^{page 229 note 1} On the spelling see p. 264. The common spelling Homon- is taken from Strabo, and Tacitus, Ann. iii, 48Google Scholar and the older published texts of Pliny v. 94 (based on conjectural emendations).

^{page 229 note 2} The premature death of Lieut. G. L. Cheesman at the Dardanelles was a grave loss to Roman historical study and to the investigation of Anatolian antiquities.

^{page 230 note 1} Chiefly Mommsen, Res Gestae D.Aug. pp. 161 ff, Cheesman, J.R.S. 1913, p. 252 f; also the numerous disputants about Quirinius, governor of Syria, Luke ii, incidentally mention the war.

^{page 230 note 2} I call the province sometimes by the Roman name Galatia, sometimes according to the oldest Greek usage Galatic (C.I.G. iii, 4001, A.D. 54, and. Acts of Apostles, xvi, 6, xviii, 23, about A.D. 51–55).

^{page 231 note 1} Syria, of course, was always a consular province, governed by a legatus Augusti pro praetore of consular rank: Galatia was a praetorian province, under a legatus Augusti pro praetore of praetorian rank.

^{page 231 note 2} Mommsen was naturally inclined to place a considerable interval between the consulship and the governorship: that keeps the cursus bonorum of Quirinius closer to the usual order and sequence. It is the rare exceptional cases that cause difficulty.

^{page 232 note 1} On this ‘Debatable Land’ see J.H.S. 1918, pp. 142-152.

^{page 232 note 2} This must not be exaggerated: the south Phrygians continued (except in the cities) to be organised after the village fashion of Anatolia, as the subjects of the god Mên-Mannes (J.H.S. 1918, pp. 148 ff), not as free men, and to use the Phrygian language, until well on in the Roman period. Yet, compared to the Pisidian robber tribes, they were more affected by new ‘modern’ fashion, which tended to be Hellenising.

^{page 233 note 1} It is true that in xii, 7, 2, Strabo quotes the term Poleis from Artemidoros, who was certainly not so precise in using this term with reference to organisation; but in §3 Strabo himself describes Selge distinctly as a Polis, and distinguishes it from the mountain Pisidians who are ruled by a series of tyrannoi (tribal chiefs?) and who are all brigands. That Selge was founded by Lacedae-monians, as Strabo mentions, is based on some Hellenistic invention of origin. More trustworthy is the older foundation by Kalchas, which he also mentions. The latter points to an Old-Ionian origin. Strabo also calls Sagalassos (Selgêssos) a polis: so that Hellenistic system was spread in his time over parts of Pisidia, and the number of autonomous poleis of that type steadily increased in the Roman period.

^{page 233 note 2} I shall not easily forget this impression. Coming from north-east, I felt giddy as if I were suddenly looking down a perpendicular precipice; and riding down the steep descent I felt as if horse and rider were likely to fall off. There was, of course, not the smallest danger, unless a young horse takes fright (as happened to me in 1913).

^{page 234 note 1} It is called Dipoiras in Kiepert's map, but the natives whom I questioned did not know this name; poiras is the common term for ‘north’ (evidently Boreas in Turkish pronunciation), and Dipoiras is probably due to popular etymology. Geological expert study of Taurus is much needed.

^{page 234 note 2} Apart from volcanic action, the chief agent in producing these cañons and ridges in the limestone plateau seems to be flowing water.

^{page 234 note 3} Something has been done by Sterrett, and, as it is believed, by a very well-equipped Austrian expedition, which travelled through part of the country in 1902, but which has never published any account of its results and explorations (except a small Vorläufiger Bericht). See the detailed account beiow in § 4.

^{page 234 note 4} Like the similar testament of the last Pergamenian king Attalus III.

^{page 234 note 5} [Public law is definite. Vassal kings were vassal, entirely dependent, and in fide populi Romani, i.e. dediti. The emperor appointed the vassal king, and fixed the terms, and terminated the appointment as he chose. Such had been the legal position (allowing for change from republican forms) since the second century B.C. In public law there was no question of accepting an inheritance: Augustus took back what had belonged to him, but had been handed over for a time to a more suitable person to look after: Anderson: see his illuminative article in J.H.S. 1910, p. 181 f]. The ‘inheritance’ was, in a sense, ‘polite camouflage’; but, as I think, importance was attached by Augustus to those religious fictions, and the bare naked facts were glossed over by such means. That was a feature of his policy, alike in Rome and in the East.

