¹ See especially Kretschmer, Einleitung, etc. (here cited as Kretschmer); Fick, Vorgriechische Ortsnamen (cited as Fick) and Hattiden und Danubier, etc.; and Pauli, Altitalische Forschungen, vol. ii. parts 1 and 2, Eine Vorgriechische Inschrift, etc. (cited as Pauli, ii. 1 or 2). Vol. iii. of the same work, Die Veneter, etc., is cited as Pauli, iii. Names from Asia Minor quoted without a reference will be found in the index to Sundwall, Die einheimischen Namen der Lykier, and from Thrace, in Tomaschek, , Die alten Thraker, ii. 2Google Scholar. Other geographical names without reference are to be found in Pauly-Wissowa.

² If Lycian were an Indo-European language, related to the Phrygian, there would be no need to go to Asia Minor for the origin of any of these suffixes. The argument must proceed on the assumption that it is not. The differences are in fact, in my opinion, fundamental and irreconcilable. The resemblances hitherto verified are not beyond the range of coincidence. But ProfessorKalinka's, belief (T.A.M. i. p. 10Google Scholar) that it is a mixed language may prove to be correct. It is probable at least that the vocabulary has been deeply affected by one or more Indo-European languages, and the same may be the case with the grammar to some extent.

³ See below, p. 53.

⁴ Kretschmer, p. 405, only mentions five instances, but he might have added at least ten more.

⁵ S, like other consonants, is doubled before i, which itself disappears, as Arnisses for *Arnisies, etc.

⁶ P. 296. It is certain that nd is always written in Lycian with a t, and that the sound almost always arises out of a t preceded by a nasal. That -νδ- in this particular suffix so arose is not proved, but it is highly probable.

⁷ As in Cocynthus, Ἀρίνθη and Ἀριάνθη, in the Illyrian region of South Italy, which regularly represent names with the common Illyrian suffix -ntia (also -ntium, etc.). For in Messapian t before i turns to θ, and the i is usually dropped. The word inθi occurs in Messapian.

⁸ See Kretschmer, p. 402.

⁹ This name (Diodorus v. 64), which is omitted by Pauli and Kretschmer, is certainly connected with that of the Phrygian tribe of the Berecyntes. The existence of a Βερεκύνθιον ὄρος in Phrygia is denied in Pauly-Wissowa (s.v. Berekyntes), but without any assigned or discoverable reason.

¹⁰ See Conway, , B.S.A. viii. p. 154Google Scholar, who himself has overlooked the name Βερέκυνθος which would have considerably helped his argument. His contention that the language of the Eteocretan inscriptions is Indo-European and allied to the Venetic-Illyrian is highly probable. It seems to me to have no kind of resemblance to Lycian. It will be seen that I accept his conclusions in general in respect of local names in Greece, though on grounds which are only partly the same as his.

¹¹ The question has not been simplified by the discovery, by American excavators, of Lydian inscriptions written in a language strikingly unlike either Lycian or Phrygian. The greater part of the proper names contained both in these and in the Greek inscriptions of the country seem to be Phrygian, and probably belong to the Maeonians, who preceded the Lydians. A few are akin to the Lycian, and may be assigned to a yet older population. The true Lydians seem to have been a race of comparatively late intruders, after the time of Homer. If so, the common worship, on which was based the belief in their blood-brothership with the Carians and Mysians, was taken over from the Maeonians as part of their title to the soil. See below, p. 72.

¹² In Lycia the commonest (after νδ and -σ-, -σσ-) are -δα, -λα, -ρα, -μα, -να These all have parallels in Phrygia,. as in Σύνναδα Μάνταλος, ᾿´Αγκυρα, Δίνδυμος, Μόσσυνα They are also to be found in Thrace and Illyria, and other European countries. By no means all such words are formed from proper names in the manner usual in Asia Minor, but there is no general test by which they can be distinguished from the Lycian.

¹³ Probably for Βαγαντα with the Phrygian -ντ- as in γχαντα, Ριμενιαντα, Θιουντα, Πεσσι νουντ, Γορβεουντ The Lycian, Cilician, and Pisidian change of nt to nd spread to the Greek dialect of Pamphylia. (Kretschmer, p. 300), and may easily have affected the Greek or the original Phrygian forms in this mixed region (see Ramsay, , Cities and Bishoprics, i. p. 286)Google Scholar. Κυλλανδιοι Κυλλοντιοι in Caria (Kretschmer, p. 301) may be explained in the same way.

¹⁴ See below, p. 62.

¹⁵ See p. 52.

¹⁶ Kretschmer, pp. 183 and 194. Mostly they represent Fιον

¹⁷ See below, p. 60, note 118.

¹⁸ It is necessary to observe that the name Lycia is used both by Kretschmer and Sundwall in a very wide sense, so as to include districts which were not Lycian until Roman or even Byzantine times. Therefore many names will be found in their works which appear to be exceptions to general statements made in this article. It is impossible in every case to enter into explanations, but on verifying the references it will, I believe, be found that such discrepancies are due to this difference of definition. A good many names are quoted in Sundwall's book from unpublished inscriptions simply as Lycian without mention of the place of discovery. Such names are quite as likely as not to come from places outside Lycia in the sense in which the word is here used.

¹⁹ Strabo, p. 520 (probably from Eratosthenes), and p. 651. See Kalinka's, remarks, Jahreshefte, viii. Beiblatt, p. 42Google Scholar.

²⁰ In Lycian, μπ always turns to μβ. Kretschmer (p. 301)Google Scholar takes the word (rightly in my opinion) to be Phrygian.

²¹ Thracum suboles Mityae (Pliny, v. 27) means no doubt that they were Phrygians, : Hecataeus, called them ἔθνος Φρυγίς, fr. 206Google Scholar. The words of Arrian about Milyas, , ἥ ἐστι μὲν τῆς μϵγάλης Φρυγίας, ξυνϵτέλϵι δὲ ἐς τὴν Αυκίαν τότϵ οὕτως ἐκ βασιλέως μϵγάλου τϵταγμένον, i. 24Google Scholar, mean that it was geographically and probably racially part of Phrygia, not that in his own day it was administratively joined to it instead of Lycia, for that was not the case. Alexander entered Milyas from the Xanthus valley, no doubt over the main pass north of Ak Dagh, and went on to Phaselis. This confirms Ptolemy's account of the Lycian part of Milyas (v. 3), which SirRamsay, W. unnecessarily doubts (Cities and Bishoprics, i. 317)Google Scholar. The country extended also far to the north (ibid.), and included the country of the Ὀρμηγϵῖς (ibid. p. 280), but it was all counted as part of Phrygia in the time of Alexander, who is described as leaving Pisidia and entering Phrygia near the Ascanian (i.e. the Phrygian) lake, (Arrian, i. 29). The country of the Ὀρμηλϵῖς, which was certainly in Milyas, is proved to be Phrygian by the proper names found there (see p. 69), and is probably the Cillanian plain of which the population was a mixture of Phrygian with a certain amount of Pisidian, Strabo, p. 629. See Ramsay, , Cities and Bishoprics, i. p. 278Google Scholar.

