^{page 163 note 1} Schwartz, E., REii. 1230–6 (= Griechische Geschichtsschreiber [Leipzig, 1957], 130–40Google Scholar). This classic survey is the basis of most later accounts of Arrian's life and career, such as Hartmann, K., Flavins Arrianus und Kaiser Hadrian(Programm, Augsburg, 1907);Google ScholarChrist-Schmid, , Geschichte der griechischen Literatur ⁶ (Munich, 1924) ii (2). 746–51;Google Scholar Stein, PIR² F 219 (up-to-date account of the epigraphic testimonia). The editions of the testimonia of Arrian's life by Roos (Flavius Arrianus: Scripta Minora et Fragmenta[2nd edn: Leipzig, 1968], lviii–lxv)Google Scholar and Jacoby (FGrH 156 T 1–7) also follow the lines laid down by Schwartz. In comwhat follows I shall refer to the testimonia and fragments of Arrian in Roos's edition, where there are marginal references to Jacoby.

^{page 163 note 2} For these dates, see Periplus 17. 3 (refer-ence to the death in 131 /2 of Ti. Iulius Cotys, king of the Cimmerian Bosporus) and Tact. 44. 3 (20th regnal year of Hadrian).

^{page 163 note 3} See particularly his exposition at RE ii. 1235; ‘Auf die Dauer geniigte das dem wackeren Marine nicht, und er empfand den unwiderstehlichen Trieb, etwas Bleibendes, das über die ephemere Broschüre hinausging, zu schaffen.’.

^{page 163 note 4} Wirth, G., ‘Anmerkungen zur Arrian-biographie’, Historia xiii (1964), 209–45,Google Scholar argues that the Anabasis and Indica were composed sometime after 147. Bowie, E., Past Present xlvi (1970), 24–7,Google Scholar argues for the first half of the 160s. Stadter, P.A., GRBS viii (1967), 160–1,Google Scholar repeats that the Anabasis and Indica were late works, com-posed after Arrian's retirement to Athens, as does Gabba, E., RSI lxxi (1959), 373.Google Scholar

^{page 163 note 5} Reuss, F., ‘Arrian und Appian’, Rheinis-ches Museum xlv (1899), 446–65Google Scholar(esp. pp. 455– 61); cf. von Domaszewski, , SB Heidelberg xvi (1925/6),Google Scholar Abh. 1, pp. 5–6 (similar conclusion but weaker arguments).

^{page 164 note 1} Phot. cod. 93, p. 73 b 4 ff. = Bithyniaca F 1. 2 (Roos).

^{page 164 note 2} Lucian, Alexander 2 = T 24 (Roos).

^{page 164 note 3} , Brinkmann and Herter, , Rh. Mus. lxxiii (1924), 373–401.Google Scholar Wilamowitz had already argued (Hermes xli [1906],Google Scholar 157 f.) that the works were the product of the second century A.D.; cf. Roos, Opera Minora, xxvii f.

^{page 164 note 4} Phot. cod. 58, p. 17 b 22 = T 2 (Roos)

^{page 164 note 5} Cyneget. 1. 4 = T 19.

^{page 164 note 6} Phot. cod. 58, p. 17 b 11–20; Suda s.v.’Appiavós = T 2. The Sudd's information is qualified as coming from Heliconius, and this datum was certainly taken over from his immediate source, the Onomatologus of Hesychius. Cf. Suda s.v. Wentzel, , Hermes xxxiii (1898), 275Google Scholar;Wirth, G., Historia xiii (1964), 507.Google Scholar

^{page 164 note 7} Phot. 17 b 15–17

^{page 165 note 1} Wirth, , Historia xiii (1964), 228;Google ScholarBowie, , Past Present xlvi (1970), 24.Google Scholar The tradition was accepted by Reuss, , Rh. Mus. xlv (1899), 456.Google Scholar

^{page 165 note 2} Phot. cod. 209, p. 165 a 31 ff; Suda sv .

^{page 165 note 3} Philostratus, VS 1. 7 (p. 8 Kayser) repeats the anecdote of the in his different, but equally sensational, Life of Dio.

^{page 165 note 4} Themistius, Orat. 34, pp. 451–2 (Dind.) = T 13; cf. Themistius 17, p. 262.

^{page 165 note 5} Marcus, Medit. 1. 7; SHA, Marcus 3. 3; Dio. 71.35.1; Them. Orat. 13, p. 212 (Dind.). For biographical details see PIR ² I. 814.

^{page 165 note 6} His consulship is attested by two brick stamps—CIL xv. 244 and 552 {Severo et Arriano cos.). Arrian is attested in Cappadocia by 131/2 (Peripl. 17. 3), and in the years immediately preceding there is room for him among the suffects of 129 or 130 (Degrassi, I fasti consolan, 37).

