1. Holleaux, , Rome 173–305Google Scholar. Holleaux presented his views more briefly in CAH III.116–37Google Scholar; for convenience I cite this English version, but for the original French see Études V.295-319.

2. E.g. Scullard, H. H., A History of the Roman World 753-146 B.C. ⁴ (1980) 215–6Google Scholar; Walbank, , Philip 68–105Google Scholar (still the best narrative); Thiel 66-8, 90-106, 115-6, 122-5, 128-30, 135-8, 152-4; Badian, , FC 55–61Google Scholar; Will, II 82-100; Errington, , Dawn 109–118Google Scholar; Lazenby, J. F., Hannibal's War (1978) 157–168Google Scholar; Ferrary, J. in Nicolet, C. (ed.), Rome et la conquête du monde méditerranéen II (1978) 735Google Scholar; Briscoe, J. in Garnsey, P. D. A. and Whittaker, C. R. (edd.), Imperialism in the Ancient World (1978) 153–5Google Scholar; Gruen 373-381.

3. Notably Balsdon, J. P. V. D., JRS 44 (1954) 31Google Scholar and Harris 205-8; see also Larsen, cited below nn. 92, 119.

4. Polybius claimed that Philip, under the influence of Demetrius of Pharos, envisaged the conquest of Illyria and intervention in Italy as steps towards world rule (5.101.6-102.1; 105.1,5; 108.4-7). It is most unlikely that Philip aimed at world rule and he may never have planned independent action in Italy, but Badian, , Studies 19Google Scholar and Gruen 375 go too far in denying that his actions in 217-6 were anti-Roman.

5. Plb.7.9; Schmitt, Staatsverträge, no. 528.

6. For such views see especially Walbank, , Philip 71–2Google Scholar and Commentary II.44,55Google Scholar; Bickerman, E. J., AJP 73 (1952) 15–6, 22–3Google Scholar; Gruen 376. For a recent protest see Briscoe, , Commentary I.73Google Scholar.

7. The treaty provided that Philip should send help to Hannibal according to need and by mutual agreement: Plb.7.9.11 . The postponement of agreement on the sending of help does not indicate reluctance on either side, but was a necessary consequence of the military situation: a Macedonian force could not be sent to Italy until each side had secured an Adriatic port and a Carthaginian fleet could ensure its safety.

8. Plb.5.109-110.

9. Livy (A) 23.32.16-7, 34.3-4.

10. Livy (A) 23.38.7-13. Livy gives the number of warships as 50 (55 here in error), but the true total was probably lower: see Appendix I. According to Livy, the senate ordered that Philip should be watched, and, if appropriate, Laevinus should cross the Adriatic to keep him in his kingdom: 23.38.9-11, 48.3; 24.10.4; 31.7.4; cf. Zon. 9.4.3. These passages are linked to the distorted annalistic tradition about Philip's treaty with Hannibal (Livy 23.33.10-2; App. Mac. 1.2; Eutr. 3.12.3; Zon. 9.4.2), but there is no reason to doubt that from the time it learnt of the treaty the senate feared an invasion of Italy.

11. Livy (A) 24.40: Zon. 9.4.4; Plut. Arat. 51.2.

12. Livy (A) 24.40.17; Plb. 8.1.6. Laevinus' command was prorogued down to 211: Livy (A) 24.44.5, 25.3.6, 26.1.12.

13. On these operations see Walbank, , Philip 80–1Google Scholar; Badian, , Studies 20Google Scholar: Hammond, N. G. L., JRS 58 (1968) 18Google Scholar.

14. Plb.8.13-4. For his gains among the Illyrians after Lissus see Plb.8.14.10, Livy (P) 27.30.13; it is uncertain how far north they extended. The position of the fragment in the excerpta antiqua shows that Polybius narrated the capture of Lissus under 213. Walbank, , Commentary II.6Google Scholar leaves open the possibility that he ran the res Graeciae for 213 and 212 into one narrative, but this is most unlikely. Polybius departed from his annalistic arrangement only rarely and when there were powerful reasons to do so (cf. Walbank, , Polybius (1972) 112–3Google Scholar, and Commentary II.2-3 for the res Siciliae of 215-4). To run two years' campaigns together could only cause confusion.