²² The Cabalian towns of Bubon, Balbura, and Oenoanda were only taken from Cibyra, and added to Lycia by Murena about B.C. 81. Four languages were spoken in the district (Strabo, p. 631) but Lycian was not one. In philological discussions Cabalian names should not be quoted as Lycian, as is commonly done.

²³ J.H.S. xv. p. 95.

²⁴ This is less true of local than of personal names, which, no doubt, were put into Greek letters originally by the bearers of them, in most cases. Names of places on the contrary were adapted, not by Lycians, but by Greeks before the hellenization of the country. In many instances this is known to have been the case, and it was probably so in all. For the proximity of every part to the sea and the nearness of Greek colonies, especially Megiste on its very shore, made the whole country familiar to the Greeks at an early date.

²⁵ Names were borrowed from Greek, Persian, and Phrygian before the Greek period, as pericle, arttum̃para, mida. Some were so entirely naturalised as to form part of native compounds, as the Phrygian Κωκος (Kretschmer, , p. 188) and ΚοταςGoogle Scholar (Tomaschek, ii. 2, p. 50) in epñ-kuka and Ερμα-κοτας. But these were subject to the phonetic laws of the native language.

²⁸ In this respect I shall follow the practice of Tituli Asiae Minoris in employing e, which though less accurate is more convenient.

²⁷ It had also the nasalised vowels ã and ẽ (which appear in Lydian likewise), and the sonant liquids m̃ and ñ. These are not found in Phrygian, nor so far as is known in Carian, where the m is apparently syllabic with a suppressed vowel, not itself a vowel.

²⁸ Though the Carian alphabet is yery obscure, the great number of different vowels would almost be enough to prove that ο and u were distinguished.

²⁹ Jahreshefte, ii. p. 55.

³⁰ Apparent exceptions aro Κτησκϵιτος, (J.H.S. xv. p. 112)Google Scholar and Σαρητιος. The first however, is hellenised so as to resemble the Greek names beginning with Κτησ-. The second has a termination -ητιος which is quite alien to Lycia, and markedly Indo-European: it is therefore probably a foreign name.

³¹ The η in Δαρηος is due to the same cause. Κϵνδηβις (Sundwall, , p. 92)Google Scholar, in an unpublished inscription, may not be from Lycia proper, as the place of origin is not mentioned (see note 18). But in any ease it probably stands for ΚϵνδϵϜϵβις, from Κϵνδϵας (for ΚϵνδϵϜας) or Κϵνδϵος, found in Pisidia, Pamphylia and Cilicia.

³² See p. 48, note 24.

³³ Σεμενδησις may perhaps be for ᾿´Σεμεν δεσσις But it is at least equally probable that it stands for ῾´ΣεμενδεϜεσις

³⁴ Ιδεβησσός does not apparently properly belong to Lycia (see p. 48).

³⁵ Τϵλμησσός, , Καρμυλησσός (only in Strabo, p. 665)Google Scholar and Ἀρτύμνησος (only in Stephanus Byzantinus). Καβῃσσός in Lycia is a misquotation; the form cited from Hellanieus is Καβασσός (St. Byz. s.v.). The name occurs in connexion with Homeric commentary, which makes it rather suspicious. Σαρδησσός (St. Byz.) ‘near Lyrnessus’ must even if the text is correct have been in Pamphylia.

³⁶ For instance, Stephanus gives ᾿Ακα λησσός though the correct form is certainly ᾿Ακαλισσός There is great uncertainty among Greek authors in general about these terminations, both in respect of the vowel, and the doubling of the σ.

³⁷ Herodotus, i. 78, 84.

³⁸ Κουρδαλι, (gen.) though ungrammatical seems certain (Reisen, ii. 7)Google Scholar. Σούρα is for the Lycian sure, but the transliteration of local names is not always exact (see p. 48, note 24).

³⁹ An apparent exception, Ουλλιας, (compared with the Cilician, Ολλις, Sundwall, , p. 227)Google Scholar is not so in reality, as the man or his father is described as Και[σ]αρϵύς, , (Reisen, ii. 107)Google Scholar, probably from Cappadocia. There was no Caesarea in Lycia.

⁴⁰ Ωαμαρας (C.I.G. 4303e) for *uwamara and Τλῶς for tlawa.

⁴¹ If Ωσσεου (gen., C.I.G. 4300d) is connected with the Carian Οσεας it is probably a Carian name. There is no certain instance of a Lycian name in εας or εας But ω may represent uwa, a common element in proper names. Κοδρωνας (quoted by Sundwall) should represent Κοδρ-οανας (^*-uwana), unless it is meant for a Latin Quadronius, like Κοδρᾶτυς for Quadratus.

⁴² For Καλλι as often in inscriptions.

⁴³ Δήμαρχος Τάρωνος Αύκιος, Dittenberger, Syll. ²183Google Scholar (inscr. found at Samos), may also have been Lycian in the wider sense only.

⁴⁴ Κομων (Ramsay, , Cities and Bishoprics, i. 337)Google Scholar, Μαριων, (Studies in History, etc., p. 326)Google Scholar are Phrygian; Νοννων is Isaurian, and probably Phrygian in origin; Μάρων is Thraeian.

⁴⁵ Ramsay, , Studies, pp. 363Google Scholar, 365, and 371.

⁴⁶ Βατων, , Βοσθων, , Βοτων, , Βωλιων, , probably Βαβιων, see below p. 60Google Scholar. Also Κοτυλων (Phrygian, and Thraeian, Κάτυς, Tomaschek, , ii. 2, p. 50)Google Scholar, Μαριωρ (see the last note), Μιννιων (Bithynian Μινας, Tomaschek, op. cit. P. 24).

⁴⁷ Kretschmer, p. 364.

⁴⁸ Kretschmer, p. 364.

⁴⁹ Iliad, v. 44.

⁵⁰ Ramsay, , Studies, etc., p. 322Google Scholar.