^{page 165 note 7} Rusticus’ earlier career is not known, aPart from his first suffect consulship in July:133 (CIL xvi. 76; JRS li (1961), 63), but it is most unlikely that he was legate of Cappadocia. At this period Cappadocia seems to have been held shortly after the consulship, and in the crucial decade after 133 that province was governed by Arrian untiI 137. and his successor, L. Burbuleius Optatus Ligarianus was in tenure after the accession of Antoninus Pius (ILS 1066)

^{page 166 note 1} Suda s.v.= T 1.

^{page 166 note 2} Wirth, G., Klio xli (1963), 221–33.Google ScholarMillar, F., Cassius Dio (Oxford, 1964), 70,Google Scholar even doubts whether Dio did in fact write a formal biography.

^{page 166 note 3} Wirth, art. cit., especially pp. 230–1.

^{page 166 note 4} Dio 69. 15. 1 = Xiphilin. 251. 27– 252. 1; Exc. Urb. 17; cf. Zonaras 11. 23–4.

^{page 166 note 5} Lucian, Alexander 2 = T 24.

^{page 166 note 6} OGIS 505. 8–9; cf. PIR V. 551 for Eurycles' subsequent career.

^{page 167 note 1} Arr. Anab. 1. 12.4–5.

^{page 167 note 2} In his edition of 1848 Krüger attributes the neuter plural interpretation to Raphelius and the masculine interpretation to Grono-vius and Schmieder, himself opting for the former. Modern scholars have generally pre-ferred the masculine plural; note the transla-tions of Müller, Robson, and de Sélincourt. Bowie, , Past Present xlvi (1970), 27, seems to prefer the neuter.Google Scholar

^{page 167 note 3} Arrian generally uses with an accusative and infinitive, as at Anab. 1. 12. 4 and 7– 20. I. There are parallels in other authors for the use of the verb with accusative and genitive plural; Lucian, Demonax 3 ( [Arist.] De Mundo 39I^a6; Xen. Anab. 3. 2. 7; Cyrop. 6. 4. 6. In all these cases the genitive, , must be understood as neuter plural. These parallels guarantee the interpretation of the Arrian passage. For the use of to mean ‘the first place’, ‘supremacy’, see Dio 40. a. 3, 42. 57. 1. As for the alternative interpreta-tion, , I can only say with Krüger, vereor ut Graece dicere licuerit.

^{page 167 note 4} Phot. cod. 93, p. 73 b II ff. (= Bithyniaca F 1, 3.

^{page 167 note 5} Cf. T4(Roos).

^{page 168 note 1} One cannot, of course, lay stress on the wording and This need not mean that Arrian was writing in his old age, as some have thought (Kornemann, Die Alexandergeschkhte des Königs Ptolemaios i. 36; Wirth, , Historia xlii [1964], 223–4)Google Scholar. The expression merely means ‘right from childhood’ (cp. Plat. Gorg. 510 d; Arist. NE 10. 1179b31), and could easily be used by a man in his early thirties.

^{page 168 note 2} Bowie, art. cit., 26–7: ‘To me this work orepresents my country and my family and my public offices.’

^{page 168 note 3} Hdt. 1. 184, 2. 161. 3, 5. 36. 4, 7. 93.

^{page 168 note 4} Phot. cod. 93 p. 73 b 11–18; see my discussion, p. 179.

^{page 168 note 5} In the Anabasis he usually refers vaguely to the work as a whole as a συγγαøή.(Prooem. 3, 6. 28. 6, 7. 3. 1, 7. 30. 3).

^{page 168 note 6} Arr. Ind. 43. 14.

^{page 168 note 7} J. H. Kent, Corinth viii, pt. 3 (The Inscriptions 1926–1950 [Princeton, 1966]), nr. 124, pp. 55–6; [øιλ]όσοø[ον…]|

^{page 168 note 8} Bowersock, G.W., GRBS viii (1967), 279–80.Google Scholar

^{page 169 note 1} For the epistula ad L. Gellium see Roos, Opera Minora, 196.

^{page 169 note 2} Hadrian is named with his imperial titles; he is not as yet .

^{page 169 note 3} The first line of the inscription might have read [¸¹λ]όσόø[οοκι συγγραøέα]. This supplement gives an adequate line length and is consistent with other dedications to literary figures, which tend to name all the genres in which the man honoured was prominent; cf. IG ii.² 3704. 12 .

^{page 169 note 4} . The inscription was first published by D. Peppas-Delmouzou, AAA iii (1970), 377–80,and the praenomen modified by Oliver, J.H., GRBS xi (1970), 338.Google Scholar

^{page 169 note 5} The full cursus is given by the celebrated Pupput inscription: CIL viii. 24094 = ILS 8973 = Smallwood nr. 236. He was consul ordinarius in 148 (CIL xvi. 95), curator aedium sacrarum in 150 (CIL vi. 855), and proconsul of Africa in 168/9 (Merlin, A., Memoires de I’Academie des Inscriptions et Belles-Lettres xliii (ii) (1941), 93–122)Google Scholar. Alfoldy, G., Fasti Hispanienses (1969)Google Scholar, 32 f., dates his tenure of L. Germany to 150–2, and places him in Spain in 161–4.

