^{page 280 note 1} Hist. Geog. p. 343. Though he does not give the modern name, the same site is referred to. He did not visit Eski-il and accepted the report of the people at the ruins.

^{page 293 note 1} The nights and early mornings are very cold in these regions even in summer.

^{page 298 note 1} Leo Diaconus, p. 170; H.G. p. 231 (cf. Class. Review, 1899, p. 137).

^{page 298 note 2} Atleast until Pissia became independent (if Piribeyli represent Pissia, as is most probable, § 5).

^{page 306 note 1} [At the time of writing there was another possibility, viz. that Pissia lay in the plain north-west of Amorion; but an examination of this district, which I made in Sept. 1899, shows that only small villages existed there. The identification with Piribeyli seems, therefore, practically certain.]

^{page 306 note 2} In 1. 2, my copy reads CYNBIOC and in 3 EAYTWN We did not see nos. xxii. and xxiii.

^{page 309 note 1} So the ensemble of rock chambers at Beybazâr is called Hissar (above i. § 8, p. 65.).

^{page 309 note 2} For an explanation of Hamilton's, Hergan Kale, see Annual Brit. School 1898, p. 71Google Scholar, n. 1.

^{page 311 note 1} The fact that I omitted to consult the map until the above argument was finished may be considered to lend some confirmation to it.

^{page 311 note 2} Annual Brit. School 1898, p. 50–1.

^{page 311 note 3} But the road from Mandri fontes (which he seems to place at In-Bunar, north-east of Mandra keui) is not ‘all down hill,’ as he declares, in seeking to account for the long day's march of 29 or 30 kilometres (20 MP.). There is a lofty pass to cross.

^{page 311 note 4} It seems doubtful if the Oroandeis are the Pisidian tribe (Orondeis). Staehelin is perhaps right in regarding them as the inhabitants of some town in the neighbourhood (Gesch. d. kleinas. Gal.). (Cf. Livy c. 19.)

^{page 311 note 5} So Polybius xxii. 20 (Schweighaiiser). Livy xxxviii. 18 says ne Tectosagis bellum inferret, which Dr. Körte prefers because Manlius ist ja seit geraumer Zeit im Gebiet der Tolistobojer. But Livy's only authority was Polybius, whom he certainly misunderstood: for, as Weissenborn pointed out, Eposognatos' ambassadors return with the news that his mission had failed and that the Gaulish chiefs had fortified themselves on Mt. Olympos,—i.e. the Tolistobogian chiefs, for the Tectosages retired to Mt. Magaba (c. 19). Polybius' meaning is perfectly clear, μή προεξαναστῆναι μηδ᾿ ἐπιβαλεῖν χεῖρας τοῖς Τολιστοβογίοις Γαλάταις [I see that Dr. Körte has now come to the same conclusion, in Woch. f. Klass. Philol. 1898, p. 5. He rightly remarks that the passage supplies important evidence as to Livy's historical method.]

^{page 312 note 1} From the fact that ‘Trojan’ pottery has been found on the site of what was probably the Phrygian city Gordion, Dr. Körte concludes that the Phrygians had conquered the country, not ‘about 900 B.C., as Ramsay thinks, …but more than five hundred, perhaps even a whole thousand, years earlier.’ (Athen. Mitth. 1897, p. 25.) But surely that conclusion does not necessarily follow.

^{page 313 note 1} For the significance of the altar see above, no. 142, and the references there given. For the importance of the door in Phrygian religious conceptions, see Ramsay, , CB. i. p. 99 ffGoogle Scholar. The door-stone continues in use even in Christian times: the raised division between the panels easily becomes a cross, and from it is developed the later stele with cross in relief occupying the centre of the field. I have noticed all stages of the transformation in Galatia.

^{page 313 note 2} Ath. Mitth. xxii. (1897), p. 49.