⁵¹ Ramsay, , Cities and Bishoprics, i. p. 288Google Scholar.

⁵² See index to C.I.L. iii.

⁵³ See p. 59.

⁵⁴ Calder, , J.H.S. xxxi. pp. 188, 190Google Scholar.

⁵⁵ Kretschmer, pp. 186, 188.

⁵⁶ Iliad, ii. 844.

⁵⁷ Cocaius, Dardanian (Naissus), Jahreshefte, iii. Beiblatt 131Google Scholar; Dalmatian Cocus, Pauli, iii. p. 365.

⁵⁸ Jahreshefte, vi. Beiblatt, p. 3.

⁵⁹ Herodotus, v. 16. The tribes round Mount Pangaeum (of whom the Orescii were probably one) were Paeonian.

⁶⁰ Kretschmer, p. 405.

⁶¹ Only known from coins.

⁶² J.H.S. xxxv. pp. 102, 103.

⁶³ The proper name Ουασος is certainly for *uwaza as in ap-uwaza, and Ερμαδεσσα for ^*erm̃medeze as in hana-daza. Οσσυβας is probably for ^*uzube, as in uz-eh[l]ẽmi; compare the Carian Παν.αβλημις. Ωσσεας is probably foreign, see p. 50, note 41. Other cases are outside the boundaries of Lyeia.

⁶⁴ See Jahreshefte, ii. p. 68. It corresponds to a Persian sh and to an Aramaic shin.

⁶⁵ In the name mullijeseh (gen.) = Μολλισιος. The engraver has twice written a Greek ι by mistake for a Lycian i.

⁶⁶ Λαλλα is not uncommon, but may like some other names be borrowed from Phrygia (Λαλα from Galatia, , C.I.G. 4123Google Scholar, also Thracian, Kretschmer, p. 352). Βαλλιων is certainly Phrygian (p. 56). Μανσωλλoς is Carian. Σολλοσος and the Isaurian-Cilician Σουλλις are probably Phrygian (compare Thraeian Sola and Sulu; Illyrian *solas, Messapia, , Soleia, Venetia, Pauli, iii. p. 358Google Scholar, Solia, Noricum, ibid. p. 376). Μολλισις, mentioned above, is also Phrygian in origin; compare Μολυξ from the Phrygian region of the Ὀρμηγϵῖς (Ramsay, , Cities and Bishoprics, i. p. 314)Google Scholar, the Macedonian Μόλυκος (Hoffmann, , Die Makedonen, p. 211Google Scholar) and Μόλων (ibid. p., 228), also found in Caria, , C.I.G. 2748Google Scholar. Compare the Illyrian, Mollico, Pauli, , iii. p. 362Google Scholar, and the Venetic *molos, molo, Mollo, Mollonius, etc., ibid. p. 328. In Lycia proper, the only other related name is mula, Μολας, in an inscription in which Carian names are mixed with Lycian, (T.A.M. 32)Google Scholar. Μολης, Μολϵσις, etc., quoted as Lycian, are really Cabalian, , Milyan, , Pamphylian, , and Pisidian, . The reading Σϵλλις (Reisen, ii. 156)Google Scholar is uncertain: Σϵδδις is perhaps more probable. Σιλλης and Σιλλ[ϵ]ας are also quoted in Sundwall's work. The last, at least, if rightly restored is foreign to Lycia, in which the ending -ϵας does not occur in native names, see note 41.

⁶⁷ E.g. J.H.S. xxxi. p. 182, in an inscription partly written in Phrygian; cf. Kretschmer, p. 351.

⁶⁸ As in J.H.S. xxxi. p. 181.

⁶⁹ Πάππος occurs also once at Olympus. Otherwise the form does not belong to Asia Minor, but is simply a not uncommon Greek name. As such only it was introduced into Lycia; Πἁπος is a variant. The derivatives Πἁππων and Πἁππων are Greek in form, and foreign to Asia Minor, especially to Lycia, where names in -ων cannot be native; see p. 50.

⁷⁰ Kretschmer, pp. 199 and 241.

⁷¹ Kretschmer, p. 345.

⁷² It seems to me certain, however, that the forms in Απϕ- from which Αϕϕ- and Αϕ are formed, are hellenised on the model of the Greek ἀπϕῦς. ᾿Απϕάριον, Ναννἁριον and Τατά ριον seem to be Greek in form, as ΜαμμἁριονΑ certainly is (Kretschmer, p. 339), and with these must be classed ᾿ Απϕαροῦς, ᾿Αμμαροῦς and Ταταροῦς with the common late Greek feminine suffix -οῦς. Except in obviously hellenised or Latin or Persian names, ϕ is generally of the rarest occurrence in Asia Minor. Neither in words allied to the Lycian nor the Phrygian does it seem to be native. Its extreme frequency in this one class of names at a late date can hardly be explained except as the result of Greek influence. Of the two other names beginning with ῾Αππ- already mentioned, ῾Αππίων seems foreign to Asia Minor.

⁷³ Kretschmer, pp. 230 and 196.

⁷⁴ C.I.G. iii. addenda 4315d. That Loew was not infallible as a copyist may be seen on the same page, by comparing 4315 and 4315b with T.A.M, i. 139 and 152.

⁷⁵ It is not improbable that the first word should also be corrected to ῾ ´ Πισιναρις on the analogy of Πισινδηλις etc. But I do not think that the initial B can actually represent a native initial p as Sundwall suggests (p. 181). No instance of such a transliteration appears to be well established, except the change of an initial pd, unpronounceable in Greek, to Βδ in the Pisidian Βδενασις which is not analogous. The change of mp to mb took place within the Lycian language. On the contrary a native b was often altered into a Greek π; see below, p. 62, note 132.

⁷⁶ St. Byz. s.v. He gives the ethnic as Τύμνιος but the typical Carian and Lycian ethnic was -εύς (ibid; s.vv. Ξύλος, ᾿ Αγἀθη etc.), which commonly has the feminine in -ίς.

⁷⁷ Also from Upper Moesia (Dardanian, and Mysian, ), Jahreshefle, iv. Beiblatt, pp. 85, 86Google Scholar.

⁷⁸ Kretschmer, p. 318.

⁷⁹ See Boeckh's, note on C.I.G. 2042Google Scholar.