^{page 169 note 6} ILS 8973 : trib. pi., praef. aerar. Saturni, item mil., cos.; cf. Plin. Ep. 10. 3a. 1 (cf. 3. 4. 3) ; Paneg. 91. I .

^{page 169 note 7} For a sceptical account of the problems involved, see Serrao, F., Atti del HI Congresso internazionale di epigrafia greca e latina (Rome, 1957), 395–413.Google Scholar

^{page 170 note 1} Eutrop. 8. 17 (redaction of praetor's edict); Dig. 4. 2. 18 (citation from Book 64 of Iulianus’ Digesta, referring to Antoninus Pius as still living).

^{page 170 note 2} ILS 8973 cui divos Hadrianus soli solarium duplicavit propter insignem doctrinam.

^{page 170 note 3} Chr. Habicht, , Istanbuler Mitteilungen ix/x (1959/1960), 109–25.Google Scholar The most accessible edition of the inscription is SEG xviii. 557. This document gives the cursus of Charax from quaestorship to suffect consulship. The date of his suffect consulship, 147, was already known from the Fasti Ostienses (I.I. xiii. 1. 207). The scanty extant fragments of his universal history are edited by Jacoby (FGrH 103). The Suda (FGrH 103 T 1) cites an introductory epigram, which proves that Charax, like Arrian, held a priesthood in his native city.

^{page 170 note 4} M. Pompeius Macrinus Theophanes, who like Charax was successively legionary legate and governor of Cilicia, moved on to a priesthood (XVvir s.f.) and the proconsulate of Africa (IG v. 2. 151 = Smallwood nr. 232). Unlike Charax he had received no marks of favour in the Quadraearlier part of his cursus.

^{page 170 note 5} For a similar decree honouring a legate of Cilicia on his promotion to the consulship, cf. ILS 8827 = OGIS 576 (C. Etrilius Priscus).

^{page 170 note 6} Habicht, art. cit. 111, suggests that the Pergamene historian was favourably impressed by the help afforded the Attalids solarium during the hostilities with Antiochus III by a contingent from Achaea (SIG ³ 606; cf. x Liv. 37. 20. 1, 39. 9).

^{page 170 note 7} Like Arrian in the preface to the Bithyniaca he refers to himself as a priest in his native city (). He clearly mentioned no office held at Rome, otherwise the detail would have been taken up by the source of the Suda, which was greedy for biographical information (FGrH 103 T 1).

^{page 170 note 8} SIGM³ 887 = FGrH 97 T 2. The man is here named .

^{page 170 note 9} Two other inscriptions of the period of Caracalla mention a consular named C. Asinius Protimus Quadratus (IG xii. 7. 267; Forsch. Ephes. iii. 127, nr. 40). Scholars have used Occam’s razor and identified the historian, Asinius Quadratus, with the consul designate honoured at Olympia and the consular, C. Asinius Protimus Quadraearlier tus; cf. PIR ² A 1244–6; Barbieri, , L'albo senatorio da Seuero a Carino, nr. 29 (pp. 21–2);Google Scholar Habicht, art. cit. 111.

^{page 171 note 1} Philostr. VS 1. 22. 3 (p. 36 Kayser). This statement about Dionysius’ multiple procuratorships is confirmed by an inscription of Ephesus honouring (Forsch. Ephes. iii. 133, nr. 47; is a possible supplement in the last line).

^{page 171 note 2} SHA, Hadr. 16. 10.

^{page 171 note 3} Arr. Peripl. 2. 4 suggests that the historian had been especially favoured by Hadrian.

^{page 171 note 4} Cf. Schwartz, RE ii. 1230–1; Wirth, , Historia xiii (1964), 228.Google Scholar

^{page 171 note 5} IGR iii. 173 = OGIS 544 = Smallwood 215. This inscription gives Severus' municipal offices only; for details of his family and cursus see Miinzer, REx. 811–30; PIR ² 1.573.

^{page 171 note 6} IGR iii. 174 = Smallwood 216:

^{page 171 note 7} IGR iii. 174–5 (cf. At 1923, nr. 4). Publicius Marcellus was legate in Syria in 132, as is attested by an inscription of Palmyra SEG xv. 849; cf. Dunant, Mus. Helv. xiii (1956), 216 ff.). It is highly probable that Severus’son, C. Iulius Severus (cos. 155), was tribune under him in the IV Scythica (IGR iii. 172). The date is at least consistent. Service as a military tribune in 132 would place him in his early forties in 155, just right for the consulship (cf. Miinzer, RE x. 817; PIR ² I. 574). In that case the father was old enough to have a twenty-year-old son when he served as legate.

