^{page 76 note 1} The map (Fig. 1) is based on that in B.M. Coin Catalogue, Lycia, etc.

^{page 76 note 2} It is our pleasant task to offer our grateful acknowledgments to the following gentlemen who rendered us assistance in various ways: Mr. G. A. Keun, H. B. M. Vice-Consul at Adalia; H. E. The Mutessarif of Adalia; Their Excellencies the Governors of Istánoz, Elmalí, and Tefénny. To many unknown friends for spontaneous and cordial hospitality and help, and particularly to Mr. Nikolas Michael Ferteklis of Adalia, an indefatigable epigraphist and numismatist, who accompanied us on the first portion of our journey, and whose local knowledge was most helpful to us.

^{page 77 note 1} The one, published in this volume (p. 185) by Mr. F. W. Hasluck, who called my attention to the existence of the stone, records the capture of Adalia on August 24th, 1361, by ‘Pierre…Roy…d'lerusalem e de Chipre.’ The other is known already (C.I.L. iii. Suppl. 6737), but see note below, p. 127. The fullest list of the inscriptions at Adalia is given in Lanckorónski, , Städte Pamphylicns und Pisidiens, i. pp. 153 ff.Google Scholar [quoted henceforward as Lanek.].

^{page 78 note 1} Travels in Lycia, etc. i. p. 228 [quoted henceforward as S.F.]; Lanck. i. p. 19, and other authorities quoted there. Cf. E. J. Davis, Anatolica, pp. 220 ff., who went over the first part of our route, but did not collect any archaeological material of value. He alludes to the site here mentioned ‘near Uzumkoyou’ Kahve, and calls it Ariassus (as = Aarassos in Strabo, xii. 7), though he adduces no evidence. See the notes that follow.

^{page 78 note 2} Evdir is the spelling adopted by H. Kiepert, Specialkarte vom Westlichen Kleinasien, (1890–91; scale 1 =250,000). In Lanck. it is spelt Ewdir; it is called both Eski Khan and Evdeer Khan in S.F. i. pp. 226, 230. The interior is figured in Lanck. ii. p. 123, Fig. 97.

^{page 78 note 3} Monatsber. der…Akademie…zu Berlin, 1874, pp. 716–718.

^{page 78 note 4} Below, No. 2. No. I had been copied already, together with some others, by Rott, see below, ad loc. C.I.G. 4341 c, d, e were copied here by Schönborn.

^{page 79 note 1} Cf. Livy, xxxviii, c. 16. Lagon, as Hirschfeld points out (loc. cit.), is not in Pamphylia, whereas this site certainly is.

^{page 79 note 2} For the true site of Lagbe see below, p. 88 and note.

^{page 79 note 3} Ath. Mitt. x. (1885), p. 343.

^{page 79 note 4} Lanck. ii. pp. 78, 223, 224; Heberdey-Kalinka, , Bericht über zwei Reisen im s.w. Kleinasien, [Denkschriften der kais. Akademie der Wissenschaften zu Wien, xlv. (1897)], pp. 36, 37.Google Scholar

^{page 79 note 5} The phrase ποταμὸς Λαγόνων in the inscription (Kaibel, Epigrammata Graeca, 808) found here by Spratt and Forbes is not a clue to the name of the site. In Lanck. i. p. 19, it is suggested that the town was a deme of Olbia or Attaleia, e.g. Uliambos or Kanavra, but Ramsay (Ath. Mitt. loc. cit.) thinks the former name merely a mistake for δήμου Ὀλβιανοῦ.

^{page 79 note 6} Kleinasiatische Denkmäler aus Pamphylien, Pisidien, etc. (1908), p. 30. He has no real evidence in support of his contention that Jovia was the earlier name of Eudokia; the latter name he supposes was substituted in the time of Eudokia (wife of Theodosius II, 421–431 A.D.). That it survives in Ew(v)de, the name of the Khan, does not seem convincing; we understood the name to be Evdir, as did Spratt and Forbes.

^{page 79 note 7} By the direct road now practicable for wheeled traffic from Adalia to Yenije-Boghaz-Kahve and Istánoz: this follows at first the route taken by Spratt and Forbes, S.F. i. pp. 225 ff., but, after passing Y.-B.-Kahve, zig-zags up the left side of the pass. The old bridle-track, which is steep and stony, lies in the actual defile (the Yenije-Boghaz), and is still used as a short cut.

^{page 79 note 8} S.F. i. p. 231; Hirschfeld, loc. cit. Dimensions of that to the S., 6·25 × 6·25 (only about three courses standing); of that to the N., 1075 (N.S. walls) × 8·40 (E.W. walls). The latter has a door ·96 wide with square hole for bolt, near the middle of the S. wall. At one corner as many as seventeen courses of masonry are still standing. A view of that to the south is given in Fig. 2.

^{page 80 note 1} S.F. i. pp. 231, 240; Hirschfeld, loc. cit. Among the sketches made by the Rev. E. T. Danieli, who accompanied Spratt and Forbes on their journey in 1842, and now preserved in the British Museum, is one (unfinished) showing the valley and the wall running across, made from a point high up on the northern side of the pass, inside the line of the wall. (Cf. Catalogue of Drawings by British Artists…preserved in the Dept. of Prints and Drawings in the British Museum, vol. ii., p. 15, No. 41.) No. 40 is a view of the site of Termessos, from the hills above the city to the W., and No. 44 is an unfinished view of Elmali. Mr. F. W. Hasluck kindly called my attention to these delightful sketches, the majority of which represent scenes and sites in Lycia. No other drawings in this series illustrate the route described in these pages.