^{page 235 note 1} The succession of Herod the Great's kingdom was subject to the sanction of Augustus: see Josephus, Ant. xvii, 3, 2Google Scholar (§ 53), and 8, 2 (§ 195). This shows; how strictly Augustus provided for the succession in the client kingdoms; see also Christ Born in Bethlehem, p. 184. Practically the choice open to Augustus was to rule Galatia himself, or to hand it over to some new client king.

^{page 235 note 2} [L. Calpurnius Piso, cos. 15 B.C., the able soldier who conquered Thrace, 13–11 B.C., was called in from Pamphylia: Dio says ΠαμΦυλἰας ἤρχϵ. Was he, as usually supposed, governor of Syria? In that case his presence in Pamphylia might be connected with preparations for the Homanadensian war, interrupted by the Thracian crisis and resumed by Quirinius. A man of his rank could not be mere governor of Pamphylia, as Klebs makes him Prosop. 249. The relation of Piso to the proposed war would be equally probable, even if he were not governor of Syria, but sent on a special mission: Anderson], This is convincing. The war was imminent in las 13 B.C. Piso planned to attack from Pamphylia, which was a bad line.

^{page 236 note 1} Obscurissima domus, Tac. Ann. iii, 23Google Scholar.

^{page 236 note 2} The same difficulties confronted Quirinius in ancient and the Italians in modern time, viz. the desert, the want of water, the warlike, predatory character of the inhabitants, and their skill in sudden attack and in eluding pursuit.

^{page 236 note 3} See his commentary on Monumentum Ancyranum, 2nd edition, p. 170 f.

^{page 236 note 4} A special expedition with a fresh army transported across the sea, such as that of Pompey in the East, is wholly contrary to imperial policy. The Provincial armies were there for all wars. A special command would not have been granted by Augustus to one of such humble standing.

^{page 236 note 5} Beyond a doubt the governor of the nearest province was charged to keep an eye on each client state, though he did not interfere except in urgent need (ad took place with the Kietai, falsely Clitae in modern texts and writers, Tac. Ann. vi, 41, xii, 55Google Scholar: they were subject to the client king Archelaos.)

^{page 237 note 1} See Christ Born in Bethlehem, pp. 105, 119, 122, 174, 184.

^{page 237 note 2} Practically universal opinion among historical scholars follows the lucid and convincing proof given by Mommsen (after Borghesi, Henzen, etc.) an his commentary on the Mon. Anc. p. 170 f. that the famous Tiburtine fragment, recording parts of the career of a Roman officer whose name is lost, belongs to Sulpicius Quirinius and was probably placed on his tomb. This officer twice governed the province of Syria. Zumpt is the champion of a different view, that the fragment speaks of Sentius Saturninus, governor of Syria, 8-6 B.C.

^{page 237 note 3} Mommsen's opinion is dictated rather by accommodation of Quirinius's career to his chronology (which is certainly wrong in several points), than by the evidence, which is slender. If it had been clear and convincing, Zumpt would not have put forward his view. We accept Mommsen in the following pages.

^{page 237 note 4} The milestones have all been discovered since 1886. Colonia Lustra was unknown until 1884, when Sterrett found the name and the site, after Waddington had published the first known colonial coin. Colonia Comama was not even a name, much less recognised as a colonia, until 1886.

