¹ It is the theme of the introductions to Books 1 and 3, put specifically in 1. 1. 5; 2. 7; 3. 1. 4, and reappears in 1. 63. 9–64. 1, in 6. 2. 3, in the introduction to the discussion of the Roman polity, and at the end of the epilogue, 39. 8. 7. The qualification in 3. 4. 3, ‘it seemed to be generally agreed’, is not to the theme of Roman invincibility, which he asserted unequivocally in 1.2. 7, but refers to his opinion that to evaluate the position of Rome one must not stop at Pydna but consider how she behaved to her subjects in the following generation, since ‘great successes can bring great disasters if states do not use their powers wisely’. Polybius' point of view appears very clearly in his interpretation of the Achaean revolt, which is comprised in the tail-piece of the whole work (e.g. 38. 1. 5; 9. 6–8; 10. 11–13; 11. 1; 12. 3; 16. 9; 18. 7–8). See further the comments of Walbank, F. W., Historical Commentary on Polybius I (1957), 40–2, 129–30, 301Google Scholar.

² Modern accounts of the foreign policy of the Anatolian kingdoms after Pydna are somewhat selective and discontinuous. Magie, op. cit. 1, covers most of the field briefly in his narrative text: 26–33, Pergamum; 193–4, Pontus; 201–3, Cappadocia; 315–18, Bithynia. Benecke, P. V. M., CAH VIII, 279 fGoogle Scholar. is remarkably brief. Rostovtzeff, M., Social and Economic History of the Hellenistic World (1941)Google Scholar, does not study political relationships in detail, but has strong and influential views about them. Will, E., Histoire politique du monde hellénistique II (1967), 319–24Google Scholar is penetrating and independent, but perforce omits much. Hansen, E. V., The Attalids of Pergamum² (1971)Google Scholar, gives a full account of the activities of Pergamum, but the interpretation is mostly derivative. Badian, E., Foreign Clientelae (1958), 99 f.Google Scholar, is brief.

³ Pol. 30. 3. 6–8; Livy 45. 34. 10–14.

⁴ Livy 45. 44. 21; Diod. 31. 14; Pol. 30. 2; 30. 28; For contemporary records of his victory, including OGIS 299, 763, and its importance, see Magie, op. cit. II, 766, n. 63.

⁵ Pol. 30. 30. 7; 31. 1. 6; 15. 10; 32. 1–2.

⁶ Pol. 32. 15–16, cf. App., Mithr. 3; Diod. 31. 35; OGIS 323, 15–22, confirming Polybius. Cf. also Pol. 33. 1. 1–2; 13. 4–5.

⁷ Pol. 33. 12–13; App., Mithr. 3.

⁸ Pol. 36. 14; App., Mithr. 4–7; Strabo 13. 4. 2 (624). Cf. OGIS 327, celebrating ‘Attalus and those who marched with him against Prusias and besieged Nicomedia’.

⁹ Pol. 31. 3; Diod. 31. 19. 7; Livy 42. 19. 3–6; 29. 4.

¹⁰ Pol. 31. 8, 32. 3.

¹¹ All this is relatively well-documented: Pol. 32. 10–12; Diod. 31. 32–32b; App., Syr. 47; Justin 35. 1. 2. Cf. Magie, op. cit. II, 1097, n. 9, on the discrepant version of Livy, Ep. 47: ‘a senatu restitutus est’.

¹² For the war against Prusias, Pol. 33. 12–13. For Priene, 33. 6. 6–8, and the fragmentary OGIS 351, in which the Senate merely instructs a Roman magistrate to write to the kings. Magie, op. cit. I, 117, ignores Polybius' statement that the Senate ‘paid no attention’.

¹³ OGIS 315, VI, with Dittenberger's notes ( = Welles, Royal Correspondence, no. 61).

