relationship between these Celts and the peoples of the lower Danube on the other. It is not within the 3. scope of this paper to resolve the latter questions, but it is necessary to mention the gaps in our knowledge which complicate the Galatian problem. Even after they had reached central Anatolia with whatever culture they had retained during their wanderings, the Galatians seem to have lived from hand to mouth; it is possible that such a way of life was reflected by a correspondingly ad hoc culture. While the Galatians were not pastoralists with a regular pattern of transhumance (and therefore did not evolve the streamlined culture of the true nomad), they took a long time to adapt to a settled life which did not require sporadic fighting and pillaging. This may explain the spottiness of the archaeological record, and the relative lack of obvious cultural traits which can be labelled Galatian with cert ainty. To sum up the difficulties involved in undertaking to establish the archaeological characteristics of the Galatians: we lack the basic archaeo logical sequence for the place and time under consideration, that is, the period between the third and first centuries B.C., and the area of modern Turkey from Sivrihissar to Yozgat, or central Anatolia. Furthermore, we are dealing with a people whose cultural identity had become blurred before they entered this territory, and who failed to produce an easily recognizable culture of their own before Galatia was made a Roman province. By and large, the study of Galatian archaeology suffers most from never having been regarded as a problem, or rather as a subject worthy of system atic investigation. Evidence for Galatian material culture has accumulated almost by accident: few archaeologists have set out to discover the Gala tians, and most, when confronted with untidy Hellenistic debris on sites in central Anatolia, identify it as Galatian and remove it in order to 4. excavate what they are really looking for. Such a situation is understand able in Anatolia, given the splendours of the Hittite and Phrygian periods, to say nothing of the classical fringes of western Asia Minor, but it is hardly a good situation from the point of view of pre-provincial Galatia. The first attempts at reconciling history and archaeology were made in the nineteenth century by scholars such as Ramsay. He was preoccupied with the Jerusalem Itinerary and Roman road systems, but he suggested several sites as possible Galatian hill-forts. Others of his era tried to do the same but were certain that the Galatians inhabited magnificent cities, and often erred in their attributions to the detriment of Hittites and Phrygians. In the early twentieth century, scholars became interested in the similarities of European Celtic pottery and "Galatian" pottery, both local offshoots of typically Hellenistic wares. The connection between Europe and Asia Minor seemed more definite. In the 1930's, Kurt. Bittel visited sites such as Pessinus and the newly located Tavium, trying to establish the nature of the Galatian occupation. Celtic fibulae turned up at Bogazkby and indicated some link with Europe. Also in the 1930's, the Karalar ex cavations were conducted and published by Remzi Og"uz Arik. Karalar was then, and is now, the only site identified beyond doubt as Galatian. It is important for its three tumuli and its hill-fort, but unfortunately provides little evidence for Galatian habitation. Arif Miifid Mansel was particularly interested in the corbelled tomb at Karalar and its possible connections with Thracian and Macedonian ex amples. Since the 1940's he has continually tried to establish the degree and extent of the influence of these areas on western Asia Minor. Since then a number of tombs has been found in Bithynia; these show that Thrac-. ian'.and Macedonian influence was probably felt in Asia Minor before the arrival of the Galatians. More work is needed in these peripheral areas to determine what non-Greek peoples were living there and to discover how closely they were affiliated with Europe itself. "Galatian" pottery continues to nag at the conscience of archaeologists. Having analyzed the type in detail, Ferdinand Maier asserted that Galatian pottery is Galatian only in provenance, making it far less easy to call the rough Hellenistic settlements on sites such as Gordion purely Galatian. Frederick Winter, who studied the Hellenistic pottery at Gordion, is in total agreement with Maier and subscribes to the now prevalent belief that the Galatians were an extremely adaptable people, and therefore difficult to pin down in terms of distinctive types of pottery and architecture. But Bittel staunchly continues to believe that "Galatian" pottery is Galatian, and the controversy remains. More recently, Stephen Mitchell has completed a D.Phil, thesis en-3 titled The History and Archaeology of Galatia. He examines the historical evidence from the point of view of the Galatians—the first time that this has been done—and proceeds to a detailed investigation of the topography and archaeology of Galatia. His study is of immense value to students of the area, since it combines sound historical research with a well-organized archaeological survey. Mitchell manages to extend our knowledge of the Galatians without losing sight of the limitations of the available evidence; he is particularly helpful on the subject of hill-forts and possible un fortified settlement sites. It is time now to consider the scope and restrictions of this paper. What follows is not a comprehensive study of pre-provincial Galatian arch aeology, for such an opus would be far beyond the competence of the present 6. writer. Rather, this paper will examine three types of archaeological evidence—pottery, burials and grave goods, and forts and settlements, to clarify the methods used to amass each of these types of evidence. The problems involved in defining the material culture of the Galatians recur in any attempt to focus on the archaeological identity of a marginal people. This is why the Kimmerian problem, discussion of which might initially have seemed irrelevant, was mentioned at the beginning of this introduction. It may not in fact be possible to abstract any general principles from the uneven evidence at hand, but it is necessary at least to try. The first step is certainly clear. In this case, the historical evid ence is relatively coherent, and so, before any examination of the archaeo logical material, a brief chronological outline will be given, in order to put the Galatians in their historical context. This will include a short note on the artistic representations of the Galatians. Then we will pro ceed to the archaeological sections on pottery, burials and grave goods, and forts and settlements. 7. FOOTNOTES TO THE INTRODUCTION 1. E.D. Phillips, "The Scythian Domination in Western Asia: its Record in History, Scripture and Archaeology", World Archaeology 4 (1973-73) 129-139. See also Rodney Young, "The Nomadic Impact: Gordion", pp. 52-57 in Dark Ages and Nomads c_. 1000 B.C., edited by Machteld J. Mellink, Nederlands Historisch-Archaeologisch Instituut, Istanbul, 1964. 2. T.G.E. Powell, The Celts, Thames and Hudson, London, 1958, pp. 22-23. See also C.F.C. Hawkes' excellent summary, "The Celts: Report on the Study of their Culture and their Mediterranean Relations, 1942-1962", pp. 61-79 in Le_ Rayonnement des Civilisations Grecque et Romaine sur  les Cultures Peripheriques, VIII Congres International d1Arch£ologie (Paris 1963), Editions de Boccard, Paris, 1965; Nora K. Chadwick and Myles Dillon, The Celtic Realms, Weidenfeld and Nicolson, London, 1967, chapter 1, pp. 1-17; Jan Filip, Celtic Civilisation and its Heritage, Publishing House of the Czechoslovak Academy of Sciences and Artia, Prague 1960 (English edition 1962); and Stuart Piggott, Ancient Europe, Aldine, Chicago, 1965, chapter 6, pp. 115-266. For the art of the European Celts, see Paul Jacobsthal, Early Celtic  Art, Oxford, Clarendon, original edition 1944, corrected edition 1969; and Nancy K. Sandars, Prehistoric Art rin.'.Europe, Penguin, Harmondsworth, 1967, chapters 8 and 9, pp. 226-284. 3. Stephen Mitchell, The History and Archaeology of Galatia, D. Phil. : .z'. thesis, Oxford University, 1974, hereafter Mitchell. 8. HISTORICAL OUTLINE The historical outline that follows is not intended to be a comprehen sive account of pre-provincial Galatian history.^ The intention is to present the basic historical data with a view to establishing a time-frame for the archaeological evidence collected in the later sections of this paper. It was mentioned in the introduction that the Galatians made a practice of looting and pillaging in order to sustain themselves, and that they . . seemed to find settled life uncongenial. Thus their early history in Ana tolia is that of a people always on the lookout for short-term profits, whether material or political. They found it easy to prey on the prosperous cities of Asia Minor, and were quick to take advantage of the shifts of allegiance among more stable populations in the area. Their first encounters with the Hellentistic world set the tone for their subsequent history. In the first third of the fourth century B.C., groups of Celts, better known as the Gauls, were driven south..from their central European homeland. Some were responsible for the sack of Rome; others travelled east and followed the Danube to Illyria and Pannonia. In 280 B.C., a second, and tripartite, migration began. One group, under Cerethrius, went to Thrace: the second, under Brennus and Acichorus, to 2 Paeonia; while the third, under Belgius or Bolgius, went to Macedonia. Brennus and some of his men made an unsuccessful attack on the sanc tuary of Apollo at Delphi, and were discouraged from further raids on 3 Greece. Another group of Celts, after various campaigns in Macedonia and 4 Thrace, was defeated by Antigonus Gonatas at Lysimacheia in the Propontis. 9. Some of the Celts, under Commontorius, founded the kingdom of Tylis on the 5 west shore of the Black Sea north of Byzantium. Mitchell points out that although, the sequence of events in Greece, Macedonia, and Thrace between (roughly) 280 and 278 B.C. is difficult to establish, the activities of the Celts were remarkably consistent. As Mitchell puts it:^ Their aim was not land on which to settle, but money or booty, which could be acquired in a variety of ways: by hiring out their services as mercenaries, by demanding protection money from rulers whose land they were in a position to ravage, by attacking wealthy cities or sanctuaries, and by plundering the countryside. All these methods clearly anticipate the more widespread Galatian activities in Asia Minor. The next major event in the history of the Galatians was the diabasis of 278 to Asia Minor, which presumably took place after the defeat of 7 Lysxmachexa. Under Leonnorxus and Lutarius, some of the Celts who had been with Brennus1 and Acichorus' group had left Thrace for the Propontis. Perhaps it was there that they first heard of the richness of Asia Minor; in any case, Lutarius and a small band obtained five boats from the local Macedonia garrison, and crossed the Hellespont independently. Leonnorius and the larger part of the two leaders' original group were engaged by Nicomedes of Bithynia, to help subdue his rebellious brother Zipoetes (and thereby lessen the threat of Bithynian annexation by Zipoetes' ally Antio-g chus I), and thus obtained their passage to Asia Minor. Whatever the precise terms of the contract between Nicomedes and the 9 Galatians were , once they had helped him to quell Zipoetes' revolt, they were permitted to raid any part of Anatolia outside Bithynian territory. At this point, the Galatians divided into three tribes, each with its own area for plunder: the Trocmi took the coast of the Hellespont, and the 10. 11 Tolistobogii chose Aeolis and Ionia, while the Tectosages were to concentrate on the inland parts of Asia Minor.^ For the next five years or so, the Galatians roamed these areas at will, leaving terror and destruction in their wake. Cyzicus was hit in early 277, probably by Lutarius and his men on their way to join Leonnorius. Ilium was briefly considered as a possible Galatian base, but rejected be-12 cause it was unwalled. An inscription at Erythrae thanks the generals of the first four months of the year for arranging "Danegeld" payments to the 13 14 15 Galatians. Miletus and Didyma also suffered attacks, nor did Thyateira 16 escape. An inscription from Priene describes Galatian raiding methods and records the measures taken by Sotas to get rid of the city's attackers. The Galatians moved into the territory of Priene, desecrated sanctuaries, cap tured citizens living outside the walls at random, set fire to houses and farms, and killed numerous people. Sotas paid volunteers to man strong 17 points in the countryside, from which attacks could be made. Eventually Antiochus I undertook to rid Asia Minor of the Galatian menace. Records of this campaign are almost entirely lacking, except for mentions of a battle in which Antiochus, with the help of sixteen elephants, x / 18 defeated the Galatians and earned the title of io>T*jp. The date of the battle has recently been moved from 275 to 272, which means that the Gala-19 tians had had fully five years in which to plunder and terrorize. . 20 Ancient historians differ as to what happened next. Appian and 21 Livy imply that because of Antiochus I, the Galatians had to leave west ern Asia Minor and take up residence in the basin of the Halys and the 22 23 Sangarius. Strabo and Pausanias suggest that the Galatians were con-11. fined to Galatia proper only after their defeat at the hands of the Attal-ids some forty years later. Mitchell points out that Livy, Strabo, and Pausanias are trying to emphasize later successes—of Cn. Manlius Vulso on the one hand who defeated the Galatians on their own ground in 189 B.C., and of the Attalids on the other. Other factors also have a bearing on this subject. One is Antiochus1 reputation after the battle; another is the available evidence for Galatian history between ca 270 and ca 230. There are no firmly dated attacks on the cities of western Asia Minor after 270; this implies that the Galatians were using some other area as their base, probably central Anatolia. Then, too, forty years of Galatian wandering, during, which the fighting men would have been accompanied by their wives and children seem difficult, unlikely and unnecessary. As Mitchell says, it is far more sensible to assume that at some time after the Battle of the Elephants, the Galatians settled on the Anatolian plateau in the three groups mentioned by ancient historians. The territory around Pessinus was inhabited by the Tolistobogii, that around Ankara by the Tectosages, while the area east of the Halys around Tavium was populated 24 by the Trocmi. In theory, then, it should be possible to find traces of the Galatians in Galatia itself dating from ca 270 B.C. or later. It should not be forgotten that Antiochus' victory over the Galatians in 272 was cemented by payments of_protection money in order to prevent future Galatian harassment in western Asia Minor. The Galatians, pre sumably established in their new settlements in central Anatolia by now, looked north and east for additional sources of income. An episode of ca 255-253 is perhaps paradeigmatic of their approach to earning a living. Ziaelas had been passed over as heir by his father Nicomedes of Bithynia, 12. the old ally of the Galatians who had procured their passage to Turkey in 278. When the citizens of Heracleia Pontica settled the dispute between Ziaelas and Nicomedes peacefully, the Galatians attacked their territory 25 and marched home with the booty. It is obvious that the Galatians did not have ahigh regard for treaties, since their raid on Heracleia was a direct violation of their agreement with Nicomedes. Relatively soon after the Battle of the Elephants, the Galatians had fought with Mithridates of Pontus and Ariobarzanes of Cappadocia against a Ptolemaic expedition in the Black Sea. The Galatians captured the anchors of the enemy ships, were awarded the territory around Ancyra, and named their new city after their naval trophies. Tidy aetiological myths of this type often contain some truth: Mitchell says that it is possible that Mithridates I Ctistes was partially responsible for settling the Gala-26 tians in the Ankara region before he died in 266. This is also another indication that the Galatians were establishing themselves in central Ana tolia before the middle of the third century. The Galatians remained on a friendly footing with Pontic rulers until the death of Mithridates' succes sor Ariobarzanes in ca 250 B.C., at which point Ariobarzanes' son Mithri dates II, who was still a boy, succeeded to the Pontic throne, and the Galatians plundered the kingdom. When Heracleia tried to help out one of the Pontic cities, the Galatians attacked it again. In the end, the usual 27 antidote to Galatian invasion was employed, and they were paid off. The Galatians became involved with the Seleucids during the "Brothers' War" in ca 241-239 B.C., when they were recruited to help Antiochus Hierax against his brother Seleucus II. Seleucus had given his Anatolian hold ings to Antiochus for his aid in the war against Ptolemy III, but after 13. peace had been made, regretted this decision and tried to win Anatolia back. He led an expedition against Antiochus' ally Mithridates of Pontus, and continued to central Anatolia, where he was thoroughly defeated by Mithridates and Antiochus fighting with Galatian mercenaries near Ancyra. The Galatians saw their chance to get rid of the Seleucids altogether, since Antiochus could not maintain his authority over them. They forced Antiochus to make them his allies, and to give them part of the spoils of the war, plus some of the tribute which Antiochus required of the cities of Asia. Once these terms had been agreed upon, they threatened the life 28 of Antiochus, who then fled to Magnesia. Now that the Galatians had manoeuvered themselves into a superior position vis-a-vis the Seleucids, they could put pressure on western Asia Minor again, particularly on Pergamon which had risen to power in the forty years since the Galatian raids in that area. Pergamon under the Attalids was still paying the Galatians money to avoid a recurrence of 29 these earlier attacks, but at this point they refused to do so any longer. The Tolistobogii set out for Pergamon, and were repulsed by Attalus in the 30 Valley of the Caicus in ca 241-240 B.C. The Tolistobogii called in their allies, the Tectosages and Antiochus Hierax, and got as far as the walls of Pergamon itself before being trounced by Attalus. Attalus contined to fight Antiochus until 229 or 228, but the Galatians accepted the victory 31 as a decisive one, and in future left Pergamene territory alone. Attalus lost no time making political capital out of this victory, which was commemorated in monuments in Pergamon and Athens celebrating the triumph of civilized Hellenism over Celtic barbarity. The Seleucids, how ever, were not so easily eliminated. Antiochus III succeeded Seleucus III 14. in 223; within a year, Antiochus1 uncle Achaeus had stripped Attalus of his recently acquired Asian territory and had restored Seleucid rule in 32 Anatolia. When in 220 Achaeus declared himself king of this Asian terri tory, Attalus pursued the undoubtedly distasteful course of enlisting the aid of the Aegosages, probably the Hellespontine refugees of the kingdom of Tylis which fell in ca 218. Attalus1 next step was to march against the cities of Aeolis, which Achaeus had encouraged to revolt against him. Ini tially the expedition was a success, but an eclipse of the moon gave the 33 discontented Celts an excuse to mutiny. Attalus solved this dangerous problem by settling them as a military colony in the region of the Hellespont, under the guidance of Lampsacus, Alexandreia Troas, and Ilium. The Galatians almost immediately turned against the cities, and stormed Ilium. The Alexandrians defeated them and drove them northeast to Arisba near Abydus. Prusias of Bithynia realized that his own kingdom was in danger; he defeated them and killed the men 34 in the battle field, and the women and children in their encampment. For the next generation, a certain calm prevaled among the Galatians in Anatolia. The Aegosages had been wiped out, and it was unlikely that any ruler of Asia Minor would make the mistake of inviting Thracian Gauls across the Hellespont for a third time. Thus the Galatian migrations to Asia Minor were at an end. At this point in his historical account of the Galatians, Mitchell takes advantage of the lull in Galatian activity after the Pergamene defeats to sum up the information on the Galatian tribes in the third century B.C. He begins with the problem of their numbers. Ac-35 cording to Livy, Nicomedes recruited only 20,000men. The figures avail able for the Gallic expedition into Greece are substantially higher. 