^{page 80 note 2} First identified by Spratt and Forbes, S.F. i. pp. 232 ff. A view of the theatre is given here, Fig. 4.

^{page 80 note 3} Lanck. ii. pp. 21–122, and Plates I.–XXI.

^{page 81 note 1} Lanek. ii. pp. 196–223; Cousin, B.C.H, xxiii. (1899), pp. 165 ff., 280 ff.

^{page 81 note 2} Heberdey-Wilberg, , Jahresh. iii. (1900), pp. 177 ff.Google Scholar

^{page 81 note 3} Lanek. ii. plan facing p. 21, L 5.

^{page 82 note 1} Smith, A. H., B.M. Sculpture, ii. pp. 237 ff, No. 1418Google Scholar; Nos. 1419, 1420, 1421 seem lo be small fragments of similar groups; see also Bernoulli, Aphrodite, pp. 245, 263, 300; Reinach, , Réf. de Sculpt, ii. p. 357Google Scholar, No. 9.

^{page 83 note 1} At a Tchiftlik close to the foot of the hills on the western edge or the valley (ten minutes from the Kahve) is a square tower of regular masonry with drafted angles, standing to the height of two or three courses only, and measuring about ten metres each way. On either side are traces of two parallel lines of wall running N.-S. at the base of the hill.

^{page 83 note 2} A few minutes before one reaches the Kahve one sees on the right, a short distance above the road, another square tower, standing to a height of about seven courses, and similar in construction to the others. It has an opening on the side towards the road. This seems to be the building noted by Cousin, B.C.H. xxiii. (1900), p. 332, who calls it a tomb. Mr. Ormerod writes (March, 1911) after a second visit, that it is one of a walled line of towers, like that at the E. end of the pass, of which he counted three. We did not see the inscription built in above the entrance, of which he gives an incomplete version (ibid.). I hope to obtain a satisfactory copy of it before long. We saw no water in the small lake marked in Kiepert's larger map immediately to the west of ‘Gioklüdje’ (sic) Kahve.

^{page 83 note 3} Called there Isionda? We could neither see nor hear of any ruins here. For the correct site of Isi(o)nda see below, § 3 ad init. 84.

^{page 83 note 4} Called by Kiepert, ibid. Eyub. The mound, which we examined subsequently and found to be a prehistoric site, can hardly be all artificial. Cf. p. 90, No. 5.

^{page 84 note 1} Ath. Mitt. x. (1885), p. 340.

^{page 84 note 2} Called in S.F. i. pp. 242 ff. Stenez. The site at Kyshlár is by them identified as Mandropolis.

^{page 84 note 3} Below, No. 4. This was perhaps copied by Spratt and Forbes, but has not been published; they mention the building into which it is built as ‘a large well-built Turkish edifice, used as a Khan, and of the same age and style as that at Lagon’ (Evdir Khan), i. p. 246.

^{page 84 note 4} See B.M. Cat. Lycia, Pamphylia, and Pisidia, Introd. pp. xciii, ff.; pp. 225, 226, Nos. 16, 17, and 18.

^{page 84 note 5} S.F. i. p. 246.

^{page 84 note 6} Nos. 3, 5, 6.

^{page 84 note 7} Andeda, at Andya, , Ath. Mitt. x. (1885), p. 337Google Scholar; Pogla at Fugla (Foula, Foglila), ibid. p. 335 (first suggested by Kiepert); Verbe at Zivint, ibid. pp. 338, 339; Komama, at Sherif, . (Shehir., ace. to Ramsay) euyuk, A.J.A. iv. (1888), p. 263.Google Scholar

^{page 85 note 1} Ramsay, , C.-B. i. p. 327Google Scholar, points out that the site of Kodroula is uncertain, but may have lain to the S.W. of Lake Kestel.

^{page 85 note 2} Cf. B.M. Cat. Lycia, etc., pp. 174, 175.

^{page 85 note 3} To left: draped female figure turned half-left: head and r. arm missing; 1. arm supported across breast in a fold of the veil which presumably covered the head as well; 1. knee bent, weight being entirely on r. leg. To right: draped male figure turned three-quarter-right: left side broken away, and head missing; long chiton reaching to the feet, over which is thrown a mantle which passes round the waist in heavy horizontal folds: r. hand (missing) rested on hip. For sarcophagi of this type see Strzygowski's, article, J. H. S. xxvii. (1907), pp. 99 ff.Google Scholar and authorities there quoted.

^{page 85 note 4} See below, p. 92, No. 12.

^{page 85 note 5} The peribolos-wall measures about 30 yards each way. The altar itself is about 30 feet long and 18 broad, the greater length being N.-S. It stands about 4 feet 6 inches high.

^{page 85 note 6} Perhaps that seen by Spratt and Forbes. S.F. i. pp. 245, 246.

^{page 86 note 1} Reisen in Lykien, etc., ii. p. 165, and P). XXIII.

^{page 87 note 1} A good view of Elmali is given by Benndorf-Petersen, op. cit. ii. Pl. XXI.

