^{page 370 note 1} Of course not until Mr. Sterrett had learned my ways of work.

^{page 370 note 2} Besides this I have impressions made by Mr. Sterrett of many of the inscriptions which he copied: in such case I still attach his initials to the text. I hoped here to be able to refer to an important series of inscriptions copied by us at Tralleis, which Mr. Sterrett is preparing for publication; but an unfortunate accident has delayed his work.

^{page 371 note 1} I omit four which have been identified in my own papers, also Ceretapa, Dionysopolis, Trajanopolis, placed in the right district but on the wrong site, and Eudocias Gand other temporary names of well known cities.

^{page 371 note 2} Writing in Smyrna I have to depend on rough notes made during a very hasty and inadequate examination of the Acta Conciliorum in the Athenian University Library. The Indices to the Acta and the lists of bishops in Le Quien, Or. Christ., are so imperfect as to be useless for my purpose.

^{page 372 note 1} ‘L'ordre d'Hiéroclès, qui est très souvent l'ordre géographique,’ Waddington, , Voy. Numism. p. 50Google Scholar.

^{page 372 note 2} Journ. Hell. Stud. 1883, p. 40.

^{page 372 note 3} I use the word city in an emphatic sense.

^{page 373 note 1} It is true that in the reign of Valens the Church had not the power which it obtained in the Byzantine period.

^{page 374 note 1} It was perhaps at this time that Cotyaion was detached from Pacatiana and assigned to Salutaris.

^{page 375 note 1} See C.I.G. 4300h.

^{page 375 note 2} See an inscr. of Attaleia published by me in Bull. Corr. Hell. 1883, p. 263.

^{page 376 note 2} However this may be, I have no doubt that the Leda of Spartan legend bears the Lycian name, Lada: the remarkable analogies which have recently been discovered between the antique art and hieratic symbolism of Sparta and of Lycia prove that interchange of religious and mythological forms between the countries is probable.

^{page 377 note 1} The general view on Pl. XXXIV, might pass for a picture of one of the ‘Geuzlar.’ I speak of Steuart's tomb from memory, not having seen the book for years: I have twice looked in vain for the tomb at Gherriz.

^{page 377 note 2} I often refer to Dr. Kiepert's views in the appendix to Franz, Fünf Inschriften.

^{page 377 note 3} The two roads in the Table meeting at Laodicea must be corrected thus:

Sardis 25 Philadelpheia 34

Tripolis 12 Hierapolis 6

Ephesus 35 Magnesia 17

Tralles 45 Antiocheia 31

The numbers are of course only approximate.

^{page 378 note 1} Moxeanoi and Mokkadenoi are the only others.

^{page 379 note 1} M. Waddington has a late coin with the legend ΜΟССΗΝΩΝ.

^{page 379 note 2} Voyage Numismat. p. 50.

^{page 379 note 3} Ova valley, Tchal a kind of soil.

^{page 379 note 4} Demirdji Keui is a Kaimakamlik: the name means Blacksmith Village.

^{page 380 note 1} He was misled by Arundel's somewhat confused language into the belief that this site (see III) was in the plain.

^{page 381 note 1} Equivalent to 232 A.D.

^{page 381 note 2} The letters in the first three lines are much larger than in the others. The shape of these stones is peculiar: it is that of a square pillar surmounted by a capital, but the pillar is only about four inches high and the capital about eight.

^{page 382 note 1} Foucart in Saglio, Dict. Antiq. s. v. Apeleutherismos: none of the inscriptions referring to this custom are accessible to me while writing.

^{page 382 note 2} See Μουσ. Σμυρν., No. τλ where unfortunately the date is mutilated. On the survival of the ancient custom of ἑταιρισμός in Lydia as late as 200 A.D., see an inscription published by me, Bull. Corr. Hell. 1883, p. 276.

^{page 382 note 3} On the interchange of β and μ compare Ahrens.

^{page 382 note 4} ΛΑΙΡ of course resembling in sound our lair.

^{page 382 note 5} We saw it in possession of an Iatros in the Khan at Kaïbazar. In line 5 M and E are liée.

^{page 383 note 1} HMH in line 6 and KA in line 7 are written liée.

^{page 384 note 1} Xenophon, Anab. i. 2, 13: Pausan, i. 4, &c.

^{page 384 note 2} Various reasons, which I cannot here specify at length, confirm the words of Stephanus.

^{page 384 note 3} The district is a great vine-growing one: this would give a local colour to the cultus of the Sungod.

^{page 385 note 1} I had a little ‘row’ with the people, and left without making a proper revision of the text.