¹⁴ For Hansen, op. cit. (n. 2), 132, the letter shows that Attalus was determined to keep the friendship of Rome, and nil ultra. Magie, op. cit. 1, 280: ‘Attalus, whose policy it was to be guided in such matters by Roman wishes’, with reference to the beginning of his reign; ibid. 27, the decision of the letter ‘was a distinct step towards closer relations with the Senate’. He adds that Attalus engaged in military activity in regions where Roman interest was not concerned, but does not probe beyond actions at Selge and Attaleia in Pamphylia.

¹⁵ In 181, the intransigent Pharnaces, during his war with the Pergamene coalition, which eventually defeated him, sent an emissary to Rome, but paid no heed to a Roman commission: Livy 40. 20. 1; Pol. 23. 9. 1; 24. 1. 1–3, 5. 1. The first recorded mission after Pydna is that implied c. 124 by Gellius, , NA II, 10Google Scholar, which is followed belatedly by that of 103–2 (see n. 61 below).

¹⁶ App., Mithr. 10. In OGIS 375 (ILS 30) ‘Mithridates M.f. Philopator Philadelphus’ records his alliance with Rome. Two coins (Recueil ² 1. nn. 2–7) and an inscription (Inscr. Délos 1555) combine to identify him as the brief successor of Pharnaces c. 159; cf. Magie, op. cit. II, 1090 nn. 46, 48, 49, and Larsen, J. A. O., Cl. Phil. 51 (1956), 157 f.Google Scholar, against the older identification with a son of Mithridates Eupator c. 80 (for which see Dittenberger ad loc).

¹⁷ cf. Magie, op. cit. 1, 20, on the Attalids (and n. 14 above); 202, the Ariarathids; 315–17 on Prusias II and Nicomedes II. Rostovtzeff, op. cit. (n. 2), 801–3, 827, regards Pergamum and Bithynia as reduced to vassalage. This all goes back to Mommsen, , History of Rome III (1894), 234 fGoogle Scholar. The subtle Will considers this a confused period in which Rome failed to exert her power consistently, though the kingdoms were dependent on her grace and favour, op. cit. (n. 2) II, 302, 312 f., 320. For Hansen, op. cit. 141, both Eumenes II and Attalus II were vassals. For Th. Liebmann-Frankfort, see n. 18 below. Only R. B. McShane, The Foreign Policy of the Attalids of Pergamum (1964), 190, briefly denied that the Attalids were ‘subservient’ to Rome in Asian affairs.

^17a See n. 75a below.

¹⁸ Liebmann-Frankfort, Th., La frontière orientale dans la politique extérieure de la République romaine (1969)Google Scholar, puts her views together. The theme is worked out inch. 1–2. The kings are turned into ‘satellites’, ‘absorbed’, and even ‘integrated’, as an alternative to annexation or permissive expansion. Cf. 101 f, Bithynia; 103 f., Pergamum; 108 f., Cappadocia. So, e.g., when Attalus II restores Ariarathes, it must be on the advice of Rome (114–5), despite Pol. 32. 10–12.

¹⁹ Pol. 21. 18–3, cf. Livy 37. 52–4.

²⁰ Pol. 21. 7–9; 22. 13; 23.2–5.

²¹ Pol. 21. 22. 15.

²² Sail., Hist. IV, fr. 69. 8.

²³ cf. e.g. the frequent missions to Macedonia, Achaea and the Anatolian kingdoms leading up to the war with Perseus: Livy 42. 17. 1; 19. 7–8, 26. 7–8, 37; 45. 1–5.

²⁴ Pol. 31. 15. 7–11.

²⁵ cf. nn. 6–8 above, and the notorious criticism by M. Cato of the mission that failed to save Prusias in 149, Pol. 36. 14. 4–5. If factional or family interests underly some of the ambiguities of Roman behaviour, this only emphasizes the absence of an overriding public interest.

²⁶ cf. Chapot, V., La province romaine proconsulaie d'Asie (1904), 5, 9, 10–11Google Scholar; Magie, op. cit. 1, 31–2, 147; 11, 780 n. 91; Rostovtzeff, op. cit. (n. 2) n, 807. Even Last, H. M., CAH IX, 103Google Scholar, thinks that there was no hesitation, though there should have been. Hansen, op. cit. 148, ‘the only logical course he could follow’. Will, op. cit. (n. 2) II, 330, as usual is more puzzled.