^{page 87 note 2} No. 15 below. This is presumably the mound noticed by Spratt and Forbes: ‘a small flat hillock or mound, on which were the foundations of ancient buildings, many loose blocks, and much broken pottery.’ There are traces of an apsidal building, of which not more than two courses are preserved above ground, and of a small cistern, both probably of Byzantine date. Besides the prehistoric sherds were many of Roman, and of more recent times, and some fragments of Byzantine carving.

^{page 87 note 3} Op. cit. ii. p. 161; cf. S.F. i. p. 290.

^{page 87 note 4} Its height above the sea is about 4250 feet: that the situation is healthy we can endorse from personal experience, as we passed a night there in a shepherd's hut during a heavy rain-storm, without evil effects.

^{page 87 note 5} Op. cit. pp. 7 ff.; some had been copied previously by Benndorf, , Reisen in Lykien, ii. pp. 166–167Google Scholar, Nos. 193–203.

^{page 88 note 1} Op. cit. ii. p. 166, No. 195.

^{page 88 note 2} Smith, A. H., J.H.S. viii. (1887), p. 253Google Scholar, No. 34 (‘Abu Faradin Yaila’); the ancient name is there given as Lagbon: cf. Ramsay, , A.J.A. iv. (1888), p. 16Google Scholar, who follows Kiepert and Waddington in identifying this site with Lagon, which is mentioned by Livy in his account of the march of Manlius. ‘Polybius may have used the form Λαγόη or Λαγούη to represent the Pisidian name; and the text of Livy ought to be Lagoen. It is also possible that Λάγβος or Λάγβον is the proper form, and that the text of Livy should read Lagbon.’ (Ramsay)

^{page 88 note 3} Height of the summit of the pass, 5400 feet.

^{page 88 note 4} See below, p. 93, Nos. 17 and 18.

^{page 88 note 5} The inscriptions from Tefénny and district, published since the original C.I.G. (Nos. 4366 w, x, and 4367), are scattered over many publications; cf. B.C.H. ii. (1878), pp. S3 ff., 170 ff., 243 ff.; Sterrett, Efig. Journey, pp. 38 ff.; B.C.H. xvi. (1892), pp. 418, 419; Ramsay, , C.-B. i. pp. 304 ff.Google Scholar ‘Inscriptions of the Killanian plain.’

^{page 88 note 6} Adji Badem = ‘bitter almond tree,’ Ramsay, , A.J.A. iii. p. 362, Note 45.Google Scholar There seems some uncertainty as to the correct spelling: it appears sometimes as Hadji Parim.

^{page 89 note 1} Ramsay, loc. cit.; Murray's Guide Book to Asia Minor, p. 120.

^{page 89 note 1} I have to thank Mr. H. R. Hall of the British Museum for various suggestions in the writing of this paper, also the Rev. G. B. Cronshaw of Queen's College, Oxford, who very kindly undertook the chemical analysis of some of the pottery. The Keeper of the Ashmolean Museum allowed me to have the photographs illustrating this part of the paper (PI. VII) taken at the Museum. Above all, I would acknowledge my indebtedness to Professor J. L. Myres for many points suggested by his articles in the Anthropological Journal, and for much advice given in the writing of the paper. I trust that I have stated his views correctly, more particularly those concerning the black- and red-faced technique in the pottery.

The sketch-map of the Istánoz Plain was drawn for me by Mr. H. H. Jewell. (Fig. 7.)

^{page 89 note 2} J.A.I, xxx. p. 253.

^{page 89 note 3} Annals of Archaeology and Anthropology, ii. p. 146, and Pls. XXVI, and XXVII.

^{page 90 note 1} Kiepert and Lanckorónski (Städte Pamphyliens und Pisidiens) place the village too far west.

^{page 92 note 1} See p. 85.

^{page 93 note 1} Spratt, and Forbes, , Travels in Lycia, i. p. 290.Google Scholar

^{page 93 note 2} Op. cit. i. p. 288; see above, p. 87, note 2.

^{page 94 note 1} A celt seen by us in the bazaar at Elmalí was said to have been found in a modern housewall, the earth for the bricks having probably been brought from a similar mound. The owner would not part with the celt, which was the only one we saw.

^{page 95 note 1} Ath. Mitt. xxiv. p. 13.

^{page 95 note 2} Ibid. p. 19.

^{page 95 note 3} It is noticeable that these stone-heaps are not a feature of the ploughed lands in the plain itself.

^{page 96 note 1} Ilios, p. 219; Troja, p. 33.

^{page 96 note 2} We may compare the experiments of M. P. Marguerite de la Charlonie on fragments of Hellenic pottery (Revue des Études Grecques, xx. p. 232, ci. Myres, , Year's Work in Class. Studies, 1908, p. 42Google Scholar). When the fragments were exposed to a wood fire it was found that the blackening caused by the smoke, which at first affected the surface only, spread gradually through the clay. On the temperature being raised the blackness again disappeared, as in Doulton's experiment (Troja, p. 33). As will be seen below, when a glaze is not used, a black-faced or red-faced result probably depends on the regulation of temperature in firing, a high temperature burning all carbonaceous matter out of the clay. M. de la Charlonie's experiment shows that a blackness can be again produced by smoking the pot, after it has been fired, over a slow fire. This process, however, as Myres points out (op. cit. p. 43), needs very careful regulation; the colour moreover tends to be grey rather than black.

^{page 96 note 3} Naqada and Ballas, pp. 36–7; Diospolis Parva, p. 13.