15. Livy (P) 27.30. Philip also consolidated his control of Dassaretis now (Plb.8.14b). On the location of the Atintanes see Cabanes, 78-80.

16. Livy (P) 25.23.9 (giving the date); 26.24.1.

17. Livy 26.24.1-15: from Polybius (see below n. 56); in my view, there is no need to suppose that Livy inserted any material from annalistic sources in this passage. Part of the treaty is preserved on an inscription at Thyrrheum (IG. IX. 1².2.241 = Moretti, , ISE 87Google Scholar). For the ample bibliography on the treaty see Schmitt, Staatsverträge, no. 536; Musti, D., ANRW I.2. 1146–1153Google Scholar.

18. See Appendix II.

19. A point made by Laevinus himself according to Livy 26.24.2 (dismissed as annalistic by those who date the treaty to 212). For the view that this consideration carried weight see Walbank, , Philip 83Google Scholar; Balsdon (n. 3); Badian, , Latomus 17 (1958) 199–200Google Scholar. Holleaux's view (Rome 195-201; cf. Lehmann 46-50) that the Romans could have had the Aetolian alliance years earlier, if they had sought it, depends on dating the treaty to 212.

20. Livy 26.24.14-5.

21. Kahrstedt, U., Geschichte der Karthager III (1913) 485Google Scholar (humanitarian disapproval of the booty provisions!); Holleaux, , Rome 212Google Scholar (indifference), CAH VIII.125Google Scholar (dislike of committing Rome to an Eastern policy). Against such views see Badian, , Latomus 17 (1958) 205–6Google Scholar.

22. An explanation along these lines was suggested by Badian, , Latomus 17 (1958) 206–8Google Scholar: the senate usually considered important matters like the ratification of treaties under the presidency of the consuls and will have wished to give Laevinus the opportunity of presenting his treaty in person, but in the event Laevinus was not in Rome long enough during his consulship in 210 for the matter to be raised and so the ratification had to be postponed until the consuls of 209 entered office. Unfortunately, there are difficulties: Laevinus would surely have had time to introduce the Aetolian envoys to the senate before setting out for his new province of Sicily, and, if the senate needed more time to complete the process of ratification, this could have been done with the urban praetor presiding. As Hopital, R. G., RHDFE 42 (1964) 26Google Scholar n. 18 and Lehmann 44-5 observe, the delay is easier to explain if the treaty is dated to 212: ratification would then have been postponed until Laevinus' return in 210.

23. For the bibliography on these clauses see the works cited in n. 17. Polybius, followed by Livy, held that the Romans were entitled to all the moveable booty, but the text of the treaty on the inscription allows the Aetolians a share on joint expeditions. In my view, Polybius was mistaken.

24. Livy 26.24.5. The League territory lost in that war comprised Ambracus (to Epirus: Plb.4.63.2-3); Oeniadae, Phoetiae and perhaps Metropolis (to Acarnania: Plb.4.63.7-8, 64.4,65.5-11); and Phthiotic Thebes (Plb.5.99-100).

25. Livy 26.24.6,8,11. The hope of acquiring Acarnania was, Livy says, a major factor in the Aetolian decision, and the treaty included the provision that darent operam Romani ut Acarnaniam Aetoli haberent. At some time in the mid third century, Epirus and Aetolia had partitioned Acarnania (references and discussion in Cabanes, 61-2,91-3). The Aetolians had taken the eastern part, which they had retained except for their losses in the Social War (n. 24). The western part had become independent with the collapse of the Epirot monarchy ca. 232, and had subsequently been the object of at least two Aetolian attacks, against Medion in 231 (Plb.2.2-4) and against Thyrrheum sometime in the 220's (Plb.4.6.2,25.3). There is confusion in Livy 26.24.6. Livy evidently supposed that all Acarnania was once Aetolian.

