^{page 102 note 1} TW is often spelt TB in the Roman period.

^{page 102 note 2} See Paribeni in Annuario della R. Sc. Arch. di Atene; the city lies nearly on a direct route from Pambuk-Ova to Selge, difficult and winding round mountains. It is about 3 hours ESE. from Djandir in Pambuk-Ova, 2 hours ENE. from Kizilli-Keui, 5 hours nearly due E. from Milli, 4 hours WNW. from Yenije on the Kestros. The estimates are rough guesses, taken from Paribeni's map.

^{page 102 note 3} Idris, ‘educated,’ as Sayce tells me: Sari (as I first spelt it), means ‘yellow,’ which seems unsuitable here: probably Sarai (minsion) is the true form. With Idris connect Medresse, school of religion and religious law.

^{page 103 note 1} Anderson places the head of the Eurymedon in Yilan-Ova. The river traverses Yilan-Ova, but does not rise there.

^{page 103 note 2} On the Pisidian rustic language, see Ramsay in Revue des Universités du Midi, 1895.

^{page 103 note 3} Amblada and Amilanda both occur in Byzantine authorities.

^{page 103 note 4} Tembris in Wadd. Cat. Tembros in B.M. Cat.

^{page 104 note 1} J.H.S. 1887, p. 504; on the variation between A and O in Anatolian names, cf. Hist. Geog. Asia Minor, p. 189, etc.

^{page 104 note 2} See below, p. 107, n. 2.

^{page 104 note 3} The names Thymbra, Thymbrara, Thymbres, Thymbrion, Thymbrios, point back to old Anatolian nomenclature, in the Troad, Lydia, Ionia: perhaps also θύνναρος, a heroised ancestor of the θυνναρίδαι, a section of the people of Synnada, Hist. Geog. Asia Minor. pp. 14, 36, also the writer's more detailed article in Strena Buliciatia, p. 662 ff. Hill rghtly distinguishes Timbriada from the old Thymbrion (which, however, was probably not in Lycaonia).

^{page 105 note 1} This road is proved by the many milestones and the boundary pillar to be the great route. The Via Sebaste certainly passed this way to Olbasa, Comarna and Cremna, though none of its milliaria have been found except at Comama; but there are many late milestones.

^{page 106 note 1} Ἀντιόχειαν τὴν Πισιδίαν, the ‘Phrygian city towards Pisidia’ (as Strabo defines it).

^{page 107 note 1} (1) The map is made and published at the special request of Professor H. Dessau.

(2) The course of the river Anthios, which flows through the gorge south of Antioch, and then is divided into a score of channels which run down the streets and in at least one drain under a street, keeping the town clean, cannot be indicated on such a sketch map, but one channel with four bridges is marked, being the most conspicuous.

^{page 107 note 2} On Oa, modern Ova, ancient OWA, see Hist. Geog. Asia Minor, p. 143: A village north of Afiom-Kara-Hissar, was Ὀηλέοντος, literally Λέοντος κώμη, interpreted as Γοὴ έοντος, wailing of Leo Phokas, who was caught here and blinded. Γ for W was rustic pronunciation. See also Hesych: ὠβάι, ὠάς, ὠγή, ὀά, οὐαί, οἰατᾶν.

^{page 109 note 1} A good example of the apparently purposeless variation between L and N; compare above, p. 104, Enevre and Elevre. The natives hardly notice the difference. Sayce divides the languages of Asia Minor into the N-languages and the L-languages.

^{page 110 note 1} It should be added that my authority and map omit the continuations of this street past the fountain, and onwards to the north and to the vineyards. The fountain is some way west of the barracks, marked as a large square building, and separated from it by graveyards.

^{page 110 note 2} Professor Robinson has not stated which (if any) of the photographs and enlargements that he publishes were made by Mr. Swain of the University of Michigan, an excellent photographer who accompanied the Expedition

^{page 110 note 3} In criticising my article, especially the reproductions, Professor Robinson has hardly taken sufficient account of the inadequacy of type, to which I refer in my commentary, for the purpose of giving an accurate epigraphic reproduction: cf. the editorial note on p. 258.

^{page 111 note 1} These attacks are mentioned by Xenophon, Anabasis i, 2, 1; they were a continual danger.