^{page 97 note 1} As he says, the remarkable mirror-like brilliancy of the black part of the surface would be assisted by the presence of small amounts of carbonyl resulting from imperfect combustion. Körte, op. cit. p. 24, states that some of these vases in the Bonn Museum are touched with graphite.

^{page 97 note 2} J.A.I, xxxiii, pp. 369 ff.

^{page 97 note 3} Man, 1901, p. 99, where the locality is wrongly given as Torres Strait; cf. J.A.I, loc. cit. An interesting case is recorded among the Malays (Wray, L., The Malayan Pottery of Perak, J.A.I. xxxiii, p. 28Google Scholar). The clay of these pots, which after firing is of a light red colour, is blackened by burying the pot, while still hot from the kiln, in a mass of ‘padi husk.’ We clearly have here a method of recalling the black colour after firing, alternative to the smoking process described above (p. 96, note 2).

^{page 97 note 4} See above, p. 96.

^{page 98 note 1} We ought perhaps to include a fourth variety, which appeared in small quantities but is of later date—a grey polished monochrome ware, parallel in form and fabric to the sub-Mycenaean wares of Hissarlik vii.

^{page 98 note 2} i. p. 246.

^{page 99 note 1} The object illustrated, Pl. VII, No 19, which resembles a small unbored hammer-head, was obtained by us in the bazaar at Elmalí. It is of pure haematite, and if used as a burnisher on the damp surface of the pot would give the effect described above.

^{page 99 note 2} It is probable that the mottled black and red surface of some of the pots from Boz-Euyuk (Körte, op. cit. p. 24) was intentionally caused by withdrawing them from the kiln or muffle before they had completely cooled. For the mottled ware from Vasiliki in Crete and other parallels, see Seager (ap. Boyd-Hawes, Gournia, p. 50.).

^{page 99 note 3} Unfortunately there was no evidence in the cutting to show to which stratum the object belonged.

^{page 99 note 4} J.A.I, xxx. p. 255.

^{page 99 note 5} J.A.I, xxvii, p. 171.

^{page 99 note 6} Cf. Schmidt, in Troja und Ilion, i. p. 246.Google Scholar

^{page 100 note 1} Myres, Cat. Cypr. Museum, p. 15.

^{page 100 note 2} Ibid. p. 14.

^{page 100 note 3} With the exception of Nos. 7 from Karalitis, and 15 from Hassan Pasha.

^{page 100 note 4} Cf. Crowfoot, , J.H.S. xix. pp. 34 ff.Google Scholar § 2; Anderson, , Studia Pontica, i. p. 71.Google Scholar

^{page 100 note 5} Cf. Welch, ap. Myres, , J.A.I. xxxiii, p. 379Google Scholar, who describes this ware as of Geometric Cypriote type. It appears also in Hissariik viii. (Troja und Ilion, i. p. 309).

^{page 101 note 1} Chantre, Mission en Cappadoce, Pl. XII. 5.

^{page 101 note 2} E.g. on the Burgon lebes.

^{page 101 note 3} E.g. Pl. VII, Nos. 2, 6, 7, and on other pieces not illustrated. It is interesting to note its appearance at Moussian (Gautier et Lampre, Délégation en Perse, vii. Fig. 165), where it does not appear to have the maeander as a prototype.

^{page 101 note 4} Op. cit. p. 390.

^{page 101 note 5} Professor Myres gives me the following note: ‘In relief, for hair of terracottas, seventh century, Cyprus; in paint, Cyprus, eighth and seventh century, especially at Tamassos.’ The design occurs also on a sherd from Gordion (Körte, Cordion, p. 178, Fig. 160, No. 2). Körte compares a ‘Cypriote’ sherd from Hissarlik (Schliemann-Sanmlung, 3482), and an amphora from Caere (Pottier, Vases antiques du Louvre, Pl. XXIX. D. 18).

^{page 101 note 6} B.M. Excavations at Ephesus, p. 219, Fig. 45, No. 10 = p. 223, Fig. 49 (? Local Ephesus fabric).

^{page 102 note 1} Op. cit. p. 385, with reference to the similar Cappadocian bowls.

^{page 102 note 2} Perrot-Chipiez, v. p. 328, Fig. 232 = Winter, , Ath. Mitt. xii. p. 226Google Scholar, Fig. 3.

^{page 102 note 3} Furtwängler-Löschcke, , Myk. Thongef. iv. 3, 17Google Scholar; Myk. Vasen, xxiii. 168.

^{page 102 note 4} Cypr. Mus. Cat. p. 48, No. 328, on the ‘white ware’ (Fabric, ii. 1, p. 38).

^{page 102 note 5} Myres, op. cit. Fig. 10, p. 385.