⁸⁰ Cf. Pauly-Wissowa, s.v. The Carian Βρησιον is similarly derived. Names beginning with Βρ- akin to Lycian are hellenised forms of originals in Mr-, just as Μλαυνδος was changed by the Greeks to Βλαῦνδος. Thus the Carian Βρύασσις corresponds to a Lycian, mru[w]asi (to be so read in T.A.M i. 55, 4)Google Scholar, compare mẽ-mruwi: as the Cilician Βλενδιος is for mlẽ-tije (like mizre-tije) compare mlẽ-tederi. There is no question in such cases of any exchange of consonants in the native languages, but only of the substitution of a possible for an impossible combination in the Greek transcription. Names in Βρ- and Βλ- may therefore be left out of consideration.

⁸¹ See below, p. 62.

⁸² See below, p. 59.

⁸³ Sundwall, p. 283. As the locality is not mentioned, it may not be Lycian in the exact sense.

⁸⁴ Βαλας is found at Thessalonica, in the native land of the Phrygian Mygdonians.

⁸⁵ Kretschmer, p 242¹; Tomaschek, ii. 2, pp. 11, 12. The Dardanian Βαλλανστρα may be for Βαλλαν-στορα, ‘stronghold of the king, see Tomaschek, ii. 2, p. 81. The root appears also in the name of the Dacian king Δϵκέβαλος (cf. Balius, Dacian, C.I.L. iii. 1629, 3)Google Scholar and the Illyrian king Βαλλαῖος: probably also in the Bithynian Zeus Βόληος, and perhaps in the Thracian Βαλιός (Dionysus).

⁸⁶ Possibly the Lydian, Βϵλϵτρος is from the same stem; see American Journal of Archaeology, xvi. p. 28Google Scholar.

⁸⁷ Kretschmer, p. 202.

⁸⁸ Also Γαρσαυρα which makes it probable tha ᾿´Ισαυρα represents Ισα-ουρα

⁸⁹ The root Aim- is Illyrian, and Paeonian, ; see Am. Journal of Archaeology, xvi. p. 51Google Scholar.

⁹⁰ ᾿ ´ Αζωρος, ᾿´Αλωρος, Γἀζωρος, Θέστωρος, Μί λκωρος, Πίλωρος, Τάρπωρον

⁹¹ Pauly-Wissowa, s.v.

⁹² Βόλβαι in Caria, a presumably older name of Heraelea (St. Byz. s.v.) does not seem related to Βάλβουρα, but rather to the town and lake Βόλβη, in Mygdonia, whence one of the Phrygian tribes migrated. If Tomasehek's derivation (ii. 2, p. 94) is correct, from the root bhol, Armenian bol-, ‘to swell, to be round,’ the Greek βολβός would seem to be borrowed from a dialect akin to the Phrygian. The islands Bolbulae (Pliny, v. 137, which should be emended to Bolbnsae) off the Ionian coast derived their name from the Greek word. But the district Βολβοσός in Cilicia Tracheia (Ramsay, , Hist. Geog. p. 371)Google Scholar no doubt had a native name.

⁹³ See p. 50.

⁹⁴ Kretschmer (p. 356) states this most distinctly. It appears to me that all the names he collects (pp. 334, seqq.) are genuinely Phrygian, but that he over-estimates the number of those which are also Lycian.

⁹⁵ In C.I.G. 4009, b, probably ὁμοίως ἀνέστη σεν καὶ Νερσιων(α) καὶ Βᾶν, Βᾶ θυγατέρα should be read instead of Βανβα The Phrygian local name Βανβουλα (p. 58) is probably a contraction of Βαναβουλα and akin to Βαναβα in Cilicia (Ramsay, , Hist. Geog. p. 371)Google Scholar.

⁹⁶ Βα appears once in Cilicia at Dalisandus. The Pamphylian Ἐλαιβαβης seems, like most Pamphylian names (see p. 68), to be unrelated to the Lycian. Otherwise no names are compounded with baba in the second part. No related names seem to be found in Caria, unless the Milesian Βαβιων be reckoned, as it probably may, as Carian.

⁹⁷ E.g. at Cotiaeum, Kretschmer, p. 223.

⁹⁸ Old Phrygian Baba, Kretschmer, p. 336.

⁹⁹ The neighbourhood of this place to Parthicopolis confirms its relationship to Bargullum, among the Parthini, Tomaschek, ii. p. 62.

¹⁰⁰ Hierocles, 641. This cannot possibly be the same as Bargullum, as suggested in Pauly-Wissowa, sub vv. It was in Macedonia, probably Pelagonia, while the other was near the Illyrian coast among the Parthini.

¹⁰¹ This may however be derived from the town of Βἐργα.

¹⁰² The north side of the Meander valley was Carian in Homer's time as far as Mycale, (Iliad ii. 869)Google Scholar, and was still so reckoned by Ephorus (see frags. 35 and 86). Later it was generally called Lydian. The population was mixed in Strabo's day (p. 648), but the Lydians were probably immigrants. Native Carian (not Lydian) inscriptions have been found at Tralies (Sayce, Proceedings of S.B.A. xxvii, Nos. 8 and 9).

¹⁰³ These two places (Ramsay, , Studies etc. pp. 361, 371)Google Scholar, together with almost all those named in the group of inscriptions in which they are mentioned, must be reckoned to Phrygia παρώρϵιος, to which Apollonia and Antiochia are distinctly assigned by Strabo (see Ramsay, , Hist. Geog. p. 397Google Scholar, and Cities and Bishoprics, i. p. 316). Late Phrygian inscriptions are found in this district (Ramsay, , Jahreshefte, etc. viii. Beiblatt, p. 85)Google Scholar. The names, local and personal, which occur there, are almost exclusively Phrygian, and I shall quote them as such. It was only under the Roman empire that this part of Phrygia was included in Pisidia.

¹⁰⁴ See Tomaschek, ii. 2, p. 63.

¹⁰⁵ Pauli, iii. pp. 374, 366.

¹⁰⁶ Ibid. pp. 371, 377.

¹⁰⁷ Hoffmann, , Die Makedonen, p. 224Google Scholar. It is also Thracian, Tomaschek, ii. 2, p. 16. The Phrygian *Βουδας in Βουδϵια is a form of Βουτας, like the Illyrian, Λυγο- Αυτο- (Kretschmer, p. 247)Google Scholar, Μϵδέων-Μϵτέων (ibid. p. 257). The Thracian Βούζης is for Βούδης as e.g. Δορ-ζϵνθης for -δϵνθης: it is identical with Βωζῆς from. Pergamum. The Phrygian town of Βοζα is probably from the same stem, if it really existed (see P.-W. s.v. Bozenos). See Kretschmer, p 199¹. By a slip in Sundwall, p. 176, Βοζα is confused with Βαζις in Cappadocia.

