² Thus Prof. Ramsay's disbelief (expressed e.g. in Cities of St. Paul, p. 250) in Hamilton's identification (Researches in Asia Minor, i. p. 474) was confirmed.

³ We intended to track it down to Antioch later, but time failed us.

⁴ See note 11.

⁵ Strabo 577.

⁶ See Fig. 14.

⁷ Seen and drawn by Sir W. M. Ramsay.

⁸ Cf. Prof.Ramsay, , Hist. Comm. Gal. p. 43Google Scholar; Pauline and other Studies, p. 163.

⁹ Pauline and other Studies, p. 163.

¹⁰ Here, as elsewhere, we do not guarantee the measurements. For accurate measurement excavation is needed.

¹¹ It needs a considerable effort of the reconstructive imagination to discover a Sacred Way in this part: the only place where its course seemed quite certain was to the north of the ravine; but the lie of the ground is decisive.

¹² Sterrett, Epig. Jour. No. 101.

¹³ Mr. Calder observed traces of charring on some, due probably to shepherds' fires.

¹⁴ See Prof.Ramsay, in Athenœum, Aug. 12, 1911Google Scholar.

¹⁵ The orientation is not exact: see infra, p. 118.

¹⁶ Their widths (starting from the corner C) are 2′3″;, 2′ 1″, 2′ 1″, 2′, 2′2″, 2′ 1″ respectively: the distances between are 21′8″, 22′7″, 21′9″, 39′ 11″, 38′ 8″. The first is 14′ 10″ distant from the break H. They were built contemporaneously with the wall.

¹⁷ In the present state of the evidence, Sir W. M. Ramsay finds himself unable to agree with Mr. Calder and myself on this point.

¹⁸ We had intended to return to photograph these when the sun should suit, but unfortunately we were always unable to get away from the ἱϵρόν until night-fall.

¹⁹ Wolff, G., Ueber die architektonische Beschaffenheit der Mithrasheiligtümer, p. 90Google Scholar; Cumont, , Textes et Monuments de Mithra, i. p. 58Google Scholar.

²⁰ It is distant 96, 76, 46, and 58 feet from the S., N., E., and W., sides respectively.

²¹ Wiegand, , Milet, ii. pp. 73 f.Google Scholar; Herzog, , Arch. Anz. 1903, p. 187Google Scholar.

²² See especially the restoration of the altar at Miletos, , Arch. Anz. 1902, p. 154Google Scholar, Fig. 10.

²³ Pauly-Wissowa, , s.v. Altar, p. 1674Google Scholar; Wiegand, and Schrader, , Priene, p. 241Google Scholar; Pergamon, viii. No. 68.

²⁴ Petersen, , Jahresh. 1906, p. 310Google Scholar.

²⁵ The chief arguments see Men in the Iranian Moon-god, Mao or Maonha (Roscher's Lexikon, s.v.) and in the god Lunus of Carrhae (Spartianus, , Caracalla, vi. 6Google Scholar; viii. 3). Men is also frequently found associated with deities of undoubted Persian origin, such as Mithra and Anaitis. See the writers in Roscher's Lexikon (s.v. Men) and Daremberg and Saglio's Dict. des Ant. (s. v. Lunus), and also Perdrizet, M. in Bull. Corr. Hell. 1896, pp. 91 ffGoogle Scholar.

²⁶ Strabo 557: Reinach, Th., Mith. Eup. pp. 240Google Scholar ff.

²⁷ Strabo 512, 559, 733.

²⁸ In Eusebius, , Praep. Evang. vi. 10Google Scholar, 16.

²⁹ Strabo 733.

³⁰ Perrot, and Chipiez, , Hist. de l'Art, tome V. p. 643Google Scholar, Fig. 396 (in the English edition, Persia, p. 244, Fig. 116).

³¹ The legends are collected by M. Cumont, op. cit. Vol. I. pp. 127–8.

³² At the time of the expedition of Cyrus. the Pisidians were independent and hostile to the Persians (Xen., Anab. III. ii. 23Google Scholar; cf. I. i. 11. The references I owe to Mr. Calder).

³³ See SirRamsay, W. M. in Expositor, Sept. 1911, pp. 260 ffGoogle Scholar. The evidence proving that Antioch was a Phrygian city has often been collected, and is conclusive.

³⁴ Nos. 1–51 are from the West Wall. The provenance of the others, when known, is stated in the commentary on each. The figures make a rough attempt to reproduce their appearance.

³⁵ [My copy shows ᾿ Α]ρλέξαν[δρος] with great hesitation. The intrusion of Ρ is unexplained. The letters after Ρ are broken, and only tops remain.—W.M.R.]

³⁶ See the following article, p. 167.

³⁷ Wolfe Exped. Nos. 369, 370, and 372.

³⁸ Published Cl. Rev. 1905, p. 419; Studies, pp. 329–330.

³⁹ Studies in the History of the Eastern Roman Prov. (often quoted below as Studies, and the inscriptions in the final paper as Q 1 etc.).

⁴⁰ Expositor, Sept. 1911, pp. 270–1.

⁴¹ Expositor, loc. cit.

⁴² See especially Nos. 13, 34, 64, 65, 68.

⁴³ Loc. cit.

⁴⁴ Both are published by Calder, in J.H.S. 1911, p. 196Google Scholar; Uoê denotes the same local god as Dionysos Iuô: the dedication is at Serai-Ini, but the tribal epithet shows that the god belongs to the Orondian mountains, where Dionysos was at home. Miss Ramsay's Report, quoted by Mr. Calder, was only privately printed: it contains the identification of Iuo, and will be published.