²⁷ Diod. 33. 14, 15; Strabo 13.4. 2(624). OGJS 330 may date this to 145, cf. Dittenberger ad loc., Hansen, op. cit. (n. 2), 139.

²⁸ Th. Mommsen, op. cit. (n. 17), III, 278. His view prevails; cf. especially Magie, op. cit. 1, 32. McShane's notion, op. cit. (n. 16) 194, that Attalus wanted to end the tension between Roman and Pergamene power is hardly intelligible.

²⁹ SEG IX 7. Cf. Will, op. cit. (n. 2), II, 305 f., for a survey and bibliography.

³⁰ App., Mithr. 71; B.C. 1. 111; Livy, Ep. 93; Eutropius 6.6. Cic., de leg. agr. 2. 40, 50, confirms Sallust, , Hist. IV. fr. 69Google Scholar. 9: ‘Bithyniam Nicomede mortuo diripuere (sc. Romani), cum filius Nysa quam reginam appellaverat genitus haud dubie esset’. He does not allege that this was a recognized and legitimate son; cf. ibid. II. fr. 71, for the rejection of his claim.

³¹ Diod. 34. 3; Justin 36. 4. 1–3. Cf. Strabo 14. 1. 39 (647), with Hansen, op. cit. (n. 2), 144, n. 55.

³² Magie, op. cit. II, 778 n. 87 minimizes the numbers of relatives. That Attalus remained unmarried for some years after the death of Berenice does not mean that he had no intention of remarriage.

³³ Sallust, BJ 9. 3.

³⁴ For the birth of Attalus cf. Magie, op. cit. II, 772 n. 76. Hansen, op. cit. (n. 2), 471 f., on the possible implications of Pol. 30. 2. 5, 33. 18. 2.

³⁵ OGIS 338.

³⁶ Plut., Ti. Gracchus 14. 2; Livy, Ep. 59: ‘cum testamento Attali regis legata populo Romano libera esse deberet.’ For discussion and bibliography cf. Will, op. cit. (n. 2), 351 f.; Magie, op. cit. I, 32–3, II, 780–1 nn. 92–4. It is possible that the bequest was limited to the ager regius and the townships other than Greek cities.

³⁷ RE I, 190 f. Rostovtzeff, op. cit. (n. 2) II, 749 f., and III, 1506, n. 15, for bibliography. Cf. the notable letter of the proconsul after a local revolution at Dyme, SIG ³ 685 (= Sherk, Roman Documents, no. 43), with Paus. 7. 16. 10.

³⁸ For a recent survey see Will, op. cit. (n. 2) II, 342 f.

³⁹ Diod. 33. 28 a 2–3; Strabo 14. 5. 2 (669). For the mission cf. Astin, A. E., Scipio Aemilianus (1967), 127, 177Google Scholar. Liebmann-Frankfort, op. cit. (n. 18), 133, oddly thinks the disappearance of menaces to northwest Anatolia now made it possible to create a province.

⁴⁰ The date of the death of Attalus III is commonly attributed to spring or summer 133 with unjustified confidence (e.g. Magie, op. cit. II, 781, n. 94). The back-dating of the Roman era on republican cistophori to September 134 seems to suggest an earlier date (CIL 1². 2, p. 761–2: Kubitschek, , RE I 637Google Scholar). The sole other criterion is the length of reign given by Strabo 13. 4. 2 (624) for Attalus II, twenty-one years, and Attalus III, five years, to be dated from c. March 159, when inscriptions indicate that Eumenes II was still alive (Hansen, op. cit. 127). It is not clear that Strabo's years are completed regnal years. If there is an overlap, the fifth year of Attalus III, counted from 139–8, must end in 134. Carcopino, J. was perhaps right about the sunstroke and the summer but wrong about the year, Autour des Gracques² (1967), 34 fGoogle Scholar.