^{page 102 note 6} Cf. Délégation en Perse, vii. p. 142. The early date assigned by the excavators is not universally accepted (see Hall, in Proceedings of Society of Bibl. Archaeology, xxxi. p. 311Google Scholar). The comparison cannot be pressed in view of the distance. We have yet to learn the full extent and importance of the wide Geometric areas recently revealed in Western Asia, at Anau in Turkestan, and Northern Persia. Their influence seems to have been felt by the Geometric Cappadocian fabrics and even in Cyprus and on the Syrian coast (cf. Hall, op. cit. who developes certain suggestions made by Professor Myres, J.A.I, xxxiii.). Mrs. Boyd-Hawes (Gournia, p. 33) claims a connection between Anau and early Minoan Crete, based on a trade in tin with Khorassan. The spreading across Asia Minor of decorative influences foreign to the Aegean, might perhaps be due to the trade in obsidian. The fragments of obsidian which we found at Rames have been examined by Professor Myres, who pronounces them definitely to be not of Melian origin. The source of the obsidian found on the Black Sea litoral and in Northern Assyria, as is suggested by Professor Bosanquet (Phylakopi, p. 229), is probably the Caucasus and Russian Armenia, as is also the case with that found in Culture iii. at Anau in Turkestan (cf. Schmidt, in Pumpelly, , Prehistoric Civilisations of Anau, South Kurgan, i. p. 151Google Scholar). Further evidence however is needed from these and other sites.

^{page 102 note 7} Bliss and Macalister, Excavations in Palestine, 41, No. 149, belonging to the ‘later pre-Israelite period.’

^{page 103 note 1} The ground-plan of a watch-tower at Sigri in Lesbos as published by Koldewey (Die antikon Bauwerke der Insel Lesbos, Pl. XXVII, 1, and cf. p. 62) shows a similar arrangement of interior walls.

^{page 104 note 1} Annals, ii. p. 146, Pl. XXVII.

^{page 105 note 1} About two hours from Kosagatch, on Lake Karalitis.

^{page 105 note 2} See above, p. 78, note 2.

^{page 106 note 1} A previous traveller, E. J. Davis, who passed by here in 1872, alludes to this cistern, but failed to notice the inscription built in, Anatolica, p. 222.

^{page 106 note 2} Weber, who publishes the inscription from an inadequate copy by Rott, treats it as incomplete on the right, and lacking only one letter on the left, except in 1. 4: this is certainly incorrect, and leads him to a different restoration in 111. 3 and 4: 10 οἱ ἀπελεύ[θεροι ∣ ἀρε]τ[ῆς καὶ εὐ]ε[ρ]γεσίης ἔνεκεν.

^{page 107 note 1} E.g. at Termessos, , B.C.H. xxiii. (1899), pp. 166 ff.Google Scholar, Nos. 12, 26, 35, 41, 42. It also occurs in Nos. 3, 10, 18, and 22 of the inscriptions contained in this article. For all these names cf. the index to Lanckorónski's work.

^{page 107 note 2} E.g. at Termessos, C.I.G. 4366 n; at Pogla, , J.H.S. viii. p. 256Google Scholar, No. 41; in the Kibyratis, Heberdey-Kalinka, Bericht über zwei Reisen in südwestlichen Kleinasien, p. 54, No. 78, and in No. 20 below.

^{page 107 note 3} The right-hand portion is made up of four fragments which fitted together exactly, and, as the reproduction shows, very little is lost between them and the left-hand portion. We were unable to squeeze or photograph the latter, as a stout tree was growing directly in front of it, which made copying a laborious business. But the copy is, as far as it goes, complete, and substantially correct.

^{page 107 note 4} I wish to acknowledge the kind help of Dr. L. R. Farnell, of Exeter College, Oxford, and of Mr. A. S. F. Gow, of Trinity College, Cambridge, who have greatly assisted me towards the restoration and interpretation of this inscription; and particularly of Mr. A. B. Cook, Reader in Archaeology at Cambridge, who has given me, at the last moment, many valuable suggestions and references. Nor must I omit to mention Miss J. E. Harrison's kind interest in the inscription.

^{page 108 note 1} In 1. 1 πῶ for πῶς, and in 1. 5 βροτον (if read aright) for βροτῶν, are due to careless engraving; in 1. 4 τοῦτο for τούτῳ in 1. 5 οὗτος for οὕτως, and ἔστο for ἔστω to indifference to the spelling of the unaccented syllables. In 1. 2 ἐνθάδ; in 1. 3 τὸ σκῆπτρōν Ἑρμοῦ ; in 1. 5 καλαῦρŏψ βπŏτ(ῶ)ν ; n 1. 6 βί̆ῦς κάμπτει ; and in 1. 9 Μορσιανὸς Ἑρμα[ιόν ?] (unless we scan the first name as a dactyl) are due to the other causes mentioned.

^{page 108 note 2} Epigrammata Graeca, 186, φίλε πυνθάνεαι ; 270, Εἴ με θέλαις άνιχνει̑ν; 408, Εί μου θέλις

^{page 109 note 1} Mr. Cook tells me of the employment of a somewhat similar symbolism on a mosaic table-top from Pompeii, which shows a butterfly resting on the Wheel of Fortune with a sceptre to the left, and a beggar's staff to the right: cf. Niccolini, , Le case ed i monumenti di Pombei, ii. Pl. XLVIII.Google Scholar; Mau, Pompeii ¹, p. 399.

^{page 109 note 2} Il. xxiii. 845; Anth. Pal. vi. 106; Anth. Plan. 74. Eustathius, 1332, 45, defines it as ῥόπαλον ἐπικαμπὲς ἑν ἄκπῳ. Arvanitopoullos, Ἐφ. Ἀρχ. 1907, p. 121, restores καλ]αύρύροπ(α) in a Tegean inscription copied by Ross, I. G. ined. No. 3.

^{page 109 note 3} By Professor J. L. Myres; cf. E. Norman Gardiner, Greek Athletic Sports and Festivals, pp. 24, 313, 314.