26. Demetrius had deprived them of Boeotia (an independent ally) and Opus (a part of the League) in 237/6 (Feyel, M., Polybe et l'histoire de Beotie au III^e siecle avant notre ére (1942) 83–105Google Scholar; Etienne, R. and Knoepfler, D., Hyettos de Béotie et la chronologie des archontes fédéraux entre 250 et 171 avant J.-C. (1976) 331–7)Google Scholar. All Phocis had once been Aetolian, but most had been lost by 224: Flacelière 286-7; Walbank, , Commentary II.617Google Scholar. Pharsalus had once been Aetolian (Plb. 18.8.9,38.3.6), but had been recovered by Macedon before 217 (Plb.5.99.4). It is usually supposed that Aetolia acquired Hestiaeotis and Thessaliotis in the disorders following the death of Demetrius. If so, they must have been recovered very soon by Doson, (Fine, J. V. A., TAPA 63 (1932) 126–55)Google Scholar; but Daux, G., Studia Antiqua A. Salac… Oblata (1955) 35–9Google Scholar seems to me to have shown that there is no good reason to suppose that the Aetolians ever held these territories.

27. Livy 26.24.9. Livy's omission of the Messenians is probably accidental (Walbank, , Commentary II. 163Google Scholar). Also listed were Rome's Illyrian friends, Scerdilaidas and his son Pleuratus, who were already at war with Philip.

28. Sparta had been on the Aetolian side in the Social War and the treaty of alliance concluded then remained in force (Plb.9.31.3,36.8). Elis had long been a more or less dependent ally of Aetolia. Alienated by Aetolian conduct in 220, the Messenians had been on the other side in the Social War: but Philip's interventions in 215-4 had driven them into open hostility towards him (Plb.8.12.1) and as a result they will have resumed their old friendship with the Aetolians. Many scholars suppose that the treaty of alliance between Aetolia and Messenia (Plb.9.30.6, 16.13.3) was concluded ca. 213 (Holleaux, , Rome 203Google Scholar n. 3; Roebuck, C., A History of Messenia from 369 to 146 B.C. (1941) 84Google Scholar; Walbank, , Commentary II.169Google Scholar; Meyer, E., RE Suppl. XV.273Google Scholar), but the Aetolians can hardly have made such a commitment to a state which was at war with Philip before their own decision to renew war with him in 211.

29. For the debate see Plb.9.28-39; although we are not told the outcome, there is no reason to doubt that it was then that the Spartans decided to enter the war (contra De Sanctis 421). Plb.9.30.6 implies that Elis and Messenia had already agreed to join, but this may be rhetoric (cf. Walbank's note). We do not know whether any hostilities took place in the Peloponnese in 210. That all three states concluded treaties with Rome can safely be inferred from Plb.18.42.7; Livy (P) 34.31.5,32.16 (Larsen, , CP 30 (1935) 210–2Google Scholar; Badian, , FC 57–8Google Scholar; Lehmann 366-71; contra, Dahlheim, W., Struktur und Entwicklung des römischen Völkerrechts (1968) 221–9)Google Scholar. These treaties probably included provisions about the profits of war, and the obscure phrase urbium Corcyra tenus ab Aetolia incipienti in Livy's version of the Roman-Aetolian treaty (Livy 26.24,11; for the text see Tränkle 212 n. 69) may imply that the clauses of that treaty assigning territorial gains to the Aetolians did not apply to the Peloponnese (cf. Stiehl, R., Untersuchungen zur römischen Geschichte (ed. Altheim, F.), I (1961) 166–8Google Scholar, who, however, absurdly supposes that the Peloponnese was excluded from hostilities altogether).

30. The Messenian claim to Pylos (Achaean by 220: Plb.4.25.4) was advanced by the Aetolians in 209 (Livy (P) 27.30.13); it is a reasonable inference that the Messenians had been given an undertaking on the subject when they decided to enter the war. In 196 the Messenians claimed Pylos and Asine; the omission of Asine from Livy 27.30.13 is not proof that the Achaeans acquired it after 209 (as Niese, 646 n. 4). It is uncertain whether the Achaeans held Cyparissia (so Niese, 411 n. 1; Larsen, , GFS 327Google Scholar; contra. Roebuck, , History of Messenia 94Google Scholar n. 124).