^{page 171 note 8} Similarly A. Claudius Charax seems to have had his first military experience as legate of II Augusta in Britain (SEG xviii. 557) and M. Pompeius Macrinus Theophanes as legate of VI Victrix in L. Germany (IG v (2). 151; cf. IGR iv. 96).

^{page 172 note 1} Arr. Ind. 4. 15–16.

^{page 172 note 2} For instance, his statement that the tributaries of the Danube are small and rarely navigable is erroneous. Pliny knew of sixty tributaries, over half of which, he says, were navigable (NH iv. 79; cf. iii. 147 f.). Similarly inaccurate are Arrian’s repeated statements that the Danube has five mouths (Anab. 1. 3. 2 ; 5. 4. 1; Ind. 2. 5; Peripl. 24. 2). This is a reflection of the more ancient tradition (Hdt. 4. 48; Ephorus, FGrH 70 F 157; Scymn. 773). From Augustus’time the regular figure for the mouths of the Danube was seven (Strabo 7. 3. 15 (305); Ovid, Tristia 2. 189; Stat. Silv. 5. 2. 136–7; Mela 2. 8); cf. Brandis, RE iv. 2118–9.

^{page 172 note 3} IG ii². 2055. For the dating see Kolbe, Athenische Mitteilungen xlvi (1921), 131 ff. esp. p. 148).

^{page 172 note 4} Arr. Anab. 3. 16. 8, cf. 7. 19. 2, 1. 16. 7, 7. 13. 5.

^{page 172 note 5} Schwartz, RE ii. 1237; PIR ² F 219 p. 139); Roos, Opera Minora, Ixiv; Wirth, , Historia xiii (1964), 224–5.Google Scholar

^{page 172 note 6} Arr. 3. 16. 8.

^{page 172 note 7} Much unnecessary confusion has been generated by Arrian’s reference both to a . It has been assumed that Arrian was referring to two distinct altars, the first at Athens and the second at Eleusis, and that Arrian mentions the Eleusinian altar to elucidate his reference to the more obscure altar at Athens (Roscher, , Lexicon der gr. und torn. Mythologie i. 2654Google Scholar; Wachsmuth, RE vi. 893–4; Wycherley, R.E., The Athenian Agora iii (Princeton, 1957),Google Scholar 94, nr. 260). But this makes gibberish of the passage. Arrian is locating the statue group of the tyrannicides by co-ordinates which he thinks will be progressively more familiar to his readers. He refers first to the area of Cerameicus, which extended from the north-west of the Acropolis to the Dipylon Gate; next to a particular landmark; the Metroon; and to the altar of the Eudanemi, which he claims will be familiar to any initiate. If Arrian is referring to one altar under slightly different names, the passage makes perfect sense, but if he means us to understand two altars, he is behaving most perversely, first referring to a very obscure altar in the agora and then adding a gratuitous footnote that any initiate will know the altar of the same hero at Eleusis. Obscurum per obscurius: Arrian could hardly have been more misleading. There can be no doubt that Arrian is refer-ring to a single altar, which was located in the agora immediately to the north-west of the Acropolis. There is no difficulty in the double nomenclature. We hear of a genos of the Eudanemi at Athens, whose activities were connected in some way with the Mysteries (Dion. Hal. Deinarch. II [p. 315 I Radermacher]). Nothing is more natural than that this genos should have administered the cult of their eponymous hero, and that the same altar should be called both ‘altar of Eudanemus’ and ‘altar of the Eudanemi’. What is more, the altar was in just that quarter of the city which would be traversed by the mystic procession to Eleusis. On the fifth day of the festival the were taken from the Eleusinion in Athens for the first stage of the journey to Eleusis (Mylonas, G.E., Eleusis and the Eleusinian Mysteries [London, 1961], 246–7Google Scholar). Now the Eleusinion was on the southern boundary of the agora just below the Acropolis—exactly where Arrian Arrian locates the statue group and altar of Eudanemus (3. 16. 8 cf. Thuc. 2. 15. 6 for this use of πόλις). I conring elude that the altar was on the Sacred Way of in the vicinity of the Eleusinion, where every initiate would pass.

^{page 173 note 1} Hdt. 8. 65. 4.

^{page 173 note 2} Plut. Sulla 26. I; SIG ³ 1125 (for the date, c. 74 B.C., cf. Cichorius, Römische Studien, 187) Cic. de Leg. 2. 35 (Cicero and Atticus); Dio 51. 4. I, 54. 9. 10 (Augustus). For Hadrian's activities in 124/5 see SHA, Hadr. 13. I; Grraindor, , Athènes sous Hadrien (1936), 5–8.Google Scholar

^{page 173 note 3} Arr. Peribl 9. I.

^{page 173 note 4} Arr. Anab. 2. 16 6; Ind. 41, 2.

^{page 173 note 5} Arr. Anab. I. 16. 4; cf. Plut. Alex. 16. 16.(FGrH 139 F 5); Just. II. 6. 13.