⁴⁵ Le Bas-Waddington, , Asie Mineure, 668, 1607Google Scholar. On Men Askaenos, see Ramsay, , Cities and Bish. of Phrygia, II. p. 360Google Scholar.

⁴⁶ Hom., Il. xiii. 793Google Scholar.

⁴⁷ The dedications belong as a general rule to the humbler classes, and a board of magistrates is not to be expected.

⁴⁸ See MrCalder, in Cl. Rev. 1910, p. 79Google Scholar.

⁴⁹ [The name in the second line seems probably to be a nomen. My notebook suggests Λούκιος Πουβλου[λ]ίου υἱός a degenerate usage for L. Publilius, L.f.; but the reading on the stone is ΠΟYΒΝΟYΑΙΟC —W.M.R.]

⁵⁰ From the Athenœum, Aug. 12, 1911.

⁵¹ [That ΑYΝΜΗC should be a native Pisidian name, like ΟYΙW No. 12 and ΛΟΛΟY is not impossible. I missed this inscription. —W. M. R.]

⁵² [I read Γάμος on the stone; and, though the change of Μ with oblique sides to Λ Λ is easy, yet the name Gamus occurs both in Greek and in Latin characters in a Bithynian bilingual inscription, published by Dr.Wiegand, in Ath. Mitth. 1908, p. 151Google Scholar, and no change is necessary. —W. M. R.

⁵³ [The full form of Φλαούιος is not inconsistent with its use as a pseudo-praenomen, as was frequent under the second Flavian dynasty. —W. M. R.]

⁵⁴ [That Dionysius was slave or freedman may be taken as certain.—W. M. R.]

⁵⁵ Prof. Sterratt took Τέκτων and Σκυτέος (for σκυτϵύς) as personal names, see Ramsay, , Cit. and Bish. I. pp. 311 fGoogle Scholar. in correction of Sterrett, , Epig. Jour. 53Google Scholar B 32, 41 A 20.

⁵⁶ [My copy suggests [Π]ο[μπ]ϵανός as possible, but this did not occur to me before the stone.—W.M.R.]

⁵⁷ The comparison is Mr. Calder's. [Compare also the discussion of the form Koundoia, in Studies in the Eastern Provinces, pp. 365 f.Google Scholar, and the examples there quoted.—W.M.R.]

⁵⁸ The double nomen would be unfavourable, though not fatal, to this opinion.

⁵⁹ [The reading is perhaps ΜΤ or more probably ΛΙΤ.—W.M.R.]

⁶⁰ Sterrett, Epig. Jour. No. 142; his transcription is corrected as above by Prof.Ramsay, in Expositor, Oct. 1888, p. 263Google Scholar. It is perhaps more probable that Ἴα is gen. of a masc. name Ias: Artemeis, son of Ias, to his wife Manto or Mantou, as Prof. Ramsay now believes.

⁶¹ [In copying the inscriptions I felt confiden that they were memorials of two freedmen of the same lady. In the second case the stone has L, carelessly engraved for C.—W.M.R.]

⁶² [This was the first inscription that I read: my companions had read it on the previous day: we ought to have gone back to it after our eyes had become used to the character and look of the letters on these stones.—W.M.R.]

⁶³ I am indebted to Mr.Calder, for the references. In Histor. Comm. on Galatians, p. 201Google Scholar, Prof. Ramsay preferred Αἴπη.

⁶⁴ Inscriptions of this class, confessions with atonement, are common in certain parts of Asia Minor. A number are given in Ramsay, , Cities and Bish. of Phrygia, i. pp. 149–154Google Scholar; and he has collected others in a series of articles in the Expository Times, 1899; but many more are now known. They were sometimes called ἐξονπλάρια, a borrowed Latin term.

⁶⁵ It must, of course, be assumed that all civitate donati received a Roman name.

⁶⁶ As Mommsen says, l.c. p. 452 gentilicia Romana abhorrent a consuetudine Graeca.

⁶⁷ The two cases are sometimes hardly distinguishable by mere names.

⁶⁸ Iconium as a colonia of Hadrian, receiving probably no Roman population, but merely higher rights than previously, presents a total difference in character from Antioch, as inscriptions show clearly. So also probably would be the case with Julia Augusta Germa in Galatia, or Julia Augusta Ninica Claudiopolis in Tracheiotis, both (as Professor Ramsay holds) founded by Domitian and named after his ill-starred niece Julia Augusta, if their epigraphy were known.

⁶⁹ (This can no longer be said; there are some names at Antioch purely Greek in form; but even as to these some doubt remains about civitas, as is stated later. W.M.R.)

⁷⁰ I have incorporated note 28 (from p. 446) and made, at Prof. Ramsay's suggestion, some slight changes, additions, and abbreviations in the text. The remark about Secundus was proved right in 1911, when we found that his fuller name was Saturninus Secundus. He governed Pisidia Provincia in the fourth century.

⁷¹ Ibid. p. 272.

⁷² Ibid. p. 278, the following paragraph on that page stating the further problem.

⁷³ The Latin votive formula was added by persons who wrote the rest of the dedication in Greek: in such cases we must understand that the household was Roman.

⁷⁴ One certainly in the early fourth century, No. 68.

⁷⁵ Ramsay, , Studies pp. 305–380Google Scholar.

⁷⁶ In Studies, p. 355 he says ‘this observation…. is now abundantly justified’ by observation during nearly thirty years.

⁷⁷ One could hardly hesitate at first sight to declare that Lyciscus son of Athenion was a simply Hellene incola; but one remembers that Nikandros, son of Menekrates, was a Roman; and hesitation begins.

⁷⁸ On the influence of Classical literature on the personal names in this district see Ramsay, in J.H.S. 1883, p. 36Google Scholar.

⁷⁹ Distinction between I. (b) and III. is hard and often impossible.