31. Philip's campaign of winter 219-8 had deprived the Eleans of Psophis, Lasion, Stratus, Alipheira and Triphylia (Plb.4.70-3, 77-80). Their claim to Triphylia was asserted in 196 (Plb.18.42.7).

32. Cleomenes had come near to achieving the hegemony of the Peloponnese. At Plb.9.30.3-4, 37.7 the Spartan aim in the Social War is represented as the restoration of .

33. Allen 68 holds that the fleet had been gradually built up since the reign of Eumenes I. He is right that no weight can be placed on the silence of our patchy sources, but he does not succeed in explaining why a purely land-based power should have gone to the expense of constructing and maintaining a war fleet.

34. The most important evidence for this relationship is Attalus' financing of the fortifications of Elaus at some time before 219 (Plb.4.65.6). See further Allen 70-1.

35. On Attalus' agreement with Antiochus see now Allen 58-62. I am not convinced by the overall conclusion of Allen's discussion (28-65) of Attalus' activity in Asia down to 216, that Attalus was throughout concerned with strengthening his kingdom rather than expanding it.

36. Holleaux's fantasy (Rome 204-8) that Attalus had long cherished plans for westward expansion through war with Philip and had already solicited Aetolian co-operation before Laevinus approached them is sufficiently refuted by Attalus' failure to arrive in Greece before 209. However, Holleaux was right to explain Attalus' entry into the war in terms of the desire for territorial expansion. Other explanations offered by older historians are quite unconvincing (Holleaux, , Rome 206Google Scholar n. 1; Allen 68-9). So too is Allen's view (65-71) that Attalus entered the war only reluctantly and ‘any notion of personal ambition on Attalus' part at this time must be discounted’ (71). Reluctance to join is not the only possible explanation of the timing of Attalus' arrival in Greece. The notion that Attalus was indifferent to territorial gain is curious: he agreed to purchase Aegina; Allen does not consider how further allied gains in the islands would have been disposed of. Attalus' part in the war was not ‘vague and undefined’ (Allen, 69): although his relationship with the Romans remained informal (Livy (A) 29.11.1; Plb. 16.25.4), he concluded a treaty of alliance with the Aetolians (Livy (P) 31.46.3). I do not understand why Allen thinks Attalus entered the war at all.

37. The election is attested by Livy (P) 27.29.10-30.1. Attalus' title may have been (Holleaux, , Rome 209Google Scholar n. 1; Heuss, A., Stadt und Herrscher des Hellenismus (1937) 43Google Scholar). A similar position was given to Antiochus after his arrival in Europe: Livy (P) 35.45.9; App. Syr. 12. Allen 69 implausibly interprets the election as a further attempt to induce the still reluctant Attalus to enter the war.

38. Cf. McShane, R. B., The Foreign Policy of the Attalids of Pergamum (1964) 109–10Google Scholar.

39. Cf. Holleaux, , Rome 207 n. 3. However, the treaty with the Aetolians (n. 36), which implied a firm commitment, must have been concluded after their treaty with Rome.Google Scholar

40. The importance which Attalus and the Romans attached to Euboea is clear from the campaigns of 208 and 200-198; Philip had already learnt that Attalus planned to attack Euboea in 209 (Livy 27.30.7). On its strategic significance see n. 45. The 208 campaign shows that Philip also held Lemnos and Peparethos. On the problem of what other holdings he may have had in the Aegean see Walbank, , Philip 13Google Scholar; Fraser, P. M. and Bean, G. E., The Rhodian Peraea and Islands (1954) 157–8Google Scholar; Huss, W., Untersuchungen zur Aussenpolitik Ptolemaios IV (1976) 213–38Google Scholar.

