¹ The inscription is identified in a brief report in Anatolian Studies XXIII (1978), 4Google Scholar. In 1984, while working on epigraphical material from the Kibyratis region, I was allowed to study the photographs and squeezes, now held in the Institute's archive in Ankara, and was encouraged by the Director, Dr. French, to publish the results. I am grateful to him for discussing the text with me, and to those colleagues who carried out the original work in Burdur, where they were assisted by Bay Kayhan Dörtlük, then Burdur Museum Director, and by his staff. Above all, our thanks are due to the Directorate-General of Antiquities and Museums of the Turkish Ministry of Culture and Tourism for help in securing the permit which made this work possible. While in draft, the article has benefited from Dr. Stephen Mitchell's advice.

² The height of the letters—3 cm.—suggests that they were designed to be read from close at hand. An altar within a sanctuary seems likely; alternatively, a large statue-base. For examples, see the material assembled by Tuchelt, K. in Frühe Denkmäler Roms in Kleinasien. Teil I. Roma und Promagistrate (Istanbuler Mitteilungen. Beiheft 23) (1979)Google Scholar, especially KIBYRA 01 (p. 159), with his discussion of the emplacement of dedications to Roma (pp. 21–45).

³ Bean (1959), 108–10.

⁴ The title σωτήρ τῆς οἰκουμένης is applied to the emperors Hadrian and Philip the Arab on honorific bases erected by Council and People in Attaleia (Bosch, E., Belleten XI (1947) 90 no. 4 and 93 no. 9Google Scholar). For separate bases which give the title to Diocletian and Maximian, see IGR 1, 789 and 790 (Heraclea-Perinthus). The title is unknown on milestones of Asia Minor of this period (information from David French).

⁵ The “large hüyük” could be the original site of Cormasa, captured by Manlius Vulso in 189 B.C. during his march from Termessus to Apameia, before he reached the territory of the Sagalasseis. See Bean's discussion of the evidence in “Manlius' Route in 189 B.C.” (Bean (1959) 113–117), which is based on his belief that Cormasa was at Eǧneş (now Çallıca), on the western side of the river (op. cit. 94). But if Hadriani/Hadrianopolis is now seen to lie west of the river at Çallıca, then Cormasa is likely to be located on the eastern side of the Lysis. Perhaps the name was later transferred to a new site at “Gâvur Ören”, while the original site became “Palaiopolis”, which Hierocles placed somewhere between Olbasa and Lysinia and which issued coins in the second and third centuries. Ramsay had proposed “Giaour Euren” for Cormasa (see AJA IV (1888) Plates II and III, facing p. 20Google Scholar and Cities and Bishoprics I (1895), 326–7Google Scholar) on good topographical grounds, but he later changed his mind and moved it to a site close to Kestel, Lake (The Social Basis of Roman Power in Asia Minor (1941), 239)Google Scholar. The evidence for Palaiopolis was given by von Aulock, H. in Münzen und Städte Pisidiens Teil I (Istanbuler Mitteilungen. Beiheft 19) (1977) 40–1Google Scholar, where he also summarized the earlier opinions about its location, with a catalogue of its coins (110–14) which date from Antoninus Pius to Severus Alexander, and which include two (nos. 1107 and 1108) carrying the words ΠΑΛΕΟΠΟΛΕΙΤΩΝ ΘΕΜΙС, with a scene of athletes drawing numbers (cf. Robert, L., Hellenica VII, 107Google Scholar n. 5; XI/XII, 363 f.)

That the name of Cormasa was preserved is shown by the bilingual funerary inscription found by Bean (in BoǧaziÇi) in front of the mosque, near the bridge (Bean (1959) 93–4 no. 42 with Plate XVI (b) and (d)), which commemorates “C. Iulius Cf. Papiria natus Cormasa”, a veteran of Legio VII. Bean argued (op. cit. 94) that, “Whether the stone belongs to BoǧaziÇi or Eǧneş, makes no practical difference; the two places are separated by 5 miles of plain, and must … have belonged to a single city. We may, I think, regard this stone as cogent evidence that Cormasa was at Eǧneş.” However, it is more likely that this tombstone, like the stones in the bridge, was brought originally from the eastern side of the river, and Bean's conclusion is therefore unsound. Furthermore, a position on the eastern bank, close to the river, makes better sense of the reference to Cormasa in the bilingual inscription which presents an edict of Sotidius Libuscidianus, governor of Galatia under Tiberius, which specifies what transport must be provided by the Sagalassenses to those on official business. “Praestare autem debebunt vehicula usque Cormasa et Conana.” (Latin text line 13, repeated in Greek in line 32) (Mitchell 1976, 107). The boundaries of the territories of these two cities represent the limits of the obligation imposed by the governor. Conana (at Gönen, see Fig. 1) had a territory which bounded that of Sagalassus to the east of Lake Askania, while that of Cormasa will have adjoined Sagalassus to the south-west. It was, according to Ptolemy V, 5.5, part of Pamphylia, or rather, of that part of Pisidia assigned to Pamphylia, and therefore lay within the province of Galatia. Since the boundary of the provinces of Galatia and Asia lay on or close to the Lysis it is likely that Cormasa lay east of the river.