^{page 238 note 1} It is remarkable that these reads were not called by a Latin word, but by a Greek word written in Latin characters: the road was a via sebaste. In the use of the Greek word one may justly see a recognition of the fact that, in the East, Latin and Greek were conjoined as the officially recognised languages, and that the imperial policy regarded the Greek civilisation and system and law as allied with it in the task of subduing the barbarism of the East (see the remarks in J.H.S. 1918, p. 143 f, based on the settlement made on the Pisidian frontier by the first governor of the Galatic province). It is also an interesting fact that this name for these roads lasted in popular use at least as late as the middle of the second century, because it is used in the more literary form Βασιλική in the legend of Paul and Thekla, which was committed, to writing some time shortly after 150-160 A.D. The literary Greek called the Roman emperors Βασιλϵῖς, scorning such barbarous terms as αὐτοκράτορϵς or Αὐγοῦστοι, or Σϵβαστοί.

^{page 238 note 2} C.I.L. iii, pp. 406, 407.

^{page 238 note 3} Ισαυροπαλαϵίτης an inhabitant of Isaura Palaia: Hamilton, no. 438, C.I.G. iii, 4393.

^{page 238 note 4} The existence of these roads was unknown to Mommsen. The first milestone (C.I.L. iii, 6974) was found on the site of the hitherto unknown Colonia Comama in June 1886, and it alone was sufficient to make him recognise that the fabric of his chronology seemed to need reconsideration.

^{page 239 note 1} The road system to the west of Antioch extended at least 122 Roman miles, as is attested by a milestone on the site of Comama. Several milestones found east of Antioch are numbered round about forty-five, but the road extended to Lystra, certainly about fifty miles further on. The weakness of such a defensive roadway, running partly through the foothills and partly through the level country on the north side of the mountains held by the Homanadeis and unsympathetic Pisidian tribes generally, needs no emphasis. It is true that part of the road immediately on each side of Antioch and Apollonia was safe, being too remote from the actively hostile mountain tribes, if we assume that the Via Sebaste went by Apoltended Ionia and not by Egerdir (see § 3); but all the rest was close to and under the mountains.

^{page 239 note 2} Strabo, p. 579, says that Amyntas captured many of their fortified posts (χωρία, Pliny's castella) and slew their tyrannos. This tyrant had seized (?) the office of priest-king, although there is some special meaning in the term, as distinguished from the dynastes or priest-king of Olba, Kennatis and Lalassis. [Dynastes was a technical term, used by Strabo in connexion with Thrace according to strict official usage, as inscriptions show; dynastes implied Roman recognition, but tyrannos connoted usurpation: Anderson.]

^{page 240 note 1} The impression left by this war is mirrored in the pages of Strabo (pp. 569 ff).

^{page 240 note 2} See Cheesman loc. sit. p. 257 f.

^{page 240 note 3} It is to be understood that, although Corbulo had superior authority, e.g. in Galatia, yet there was a legatus praetorius looking after the ordinary business of the province in the usual fashion. Cappadocia at that time was administered by the emperor's own procurator.

^{page 240 note 4} While Nero granted such over-governorship to Corbulo under the pressure of extreme danger, his jealousy and apprehension were roused, and Corbulo paid the penalty of his pre-eminence.

^{page 241 note 1} Two ladies, Miss Buxton and Miss Jebb, travelled over Anatolia in winter, though warned that travelling at that season was practically impossible.

^{page 241 note 2} See Diod. xix. 69, 2. A German writer puts the event on Nemrun pass, but without reason. The Cilician Gates was the the route for an army.

^{page 242 note 1} [Tac. Ann. xiii, 35Google Scholar, describes the rigorous measures adopted by Corbulo to harden his troops in Armenia: Anderson.]

^{page 242 note 2} 25 B.C. was the official date of the organisation of the Galatic province and of the great military colony, yet considering the distance, the time needed in conveying the news (even by a pinnatus nuntius), and the necessary preparations, 24 B.C. is the earliest possible date for actual work, if Amyntas was killed in 25; for his death must have taken place in summer, when alone warfare was possible among the Homanadensian mountains. If, however, 26 was the year of defeat, then 25 was perhaps recorded as the date of actual organisation of the province. Most probably the province was considered to begin from the death of Amyntas without any interval in 25 B.C., though permanent occupation could not begin until the year following. On the rate of imperial messengers see my article on ‘Roads and Travel in New Testament Times’ (Hastings, , Dict. Bib. v, p. 386Google Scholar). Naturally the arrangement of a military colony would require time.