¹⁰⁸ C.I.L. iii. 4944.

¹⁰⁹ Kretschmer, p. 245.

¹¹⁰ Pauli, iii. p. 370; also Bataro, p. 369, Bateia and Batelus are found in Noricum. ibid. p. 373.

¹¹¹ Conway, Italic Dialects, index.

¹¹² The Isaurian or Cilician Βαδας is probably for ᾿´Βατας as Βουδας for Βουτας (note 107). With it is connected the Cappadocian feminine name Βαζεις and the town of Βαζις in the same way as Βουζης with Βουδας. The dative Βαδι in Heberdey-Kalinka, p. 7, from Milyas, is probably from Βᾶς genitive Βᾶδος as Παππᾶς Παππᾶδος

¹¹³ Καλαβώτης found in Caria is certainly a Greek word ‘lizard,’ a known form of ἀσκαλα θώτης. Thence, I believe, is derived the Lycian place Καλαβατία a corruption arising from the fact that the Lycians could not pronounce the Greek ω (see p. 50). Most places on the coast had Greek names. The river Κολοβατος is not likely to be a compound, as local names are almost always formed with a suffix. It is probably from the same stem as the neighbouring town of Κολβάσα. That district, afterwards reckoned Pisidiau, was originally part of Phrygia.

¹¹⁴ See p. 51. The name does not, I believe, occur in any compound proper name of the Lycian type. Even for one of these, the supposed Cabalian Μολωβουλουβασιος (gen.) would be too long. It is certainly a double name, Μολεβης Λουβασις such as are common enough in Asia Minor (see Sundwall, p. 265). In an inscription of the same family we find Μόλης δὶς τοῦ Λουβασιος (Heberdey-Kalinka, p. 47).

¹¹⁵ Through Strabo, p. 649, and Hesychius call them Thracians, Stephanus Byzantmus, in describing them as Illyrian (sub vv. Τραλλία Βῆγις and Βόλουρος), cites the better authority of Theopompus, and Livy (no doubt following Polybius) expressly mentions several times that they were Illyrians (xxvii. 32; xxxi. 35; xxxiii. 4). They were much used in the armies of the Macedonian kings, who are more likely to be the βασιλεῖς referred to by Hesychius than the kings of Pergamum. See Ramsay, , Hist. Geog. p. 112Google Scholar.

¹¹⁶ Also in the Illyrian tribe of Bulini and the district of Bullis or Βυλλίς;

¹¹⁷ Βιττω (p. 56), Βολβαι (note 92), Βολβαι (p. 58), Βάργασα (ibid.), Βαργυλία (ibid.), Βρίουλα (ibid.), and Βάλαγρος (ibid.) have been already discussed.

¹¹⁸ See p. 47.

¹¹⁹ It is very likely that Μολυνδεια quoted from Alexander Polyhistor (St. Byz. s.v.), may be formed in the Phrygian way from a proper name ῾῾ Μολυνδας. But, if so, it was probably in Milyas (see p. 48), which in Alexander's time was part of Lycia, and where names in Μολ- are remarkably common: in Lycia they are almost unknown and probably foreign, see note 66. The termination -νδας is also especially Milyan. The only Lycian place in -ια, Καλαβατια is probably meant for Greek, see note 113. The suggestion (Sundwall, p. 175) that Βερραβλωιον is derived from a possible Lycian *para-pruwa with a suffix -ija is not therefore convincing enough in itself to give any support to the view that a Lycian p may be rendered by Β.

¹²⁰ St. Byz. s.v. Βέροια

¹²¹ Tomaschek's derivation (ii. 2, 58) from the root bher, in the sense of fertile, is peculiarly suitable to the garden of Midas, Herod. viii. 138.

¹²² Grote's argument is very convincing and has other support, Pauly-Wissowa, p. 306 (2).

¹²³ The change of b to m is Thracian, Kretschmer, p. 236.

¹²⁴ As Carian Απ-δυβερις compared with Lycian Τυβερις Τυβερις in Τυβερισσός and Περπεν-δυβε ρις

¹²⁵ In the region of Antiochia Pisidiae, which was certainly really Phrygian (see p. 58, note 103). The proper name Βούβαλος occurring in the same district is a Greek word ‘antelope,’ but is almost certainly hellenised from ΒοϜαλος for fashion's sake, as often happened.

¹²⁶ F.H.G. iv. p. 358.

¹²⁷ Boatius also occurs as a gentile name in Italy, but only in the Illyrian Daunia (Conway, The Italic Dialects, ii. index).

¹²⁸ The Venetic φohiios ( = boiios) is taken by Pauli, (iii. p. 350Google Scholar) for Gaulish, but it might be Illyrian for Bovios, as arahos for aravos, ibid. p. 386.

¹²⁹ This however may very easily represent *Βοιης as there is ample precedent for the conversion of a native b into π, though not for the reverse change.

¹³⁰ P. 51.

¹³¹ The Lycian Κινδ-αβυρι (Kñt-abura) is compounded with the word abura (T.A.M. 55, 2), whence also by a common change of vowel Κεν-οβορ[ι]ς

¹³² The change of b to m in Thracian, especially where an n follows, is well-establislied (Kretschmer, p. 236). The substitution of a Greek Π for a foreign B is not uncommon in regions bordering on Paeonia. The people of Βριαντική in Thrace were also called Priantae: the Macedonian Πέργαμος is almost certainly for Βεργ Παλλήνη represents a Macedonian Βαλληνη (L. & S. s.v.). M t. Βόιον on the Epirotic frontier is also called Ποῖον On the Asiatic side we find Παρταρας for Bartaraś in a Lydian bilingual at Pergamum, Πάργασα for Βάργασα, Πρίουλα for Βρίουλα and other instances.

¹³³ Some held that the Paeonians were a colony from Phrygia, others that the Phrygians were a colony from Paeonia (Strabo, p. 331). The Mygdonians, from whom a Phrygian tribe were descended, were a people of Paeonia (Pliny iv. 10). Herodotus believed that the Paeonians were descended from Teuerians, by which he meant Trojans (Dardamans), as appears from ii. 114 and 118; from a comparison of vii. 20 with v. 13 it is to be inferred that Mysians were mixed with these Teucrians. Hellanicus (fr. 46) says that in the time of Macedon, son of Aeolus, the Mysians were the only inhabitants of the country besides the Macedonians. The true country of the European Mysians or Moesians was the district about Ratiaria. There they bordered on the Dardani, whom they probably accompanied or followed in their migrations. The neighbours of the Dardaniaus on the south and south-east were Paeonians.