⁶ His sources are unknown, but certainly Greek.

⁷ .

⁸ Cf. How, W. W. and Wells, J.: A Commentary on Herodotus (2 vols.) (1912), Vol. 2, 158Google Scholar, ad loc: “The retention of the plaid and fibula is the characteristic noted, appropriate to Highlanders.” Was this based on an eye-witness account? These Milyae presumably reached Greece and fought there under Badres.

⁹ Stein wanted to put Πισίδαι in the lacuna at the beginning of ch. 76. How and Wells suggest that it may be better to take from III, 90 the name of one of the smaller tribes, since “… Herodotus never mentions these unruly mountaineers, who probably never as a people acknowledged Persian sovereignty” (VII, 76, ad loc.); but this is a weak argumentum ex silentio.

¹⁰ The Kabalis extended westwards from the Milyas as far as Caria (see Fig. 1), and the later Tetrapolis, headed by Kibyra, seems to have included most of it. It is the area north of Massicytus (Ak Daǧlar), centred on the plain of the upper Indus and Lake Gölhisar.

¹¹ For the relationship between Ptolemy's history and the Anabasis, see most recently Bosworth's, A. B.A Historical Commentary on Arrian's History of Alexander Vol. I (1980), 16–34Google Scholar, “Arrian's historical methods.”

¹² Anabasis I, 24–5Google Scholar.

¹³ Alexander's close relationship with his chief seer, Aristander of Telmissus, for whose prophecies he had great respect, probably assisted this part of the campaign. (Bosworth, op. cit. 97–8; 186)

¹⁴ The absence of Tlos from this list is unlikely to be significant.

¹⁵ . I, 24.5.

If the Milyas extended north of Burdur Lake (Askania lacus) as the evidence suggests (see Fig. 1), then it was contiguous with Phrygia, and could easily have been administered from Kelainai.

¹⁶ The subordination of the Milyas to Lycia may have been the Persian King's recognition of the conquests of Pericles of Limyra in the 370s B.C. He had defeated Artembares in Western Lycia; had nearly captured Phaselis; and probably made inroads into the Milyas. For a recent discussion of the evidence for this period of Lycian history, see Childs', W. P. “Lycian Relations with Persians and Greeks in the fifth and fourth centuries re-examined.” (Anatolian Studies XXXI (1981), 55–80, especially 72–7.)CrossRefGoogle Scholar

¹⁷ J.H.S. LXXVIII (1958) 102–20Google Scholar, “Alexander's March from Miletus to Phrygia”. The article is also published as an Appendix to her book Alexander's Path (1958).

¹⁸ In the fourth century B.C. the Kibyrateis may not yet have been concentrated in an urban centre, although the site on the island of Lake Gölhisar—which I identify as Sinda (see “The Site of Sinda”, Anatolian Studies, forthcoming)—was certainly occupied at this date. It took Lydian and later Pisidian invaders to organize city government at Kibyra.

¹⁹ Most travellers and scholars since Ramsay, (Ath. Mitt. X (1885), 340)Google Scholar have concluded that Isinda must lie close to Korkuteli. This was accepted by Bean, G. E. in his entry in The Princeton Encyclopedia of Classical Sites (1976), 417Google Scholar—ISINDA (Kışla)—while admitting that the remains are slight. David French will soon publish his own interpretation of the evidence, and so I have omitted the site from Fig. 1.

²⁰ The modern road from Fethiye to Korkuteli follows this route.