^{page 111 note 2} I have described in J.R.S. xv the extent of Pisidia, which is known from the enumeration of its cities by Artemidorus c. 100 B.C.

^{page 111 note 3} As it is, he gives no reasons, but merely states bis opinion, ex cathedra.

^{page 111 note 4} Pliny, , N.H. v, 95Google Scholar, who mentions only Colonia Caesarea, uses an authority older than tbe map of Agrippa, but later than 20 B.C.

^{page 113 note 1} Use of a pickaxe would leave fragments such as were found, but would not account for their disappearance. Robinson maintains that all surviving pieces have been found and I agree that this is probably so. A great fire would lick up and destroy many fragments and account for the facts. Many of those which were found were in places sheltered from the worst fire.

^{page 113 note 2} The Greek has τριῶν ἀνδρῶν ἐγενόμην κτλ.: duumviru[m] is gen. pl. The Mon. Antioch. has this fragment almost complete, with reading certain.

^{page 113 note 3} I ascribed its discovery to Mr. Peterson, as he and Professor Kelsey, sitting in my house in 1924, agreed was the case.

^{page 114 note 1} It was found north of Egerdir, on the west bank of the Lake. What Kah Yazla means is unknown to me. The ruins I would call S ljuk. Egerdir was captured by Tamer ane in 1402 on his march from Smyrna to Konia.

^{page 114 note 2} Egerdir (Akrôtêri) is about fifteen hours distant from Antioch, and is separated from it by the territories of Ouramna and Timbriada, and the mountains which are crossed by the Pass of Demir-Kapu.

^{page 114 note 3} The Greek text of Res Gestae D. Aug. at Ancyra and at Apollonia is a real translation from a Latin original.

^{page 114 note 4} I omit others who were with me in those unsuccesful attempts: they probably took no further interest in an apparently hopeless inscription.

^{page 114 note 5} Professor G. A. Harrer has discussed column i in excellent style in A.J.A., vol. xxix (1925), 429 ff.Google Scholar, correcting Robinson's errors.

^{page 115 note 1} There would have been more errors in it had it not been for Professor Stuart Jones's great care, for which I am profoundly thankful to him. I had to telegraph to him about the proofs from Naples in June or July, 1925, and forgot in my haste to sign the telegram, which reached him, but was difficult to comprehend.

^{page 116 note 1} Chapot, Prov. d'Asie does not admit him (as Wadd. does, no. 108) as proconsul on the authority of a much worn coin of Domitian. I found at Antioch in 1926 a broken inscription:

≡VSINIO

≣VEL·RVFO

^{page 116 note 2} I have stated some facts about famines in some parts of Asia Minor in the second series of my Gifford Lectures, now in the press. Further, a famine in the low Asian ground near the sea is unlikely to extend to the central plateau, and vice versa.

^{page 116 note 3} Rostovtzeff's quotation from Revelations is apt; but there is moie to say, which I have indicated elsewhere.

^{page 116 note 4} This late winter weather on the plateau was not common to the lowlands of Asia, where mildness prevailid.

^{page 116 note 5} The storks, which come in late March to the plateau, found nothing to eat and died in numbers; 70 per cent, or 80 per cent, of the young lambs died. I heard both estimates. This was in 1908; in 1909 the storks delayed their arrival six weeks.

^{page 117 note 1} I have often in conversation called white limestone by the title marble: this is the result of asking the Anatolians frequently about mermertash, marble stone; but I have avoided the wrong term in print except in one case in my article.

^{page 117 note 2} The legati usually remained in office several years, often three, in one case five or six; but there is not sufficient evidence to make the dates precise and full. I give only dates that seem certain.

^{page 117 note 3} There is no proof that Sergius Paullus was a consular, but the joint province Capp. Gal. was not so large then as later; and praetorian legati, were perhaps not sent at that time. To replace him Nonius Calpurnius Asprenas was sent by Galba from Pamphylia to take over the duty, which implies his death. Sergius is called filius, which is an imitation of the Greek νέος or νεώτερος, used in our sense of ‘junior.’

^{page 118 note 1} Mommsen had never seen any copy of mine, as I did not go to Sidgilli until 1912. He depended on Sterrett.