^{page 396 note 1} This work will be referred to as CB. The other abbreviations need no explanation.

^{page 396 note 2} Although exploration can hardly ever claim to be final, this will be regarded as an indication in favour of the alternative view which would place Daldis in the Hermos valley near Hierocaesareia, (CB. i. p. 179)Google Scholar. See also Buresch, , Reisebericht in Ber. der Kgl. Sächs. Ges. d. Wiss. (Leipsic), 1894, p. 91Google Scholar: and cp. Imhoof-Blumer, M. in Rev. Suisse, 1897, p. 211Google Scholar f.

^{page 397 note 1} Heights estimated by aneroid (except in the Map, Pl. XII.).

^{page 398 note 1} Prof. G. Radet, in the map attached to En Phrygie (1895), gives up his former identification of Karoura and Kydrara, and now places Karoura here, while still leaving Kydrara at Serai keui, where no remains exist. On the recent growth of Serai keui CB. p. 164, 168.

^{page 398 note 2} Traces of pavement were also seen quite near Gumuldjak.

^{page 399 note 1} Inscr. of Attouda, C.I.G. 3949 ff.; Le Bas-Wadd. 743; B.C.H. 1887, pp. 348 f. and 1890, pp. 238–9; CB. i. p. 181–3.

^{page 400 note 1} At Apameia there is a T̈rbe on the hill above the town, to which the natives go up to pray when they want rain.

^{page 401 note 1} An interpretation of a coin (described p. 166), which will be seen to be justified.

^{page 401 note 2} The rule is stated e.g. p. 168 n 1. M. Radet places Trapezopolis at Kadi keui, but without any proviso. At Kadi keui we did not see the slightest trace of an old settlement. In this district between Tchibuk Dagh and Laodiceia he crowds together a host of towns, several of them on sites where there is no vestige of ancient remains. This is wrong method.

^{page 401 note 3} The name is so given by the Railway Company, but it is always pronounced Sara keui, a name which occurs elsewhere, and seems to be the correct form here.

^{page 401 note 4} This was given me as the right form of the word: I was corrected when I used the form Djebe.

^{page 401 note 5} The suggestion that the name was probably significant is thrown out in CB. i. p. 172.

^{page 401 note 6} The nearest village is Seine keui at the foot of the Dere by the river side. It contains no remains: it would be very difficult to transport heavy blocks down the steep side of the caῆon, but even those blocks which have fallen down have not been carried away.

^{page 403 note 1} Cod. Iust. i. 64, 3: Curator rei publicae qui Graeco vocabulo logista nuncupatur. is the natural rendering of curator (rei publicae), and the adoption of was probably due to the fact that had almost become specialised in the sense of curator operis (as, for example, in 1. 10 of this inscr. )

^{page 403 note 2} He may have been a Trallian, brother of Claudianus Damas (his Latin name being probably M. Ulpius Damas Claudianus) who left a large bequest to Tralleis to found games in the reign of Antoninus Pius: see Pappaconstantinos Nos. 30 and 31. The two brothers in that case were probably sons of a Traillan named Damas. Antoninus Pius sent a native of Aizanoi as curator to Aphrodisias, (C.I.G. 3834Google Scholar, 2741).

^{page 404 note 1} He himself says ‘La carte hydrographique de ce district reste à faire. Toutes les cartes existantes fourmillent d'erreurs.’ (Rev. des Univ. du Midi, 1896, p. 22, n. 2).

^{page 404 note 2} It is sometimes called Bashli Tchai, but Bash Bunar Tchai is clearly right, ‘the stream that flows from the Head Source.’ In the map (Pl. XII.) the stream ought to be represented by a much lighter line.

^{page 404 note 3} Radet, M. justly remarks (Rev. Univ. Midi p. 22Google Scholar) that ‘le site de Dénizly, l'un des plus frais, des plus enchanteurs qui soient dans'la péninsule, n'a certainement jamais (i.e. before 314 B.C.) été inoccupé.’

^{page 405 note 1} Revue des Univ. du Midi, 1896, pp. 20–22; map in En Phrygie.

^{page 405 note 2} Strabo's authority is especially high here, where he was probably an eye-witness.

^{page 405 note 3} Quoted by M. Radet, l.c.: see Bonnet, , Narr. de miraculo a Michaelo archangelo Chonispatrato (Paris, 1890)Google Scholar. It belongs to the eighth or ninth century (Church in R. E. c xix.).

^{page 406 note 1} As Waddington explains it (No. 1693a): une des deux rivières qui traversaient le territoire de Laodieée s' appelait le Caprus, et le nom du village en est dérivé.

^{page 406 note 2} This is reported to me by W.M.R.

^{page 406 note 3} The stream called Kale Tchai is a mere torrent-bed, which never flows except during heavy rains.

^{page 406 note 4} Waddington (on 1693a) identifies Eleinokapria with Badjali, i.e. the village (tchiftlik), on the left bank of the Dere keui stream, in the corner between it and the Lycos. The Kaive stands close to the Railway, a very short distance E. of the Station.

^{page 407 note 1} This was pointed out to me by Prof. Ramsay himself after I had been making a fruitless search for another duden, being unable to accept the identification of Kapros with Bash Bunar Tchai. I had not a copy of Strabo with me.

^{page 407 note 2} Hence it is not quite accurate to say that ‘the disappearance actually takes place at Kara Göl as well as on the Lycus’ (J.H.S. l.c.). I did not hear the water from the duden ‘flowing from the side of the deep gorge and falling down to the bed of the river’; I thought the noise was merely the roar of the river, but the fact may be as stated, for the water must issue somewhere.