²¹ Lane-Fox, R.—Alexander the Great (1975 ed.) 144Google Scholar—agrees that Alexander was anxious to make contact with Parmenio.

²² See Fig. 1.

²³ Arrian, , Anabasis I, 28Google Scholar.

²⁴ Diodorus Siculus XVIII, 44–7.

²⁵ For a recent attempt to establish the position of Cretopolis, which he identifies with Keraitai, see “Keraitae Araştırma Raporu” by Dörtlük, K., Türk Arkeoloji Dergisi XXIII (1976) 17–23Google Scholar. The name ‘Cretopolis’ may reflect the legendary Cretan origin of the Lycians (Herodotus, I, 173, quoted above), although the place is far removed from the historical Lycia.

²⁶ See Strabo's definition (XIII, 631) below.

²⁷ For comments on this passage, see F. W. Walbank (1957), 597.

²⁸ No site has yet been confirmed for Pednelissus. In any case, Garsyeris and his army had to cross the Cestrus at some point.

²⁹ .

³⁰ Information supplied by David French.

³¹ Bean has shown that some minor cities existed in this area in the Hellenistic period. (Bean 1960). An inscription mentioning Pogla shows a rudimentary city organization (op. cit. p. 56 no. 103, found in Çomaklı): Comama, Verbe, Andeda and Sibidunda are the other cities, but there was no large grouping comparable to the Sagalasseis. The region could be treated like a corridor by invading armies.

³² For discussions of the meaning and significance of this phrase, see Meyer, Ernst, Die Grenzen der Hellenistischen Staaten (1925), 145–56Google Scholar “Kleinasien nach dem Frieden von Apamea”; D. Magie (1950) II, 755–8; and Walbank (1979), 157.

³³ σον Polybius XXI, 45.10

³⁴ Hieronymus, in Dam. XI, 15Google Scholar.

³⁵ Fragmente der Griechischen Historiker 86 F 16 (Jacoby)Google Scholar.

³⁶ Meyer, Eduard, Geschichte des Königreichs Pontos (1879), 53, 1Google Scholar.

³⁷ Pliny, N. H. V, 95. See below.

³⁸ For Rhodian control, see Livy XXXVII, 56, 5 f.

³⁹ Soon to appear in Epigraphica Anatolica. An Attalid ruler, probably Attalus II, addresses the authorities of Olbasa soon after the end of a Pisidian War, during which the affairs of the city had been seriously affected.

⁴⁰ For a brilliant description of the Lycus valley, see Ramsay (1895) Ch. I.

⁴¹ David French has recently confirmed that its caput viae was Pergamum, not Ephesus (Anatolian Studies XXXIV (1984), 11Google Scholar). The relationship of the various roads in this area, on both sides of the Lysis, is not yet clear, but the line of the Via Sebaste, as traced by French in 1975, is set out by Stephen Mitchell (1976), 122 and n. 90. It crossed the Lysis at Boǧaziçi and passed between there and Elmacık before crossing the mountains to Aziziye on the way to Comama.

⁴² If the Pisidian Milyas was still attached to Greater Phrygia, it may have been given to Ariarathes. The unsatisfactory nature of this arrangement became visible later. An attempt to show that the Romans soon came to regard Pamphylia as attached to Asia is made by Sherwin-White, A. N. in J.R.S. LXVI (1976), 1–14Google Scholar, in “Rome, Pamphylia and Cilicia.”

⁴³ That the Araxa decree must be dated after 167 B.C. is demonstrated by Sherwin-White, A. N. in Roman Foreign Policy in the East 168 B.C.–A.D. 1 (1984), 49–51Google Scholar. For Kibyra's treaty with Rome, see OGIS 762 and Magie (1950) II, 967–8, 1122–3. The original connexion of Termessus with Rome cannot be much later than 167 B.C.

⁴⁴ The name may have been used by this period to distinguish the inhabitants of the Pisidian Milyas from those parts of the Milyas further south whose cities were now attached to the Lycian League, while the term “Milyae” may have been retained to cover both groups.

⁴⁵ Magie (1950) II, 1165–6.

⁴⁶ Among whom E. Will believes that Pamphylia and the Milyas appear “incontestablement comme le centre de la nouvelle province.” (Histoire politique du Monde Hellénistique Vol. 2 (1967), 409Google Scholar) and in view of the strategic importance of the region for communications between Asia and Pamphylia, this seems likely, although it is no doubt true, as Magie suggests, that Verres “took grain from outside the limits” of the province (Magie (1950) II, 1166).

