^{page 100 note 1} The paper was intended for publication in the last Number of this Journal, and was completed in all essential features last September; it was delayed in order to see whether a second journey in Phrygia might modify or add to the views here expressed. The delay has enabled me to speak of the Maeander route from personal knowledge, to add inscrr. 4, 5, 6, 11, and 12, and to insert many corroborative details.

^{page 101 note 1} Anava (Herod, vii. 24) was situated at the modern village Sari Kavak, ‘Yellow Poplar,’ an hour east of Chardak, overlooking the lake whence people still get salt as they did in the time of Herodotus. Considerable traces of the city remain, including a curious old doorway of pre-Hellenic style.

^{page 101 note 2} In Caria 740 st., towns Ephesus, Magnesia, Tralles, Nysa, Antiocheia, Caroura; in Phrygia, 920 st., towns Laodiceia, Apameia, Metropolis, Chelidonion, Holmoi; in Phrygia, Paroreia rather more than 500 st., towns Philomelion, Tyriaion; in Lycaonia, 840 st., towns Laodiceia Katakekaumene, Koropassos; thence over Cappadocia by Garsaoura (120 st.), Soandos, Sadakora to Mazaka, 800st. (Strab. p. 663). Pliny refers to this road, ii. 172, v. 106, xvi. 240; distance from Mazaka to Ephesus ccccxv M. P. Xenophon gives the distances—Sardis to Apameia 50 parasangs (Sardis-Apameia = Ephesus-Apameia), Colossae to Apameia 20, Peltae 10, Keramon Agora 12, Kaustrou Pedion 30, Thymbrion 10, Tyriaion 10, Iconion 20, through Lycaonia 30, through Cappadocia to Dana 25; whence he crossed the mountains to Cilicia. The parasang is 30 short stadia of 480 feet measured by a bematistes (Hirschfeld, , Apameia-Celaenae, p. 8)Google Scholar.

^{page 101 note 3} When the Hermus valley railway was extended to Philadelpheia, trade from the interior was diverted to this route. Now when the Ottoman line is extended to the Lycus junction, trade will resume the old path.

^{page 102 note 1} I have no sympathy with the view that recognizes in the Phrygia of the Homeric poems merely the district beside the lake Ascania. Hiad iii. 186, and the phrase of the Hymn to Aphrodite, Φρυγíης εúτειχήτοιο, seem to me decisive against it.

^{page 102 note 2} No one who has travelled both will accuse me of overstating the difficulty of the Hermus route or the ease of the Lycus route.

^{page 103 note 1} Voyage Archéol. en Galatie, etc., p. 323 ff. Stein on Herod, i. 76 doubts that Boghaz-keui is Pteria. In that case one would simply have to read throughout the present article Boghaz-keui instead of Pteria; the reasoning is not dependent on the name. The ruins of Boghaz-keui show what its character was; according to Stein, Croesus must have passed by this chief seat of his enemy and attacked some place on the coast near the mouth of the Halys. But (1) it is in the highest degree improbable that a great Oriental city so situated should either have disappeared completely or escaped the notice of travellers; (2) Herodotus implies that Croesus attacked the strongest seat of the Oriental power; (3) it seems to me quite certain that Boghaz-keui was the chief seat of that power. Against these arguments it seems injudicious to press so much the force of κατά, even admitting the interpretation of Stein, which I think misses the true character of the preposition. Stephanus has the form Pterion, Herodotus Pterie.

^{page 104 note 1} The road between Sinope and Pteria probably went, like the modern path, due south by Boiwad. It traverses a very difficult country.

^{page 104 note 2} According to Herodotus, Xerxes crossed the Halys into Phrygia, but here he diverged south to Celaenae; if this account is correct, Xerxes followed the “Royal Eoad” for some time, but came round by Celaenae in order to avoid the Hermus rsoute, which was impracticable for an army.

^{page 105 note 1} Berl. Monatsb. 1857, p. 126 f.; Hirschfeld, , Apameia-Celaenae, p. 7Google Scholar (extr. fr. Berl. Abhandl. 1875).

^{page 105 note 2} It is now admitted that the eastern half of the Royal Road existed long before the Persian rule; see Stein on Herod, v. 52, p. 52.

^{page 106 note 1} The tomb of Mygdon, at Stectorion near Apameia, Paua. x. 27.

^{page 107 note 1} I hope hereafter to trace the southern road across the peninsula by Iconium, and to show that it also grew through the spread of religion and civilization, not through foreign conquest.

^{page 107 note 2} Herodotus distinguishes throughout those rivers that were crossed by boat, νηυσ) πɛρητóσ, and he tells us that Croesus had a bridge to cross the Halys (i. 75).

^{page 108 note 1} The remark was made with reference to the Roman period, when the principle of the arch was thoroughly developed. Close to the eastern hank there is a large artificial mound, such as are very common in Cappadocia and Lycaonia, and were called hy the ancients “mounds of Semiramis.” Strabo attributes to the Syrians the invention of roads, bridges, and artificial mounds, p. 736.

^{page 109 note 1} “Gordium, Pessinus, Sivri Hissar,” in Munch. Gel. Anz. 1861; Perrot, p. 154.

^{page 110 note 1} Journ. Hell. Stud. 1882, pt. i. “Inscrr. fr. Nacoleia.”

^{page 110 note 2} I have unfortunately no note of the breadth between the wheels. On the breadth of Greek waggon-roads see Mitth. Inst. Ath. iii. p. 29.

