^{page 172 note 2} In answer to queries it is here mentioned that Antiocheia is the Greek form of the name, while Antiochea is always used in Latin inscriptions.

^{page 172 note 3} On this road see J.H.S. 1920, 89 ff.

^{page 172 note 4} J.R.S. vi, p. 94. It is possible that there were miscellaneous veterans (besides those of leg. V Gallica), who are not known to us except two of leg. VII.

^{page 173 note 5} Traces of the Phrygian language are most scanty, and none have been found in the native sanctuary above Antioch on the hill of Kara-Kuyu.

^{page 173 note 6} See my Cities of St. Paul, pp. 169 ff.

^{page 174 note 1} See C.R. 1919, p. 1 ff, ‘A noble Christian Family of the fourth century.’

^{page 174 note 2} See the important inscription found at Laodicea of Lycaonia by Calder, and published by him in Klio.

^{page 174 note 3} J.R.S. viii, 117 ff.

^{page 174 note 4} The date is determined by the coins, which cease then.

^{page 175 note 1} The water conveyed by this aqueduct is remarkably pure and good. It is conveyed underground for some miles; but near the city the conduit is on the top of arches.

^{page 176 note 1} They belong to Antioch unless a different origin is stated.

^{page 176 note 2} I call these the Scalae Augustae, as a name is needed.

^{page 178 note 1} J.R.S. viii, p. 141.

^{page 178 note 2} Owing to natural character a bountiful harvest must be universal in the valley, and not confined to one estate.

^{page 179 note 1} Anderson, in J.R.S. iii, 267Google Scholar f.

^{page 179 note 2} Restorations like [ex agri su]i messe are too short: something similar to [abundant]i is needed.

^{page 179 note 3} J.H.S. 1918, p. 174; J.R.S. vi. 134.

^{page 179 note 4} I prefer superabundanti as most probable; but the meaning is certain. Some might date the inscription about A.D. 50; and that was my original impression.

^{page 182 note 1} The two Plateae, Aug. and Tib., were the glory of the city. Between them were the Scalae Aug.

^{page 182 note 2} A detachment fought at Bedriacum on the side of Vitellius in 69.

^{page 183 note 1} A poppy is carved on an altar of Hermes: the native name of opium was papa; papaver (cp. cadaver) is of Anatolian origin. Pappa meant father.

^{page 184 note 1} This would explain why A it shorter than B.

^{page 185 note 1} Also C.I.L. vi, 1410, 312; xiv, 251, Bücheler, , Carm. Epigraph. ii, 868Google Scholar; see also iii, 4120, 4617, 11323, 4638, 4642, 14203^8.9; vi, 2003; xv, 74475; I.G.R.I. 138 De Rossi Inscr. Chr. U.R.I, 4; Ath. Mit. xxiii, 1898, 165Google Scholar f., Digest, i, 12, 1, 15, 5; xlviii, 19, 8, 5; 22, 6, 1; C.I.L. xiii (2), 5330; xv, 7447: also Hist. Aug. Comm. 17, 4 and 20, 1: Carac. 3, 2; 4, 5, etc. (taken from P. W.).

^{page 186 note 1} There is much irregularity in the titulature of Severus, but the want of his oriental titles favours 195.

^{page 187 note 1} Probably the error Augg. in 1142 in respect of Galatia is due to the right use of Augg. in respect of Pannonia.

^{page 188 note 1} Memmia Paulla was perhaps daughter of Oppia and Memmius: the oldest daughter takes her mother's name, perhaps because Oppia had the money. The younger takes her name from her father, L. Memmius Paullus.

^{page 188 note 2} See Dessau, in J.R.S. III, 302Google Scholar.

^{page 189 note 1} The spelling is unusual: the Greek has ου, Latin u.

^{page 190 note 1} Divined by Professor Knox McElderry, accepted by Dessau in his Index and proved by an inscription of 1914, which gives the name Caparcot. more fully (J.R.S. vi, 130).

^{page 190 note 2} Vettius, L. Fulvius Rusticus Secundus in P.I.R. ii, p. 98Google Scholar.

^{page 191 note 1} In Britain honours are sometimes obtained by money; but the money goes to party funds or other objects and does not in any way lessen the burden of taxation.

^{page 192 note 1} Probably 92: often mentioned in Martial vii and viii, dated Dec. 92 and 93. In ix, 101, 17 f. (published A.D. 94), he speaks of three Suebian and Sarmatic expeditions. Suetonius speaks of two Dacian wars.

^{page 192 note 2} I omit Syria on the south-eastern frontier, which was, of course, a highly important sphere of duty, connected more with the Arabs and the Parthians, who were always pressing more or less on the civilized and orderly Roman Empire.

^{page 192 note 3} I am not convinced by Ritterling's theory that about 93 or 94, Domitian disjoined Galatia and Cappadocia, leaving Pontus and Armenia to Galatia. Pomponius Bassus 95-101 was governor of Galatia and Cappadocia, Antistius Rusticus 91-93 (died in office). Perhaps Sospes fell between them (as Ritterling says). Capp. is probably omitted by the engraver. These lists are never complete.

^{page 192 note 4} Praef. fr. dandi ex. s.c. hardly occurs between Claudius and Trajan (Anderson); but it occurs before and after.

^{page 200 note 1} The west side of Karalis may be deep but the mountains rise steep from the water, and no road or villages exist, as we were informed when I thought of exploring that side.

^{page 200 note 2} Perhaps Pamphylia extended up to the SW. bank of Lake Karalis in Byzantine time; and Karallia might then be near Ueskeles or Parisbeleni, two neighbouring villages with a fortified site between them. The ancient topography is very uncertain and difficult.

^{page 201 note 1} I have a note that the Preliminary Survey made by the German engineers differs slightly from the map, as they finally plotted it at the conclusion of the enterprise; but I cannot state details, as my copy of the map was lost (or taken away) on board ship going to New York in 1921.

^{page 202 note 1} Personally, I thought TI, from the spacing; and I think that Anderson had the same opinion.

^{page 202 note 2} Some such suspected names in Ptolemy are guaranteed by one other solitary allusion: e.g. Talbonda is once mentioned as a bishopric identical with Tumandos (H.G.A.M. p. 419, where κομμακον is supposed to be an error for Comama Col.). It is strange to find so many similar names in this region, Conana, Comama, Commacum; and frequently confusion amomg them is caused to modern scholars.

^{page 202 note 3} Also J.R.S. vi, 96, where by a slip cohortes is used instead of alae. On Ala VII Phrygum, see Cheesman, Auxilia, p. 61.

^{page 202 note 4} In the last two the title is Germanica, showing that its origin and meaning had already been for gotten, and that it was connected with Germany, not with Germanicus (as was correct). The title was probably given to Ala Antioch.