15. Pausanias says that there were 152,000 foot-soldiers and 20,400 cavalrymen 3 6 each with two mounted servants. Pompeius Trogus mentioned 150,000 foot-37 soldiers and cavalrymen altogether, while Diodorus reckons their strength at 150,000 foot-soldiers, 10,000 cavalrymen, and a baggage-train of 2,000 38 vehicles. While these figures are large, Mitchell points out that they are consistent and explain why the Gauls caused so much terror when they • * • „ 39 arrived in Greece.-There is still some discrepancy between these figures and Livy's, which are much lower. Many of the Gauls were of course killed in Greece and Macedonia, while others settled in the kingdom of Tylis near the Black Sea. It is possible that some of the Gauls migrated to Asia Minor after Nico-. 40 medes 20,000 crossed the Hellespont. As for their strength in Anatolia, 41 this was apparently ensured by rapid reproduction, mentioned by Livy, .42 and by Justin: quamquam Gallorum ea tempestate tantae fecunditatis iuventus fuit, ut Asiam omnem velut examen aliquod implerent. Danique neque reges Orientis since Mercennario Gallorum exercitu ulla bella gesserunt, neque pulsi regno ad alios quam ad Gallos confugerunt. Mitchell cites Launey's work on the armies of the Hellenistic period as confirmation of Justin's second statement. Galatians did serve in the army of every king in the eastern Mediterranean, according to Launey, and they enlisted from Asia Minor. Thus the numbers of the Galatians and 43 their military prowess are emphasized. Mitchell next addresses himself to the question of the extraordinary reputation of the Galatians. Hellenistic soldiers were well-trained pro fessionals, and the Galatians who fought as mercenaries with them grad-44 ually adopted standard Greek armour and equipment. When they fought on 16. their own, their military methods may have been different. According to Livy, they did battle naked, armed with large shields, long swords, and any Galatians tend to show them naked or only partly clothed, fighting without the help of Hellenistic weapons or equipment. Whether the Galatians always fought like this, or whether these descriptions simply make them conform to the conventional image of barbarians, it is hard to say. In any case, Mitchell concludes that they had two advantages over their opponents. First was their formidable and probably exaggerated reputation for barbaric cour age; add to this their unusual appearance—the "procera corpora, promissae 46 et rutilae comae, vasta scuta, praelongi gladii" mentioned by Livy — and the psychological advantage of the Galatians over their terrified opponents becomes clear.47 Their second advantage, Mitchell says, was that they did not fight like Greeks. They did not use the phalanx, which required a flat and un impeded field of battle; rather, they used the tactics of guerrila warfare, 48 and made the most of rough ground and quick skirmishes. In the early second century B.C., the Galatians were active again. 49 In 197/6 they were giving the people of Lampsacus trouble, and at roughly the same time, they attacked Heracleia Pontica yet again, in order to gain access to the sea, and perhaps to their kinsmen in the Danube basin; they were foiled in their attempt.Some Galatian activity in Paphlagonia during this period may also be indicated, if Mitchell is cor rect in assuming that at this point the Celtic noblemen Gezatorix, whose name is mentioned by Polybios, himself acquired the district in Paphla-« r X ' 51 gonia known as *j t€^e(ropi^o£. available stones. 45 Statues and representations in other media of the 17. The major event of the second century B.C. from the Galatian point of view was the expedition led against them by Cn. Manlius Vulso in 189 B.C. One of his first steps after he took over the consulship and the army at Ephesus was to declare his intention of subduing the Galatians for good, as part of a general campaign to pacify Asia up to the Halys. Manlius marched up the Maeander valley, through Phrygia and Pisidia, south to Pamphylia, and north through eastern and central Phrygia to the frontiers 52 of the Tolistobogii, accompanied by Attalus, and aided by Eumenes. After a brief and fruitless attempt at a diplomatic settlement in which Eposognatus, one of the Galatian reguli, tried to persuade the Gala tians to surrender without a fight, a battle was fought on Mount Olympus somewhere between the Sangarius and Ancyra. The Tolistobogii, helped by some of the Trocmi, were badly defeated, and 40,000 captives were taken. When the Romans proceeded to Ancyra and tried again to negotiate, the Galatians responded with an ambush on Manlius, which failed. The result was another battle, this time at Magaba, where the Tectosages and the rest of the Trocmi, aided by Ariarathes III of Cappadocia and Morzius, a Paphlagonian dynast, were defeated. 8000 Galatians were killed, and those that survived fled east across the Halys. Manlius marched back to Ephesus 53 with the spoils before winter set in. Manlius concluded two separate settlements, one with Eumenes in Eph esus, and a second with the Galatians in Lampsacus. Although Eumenes re ceived all of Antiochus1 former possessions, Galatia was not one of them. In the meeting with the Galatians, Manlius did not require them to become citizens of Pergamon, nor did they have to pay an indemnity, as Ariarathes did. They were simply to keep the peace with Eumenes and to contain them selves within the boundaries of their territory. Mitchell concludes from 18. this evidence that "the Romans were reluctant to crush the Galatians once and for all, but already saw them as a potential counterweight to the king-54 dom of Pergamum". For the next generation, the principal role of Galatia was in fact that of a buffer state between Pergamon and Pontus. Ortiagon, one of the reguli of the Tolistobogii, played a leading part in the disputes of the 180's, which consisted mainly of wars between the Attalids and the kingdoms forced by Rome to yield territory to them. Ortiagon was particularly involved with Prusias of Bithynia in a war with Pergamon from ca 186 to 184. Mitch ell notes that in an inscription from Telmessus, Prusias and Ortiagon are named as equal partners in the war, and that the Galatians are called IctfAetTrtt rather than being mentioned in tribal divisions. These facts, Mitchell says, give credence to Polybios' statement that Ortiagon had 55 managed to unite the Galatians after the defeat at the hands of Manlius. Eumenes was able to put a stop to the alliance of Ortiagon and.Prusias, but he was not so lucky with a coalition formed by Pharnaces of Pontus and certain eastern rulers such as Mithridates of Armenia, formed in 183. In 179, he did defeat Pharnaces, but not before a Pontic attempt to ravage Galatia. As Mitchell points out, even after the treaty between Pharnaces and Eumenes was signed, Galatia was in a difficult position, since she was 56 open to attack from both sides. There follows a ten year gap in Galatian history. In the mid 160's however Galatia enjoyed a temporary resurgence of power. In 168 the Galatians actually rose against Eumenes, forcing him to ask Rome for help. P. Licinius was sent to Asia Minor to negotiate, but he failed to estab lish any basis for a lasting peace. As Polybios indicates, this failure was no accident, but the result of the Roman desire to use the Galatians 57 as a check on the power of Pergamon. The same strategy is evident in the Roman reaction to Prusias' (of Bithynia) accusation that Eumenes was occupying Galatian territory illegally. 58 Rome in this case was content to re-affirm Galatia's independence. At the same time, again during the mid 160's, the Trocmi were making trouble 59 for Ariarathes of Cappadocia. In addition, Mitchell dates the acquisi tion of much of Lycaonia by the Galatians to this period.^ The spread of Galatian influence in the middle of the second century B.C. can be seen in another area as well. A series of inscriptions from Pessinus preserves the correspondence between Eumenes and Attalus II, and the high priest of the sanctuary, Attis. The first letter in the series has been dated to 163 B.C. The second mentions that Aioiorix, the brother of the high priest, was accused of some crime against the temple of the Mother Goddess. The name Aioiorix reveals that after the expedition of Manlius, the Galatians had taken over a hitherto exclusively Phrygian .- .. 61 privilege. After this brief renaissance, Galatia becomes less and less important, as the scanty evidence for the next century seems to indicate. No evidence has survived for any Pergamene transactions with the Galatians during the second half of Attalus Ii's reign, nor during the reign of his successor. The Galatian decline into obscurity can be inferred from other evidence. When Asia was made a province in 129, Phrygia was turned over to Mithri dates V of Pontus. Since Mithridates could only have communicated with Phrygia through Galatia, Jones suggests that by this time the area was 62 controlled by Pontus. Later in 96 B.C., Mithridates VI was forced to give up his control of Galatia, as well as of Paphlagonia and Cappadocia, 20. when Nicomedes of Bithynia shifted his allegiance from Mithridates to Rome. In Mitchell's opinion, from this time onward the Galatians were simply the allies of Rome in the Mithridatic wars.^ The Galatians turned against Mithridates because of what he did after his defeat by Sulla. He had apparently decided to eliminate any future threats from the Galatians, so he invited the tetrarchs to Pergamon on some friendly pretext, and had all but one of them killed. Mithridates then sent men to Galatia to take care of the tetrarchs who had not come to Per gamon. The three tetrarchs who survived the massacre, one of whom was Deiotarus, promptly threw out Mithridates' satrap. After this, there was no question of loyalty to Mithridates, and the Galatians went over to Rome. Mitchell makes a number of perceptive comments about second century B.C. Galatian history. During this period the description of the Galatians becomes less and less accurate, as they learned to use politics and diplom acy as much as military ability. As Galatian diplomatic activity increased, the names of individuals such as Ortiagon begin to appear. Mitchell assumes that by this time the aristocracy had learned some Greek, but he says that "there is little evidence that the cultural aspects of Hellenic civiliza tion were already being adopted by the Gauls. During this period, the name "Galatian" seems to be applied as a reg ional rather than an ethnic term. Mitchell cites as evidence for this change the names of two Galatian slaves mentioned in inscriptions from Delphi: Sosias, a boot-maker, and Athenais, an artisan. The name of an-other Galatian slave, r\*i<f#TV}S, probably had an Iranian origin and implies connections with eastern Anatolia. Intermarriage of Celts with the native Phrygian population can also be assumed. 21. There is also evidence for Galatian assimilation into local Anatolian cults, particularly the cult of Cybele practised at Pessinus. The high priest of the sanctuary during the reign of Eumenes had a brother named Aioiorix, and was presumably himself a Celt, as has already been mentioned That a Galatian could hold this office proves that the Galatians had made 6 7 a place for themselves in the organization of the temple. This is addi tional confirmation of the gradual absorption of local traditions by the 68 Galatians in the second century B.C. It is possible to speak of the Galatians collectively, even though politically they were not unified. Strabo's description of their constitu tion helps us to understand why the Galatians found group action difficult Each of the three tribes, the Tolistobogii, the Tectosages, and the Trocmi, was made up of four parts, each. with, its own tetrarch. Each tet-rarchy had in addition a judge, an army commander, and two subordinate officers. Mitchell finds it unlikely that this constitution was adhered 22. to throughout Galatian history; on the other hand he sees no difficulty in equating the tetrarchs mentioned by Strabo with the four reguli who ruled the Tolistobogii at the time of Manlius' invasion in 189. At times the tetrarchs seem to have acted independently of one another, a situation which 70 would not foster Galatian unity. The existence of the twelve tetrarchies provides an explanation for the mention of tribal names other than those of the three principal tribes in the sources. These include the Tosiopae, who had a tetrarch in 73 B.C., the Ambituti and the Voturi, who were connected with the Tolistobogii, and 71 the Teutobodiaci, linked with the Tectosages. The meeting place known as the Drynemeton mentioned by Strabo is a particularly interesting Galatian institution. .The word nemeton is found in other place names in western Europe such as Nemetobriga in Spanish Gal-acia, Nemetacum in northeastern Gaul, and Nemetodurum which became the modern Nanterre; it signifies the sacred grove in which important political and religious business could be transacted. In the case of the Galatian nemeton, a connection with oak trees is almost certainly indicated by the 72 drys-prefix. The lack of a clear distinction between the religious and the political which the existence of this institition seems to imply is 73 important, as Mitchell notes. As he puts it: The Galatians had rapidly taken over the superstitious beliefs and cults of their new home, but retained those parts of their religion which were indivisibly linked to their peculiar social and communal life. The cult of Rome and Augustus came as a substitute for these practices; the indigenous cults continued to be maintained as before. The Drynemeton is a rare example of Galatian retention of the Celtic cultural heritage which must have accompanied them during their travels through eastern Europe and into Anatolia. Archaeologically, of course, such an open-air sanctuary would be impossible to locate. We must continue to rely on other types of evidence for meeting-places like these. So far we have been considering the evidence from the literary sources for the political and social background of the Galatian phenomenon. Galar-tian religious customs have also been mentioned, although- the evidence for these is fairly slim. Before plunging into the history of the end of the pre-provincial period, it seems appropriate to give a very brief account of 74 some of the artistic representations of the Galatians. The psychological advantage which the appearance and demeanour of the Galatians gave them over their enemies has already been discussed. The attitude of the beleaguered Hellenistic opponents toward the Galatians may perhaps be revealed in the remnants of two Attalid sculptural dedications. The date of both monuments is controversial, but the identification of the principal figures at least has not been challenged. The first of these two monuments was a large circular group of statues erected in the precinct of 75 Athena at Pergamon. In the centre of the composition stood a Gaul com mitting suicide after killing his wife, whose body he supports on his left arm. He has short, coarse hair and a moustache, while she is modestly at tired in a longish dress. The man and his wife were surrounded by indiv-r idual statues of dying Gauls, the most famous of which is the Dying Gaul or Trumpeter in the Capitoline Museum. Bieber remarks on the tore around his neck, his leathery skin, his moustache, and "the greased hair standing stiffly around his forehead and cheeks" as characteristic of the Gala tians.^ The Dying Gaul's tore is a striking example of a direct link with 77 Celtic Europe. There is no agreement among scholars on the total number 78 of figures in this monument. 