⁴⁷ χώρας

⁴⁸ See Levick (1967) Appendix V. The Homanadensian War. (pp. 203–14) and Hall, A. S. “The Gorgoromeis”, Anatolian Studies XXI (1971), 127Google Scholar, Fig. 2, “The Homonadeis and their neighbours”.

⁴⁹ Under excavation since 1971 by an Ankara University archaeological team, led by Professor Cevdet Bayburtluoǧlu. For the most recent of preliminary reports, see VII. Kazı Sonuçları Toplantısı (Ankara 1985), 357–71Google Scholar.

⁵⁰ Hellenica XI–XII, 353 n. 4, 596, for bibliographyGoogle Scholar.

⁵¹

⁵² Who lived in the plain of Karayük, north of Kibyra.

⁵³ Tabai and Kibyra were and the legendary founders of Kibyra, Tabai δῆμοι ἀδελφοι and Kidrama were brothers (La Carie II, 73–76, 88, 136—7Google Scholar). Kibyra had a Θεά πισιδική on its coins, and the Θεά πισιδικοί are recorded throughout the region.

⁵⁴ Ramsay (1895) 280–6.

⁵⁵ “Myodia” may be a distortion of “Miluadika”.

⁵⁶ The history of the Imperial estate in the area of Tefenni, which belonged in the second and third centuries A.D. to Annia Faustina and her relatives, is well shown by inscriptions. For a list of them and a review of the interpretations of Ramsay and other scholars, see RE. “Ormeleis” (Ruge).

⁵⁷ Magie (1950) II, 761–2 gave an admirable review of previously published scholarship. Bean (1959, 1960) considered that the territory covered by the first of his two articles (1959) “coincides approximately with the Milyas as defined by Strabo 631 and nowadays with the Eastern half of the vilayet of Burdur” (1959, 67).

⁵⁸ RE “Lykia” (Jameson).

⁵⁹ Zgusta, L., Kleinasiatische Ortsnamen (1984)Google Scholar gives examples.

⁶⁰ For the history of Pisidian settlement in this area, see Coulton's, J. J. “Termessians at Oinoanda” (Anatolian Studies XXXII (1982) 115–31CrossRefGoogle Scholar. For the foundation of pan-Lycian games by M. Aurelius Artemon and his wife, early in the third century A.D., to which the citizens of Termessus (συνένίς πόλις) were invited, see the text published in BCH 1900, 338–41, no. 1Google Scholar: … τῆς ἀνέκαθεν συνγενίδος ὑπαρχούσης Τερμσσέων τῆς Παμφυλίας (lines 17–19).

⁶¹ BCH XVI (1892) 436–8Google Scholar.

⁶² The Calder, /Bean, Classical Map of Asia Minor (1957)Google Scholar placed it without a query on this site. Doubts were later expressed by Bean (1960), 79–80 with nn. 74–78, and emphasized in his brief description in the Princeton Encyclopaedia of Classical Sites (1976), 570Google Scholar: “The location at Melli depends wholly on the similarity of names.”

⁶³ Die Grenzen der Hellenistischen Staaten (1925), 4Google Scholar.

⁶⁴ Previously known evidence for the presence of negotiatores in this area is summarized by Mitchell (1976, 116–7) and by Levick (1967, ch. IV–V). The nearest are at Apameia and at Conana. We see that a substantial group must also have been present in the Lysis valley. They were reinforced by the settlement of veteran colonists not only in the official Augustan colonies but also in other cities, such as Apollonia, demonstrated convincingly by Mitchell (1978 = “Roman Residents and Roman Property in Southern Asia Minor”, Proc. Xth International Congress of Classical Archaeology 1973 (1978) Vol. I, 311–318), where he defines the negotiatores as “the merchants and businessmen who seized the commercial opportunites offered by trade, money-lending and even light industry in the new territories. They tended to form communities within the larger and more prosperous cities”; with reference to Hatzfeld's, J.Les Trafiquants italiens dans l'Orient hellénistique (1919)Google Scholar. The negotiatores in the Lysis valley may have arrived only after the creation of the province of Galatia, although their presence in Attaleia points to an earlier date.

⁶⁵ MAMA VIII no. 350.