^{page 385 note 2} The reading is not certain.

^{page 386 note 1} In his Mélanges de Numismatique, I., 103. The emendation was only a restoration of the MS. reading, which had been unanimously altered by editors. Bargylia is a well-known town: Hyrgaleia is never mentioned in any other literary authority, not even in any Byzantine list. The MSS. therefore must be corrected, and we had to read Bargyleticos. It is true that Bargylia is a coast town of Caria, far from the Maeander, but that only showed what was already well known—the ‘inaccuracy’ of Pliny.

^{page 387 note 1} There were never any letters in the gap in l. 5. The reading Οὐενῶκα is suggested to me by M. Waddington.

^{page 387 note 2} M. Waddington is misled by Hamilton's rather ambiguous language into the belief that Demirdji Keui is in the Baklan Ova: it is in the Tchal Ova. Hamilton found his inscription fully three hours from Demirdji Keui, after crossing the range of hills that divides Baklan Ova from Tchal Ova.

^{page 388 note 1} On Μοτελληνός, see below No. 14. The date is probably 126 A.D., see No. 14.

^{page 388 note 2} Blaundos must be fully twelve hours' journey from Hierapolis.

^{page 389 note 1} Mosyna was at this time not a separate city, but a village of the territory of Hierapolis.

^{page 389 note 2} On their situation see below ss. vv.

^{page 389 note 3} Only a person who has wandered over all these roads, who has looked from any hillock in the Dionysopolitan valley across to Blaundos and from Blaundos seen the view stretching unbroken to the Demirdji Keui hills, will appreciate the certainty of this reasoning. Blaundos is so placed as to communicate both with Hierapolis, twelve to fourteen hours, and with Philadelpheia, sixteen hours.

^{page 390 note 1} C.I.G. 2698b, Waddington, , Fastes, p. 102Google Scholar. Censorinus was dead, and no longer proconsul when the inscription was engraved: the circumstances of our inscription might explain the omission of the title, if it is omitted.

^{page 390 note 2} Besides numerous coins of Hierapolis, I saw a good many coins of Blaundos, Dionysopolis, Laodiceia, and Tripolis, fewer of Sala, one each of Hyrgaleia and Marcianopolis (in Thrace, a coin in most beautiful condition).

^{page 391 note 1} Only the second half of Π at the beginning of line 2 remains: the Δ for Λ in this line is an error of the engraver.

^{page 392 note 1} Omitting the district of the Ancyra bishoprics, which are in these Notitiae attached to the metropolis Hierapolis.

^{page 393 note 1} The large cross evidently marks the middle of the stone, so that exactly half of the inscription is preserved.

^{page 393 note 2} The reading ΑΔΕΛφΙΔ⊏Ι is quite certain. At least four lines are lost at the end, having been wilfully erased. In line 7 I read on the impression ΓΛ·ΚΕ or ΓΑ·Κ⊏.

^{page 394 note 1} Σπουδασάντων of the members of an association in an inscription of Apameia which I published Bull. Corr. Hell. 1883, p. 307.

^{page 394 note 2} I copied at Apameia an inscription in which two lines are transposed, see Bull. Corr. Hell. 1883, p. 308.

^{page 394 note 3} See below, No. 43.

^{page 394 note 4} Opuscoli di Modena VIII. 176: de Rossi, , Roma. Sott. I. 106Google Scholar: I have not seen these comments, but take the references from Duchesne, M. in Rev. Quest. Histor. July, 1883, p. 31Google Scholar. See also Waddington on Lebas, 1687.

^{page 394 note 5} On the opposite bank stands the village of New Aidan.

^{page 395 note 1} The three successive names Konioupolis, Sitoupolis, Krassos, are corrupt to an extent almost unexampled in the list.

^{page 395 note 2} Theoph, . Chronogr. I. p, 746 (p. 406)Google Scholar.

^{page 395 note 3} Under the form

^{page 396 note 1} Hamilton, No. 348: restored in C.I.G. [Τραλ]λέων, in Waddington Mel. Numism. I. p. 105

^{page 396 note 2} In 5, ΝΓ liée; so Μ⊏ in 6, ΝΕ and Η⊏ in 7, ΝΚ in 8, ΗΝΠ in 10. The last letter in Απολλοδοτος is according to the copy a monogram of Ν and ⊏, probably it is Ν corrected to ⊏.

^{page 397 note 1} Also Conni Metropolis, called in the Byzantine lists Conni Demetriopolis.