24. The date of this monument has also been a bone of contention. Schober, who did the original reconstruction, dated it to the last third of the third 79 century B.C. Bieber said more precisely that the monument was done in 80 228 B.C. Carpenter thought that the statues in question came from a monu-81 ment erected after the Great Altar in 180 B.C. by Attalus II, Havelock suggested that two master sculptors, one old-fashioned, one with newer . . ideas, could have contributed to the original monument, and proposes 200 B.C. as a compromise date. Attalus* dedication in Pergamon shows the Galatians in a pleasing . state of subjugation. Another Attalid dedication was set up at the south 83 wall of the Acropolis in Athens. The figures, on a smaller scale than the various dying Gauls discussed above, represented the war of the giants who lived in the area of Thrace, the battle of Athenians and Amazons, the conflict with the Persians at Marathon, and the destruction of the Gauls 84 in Mysia. Modern reconstructions include a total of sixty to eighty figures; the individual statues were placed on platforms or a series of steps so that they could be seen more easily. Twelve Galatians have been 85 attributed to this monument. Most of these male figures are shown fight ing defiantly. Most are naked; they have the short, coarse hair mentioned 86 above, and moustaches. One man is bearded and wears a short tunic. One of the most important aspects of this dedication is the equation of the Galatians with the traditional barbaric enemies of Greece. The dedication must have been the result of an artistic reflex action on the part of the Attalids after they had defeated the Galatians. The hairstyle and mous taches were presumably enough to indicate a specific type of barbarian, since there are no tores or other particularly Celtic pieces of identifica tion. 25, The same range of dates has been suggested for the Athenian monument as for the large circular composition discussed above. 228 B.C. seems early for this monument; most scholars favour 200 B.C. as a terminus post quern for the statues, since this was the year that the king actually visited Athens. Some prefer a more definite second century date, and connect the Athenian dedication with the Great Altar in Pergamon. Obviously this is 8 7 not a problem with an.easy solution. It is time to conclude this historical introduction. We left the Galar-tians at the point of their abandonment of Mithridates. The next event of any consequence to the Galatians was Pompey's eastern settlement of 63 B.C. Pompey determined that Galatia should be ruled by three tetrarchs, one for 88 each tribe. Mitchell points out that having three tetrarchs instead of the original twelve would promote Galatian unity, and avoid disputes within the tribes. This was important if the Galatians were to be of any use in Roman foreign policy. The three tetrarchs chosen from the survivors of the Mithridatic slaughter were Deiotarus, son of Sinorix of the Tolistobogii, Brogitarus son of Deiotarus of the Trocmi, and an unknown Tectosagan. As was the custom in Anatolian ruling families, the Galatian aristocracy had inter married: Brogitarus was married to one of Deiotarus' daughters, and Castor 90 Tarcondarius, a Tectosagan leader of the 40's, married another. The territory of Deiotarus and Brogitarus was increased by Pompey. Brogitarus acquired the fortress and territory of Mithridates which was 91 adjacent to the eastern territory of the Trocmi. Deiotarus received a larger area with more responsibility, including the Gazelonitis, a part of 92 eastern Paphlagonia, and the district near Trapezus and Pharnaceia. 26. Pompey was interested in Galatia primarily from a military point of view, and the Galatians proved useful to him on a number of occasions. Deiotarus in particular provided military support for Sulla, Murens, Serv-93 ilius, Lucullus, and Pompey in their Anatolian campaigns. In later years, 94 as Mitchell indicates, The Galatians provide the essential element of continuity between Pompey and Antony. At both periods it was they who provided the military backbone of Roman foreign policy in Anatolia. As for the political situation within Galatia, Brogitarus died some-95 time in the late 50's, and Deiotarus took over the Trocmian tetrarchy. His power had certainly increased by the time he supported Cicero in the 96 campaign to stop the Parthians in 51 B.C. The Tectosagans were still independent of Deiotarus and the Tolistobogii, since they also sent cavalry 97 to fight for Pompey at Pharsalus. When Deiotarus failed to withstand the invasion of Pontus and Armenia Minor by Pharnaces, he was deprived of his Trocmian lands by Caesar as punishment. Caesar himself came to Asia Minor to defeat Pharnaces, and returned after the campaign through Galatia to Bithynia in 47 B.C. Soon after Caesar's visit, Deiotarus reoccupied the Trocmian territory, and began putting pressure on the Tectosages as well. In 45 B.C., his grandson Castor, who was also the son of the Tectosagan Castor Tarcondarius, accused Deiotarus of trying to murder Caesar during his visit to Galatia. His accusation was the reason for Cicero's defence, 98 Pro Rege Deiotaro, although Deiotarus may have have been tried. Between 43 and 40 B.C., when he died, Deiotarus seized the territory of the Tecto sages, and had Castor Tarcondarius and his wife killed at their fortress 99 Gorbeous. Thus Galatia was united under one man, although the epitaph of Deiotarus' son found near Tomb B at Karalar, describes his father as king only of the Tolistobogii and the Trocmi.100 When Deiotarus died in 40 B.C., he was succeeded by none other than his accuser and grandson Castor.10"*" Castor's tenure as rule of the Gala-102 tians was short, since Amyntas was made king of Galatia in 37/6 B.C. In 25 B.C. when Amyntas was killed, the kingdom became part of the Roman . u. L ,103 empire as his son was too young to govern in his stead. The account of Galatian history from Pompey's settlement of 63 to the foundation of the Roman province in 25 B.C. is admittedly compressed, but the events of this period can be more easily seen the introduction to the history of the empire in Asia Minor, rather than the culmination of the pre-provincial period. It is hoped that this historical outline, despite its inadequacies, will provide a suitable background for the archaeological discussion that follows. The Galatians were part of the widespread Celtic migrations in eastern Europe. They crossed the Hellespont and terrorized the Asia Minor coast; when they were stopped, they settled in the area between the modern Sivrihissar and Yozgat. Many of them were mercenaries for eastern rulers, while others were embroiled in Anatolian politics. The Galatians did not really begin to assimilate either Graeco-Roman or local customs in any significant way until they had been subdued by Manlius and by Eumenes. During the first century B.C. they became the allies of Rome, and eventual ly their territory was incorporated as a Roman province. The Galatians established no large cities, and they assembled for worship outdoors. The historical information does not seem to indicate that they retained much of their Celtic heritage. They cannot have led very settled lives even after they arrived in central Anatolia, with con-28. stant tribal disputes, wars with Pontus and Pergamon, and a certain amount of "absenteeism" because of the men working in more distant campaigns. Galatian social organization seems to have been fairly loose in any case. The archaeological evidence will perhaps enlighten us on some of the aspects of Galatian material culture in Anatolia from the third to the first centuries;B.C. FOOTNOTES TO THE HISTORICAL OUTLINE 1. Useful sources for this period include F. Stahelin, Geschichte der  Kleinasiatischen Galater, 2nd edition, Leipzig, 1907; David Magie, Roman Rule in Asia Minor, Princeton University Press, 1950; Roger McShane, The Foreign Policy of the Attalids of Pergamum (Illinois Studies in the Social Sciences No. 53), University of Illinois Press, Urbana, 1964; Esther V. Hansen, The Attalids of Pergamon, 2nd edition, Cornell University Press, Ithaca and London, 1971; A.H.M. Jones, The  Cities of the Eastern Roman Provinces, Oxford, Clarendon, 1971; and chapters..1 and 2 in Stephen Mitchell's D.Phil, thesis, The History and  Archaeology of Galatia (Oxford University, 1974). 2. Hansen, p. 28. 3. Pausanias, X.19.5-23; Diod. Sic. XXII.9; Justin XXIV.5-6. 3 4. Justin XXV.102; Diog. Laert. II.141-2; SIG II no. 207 (Athens) cited by Mitchell, p. 483, n. 5. 5. Pomp. Trog. Prol. XXV (Otto Seel, Pompei Trogi Fragmenta, Teubner, Leipzig, 1956); Polyb. IV.46, VIII.22. 6. Mitchell, p. 2. 7. The principal authorities for the diabasis are Livy 38.16, and Memnon, FGH IIlb no. 434.11; cf. also Justin XXV.2.7f., Pausanias 1.4.5, Strabo XII.5.1, 566. Mitchell points out that since Livy and Memnon considered the situation from the point of view of Asia Minor, and not that of Europe, it is difficult to align their accounts chronologically with the sources on the Celts in Thrace and Macedonia. Justin, Mitchell says, implies that the crossing followed on the Lysimacheian defeat (p. 2). 8. Livy 28.16. M. Launey, "Une episode oubliee de 1'invasion galate en Asie Mineure", REA XLVII (1944) 217-234, and Hansen, p. 29. 9. Memnon, FGH Illb, no. 434.11. 10. Livy 38..16.11. This is the first mention of the tribal names. 11. M. Launey, Recherches sur les armees hellenistiques I (1949), p. 217f., cited by Mitchell, p. 484, n. 21; cf. G.T. Griffith, Mercenaries of  the Ancient World, Cambridge University Press, 1935. 12. Hegesianax in Strabo XIII.1.27. 13. This inscription is interesting because Leonnorius may be mentioned in it; SIG I no. 410, cited by Mitchell, p. 484, n. 23). 14. Mitchell, p. 5 and n. 25. 30. 15. A. Rehm, Didyma II: die Inschriften (1958), 275f., no. 426.7, cited by Mitchell, p. 484, n. 26. 16. Hansen, p. 30. 17. Hiller von Gaertringen, Inschriften von Priene (1906), no. 17, cited by Mitchell, p. 484, n. 24. 18. Appian Syr. 65 and Hansen, p. 30. 19. B. Bar-Kochva, Proc. Cam. Philol. Soc. n.s. 10 (1973), p. 1, n. 5, cited by Mitchell, p. 484, n. 30. 20. Appian Syr. 65. 21. Livy 38.16.8. 22. Strabo XII.5.1, and cf. IV.1.13. 23. Pausanias 1.4.5. 24. Jones, p. Ill; Mitchell, pp. 6-7, n. 39. 25. Mitchell, p. 485, nn. 45, 46. 26. Mitchell, p. 9; Stephanus of Byzantium, sv Ancyra; cf. Pausanias 1.4.5. 27. Mitchell, p. 9. 28. Justin XXVII.2. 29. Livy 38.16.14. 30. Hansen, p. 31; Jones, p. 113. 31. McShane, p. 60. 32. Edouard Will, Histoire politique du monde hell£nistique, II, Nancy, 1967, pp. 10-12. 33. Mitchell, p. 14. 34. Mitchell, pp. 15-16. 35. Livy 38.16.2. 36. Pausanias X.19.9. 37. Justin XXIV.6.1. 31. 38. Diod. Sic. XXII.9. 39. Mitchell, p. 16. It is possible, of course, that these sets of figures are derived from a common source. 40. Mitchell, p. 17. 41. Livy 38.16.13 42. :Justin-.XXV.2.8-9. 43. Launey as cited in n. 11 above, pp. 490-534; Jones-, p. 114. 44. Mitchell notes, however, that the large oval shield used by the Gauls known as the thureus was adopted by Greek and Asiatic professional soldiers (Launey, as in n. 11, 529f.). 45. Livy 38.21. 46. Livy 38.17.3. 47. Mitchell, p. 19. 48. Mitchell, pp. 19-20. 3 49. SIG II no. 591, cited by Mitchell, p. 487, n. 94. 50. Memnon, FGH Illb, no. 434.20. 51. Strabo XII.3.41; Polyb. XXIV.14.6; Mitchell, p. 21. 52. Livy 38.12-15; Mitchell, p. 22. 53. Livy 38.27.9; Polyb. XXI.37; Hansen, pp. 88-91; Mitchell, pp. 22-23. 54. Mitchell, p. 24. 55. Polyb. XXII.21; OGIS no. 298, cited by Mitchell, p. 488, n. 111. 56. Polyb. XXIII.9.1-3, XXIV.14-15, XXV.2; and Livy 40.2.6, 40.20.1-4; and cf. Mitchell, pp. 25-27. 57. Polyb. XXIX.22.4, XXX.1, and especially XXX.19-12-14. 58. Polyb. XXX.30.2f. 59. Polyb. XXXI.8.2. 60. Mitchell, p. 28. 32. 61. C. Bradford Welles, Royal Correspondence in the Hellenistic Period, Yale University Press, New Haven, 1934, nos. 55-61. Hansen sees Aioiorix. ' brother the priest as the leader of the Pergamene faction in Galatia (Hansen, p. 126 and n. 181). 62. Jones, p. 116. 63. Mitchell, p. 29. 64. Appian Mithr. 46; Plut. De mul. virt. 259a-b. 65. Mitchell, p. 31. 66. W.M. Ramsay, A Historical Commentary on St. Paul's Epistle to the  Galatians, Hodder and Stoughton, London, 1899, pp. 81-83. Ramsay sug gests that "the trades and handicrafts in Galatia were wholly in the hands of the subject population, while .the Gaulish aristocracy had war as their only trade" (p. 83) . For the name MerKpcfnjy see also L. Zgusta, Kleinasiatische Personennamen, Verlag der Tschechoslovakischen Akademie der Wissenschaften, Prague, 1964, §846, p. 280. 67. Mitchell, p. 353. It is worth noting that in the mid first century B.C. Deiotarus and Brogitarus, two Galatian tetrarchs, fought over the right to appoint the high priest at Pessinus. 68. Mitchell, p. 32. 69. Strabo 12.5.1. 70. Jones, pp. 114-115; Mitchell, p. 33. 71. Plutarch De mul. virt. 259a; Pliny NH V.146; Mitchell, p. 33. 72. T.G.E. Powell, The Celts, Thames and Hudson, London, 1958, pp. 140-141; and cf. Anne Ross, Pagan Celtic Britain, Routledge and Kegan Paul, London, 1967, for British examples, pp. 36-37. 73. Mitchell, p. 358. 74. The principal sources for these are P. Bienkowski, Die Darstellungen  der Gallier in der hellenistischen Kunst, Alfred Holder, Vienna, 1908; Margarete Bieber, The Sculpture of the Hellenistic Age, revised edi tion, Columbia University Press, New York, 1961; Hansen, pp. 299-314; and Christine Mitchell Havelock, Hellenistic Art, Phaidon, London, 1971. Cf. Bienkowski's Les Celtes dans'les arts mineurs (Krakau 1928), especially p. 65 for painting. I was not able to see this book; the title and page reference were brought to my attention by Prof. E.H.. Williams. 33. 75. Hansen, p. 303. This monument may be contemporary with a long monu ment commemorating the successes of Attalus during the first fifteen years of his reign. The victors, on horseback, were shown triumphing over Galatians and Persians in the Seleucid army, the figures being arranged in groups. Too little of this monument remains to say much more than that the Galatians were, as usual, equated with the tradi tional barbarian enemies of Greece (Hansen, p. 306). Pausanias mentioned (1.4.6) that there was also a painting of the struggle between Pergamenes and Gauls (.cf. again Hansen, p. 367). 76. Gaul and wife: Bieber, figs. 281-283, 424; Dying Gaul: Bieber, figs. 426-427; pp. 108-109. Diodorus says that it was customary for Celtic noblemen to grow shaggy moustaches, but that as a rule they did not have beards. As for the "greased" hair, Diodorus mentions that it stood away from their faces like a horse's mane (Diodorus V.28). Cf. Powell, Celts, p. 68. 77. Powell notes that tores first appear in the early La Tene period during the mid fifth century B.C., and that they may be derived from Persian exemplars (Powell, Celts, pp. 71-72). 78. Bienkowski, pp. 1-36; Bieber, p. 108; Hansen, p. 304. 79. Arnold Schober, "Das Gallierdenkmal Attalos I. in Pergamon", Mitteil- ungen des deutschen archaologischen Institut's. Romische Abteilung 51 (1936) 104-124. 80. Bieber, p. 108. 81. Rhys Carpenter, Greek Sculpture, University of Chicago Press, 1960, p. 224. 82. Havelock. pp. 145-147. 83. Bieber, p. 109; Hansen, pp. 311-314. 84. Pausanias 1.25.2. 85. Hansen, pp. 310-311. 86. Bearded man: Bienkowski, pp. 40ff. , figs. 53-55; other figures'.L-tributed to this group: Bieber, figs. 430-431, 435-436, and p. 109. 87. Hansen, pp. 311-314. 88. Strabo XII.5.1. 89. Mitchell, p. 36. 90. Mitchell, p. 36. 34. 91. Strabo XII.5.2. This fortress has been identified with- Kerkenes Kale southeast of Yozgat which was inhabited between the first millennium B.C. and the Selcuk period (Mitchell, p. 36). 92. Strabo XII.5.2. Mitchell discusses the boundaries of this area, pp. 37-38. See generally his chapter 2 and especially chapter 3 for dis cussion of the boundaries of Galatia in this and later periods; cf. also Jones, pp. 121-122, and Magie, chapter XIX, pp. 453ff. 93. Cicero Phil. XI.33.4; Mitchell, p. 44. 94. Mitchell, p. 61. 95. Cf. Strabo XII.5; Mitchell, p. 44. 94. Mitchell, p. 61. 95. Cf. Strabo XII.5; Mitchell, p. 44. 96. Cicero Ad Att. VI.1.14; Mitchell, p. 42. 97. The Tolistobogii also sent support; see Mitchell, p. 42. 98. Cicero Pro Rege Deiotaro; Mitchell, pp. 46-48. 99. Strabo XII.5.3; cf. Suda, sv Castor Rhodius. 100. See the section on burials and grave goods, and Coupry in Arik, RA 6 (1935), pp. 140-151. 101. Dio 48.33.5. 102. Dio 49.32.3; 53.26.3; Strabo XII.5.1; Mitchell, pp. 55-56. 103. Mitchell, p. 64. 