41. Plb.9.42.5-8, 22.8.10. On Attalid rule in Aegina see Allen, , ABSA 66 (1971) 1–12Google Scholar. Niese 484 n. 5 wrongly inferred from OGIS 281 that Pergamene forces participated in the capture of Aegina; the receipts from the island attested by the inscription must be later tribute payments – so rightly Cardinali, G., ll regno di Pergamo (1906) 178Google Scholar; Allen, , ABSA 66 (1971)Google Scholar n. 3. On the Aetolians' reasons for selling Aegina see Flacelière 300 n. 2; McShane (n. 38) 107; Walbank, , Commentary III.190Google Scholar; Allen 69. None of these writers considers what territorial promises may have been made to Attalus before the island was captured.

42. Cf. Holleaux, , Rome 217Google Scholar n. 2; CAH VIII.130Google Scholar. That the moveable booty was to be shared is shown by Livy (P) 28.7.4. Lehmann 369-70 argues that all territory gained by the combined fleets would have been assigned to Aetolia, but in view of the way in which the Aetolians had disposed of Aegina this seems unlikely (the reference to Oreus in Plb. 11.5.8 is irrelevant). In their treaty with Attalus (n. 36) the Aetolians may have waived such rights to island territory gained by the fleets as they may have enjoyed under their treaty with Rome. Opus, on the mainland and a former member of the League (n. 26), was a different matter; if it had been retained after its capture in 208, it would doubtless have been assigned to Aetolia. The same policy was followed with regard to Attalus at the beginning of the Second Macedonian War: Andros and Oreus were assigned to him when taken by the Roman and Pergamene fleets in 199 (Livy (P) 31.45.7,46.6), although Flamininus' programme later required that Oreus, along with the rest of Euboea, should be declared free.

43. Cf. Holleaux, , Rome 157–60, 208–9Google Scholar.

44. Cf. Plb.9.30.5-9. In addition to the war in Greece, Philip also faced threats on his northern borders from Rome's allies Scerdilaidas and Pleuratus, and from the Dardanians and Thracian tribes, ever ready to take advantage of Macedonian preoccupation elsewhere.

45. On these routes and their significance see Picard, O., Chalcis el la confédération eubéenne (1979) 212–8, 257–8, 276–7Google Scholar. The sea route through the Euripus was controlled by Chalcis with its bridge(Livy (P) 31.23.12; Picard 252-8). The routes through Euboea were used to transport armies by Doson in 224 (Plb.2.52.7-8: the Aetolians were refusing him passage through Thermopylae) and by Philip in 219 (Plb.4.67.7); Philip presumably sent his troops home this way in 218 (Plb.5.29.5). Oreus was the crossing point, and there was then a choice of route. Philip in 219 cut across the western tip of Euboea and reembarked to cross the Euripus, landing at Cynus in Locris; on this route see Larsen, , Phoenix 19 (1965) 116–128CrossRefGoogle Scholar. Alternatively, there were roads across Euboea to Chalcis (Picard, 213, 257), and from there the mainland could be reached by the bridge. The bridge enabled the great fortress of Chalcis to dominate not just Euboea but the adjacent mainland, and it was this which qualified it to rank as one of Philip's three Fetters of Greece (Plb. 18.11.5-6).

46. Cf. Badian, , Studies 20Google Scholar; Hammond, , JRS 58 (1968) 18Google Scholar.

47. So rightly Walbank, , Philip 81–2Google Scholar, contra Holleaux, , Rome 198–201Google Scholar and CAH VIII.123Google Scholar.

48. In Livy's account keeping Philip out of Italy remains the aim of Roman policy east of the Adriatic: 26.24.16 (in a Polybian context, but possibly influenced by the annalistic tradition), 26.28.2, 26.35.10 (annalistic). Their link with the distorted annalistic tradition on Philip's treaty with Hannibal (n. 10) makes these statements of doubtful value. The account of recent warfare and perhaps also the statement on the Roman legion in 26.28.2 are false: see Appendix I. On Roman fear of Philip see Plb.5.105.8 (perhaps referring to the Aetolian treaty), 8.1.6, 9.22.5.