¹³⁴ Strabo, p. 553. The derivation of Bagadaonia in Cappadocia is obscure.

¹³⁵ On this disputed name, see Kretschmer, p. 198. Torp's objection to the derivation from bhága-, on account of the suffix, does not seem to me so irresistible as to Kretschmer. For several parallel cases are found in Asia Minor, as Zeus Παππῶος by the side of Πἁπας (ibid. p. 199), the goddess ῾Αμμαία by the side of ᾿Αμμά (Pauly-Wissowa, s.v. ᾿Αμμάς), and possibly Σαβάζιος by the side of Σάβος (Kretschmer, p. 196), and Κακασβεύς by the side of Κακασβος (ibid. p. 351).

¹³⁶ Littmann, , in Sardis, vol. vi, pt. i, p. 39Google Scholar.

¹³⁷ It is uncertain if the Cilician local name Βάκα (Ramsay, , Hist. Geog. p. 386Google Scholar) has any connexion with this.

¹³⁸ Hoffmann, , Die Makedonen, p. 53Google Scholar. The name occurs at Pharsalus, but the bearers were not natives.

¹³⁹ Genitive of Βιλλις, according to Sundwall, p. 61.

¹⁴⁰ C.I.G. 4322; see also Addenda. Müller, who found the inscription among Beaufort's, papers, evidently mistook Chelindreh, by which Beaufort meant Celenderis (Karamania, p. 201)Google Scholar for Chelidoniae. No ancient remains seem to have been found by Beaufort on those barren rocks (p. 35), but he noted inscriptions at Celenderis (p. 201). This particular one seems to have been copied at Celenderis, by three other travellers (C.I.G. iii. p. 1152)Google Scholar.

¹⁴¹ It may be borrowed from the Greek, like the Macedonian Βίλιππος for Φίλιπ πος Such a change could not take place in Lycian, which had no initial β, and would tend to prove that Phrygian was spoken at Celenderis. But it is more probable that a Macedonian colony was at some period settled there, and retained traces of its original dialect in colloquial phrases. If so, Βιλλος has no direct connexion with the Phrygian Βιλλις which in that case may be better compared with the Illyrian, Venetic φila (bila), Pannonian Bilisa, Messapian, billas, biliovas, etc. (Pauli, iii. 344)Google Scholar.

¹⁴² Pauli, iii. p. 344.

¹⁴³ See the index to C.I.L. iii.

¹⁴⁴ As in Σαβάζιος for ΣαϜαζιος, Kretschmer, , p. 195Google Scholar.

¹⁴⁵ Ramsay, , Cities and Bishoprics, i. p. 324Google Scholar.

¹⁴⁶ Ramsay, op. cit. p. 326.

¹⁴⁷ The Isaurian or Cilician Βανις may be for Vanis, but in a native Pisidian inscription (Ramsay, , Revue des Universités du Midi, i. p. 360, No. 10)Google ScholarΟυα Νις Βαβου[ς] should probably be read, not Ουανις

¹⁴⁸ On p. 350, Pauli says that the Venetic name is Gaulish, on account of its resemblance to Celtic names formed with Vass-. But the Messapian form cannot be explained in that way, and there are many resemblances between Celtic and Illyrian names which are due to relationship, not to borrowing (ibid.).

¹⁴⁹ See Ramsay, , Hist. Geog. p. 398Google Scholar. There is no sufficient reason to suppose that ᾿Ορο αννεύς (Hall, , Classical Review, xii. p. 276Google Scholar) is synonymous with ᾿Ορονδεύς. The supposed assimilation of δ (Kretschmer, p. 307) depends on a mistaken identification of Τρέβενδα near Myra in Lycia with Τρέβεννα in Pamphylia. The two places had the stem in common, but not the suffix.

¹⁵⁰ P. 57.

¹⁵¹ P. 45, note 1.

¹⁵² P. 56, note 80.

¹⁵³ The number of examples in Phrygia is of course far greater. The author's object was only to include such names from Phrygia as he believed to be of Lycian or kindred origin. For the districts mentioned, see pp. 48, and 58, note 103.

¹⁵⁴ Names from lsauria proper are indistinguishable from those of Lycaonia.

¹⁵⁵ In J.H.S. xxxv. p. 100, I have shewn that the supposed θ in Lycian is a sibilant.

¹⁵⁶ For the Messapian names see Deecke, Rheinisches Museum, xxxvii., where many other examples are given. For the Latin equivalents, see the index to Conway, The Italic Dialects, vol. ii. The same change in Thracian is proved by the names Bitus, Bitius, Βιθυς -centus, -κεντιος, -κενθος Cuta, Cuties, -κυθης, Κουθεις and other examples.

¹⁵⁷ In some cases the Messapian o certainly represents u, which is otherwise wanting in the inscriptions, and it may be questioned whether it does not always do so, as an original o regularly turns to a; Kretschmer, p. 263, holds that o stands for ō in some instances.

¹⁵⁸ Pauli, iii. pp. 374, 377, 353, and 368.

¹⁵⁹ Tomaschek, ii. 2, p. 23.

¹⁶⁰ All names in Lycia containing θ are either hellenised or unquestionably foreign, except Παναθιατις. This should very probably be divided into Πανα Θιατις in which ease the woman, who may have been descended from Phrygian immigrants, bore a Lycian as well as a Phrygian name. Τατας and all the related names seem to be of Phrygian origin, as they are common in all parts of that country, but are of the greatest rarity in Lycia and Cilicia. None of the examples quoted by Kretschmer and Sundwall, where the place of discovery is given, are found in Lycia proper except Τατας ροῦς which in form is Greek (see p. 55, note 72), and Τατασιον of which the termination at least is hellenised. The native name tettm̃pe is correctly divided by Sundwall (p. 210) as te-ttm̃pe, and probably has no connexion with τατας

¹⁶¹ Tomaschek, ii. 2, p. 36.

¹⁶² Tatta, from Servia (Kretschmer, p. 348)—that is, Upper Moesia, is probably Dardanian.

¹⁶³ Pauli, iii. pp. 365, 370, 372, 374, 375.

¹⁶⁴ The derivation of this name from tentā (Tomaschek, , ii. 2)Google Scholar is considered certain by Kretschmer, p. 228. Compare the Lithuanian tanta, country. Philipon, , Les Ibères, p. 25Google Scholar, besides several names formed from taut- in Spain, quotes the Armenian proper name Tautukas, which may be taken as evidence with regard to the Phrygian form, if the relationship between those two languages is real (Kretschmer, p. 208).