^{page 109 note 4} A wooden καλαυ̑ροψ has the further advantage that one could not only throw it ἑλισσομένη διὰ βου̑ς ἀγελαίας to round-up stragglers, but use it simply for driving the herd, which not even the most skilful of shepherds, in Homeric times or modern, could do with success with a stone on the end of a string. The wolf-hunter, , Anth. Pal. vi. 106Google Scholar, 11. 3, 4, dedicates to Pan ῾καὶ τὰν ἐκ κοτίνοιο καλαύροπα, τάν ποκα τη̑νος πολλάκι ῥομβητὰν ἐκ χερὸς ἠλροβόλει.᾿

^{page 109 note 5} Ἐντυπάζω does not, I believe, occur elsewhere; but cf. the use of τυπάζω by Oppian in this sense, C. i. 458. The uses are often confused; cf. L. and S., s.v.

^{page 110 note 1} As Mr. Cook suggests to me.

^{page 110 note 2} I owe this restoration to Dr. Farnell.

^{page 110 note 3} Kaibel, Epigr. Gr. 1023, 1. 5 (restored); B.C.H. xxv. (1901), p. 276, No. 17, on a boundary-stone from Tegea, inscribed Πανὸς Λυκείον προκαθηγετου̑ (so the editor accents the word); date, 2nd century A.D. I am indebted to Mr. Tod for this reference, and for that to the καλαυ̑ ροψ in a Tegean inscription. For such titles of deities as προηγεμών,προηγέτης ,προκαθηγέτις, etc., see the instances collected by Gruppe, , Gr. Mythologie und Religionsgeschichte, (in Iwan von Müller's Handbuch), ii. p. 1295Google Scholar, Note 1. Add to them the phrase in an Alexandrian hexameter-fragment published by ProfessorGoodspeed, , J.H.S. xxiii. (1903), p. 240Google Scholar, A. 1. 17, [ἀ]θανάτων προκαθηγ[ητήρ.]

^{page 110 note 4} As Mr. Gow has convinced me.

^{page 110 note 5} Suggested by Mr. Cook: cf. Eur. Hel. 1666 (ed. G. Murray, 1909), ὅταν δὲ κάμψῃς καὶ τελευτήσῃς βίον.

^{page 111 note 1} If we may credit the poet with a reminiscence of the classical contrast of παίζειν and σπουδή, e.g. Plato, Legg. 887 D, μετά τε παιδια̑ς καὶ μετὰ σπουδη̑ς.

^{page 111 note 2} Cf. Stephanus, Thesaurus, s.v.: this I owe to Mr. Cook.

^{page 111 note 3} The name Μορσιανύς does not seem to occur elsewhere.

^{page 111 note 4} Cf. Kern, O., in Pauly-Wissowa, iii. pp. 1013–1017.Google Scholar I am indebted to Mr. Cook for this reference and interpretation. May we see in this use of the καλαυ̑ροψ. in Asia Minor at least, an indirect survival from Hittite practices? Cf., for instance, the καλαυ̑ροψ.-like staves carried in the procession in the Feraktin relief: Garstang, Land of the Hittites, p. 150, Pl. XLVII.

^{page 112 note 1} Cf. C.I.G. add. 4300 m: 4321 b, c; 4341 e.

^{page 112 note 2} Cf. C. I. G. 4366 k, n (Αὐρ.Τροκόνδας) ; 4367 g. There is an epitaph of a Τροκόνδας ot Termessos, in Athens, , I.G. iii. 2, 2937.Google Scholar

^{page 112 note 3} J.H.S. xxii. p. 124, No. 58.

^{page 112 note 4} E.g. at Tefénny, C.I.G. 4366 w; the genitive Ἀττέους at Pogla (Foghla), , A.J.A. 1888, p. 10Google Scholar, may come from either form.

^{page 112 note 5} Cf. C.I. G. 4226, 4248; Ἀρτίμας is still more common: cf. C.I. G. add. 4315 i, 4321 d, etc.

^{page 113 note 6} I.e. the name of which the genitive is contained in 1. 12.

^{page 112 note 7} Or rather ὑὸν, if only four letters are lost.

^{page 113 note 1} For allusions to θρεπτοί in S.W. Asia Minor, cf. Heberdey-Kalinka, op. cit. p. 26, No. 22 (a mausoleum at Patara is built by two people ἑαυτοι̑ ς καὶ τοι̑ς τεκνοθρέπτοις ἀπελευθέροις) p. 27, No. 26, ll. 13–15 (also at Patara, ἰς δὲ τὰ ὑποσ[ό]ρια κατατεθη̑ναι τοὺς θρεπτοὺς καὶ θρεπτὰς αὐτον̑).) These privileges toθρεπτοί are seldom found inscribed on sarcophagi, so we may assume that they were comparatively rarely given. For θρεπτοί in other parts of Asia Minor, in addition to the instance quoted above at Cyzicus, cf. Ramsay, , C.-B. i. pp. 147, 350Google Scholar; ii. p. 546; and, above all, Plin. et Trai. Epist. lxv. In Athens, Anderson, J. G. C., S.S.A. iii. (1896–1897), p. 120.Google Scholar I am indebted to Mr. Anderson for these and a few other references, and for his help and advice on several points.

^{page 113 note 2} See below, No. 22, 1. 5.