^{page 397 note 2} Until M. Waddington restored the true reading in Pliny, there was no other evidence than style to prove that Hyrgaleia was in Phrygia.

^{page 397 note 3} The sixth, seventh, and eighth letters doubtful: perhaps ΛΛΛ: the name is perhaps Κλ. Λαώβροτος.

^{page 398 note 1} See my paper in Journ. Hell. Stud. II. p. 286.

^{page 398 note 2} The statement made by me in Rev. Archéol. Sept. 1883 (in the Chron. d'Orient), with regard to Xenophon's route is wrong. To leave no doubt, I made another journey across the district after that statement was written. I observed that Peltai must be several miles further north than I at first thought.

^{page 400 note 1} In the same inscription and in another from Apameia I have published these Bull. Corr. Hell. 1883, p. 310 and 312.

^{page 400 note 2} I published this inscription in Bull. Corr. Hell. 1882, p. 516; are the δαίμονας devils, or is the inscription a mixture of pagan and Christian phraseology?

^{page 401 note 1} I see that M. l'Abbé Duchesne holds the same opinion, Rev. d. Quest. Hist. July, 1883, p. 31. But it would not be safe to assume the point without proof: ὁ θεὸς and ἡ θεὸς are common in pagan Phrygian inscriptions.

^{page 402 note 1} ΝΝ liée: the impression is very faint, and I do not feel certain that the reading is correct: possibly ΚΝ.

^{page 402 note 2} I have known large inscribed stones transported to a greater distance.

^{page 403 note 1} Manuel, marching from the Rhyndacus valley, p. 298. Compare Nicetas de Man. vi. p. 229, (towards Dorylaion)

^{page 403 note 2} Taba means rock in Carian.

^{page 404 note 1} If this be so, Hodjalar must be Tymion.

^{page 405 note 1} I give the statement on the authority of Smith's Dict., having no means of verifying it in Smyrna.

^{page 405 note 2} ΜΕ in 1, ΗΠ in 2, ΤΕ in 3, ΗΡ in 4, ΗΚ in 5, ΜΝ, &c., in 12, ΝΕ and ΝΚ in 14, liée.

^{page 406 note 1} Steph. Byz. s.v. Μεσ-ημβρία.

^{page 406 note 2} Assos, probably ak-yo-s, the peak.

^{page 406 note 3} Compare Thyessos, the peak of Thya.

^{page 407 note 1} The river Kissos is mentioned on a coin of Tomara in the collection of Mr. Lawson.

^{page 408 note 1} HM in 7, ME in 9, probably HN in 10, NE and HNE in 11, NE twice in 14, liée.

^{page 408 note 2} Incomplete at bottom. [H]E in 6, and HN in 7, liée.

^{page 409 note 1} See my paper on Abercius, in J.H.S. iii. p. 349Google Scholar.

^{page 409 note 2} My remark in the last number of this Journal, p. 36, note 2, must therefore be corrected. Phrougios also at Cotyaion and Aizani.

^{page 409 note 3} Seljükler, the Seljuks; Bounarbashi, head of the springs.

^{page 409 note 4} It must be said, in justice to Arundel, that he placed Sebaste only three or four miles to the west of Seljükler.

^{page 409 note 5} The modern form is evidently due to ‘false analogy’: ‘-lu or -lü,’ ‘endowed with,’ is an exceedingly common termination in Turkish.

^{page 410 note 1} No. 731 is correct.

^{page 410 note 2} The first letter in 7 is certainly ω.

^{page 411 note 1} The writer in the Bulletin remarks ma copie et mon estampage portent Νοντάνου. My copy and the impression now before me bear Μοντάνον distinctly. In order to leave no doubt on the numerous errors of the writer in the Bulletin, Mr. Sterrett undertook a two days' journey to Sebaste to compare once more the text of the Bulletin with the stones.

^{page 412 note 1} See Hirschfeld, , in Monatsb. Berl. Akad. 1879, p. 312Google Scholar.

^{page 414 note 1} Besh Sheher on Kiepert's map, on the authority of Arundel.

^{page 414 note 2} From the Lydo-phrygian Alucomes Aloudda, as from Attu- Attoudda. Alu-, the impetuous, Sanskrit arvan, is the Greek Aru-. Alu and ala, horse in Carian, are closely related.

^{page 416 note 1} Mordtmann, Marmora Ancyrana; Franz, Fünf Inschriften.