^{page 173 note 6} Vell. I. II. 4; Pliny, NH 34. 64.

^{page 173 note 7} Wirth, , Historia xiii (1964), 231,Google Scholar argues that Arrian found it embarrassing to mention the depredations of Metellus in the Anabasis where he is generally enthusiastic towards Rome. But in that case why did Arrian even mention Lysippus‘ statue-group in this context? It is not germane to his narrative, and, if he was so careful of Roman sensibilities, he could easily have omitted all reference to it.

^{page 174 note 1} Arr. Anab. 2. 4. 2.

^{page 174 note 2} Curt. 3. I. 24, 4. I. He seems to be drawing ultimately on the same source as Arrian.

^{page 174 note 3} Arr. 7. 16. 3. cf. Reuss, Rh. Mus. (1899), 459; Domaszewski, von, SB Heidelberg xvi (1925/1926),Google Scholar Abh. i, p. 5.

^{page 174 note 4} Strab. II. 14. 13 (531); FGrH 139 F I. Here Strabo has no doubts about the course of the Araxes (II. 14. 3), nor has the Elder Pliny (NH 6. 26). After the campaigns of Pompey the only problem was whether or not the Araxes flowed into the Caspian independently of the Cyrnus (Plut. Pomp. 34. 4; App. Mithr. 103, 480).

^{page 174 note 5} Arr. Anab. 1. 12. 5.

^{page 174 note 6} This fact has been an embarrassment to upholders of a late date for the Anabasis. Bowie, art. cit. 27, omits the crucial passage altogether. Stadter, , GRBS viii (1967), 161,Google Scholar mistranslates as ‘in my life’. Wirth, art. cit. 224, resorts to interpretation, claiming that there is a sharp distinction between the first designating Arrian's homeland, Bithynia, and the second his adopted city, Athens. I fail to see here any more than a simple case of variatio. If Arrian meant to draw a distinction here between his birthplace and his city of residence, he could hardly have expressed himself more obscurely. No uninformed reader could be expected to infer that there was a distinction implied here. The same objection is valid against the alternative suggestion, that Rome is what Arrian intended to denote by his expression Arrian's passion for variatio was such that he was almost incapable of using the same word twice in close proximity; compare I. 12. 8, where is patently used as a variant for which occurs in the previous phrase. Similarly, I. 12. 5 is unintelligible unless the expressions, and denote the same object. Now it is clear that Greeks of the second century A.D. tended to regard their birthplace rather than Rome as their Appian explicitly contrasts Rome with his Alexandria (praef. 15. 62). Dio also regards Nicaea as his (76. 15. 3, 80. 5. 3); Italy he describes merely as (F I, 3). The same is true of Arrian, who composed his Bithyniaca explicitly as a gift for his Nicomedia (Bithyniaca F I, 1 and 3). In the Anabasis too the probability is overwhelming that Arrian used and as variants to describe his birthplace, Nicomedia.

^{page 175 note 1} Cf. Millar, Cassius Dio, 7 ff.

^{page 175 note 2} App. Prooem. 15. 62 cf. Herodian I. 2. 5. By contrast Claudius Charax, like Arrian, mentioned his priest-hood at Pergamum but did not hint at a career in Rome (see p. 170 n. 7).

^{page 175 note 3} Compare the case of C. Iulius Philo-pappus, a member of the deposed royal house of Commagene (PIR ² I. 151). He was descriparchon at Athens shortly before 87/8, but it was not until the reign of Trajan that he catawas adlected to praetorian rank. He then moved immediately to the consulship and a priesthood (ILS 845).

^{page 175 note 4} Arr. Anab. 7. 15. 5 (cf. 7. 1. 3), 3. 5. 7.

^{page 175 note 5} Arr. 5. 7. 3–5. This passage has been taken as irrefragable evidence that Arrian had extensive military experience. Schwartz, RE ii. 1237; von Domaszewski SB Heidel-berg, 6.

^{page 175 note 6} 6 Arr. 5. 7. 2; cf. Hdt. 7. 33–6.

^{page 175 note 7} Suda s.v. = Dio 71. 2. 3 (cf. Boissevain, , Hermes xxv [1890], 338 n. 1Google Scholar).

^{page 175 note 8} The version in the Suda adds that the boats used were generally flat-bottomed and is rather more explicit that they were launched upstream from the intended crossing place, Unlike Arrian, the Suda ends with a descriparchon tion of the ship nearest the enemy bank, equipped with towers, archers, and catawas pults. Arrian ends with a purple passage, describing how despite the noise the Roman discipline was unimpaired (5. 7. 5; some of the colour seems borrowed from Thucydides 7. 70. 6–7). But apart from these variants the two descriptions are vir-tually identical and must stem from a common source.

^{page 176 note 1} Dio 68 26 1–2

^{page 176 note 2} App. Syr. 56. 288–91; BC 2. 152. 639 Schwartz, RE ii. 1245 took it as self-evident that Appian had read the Anabasis; Reuss, art. cit. 446–50 argued at length for the borrowing.