^{page 243 note 1} Decreto decurionum publice [in formal language, as is stated in the following words of the inscription: Anderson.]

^{page 243 note 2} When an emperor accepted the honorary magistracy, he had no colleague, and his praefectus was sole supreme magistrate for the year. When any other than the emperor was appointed a colleague was elected in the usual fashion.

^{page 244 note 1} C.I.L. iii, 605.

^{page 244 note 2} C.I.L. iii, 6809.

^{page 244 note 3} C.I.L. ii, 3417, and Dessau, Inscr. Lat. Sel. 840 (who cites other evidence).

^{page 244 note 4} C.I.L. iii, 14712.

^{page 244 note 5} ‘Ti. Statilius Severus (otherwise unknown) is named on an Italian inscription not earlier than the second century A.D. as having a praefectus. But it is not clear that he was a duumvir, or that the praefectus was a deputy to a duumvir (C.I.L. x, 3910).’

^{page 244 note 6} Historical record, geographical facts, and commercial advantages, agree as to this. At the present day the interests of Spain in Mauretania are admitted, and constitute an international difficulty in respect of the French position.

^{page 245 note 1} Cheesman, indeed, suggests that Quirinius and Servilius were elected for successive years; but this would spoil the quasi-dictatorship, which (as he rightly sees) was held by C. Caristanius.

^{page 245 note 2} The emperor as duumvir and his praefectus stood alone.

^{page 245 note 3} This succession of great names as honorary duumvirs would even in itself alone sufficiently prove that Antioch in the first fifty years of its history was a place of imperial consequence.

^{page 246 note 1} Drusus cannot have been elected duumvir for 9 and 8 B.C.; it is indeed true that his death on 14th September would be unknown in Antioch at the time of election in 9 B.C.; but the inscription would in that case naturally contain some reference to his death, which would be known in Antioch by 1st January, when his praefect entered office. C.I.L. iii, 6843.

^{page 247 note 1} Cheesman also compares Tacitus, , Ann. iv, 46, xii, 16Google Scholar.

^{page 247 note 2} It was essentially obscure to me till 1909, when I had the opportunity of visiting lake Trogitis and traversing the Aulon or cañon from it to the plain of Konia. Yet I had been learning by hearsay and reading much about the lake, and had seen the neighbouring regions north and east.

^{page 247 note 3} The suggestion (made by Jüthner, etc. in their Prelim, account of their journey of 1902) that it was the valley of Kembos Göl cannot be sustained.

^{page 248 note 1} Wild and fanciful tales are current, both among the old Arab geographers and in modern talk, about localities and events in Anatolia. There is always some real fact of remarkable character behind the tales.

^{page 248 note 2} This scheme was mentioned by Said Pasha (Ingleez Said) at Konia in 1882 in my hearing; he had instructed engineers to take the levels, which, proved the possibility of the enterprise.

^{page 248 note 3} It is made in Pisidia-Lycaonia, p. 265.

^{page 249 note 1} εἰσὶ δὲ καὶ λίμναι … ἐνταῦθα δέ πον καὶ τὸ ’Ικόνιον ἔστι. Contrast this with the account of Savatra (Soatra) which is that of one who had seen what he tells.

^{page 250 note 1} Calder suggests Kabuklu, I follow St. (perhaps Kabaklu).

^{page 251 note 1} [What about the northward extension of Kalikia in Herodotus? Any connexion? Anderson.] Kilikia was one of the ten Cappadocian strategiai.

^{page 252 note 1} Vasada is mentioned by Ptolemy, v, 4, 9, who probably got it from Agrippa's lists. It was in the province Galatia. Being further north than the Homanadensian fortresses it was better known, and nearer the sphere of Hellenisation, but it never struck any coins (so far as known). Dalisandos in the plain of Trogitis was not strongly held by the tribe when freely marauding, but became important in the settled peace of the Roman empire.

^{page 252 note 2} Lycaonia was not a province until A.D. 371 but those two towns from 295 to 371 were in the province Isauria, not in Pisidia; and the north-west and north part of Isauria was assigned to the new province Lycaonia by Valens about 371.