⁴¹ Strabo 14. 1. 38 (646) distinguishes the stages of Roman intervention carefully, pace Magie, op. cit. II, 1037 n. 10. The mission of five arrives after the kings and cities have taken action against Aristonicus and before the despatch of the consul of 131. So too Liebmann-Frankfort (op. cit., 140) observes the tactical delay. Eutropius 4. 20, Orosius 5. 10. 1–2 are less exact.

⁴² Strabo loc. cit., Livy, Ep. 59. Cic., Phil. II. 18, for the quarrel. This led to a iudicium populi and to the reallocation of the consular province by a lex that instituted a direct and open election (surprisingly), at which P. Crassus was preferred to Scipio Aemilianus, although Crassus, famous for civil virtues, lacked any military ability and as pontifex maximus should have not have left Italy (Gellius, , NA I. 13. 10Google Scholar; Livy, Ep. 59).

⁴³ OGIS 435. Though it would suit my view to take the magistrates for commissioners, the term στρατηϒοί can only mean praetores in a public document of this date, pace Vogt, J., Atti del terzo congresso int. epigr. gr. Lat. (1959), 45 f.Google Scholar, whose argument that the phrase εἰς Ἀσίαν πορευόμενοι must refer to legati is contradicted by the Cnidian and Delphian texts of the Law, Piracy, JRS 64 (1974), 204Google Scholar, col. IV 9–10; FIRA ² I, 9. B. 28–9. The date may be any year after 133 when both consuls were out of Rome in September and October, since a praetor, otherwise unknown, presides. Cf. Magie, op. cit. II, 1033, n. 1; Broughton, MRR I, 496Google Scholar; Drew-Bear, T., Historia 21 (1972), 75Google Scholar.

⁴⁴ Justin 37. 1.2; 38.5. 3; App., Mithr., 57. The text of Justin 37. 1. 2, which has Syria for Phrygia, also adds Cilicia to Lycaonia, which makes no sense, unless one boldly amends it to Pisidia. Cf. Magie, op. cit. II, 1044 n. 28; A. H. M. Jones, CERP ², 131.

⁴⁵ Magie, op. cit. I, 125 f., 276–7.

⁴⁶ cf. Magie, op. cit. II, 1042 n. 26, 1048 n. 39; CIL 1², 646. There is no other direct evidence at this date. By 113 Pisidian Prostanna was within the province, cf. Inscr. Délos 1603; Magie, op. cit. 11, 1161, n. 12.

⁴⁷ Magie, op. cit. 1, 157–8; II, 1048, nn. 39–40. For the milestones, CIL I². 3, 646–51. A new stone from the Burdur region adds nothing substantial; cf. Annual Report of the British Institute of Ankara 1975, 10.

⁴⁸ Cyzicus, a city state with extensive territory, was free c. 133 (IGRR IV, 134, ll. 18 f.), and reappears as free after the first Mithridatic war (Plut., Luc. 9. 1; App., Mithr. 73; Diod. 38/9. 8. 3; Strabo 12. 8. 11 (575–6); Magie, op. cit. II, 1111 n. 4). Lampsacus, free earlier and never part of the Pergamene kingdom, became provincial after that war: Livy 43. 6. 8–10; SIG ³ 591; Cic.;, Verr. 2. 1. 81; Magie, op. cit. II, 947 n. 51. For Ilium cf. Magie, ibid. 950, n. 60; whether its freedom is earlier than Sulla's settlement is not certain. See Strabo 13. 1. 27 (594–5); IGR IV, 194, cf. Jones CERP ², 60–3, 86–7.

⁴⁹ See Sherwin-White, A. N., ‘Rome, Pamphylia and Cilicia 133–70 B.C.,’ JRS 66 (1976), 3Google Scholar n. 6.

⁵⁰ If Q. Mucius went to Asia after his consulship, as Badian, E. argues, Athenaeum N.S. 34 (1956), 104Google Scholar, it was not a regular assignment lege Sempronia, because he was at Rome after the campaigning season of 95, when he vetoed his colleagues' triumphus (Asc. 14C; Cic., de Invent. 2. 111). See now Marshall, B. A., Athenaeum N.S. 54 (1976), 117Google Scholar, against Badian.