49. This view of the significance of Tarentum appears in the annalistic tradition: in 214 Hannibal is said to have wanted to acquire it because it was in Macedonian opportune versam regemque Philippum hunc portum, si transiret in Italiam, …. petiturum (Livy (A) 24.13.5). Hannibal took the town in 213/2. While the Roman garrison retained the citadel, it could deny him the use of the Mare Piccolo, but it must have been generally assumed that the citadel would soon fall. In the event, the garrison held out until the Roman recapture in 209.

50. Livy (P) 25.23.8-9.

51. The problems raised by the Roman forces in the war are discussed in Appendix I.

52. Plb.9.37.7-10, 10.25.1-5, 11.5.9-6.8. On the authenticity of these speeches see Walbank, , JRS 53 (1963) 8–11Google Scholar, Speeches in Greek Historians (1965) 16–8Google Scholar, Commentary II.163Google Scholar, Polybius (1972) 45Google Scholar; Pédech, P., La méthode historique cle Polybe (1966) 264–6Google Scholar; Lehmann 135-49; Deininger, J., Der politische Widerstand gegen Rom im Griechenland (1971) 29–34CrossRefGoogle Scholar. Polybius asserted the principle that, although historians might supply the words of speeches, the substance should be an accurate version of what was actually said (12.25b.1; 29.12.10; 36.2.7). I suspect that in practice he may sometimes have allowed himself to supply arguments which seemed appropriate, but he must have had evidence for at least the overall gist. There is no reason to doubt that he was well informed on the First Macedonian War speeches (Pédech's suggestion that his sources were contemporary pamphlets is superfluous).

53. On Polybius' doctrine of the Roman aim of world rule see Walbank, , JRS 53 (1963) 1–13Google Scholar; Musti, D., Polibio e l'imperialismo romano (1978)Google Scholar; Harris 107-117; Derow, P. S., JRS 69 (1979) 1–15Google Scholar; Richardson, J. S., PBSR 47 (1979) 1–11.1Google Scholar hope to show elsewhere that Polybius held that this began to be the Romans' aim from the beginning of the Hannibalic War, not the end of that war as Derow argues.

54. Cf. Rome 233.

55. Rome 238-45; CAH VIII.126–31Google Scholar. Cf. now Gruen 379-80, arguing that the Romans did the minimum necessary to keep the war in Greece alive.

56. Fragments of Polybius survive, but most of our information on this period of the war is provided by three sections of Livy, dealing with events from autumn 211 to spring 210 (26.2-26.3) and with the campaigns of 209 (27.29.9-33.3) and 208 (28.5-8). The source of these passages is Polybius, as their character and their relationship to the extant fragments make clear. Most scholars suppose that Livy was using Polybius directly here, but the view that he was following an intermediary has recently been championed by Tränkle 211-28 (see also Lehmann 21-44). For the date of the first section see Appendix II. The other two occur in Livy's account of the consular years 208 and 207, but it is clear from the order of the extant fragments of Polybius and from the reference to the Nemean and Olympic Games that he dated each of them one year too late (Walbank, , Commentary II.15Google Scholar). We are worst informed on the campaign of summer 210, which Livy omitted, but even where we have Livy there are serious deficiencies in our knowledge. Polybius appears, as on the Social War, to have accorded fullest treatment to the activities of Philip and the Achaeans, on which he will have been best informed, and parts of his account have clearly been heavily compressed in Livy's version. As a result, the activities of the Romans and their allies are sometimes reported with extreme brevity (see e.g. nn. 57-9) and sometimes omitted altogether (e.g. the Roman fleet in 209 before its arrival at Naupactus or in 208 after Attalus' withdrawal). Very little is said about the reasons for the allies' actions.

57. Livy 26.24.15: et Aetoli extemplo moverunt adversus Philippum bellum, et Laevinus Zacynthum – parva insula est propinqua Aetoliae, urbem unam eodem quo ipsa est nomine habet: eam praeter arcem vi cepit – et Oeniadas Nassumque Acarnanum captas Aetolis contribuit. The Aetolian acquisition of Oeniadae and Nasus is attested by Plb.9.39.2.