^{page 252 note 3} A triumph was originally purely religious, and the triumphing general appeared as the god on his car and wearing his dress. It became more political and less strictly religious, but something of the religious character persisted.

^{page 252 note 4} On the other hand, it is also quite probable that the level ground round Trogitis, easily accessible from Antioch, was quickly overrun in the first year, and thereafter the real difficulties of the war began when the forty-four fortified villages among the mountains had to be captured, and the difficult narrow passes to be forced. I have seen one important entrance, ascending as by a ladder of rock 2,000 feet, which a few determined men could hold for a long time.

^{page 253 note 1} His references to the Homanadeis, pp. 668, 679, have no immediate bearing on the war.

^{page 253 note 2} Unless this cordon were maintained on the southern side, those hill tribes, fighting a guerilla warfare, might take refuge in small detachments and singly among the anti-Roman mountaineers of Cilicia Tracheia (subject to Archelaos), and so be ready to reunite and fight again. See also p. 235, note 2.

^{page 253 note 3} It is common in Anatolian municipal inscriptions to find a clear separation drawn between the imperial service and the municipal service.

^{page 254 note 1} Caesarea colonia was, in a sense, a fragment of Rome itself, separated in space, but peopled by Romans of the tribus Sergia.

^{page 254 note 2} There were, as stated already, few troops in the province Galatia; but the client kingdom of Bosporus was under a certain tutelage to the Galatic governor, and its troops were summoned.

^{page 254 note 3} [Bosporus was part of Polemo's kingdom, and the obligation to provide soldiers for the Roman army was evidently imposed by Agrippa; cf. Rostovtseff in J.R.S. 1917, p. 43, note 4. It is natural that they should have been utilised, as the Roman army had its hands full elsewhere: Anderson.]

^{page 254 note 4} [So also Stille, Historia leg. auxiliorumque inde ab excessu divi Augusti, p. 87: ‘Avec la plus grande vraisemblance’ (Cagnat): Anderson.]

^{page 255 note 1} Borghesi's opinion that Leg. XII wasin Germany at this time must be definitely set aside.

^{page 255 note 2} In J.R.S. 1916, p. 96, I used by a slip the term ‘Cohortes’ in place of Alae.

^{page 255 note 3} For some reason many soldiers were drawn from Eumeneia, and a few from Apameia (C.B. Phr. ii, 379); but elsewhere in Asia they are few; and no auxiliary troops take their name from senatorial Provinces.

^{page 255 note 4} In 74 great part of Galatic Pisidia was transferred to Pamphylia; the remainder was perhaps classed with Galatic Phrygia as one region. Some slight epigraphic evidence points to this.

^{page 257 note 1} It is true that Strabo about A.D. 20 does not mention Neapolis as one of the cities of Pisidia, but his list is taken direct from Artemidoros and therefore really belongs to an earlier time: he had not himself seen or travelled in this region of Pisidia, and he was depending on the excellent authority of Artemidoros. In a sense Anaboura and Neapolis are sister-towns which belong to the same district, lying between the valley of the Anthios and the lake country of Karalis, and traversed by a Via Sebaste of Augustus, made about 6 B.C. which here followed the long-existing line of an ancient highway very important in the Hellenistic time. The remarks which I have made on the date of Neapolis in Athen. Mitth. 1883, p. 75 f. require therefore to be modified; and the town of Neapolis may be older than there allowed on the wrong assumption that Strabo's list is authoritative in respect of 20 A.D. Perhaps Neapolis and Anaboura represented the newer Hellenistic and the older Pisidian tendency respectively; but the inscriptions of Anaboura (all late) are numerous and entirely Graeco-Roman in character.

^{page 257 note 2} I have often had occasion elsewhere to speak of the remarkable steepness with which the Taurus rises like a wall from the Lycaonian part of the plateau. There are few breaks in the wall, and these present only difficult access.

^{page 258 note 1} The reading of the proper names is treated in § 8. I give Mayhoff's text.

^{page 258 note 2} Acta (not Triumph.) are quoted as an authority in vii: doubtless Acta P.R.