35. POTTERY The earliest attempt to find archaeological evidence for the Galatian presence in Anatolia involved the definitions of a specifically Galatian type of pottery. This was only natural, given the great durability of pottery, and the tendency for styles in shape and decoration, at least in the fine wares, to succeed one another in chronologically recognizable phases. If the pottery of the Galatians could be defined, then Galatian sites could be easily identified through field reconnaissance. The location of the sites thus identified could provide useful information on settlement patterns, and this evidence in turn would fill in the gaps in the historical record. Such a way of proceeding is perfectly acceptable in terms of arch aeological method: the identification of Galatian pottery, if there were such a thing, would be invaluable to the archaeologist. In his article on the "Galatian" pottery from Bogazk5y, Maier has summarized earlier theories on the subject."'" In 1907, he reports, Zahn noticed a similarity between a little-known Anatolian ceramic type and late La Tene pottery from Central Europe. The sites involved were Bogazkoy and Gordion on the one hand, and Mont Beuvray and Hradischt von Stradowitz on the other. Zahn postulated that the Celts (or Galatians) made.their way to Asia Minor, discovered Hellenistic pottery, and liked it enough to in spire its manufacture in Europe. Their newly acquired skill in pottery-making would have been transmitted over the Black Sea, up the Danube, and 2 into the old Celtic homeland. There were several reasons which made it difficult to adopt this theory. First, the development and spread of the "Galatian" pottery had not been sufficiently studied. Second, evidence was lacking for the intermediate Balkan and Danubian connections. Third, as Dechelette and Behrens have pointed out, late La Tene was modelled on local European wares, in particular early La T&ne pottery of the Marne type. In addition, the painting techniques of the late La T£ne could have been derived from a 3 similar technique used in the later Hallstatt period of central Europe. Bittel probably came closer to the truth when he suggested that both' the La TSne and "Galatian" types were separately inspired by Hellenistic 4 models which were in wide distribution throughout the Mediterranean. It is worth noting that the early theories concerning "Galatian" pottery were proposed at a time when no Galatian site had been identified, let alone excavated. Not enough was known about the pottery of the Hellenistic age generally, so that the "Galatian" variety could not have been the local, but derivative, product that Maier believes it to be. Furthermore, Zahn's theory assumes that the Galatians maintained contact with the Celts in Europe even after they settled in central Ana tolia. The assumption that the Galatians still possessed a strong Celtic identity after years of migration and cultural instability is one that needs to be challenged. Zahn's theory reflects a tendency, which has been remarkably persistent, to think in terms of Celtic unity, rather than to allow for Galatian adaptability. Maier in the first detailed study of "Galatian" pottery, deals mainly with the material from Bogazkoy, since it has provided the largest sample to date. In addition, he discusses similar finds from six other sites: Alaca Huyuk, Alisar, Buyuk Nefeskoy/Tavium, Gordion, Kirsehir Hiiyiik, and Pazarli. He does not mention the pottery from Karalar, which had of 37. course been excavated before he wrote his article in 1963. Since then, other sites have produced additional pieces. These are Yalincak outside Ankara, Asarcik Huyuk/Ilica more or less equidistant from Gordion and 5 Ankara, and Kululu, located south of the Halys and northeast of Kultepe, The more recent finds are important in that they make it less tempting to view "Galatian" pottery as a phenomenon confined to the Halys bend. Maier himself notes that the distribution of this pottery is not restricted to traditional Galatia, but that it extends beyond those areas, particularly in the cases of Alisar and Pazarli.^ As for the pottery itself, the forms are broad and low for the most part, although amphorai, kraters, and perhaps pitchers of the lagynos type have been found. The typical shapes (see fig. 2, p. 81) include bowls with inturned rims and a low base or foot, flat-bottomed beakers, and large plates with everted rims. The pottery is made, in good Hellenistic fashion, of fine hard-fired clay which shows red or reddish-brown with a grey-black centre in the break. The surface is light leather brown, often burnished so as to give the appearance of paint. The broad, low vessels have part-7 lcularly thin walls. The decoration of these pots is characterized by overlapping or almost touching stripes on a plain background (.see fig. 1, p. 80) ; often the stripes divide the pot into zones suitable for further decoration. Colours include sepia, red, yellow, and pale brown, with all the shades in between, and one extremely pale shade which may be a late Phrygian legacy. In taller vessels, more attention was paid to the shoulder, while in plates and bowls the rims were decorated. Both dark- and light-8 ground pots are found; the light-ground variety predominates. 38. Vegetal motifs were often combined with stripes. Thus as well as diagonal stripes, herringbone pattern, simple and stripe-filled zigzag bands, crosses, triangles, and hatching, the pot-painter's repertoire included tendrils, sometimes with birds, palms, or blossoms. One very 9 common motif is the schematised ivy, laurel, or grape leaf tendril. That this pottery is Hellenistic is confirmed by two factors, accord ing to Maier. First, the pottery is delicate; second, paint is of three main colours applied to a burnished surface. The overlapping or almost touching stripes are apparently inherited from the later Phrygian period. "Galatian" ware, especially at Alisar or Bo§azkSy, is found in purely Hellenistic and early Roman context, that is, with broken Megarian bowls, coarse grey ware with thick rims, and a "Roman" pottery with red overglaze; terra sigillata is sometimes found in association with it."^ Maier found it impossible to evolve any precise chronological sequence for "Galatian" pottery, although he was able to define the upper and lower chronological limits for the period during which it was produced. The style began at the end of the fourth century when it absorbed a number of later Phrygian features, such as the leafy tendril. It continued through out the Hellenistic period incorporating various other features from the Lagynos Group and West Slope Ware. It was in fact "a desiccated version of the familiar flora" of the Hellenistic period.^ No intimate connec tion with late La T^ne pottery is imaginable, according to Maier. Maier admits that the date and origin of the sherds make it possible that the pottery was used by the Galatians, but he cannot say that it was made by them. He suggests that the monochrome coarse ware often found in conjunction with "Galatian" pottery is the true Galatian pottery; this 39. 12 still awaits investigation. As for late La Tene ware, Maier says that European contact with the appropriate Hellenistic types could have taken place in the Balkan peninsula. Recent excavation has produced a small body of Celtic material from the Woiwodina region and the area of the central Sava. It is significant that these Balkan Celts retained their European identity; their pottery does not resemble that of the Galatians, but that of the inhabitants of the south western Slovakian plain, and their metal work is Celtic beyond doubt. Thus after a certain point there was no contact between Galatians and Europeans; any European Celtic contact with the Hellenistic world was entirely inde-13 pendent of the Celts m Asia Minor. Maier has tried to demonstrate that on the one hand Anatolian pottery of the third to first centuries followed externally imposed models, and that on the other hand there is no known pottery which can be shown to have been exclusively Galatian rather than generally Hellenistic. Mitchell, too, has stressed the limitations of the evidence. The chronology of the pottery of the Hellenistic, Roman, and Byzantine periods in Anatolia is still sadly incomplete. Megarian bowls, black glazed and fine red slip wares, which were common throughout Asia Minor, are the only 14 definite indicators for dates in the third and first centuries B.C. Bittel, however, is determined to find a direct link between the Galatians and the pottery under discussion. In an article roughly con temporary with Mitchell's thesis and therefore ten years after Maier's article, he presents a rather different point of view."^ Bittel re-examines the pottery from BogazkJJy, other sites east of the Halys, and the sites west of the Halys mentioned by Maier, and reaches 40. the conclusion that the pottery west of the Halys does not really belong in the same category. Gordion has none; neither do Ancyra, Pessinus, or Karalar. The evidence from Pazarli and even Alaca is likewise not reliable. There is, of course, a great deal of "Galatian" pottery from Tavium and Bogazkoy, and small sites nearby. Bittel dates this pottery to the second and first centuries B.C. and says that it might continue into the first century A.D. He suggests that since there are no definite findspots west of the Halys the pottery was produced east of the river, possible at Tavium, the Trocmian capital. From there it could have been distributed to east Pontic and Cappadocian sites. He is skeptical of the late Phrygian elements which Maier saw in the pottery, and indeed his later dating precludes the absorption of such elements into "Galatian" pottery. Bittel concludes by saying that more research should be done on late Hellenistic and early • 16 Roman east Galatia. To a large extent Bittel'sconclusions nullify Maier's, since they disagree on the pottery's distribution and date. Thus the task of the field archaeologist has in no way been made easier, because there is still no reliable definition of the pottery, either as Galatian or as "Galatian". In other words, we lack one of the principal means of identifying possible Galatian sites. One way of alleviating this is to excavate Tavium, which should yield a good pottery sequence. Such an excavation would also test Bittel's interesting theory that the pottery might have been made there, and the implication that after a certain point the Trocmi were leading an exist ence culturally independent of the other two tribes. Then, too, Maier1s earlier suggestion that Galatian pottery is really the monochrome coarse ware found with the decorated wares we have been discussing should be thoroughly investigated. Thus the question of Galatian pottery is far from closed. 42. FOOTNOTES TO POTTERY 1. Ferdinand Maier, "Bemerkungen zur sogenannten galatischen Keramik von BogazkCy", Jdl 78 (1963) 218-255. 2. Zahn, AA 1907, pp. 225 ff., cited by Maier, p. 218. 3. J. De'chelette, Manuel d' arche'ologie prihistorique, celtique, et gallo- romaine II, 1914, pp. 1488ff.; G. Behrens, in Beitrage zur alteren  europSischen Kulturgeschichte, Festschrift R. Egger, I. 1952, pp. 52 .'. ff., cited by Maier, p. 218. 4. Kurt Bittel, Kleinasiatische Studien, IstMitt 5 (1942) 31 ff. 5. For Yalincak, see Burhan Tezcan, Yalincak Village Excavation 1962-1963, Middle East Technical University Archaeological Publications No. 1, Ankara, 1964, pp. 18-19. I am grateful to Sevim Buluc of the METU Museum for letting me see the Yalincak pottery in October 1973. For Asarcik Hiiyuk/Ilica, see Winfried Orthmann, "Untersuchungen auf dem Asarcik Huyuk bei Ilica", IstMitt 16 (1966) 27-78. For Kululu, see Jones in Anadolu XIII, a reference cited by F. Winter, letter, December 1, 1973. Prof. Winter also mentioned that some "Galatian" pottery had been found in Kara Samsun, but neither he nor I have been able to find references; cf. Maier, p. 244, n. 38. 6. Maier, pp. 220, 223. 7. Maier, pp. 223-224; figs. 1 and 2; and see.figs. 1 and 2 here, pp. 80-81. 8. Maier, p. 225; figs. 3, 4, 8, 9. 9. Maier, pp. 226, 228. 10. Maier, pp. 230, 237. 11. R.M. Cook, Greek Painted Pottery, second edition, Methuen, London, 1972, p. 203. 12. Maier, pp. 251-253. The monochrome coarse ware could also be a contin uation of the local Phrygian variety. Cf. J. Schafer's remarks on "graue polierte aiolische Keramik" in Hellenistische Keramik aus  Pergamon, Walter de Gruyter and Co., Berlin, 1968 (Pergamenischer  Forschungen Bd. 2), pp. 14, 29. 13. The Woiwodina and Sava evidence comes from cremation graves in these areas. Metal objects include brooches, fibulae, chains, swords, and knives. Pots are strip-painted with, geometric designs. The cruder wares are sometimes straw-tempered, and belong to the middle and late La Tene periods (Maier, pp. 253-254. 43, 14. Mitchell, pp. 395-396. 15. Kurt Bittel, "Bemerkungen zur sogenannten galatischen Keramik" in Melanges Mansel, Turk Tarih Kurumu Yayinlarmdan Dizi VII, TTK Basimevi, Ankara, 1974, pp. 227-237. 16. Bittel, pp. 235, 237. 44. BURIALS AND GRAVE GOODS Distinctive burial customs, like distinctive styles in pottery, can be a useful indication of the presence of a cultural group in a specific region during a specific time period. As we have seen, pottery so far is not a reliable index of the Galatian presence; we will encounter a somewhat simi lar situation in the following investigation of Galatian burials and grave goods. The Hellenistic period, as Kurtz and Boardman remark in their book on Greek burial customs, saw the introduction of monumental tombs for the elite. These tombs were of two main types, chamber tombs and mausolea. In the same period, the common people continued to use the sarcophagus, cist, tile grave, or ash urn, all of which are types known from earlier periods.1 Thus while it would normally be difficult to identify the graves of ordinary Galatians, there might be some chance of locating those of . important personages, such as the tribal tetrarchs. As it happens, we may have evidence for both kinds of burial among the Galatians, in the form of unsophisticated cist graves and pithos burials from BogazkSy, and more elaborate chamber tombs from western Galatia, in the area between the Sangarius and the Halys. These two broad categories— chamber tombs and humbler graves—will be examined separately, on the basis of the architecture and its variants. This will be followed by a general discussion of the grave goods found in each type of burial, and the value of certain objects for the identification of a Galatian tomb or grave. The map on p. 91 shows the locations of Galatian, "Galatian", and other tombs. 45. I. Chamber Tombs The evidence that the Galatians occasionally buried their dead in chamber tombs comes principally from Karalar, where three tombs were ex cavated in the 1930's. Near one of these structures was found the epitaph of Deiotarus II; all three were therefore identified as Galatian. It is important to remember that the inscription was the basis of the identifica tion, rather than the architecture of the tombs. All three structures are chamber tombs, one with a barrel vault, one with a peaked roof, and one in which a special kind of corbelling was used. Each of these tombs resembles others in Asia Minor, both because of architectural likenesses, and because of certain parallels in the grave goods. In the past it was sometimes assumed that such tombs were also Galatian, although there are still no definite reasons for so identifying them. Rather, it seems that the tombs in question, including those at Karalar, should be considered a phenomenon of the Hellenistic period. The introduction of the three architectural types into Anatolia, and particu larly Galatia, as well as the inclusion of certain types of grave goods, should not be connected with the arrival of the Galatians. With their usual adaptability to new surroundings, the Galatians at Karalar happened to use three of the architectural forms current in the Hellenistic period for the burial of some of their prosperous leaders. Once more we find ourselves floundering for lack of a solid background: there is no thorough study of the Hellenistic tombs of northwestern Ana tolia, and thus there is no chronological or typological framework within which these examples can be placed. Nonetheless the inscription at Karalar gives us some sort of starting-46. point. In the section that follows, each of the three architectural types at Karalar will be discussed, after which there will be a brief description of the Karalar example, and any other appropriate examples from the area. 1. The corbelled tombs Our first task is to explain what is meant by the term "corbelled" in this context. The corbelling technique under discussion consists of laying successive courses of slabs diagonally across the corners of the space to be enclosed, until one slab will complete the vault. Lawrence was certain that this method of roofing a small area had been used in wooden prototypes.' He suggested too that Karalar C derived from a type of royal tomb used in Central Asia in the Bronze Age, and further, that the Phrygian tombs of the seventh and sixth centuries B.C., the classical tholoi of Asia Minor, and Karalar "may be reproduction of local types of tomb or underground dwell ing" .3 This brings us to the question of the origin of the corbelled chamber tomb. One explanation, put forward by Lawrence, is that the type was a local Anatolian one, and that the Galatians simply adopted it when they arrived. The alternative explanation is of course that the corbelled chamber tomb was not a local type, and that it was introduced from some other area, either by the Galatians themselves, or by some other group of people. The first step toward determining the origin of the corbelled chamber tomb is to list the known examples of the type in Turkey, and their dates, 4 starting with- Karalar and proceeding (roughly) westward: Karalar C: 1st century B.C. Gordion I (Tumulus 0): 2nd century B.C.? 47. Igclir: 4th century B.C. Tepeeik/izmit: 3rd century B.C. Gemlik: 4th century B.C. Mudanya: 5th century B.C. Kepsut: no date Kirkagac: no date Pamukkale (2): late Hellenistic/Roman Belevi: 6th century B.C. (Milas:. late Hellenistic/Roman variant) An interesting pattern emerges from this list: most of the tombs were con structed outside Galatia, and the earlier ones are those which lie further away, toward the west. This suggests that the tomb type was introduced ... from some area to the west of Galatia. Given the early date of the Belevi tomb, and the locations of Mudanya, Gemlik, and I^dir, it seems possible that the corbelled chamber tomb might have evolved in western Asia Minor, and that later on, similar tombs were built in areas to the east. It does not seem possible that the Galatians introduced the type to Asia Minor, since at least four of the known examples were built before they crossed the Hellespont. In an article on the Gemlik tomb, Mansel discussed a group of corbel led tombs in Thrace, the majority of which can be dated to the fourth, century B.C. Most of the Thracian examples are corbelled in beehive fashion, rather than in the irregular manner used in the Turkish tombs. The tomb of Kurt Kale near Mezek combines both types; it has an irregularly corbelled antechamber leading to a beehive tomb chamber. Builders in Turkey seem to have preferred the irregular corbelling, although a fourth, century beehive tomb is known at Kutluca in the Propontis.^ There exists the possibility that the corbelled chamber tombs of Asia Minor may have inspired the corbelled antechamber at Mezek. While there is still work to be done on this subject, the information 48. at hand can be used to clarify a few points with respect to the Galatians. The corbelled chamber tomb seems to have originated in Turkey, perhaps as early as the sixth century B.C. There several fourth century examples in Bithynia, which the Galatians might have seen on their way to settle in central Anatolia. One of these could have influenced the choice of architectural form at Karalar C. There is certainly no basis for saying that any corbelled tomb must be a Galatian tomb. The only definite ex ample is Karalar C; another possible Galatian example is the corbelled tomb at Gordion, because of its location in Galatia. 