^{page 416 note 2} So Hierocles makes out of ὁ Βριάνων a town Briana, see above, XVIII. Such facts as these prove that Hierocles constructed his lists from the bishoprics of his day. His Tianai or Tiarai of Asia is formed from ὁ Τιανῶν: the town is Tia, i.e. Attea: the bishop is in the Notitiae ὁ Σιών. The bishop of Kolose or Koloe is sometimes ὁ Κολώνης.

^{page 417 note 1} By a correction of the text of Strabo: see Waddington on Lebas, No. 668.

^{page 418 note 1} The place is referred to by Nicetas Choniates in his history of the Emperor Manuel, under the name Charax: I know no other place where it is mentioned.

^{page 420 note 1} Perhaps τί[ε]ιν.

^{page 420 note 2} With line 2–4 cp. Lebas, 805, at Cotyaion.

^{page 421 note 1} A town of Phrygia in which every inhabitant and magistrate was Christian, is said by Eusebius, H. E. viii. 1, to have been burned in the time of Diocletian.

^{page 421 note 2} I published it in Bull. Corr. Hell. 1883, p. 310.

^{page 421 note 3} Steph. Byz. s.v. Acmonia. Manes and Men I believe to be the same word: Journ. Hell. St. 1883, p. 31.

^{page 422 note 1} Prof. Gardner, who kindly sends me a note of this coin, adds that the emperor may be Caracalla.

^{page 422 note 2} ΓΕ liée in 9, Κ very large; does it stand for Κυρίου, guardian of his son in 10 ΠΕ liée.

^{page 423 note 1} It is doubtful whether the guttural has been dropped between two vowels (see above, XXV.), Δόκελα = Do'ela = Dola, or whether it has been softened to gh, which is silent before l.

^{page 424 note 1} This explanation did not occur to me for many months after finding the inscription: from the moment of finding it I always assigned the date, on account of the nomenclature, as 115–30 A.D.

^{page 425 note 1} Lines 11–12 erased; the tops of the letters in 11 and 18 alone are preserved, and the bottoms of the letters in 12.

^{page 425 note 2} See Duchesne, , Saint Abercius, in Rev. des Quest. Histor., July 1883, pp. 1–33Google Scholar: de Rossi, Bull. d'Archéol. Chrét, 1882, p. 79Google Scholar: Duchesne, , Bull. Critique t. iii. p. 135Google Scholar: also Journ. Hell. Stud. 1882, Pt. II.

^{page 426 note 1} The only one available to me at the time of writing.

^{page 426 note 2} So Pitra and Dübuer: Duchesne accepts the general sense obtained by them, though doubting their exact reading.

^{page 427 note 1} See my paper on the Phrygian Language in Journ. Roy. As. Soc. 1883, p. 32 of the reprint.

^{page 427 note 2} Βασιλῆαν thus interpreted disagrees with M. Duchesne's allegorical interpretation.

^{page 428 note 1} The conjecture of M. de Rossi was unknown to me till after my return from Phrygia in October. The remarks made by me in Rev. Archéol. 1883, September, on the authority of Mr. Sterrett, require correction in this one point. In all others Mr. Sterrett's reading and measurements were perfectly accurate. I saw the stone in October. The first hasty copy of the stone made by me in November 1881, and published in Bull. Corr. Hell. 1882, p. 518, was accurate in every point except the reading I for P; but I did not observe that lines 3–4 were longer than lines 1–2.

^{page 428 note 2} In 2 Τ and Ε liée, in 9 the second ή of is a very small letter inserted as a correction between and Ι, in 10N and M liée, in 12 ON at the beginning are certain, but the left side of the following letter is blurred and may be either ρ or φ or a lettre liήe.

^{page 430 note 1} Copied by Hamilton at Sandykly: commented on by de Rossi, , Roma. Sotter. i. p. 108Google Scholar.

^{page 430 note 2} Trois Villes Phryg. in Bull. Corr. Hell., July, 1882.

^{page 431 note 1} Apparently the Λ of Γαλ(έριος) was wrongly engraved.

^{page 432 note 1} Mionnet gives one such coin. The ruins of Hieropolis are far more imposing than those of Otrous or Stectorion.

^{page 432 note 2} Β in line 4 is of peculiar shape, Υφ and ΥΠ both in monogram.

^{page 433 note 1} Finlay refers to Niceph. Pat. 3 which I have no means of verifying.

^{page 433 note 2} in 8; a whole line omitted after 9.

^{page 434 note 1} There the governor is a ἡγεμών, in Hierocles he is a ὑπατικός.

^{page 434 note 2} τ′ is not common in C. I. G., but I know many unpublished examples.

^{page 435 note 1} See Perrot, , Voy. Archéol. p. 126Google Scholar.