^{page 259 note 1} [Strabo was not known to Pliny: Anderson.]

^{page 260 note 1} It is necessary with Mommsen and Nipperdey to read super Ciliciam and not per Ciliciam. The latter words would imply that the Homanadeis extended over various parts of Cilicia, but it is clear from the words of Strabo and Pliny that they were situated not exactly in Cilicia but on the Cilician frontier and on very high ground which might justifiably be called ‘above and outside of Cilicia.’ Nipperdey quotes the expression from Nepos Dat. 4, 1, Cataoniam, quae gens iacet supra Ciliciam: the situation is similar. The reading super is also justified by the expression of Strabo, 668, ἧς ὑπέρκειται ὁ Ταῦρος (i.e. Κιλικίας).

^{page 260 note 2} He probably took castella from the same excellent authority from which Pliny derived that technical term. Tacitus certainly used authorities of that class.

^{page 261 note 1} The Roman term Kastron, also in Palaiokastro, Neokastro, etc. is frequent in the Greek parts of Anatolia.

^{page 261 note 2} He united them with Galatia (diminished on the south-east, where a part of Pisidia was included in the praetorian province Lycia-Pamphylia), and this huge consular province lasted c. 74–115 A.D.

^{page 263 note 1} Pisidia-Lycaonia, p. 265.

^{page 263 note 2} See § 8, paragraph 2.

^{page 263 note 3} H.G.A.M. p. 335; Sterrett, W.E. 240.

^{page 263 note 4} Even the unified Athenian state in the period of most perfect synoikismos was divided into demoi; but among the newly conquered tribe Quirinius was probably not desirous of establishing perfect unity. The demoi were not parts of a true polis.

^{page 264 note 1} The text of Pliny (which has the genitive form here, gens Omanadum) is very uncertain as regards the spelling of proper names. Mayhoff's text is given above, §7; there is manuscript authority for Humanadum (which comes close to the Byzantine Oumanada, Cumanada, see p. 269).

^{page 264 note 2} Perhaps Str. should be interpreted as meaning (what is very probable in itself) that Cilician tribes of Tracheia allied themselves with the Homanadeis against Amyntas: the sympathy of those Tracheiotic mountaineers was certainly against Amyntas and the Romans, and this was a difficulty against which Quirinius had to be prepared.

^{page 264 note 3} Waddington considers that the names Katenna and Hetenna are varieties of the name of one town or tribe; but the Byzantine proof is complete that there were two bishoprics. Hirschfeld follows Waddington (Reisebericht in Berl. Monatsh 1875).

The omission of the Semitic initial CH in various Greek words has been discussed by Lewy, , Semitische Fremdwörter in Griech. Berlin, 1895Google Scholar, and Keller, Berliner Phil. Woch. 1896, p. 118: also Keller, Volksetymologie, p. 196. Compare the Lycian Idubessos with the central Phrygian Kiduessos (i.e. Idu Wessos Kidu Wessos). Sayce in a private letter regards the Lycaonian personal name Tarasis as TarHasis for TarKasis a derivative from the divine name Tarku, and the modern Marash as an ancient form Markas.

The subject is too large for a footnote, or for the writer's knowledge; but all writers on the relation of West Asian and Greek names must take account of it.

^{page 265 note 1} Οὐμάνδων, Ὀνομἀνάδων, Νοουμανάδων, are mere clerical blunders. Μανάδων, may be a true form, shortened from the older form given above, as mere popular pronunciation dropped the unaccented first syllable.

^{page 265 note 2} Cumanada occurs, see p. 269.

^{page 266 note 1} Compare κωρυκιώτης, Ἀκρασιώτης, Σιδαμαριώτης, Κιβυῤῥαίωτης.

^{page 266 note 2} The lists at the councils are taken from, and commonly expressed as, the signatures written by the bishops.

^{page 266 note 3} The pronunciation sometimes faded to INDA in this suffix.

^{page 267 note 1} See Fraser, Trans. Phil. Soc. Camb. 1912.