49. FOOTNOTES TO THE CORBELLED TOMBS 1. Donna C. Kurtz and John Boardman, Greek Burial Customs, Thames and Hud son, London, 1971, chapter XVI, pp. 273ff. The chronology of Hellenistic tombs may well be affected by the recent discovery of the unplundered burial at Vergina, said to be the tomb of Philip of Macedon. Cf. Nicholas and Joan Gage, "Treasures from a Golden Tomb", New York Times Magazine, December 25, 1977, pp. 14-19, p. 32. 2. A.W. Lawrence, Greek Architecture, Penguin, Harmondsworth, 1967, second edition, p. 230. Cf. A.M. Mansel, "Gemlik Tumulus Mezari", Belleten 38 (1974) 181-189, and Rodney Young, "The Campaign of 1955 at Gordion: Preliminary Report", AJA 60 (1956), p. 252. 3. Lawrence, p. 302, chapter 6, n. 1. 4. References for the corbelled tombs at Karalar and Gordion may be found in the catalogue which follows. Ig'dir: Diindar TSkgSz, "I§dir Kazisi Raporu", TflrkArkDerg 22 (1975) 151-153. Tepeeik/izmit: Nezih Firatli, "Jewellery Found at izmit", Istanbul Arkeoloji Miizeleri Yilligi 15-16 (1969) Gemlik, Kepsut, MilSs: A.M. Mansel, "Gemlik Tumulus Mezari", Belleten 38 (1974) 181-189. Mudanya: A.M. Mansel, "Mudanya Mezar Binasi", Belleten 10 (1946) 1-12. Kirka^ag: M.J. Mellink, "Archaeology in Asia Minor", AJA-67 (1963), p. 189. Pamukkale: Eugenia Schneider Equini, La Necropoli di Hierapolis di  Frigia, Accademia Nazionale dei Lincei, Monumenti Antichi, Serie. Miscellanea 1.2, Rome, 1972, cited by Jean Carpenter and Dan Boyd, "Dragon-Houses: Euboia, Attika, Karia", AJA 81 (1977) 179-215, p. 201, n. 111. Belevi: Mansel as cited for Gemlik, etc.; see also Hermann Vetters, "Ephesos. Vorla'ufiger Grabungsbericht 1971", AnzWien 109 (1972) 85-88, figs. 2-3. Cf. also the 1st century B.C. Roman tomb at Kenchreai, for which a sim ilarly corbelled roof has been restored (W. Willson Cummer, "A Roman Tomb at Corinthian Kenchreai", Hesp 40 (1971) 205-231. According to Prof. E.H. Williams, the grounds for this restoration are questionable. 5. See A.M. Mansel, Trakya-Kirklareli Kubbeli Mezarlari ve Sahte Kemer ve  Kubbe Problemi, Turk Tarih Kurumu Yaymlarindan VI Seri, no. 2, Ankara 1943; "Gebze Yoresince Kutluca Kubbeli Mezari ve onun Trakya Kubbeli Mezarlari Arasinda Aldigi Yer", Belleten 37 (1973) 143-158, chart after p. 158, and map, fig. 28; and Mansel"s article on Gemlik, cited above in notes 2 and 4. For the tombs at Mezek, see B. Filov, "Die Kiippelgraber von Mezek", Bulletin de 1'Institut d'archgologie Bulgare XI (1937), and B. Filov, "The Bee-hive Tombs of Mezek", Antiquity XI (1937) 300-304. While 50. Kurt Kale had been thoroughly pillaged, its neighbour at Maltepe had merely been rifled in antiquity. According to Megaw, it contained either an intrusive Celtic chariot burial (there were horse bones buried in the dromos), or the trophies of local contact with Celtic warriors. The finds included two linch-pins and five terret rings of Celtic type, numerous Greek objects of the Hellenistic period, a local bronze figurine of a great boar, and a third century imported Italic bucket. The Celt (or possible Celts) involved here may have been a veteran of the raid on Delphi (J.V.S. Megaw, Art of the European Bronze  Age, Adams and Dart, Bath, 1970, pp. 19, 112; fig. 170). Megaw also mentions that in a third beehive tomb at Kazanlik there are paintings which may depict Celtic warriors (Megaw, p. 112). 51. Catalogue of corbelled tombs 1. Karalar C (figs. 7a and b, p.95) Karalar C is oriented roughly north-south and consists of two square chambers and a short dromos. The burial chamber is much larger than the anteroom; both roofs are corbelled. There were burials in both chambers, but robbers had disturbed and partially removed the contents. Finds in cluded fragments of a caligula speculatoria, pieces of iron mail, and sections of a gold tore set with precious stones. An altar was set up toward the north, as with Tomb B; it was fairly well preserved and had a krepis of five steps.1 1. Remzi Oguz Arik, "Karalar Hafriyati", Turk Tarih Arkeoloji Etnografya  Dergisi 2 (1934) 102-167; R.O. Arik, "Les Tumuli de Karalar et La Sepulture du Roi Deiotarus II", RA 6 (1935) 133-140, pp. 137, 140; there are no photographs of the tore sections. 2. Gordion I/Tumulus 0 Gordion I, or Tumulus 0, was discovered by shepherds in 1954 and excav ated under the direction of Rodney Young in 1955.1 It had already been robbed at an earlier date so there were no finds, apart from some iron nails on the floor of the main chamber, and the fragments of a terra cotta larnax. The tomb is similar in type to Karalar C. The tomb was oriented approximately east-west with, the entrance at the east, and consists of a square inner chamber and a smaller, rectangular antechamber connected by a door. A bedding for the tomb was prepared by packing a layer of pebbles over hardpan. The floor-slabs lie directly over the pebble layer and pror-vide a platform for the walls. The blocks for the walls are of soft lime stones (poros), neatly cut and fitted on the inside, but uneven and roughly-dressed on the outside. Mortar was used to fill unaesthetic gaps on the inside. The interior was once covered with a thin layer of white stucco. A plain moulding separates the walls from the roof. 2 Each, chamber has a corbelled roof of hard limestone slabs. six cor belled layers reduce the roof opening to an oblong small enough to be covered by a capstone. Since the courses of the chamber roof are slightly higher, the roof itself is slightly higher. Cuttings were found for pivot-sockets, bolt-holes, and sills for the doors to close against. No trace of either door was found, unless the iron nails in the inner chamber were used in a wooden door. Young says that the doors may never have been installed, and that the stones blocking the 3 outer doorway may have seemed sufficient to repel robbers. Particularly interesting is the fact that the tomb was laid out on a 4 set unit of measure, .165 metres or .33 metres. Young suggests that a "foot" of .33m was in use in Phrygia during the period, and adds that this unit may have been in use from Persian times in the third century. If this is so, the builders of this tomb, whether they were working for the Gala tians or for someone else, must have been familiar with local building traditions. The dating of this tomb is difficult, since no helpful inscriptions or grave goods were found. It must have been built before the first century B.C. at the latest since there is a pit of this date cut into it. The terminus post quem is much harder to establish. It is conceivable that the tomb is as old as the fourth century B.C., since there are other corbelled tombs in Anatolia of that date. The tomb might have been re used in the Roman period, since other terra cotta sarcophagi of that type have been dated to the Roman period. 53. 1. Information in this section is condensed from Rodney Young, "The Cam paign of 1955 at Gordion: Preliminary Report", AJA 60 (1956), pp. 250-252; plan and section, pi. 81, fig. 3. 2. Young, pi. 81, figs. 4, 5; pi. 82, figs. 6, 7. 3. Young, pi. 82, figs. 7, 8. 4. Young, p. 251. 5. Cf. for example the terra cotta sarcophagi on view in the Afyon Museum. 2. Barrel-vaulted Tombs Like the corbelled tombs, barrel-vaulted tombs were probably introduced from the west at a date earlier than the arrival of the- Galatians. Mace donia is the most likely source for this type, although the Anatolian ex amples are humbler than the ornate tombs found in Greece."*" The catalogue below includes Karalar A, and the tomb at Kucucek as a comparison. 1. See Kurtz and Boardman, and Lawrence, p. 211, as cited in notes 1 and 2 on p. 49. Catalogue of Barrel-vaulted Tombs 1. Karalar A (figs. 8 and 9, pp. 96 -97). Karalar A consists of a square, barrel-vaulted chamber with a dromos, built in regular well-dressed courses. The dromos was not centred with respect to the chamber, and was filled with large stones. More stones were piled outside the dromos itself. The tomb had been robbed before excavation, but there were some offerings left. Among these were a gold necklace set with precious stones, part of a floral diadem, also gold, a 54. bronze fibula, and a guttus of Pergamene style painted with flowers in the manner of "Galatian" pottery.1 This tomb lies some distance from Tombs B and C; its date of construc tion may be different from the first century date assigned to the other two on the basis of the inscription near Tomb B. 1. See Arik, as cited on p. 51, n. 1. For the fibula see the earlier of the two articles, pi. 9, fig. 18, and p. 122. 2. Kugucek/Aykazi (fig. 10, p. 98) Here the vaulted chamber and the dromos are centered.on the same axis. The tomb was apparently built in the first century B.C. and re-used in the second century A.D. The grave goods remaining consisted of a leafy gold diadem, lagynoi, and lamps. Finds of a later date included a Roman imperial coin which led to the original identification of the tomb as a Roman struc ture .1 1. Nezih Firatli, "Bitinya Arastirmalarma Birkac ilave", Belleten 17 (1953), pp. 22-25. 3. Tombs with Peaked Roofs There has apparently been little or no discussion in print of this architectural type; since there has been no general study of Hellenistic and later tombs in Galatia and nearby areas, it is impossible to situate these tombs properly. As in the case of the corbelled and barrel-vaulted tombs, it is not possible to say that peaked roofs indicate a Galatian burial. The catalogue here includes Karalar B, the tomb with Deiotarus II' s epitaph.; another tomb at Gordion, which, could be Galatian because of its location; Bolu East, which is a borderline possibility; and Begevler, 55. which is almost certainly not Galatian, but combines a peaked roof and a barrel vault with a relatively early date and an interesting (Bithynian) location. Catalogue of tombs with peaked roofs 1. Karalar B (fig. 11, p. 99 ) Karalar B consists of a rectangular chamber and a very short dromos. The roof of the burial chamber, which faces northwest, is not vaulted as in Karalar A, nor corbelled as in Karalar C, but peaked. The roof consists of twelve large blocks, inclined and leaning against each other. This tomb had also been robbed, but contained a porphyry offering table, a glass vase with gold ornament, and mysterious traces of purple colouring, per haps connected with the burial. On the north side of the tomb, there was a ruined white marble altar, perhaps used in a funerary cult. Fragments of a sculptured lion and of a trophy were found in front of the tomb. The epitaph of Deiotarus II was also found in this area:'*" C&fytoTeipoo hkOQO/iotCov D<3*t Y<*kiTcjv ToJUrTo&ojc-Coupry restored and commented on the text. Deiotarus the elder we know 2 was a basileus; but Deiotarus, his son, also appears to have held the 3 title: he was declared rex by the Senate. 56. The inscription and therefore, presumably, the tomb can be dated fairly precisely. Cicero, writing in March. 43, refers to Deiotarus f ils_ as living.' At the battle of Philippi, Deiotarus does not appear, so that he may have died shortly after Cicero wrote his letter. His father did not die until 41/40.5 Thus we have that rare phenomenon in Galatia, a clearly identified and securely dated monument. 1. Jacques Coupry, section on the inscription in Arik's article, "Les Tumuli de Karalar et La Sepulture du Roi Deiotaros II", RA 6 (1935), pp. 140-151; for the tomb generally see the earlier part of Arik's article, and Arik, "Karalar Hafriyati", Turk Tarih Arkeoloji Etnografya  Dergisi 2 (1934) 102-167. 2. Strabo 12.3.13. 3. Cicero Ad Att. V.17.2. 4. Cicero Phil. XI.12.31. 5. Coupry, p. 147. 2. Gordion II (fig. 14a, p.102) Like Gordion I, this tumulus had also been robbed, and the tomb itself is in a fairly ruinous state.1 It was a plain rectangular tomb chamber set on a stone platform, similar in type to Karalar B. The section shows a raised area at one end. The skeletal remains of two individuals were found along the two long sides, so that the raised area may have been an offering table rather than a kline. It is difficult to see from the pub lished drawing what grave goods (if any) were buried with the dead. Ed wards remarks that the tomb "was laid out on a set unit of measure", but does not say whether the units were the same as those used in Gordion I. No date more specific than Hellenistic was given to the tomb. 57. 1. G.R. Edwards, "Gordion 1962", Expedition 5 (JL963), pp. 47-48. The tomb is mentioned only briefly, so that this description is necessarily short. There is apparently no more detailed account of the tomb in print. 3. Bolu East Cfig. 12, p.100) Bolu East is one of two plundered tumuli south- of Bolu excavated in 1964."'' Like its fellow, it was set on a terraced hilltop. Around the tumulus is a krepis of local andesite blocks. The stone pavement of the rather long dromos is now gone; it led to a rectangular burial chamber, .. also constructed of local andesite. The blocks were carefully shaped and no mortar was used. Large andesite slabs formed the peaked roof. Above these were three courses of limestone and andesite blocks forming a false arch. The tumulus consisted of rocks and earth piled on top of the roof. The dromos is lined with stone blocks. The upper blocks in the side walls project inward so that the width of the passage is narrowed. The walls in the burial chamber were plastered but unpainted. No small finds remained.3 1. Nezih Firatli, "Two Galatian Tumuli in the Vicinity of Bolu", AJA 69 (1965) 365-367. 2. Fxratli, p. 366, pi. 94, figs. 3, 4, 5. 3. Firatli dates both tombs to the years between 278 and 189. He theor izes that since Bolu (Claudiopolis) lies on the northwest fringe of Galatian territory, the tomb must have been built fairly early on, be fore the defeat of the Tolistobogii by Manlius in 189. This is fairly tenuous reasoning; there is no possible way of saying definitely that the tomb was so early, even if it was Galatian. The dating is at least partially based on the finds from Bolu West, which was so badly destroyed that its architectural type is no longer recognizable. The treasure hunters had uncovered a two metre section of andesite paving, beyond which- was a rough sarcophagus of pinkish andesite. The grave goods handed over to the authorities included articles of gold, silver, and bronze; some iron objects had also been found but were thrown out because they were so badly oxidized. Appar-58. ently there was no pottery. Among the gold objects was a buckle with the face of a man in relief (fig. 13, p. 92), which will be described in greater detail in the section of grave goods; there were also two gold tores, two gold brace lets, and a pair of golden finger or ear-rings. Silver objects included a bowl with relief decoration, imitating the shape of ceramic Megarian bowls, and a patera with leaf ornament and omphalos. There were in addition a bronze horse-bit and ring. (Firatli, pp. 366-367; pis. 93, 94, 95, 96.) 4. Be|evler (fig. 14b, p.102 ) Begevler consists of a barrel-vaulted chamber, and a dromos with a peak ed roof, built for the most part of shelly limestone with some sandstone blocks. The roof of the rectangular chamber is a true vault, built up over a "barrel" of heaped earth. The masonry is rather rough and no clamps were used in the construction. It was plundered so thoroughly that no grave goods remained. The excavator of the tomb compares it to the tombs at Bolu, and to Karalar B; he dates it to the third century B.C. without mentioning spec-... 1 lfic reasons. 1. Wolfram Hoepfner, "Kammergrab in bithynisch-paphlagonisch Grenzegebiet", AthMitt 86 (1971) 125-139. The combination of the two roof types is not unique, as another tomb of this kind—vaulted chamber, dromos with peaked roof—was found 6 km northwest of iznik. It was built of local marble; two painted klinai were found within. See M.J. Mellink, "Archaeology in Asia Minor", AJA 75 (1971), p. 179. It seems as if Anatolian builders were experimenting, albeit on a somewhat humble level, with the various architectural types imported from Macedonia and Thrace. II. Cist Graves and Vessel Burials All the evidence for these two types comes from Bo§azkSy. The two cist graves were called Galatian because each, contained a fibula; these 59. will be dealt with, in the general discussion of the grave goods. Vessels (almost all pithoi) were used chiefly for child burials; these have been identified as Galatian because plates or bowls of the Galatian or "Galatian" type were used to cover the mouths of the vessels. It is suspected by the excavators that the custom of burying the dead in pithoi or other vessels may have been borrowed from the inhabitants of north Cappadocia.1 Thus the vessel burials could be another example of the Galatian ability to adapt to local custom. 1. Kurt Bittel and R. Naumann, Bogazkoy-HattuSa I, Kohlhammer Verlag, Stuttgart, 1953, pp. 120-121. Cist Graves 1. Grave 12 in Section 3, Area XII, of the South Area at Bo§azk"6y was built of stone slabs. Its owner was an adult male, who was buried with a silver coin of Ariobarzanes I (95-62 B.C.), a fine red brown jug, a Meg-. arian bowl, a small iron ring, his sandals (judging from the 25 iron tacks found at his feet), an unrecognizable lump of.iron, and an iron fibula of Middle La Tene type with a high curve. The date assigned is Hellenistic.1 1. Hartmut Kuhne, "Die Bestattungen der hellenistischen bis spat-kaiserzeitlichen Periode", pp. 35-45 in Kurt Bittel, Bogazkoy IV. Funde aus den Grabungen 1967 und 1968, Mann Verlag, Berlin, 1969. For a photograph of the grave see pis, 23 c-d, 24 a-c. For the fibula see pi. 24d and fig. 10a. Cf. also Kurt Bittel, "Bemerkungen zu einigen spathellenistischen Grabfunden aus den sogenannten Sudareal im Bezirk des Tempels I in Bogazkoy", also in BoffazkOy IV, pp. 45-49. Bittel says that we must hesitate on the brink of calling this the grave of one of the Trocmi. 