^{page 176 note 3} Wirth, , Hutoria xiii (1964), 210–20.Google Scholar

^{page 176 note 4} Arr. 7. 22. 2–5 = Jacoby, FGrH 139 F 55.

^{page 176 note 5} App. Syr. 56. 288–91.

^{page 177 note 1} Compare his version of the death of Callisthenes (Arr. 4. 14. 3 = FGrH 139 F 33), and his apology for Alexander's notoriously intemperate drinking habits (Arr. 7. 29. 4 = FGrH 139 F 62).

^{page 177 note 2} Arr. 7. 21. 1–7; Strab. 16. I. 11 (741).

^{page 177 note 3} App. BC 2. 153. 644. There is a slight divergence between Appian and Arrian in their spelling of the name of the canal. Appian's manuscripts read while the consensus in Arrian is This is no problem. The name of the canal is represented in Babylonian documents most regularly as Pal-lu-kat (Meissner, , Mitteilungen der vorderasiatischen Gesellschaft i [1896],Google Scholar 183 ff.), and the spelling in the manuscripts of Pliny, NH 6. 30 oscillates around Pollaconta. Appian0027;s form is the most correct, but that does not disprove the hypo-thesis that he drew from Arrian. All our extant manuscripts of the Anabasis are known to be copies of Roos's codex A (cf. Roos, Anabasis, v–xi), and a corruption in this, the archetype, would naturally have been passed on to all surviving manuscripts. The corruption of into -π- is very easy.

^{page 177 note 4} Arr. 7. 21. 3–5.

^{page 177 note 5} The details are given in Strabo 16. 1. 9–11 (740 f.); all his account of Alexander's operations on the Euphrates seems taken from Aristobulus. The details are extracted by Arrian, but distributed over his narrative. His description, for instance, of the demoii-tion of the Euphrates cataracts is placed much earlier (Arr. 7. 7. 7 = Strab. 16. I. 9).

^{page 178 note 1} Strab. 16. 1. 11 (741) = FGrH 139 F 56; Arr. 7. 19. 6–20. 2.

^{page 178 note 2} Arr. 7. 22. 1 = App. 2. 153. 644–5. For discussion of the context, see Reuss, art. cit. 448–50.

^{page 178 note 3} For the terminus post quern see Gelzer, M., Kleine Schriften, iii. 291.Google Scholar Appian seems not to know of Marcus' reintroduction in 163 of Hadrian’s division of Italy into regions (BC 1. 38. 172; cf. SHA, Marcus 11. 6), or of Verus' conquests in 165 (Prooem. 4); cf. Schwartz, RE ii. 216.

^{page 178 note 4} Wirth, , Historia xiii (1964),Google Scholar 235 ff., argues that Lucian's criticism of the luxuriant fantasies of the Verus historians in his treatise is a veiled criticism of Arrian's Anabasis. Unfortunately there is no overt reference to Arrian in the entire treatise, and Wirth's examples of attacks on specific portions of the Anabasis are most unconvincing. There is absolutely no reason to think that the to objects of Lucian's attack are not those he explicitly states them to be.

^{page 178 note 5} Phot. cod. 131, p. 97 a 10 ff. = FGrH 150 T 1 Cf. Arr. I. 12. 4–5.

^{page 178 note 6} Phot. cod. 93, p. 73 a 32 ff. = Bithyniaca F 1.

^{page 179 note 1} Phot. cod. 83, p. 65 a 2 ff. = Dion. Hal. 1. 7. 2.

^{page 179 note 2} Phot. cod. 70, p. 35 a 14–26 = Diod. 1. 4. 1–7 (Photius rearranges the order of his original but keeps very faithfully to the terminology).

^{page 179 note 3} Phot. cod. 71, p. 35 b 10–ai = Dio 80. 5. 1 ; 79. 7. 4 ; 80. 1. 2–3, 5. 2–3.

^{page 179 note 4} Phot. cod. 57, p. 17 a 13–15 = App. Prooem. 15. 62.

^{page 179 note 5} Phot. cod. 93, p. 73 b 14–18 = Bithyniaca F 1, 3.

^{page 179 note 6} Anab. 1. 12. 5, cf. Prooem1. 3.

^{page 179 note 7} For this procedure see Lucian, 47–8; Cic. ad Att. 2. 1. 1–2; Arr. Epist. ad L. Gellium 2 ; Avenarius, G, Lukians Schrift zur Geschichtsschreibung (1956),Google Scholar 85 ff.

^{page 180 note 1} Wirth, , Historia xiii (1964) 227.Google Scholar Once he has dismissed any reference to the Anabasis in Photius, he is able to argue for the sequence Parthica-Anabasis-Bithyniaca.