¹⁶⁵ In that case the Isaurian Θουθ and Θαθ for Τουτ and Τατ would be analogous to the European-Dardanian, Thithi (nom. masc. Jahreshefte, iv. Beiblatt, p. 85)Google Scholar, as compared with the Lycaonian Τιττις(Kretschmer, p. 349), Illyrian Titio, etc.; see C.I.L. iii. index. Cf. also the Thracian Θιθι-σαττα (Tomaschek, ii, 2, p. 48) and Βρυ-θειεος (ibid. p. 37) with Tithutes and Nusa-tita.

¹⁶⁶ The Phrygian ‘youths,’ Τόττης and ᾿Οννης who brought the mysterious ίερά of the Cabiri to Assesus in a chest and introduced their worship into Miletus, , were evidently the two male Cabiri themselves, the son and father, F.H.G. iii. 388Google Scholar. ᾿Οννης, Νοννος and Νανας are the masculine forms of ᾿Αννα, Νοννα and Νάνα variant-names derived from the mother-goddess, ᾿Αρτεμις Νάνα (Kretschmer, p. 355). ᾿Οννης was her mate, the father-god. Τόττης like Τάττης is a variant of the name of her son, ᾿Αττυς who is identical with ᾿Αττης

The words αυτος κε ους κε ποκα γεγαριτμενος αιβαταν τευτους in a late Phrygian, epitaph (J.H.S. xxxi. p. 181Google Scholar; cf. p. 183) do not seem to be connected with the curse which precedes them. For γεγαριτμενος should represent the Greek κεχαρισμένος and if so, αυτος probably refers to the dead man, and the phrase either commends him to the favour, or describes him as the favourite, of some god. Τευτους in that case would be the son-god Τὁττης

¹⁶⁷ See p. 63.

¹⁶⁸ J.H.S. xxii. p. 118. The name may more probably be Τιει dative feminine from Τιϝις

¹⁶⁹ Strabo, p. 304.

¹⁷⁰ Kretschmer, p. 207.

¹⁷¹ These may be hellenised, to resemble the Greek Θὁας There is also a Lycian name tuwada, of which the stem is found in Ερμαν δοας and the Pisidian Νανι-τοας etc. The resemblance to the Paphlagonian and Illyrian names is, I believe, merely a coincidence. The Cilician Βαργαθοης must on the other hand be considered as a Phrygian name, not only because the first part is Phrygian, not Lycian (see above, p. 58), but also because the change of t to θ is foreign to Ciucia as well as to Lycia; no other Cilician name contains a θ, except the Phrygian Βιθυς (see p. 55).

¹⁷² Such as Θύμβρς Θεμίσων Θύρα ᾿Ιθαρύς Ορνίμθος etc., etc.

¹⁷³ Θαθεας Θαθοις Θαννις Θιης Θουας Θοων Θουθους Θουθου Ιμμοθις already discussed: Βχθθις (p. 59), Βαλαθθις (p. 56), Βαρατθα (p. 63). All these appear to be Phrygian names. Μαθουν and Γουλαθεις (if correct) may be Phrygian adaptations of Lycian names, but the evidence is insufficient. ΘήΕασα is probably hellenised on the model of Θῇβαι

¹⁷⁴ See p. 68.

¹⁷⁵ Βοσθαν (p. 60), Βερθας (ibid.), Θυσσος and the local names Θασθαρα Θεμβρια Thymnias, Thabusioii, and Συμμαιθος seem native. Θεμισσός Ορθονδουωκα and Θυησσός may be partly hellenised. Θεκυιλωνης is omitted.

¹⁷⁶ -Θιατις (p. 65) and Βαργαθοης (p 58). Βειθυς and Βιθυς should be added (p. 55).

¹⁷⁷ J.H.S, xxxi. pp. 161–215, Nos. xxxi., xlviii., and possibly lxv. The borrowed βαλαμει (No. iv.) goes far to prove that the sound was native, as in other languages θ in words taken from the Creek usually appears as t. The frequent substitution by native engravers of θ for τ in writing Greek (ibid. p. 211) suggests that they were accustomed to such a substitution in their own language. In Lycia, where there was no θ, such an alteration is, I believe, unknown; it is certainly most exceptional.

¹⁷⁸ The alphabet in this is not merely similar but identical, and unless the Phrygians obtained theirs from Lemnos, which is most unlikely, they must themselves have taken θ from the Greek parent-alphabet, presumably because they required it.

¹⁷⁹ As a test I have taken at random fifty names from J.H.S. xix., xxii., xxiv., and xxv., and B.C.H. x., which happened to be at hand. Of these seventeen are certainly or probably allied to the Lycian and twenty-three to the Phrygian: ten are doubtful.

¹⁸⁰ Λύστρα, Κύβιστρα and ᾿Ιλιστρα have the same ending as Λαπιστρα and Πλουριστρα in the region of the Phrygian Antiochia, Sostra, and Κάναστρα in Thrace, Βαλλανστρα in Upper Moesia or Dardaniu, and ᾿Αλιστρος in Illyria. Δέρβη is synonymous with Derba in Dalmatia, and probably with the Thraeian Ζαρβα and Zervae. Βάρατα has also an Illyrian name (p. 63). No local name has any special Lycian affinities, unless it be Κοροπασσός

¹⁸¹ See pp. 54 and 55.

¹⁸² As Βαναβα (note 95), Βολβοσος (note 92), and possibly Βακα (note 137). ᾿Ανα-ζαρβά seems to be compounded from the Thraeian Ζαρβα with the preposition ἀνἀ The older name Κύινδα is Lycian.

¹⁸³ P. 63.

¹⁸⁴ P. 54.

¹⁸⁵ The name is probably formed in the Phrygian way (p. 47) from a proper name Μεστος or Μεστιος (p. 61), slightly hellenised.

¹⁸⁶ Probably from ἀνἀ and βουρα, ‘house’; see Fick, p. 95.

¹⁸⁷ Ramsay, , Revut des Universités da Mudi, i. p. 356Google Scholar. Sundwall is, I believe, right in stating (p. 255) that the nominative both in masculine and feminine names ends in a vowel, and that the genitive ends in -s in both genders. In both these points Pisidian agrees with Lycian, for the Lycian, ‘genitive’ in -h represents an earlier -s (J.H.S. xxxv. p. 106)Google Scholar. If there is no grammatical gender, the agreement is more significant.