2. The second cist grave was found below the retaining wall which is con nected to the gate chamber of the Hittite wall to the east. Like the 60. first grave described above, it is built of stone slabs. The skeleton was almost completely gone. The finds included an iron sword, described as the only non-Hellenistic find, a bronze fibula of which- the upper part was shaped like a dolphin, several gold leaves, a Megarian jug, and four other undistinguished pots. No further comment was made about the fibula.1 1. Bittel, Bogazkoy IV as above, p. 121. Fibula, fig. 36a; sword, fig. 36b. Vessel Burials 1. In the Hittite residence quarter was found a broken column krater of Asiav Minor..origin, in which a child had been buried. It is solely on the basis of the krater that the burial is called possibly Galatian.1 1. Kurt Bittel and R. Naumann, Bogazkoy-Hattusa I, Kohlhammer Verlag, Stuttgart, 1953, pp. 120-121. 2., 3. There are five later burials from the area of the Great Temple, two of which are said to be Galatian because of the Galatian pots covering the mouths of the pithoi in which the remains of children had been placed. The pithoi were not consistently oriented. One of these two burials contained two silver earrings.1 1. Kuhne, Bo§azk'6y IV as above, p. 36. 4., -8. There are five pithos burials from the South Area, all with Galatian plates or bowls as covers. Again, there is no consistent orient ation of the pithoi. The skeletal remains where preserved were of child ren. Three of these five burials contained no grave goods at all. One 61. contained a silver armband, a silver finger-ring, and flat hammered copper coins. The other contained a possible amulet consisting of a small disc with short lead shank or handle, and a bowl."*" 1. Kuhne, Bogazkoy IV as above, p. 37. III. The Grave Goods There are two factors which imperil the accuracy of any discussion of the grave goods from the burials under investigation. First, the lack of a general study of the burial customs in central Anatolia during the Hellen istic period, which means that we have no concept of the "normal" assemblage of grave goods in a tomb of this date. The second is that many of the . structures discussed were looted and even partially destroyed; thus we have no way of knowing what may be missing from the groups of objects recovered. Nonetheless the objects can be divided into two major categories: those which might occur in any reasonably prosperous Hellenistic burial, and those which seem Celtic. The latter are of particular interest because of their potential use in identifying Galatian burials. We will begin with the non-Celtic grave goods. In the architectural section above, a number of tombs were described partly because they might be Galatian and partly because they were good comparisons both for archi tecture and for the grave goods they contained. The evidence from the chamber tombs west of the Halys suggests that the standard set of grave goods included jewellery (leafy diadems in gold, ear-.and finger-rings, bracelets) and objects in clay (lagynoi, Megarian vessels, lamps). The burials at Bogazkoy, although not as rich as the Galatian and Bithynian examples, contain objects of similar type: leaves from a gold 62. diadem, silver earrings, finger-ring and armband, two Megarian vessels, the possible amulet, plus the coin of Ariobarzanes I. Thus while the Hellen istic inhabitants of Bo§azkoy were apparently uninterested in chamber tombs, they were sufficiently in touch with their neighbours to the west to find the same sorts of grave goods appropriate. A number of tombs in Bithynia and elsewhere Csee map, fig. 3, p. 91 ) contain similar finds. The ruinous tomb at Diizce contained two diadems, a pin, a cup, and a ring, all in gold; two strigils, a bracelet, a rod or pipe, and an instrument, all in silver; and a bronze mirror.'*" The cist grave at Yaylapinar contained a gold diadem and a fusiform unguentarium of 2 the late second/early first century B.C. The barrel-vaulted tomb at Tepecik/Tersiyekoy contained fragments of a gold diadem, five lagynoi, six fusiform unguentaria, two lamps, and a silver urn.^ In the first century barrel-vaulted tomb at Kocakxzlar near Eskisehir were found fragments of a gold diadem, gold jewellery, lagynoi, unguentaria, and terra cotta figurines.^ Farther west, another barrel-vaulted tomb dated to the second century B.C. at contained a gold diadem, a gold medallion with the impression of a Lysimachos stater, fusiform unguentaria, and lamps.^ The third century corbelled tomb named Tepecik, also in izmit, yielded 6 several pieces of gold jewellery. And, finally, there is the tomb at Dardanos in the Troad, built in the fourth century B.C. and re-used throughout the Hellenistic period, which, is exceptional because it seems not to have been robbed. It contained a large number of grave goods including gold wreaths, diadems, necklaces, and pendants; ceramic cups, jars, and bowls; bronze urns, two of which bore the names of the cremated 63. ' dead; bronze mirrors, pins, and bracelets; alabaster bottles; bone pins, spoons, and rings; a stringed instrument of wood; some silver; and remains 7 of shoes, combs, textiles- and baskets. This quick survey of scattered tombs and burials indicates a certain uniformity in funerary practices in Anatolia from the fourth to the first centuries B.C., at least as far as grave goods are concerned. Jewellery, vessels, and utensils of various kinds, often of precious materials, con stitute the majority of the grave goods. While the material obviously deserves more attention, we can say that if the Galatians used Hellenistic architectural tomb types, they also provided their dead withuthe objects found in other Hellenistic burials found in the area, and generally in western Anatolia. Are there any objects from the burials in question which are not so typically Hellenistic? From Karalar A we have a bronze fibula; from Kar alar C, pieces of iron mail and tore fragments. From Bolu West, we have two tores, a buckle showing a man with a beard and moustache, bracelets with animal head terminals, and a bronze horse-bit. From Bo^azkSy we have two fibulae, 1 bronze and 1 iron, and an iron sword. Tores and fibulae immediately remind us of European Celts; a moustachio'd man seems unusual for the habitually clean-shaven Greeks; weapons and a horse-bit contrast strangely with the leafy diadems and unguentaria from these -burials. Tores have already been mentioned in conjunction with the sculptural representations of the Galatians as obviously Celtic objects. The brace lets with animal head terminals have parallels in Iran and in Germany, which is initially confusing. It must be remembered, however, that one of the major influences on the art of the European Celts was the Near 64, East. Thus tores and animal head terminals are part of the Celtic reper™ Q toire although, originally inspired by Persian and Scythian models. The gold buckle from Bolu West Csee pi. 13). shows a man's face in re lief. The man's hair is flowing and curly; he wears a band or diadem. Small incised lines indicate his beard and moustache. Incision was also used for the decoration of the buckle's rim, with an ivy and bud and leaf pattern. Firatli cites parallels for this ivy pattern in "Galatian" pot tery, but does not comment on the beard and moustache, which are used in sculpture to indicate barbarians. He does suggest that the buckle belonged 9 to "an important person buried in the tumulus". While the bronze horse-bit resembles some from the treasury at Perse-polis and others from Transcaucasia, it should be remembered that the European Celts acquired their knowledge of horse-trappings from the Near 10 East. The Bolu West horse-bit deserves further study by Celtic special ists, as do the iron sword from Bogazkoy and the iron mail from Karalar C. The fibulae found at Karalar A and Bogazkoy are potentially Celtic objects. Bittel discusses the iron fibula from Bogazkoy in conjuntion with other (bronze) fibulae from Asia Minor, all of which lack a proper strati-graphic context; two of them have no provenance at all. The sites rep resented are Priene, Kayseri, Mersin, and western Asia Minor, of which Mersin is the only surprising location. Bittel does not seem to mention the fibula from Karalar, nor does he comment on the other fibula from Bogazkoy, whose upper part was shaped like a dolphin. Furthermore, he does not speculate on the presence of the fibulae at Mersin, or at the remaining sites, although, he dates them to the mid first century B.C., 11 r with one possible later exception. Doubtless these and other details 65. will be discussed in his paper on Celtic finds to be published with the proceedings of the Tenth International Conference of Classical Archaeology-held in Ankara and Izmir in 1973. The real question here is this; how can we use the evidence of these possible Celtic objects? Let us restate the terms of the problem. We know from the historical sources that the Galatians were Celts, and that they settled in central Anatolia. European Celts used tores and fibulae. We have three tores and three fibulae from three different sites, one of which is known to be Galatian from an inscription. At this point all we can say is that the presence of such objects at Karalar, Bolu West, and Bogazkoy makes it possible that Galatians had something to do with these sites. We have no Celtic objects from domestic contexts, which means that these tores and fibulae were conceivably heir looms or booty. In other words, they were probably not made in Anatolia, but were perhaps brought in at some point, either by the Galatians them selves when they crossed the Hellespont in the third century B.C., or through some later indirect contact with Europe. If we did not know from the historical sources that a group of Celts had settled in Turkey, we would probably say that tores and fibulae had somehow trickled into Anatolia from the west and found their way into .. local burials of the Hellenistic period. There are other Celtic objects from Greece, Egypt, and Russia left there by mercenaries or brought in by 12 traders. No one would rely on these isolated finds to identify a Celtic settlement or burial unless there were additional evidence. The Celtic objects from Karalar C, Bolu West, and Bo§azk'6y were found in Hellenistic burials which resemble other burials in the area. Historical information 66. and the Karalar epitaph suggest that they might be more than stray finds and that they might be directly connected with, the Galatians, an originally Celtic people. The tores and fibulae help to confirm information supplied by ancient authors, but cannot be used by themselves to identify Galatian sites. We are therefore forced to limit our conclusions. Karalar B and C are certainly Galatian; Karalar A, which lies some distance away, probably is too. Bolu West is a definite possibility, as are some of the burials at Bogazkoy. We can also say that there is no such thing as a Galatian tomb type; the Galatians were happy to use local architectural styles and burial cust oms. Certain objects, for example tores and fibulae, recall the European origin of the Galatians. East of the Halys it seems that less elaborate preparations were made for the burial of the dead, but the grave goods, although fewer in number, are the same kind as those from the tombs west of the river. 67. FOOTNOTES FOR GRAVE GOODS 1. Nezih, Firatli, "The Tumulus of Tersiyekoy near Adapazarx'', Istanbul  Arkeoloji Muzeleri Yxllxgx 9 (19601, pp, 75-76, n. 2. The grave goods are listed in f AMY 7 (1956),, p. 12, nos. 6199-6216. A year after the recovery of the Duzce material, the Istanbul Museum acquired some paint ed terra cottas reputed to be from the Duzce region. Other such pieces were acquired by H. Kocabas in Istanbul, and by museums in Brussels and Munich. There are several types: busts of Aphrodite and other females; applique"es of Medusa, heads of women; shields, flowers, rosettes, and Ionic capitals. The colours are maroon, yellow, and white, and some pieces are gilded. See A.M. Rollas, "Terres Cuites colonies provenant de Bithynie", IAMY 13-14 (1966) 121ff. 2. Nezih Firatlx, "Bitinya Ara^txrmalarxna Birkac Ilave", Belleten. 17 (1953), pp. 16-18ff. and pi. 8. 3. Nezih Firatlx, "The Tumulus of TersiyekSy near Adapazarx", IAMY 9 (1960) 73-76. 4. Silmer Atasoy, "The Kocakxzlar Tumulus in Eskisehir, Turkey", AJA 78 (1974) 255-263. 5. Yildxz Mericboyu and Siimer Atasoy, "The Kanliba§ Tumulus at izmit ", IAMY 15-16 (1969) 67-95, English, summary, pp. 91-95. 6. Nezih Fxratlx, "Jewellery Found at izmit", IAMY 15-16 (1969). 7. Riistem Duyuran, "D^couverte d'un tumulus pres de l'ancienne Dardanos", Anadolu V (1960) 9-12; Zafer Taslxklxoglu, "Dardanos Jehri Yakxnxndaki Tumuliiste Yeni Bulunan Grekce Kitabeler", Tarih Dergisi XIII. 17-18 (1963) 160-173; J.M. Cook, The Troad, Oxford, Clarendon, 1973. 8. Jacobsthal remarked that the tore "is the ornament of man in Persia and Scythia: (Early Celtic Art, Oxford, Clarendon, original edition 1954, corrected edition 1969, p. 156). Fourth century tore with animal head terminals from a grave near Susa: Edith Porada, The Art of Ancient Iran, Methuen, London, 1965, pi. 51 below; cf. also O.M. Dalton, The Treasure of the Oxus, originally published by the British Museum at the Oxford University Press in 1926, republished by Lund Humphries, London, 1964, pi. XX, no. 137, for a penannular armlet ending in winged goats. These references were brought to my attention by Prof. Fred Winter, letter, December 1, 1973. 9. Nezih Fxratlx, "Two Galatian Tumuli in the Vicinity of Bolu", AJA 69 (.1965) 365-367, p. 366, pi. 95. 10. Cf. Johann Potratz, "Die Pferdetrensen des alten Orient", Analecta  Orientalia 41 (1966), fig. 47; Jacobsthal as cited in n. 8, p. 156. 68. 11. Kurt Bittel, "Bemerkungen zu einigen spathellenistischen Grabfunden aus den sogenannten Sudareal im Bezirk des Tempels I in Bogazkoy, Bogazkoy IV, pp. 45-49. 12. J.V.S. Megaw, "Two finds of the Celtic Iron Age from Dodona", pp. 185-193 in Liber Iosepho Kostrzewski octogenario a, venatoribus dicatus, edited by Konrad Jazdzewski, Wrociaw, 1968, The two finds from Dodona are an iron sword and a bronze fibula (p. 186, figs. 1, 2).. Other Celtic finds from Greece are the bronze anklets from Isthmia (John L. Caskey, "Objects from a Well at Isthmia", Hesp 29 (1960) 168-176), and a fibula in the Delos Museum (p. 191). The Celtic fittings from Maltepe near Mezek have already been mentioned (p. 4 9, n. 5). As well as the fibulae from various sites in Turkey mentioned by Bittel, there are Knotenringe like those from Isthmia from Cyzicus and Priene. A laminated birch-wood shield from Kasr el Harit in the Fayum and a small terret with southern British affinities, also from the Fayum, constitute the Celtic finds from Egypt (pp. 191-192). Bittel as cited in n. 11 above mentions a fibula from Gezer; this is the only piece of Celtic evidence from Palestine so far. A fourth/third century grave on the Dnieper has produced an early La Tene sword (p. 192); fibulae have also been found in Russia (p. 187). 69. FORTS AND SETTLEMENTS To speak of cities in Galatia is to speak almost exclusively of Roman Galatia. During the pre-provincial period, the major population centres in central Anatolia were Pessinus, Gordion, Ancyra, and Tavium, which were native in origin. Gordion, as we shall see,declined after the Galatians arrived. Of the three cities that were to become the tribal capitals in the Roman period, Pessinus was easily the most solidly established, undoubt edly because of the shrine of Cybele which was located there. The story of the Galatian high priest and his brother Aioiorix, already cited, shows that the ethos of the city and its temple were not fundamentally affected by the Galatian penetration of the holy office. Of Ancyra and Tavium more will be said later. That the Galatians were not great city builders is obvious from the historical sources; they were too preoccupied with military matters and tribal disputes. Their settlements consisted mainly of "isolated fortress es, usually at remote and well-protected sites in the hills, chosen for strategic purposes and not designed to act as economic, social, or cultural centres".1 As part of his investigation of the history and of Galatia, Mitchell set out to locate possible Galatian sites in central Anatolia. Using the historical sources, the accounts of various early travellers and scholars, and the techniques of field reconnaissance, he has collected evidence for possible Galatian forts and villages in north Galatia. As always with the Galatians, it is not easy to draw up guide lines for the identification of sites, and Mitchell's work should be re garded as a prelude to further surveys and above all excavation, rather than a series of hard-and-fast conclusions. 70. Mitchell distinguishes two types of settlement: hill-forts, and pos sible market towns and commercial centres. Some sites, such, as Ancyra and Tavium, may have incorporated Both, these functions; where this happens, the population wa.s probably not pure Galatian, but a, mixture of Galatian and Phrygian elements. For the most part, however, the forts are separate from the market towns. We shall discuss the hill-forts first. Mitchell describes the hill-forts as "small fortified enclosures, sited on easily defended hill tops, usually well off the main lines of 2 communication". The forts were designed for protection and defence rather than accessibility, but supplies would have been available from farmsteads and villages in the plains and valleys below. They confirm the general impression obtained from the historical sources that the Galatians, sus picious of each other and of outside enemies, eschewed living together in large groups, and that Galatia from the third to the first centuries B.C. 3 was politically fragmented. The construction of the forts, with the exception of Deiotarus' forts at Blucium and Peium, owes nothing to the military architecture of the Hellenistic period. There is no consistent ground plan, nor are building materials and methods necessarily the same from one site to another. None of the hill-forts proper, that is excluding Ankara and Tavium, was built 4 on an earlier settlement, as far as we know. The only factor which links the hill-forts is the choice of site—high, hard to reach, and commanding a good view of the surrounding area. Evidence for a date in the pre-provincial period is often restricted to Hellenistic sherds visible on the surface of the site. There are grounds for scepticism in a considerable number of cases; again only more field work can clear up some of the 71. doubts. Ancyra and Tavium; Blucium and Peium, (the modern Karalar and Taban-lxoglu Jiftlik): these are far and away the most definite candidates for the term "Galatian", The presence of Hellenistic sherds at Karacakaya, Sirkeli, and Yaraslx make their identification as Galatian forts probable. Mitchell has proposed that Dikmen, Tizke, andAssarlikaya be added to the list of possible Galatian forts by virtue of the relatively extensive re mains of walls. These ten sites will be discussed individually. There remain five other sites which at one time or another have been identified as Galatian: Karaviran, Cagnxk, c^anakci, Basrx, and Giizelce. Karaviran has been described by only one visitor, Ainsworth, whose account 5 Mxtchell quotes: It was a rude and primitive structure, consisting of a single wall, built of large stones, put together without mortar, and enclosing a space of 127 feet in diameter. Not far distant, upon a neck of rock below, was a fort of similar description. This is simply not convincing enough. The reports on the next three sites have not been checked since they were visited by travellers at the turn of „ 6 the century. At Cagnxk, stone foundations were noticed ; C^anakgx is 7 succinctly described as "ein altes (galatisches) Kastell" ; only a mention Q of Basrx is made. Giizelce is cited as a Galatian fort on the basis of a 9 sloppily built enclosure at the top of a hill; here too the evidence seems insufficient. In addition to the sites which, will be described in the catalogue below, there are three other fort sites mentioned by Strabo for which definite locations have not been found. Two are Mithridatium and Danala, in the territory of the Trocmi.''"0 The third is Gorbeous, the royal 72. residence of Castor Tarcondarius, chief of the Tectosages; the fpoopiov here was pulled down when Deiotarus had Castor and his wife killed.