^{page 266 note 2} It is quite probable (as in H.G.A.M. p. 334) that in the Hellespontine Blados and the Lydian Blaundos the original Anatolian is MLADA for AMLADA, which appears in many variations as Amblada, Amilanda, Ampelada, etc. Polychalandos of Lydia (ch for b) occurs A.D. 359 (H.G.A.M. p. 128).

^{page 266 note 3} We know most about the south. The truth is doubtless universal in the peninsula.

^{page 268 note 1} Perhaps also Prostanna, Prostana, in Ptolemy Prostama (the latter form is not a mere scribal error: Prostamensis at Constantinople A.D. 381, defends Ptolemy, H.G.A.M. p. 407).

^{page 268 note 2} This remains the critical question in the early history of west Asia: so I wrote to Wilamowitz many years ago.

^{page 269 note 1} Possibly Οὐμανάνδρων. The intrusion of ρ seems due to some peculiarity of Anatolian pronunciation, a soft semi-vocalic R, differing from the Greek rho, involving difficulty and variation of expression in Greek letters: take for example Siniandos, Sinethandos, Sintriandos, on which see my paper Pisidia-Lycaonia.

^{page 269 note 2} The elements ANDOS, ONDA, are identical (see p. 267). There remains TALB and TUM, in which B = M and L may be (like R, see preceding note) intruded.

^{page 269 note 3} That there was only one bishop and one bishopric is the view taken by Turner, who says ‘idem est episcopus Quintus Quirillus eiusdem civitatis comanadensis timanodorum.’ This opinion is the simplest; but it is not always the simplest that is right in such matters. [Since I wrote as precedes Mr. Turner tells me that he is disposed to take my view, which is based on I. the best manuscript; I. often stands alone and right.]

^{page 269 note 4} Codex Ingilrami ep. Teatini nunc Vaticanus Reginae, 1997: saec. IX (C. H. T. who quotes it as I).

^{page 269 note 5} The Greek original was perhaps Κύριλλος Μαναδεύς or Μανανδεύς.

^{page 269 note 6} The manuscripts of this Class V (even those which give here some corrupt form of Vasada) also insert at the end of Pisidia (C. H. T. no. 155) Theodorus of the bishopric uis or viis. Classes III, IV give a similar Theodorus as no. 155. Classes I, II omit him.

^{page 269 note 7} Another very common form is exemplified at the council held in 536, ὁ Οὐμανενδεωτῶν, the (bishop) of the people of Houmanenda.

^{page 270 note 1} The generic term becomes a personal name (according to a frequent usage in every society, Miller, Baker, etc., so in Anatolia Devrish in modern times for Dervish, Λαχανᾶς in ancient times for λαχανοπύλης).

^{page 270 note 2} See my article on the ‘Religion of Greece and Asia Minor,’ in Hastings, , Dict. Bib. v, p. 129Google Scholar: also on ‘Phrygians’ in Hastings, , Dict. of Eth. and Relig. ix, p. 902Google Scholar.

^{page 271 note 1} The plain west of Aleppo (as Professor Sayce tells me) was called Ouana. The wine-god was Ouan: hence Vangdibassin, Vangdamoas etc. personal names.

^{page 271 note 2} There were many exceptions to the custom that legati of the emperor held a province for three years; and there is no difficulty in accepting a shorter tenure of office for any of these governors, if any argument for this could be found.

^{page 272 note 1} A boundary stone is set up ‘per Rutilium Gallicum cos. pont. et Sentium Caecilianum praetorem legatos Aug. pr. pr.’, Cagnat, , Comptes rendus de l'Acad. xxii (1894), p. 43Google Scholar; A.E. 1912, no. 151. See Cheesman, p. 257. M. Bour (see §10) gives other examples showing great freedom of government in the provinces.

^{page 273 note 1} This cannot be inferred even from the statement that Quirinius left the land ἕρημον Τῶν έν ἀκμῆ, for 4000 captured men were settled in neighbouring cities, and the tyrannos might have been spared as a captive. This is indeed improbable; but Strabo has no statement anywhere that he was killed by Quirinius. His fate is left to conjecture. Augustus spared and treated well some kings who fought against him: this doubtless was skilful policy, as they might if spared be useful agents.