⁵¹ App., Mithr. II, 17, 19: Cassius and Aquilius in 90–89 raise a great army of ‘Phrygians and Galatians’ from ‘Bithynia, Cappadocia, Paphlagonia and Galatia’. Cassius also had a small army, but no legions are mentioned. Cf. Memnon, , FGrH 434, (22), 7Google Scholar, ‘with few Romans’; Justin 38. 3. 8, ‘Asiano exercitu instructos’. So too L. Sulla assisted the restoration of Ariobarzanes ‘with few troops of his own but eager allies’, Plut., , Sulla 5. 7Google Scholar. For the possibility that he was the regular propraetor of Asia see my discussion, op. cit. (n. 49), 8–9.

⁵² ibid., 4–5, with bibliography.

⁵³ Strabo 14. 5. 2 (668–9). For the new text of the Piracy Law, cited as ‘Lex Cnidia’, see Hassall, M., Crawford, M., Reynolds, J., JRS 64 (1974), 195 f.Google Scholar, and for the text from Delphi, ibid, and FIRA ² 1, 121, no. 9. Whether this is one law or two laws of the same date is here immaterial. For Cilicia, Lex Cnidia III. 30–40, completing Lex Delphica B. 7–8.

⁵⁴ See my detailed discussion op. cit. (n. 49), 6–8, and 4, n. 11. Add that the Livian tradition of Eutropius 6. 1. 1 and 3. 1 succeeds in distinguishing between Cilicia and Pamphylia in the context of pirate wars.

⁵⁵ cf. Sherwin-White, op. cit. (n. 49), 2–3.

⁵⁶ For Phrygia, App., Mithr. 57; Justin 38. 5. 3, with OGIS 436, which may date its removal to 119 or 116 according to the restoration of the name or names of the presiding magistrate(s). Cf. T. Drew-Bear, op. cit. (n. 43), 79 f. For Lycaonia see Lex Cnidia III, 22 f., where its resumption precedes the law of 101–100.

⁵⁷ cf. the dispute, in 129 or 101, in the s.c. de agro Pergameno, Greenidge and Clay, op. cit. 278; R. K. Sherk, Roman Documents from the Greek East, no. 12; cf. H. B. Mattingly, AJP 93 (1972), 412 f.; at Priene c. 98–1, Inschr. Priene, no. 111, 1.112 f.; at Oropus in 73, Syll ³ 747, 24–30, and Ilium, ILS 8870. Cf. the new evidence for publicani and annexation in Thrace, Lex Cnidia IV, 15–18; also, later, Memnon, FGrH 434, (27), 5–6.

⁵⁸ Reinach, Th., Mithridate Eupator (1890), 49–106Google Scholar, is still the basic reconstruction, enlarged only for the Crimea by M. Rostovtzeff, CAH IX 225 f., and summarized by Magie, op. cit. I, 195 f.

⁵⁹ Strabo 7. 3. 18 (307), 4. 7 (312); 12. 3. 1 (541), 28 (555); cf. also 2. 1. 16 (73); Justin 37. 3. 2 and Prol. 37. For the record of Diophantus, Syll. ³ 709.

⁶⁰ Memnon, FGrH 434, (22), 3–4, cf. Reinach, op. cit. (n. 58), 95–6.

⁶¹ Justin 37. 4. 4–9. The traditional date c. 104 depends on connecting Paphlagonia with the embassy of Mithridates to Rome known from Diod. 36. 35, one or two years before the second tribunate of Saturninus. Cf. Reinach, op. cit. 95 f.; G. Daux, BCH 57 (1933), 81.