^{page 113 note 3} A similar confusion may be quoted on a grave-stone near Lake Karalitis, κακας χι̑ρες, Heberdey-Kalinka, op. cit. p. 7, No. 20.

^{page 114 note 1} Dittenb. Syll. ² 676.

^{page 114 note 2} I.G. ad res Rom. pert. iii. 342, 1. 1.

^{page 114 note 3} Cf. Ramsay, , Ath. Mitt. x. (1885), p. 340Google Scholar, and above, p 84.

^{page 114 note 4} B.M. Cat. Lycia, etc., Inlrod. pp. xciii, ff.; pp. 225, 226.

^{page 114 note 5} In the upper surface three holes are cut, as though to receive the feet of a tripod.

^{page 115 note 1} I owe this suggestion to Mr. D. G. Hogarth. Cf. Heberdey-Wilhelm, Reisen in Kilikien, p. 77, No. 155, viii. 1. 6, Ἀπολλω̑τος (gen.); 1. 88, Μω̑ς; 1. 94, Πκω̑ς; 1. 108, Κλον̑ς 1. 151, Λον̑ς 1. 152, Κνω̑ς.

^{page 115 note 2} Cf. C.I.G. 4402, 4410: the genitive was apparently Κούου.

^{page 115 note 3} Cf. Heberdey-Kalinka, op. cit. pp. 137 ff., Nos. 225, 228, 229, 230, 232.

^{page 115 note 4} Cf. Heberdey-Kalinka, op. cit. p. 55, No. 80, for the name Ὀρνεπει̑μις at Kadyanda.

^{page 115 note 5} B.C.H, xxiii. (1899), pp. 172, No. 17; 173, No. 18; 180, No. 34; 184, No. 44; 189, No. 54.

^{page 115 note 6} The confusion of ou and av can be paralleled by Οὐολλανσίας ( = Volusiae), I.G. ad res Rom. bert. iii. 829, as Mr. Tod informs me.

^{page 116 note 1} J.H.S. xxiv. (1904), p. 32, No. 43; xxv. (1905), p. 61, No. 24.

^{page 118 note 1} See Pape-Benseler, Wörterbuch der gr. Eigennamen, s.vv.

^{page 118 note 2} A.J.A. iv. (1888), p. 12, No. 9, 1. 7. Fugla is only about two hours distant from the village of Yerten-keui.

^{page 118 note 3} I am of course in error in saying there that ‘῾συνστέφω is not known elsewhere.’ I regret that this inaccuracy has remained so long uncorrected.

^{page 118 note 4} I.G. iii. 128, 1. 30; B.S.A. xii. p. 466, No. 19; xiv. pp. 106 ff., No. 5 and notes ad loc.

^{page 118 note 5} Bas-Waddington, Le, Voyage Archéologique, iii. (Inscriptions d'Asie Mineure), pp. 297–8.Google Scholar

^{page 120 note 1} Willing hands quickly removed the earth which hid the rest of the inscription, so that we were able to read every letter preserved on the stone.

^{page 121 note 1} As, for instance, C.I.G. add. 4554, and probably C.I.G. 4997, though the editors note that the latter may refer perhaps to Elagabalus.

^{page 121 note 2} Titles and name of his mother Julia Mamaea erased as well.

^{page 121 note 3} His titles have been erased and subsequently re-inserted.

^{page 121 note 4} Ath. Mitt. x. 1885, pp. 334 ff.; A.J.A. 1888, pp. 13 ff.; see also above, p. 84.

^{page 121 note 5} Ath. Mitt. x. 1885, p. 338 (No. 3); see Ramsay's notes ad loc.

^{page 122 note 1} We were told that Nos. 13 and 14 were found together in a neighbouring building, though we saw no ancient masonry.

^{page 122 note 2} If Zivint is an ancient site and is not Verbe, could it be Ευλίνη Κώμη? Cf. Livy xxxviii, c. 15, where Manlius ‘ex Pamphylia rediens ad fluvium Taurum primo die, postero ad Xylinen quem vocant Comen posuit castra. Profectus inde continentibus itineribus ad Cormasa urbem pervenit.’ If he crossed the Tauros (Istánoz Tchai) near Euyuk and halted there, a day's march in a N.N.W. direction would have taken him close to Zivint; but Ramsay, (A.J.A. iv. p. 273Google Scholar) may be right in placing Xyline Come further north, near Pogla: two years before he had placed it further south, near Ali-Fakhreddin, (A.J.A. ii. pp. 130, 131Google Scholar).

^{page 124 note 1} E.g. at Oenoanda, Heberdey-Kalinka, op. cit. p. 49, No. 65. Δόγματι is still commoner in this context.

^{page 124 note 2} A.g. at Pergamon, Dittenb. O.G.I. 484, 1. 32.

^{page 125 note 1} For this name we have a large choice: Διλιτριανός, which is known at Pogla, is rather too long for the space. The name of his father may have been Ἀρτέμων or Ἀρτείμας, which are common in this district.

^{page 125 note 2} The letters μουat the beginning of 1. 10 seem to do double duty, as the end of both Ὀνησίμου and ἀγορανόμου.

^{page 125 note 3} Ath. Mitt. x. (1885), p. 337. On coins, B.M.C. Lycia, etc., pp. 174, 175.