^ Nineteenth, century travellers Ramsay, Anderson, and von Diest placed Gorb-12 eous near the village of Beynam; French, has since proposed a different 13 site in the same general area. We pass now from the hill-forts to sites which were exclusively market towns and commercial centres, even if only on a small scale. In many of these cases, it would be impossible to prove that they were occupied only by Galatians; but obviously not all Galatians can have been living in the hill-forts. The identification of some unfortified Galatian sites is im portant for the overall settlement pattern. Of the•five sites listed below, Bo^azkoy is perhaps the best candidate for a Galatian settlement, because of the two burials containing Celtic fibulae, and the Galatian or "Galatian" pottery found there. Gordion, where there are two possible Galatian tombs, should have had Galatians among its Hellenistic inhabitants; Yalincak might have had. Tolgeri Hiiyiik and Golhuyiik have been suggested by Mitchell solely on the basis of their location and the presence of Hellenistic pottery. In terms of distribution, these sites may hold to the same pattern as the hill-forts, in that fewer people seem to have settled east of the Halys. Additional field excavation may show that this "pattern" is acci dental . Unfortunately, none of the five possible settlements listed below can be definitely identified as Galatian. We cannot therefore speak with con fidence of the characteristics of a Galatian habitation site, and indeed the Galatian gift for assimilation may have prevented the development of such characteristics. It may never be possible to distinguish, the Galatian elements in a Hellenistic town or village in Anatolia, although, the dis covery of Celtic objects such, as the fibulae and tores known from burials at Karalar, Bolu, and Bog'azkoy, and of inscriptions unequivocally mention ing Galatians could alter this situation. What is most needed, however, is excavation. As suggested before, Tavium would be an excellent site to excavate, as it would yield a good sequence from the Hellenistic to the Roman period, and later. Good evidence could be recovered for local religions, domestic and defensive architecture, and Professor Bittel's hypothesis concerning "Galatian" pottery could be tested. Once again, until more excavation has been done, most speculation about the Galatians will remain exactly that. The map on p. 103 shows the locations of the various forts and settle ments under discussion in this section. 74. FOOTNOTES 1. Mitchell, p. 178. See his chapter 6 on the cities of the Roman period. 2. Mitchell, p. 34, 3. Mitchell, p, 426. 4. Mitchell, pp. 474-475. 5. W.F. Ainsworth, Travels and Researches in Asia Minor, Mesopotamia, Chaldaea, and Armenia, London, 1842, vol. 1, p. 126, cited by Mitchell, p. 425. 6. W. von Diest and M. Anton, "Neue Forschungen in nordwestlichen Klein-asien", Petermanns Geographisches Mitteilungen, Erganzungsheft no. 116 (1895) 28-29; map. pi. Ill, cited by Mitchell, p. 434. 7. Mordtmann, Skizzen und Reisebriefe aus Anatolien, p. 370, cited by Mitchell, p. 435. 8. J.G.C. Anderson, "Exploration in Galatia cis Halym", Part II, JHS XIX (1899) 52-132; 280-318, p. 63. 9. Ainsworth as cited in n. 5, p. 146; G. Perrot, Exploration archeolo- . gique de la Galatie et la Bithynie, 1862, p. 278, cited by Mitchell, p. 441. For related problems in determining the locations and plans of classical fort sites in Greece, see James R. McCredie, Fortified Military Camps  in Attica, Hesperia Supplement XI (1966), pp. 50 and 94. I owe this reference to Prof. E.H. Williams. 10. Strabo, 12.5.2. 11. Strabo, 12.5.3. 12. Anderson, JHS XIX (1899), p. 102, quotes Ramsay's opinion; W. von Diest, "Von Tilsit nach Angora", Petermanns Geographisches Mitteilungen, Erganzungsheft no. 125 (1898), pi. III. 13. Mitchell, p. 451. 75. HILL-FORTS 1. Ancyra, The ^bofxjav at Ancyra mentioned by Strabo1 was undoubtedly located on the Kale, where any traces of Galatian occupation have been obscured by the Byzantine fortifications. Livy, speaking of Manlius' campaign in the area 2 in 189, calls Ancyra an urbs nobilis, but Mitchell feels that Strabo's designation of the site as a fort is more likely, although the settlement, which antedates the arrival of the Galatians, was certainly an important 3 market centre. 1. Strabo, 12.5.2 2. Livy 38.18ff.; Mitchell, p. 179. For general historical summaries of the Celtic occupation of Ankara, see Afif Erzen, Ilk Cagcla Ankara, Turk Tarih Kurumu Yaymlarmdan, VII seri, no. 12 (TTK Basimevi Ankara 1946); Clemens Bosch, "Die Kelten in Ankara", Jahrb. fur kleinasiatische  Forschung 2 (1953) 283-293. 3. Mitchell, p. 433. 2. Tavium Tavium, the modern Bviyuk NefeskSy, was a iffoupov and market town, like Ancyra, and like Pessinus, the focus of a cult, in this case of Zeus.1 The combination of three settlement functions--fort, market, and cult centre—is^exceptional among the Galatians. The site is located on the more westerly of two natural mounds at the edge of the hilly country be tween the open plateau and the plain of Sungurlu. Visible remains include 2 traces of late Roman or Byzantine fortification, and plentiful pottery. 76, 1. Strabo, 12.5.2; Mitchell p. 179., who calls attention to the issue of autonomous coinage in the first century B.C. at Tavium. 2. K. Bittel, Kleinasiatische Studjen, Istanbuler Mitteilungen 5 (1942). ch. 1, "Beobachtungen und Funde in Galatien", Cf, Mitchell, p. 460, 3. Tabanlio§lu Kale/Peium (fig. 16, p. 104) Strabo, describing the topography of Galatia, refers to two forts of the Tolistobogii:1 (fQoupioi S'duruv 6<rrt To T€ Bbouuiov ml To TTijiov, &v TO £v ftxoYAeco* Av)u>rJtpooi TO t<*$°9"Ac< KIOV. Cicero also refers to the establishments of his friend Deiotarus, but owing to confusion in the manuscripts, gives the name of both as Luceium (an obvious corruption of Blucium). It is probable that the name of the second fort was Peium. In the excavations at Karalar in the first third of this century an inscription was found which revealed that the site was the burial place of King Deiotarus, son of Deiotarus, both rulers of the Tolistobogii and the Trocmi. The elder Deiotarus was the friend of Cicero, 3 and his son, who died before him, was known from other ancient references. 4 Picard therefore identified Karalar with Blucium or Peium, while Coupry stated outright that-Karalar was Blucium.~* It remained to find a site for Peium. During his investigation of the Pilgrim's Route between Iuliopolis and Ancyra, Anderson identified a presumably Galatian hill-fort with the Ipetobroge of the Jerusalem Itinerary which describes the road from Istan bul, through Tarsus, to the Holy Land. The road itself was probably-built in the Hadrianic period along earlier paths.^ The site lies 90 km west of Ankara, on the opposite side of the Girrnir ^a,y from Dikmen, Anderson describes its situation as follows: 77. From the floor of the canon, in which the river flows, there rises a conical hill joined only by a low saddle to. the high, left bank; round this hill the river makes a bend exactly in the shape of an A, and its summit is crowned by a castle which commands a fine view of the valley below. The fortifications were naturally strongest on the side away from the river, where the towers still stand as - . • they were rebuilt in late Roman times. The southern one is shaped like an open hexagon, faced on the outer side with, old stones, —marble door-stones, and other rectangular blocks— and backed with opus incertum (small stones laid in beds of mortar). The other is of triangular shape and in its higher courses contains numerous old blocks; but the lower half of one face is of beautiful Greek work, built of rectangular blocks, squared along the edges and left 'free' in the middle, and laid in regular courses without cement (the three or four lowest courses projecting slightly in step fashion and being admirably fitted into the rock). On the sides overhanging the river the remains are purely Byzantine. It was disap pointing to find no inscriptions exposed to view.7 Mitchell agrees with Anderson that the site is Galatian and states that Anderson's Petobriga was definitely Strabo's Peium, the treasury of 8 Deiotaros. He strengthens his identification using evidence from other ancient sources. Cicero in the Pro Rege Deiotaro wishes to prove that Deiotaros did not try to murder Caesar. The latter was returning from 9 Zela in 47 B.C. and stayed at Deiotarus' fortresses on successive nights. The forts must therefore have been a day's journey apart. Dio tells us that Caesar was on his way to Bithynia;10 this information is also given in the Bellum Alexandrinum which states explicitly that "per Gallograeciam Bithyniamque in Asiam iter fecit"."'''1" Mitchell interprets this in the strictest sense—Caesar did not reach the province of Asia before he cros sed Bithynia. Therefore he could have taken only one road as he headed west: the Pilgrim's Route from Ankara to Iuliopolis through the north western part of Galatia, Peium, says Mitchell, must lie in northwestern Galatia, one day's journey from Blucium. Blucium we know to be Karalar, and Tabanlioglu 78. Ciftlik is 50 km away. The two forts are situated to the north, of the later main road. Further evidence comes from the life of a later Galatian (sixth or seventh, century A-D.1, St, Theodore of Sykeon, TTis<£v or IT^uv was a n 13 small village near the Ytoqiov SvKpcttZv t*fc Act£etvTt\/Hf - Eukraa remains obscure but the Lagantine is the area surrounding Lagania, a town on the Ancyra-Iuliopolis route, and perhaps to be identified with Dikmen Huyuk, a site south of Tabanlioglu Ciftlik. Thus St. Theodore's TTeZtv must be the TTtfiov of Strabo and the fortress must be situated near Lagania. 14 Tabanlioglu Kale satisfied both these conditions. 15 Mitchell gives a detailed description of the construction and masonry. As the passage quoted from Anderson indicates, fortifications were necessary only where the peninsula was not cut off by the river. Although there were subsidiary installations on the north and south sides of the hill, effort was concentrated on the unprotected east side. There are two styles of masonry, the careful Hellenistic and the irregular Byzantine. One of the earlier terrace walls is built of pseudo-isodomic, quarry-faced ashlar -~ blocks, laid as headers and stretchers, without mortar. The south tower, also of Hellenistic construction, was not as well built, but it is easily distinguishable from the Byzantine additions which contained re-used blocks of various periods with mortar and smaller stones. Mitchell reconstructs the Hellenistic fortifications as follows: there was a gateway approximately 2.65 m wide, flanked by two symmetrically placed polygonal towers about 14.50 m apart. Another polygonal tower lay 18 m to the north., with, its lower courses bonded into the terrace wall, so as to prevent access from a gully on the north side of the hill. One 79, tower is definitely hexagonal and it is assumed tha,t all three were built to the same plan. The thickness of the wall, which can be measured only at the gate, is 1.88m, The quality of the construction is extraordinarily high, Mitchell compares it with the fortifications at Isaura in the Taurus built by Amyntas, Deiotarus1 successor, in 25 B.C. Here too is a system of polygonal towers connected by a curtain wall; the gateways are similar in plan: "both are single arches relieved by a simple moulding at the point where the arch joins the upright".'''^ In both cases Mitchell concludes that the construction was accomplished by Greek workmen at the command of the local Galatian ruler. At Tabanliog'lu Kale Deiotarus would have given the orders for the stronghold suitable for a relatively small garrison to be built, sometime in the middle of the first century B.C. 1. Strabo, 12.5.2. 2. Cicero Pro Rege Deiotaro 17.21. 3. Cicero Ad Atticum V.17.3; Phil. XI xiii 31; Pro Rege Deiotaro 36. 4. Picard, Comptes Rendus de 1'Academie des Inscriptions et Belles-Lettres (1935) 42-44. 5. Coupry, RAVI (1935) 142f. 6. Anderson, JHS IX (1899) 53-54. See also plate IV, the map of Galatia. 7. Anderson, op. cit., pp. 63-64. 8. Mitchell, "Blucium and Peium: the Galatian Forts of King Deiotarus", AnatSt 24 (1974) 61-75,p. 73, n. 22. 9. Cicero Pro Rege Deiotaro 17,21. 10. Dio XLII.49-. 11. Cicero Bellum Alexandrinum 78. 80. 12. Mitchell, AnatSt 24 (1974) , p. 72. 13. Vita, St Theodore 118.2, ed. A.J, FestugiSre, 1970, cited by Mitchell, AnatSt 24 (19741, n. 17. 14. Mitchell, AnatSt 24 (19-741, pp, 72-73, See also n, 22 of the same article, in which Mitchell discusses the possible transference of the name Peium to the site actually on the Pilgrim's Route, a little south of the Galatian hill-fort, which, was known as Ipetobrogen or Petobriga. 15. Mitchell, pp. 4-6. 16. Mitchell, p. 7. Isaura is the modern Zengibar Kalesi, near Bozkir, southwest of Konya. For a photograph of one of the Isaura towers, see fig. 60 in F.E. Winter, Greek Fortifications, Phoenix, Supplementary Volume IX, University of Toronto Press, 1971. See also figs. 149 and 201 for plans of the towers, and 202 for the arch spanning the acrop olis gate. 4. Karalar/Blucium (fig. 17, p.105 ) The identification of Karalar with Blucium was discussed above with that of Peium. Karalar has been fairly systematically excavated, so that plans of the fort are available; the relationship of the fort (Asar) to the tumuli can be seen from the map of the site (fig. 6, p. 94 ). On the hill there are numerous cutting and below there is a man-made tunnel with steps leading down to a spring. Thus water was available if the fort was under siege. The palace of King Deiotarus was probably located beside his tomb on the opposite hill.1 Arik describes the fort in somewhat greater detail. The fort con sists of a rocky triangular prominence shaped like a camel's hump. In excavations at the southwest end, piles of cut stone, fragments of Roman and Byzantine pottery, roof-tiles, and ashes were found. Also found were the foot and base of a female statue in marble, of Graeco-Roman type, 2 and fragments of Hellenistic pottery, A number of coins emerged, mostly 3 Byzantine, a few Trajanic. Little sense can be made of the confused walls 81. and cuttings, which cannot all belong to the same period. The underground staircase (M) is located near a particularly complicated set of walls (A, B, C, etc.).. It originally had 56 steps, but 4 were eliminated to give 4 a larger landing at the foot. One thing is certain from the evidence of the pottery: occupation of the fort is at least Hellenistic if not earlier.' This fact and the other evidence cited in the section on Peium makes the identification of Karalar with Blucium virtually definite. 1. Mitchell, unpublished guide to the vilayet of Ankara, pp. 33-34. 2. Remzi Og"uz Arik, "Karalar Hafriyati", Turk Tarih Arkeoloji Etnografya  Dergisi 2 (1934), p. 152. 3. Arik, p. 154. 4. Arik, p. 158. 5. Arik, p. 163. 5. Karacakaya Karacakaya was first visited by Anderson who knew it as Soman Hissar.1 The pottery from the site was recognizably Hellenistic—a worn fragment of a Megarian bowl, and a piece of black glaze. Illegal excavation has re vealed part of a small enclosure of large roughly cut limestone blocks, set two thirds of the way up the hill. At the top of the hill are the remains of additional buildings, including a cistern lined with plaster to ensure a supply of water during any emergencies, Mitchell theorises that so small a site was a residence for a small group, perhaps a ruling family 2 in fear of neighbourly violence. 1. Anderson, JHS 1899, p. 87. 82. 2. Mitchell, p. 409. 6. Sirkeli Sirkeli is known locally as the.castle-caye or hissar magVrasi, The remains consist of traces of rock cuttings on the summit of the hill, and a thick scattering of pottery, some of it obviously Hellenistic."'' 1. Mitchell, p. 426. 7. Yarasli The position of Yarasli resembles those of Karalar and Sirkeli, in ... that all three of them are placed not on the highest available hill, but instead on the summit of a somewhat lower prominence. Yarasli is essen tially a plausible fort site covered with wall foundations and Hellenistic 1 and Roman pottery. 1. Mitchell, p. 440. 