^{page 407 note 3} For this section I am indebted to the kindness of Mr. S. Watkins of the Ottoman Railway.

^{page 408 note 1} In CB. No. 5 (= C.I.G. 3949), the fragments of which I copied hurriedly, read with Prof. Ramsay), which exactly fills the erasure: the next has no iota adscript. The inscription is engraved on architrave blocks below the triglyphs and above it there was a Latin inscr. of which one fragment remains, Dedi cant Esex (carved on the metopes) i.e. dedicante Sex[to……. pro consule]. The stones, which have suffered somewhat since they were copied, now lie at the south-east extremity of the ruins beside the Bash Bunar Tchai. It is possible, therefore, that they belonged to the Syrian Gateway, but they cannot have been part of the Ephesian gate, as Prof. Ramsay supposes (from the inaccurate accounts given about them).

^{page 409 note 1} In 1. 6 (end) I read 1. 8 1. 10 1. 11 there appeared to be nothing inscribed below the middle of this last word.

^{page 410 note 1} I had restored and Ath. Mitth. shows an where my copy has a vacant space.

^{page 410 note 2} L. 16 of inscription published Ath. Mitth. 1895, p. 207, should be restored by comparison with Br. M. 421.

^{page 411 note 1} Of the value of such statements, which are often dismissed (e.g. by Von Diest) as stets unzuverlässig, a striking proof will be given under meros.

^{page 412 note 1} Cp, especially No. 29.

^{page 412 note 2} The expression in CB. p. 125, might suggest that a single paraphylax was appointed for the whole Hierapolitan territory: that, however, was not the case: there was probably one for each These are also mentioned in the mutilated decree found at Develer and published by Hogarth, in J.H.S, 1887, p. 392Google Scholar (no. 21).

^{page 413 note 1} Earthquakes still occur in the valley: a rather violent one took place during one of my visits to Laodiceia.

^{page 414 note 1} σαναός Strabo p. 576, σάναος Hierocles, σανίς Ptolem. v. 2, 26, συναός or σιναός Notitiae,

^{page 414 note 2} seems curious: it is possibly a misreading due to the notches in the stone for as in next line, but probably it is used to denote

^{page 415 note 1} Radet, , En Phrygie p. 112Google Scholar; Ramsay, , CB. i. p. 243–4Google Scholar.

^{page 416 note 1} It seems probable however that No. 19 came from this spot.

^{page 416 note 2} Continuous heavy rain prevented a proper examination of the village, but (apart from the declarations of the natives, who were kindly disposed) three previous expeditions found nothing.

^{page 416 note 3} Perhaps a name like

^{page 417 note 1} An exact parallel occurs at Hierapolis J.H.S. 1885, p. 346, No 75).

^{page 418 note 1} Better (Τ)οΤεανῶν, see infra.

^{page 420 note 1} The name occurs with one T or with two indifferently (cp. Hist. Geog. p. 240 &c).

^{page 420 note 2} Iniuriose is the technical term (Dig. i. 19).

^{page 421 note 1} This agrees with the situation assigned to the Moxeanoi, in CB. ii. p. 631Google Scholar f.—a situation indeed already confirmed by epigraphic evidence (No. 615). I should now, however, prefer to say ‘the people of Soa and of Tottoia’; but the suggestion was received too late to be incorporated in the text. The change does not, however, affect what is said about the boundaries of the estate.

^{page 421 note 2} I heard that Gumulu and Hassan keui are villages near the Devrent (on the eastern side) between Otourak and Islam keui: but I have not seen them.

^{page 422 note 1} The Imperial Domains and the Colonate,London, 1890.

^{page 422 note 2} Proved for Africa by inscriptions, especially C.I.L. viii. 10570 (discussed by Mommsen, , Hermes xv. 1880, p. 385Google Scholar ff.) and for the Milyadic or Killanian Estates in Asia Minor by Ramsay, (CB. i. p. 284Google Scholar). The lex Hadriana undoubtedly applied to all the other estates. Prof. Pelham points out (p. 18) that the idea of the new system originated with Vespasian and Trajan, and this is confirmed by the African inscription recently published in Comptes Bendus de ľ Aead. des Inscr. 1897 p. 146 ff.

^{page 422 note 3} The organisation of Milyadio Estates is alsodescribed CB. i. p. 281 ff.

^{page 422 note 4} πάροικοι, sojourners, resident foreigners, as in C.I.G. 1625, 45; 1631; 2906, &c.

^{page 422 note 5} in Hierocles and the Notitt, in Not. Basilii and Not. Leonis (ed. Geizer) and Const. Porphyr, , de Thematibus i. pp. 14Google Scholarand 25.

^{page 422 note 6} E.g. the Lion Tomb and Palace described by Prof. Ramsay, in his Study of Phryg. Art (J.H.S. 1889, p. 176Google Scholar ff.).

^{page 423 note 1} At Kara Agateh ören (spore), north-east of Altyn Tash, we were likewise told that many stones had been taken thence to Kutāya.

^{page 423 note 2} The form Miros occurs in 536 A.D. (Labbe, viii. p. 974).

^{page 423 note 3} I do not mean that all Turkish statements re true, but the traveller can distinguish. If e.g. one is visiting a frequented district, and were to ask whether a well-known inscription, in the possession of a villager, had been copied before, your friend (in expectation of bakshish) would of course answer ‘No.’ But when there is no motive for deception, there is less reason to disbelieve.

^{page 424 note 1} See, for example, CB. p. 583 and Nos. 498,499.

^{page 424 note 2} As Radet, M. says, Rev.Univ.Midi, 1896, p.6Google Scholar.

^{page 424 note 3} Not of Nakoleia, whose territory could hardly extend to the west side of the mountains.