⁶² Nicomedes still held Paphlagonia at the time of the replacement of Ariarathes IX by Ariobarzanes in c. 96 (n. 67 below), Justin 38. 4. 6–7. I omit consideration of Mithridates' occupation of Galatia at this time, according to Justin loc. cit., for lack of supporting evidence. Possibly this refers to the territory of the Trocmi beyond Halys, in which Mithridates eventually built Mithridation (Strabo 12. 5. 2 (567)).

⁶³ Justin 38. 1, with Memnon, FGrH 434, (22), 1, elucidated by Reinach, op. cit. 97 f. Cf. Magie, op. cit. 1, 203, and his notes.

⁶⁴ Justin 38. 2. 3–7. I omit the vain attempt of the murdered king's brother (Ariarathes VIII) to expel the false Ariarathes, ibid. 1–2.

⁶⁵ The Cappadocian coinage has been resurveyed and recatalogued by B. Simonetta, Num. Chron. 1961, 9 f., with some modification of the data on which Reinach established his chronology, cf. n. 67 below.

⁶⁶ The date results from OGIS 353 and Inscr. Délos 1576, 1902, cf. G. Daux, op. cit. 81 f.

⁶⁷ Coins record his regnal years 2–5 (but not 6), 12, 13 and 15, Simonetta, op. cit. 18. All scholars seem to take the break after ‘five’ to mean that Ariarathes IX was expelled in or after his fifth year and restored not later than his twelfth year (c. 90–89). This is not affected by the radical arguments of O. Mørkholm about other aspects of the coinage, Num. Chron. 1962, 407 f.; 1964, 21 f.; 1969, 26 f.

⁶⁸ Justin 38. 2. 6–7, with Strabo 12. 2. 11 (540).

⁶⁹ The arrangement of Badian, E. (Athenaeum N.S. 37 (1959), 279 f.Google Scholar, reprinted in his Studies in Greek and Roman History (1964), 56 f.) implies that the Senate took a strong line with Mithridates in Cappadocia from the start, using force instead of diplomacy. I have criticized it in a forthcoming article to appear in CQ 1977. For the older view, cf. Magie, op. cit. 1, 206, following Reinach, op. cit. (n. 58), 105.

⁷⁰ Plut., , Sulla 5. 6–7Google Scholar is the principal source, with Livy, Ep. 70; App., Mithr. 57.

⁷¹ Justin 38. 5. 6, confirmed by App., Mithr. 57, not noticed by Badian, op. cit. (n. 69).

⁷² Granius Licinianus 35. 30 (F), badly summarized by App., Mithr. 10; Memnon, FGrH 434, (22), 3. Cf. Magie, op. cit. 1, 207, 11, 1099 n. 19. Reinach, op. cit. (n. 58), 114, did not know the Flemisch text of Licinianus, which still leaves much obscure.

⁷³ App., Mithr. 10; Justin 38. 3. 4.

⁷⁴ This summarizes the story of App., Mithr. 11–19. The Livian epitomators (Ep. 76; Florus 1. 40. 3–6; Eutropius 5. 5; Orosius 6. 2. 1–2) are very thin, with a different emphasis, omitting the role of Aquilius almost entirely, which reappears briefly in Justin 38. 3. 4 and 8, Memnon, FGrH 434, (22), 7. Cf. n. 86 below.

⁷⁵ cf. the speech attributed to Mithridates by Pompeius Trogus in Justin 38. 4–7, on the theme ‘Romanos posse vinci’, citing Aristonicus (6. 4), Jugurtha (6. 6), Cimbri (4. 15).