^{page 313 note 3} The Hellenization was little more than external: it did not really change the nature of Phrygian ideas. The god, for example, may be called by a Greek name (since Greek was spoken), but he remains the Phrygian god; ho is not a Hellenic deity, as the reliefs show. [In the great centres of life pure Greek ideas may be introduced, but that is a different thing.] Thus there is nothing to astonish us in the fact that in spite of Hellenization Phrygian conceptions maintain themselves with tenacity.

^{page 313 note 4} More fully in the article on Galatia already quoted.

^{page 313 note 5} Livy xxxviii. 16.

^{page 314 note 1} Ramsay l.c., improving on Van Gelder, , De Gallis in Graecia et Asia (1888), p. 183Google Scholar (who followa Perrot, , Mémoires d' Archéologie, 1875, p. 233–4)Google Scholar.

^{page 314 note 2} Van Gelder, p. 182.

^{page 314 note 3} Cf. III. § 6 above.

^{page 314 note 4} τριῶν δὲ ὄντων ἐθνῶν ὁμογλώττων καὶ κατ ἄλλο οὐδὲν ἐξηλλαγμένων (Strabo, p. 567).

^{page 314 note 5} Körte, , Ath. Mitth. 1897, p. 16.Google Scholar

^{page 314 note 6} Ramsay's interpretation (Class. Rev. l.c.) of an inscr. published by Körte, op. cit. p. 39. The suggestion is accepted by the latter.

^{page 314 note 7} Cf. Ramsay l.c. Körte, on the other hand, infers from Liv. xxxviii. 18 and 24 that Gordion and Ancyra were already conquered in 189 B.C. (Wochemch. f. klass. Philol. 1898, p. 3). But there is nothing to lead to this conclusion. As regards Gordion, Livy indeed says that Manlius found the town desertum fuga incolarum, refertum idem copia omnium rerum. But that only means that the Phrygian natives fled in fear, just as they had done in Phrygia proper (c. 15, metu omnibus circa oppidis desertis). The Romans did not show much consideration towards the towns they passed on the march.

^{page 314 note 8} A. E. Mitth. 1884, p. 95–101 (= Ch. Michel's Recueil no. 45): a Celt Attis (his brother is Aioiorix) is priest at Pessinus. (Attis is the permanent name of the chief priest.)

^{page 315 note 1} See below.

^{page 315 note 2} In the beginning of the first century: C.I.L. XIII. Part I., 3067.

^{page 315 note 3} De Boissieu, , Inscr. Antiques de Lyon, p. 96Google Scholar (Rushforth, Latin Hist. Inscr. no. 16), [now published in C.I.L. XIII. Part I., no. 1036]. Momnisen has pointed out (Rom. Prov., Eng. tr. i. p. 341) that the process of Hellenization is more rapid in Galatia than in Gaul, because of the influence of Greek neighbourhood: ‘the conception of the community as [)(ἔθνοσ] gains the predominance earlier than among the European Celts.’

^{page 315 note 1} Ath. Mitth. 1897, p. 47: cf. C.I.G. 4072 (Ancyrft.

^{page 316 note 1} Roman Prov., Eng. tr. p. 341.

^{page 316 note 2} Preface to Comment, ii. in Epist. ad Galatas (p. 430).

^{page 316 note 3} Jerome's statement is accepted by Mommsen, Ramsay, and Mitteis.

^{page 316 note 4} De Galatia prov. Rom. (1867), p. 87–90, 168–170; Revue Celtique, tome i.=Mémoires d'Archéologie (1875), p. 229 ff.

^{page 317 note 1} Lucian, Αλέξ ἤ ψευδόμαντ c. 51.

^{page 317 note 2} ‘Galatian’ merchants would be Greek-speaking natives (Phrygians, Greeks and Jews).

^{page 317 note 3} Cf. Annual Brit. School 1898, p. 60.

^{page 317 note 4} Yet Celtic would not appear much, if at all, on the tombstones, for, as Ramsay says, ‘All who wrote, wrote in Greek; the Gaulish language was a proof of barbarism and a reason for shame ‥ and it may be doubted whether any one could write who spoke only Gaulish’ (Art. on Galatia, quoted above).