^{page 111 note 1} It perhaps passed through Bennisoa (Altuntash), with its priestly college, Benneitai, of Zeus Bennios. This supposition connects it with a religious centre, and coincides with a route marked by nature and still followed, by Ushak and Koula. It may also have gone through Konni, a still unknown site, which was certainly not very far from the line of the road. The name may be an old religious one, like Apollo Kunneios with his hereditary priests the Kunnidai at Athens. The most probable route however is by Ancyra, and thence down the Hermus by Kadoi to Coloe.

^{page 112 note 1} Ephesus takes the place which Miletus or any port on the Maeander valley would have held had their harhours remained open. In later time Ephesus also lost its harbour, and trade passed to Smyrna.

^{page 113 note 1} These notes are printed solely from the wish to call attention to a remarkable series of sculptures, which have as yet been almost completely neglected. In our hurried journey there was no opportunity of examining them sufficiently. Now Herr Hermann has been charged with the duty of bringing casts to the Berlin Museum, and there is every reason to hope that the sculptures will soon be accessible to study.

^{page 113 note 2} This view was not suggested by a preconceived theory; in reality it suggested ideas which have gradually led to the general theory of early history in Asia Minor here and elsewhere expressed.

^{page 114 note 1} Perrot, , Voyage Archéologique, pI. 52Google Scholar.

^{page 114 note 2} I must here assume unproved that theory of the character of Phrygian religion which seems required by the facts of its history.

^{page 115 note 1} Perrot, pl. 42, 47, 50, 61, 56. Euyuk is five hours north of Pteria. Here, out of the side of one of the large artificial “mounds of Semiramis,” appear the doorway and front, covered with sculptures, of some great palace or temple.

^{page 115 note 2} p. 337.

^{page 115 note 3} I must however add that at Ibriz both the husbandman-god and his bearded priest wear earrings. Lydian men wore earrings (Xen, . Anab. iii. 1, 31)Google Scholar.

^{page 118 note 1} One who looks at the plates in Perrot, 50 and 51, will at once say that I am wrong on this point, and that the fignre is certainly female. But before judging, one should bear in mind that the photograph on pl. 51 is useless, and that the drawing on pl. 50, being made by one who thought the figure male, loses all the feminine character.

^{page 118 note 2} See Inscr. published by Mordtmann, “Gordium, Pessinus, and Sivrihissar,” Munch. Gel. Anz. 1862.

^{page 118 note 3} Str. p. 557: Curtius on Ephesian history, Beitr. Gesch. Kleinas.

^{page 120 note 1} Steuart, Ancient Monuments of Lydia and Phrygia; Mordtmann, , Sitsungsb. Bair. Akad. 1862, p. 35Google Scholar; Texier, Asie Mineure, As the Phrygian alphabet does not distinguish long and short vowels, the inscriptions cannot be transliterated by the Greek symbols; I have therefore used the Roman character to transcribe them. I shall often refer to M. Schmidt's remarks on these inscriptions (Neue Lyk. Stud. 136), and to Fick's discussion of the Phrygian glosses in the last chapter of his Ehemalige Spracheinheit. I may add that in every discrepancy between my reading and the published copies, the reader may understand that I specially compared the older copy with the stone. I made my own copies of 1, 2, 3, 7, 8, 9, compared them with the older copies, and then compared each with the stone. Time failed me in the case of 4, 5, 6.

^{page 121 note 1} The form for lambda occurs in an archaic inscription of Arcesine in Amorgos, a Milesian colony (Bull. Corr. Hell. 1882, p. 187): and were used for lambda in Argos.

^{page 121 note 2} The second form may possibly be more complicated, as the stone is worn.

^{page 122 note 1} Num. Chron. viii. p. 59: Annali, 1861, p. 149.

^{page 122 note 2} Even those which are later go back nearly to the sixth century: see Fränkel, , Arch. Ztg. 1879, p. 84Google Scholar.

^{page 123 note 1} I do not think that the inscription from Euyuk in symbols partly Greek given by Hamilton, Travels, i. p. 329, is to be relied on for the forms of the letters. Hamilton is an accurate observer in most respects, but he had not realized the value of such details, as is evident from his Greek and Latin inscriptions.

^{page 125 note 1} Its trade with Miletus is probably alluded to by Hipponax, fr. 36, Bergk.

^{page 125 note 2} The ideas here stated are exactly those at which arrives, Matzat, “Herodotot's Angaben über Asien,” in Hermes, vi. pp. 392–486Google Scholar.

^{page 126 note 1} Sinope was a colony of Miletus.

^{page 128 note 1} It does not read upwards, as Leake says: a mistake onLeake's part is hardly known, but I had his book in my hand before the Tomb.

^{page 130 note 1} If we consider the situation of these two inscriptions, one on the carved work of the Midas-tomb, the other facing the spectator as he ascends a flight of steps surrounded by rock-sculptures and rock-altars, the possibility suggests itselfv that both refer to the construction of the monuments around: then the inscription over the tomb of Midas (No. 1) is the dedicatory formula, that at the side is the record of the building—“the artist's signature.” Siheneman is then the designation of the tomb; akaralasun of the place where most of the interesting rock-monuments of the city are collected. No. 5 is carved on a panel, which has been left simply to strike the eye of the spectator; it has no connexion either with an altar or a tomb, but stands upright and isolated at the side of the steps. I need hardly add that this is a suggestion to which I attach no special value.

^{page 131 note 1} I had his copy before me, and compared it with the stone to verify this point.

^{page 132 note 1} Texier's plates make it easier to understand the character of this tomb and of the Midas tomb.

^{page 134 note 1} Mordtmann saw one of these columns.

^{page 135 note 1} The Lydian language had disappeared in Lydia before the time of Strabo, though still used in the remote district of Cibyra (p. 631).

^{page 135 note 2} Rendered in Greek letters of course σíμουν.

^{page 135 note 3} As Prof. Sayce suggested.