^75a From 190 to 168, according to Afzelius' evaluation of the detailed evidence of Livy, from eight to ten legions, with their allied complement of five to eight thousand men apiece, were regularly deployed in the two Spains, north Italy, and in some years Sardinia, in consular and praetorian commands. The figure rises to twelve legions during the oriental wars, which required armies of four legions in some years, found in part by cutting down the garrison of north Italy, while two legions remained around Rome as a short-term strategic reserve. This figure, with the Italian complement, gives the maximum potential of Roman manpower under the traditional system. After the termination of Livy, statistical information disappears. A few isolated figures suggest that the standard consular and praetorian armies remained much the same in size down to 91. Pressure did not abate after 150, when the African war required five consuls out of six from 149 to 147, and renewed troubles in Spain took two consular armies each year, under consuls and proconsuls, from 143 to 134. Meanwhile Macedonia became a praetorian commitment from 146. Between 125 and 120, the conquest of Transalpine Gaul occupied four consuls, conjointly in some years. These overlapped with two consuls operating in Sardinia (126–2) and Nearer Spain (123–c. 121). Macedonia required consular attention from 114 to 107, overlapping with the Numidian war from 111 onwards, and with consular commands in north Italy in at least 113 and 109. The Numidian command in turn overlapped with consular activity in Gallia Transalpina from 107 to 105 when two armies were on foot in Gaul and a third in reserve in north Italy. Thus the sole intervention in Asia (131–29) fell in a rare quiescent period: the annual Roman requirement from 125 onwards could hardly be less than nine legions. See, for the period before 167, Afzelius, A., Die römisch Kriegsmacht (1944), 47 f., 62 f., 78–9Google Scholar. For 146–101, Ilari, V., Gli Italici nelle strutture militari romane (1974), 167Google Scholar (with Strabo 4. 1. 1. added); Brunt, P. A., Italian Manpower (1971), 427 fGoogle Scholar. For the consular commands from 150 on, see Broughton, T. R. S., Magistrates of the Roman Republic² 1Google Scholar, under each year. For consular armies see n. 78 below. For praetorian commands, Livy 39. 30. 12; 40. 36. 8; 41. 5. 6–7, 21. 2, suggests a norm of one legion with socii, which frequently escalated by the retention of legions from year to year.

⁷⁶ cf. Justin 38. 4. 16: ‘etiamsi singula bella sustinere Romani possint, universis tamen obruantur ut ne vacaturos quidem bello suo putet’, a view that Trogus or his source evidently found tenable.

⁷⁷ Sulla has six legions in Campania in 88 (App., B.C. 1. 57), and takes five legions and some extra units to Greece and back (Mithr. 30; B.C. 1. 79). In Greece he acquires an extra legion mysteriously in 86, probably withdrawing it from Macedonia, Plut., , Sulla 15. 4–5Google Scholar; Memnon, FGrH 434, (22), 12. Valerius Flaccus took out only two legions, which Sulla left behind for Murena (App., Mithr. 51, 64). L. Lucullus took out only one legion to add to four legions then stationed in Asia and Cilicia, Plut., Luc. 7. 1, 8. 4; App., Mithr. 72. Aurelius Cotta seems not to have a consular army in addition to his fleet. After infantry losses of three to five thousand men at Chalcedon, little was left to join Lucullus: App., Mithr. 71; Plut., Luc. 8. 2; Memnon FGrH 434, (27), 7–8.

⁷⁸ Pol. 6. 19–20, 21. 4, 6, 26. 3, 7 f. Cf. A. Afzelius, op. cit. (n. 75^a), 34 f., 62 f.; V. Ilari, op. cit. (n. 75^a), ch. VI; P. A. Brunt, op. cit. (n. 75^a), 681 f.

⁷⁹ cf. my discussion op. cit. (n. 49), 4–5, nn. 9–14. Sulla, arriving in Greece in 87 without a fleet after the surrender of the Asiatic flotilla in 89, was confined to land operations in Achaea until his quaestor Lucullus returned, in the winter of 86–5, with a naval force collected with difficulty from Syria, Rhodes and Pamphylia: App., Mithr. 17, 19, 33, 51; Plut., Luc. 2. 2–3, 3. 1–3, 4. 1.

⁸⁰ App., Mithr. 13.

⁸¹ App., Mithr. 51.

⁸² App., Mithr. 19. At Chaeronea the sources criticize the tactics of Archelaus, but not the valour of his men, who crack only in the final rout: App., Mithr. 42–4; Plut., , Sulla 17. 9–19. 8Google Scholar. Their skill in siege warfare was outstanding, App., Mithr. 34–7, 40.