^{page 180 note 2} If there is a scribal interpolation in this passage, I prefer, with Bekker, to identify it as the phrase It is otiose and hinders the flow of the passage, just the sort of marginal gloss that cittaa scribe would insert to explain the neuter expression Such elucidation of the obvious is more characteristic of scribal marginalia than the highly sophisticated interpolation required by Wirth's theory. I am arguing for a double corruption, first the intrusion of a marginal gloss into both the principal manuscripts of Photius, and then the omission in A only of the first phrase. It is a complex corruption but neither improbable nor implausible.

^{page 180 note 3} Cf. Millar, Cassius Dio, 190–1. Gabba, , RSI lxxi (1959), 373.Google Scholar The strong Greek flavour, coupled with the presupposition that the Bithyniaca was a work of Arrian's old age, led Gabba to suppose that Arrian was writing in disillusionment after virtual dismissal by Antoninus Pius (‘Arriano perde ai nostri occhi tutto il suo aspetto di cittaa dino romano’). This theory that Arrian ended his public career under a cloud goes back to Schwartz, RE ii. 1231, and has been widely accepted (cf. Wirth, , Historia xiii [1964], 508–9).Google Scholar There is, however, no positive evidence for the hypothesis, and the noticeably Greek colour of the Anabasis and Bithyniaca is far more easily explained on the assumption that they were written before Arrian's career at Rome got into full swing.

^{page 181 note 1} Steph. Byz. s.v. = Parthica F 17.

^{page 181 note 2} F. A. Lepper, Trajan's Parthian Wars, 202–4, argues that the Parthica was produced under Hadrian and retailed the official view of the campaign.

^{page 181 note 3} Lucian, Alexander 2; the date is supplied by the reference to at Alexander 48.

^{page 181 note 4} IG ii2. 2055.

^{page 181 note 5} IG ii2. 1773, 1776.

^{page 181 note 6} The list of prytanies for the year 167/8 is fairly detailed (IG ii2. 1774), and many of the men named can be identified. None of their ephebates fall before 142/3 and none much after 150. Among the colleagues of the Flavius Arrianus, prytanis in 166/7 was a Claudius Xenocles, who is known to have been ephebe in 154/5 (IG ii2 1773 6 ; cf. 2067. 40); in other words he was born at the time when Arrian the historian held praetorian rank at Rome. For general details about the Athenian prytanies in the Empire, see Geagan, D.J. ‘The Athenian Constitution after Sulla’, Hesperia, Suppl. xii (1967), 75 with n. 52.Google Scholar

^{page 182 note 1} Compare M. Ulpius Eubiotus, a dis-tinguished consular of the early third century A.D., who after his consulship was honoured in Athens with the eponymous archonship, and enjoyed the additional dis-tinction of having his sons serve under him as (IG ii². 3697–3702). There is no suggestion that he held any Athenian magistracy other than the archonship. Similarly C. Iulius Philopappus is attested archon and agonothetes shortly before 87/8 (IG ii². 3112; for the date see Graindor, , Chronologie des archontes atheniens sous I'Empire [1920], 98–100Google Scholar). He cannot have been in residence long. Josephus, BJ 7. 240–3 indi-cates that his family spent long periods in Sparta and Rome after their expulsion from Commagenein 72. There is no record of any magistracy other than the archonship held by Philopappus.

^{page 182 note 2} xIG ii.² 4251–3 = T 25 (Roos).

^{page 182 note 3} As argued by Graindor, , ‘Marbres et textes antiques d’époque impériale’, Recueil de travaux publiés par la Faculté de philosophie et lettres: Université de Gand, Fasc. 50 (1922), 49–52.Google Scholar Stein, PIR ² F 219 (p. 138) is more sceptical.

^{page 182 note 4} L. Annius Arrianus (cos. 243); L. Claudius Arrianus (inc. ann.); Arrianus Aper Veturius Sabinus (inc. ann.).

^{page 182 note 5} Phot. cod. 58, p. 17 b 12; Suda s.v. = T 2 (Roos).

^{page 182 note 6} On 1 Dec. 147 Marcus received tribunician power (renewed on 10 Dec), I’Empire supplemented by proconsular imperium outside Rome and the ius quintae relationis (I.I. xiii. I. 207; SHA, Marcus 6. 6). From that time onwards he was virtual co-regent with Antoninus (cf. Birley, A.R., Marcus Aurelius [London, 1966], 134–5CrossRefGoogle Scholar). The Suda interestingly claims that Arrian was at Rome and this inversion of the chronological order has been variously emended (—Jacoby and Roos; cf. Wirth, , Historia xiii (1964), 506–7Google Scholar). It could just be that the original dated Arrian to the reign of Hadrian and the joint reign of Marcus and Antoninus Pius.

^{page 183 note 1} T 22 (Roos); Jerome dates Arrian to the 12th regnal year of Pius, the Armenian version to the 11 th.

^{page 183 note 2} cf. Schwartz RE ii. 1232.

^{page 183 note 3} Ep. ad L. Gellium 2–4.