¹⁸⁸ The derivation from bherǧ is too intrinsically probable to be easily set aside; see p. 62, note 132.

¹⁸⁹ The list of nearly thirty names given by Ramsay, , Cities and Bishoprics, i., p. 314Google Scholar, contains six or seven which seem to be of Lycian origin. The remainder are, I believe, rightly claimed by him as Phrygian.

¹⁹⁰ In the inscriptions published by Sayce in S.B.A. Transactions, ix, and Proceedings, xvii., xxvii., and xxx., B only occurs in xxx., No. vii. (if this is Carian). A letter which is taken to be Β, is found twice, ix. i. 1 and ix. ii. 4, but in the same name, in which elsewhere it is replaced by the vowel (e.g. ix. i. 7). is found in ix. i. 7; xxvii. ii.(?); and xxx. i. But in all cases the writing is so irregular that exceptional forms are suspicious.

¹⁹¹ The Carian alphabet appears side by side with the Ionic at Abu-Simbel, already fully developed and so unlike any Greek alphabet as to indicate a separate evolution of some duration. It must therefore be derived from a very primitive form of the Greek.

¹⁹² The commonest name (occurring in various forms at least seven times), and one of the most legible, is M(e)sna., which resembles the Venetic (Illyrian, ) mesne, Pauli, iii. p. 327Google Scholar.

¹⁹³ Kretschmer, , p. 382. The theory of Sundwall that it is a guttural is, in my opinion, untenable; J.H.S. xxxv. p. 104Google Scholar.

¹⁹⁴ In this respect the Pisidian seems to agree with it (p. 68). The Lydian has a patronymic in -l-, quite unlike the Carian, and unknown in Lycian. It has also, I believe, an ethnic in -m-, equally foreign to Lyeian. But the subject of Lydian cannot at present be discussed, as the material is largely unpublished.

¹⁹⁵ If my contention is right that the Lycian h represents an earlier s, and that the change was still in progress about B.C. 300 (J.H.S. xxxv. p. 104), it would be surprising that h should appear in Carian at Abu-Simbel about 300 years earlier. Another difficulty is the great frequency of s in the native and Greek inscriptions.

¹⁹⁶ The letter which has the shape of θ is probably required for that sound, and in any case is not common enough for t, and the same may confidently be said of various consonants of unknown value.

¹⁹⁷ From proper names Μνιεσις (*mñn-ijesi, cf. mñn-uhe and aruwãt-ijesi) and *Γαρβεσις (*uw-erbbesï, cf. uw-ita and Carian ᾿Αρβησις). The names are Lycian, but the suffix -υτης is quite unknown in that language.

¹⁹⁸ The Carian letter ν may not be u. It might, for instance, be l.

¹⁹⁹ Strabo, p. 678, διὰ τὸ ἑτέροις γένεσι περι έχεσθαι ὡς Ιδριεῖς μὲν καὶ Τερμίλαι Καρσὶ Δολίονες δὲ καὶ Βέβρυκες Φρυξί

²⁰⁰ ᾿Απολλόδωρος ὁ θαυμασιώτατος St. Byz. s.v. ᾿Ωρικός He appears to have been also one of the sanest and most scientific of ancient critics, as might be expected of the pupil of Aristarcluis and follower of Eratosthenes.

²⁰¹ The Τέρμερα of Stephanus in Lycia is not an error, but a reference to Asclepiades of Myrlea, (F.H.G. iii. p. 300)Google Scholar, whom he quotes elsewhere (s. vr. Μύρλεια and Νίκαια). The myth probably refers to the foundation of Patara (cf. St. Byz. s.v. ᾿Απτερα).

²⁰² F.H.G. iv. p. 475.

²⁰³ In J.H.S. xvi. p. 207, the Τερμίλη of Stephanus is identified with a fort at Tremil. It is very probable that his tentative identification with Τέρμερα is wrong, but possibly the name of the district rather than a town may have survived at Tremil. The archaeology of the region is discussed in two valuable articles by Paton and Myres in J.H.S. xvi.

²⁰⁴ P. 611. In their earlier days they spread and multiplied greatly, ὔστερον δ᾿ ἄμα τοῖς Καρσὶ στρατευόμενοι κατεμερίσθησαν εἰς ὄλην τὴν ῾Ελλάδα λαὶ ἠφανίσθη τὸ γένος He certainly means that this dispersal took place in prehistoric times. Therefore, when he goes on to say that six of the eight towns were joined to Halicarnassus by Mausolus, he cannot imply, as is sometimes supposed, that they were then inhabited by a people bearing the name of Leleges. The whole story, which is a continuous narrative, seems to come from the Homeric commentary of the notoriously untrustworthy Callisthenes (Strabo, pp. 680, 814, etc.), who told a similar legend about the expulsion of the Homeric Cilices and their migration to Pamphylia (ibid. p. 667). Eratosthenes (Pliny, v. 30) described the Leleges as an extinct race, and it may be noted that Apollodorus does not mention them among the historical peoples of Asia Minor. If Herodotus had known of existing Leleges near his native town, distinct from the Carians, he could hardly have speculated on the relative accuracy of Cretan and Carian myths about their identity (i. 171). It is, in fact, evident that he considered them extinct as a people.

²⁰⁵ F.H.G. iv. p. 475.

²⁰⁶ The story about Leleges and Minyae who once existed as a degraded caste at Tralles (Plutarch, , Quaest. Gr. 46)Google Scholar represents them as originally invaders. It is frankly archaeological, like the legends about the foundation of Aphrodisias by Leleges (St. Byz. s.vv. Νινὀη and Μεγάλη πόλις), but may well contain elements of real tradition.

²⁰⁷ St. Byz. s.v.

²⁰⁸ Ibid.s.v. Ιδριάς The statements here given without the author's name are ascribed to Apollonius under Χρυσηορίς and Εὔρωμος

²⁰⁹ St. Byz. s.v. Μύλασα Tlie kinship with the Mylasians which was claimed by the Pisidians of Termessus (Kretschmer, p. 395) was probably based on a similar genealogy. The name of Termessus seems to be connected with trm̃mis, which is, I believe, the adjective corresponding to the substantive trm̃mili, ‘Lycian.’

²¹⁰ First published by M. Haussoullier, whose learning continues after the lapse of nearly forty years to throw light on these obscure studies.

²¹¹ Kretschmer, p. 391.

²¹² Xanthus Lydus, frag. 8.

²¹³ Littmann, , in Sardis, vol. vi. pt. i. p. 75Google Scholar.