⁸³ App., , Mithr. 65, 89Google Scholar; Plut., Luc. 35. 1–2. The loss of twenty four tribunes and one hundred and fifty centurions at Zela indicates a major disaster, even if exaggerated by the friends of Pompeius.

⁸⁴ After increasing trouble with the Scordisci from 118 onwards (SIG ³ 700), Macedonia became a consular province from 114 to c. 107: Livy, Ep. 63, 65; Florus 1. 39. 4–5; ILLRP 1 337; Fasti Triumph. for 106. After the campaign of the praetorian T. Didius c. 102–1, the territory of the Caeni was annexed, cf. Lex Cnidia IV, 5–30.

⁸⁵ Above, p. 69 f.

⁸⁶ The alert will notice a revision of the chronology of Reinach for the beginning of the war, hitherto unchallenged despite the difficulties that it creates (cf. Will, op. cit. n, 398-9). Reinach, op. cit. 112 f., attributed the campaigns of Mithridates in Anatolia, the siege of Rhodes and the invasion of Achaea to 88, when the crisis of the Social War was passed, so that in 90–89 Mithridates, taken by surprise, ‘missed the bus.’ Reinach based his dates on the serial order of events in Livy, Ep. 76–9, despite its ambiguities and his rejection of one of its statements. He ignored the order of events in the detailed narrative of Appian (Mithr. 17–21), which places the warfare in Anatolia and the occupation of Asia before the consular elections of 89 and the assignment of Asia as a consular province; while the siege of Rhodes, mopping-up in Lycia and Paphlagonia, the despatch of Archelaus to Achaea and his clash with the proconsul of Macedonia, are set in the year of Sulla's consulship (88), when political events at Rome prevented the normal departure of the consul with his army for the campaigning season. Appian's order of events makes much better sense of this protracted series of campaigns, but the matter needs discussion elsewhere. Livy's Epitomes and the other subsidiary sources are susceptible of various interpretations; so too the amended Olympic date in Mithr. 17. Orosius (5. 19. 2.), using Livian compendia, was justly puzzled about the year 88: ‘utrum abhinc primum coeperit an tunc praecipue exarserit (bellum).’

⁸⁷ Plut., Marius 31. Marius in 99–8, during his unofficial visit to Cappadocia and Galatia religionis causa, addresses Mithridates thus. Too much has been made of this ‘secret history’ by Luce, R. J., Historia 19 (1970), 162 f.Google Scholar, following E. Badian, op. cit. (n. 69), 279 f.

⁸⁸ App., Mithr. 15, 17, 19.

⁸⁹ Sallust, Hist. IV fr. 69. 10. Cf. Floras 1. 40. 3; Dio fr. 97; Justin 38. 5. 10.

⁹⁰ Oppius, whose provincia covered southern Phrygia and Lycaonia, was probably intended, like Sulla, to restore Ariobarzanes to Cappadocia (though he is not named in App., Mithr. 11), while Cassius in Lydia and northern Phrygia was wel l placed to assist Nicomedes in Bithynia. Cf. App., Mithr. 17, 20. This may be the first occasion of the division of the provinces, cf. my discussion op. cit. (n. 49), 9.

⁹¹ cf. Justin's technical language (38. 3. 4): ‘decernitur in senatu ut uterque in regnum restituantur, in quam rem missi M’ Aquilius etc.', confirming App., Mithr. 11. The use of force was authorized against Socrates (Justin 38. 5. 8): ‘regem Bithyniae Chreston in quem senatus arma decreverat’. There is not one word about war with Mithridates. Whe n Pelopidas, in the prolonged negotiations with Aquilius and Cassius, eventually proposes that the Senate should be consulted, the Romans promptly dismiss him and organize their offensive without referring the request to Rome, App., Mithr. 16–17. Earlier they pressed Nicomedes and Ariobarzanes to attack Mithridates' territory precisely because they lacked direct authority for this themselves, App., Mithr. 11: ‘and so to provoke Mithridates to war, because the Romans would support the kings as allies if they were at war.’