^{page 273 note 2} The theory stated there was published in the for Expositor 1895, and is mentioned by R. S. Bour in his work published before my book. The two treatises were written independently; but each thus can quote the other.

^{page 273 note 3} The chronology is there stated in a more tentative way than here (as the writer's first essay in a difficult subject); and it followed the principle of making the smallest possible change in Mommsen's conclusions.

^{page 274 note 1} Mommsen omits the probable, if not certain, fact that Marcius Censorinus governed Asia as proconsul in A.D. 2. This is generally admitted (see Prosopograpbia Imp. Rom.): he died in Asia about the same time as Lollius in Armenia, A.D. 2. Quirinius could not be his successor immediately, as he was sent to Armenia 2–4 on the death of. Asia Lollius.

^{page 275 note 1} Quoted from Christ Born at Bethlehem, p. 235.

^{page 276 note 1} That there were two towns named Dalisandos is certain (H.G.A.M. p. 335). The Isaurian is called Lalisandos by Stephanus.

^{page 277 note 1} See ‘Topography of Nov. Isaura’ in J.H.S. 1905, p. 163: on its art see Miss Ramsay in Studies in Eastern Roman Provinces, pp. 1–90.

^{page 277 note 2} In truth Pliny's conception of the facts and localities was inexact, and his words express his vague thoughts.

^{page 278 note 1} Oppida eius intus Isaura, Clibanus, Lalasis: these words fill up the gap left in the quotation from Pliny on p. 258.

^{page 278 note 2} See H.G.A.M. pp. 417–9: cp. 371, 383.

^{page 278 note 3} I should not like to quote examples, as some of the writers are Φἰλοι ἅνδρες. It is always unpleasant and invidious to mention inaccuracies in detail, unless this is absolutely forced on one in the interests of truth.

^{page 279 note 1} Polemon is called by Appian king of a part of Cilicia (Bell. Civ. v, 75).

^{page 279 note 2} Strabo regards Isaura in the feminine singular as the proper name of each of the villages, one Palaia and the other Nea. He uses the expression Τὰ Ἰσαύρα (so, not Ἰσαυρα) as a dual, not as a neuter plural (Ἰσαυρα). The idea of two villages is so strong in the passage that the dual here is practically necessary (though it was then archaic): possibly at was taken by Strabo from an older authority.

^{page 279 note 3} Isaura Nea is the town which was captured by Servilius Isauricus in 79 B.C. It was evidently a place of very considerable military strength, and as we may add of considerable size: the strength and the size are both obvious from the situation, described in J.H.S. 1905, p. 163. Strabo probably Τὴν μὲν παλαιὰν καλουμένην, εὐερκῆ, [τὴν δὲνέαν], regarding the lacuna as misplaced.

^{page 279 note 4} Besides this supply there is a splendid spring, Bel-Bunar, outside Isaura, reached by a sharp descent from the south-east gate: Sterrett W.E., pp. 122, 151. So also Calder.

^{page 280 note 1} So written in an inscription published by Miss Ramsay, Studies in Eastern Roman Provinces, p. 47.

^{page 280 note 2} Garb, points back to an older Gdarb- or Gderb-, which may be the Anatolian equivalent of the Arabic Darb: cp. Anatolian Kabala, Arabic Gebel (J.H.S. 1918, p. 163). Initial gd in Gdanmaa (see p. 267). Steph. Byz. calls Derbe λιμήν, i.e. a frontier customs-station and guard-house.

^{page 280 note 3} Parlais was situated at the important bridge crossing the Irmak where it issues from Karalis: see Pisidia-Lycaonia, p. 261.

^{page 280 note 4} I traversed this road in 1886. The stream at Bulumia flows to Khatyn-Serai, and comes from near Kizil-Euren and Siniandos: the watershed extends west from Loras Dagh, on whose north slope these places are situated.

^{page 282 note 1} The name is doubtful.

^{page 283 note 1} Perhaps Prochoros and Erechtheus. The latter does not look Christian, but it may be an example of the class described in J.H.S. 1918, p. 149.