8. Dikmen Kale Dikmen is 90 km west of Ankara, north of Dikmen village, in the ter ritory of the Tolistobogii. It was first identified as a "Gaulish castel-lum" by J.G.C. Anderson in 1899.''" Anderson reported that its altitude-is 1700 ft.; thus the fort at the top of the conical hill provided an excel lent look-out post. He described the fort as a triangle with an entrance at the southern apex. In addition thare are three semi-circular bastions in the western wall, one in the middle of the wall, and one at the north west end.. The plan of the fort was partially dictated by the shape.of 83, the hill-top; loose fragments of the outcrop provided the building material. The walls are double: two separate.walls, fairly well constructed, with, a rubble core seven feet wide between them. They are preserved to a maximum height of eight feet.. The fort is small, measuring only thirty yards across, Anderson observed no pottery; nor did Mitchell, who has also 2 proposed the identification of Dikmen Kale as a Galatian hill-fort. An exact date is impossible to assign. The strategic importance of the fort is obvious from its position, and it is invisible from below, owing to a screen of oak trees and to the configuration of the hill itself. 1. Anderson, 1899, p. 64. 2. Mitchell, p. 417. Mitchell adds that "west of the Kale itself, is a grassy saddle enclosed by a second fortification, of uncertain date." 9. Tizke Tizke consists of a double circuit of rough stone walls, enclosing an area ca 28 by 20 metres. No distinctive sherds were found, either by 1 2 Anderson, who first suggested it as a Galatian hill-fort, or by Mitchell. 1. Anderson, JHS 1899, p. 63. 2. Mitchell, pp. 417-418. Mitchell adds that either Dikmen or Tizke could be the 'fytppo^ejjg XtofiiOv mentioned by St. Theodore, which from its name must have been a Galatian site (Vita St. Theod. 26a). 10. Assarlikaya While Mitchell has given good reasons for thinking that Blucium and Peium should be identified with. Karalar and Tabanlio^lu £iftlik respect ively, Anderson identified Blucium and Peium with two other sites, Basri, 84. a rather dubious Galatian site mentioned above, and Assarllkaya; both are located along the road from Pessinus to Ancyra.1 The fort consisted of two concentric dry-stone walls, enclosing a more or less circular area of 45 metres across-. There were several entrances, and, as well, numerous internal walls dividing the area .into small rooms. As usual, a hill-top provided the location for this Galatian lookout post. The double walls are reminiscent of those at Dikmen and Tizke, while the 2 internal divisions recall those at Karalar. 1. Anderson, JHS 1899, p. 94. 2. Mitchell, p. 433. SETTLEMENTS L. Gordion There are traces of a possible Galatian occupation at Gordion, which by the Roman period had received the Celtic name Vindia.1 Gordion at the 2 time of Manlius' visit was an emporium "celebre et frequens"; it declined after the Galatian defeat in 189. Level 2 of the Hellenistic strata has been recognized as the probable Galatian level. It contains mostly light buildings suitable perhaps for a farming village. There are no monumental buildings, and no structures with any conceivable public use. The city was unwalled at this point, the circuit walls having been neglected and abandoned after Alexander, and in some cases quarried for the stone blocks. Associated with, the scattered houses are grinding stones, agricultural implements, clay bee-hive ovens, pithoi, imported Greek pottery, terra cotta Cybele figures, and several coin hoards which, may have been Galatian 85. booty. There was definitely a gap between the Hellenistic settlement, 3 and the Roman road station. 1. Mitchell, p. 435. 2. Livy 38.18. 3. Rodney S. Young, AJA 68 (19641, pp. 279-280; AJA 64 (I960), p. 232; AJA 60 (1956), pp. 249-250. Cf. the tombs discussed in the previous section. 2. Tolgeri Huyiik Tolgeri Hiiyuk lies due west of Ankara; it was occupied from the Bronze Age onward, and may have been a medium-sized market town used by the Gala tians. The Hellenistic pottery found there includes a fair proportion of high quality fine red wares, not usually found on Galatian sites, with the exception. of Tavium. 1 1. Mitchell, p. 418. 3. Yalincak At Yalincak near Ankara, a possible Galatian village site has been ex cavated. Here there was apparently no gap between Hellenistic and Roman occupations. The pottery includes some Megarian bowl fragments, good in dicators of a Hellenistic date. Architectural remains consist of small rectangular houses and stone foundations, the mud brick super structure of which- is now lacking,1 1. Burhan Tezcan, Yalincak Village Excavation 1964, METU Archaeological Publications, Ankara 1966, pp. 14-15. 86, 4. Golhilyuk Go'lhtiyuk, which, lies south, of Ankara near Garbeous, was also inhabited in the Bronze Age, but the Hellenistic material, covering a large area around the mound as well as on it, is more conspicuous. The sherds are of good quality, and the Hellenistic occupation is part of a long habitation sequence. This suggests that the Phrygian elements of society may simply have absorbed some of the Galatian population into their own comfortable existence.1 1. Mitchell, p. 451. 5. BogazkSy As mentioned in the preceding chapter, there are burials at Bogazkoy which may be tentatively classed as Galatian. Evidence for possible Gala tian habitation is also forthcoming. The remains of Hellenistic buildings are plotted on an early sketch map of the site, and a general stratigraphic table for BogazkSy published some twenty years later refers to it as a 2 Trocmian outpost. In addition, houses with several rooms built at random on a slight slope have been found, and dated to the Hellenistic period. 3 These may be Trocmian residences. 1. K. Bittel and R. Naumann, Bogazkay-Hattusa I, Kohlhammer Verlag, Stuttgart 1952, p. 28 fig. 2, and cf. p. 34. 2. K. Bittel, Hattusha, The Capital of the Hittites, Oxford University Press, New York, 1970, p. ix, 3. Wulf Schirmer, Die Bebauung am unteren Buyukkale-Nordwesthang in Bogazko'y, Gebr. Mann Verlag, Berlin, 1969. WVDOG 81, Bogazkoy- Hattusa VI, pp. 12-14, Beilage 3 and 4. 87. CONCLUSION At the beginning of this- paper, the problems of Galatian archaeology-were described as endemic to the field. It is very difficult to work from literary sources, however detailed they may be on the subjects of history and politics, if they give little or no information of the ethnographic variety. The Galatians will remain an archaeological.conundrum until the evidence from the up by surveys and excavation. Through the investigation of pottery, burial customs, and forts and settlements, we have seen that certain methods are not necessarily helpful in this context. It is not possible to count on a strong Celtic heritage for the Galatians because of their long migrations. Thus valid connections with Europe will always be hard to find. Because so little work has been done on the Hellenistic period, placing the Galatians in the context of the Hellenistic period in Anatolia will be virtually impossible until this situation is resolved. The Galatians may or may not have made their own pottery; they may have buried their dead in a variety of ways; their settlements do not bear a definite architectural stamp which can be interpreted as exclusively Galatian. If we were dealing with a prehistoric period, it is possible that the presence of the Galatians in Anatolia would have gone undetected. The few fibulae and tores which seem definitely Celtic do not, for the most part, come from sites which- can definitely be called Galatian; these anomalous finds could easily be'explained as imports from outside the area. Since there are still no reliable indicators of Galatian material culture, archaeology can add little to the evidence provided by the hist-88. orical sources. This is not for lack of trying; in some instances scholars have been too optimistic in their attempts to find traces of the elusive Galatians. The evidence such, as it is does not at the moment permit a full archaeological reconstruction of the life and times of the pre-provincial Galatians. The future need not be so bleak, however.. As evidence accumulates on the general nature of life in Hellenistic Anatolia, it will become easier to situate the Galatians in the appropriate context. A somewhat similar case may serve as encouragement. The Scordisci were another group of Celts who settled in what is now Yugoslavia in the early third century B.C. They are mentioned in the historical sources less frequently than the Galatians, but it has been possible to reconstruct sev eral phases of their existence in Yugoslavia from their arrival until the area was absorbed into the Roman Empire. Detailed information now exists on Scordiscan pottery, burial customs, settlements, and even social organ ization.1 With continuing field work, it should be possible to fill in some of the gaps in our knowledge of pre-provincial Galatia. For the present, it is hoped that this paper has at least shown what the terms of the problems of Galatian archaeology are, and what obstacles must be overcome before they can be resolved. 1. Jovan Todorovic, Skordisci. Istorija i_Kultura, Institut za IzuSavanje Istorije Vojvodine Savez ArheoloSkih. DruStava Jugoslavije, Beograd, 1974, with. English, summary. 89. 1. :,Galatian" pottery 90. 2. Shapes of "Galatian" pottery .Belevi •I Mi las 4b, I§dir 4c, Gemlik 5b. Kirka§aj 6. Map of Karalar 9-5, 96. O 98. 100. 101. 13. Buckle from Bolu West 102. AJTEfc.S-.MITCHELL 0*FORD|C>.PHIL. I3>4-/ / 104. PEIUM (TABANLIOGLU KALE) "*(/>' "Ul •' V •' 'l'A\\'' SKETCH PLAN OF THE •! mi HELLENISTIC FORTIFICATIONS '''''' Position of fortress ( no scale) 16. Peium 105. 106. BIBLIOGRAPHY Standard editions have been used for the ancient authors cited, except where otherwise noted; similarly, standard abbreviations have been used for the titles of journals. J.G.C. Anderson, "Exploration in Galatia cis Halym", Part II, JHS XIX (1899) 52-132; 280-318. Sumer Atasoy, "The Kocakizlar Tumulus in Eskisehir, Turkey," AJA -78 (1974) 255-263. Remzi Oguz Arik, "Karalar Hafriyati," Turk Tarih. Arkeoloji Etnografya Dergisi 2 U934) 102-167. , "Les Tumuli de Karalar et la Sepulture du Roi Deiotaros II", RA 6 (1935) 133-151. Margarete Bieber, The Sculpture of the Hellenistic Age, revised edition, Columbia University Press, New York, 1970. P. Bienkowski, Die Darstellungen der Gallier in der hellenistischen Kunst, Alfred HSlder, Vienna, 1908. Kurt Bittel, Kleinasiatische Studien,. IstMitt 5 (1942). , BogazkOy IV. Funde aus den Grabungen 1967 und 1968, Mann Verlag, Berlin, 1969. , Hattusha, The Capital of the Hittites, Oxford University Press, New York, 1970. , "Bemerkungen zur sogenannten galatischen Keramik", pp. 227-237, in Melanges Mansel, Turk Tarih Kurumu Yaymlarindan Dizi VII, TTK Basimevi, Ankara, 1974. . and R. Naumann, Bogazkoy-Hattus'a I, Kohlhammer Verlag, Stuttgart, 1952. Clemens Bosch, "Die Kelten in Ankara," Jahrbuch fur kleinasiatische Forschung 2 (1953) 283-293. Jean Carpenter and Dan Boyd, "Dragon-houses: Euboia, Attika, Karia," AJA 81 (1977) 179-215. Rhys Carpenter, Greek Sculpture, University of Chicago Press, 1960. John L. Caskey, "Objects from a Well at Isthmia," Hesp 29 (I960) 168-176. Nora K. Chadwick and Myles Dillon, The Celtic Realms, Weidenfeld and Nicolson, London, 1967. J.M. Cook, The Troad, Oxford, Clarendon, 1973. R.M. Cook, Greek Painted Pottery, second edition, Methuen, London, 1972. W. Willson Cummer, "A Roman Tomb at Corinthian Kenchreai," Hesp 40 (1971) 205-231. G.M. Dalton, The Treasure of the Oxus, originally published by the British Museum at the Oxford University Press, 1926; republished by Lund Humphries, London, 1964. J. De*chelette, Manuel d'arch^ologie prShistorique, celtique, et gallo- romaine II, 1914. Walter von Diest, "Von Tilsit nach. Angora,u Petermanns Geographisches  Mitteilungen, Erganzungsheft no. 125 (289-8).. 107. Rilstem Duyuran, "Decouverte d'un tumulus pre-s de l'ancienne Dardanos," Anadolu V (1960). 9-12. G.R. Edwards, "Gqrdion 1962, "Expedition 5 (1963). 47-48, Afif Erzen, ilk, gagda Ankara, Turk Tarih- Kurumu Yayinlarmdan, VII Seri, no. 12, TTK Basimevi, Ankara, 1946, Jan Filip, Celtic Civilization and its Heritage, Publishing House of the Czechoslovak. Academy of Sciences and Artia, Prague, 1960 (English edition 1962).. B. Filov, "Die Kuppelgr'aber von Mezek," Bulletin de 1' Institut d'Archeologie Bulgare XI (19-37) 79-84. __, "The Bee-hive Tombs of Mezek," Antiquity XI (.1937). 300-304. Nezih. Firatli, "Bitinya Ara|tirmalarina Birkac llave," Belleten 17 (1953) 22-23. , "The Tumulus of Tersiyekoy near Adapazari," Istanbul Arkeoloji Muzeleri Yilligi 9 (1960) 73-76. , "Two Galatian Tumuli in the Vicinity of Bolu," AJA 69 (1965) 365-367. , "Jewellery Found at izmit," Istanbul Arkeoloji Muzeleri Yilligi 15-15 (1969) Nicholas and Joan Gage, "Treasures from a Golden Tomb," New York Times Magazine, December 25, 1977, pp. 14-19, p. 32. G.T. Griffith, Mercenaries of the Ancient World, Cambridge University Press, 1935. Esther V. Hansen, The Attalids of Pergamon, second edition, Cornell Univer sity Press, Ithaca and London, 1971. Christine Mitchell Havelock, Hellenistic Art, Phaidon, London, 1971. C.F.C. Hawkes, "The Celts: Report on the Study of their Culture and their Mediterranean Relations, 1942-1962," pp. 61-69 in Le Rayonnement  des Civilizations Grecque et Romaine sur les Cultures Peripheriques, VIII CongrSs International d'Archeologie (Paris 1963), Editions de Boccard, Paris, 1965. Wolfram Hoepfner, "Kammergrab in bithynisch-paphlagonischen Grenzegebiet," AthMitt 86 (1971) 125-139. Paul Jacobsthal, Early Celtic Art, Oxford, Clarendon, original edition 1944, corrected edition 1969. Felix Jacoby, Die Fragmente der Griechischen Historiker, vol. Illb, Leiden, E.J. Brill, 1950. A.H.M. Jones, The Cities of the Eastern Roman Provinces, Oxford, Clarendon, 1971. Donna C. Kurtz and John Boardman, Greek Burial Customs, Thames and Hudson, London, 1971. M. Launey, "Une Episode oublie"e de 1'invasion galate en Asie Mineure," REA XLVII (1944) 217-234. A.W. Lawrence, Greek Architecture, Penguin, Harmondsworth, 1967, second edition. David Magie, Roman Rule in Asia Minor, Princeton University Press, 1950 Ferdinand Maier, "Bemerkungen zur sogenannten galatischen Keramik von Bogazkoy, Jdl 78 (1963). 218-255, Arif Mtifid Mansel, Trakya-Kirklareli Kubbelj Mezarlari ve Sahte Kerner ve Kubbe Problem!, Turk Tarih. Kurumu Yayinlarmdan VI Seri, no. 2, Ankara, 1943. 108. , "Mudanyal.Mezar Binasi", Belleten 10 (1946) 1-12. , "Gebze YBresince kutluca Kubbeli.Mezari ve onun Trakya Kubbeli Mezarlari Arasmda Aldigi Yer," Belleten 37 (1973) 143-158. 't- , "Gemlik Tumiiliis Mezari", Belleten 38 (1974) 181-189.' James R. McCredie, Fortified Military Camps in Attica, Hesperia Supple-. .. ment XI, 1966. Roger McShane, The Foreign Policy of the Attalids of Pergamum (Illinois Studies in the Social Sciences No. 53), University of Illinois Press, Urbana, 1964. J.V.S. Megaw, "Two Finds of the Celtic Iron Age from Dodona," pp. 185-193 in Liber Iosepho Kostrzewski octogenario a_ veneratoribus dicatus, edited by Konrad Jazdzewski, Wrociaw, 1968. , Art of the European Iron Age, Adams and Dart, Bath, 1970. Machteld J. Mellink, "Archaeology in Asia Minor," AJA 65 (1961) 51-52. , "Archaeology in Asia Minor," AJA 67 (1963) 189 , "Archaeology in Asia Minor," AJA 75 (1971) 179. Yildiz Mericboyu and Sumer Atasoy, "The Kanlibag Tumulus at Izmit," Istanbul Arkeoloji Muzeleri Yilligi 15-16 (1969) 67-95; English summary pp. 91-95. Stephen Mitchell, The History and Archaeology of Galatia, D. Phil, thesis, Oxford University, 1974. , "Blucium and Peium: The Galatian Forts of King Deiotarus," AnatSt 24 (1974) 61-75. Winfried• Orthmann, "Untersuchungen auf dem Asarcik Hviyuk bei Ilica," IstMitt 16 (1966) 27-78. E. D. Phillips, "The Scythian Domination in Western Asia: its Record in History, Scripture and Archaeology," World Archaeology 4 (1972-72) 129-139. Stuart Piggott, Ancient Europe, Aldine, Chicago, 1965. Edith Porada, The Art of Ancient Iran, Methuen, London, 1965. Johann Potratz, "Die Pferdetrensen des alten Orient," Analecta Orientalia 41 (1966). T.G.E. Powell, The Celts, Thames and Hudson, London, 1958. W.M. Ramsay, A Historical Commentary on St. Paul's Epistle to the Gala  tians, Hodder and Stoughton, London, 1899. Anne Ross, Pagan Celtic Britain, Routledge and Kegan Paul, London, 1967. Nancy K. Sandars, Prehistoric Art in Europe, Penguin, Harmondsworth, 1967. Jorg Schafer, Hellenistische Keramik aus Pergamon, Walter de Gruyter and Co., Berlin, 1968 (Pergamenische Forschungen, Band 2) Wulf Schirmer, Die Bebauung am unteren Biiyukkale-Nordwesthang in Boffazkoy. Bog*azkSy-Hattusa VI, Mann Verlag, Berlin, 1969. Arnold Schober, "Die Gallierdenkmal Attalos I. in Pergamon," Mitteilungen  des deutschen archaologischen Instituts. Romische Abteilung 51 (1936) 104-124. Otto Seel, Pompei Trogi Fragmenta, Teubner, Leipzig, 1956. F. Stahelin, Geschichte der kleinasiatischen Galater, second edition, Leipzig, 1907. Tadeusz Sulimirski, "The Celts in Eastern Europe," pp. 182-190 in To Illustrate the Monuments. Essays on Archaeology presented to Stuart Piggott, ed. by J.V.S. Megaw, Thames and Hudson, London, 1976. 109. Zafer Tagliklioglu, "Dardanos gehri Yakmindaki Turauliiste Yeni Bulunan Grekce Kitabeler," Tarih Dergisi XIII.17.18 (1963) 160-173. Burhan Tezcan, Yalmcak Village Excavation 1964, Middle East Technical University Publications, Ankara, 1966. J. Todorovic, Skordisci. Istorija i_ Kultura, Institut za IzuEavanje Istorije Vojvodine Savez ArheoloSkih DruStava Jugoslavije, Beograd, 1974; with English summary. Diindar Tokgoz, "Igdir Kazisi Raporu," TiirkArkDerg 22 (1975) 151-153. Hermann Vetters, "Ephesos. Vorlaufiger Grabungsbericht 1971," AnzWien 109 (1972) 85-88. C. Bradford Welles, Royal Correspondence in the Hellenistic Period, Yale University Press, New Haven, 1934. Edouard Will, Histoire politique du monde hellenistique, II, Nancy, 1967. F.E. Winter, Greek Fortifications, Phoenix, Supplementary Volume IX, University of Toronto Press, 1971. Rodney Young, "The Campaign of 1955 at Gordion: Preliminary Report," AJA 60 (1956) 249-252. , "The Nomadic Impact: Gordion," pp. 52-57 in Dark Ages and Nomads c_. 1000 B.C., edited by Machteld J. Mellink, Nederlands Historisch-Archaeologisch Instituut, Istanbul, 1964. Ladislav Zgusta, Kleinasiatische Personennamen, Verlag der Tschechoslovak-ischen Akademie der Wissenschaften, Prague, 1964. 


Citation Scheme:


Usage Statistics

Country Views Downloads
United States 9 2
Germany 9 64
China 8 0
Turkey 6 4
Australia 3 0
United Kingdom 3 1
Spain 2 0
Ukraine 2 0
Japan 2 0
Serbia 1 0
Russia 1 0
Romania 1 0
City Views Downloads
Unknown 14 69
Guangzhou 4 0
Mountain View 4 0
Beijing 3 0
Bochum 3 0
Leipzig 2 0
Ashburn 2 0
Tokyo 2 0
Sydney 2 0
Bolu 2 0
A Coruña 2 0
Redmond 1 0
Gladesville 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items