^{page 183 note 4} As argued by Wirth, Th., Mus. Helv. xxiv (1967), 149–61.Google Scholar

^{page 183 note 5} Cf. Schwartz, RE ii. 1332–3; Pelham, H.F., Essays on Roman History (1911), 22O–6.Google Scholar The PeriPlus falls neatly in to two halves, the report to Hadrian ending at 11. 5, to be followed immediately at 12. 1 with the transcription of the geographical treatise.

^{page 183 note 6} Cf Schwartz, RE 11. 1233. Kiechle, Fr., Berlcht der rörmsch-germamschen Kommission des deutsch. archäol Inst. xlv (1964). 108–14Google Scholar, argues that there were reasons other than antiquarianism for this resuscitation of Hellenistic tactical theory, and suggests that phalanx tactics were reintroduced to counter the barbarian cavalry hordes which menaced the eastern frontier. The opening of the Tactica is lost and there is no means of proving the theory; but, for what it is worth, Arrian does claim a utilitarian function for the Periplus (17. 3).

^{page 184 note 1} For this dating see Millar, F., JRS lv (1965), 142Google Scholar. The crucial peg is the identification of the Maximus who appears in Epictet. 3. 7 3 and 10 as …, with the corrector Sex. Quintilius Valerius Maximus, recipient of Pliny 8. 14 (written in 108).

^{page 184 note 2} Plassart, A., Fouilles de Delphes iii (1970)Google Scholar 38 ff. (nrs. 290 and 294). Plassart's discussion of Avidius Nigrinus’ boundary regulations is now fundamental (see Magie, RRAM ii. 1453, n. 11, for earlier literature). In particular, his fragment 4961 (reproduced PI. ix) contains the remnants of three names, each clearly punctuated: ]us Pollio, Q..Eppius, and Fl. [Arrianus]. The punctuation on this fragment plainly distinguishes Q. Eppius from Fl. Arrianus, and clearly two individuals are at issue. The composite Q,. Eppius Fl. Arrianus, who figures in earlier editions of the stone, must therefore disappear from scholarship. The date of Nigrinus' activity cannot be fixed exactly, despite Plassart’s arguments (p. 41). In the Latin text of the inscription Trajan is constantly described as Optimus Princeps, but unfortunatelv that is insufficient reason for Quintilius 4) dating the dossier to autumn 114, when Trajan was voted the agnomen, Optimus. Trajan’s full imperial titulature does not appear in any of the texts of Nigrinus’ decisions, and Optimus is never used as a name. Throughout the emperor merely figures as optimus princeps, and Trajan had been so described, officially and unofficially, ever since the beginning of his reign (Pliny, Pan. 2. 7; 88. 4–6; Ep. 2. 13. 8, 4. 22. 1; Mattingly- Sydenham, RIC ii. 250 ff., nrs. 91 ff.). The vacillation of the Greek texts between (nr. 293) and (nr. 295) is easily explained. Trajan, like Otho, sometimes figures as optimus maximusque princeps (ILS 6675; cf. ILS 5947), and there is no problem in the appearance of both epithets individually in the Greek text of the Nigrinus dossier. The most probable date for Nigrinus’ activity as legatus pro praetore in Achaea remains the period immediately before his suffect consulship of no, just the time at which Arrian is known to have studied with Epictetus (von Premerstein, SB Bayer. Akad. 1934, Abh. 3, p. 42 n. 3). The consensus of opinion, endorsed by Plassart (p. 46), is that the Flavius Arrianus of the consilium should be identified with Arrian the historian. Stein, but PIR ² F 219, doubted the identification: ‘Neque facile Arrianus … qui imperatoris Marci aetate etiamtum prytanis Atheniensium fuit, sexaginta annis ante in consilio legati esse potuit.’ True enough, but once one abandons the identification of the prytanis of 166/7 with Arrian the historian, these chronological difficulties disappear.

^{page 185 note 1} For testimonia of the legateship of Cappadocia see T 8–17 (Roos); PIR ² F 219. Arrian is last attested in the province in 137 (IGR iii. i n = Smallwood 204); by the time of Pius' accession he had been replaced by L. Burbuleius Optatus Ligarianus (ILS 1066 = Smallwood 194).

^{page 185 note 2} The mysterious may be a late work, composed after Arrian's victory over the Alani in Cappadocia. This book seems to have contained a discussion of the Caspian Gates, where Arrian is known to have operated (Themist. Oral. 34, p. 451 D; Dio 69. 15. 1); cf. Lydus, de mag. 3.53 = Parthica F 5. If this work included the extant (so Schwartz, RE ii. 1233– Jacoby FGrH ii. D. 563), we should know for certain that it was composed after the legateship; but unfortunately it cannot be assumed that the “EKTatjis came from historical work in the first place (Roos, Opera Minora, xxxi), and there is no certain extant fragment of the . All that be said is that the title strongly suggests composition after 135.