⁶⁶ For Macedonian settlers in this region, see the evidence assembled by Getzel Cohen in The Seleucid Colonies (Historia Einzelschriften Heft 30, 1978) and for a summary see 9–11, 14–16.

⁶⁷ Jones, Cities², 411.

⁶⁸ Levick (1967) 17, n. 6.

⁶⁹ REG. LXXI (1958), 320, no. 467Google Scholar.

⁷⁰ Mitchell (1978) 316.

⁷¹ It probably took the line later followed by the Via Sebaste as far as Comama (see Mitchell 1976, 118, Fig. 1, based on work by David French).

⁷² For Thracians at Apollonia, see above, n. 64 and comments by Robert, L. in Noms indigènes dans l'Asie Mineure Gréco-Romaine I (1963) 357 n. 2Google Scholar. At Olbasa, a dedication to the Thracian god Maron (IGR, 3, 410). Σευθος was recorded at Yazı Köy, near the south-west tip of Burdur Lake, by Ramsay (1895, 336, no. 168) who asked the shrewd question, “Was a settlement of Thracian mercenaries made on Lake Askania in the Pergamene period?” They may have been planted there to help cope with Pisidian incursions during the Πόλεμος πισιδικός around the middle of the second century B.C.

⁷³ For a detailed consideration of how this was carried out, see Barbara Levick's monograph, (1967), to be supplemented by Stephen Mitchell's article in Historia XXVIII (1978) 409–438Google Scholar, “Iconium and Ninica,” which adds these two places to the list of Augustan colonies.

⁷⁴ By Cornutus Arruntius Aquila. See Levick (1967), 38–40, with notes and references to milestones.

⁷⁵ For the date of the final conquest of the Homonadeis, see Levick (1967) Appendix V (2), 206–13, who prefers a date later than 6 B.C., and Syme, R., “The Titulus Tiburtinus,” Vestigia 17 (1973) 592Google Scholar, “close to 4 B.C.”

⁷⁶ Studia Pontica III no. 66, pp. 75–86Google Scholar. No absolute proof can be adduced for the existence of a Κοινὸν Παφλαγονίας. Kornemann proposed such a body in RE. Suppl. IV (1924) 929–41. Deininger, J. accepted the proposal, Die Provinziallandtage der römischen Kaiserzeit (1965) 69Google Scholar, in dealing with “Die kleinasiatische koina,” and comments (69, n. 2), “Dass Paphlagonien, Pisidien und Isaurien eigene (landschaftliche) Koina bessassen haben (Kornemann [1924] 932, nr. 5) besagt noch keineswegs, dass sie nicht auch dem Provinziallandtag ausgeschlossen gewesen wären; wie es denn z.b. in der Provinz Asia auch in der Kaiserzeit noch eine ganze Reihe landschaftlicher Koina gegeben hat.”

⁷⁷ Mellor, R., ΘΕΑ ΡΩΜΗ. The Worship of the Goddess Roma in the Greek World (Hypomnemata Heft 42, 1968)Google Scholar; and “The Goddess Roma” (ANRW II. 17.2)Google Scholar.

⁷⁸ C. P. Jones suggested in his review of Mellor (1968) that the Smyrnaeans could claim the first temple, but need not have been the first to introduce the cult itself.

⁷⁹ For the inscription listing these Galatian landowners who held priesthoods in the Ancyra temple of Roma and Augustus, with their contributions, see OGIS 533 = Bosch, C., Quellen zur Geschichte Ankaras im Altertum (1967) 35–49Google Scholar no. 51.

⁸⁰ Waelkens, M.—Devreker, J., Les Fouilles de la Rijksuniversiteit te Gent a Pessinonte 1967–1973 I (1984)Google Scholar.

⁸¹ S. Mitchell's report on the progress of a survey of Pisidian Antioch, “Pisidian Antioch 1983”, Anatolian Studies XXXIV (1984) 9Google Scholar.

⁸² The best recent discussions of Augustan ideas and symbols are those by Fergus Millar, in his article “State and Subject: the Impact of Monarchy” (in the collection Caesar Augustus: Seven Aspects (1984) ed. by Millar, F. and Segal, E.)Google Scholar who emphasizes that “the sudden outburst of the celebration of Octavian/Augustus was a new phenomenon” (53); and by Price, Simon in Rituals and Power: the Roman Imperial Cult in Asia Minor (1984)Google Scholar.

⁸³ Information from David French.