^{page 125 note 4} E.g. Μόλλεσις at Bubon, C.I.G. add. 4380 k, 1. 4; Μόλεσις at Oenoanda, Heberdey-Kalinka, op. cit. p. 40, No. 57; p. 50, No. 66; p. 54, No. 78. Μόκκισις at Charmylessos (Levisi) in Lycia, C.I.G. 4224 f. Compare also Μόλης, Heberdey-Kalinka, op. cit. p. 47, No. 62; B.C.H, xxiii. (1899), p. 171, No. 11, and Μολλιανός at Termessos, C.I.G. 904.

^{page 127 note 1} Δαμα̑ς is a common name, cf. C.I.G. 2562 (Crete); ibid. 2869 (Didyma); ibid. 2880 (Branchidae, ) J.H.S. viii. (1887) p. 257Google Scholar, No. 42 (Karibtche in Pisidia). But Λαμα̑ς does not appear to exist.

^{page 127 note 2} For these references I am indebted to Mr. J. G. C. Anderson.

^{page 127 note 3} Minute differences merely in the legibility of a few letters are not worth citing here. Though our copy had to be finished by candle-light it is slightly fuller than that originally published by Ramsay, (B.C.H. vii. (1883), pp. 258–9)Google Scholar; in fact I could see (completely or in part) every letter except the third and fourth letters of the word Arruntius, where a piece has gone from the surface.

^{page 129 note 1} B.C.H. 1892, p. 419, No. 43.

^{page 130 note 1} I describe as unpublished those coins which I have been unable to find in the following works: British Museum Coins, Asia Minor (by provinces) [ = B.M.C]; Imhoof-Blumer, Monnaies Grecques; id. Griechische Münzen; id. Kleinasiatische Münzen; id.Zur gr. und röm. Münzkunde, 1908; Babelon, Inventaire de la Collection Waddington [ = Coll. Wadd. InvĄ; Macdonald, Hunterian Collection; Waddington, Babelon et Reinach, Recueil général des Alonnaies grecques d'Asie mineure [=Wadd. Recueil] Where a coin is represented in B.M.C, no further references are given. Mr. G. F. Hill has kindly helped me with identifications and references.

^{page 130 note 2} Forgeries, some of good workmanship, abound. I saw, in 1909, several specimens of the Themistokles-stater of Magnesia, and of the silver alliance coins of Kremna and Keraïlae (Head, Hist. Num. ² p. 707), which could not inspire confidence. A signed Euainetos-dekadrachm of Syracuse was a merely lamentable production, being thin and having a saucer-like flan. Even Greek copper coins are forged to a small extent, notably the ‘large brass’ of Gordian III struck at Patara, with the Apollo Patareus on the reverse: of this forgery I have seen at least three separate specimens.

^{page 131 note 1} Perhaps intended for Aeneas, Anchises, and Ascanius, as on the coins of Dardanus; cf. Head, Hist. Num. ² p. 544.

^{page 132 note 1} Cf. Borrell, , in Num. Chron. x, p. 95Google Scholar for reverse type, arid Pinder-Friedländer, , Beiträge, i, p. 81.Google Scholar

^{page 133 note 1} The specimen in the British Museum is from the same dies. This seems a rare coin.

^{page 133 note 2} Formerly attributed by Svoronos, loc. cit., and others to Crete, this series is now shown by Imhoof-Blumer to belong to Lycia, , Zur gr. und töm. Münzkunde, pp. 21 ff. [1908].Google Scholar

^{page 134 note 1} The only coin, which I know, of Attaleia, struck under Faustina.

^{page 134 note 2} Legend on R. is ΠΕP[ΓΑΙ]ΩΝ ; on B.M. specimen ΠΕP[Γ]Η.

^{page 134 note 3} Now in B.M.

^{page 134 note 4} There are several points of difference: the legend on the ohv. is different, and that on R. is differently divided, on the B. M. specimen and the laurel-branch there is underneath Apollo's left arm. The R. type more nearly resembles Imhoof-Blumer, , Kleinas. Münz, ii, p. 336Google Scholar, No. 13, but is not otherwise known on coins of Hadrian.

^{page 135 note 1} Apparently from same dies as my specimen.

^{page 135 note 2} For the R. cf. a coin of Timbriada in Pisidia, B.M.C. 2, Pl. XLII, No. 8.

^{page 135 note 3} Now in B.M.

^{page 135 note 4} There is no specimen of this type in the B.M.C, with an earlier date than ΚΕ = 46 B.C., the era of Termessus being 71 B.c., cf. Head, Hist. Num. ² p. 712).

^{page 135 note 5} For another specimen, Coll. Wadd. Inv. No. 4034.

^{page 135 note 6} Cf. Imhoof-Blumer, Mtmn. Gr., p. 347, No. 118. The legend should read IVL AVG COLPARLAIS. I obtained in Athens a coin of this city struck under Commodus, of which the R. legend begins 101. AVGHA CO—: the same form occurs on one published by Imhoof-Blumer, Zur gr. und röm. Münzkunde, p. 200. The letters HA point to a re-founding of the colony by Hadrian. Cf. also Coll. Wadd. Inv. No. 4792, and Head, op. cit ² p. 714.

^{page 135 note 7} The owner would not give precise details of the finding-place.

^{page 136 note 1} I secured these three, among others. The types are as follows: Pupienus, Rev. Concordia Augg. Cohen,² 7; Trajanus Decius, Rev. Genius Exerc. Illyriciani, Cohen,² 53; Etruscilla, Rev. Fecunditas, Cohen,² 12.