58. Zacynthus was under Macedonian control again by 207: Livy (P) 36.31.11 (cited n. 134). DeSanctis (418, 430 n. 87) and Holleaux, (Rome 218 n. 6Google Scholar; CAH VIII. 126, 134Google Scholar) held that Laevinus assigned it to the Aetolians and Philip reconquered it in 207. Klaffenbach, , IG IX. 1²1 p. xxxGoogle Scholar doubted whether the Aetolians received the islands and Walbank, , Philip 84 n. 5Google Scholar leaves the matter open. Livy 26.24.15 does imply that Zacynthus was assigned to the Aetolians, but he (or an intermediary) was probably compressing Polybius heavily here, and the omission of Zacynthus from Plb.9.39.2 suggests the contrary. The town would have had to be garrisoned against the Macedonians in the citadel and it seems unlikely that Laevinus thought it worth the trouble.

59. Aetolian participation is not explicitly attested, but is reasonably inferred by Niese 478; Walbank, , Philip 84Google Scholar. Livy's words Aetoli extemplo moverunt adversus Philippum bellum have been taken to imply that the Aetolians launched a brief offensive of their own in autumn 211 while Laevinus attacked Zacynthus, and the suggestion of Niese 478 that they raided Thessaly has found support (Costanzi, V., Studi storici 1 (1908) 42–3Google Scholar; Walbank, , Philip 84Google Scholar; Lehmann 18). But Livy would surely have mentioned such a raid explicitly, and there would hardly have been time for it before the assault on Oeniadae.

60. Livy 26.25.9-17; Plb.9.40.4-6.

61. Livy 26.26.1-3; Plb.9.39.2-3; Paus.7.7.9, 10.36.6. Livy tells us that Laevinus left Corcyra veris principio. If he has reproduced Polybius accurately, this indicates a date in early March: Pédech (n. 52) 464; Derow, P. S., Phoenix 30 (1976) 367Google Scholar. Livy's statement that Anticyra was in Locris must be an error (Walbank, , Commentary II.179Google Scholar). For Philip's recovery of Anticyra in 208 see below n. 87.

62. For the importance which Philip attached to Oeniadae and the fortifications which he constructed in consequence see Plb.4.65.8-11 with Walbank's note. Anticyra's significance for him was that it was the terminus of the route south through Phocis and a good crossing point for the Corinthian Gulf: cf. Livy (P) 32.18.4-5; Walbank, , Philip 87Google Scholar; Larsen (n. 45); Chamoux, F., BCH 89 (1965) 221–2CrossRefGoogle Scholar.

63. Holleaux, , Rome 239Google Scholar, CAH VIII. 126–7Google Scholar; Thiel, 103. Holleaux dated the Roman-Aetolian alliance to 212 and criticized Laevinus for inactivity during the campaigning season of 211. At Rome 239 Holleaux appears to follow Niese 477-80, who held that the events narrated in Livy 26.25-26.3 occupied the campaigning season of 211; he gives a different reconstruction at CAH VIII.126Google Scholar. There is in fact no reason to doubt Livy that Anticyra fell in the spring following the conclusion of the treaty. Livy states that Laevinus returned home after the fall of Anticyra. If Laevinus in fact stayed on for the rest of the campaigning season, Livy (or his intermediary) will have rejected what he read in Polybius. Those who accept this must resign themselves to ignorance about Laevinus' later activities; there would be no warrant for supposing that he did nothing. See further Appendix II.

64. The investing of Echinus: Plb.9.41. Since Philip's advance in 217 had stopped at Phthiotic Thebes (Plb.5.100), we can infer that the intervening cities on the coastroad came over now; at Larisa Cremaste there was some resistance (Livy (P) 31.31.4). For the timing of Philip's advance see Livy 26.25.2: Philip devoted winter 211/0 to securing his northern frontiers quia primo vere moturus exercitum in Graeciam erat.

65. Plb.9.40.2-3 comes from the report of a speech on behalf of an unnamed state appealing for help from a treaty ally. Since the fragment occurs in the excerpta antiqua between the debate at Sparta, which followed Laevinus' capture of Anticyra (Plb.9.39.2), and the siege of Echinus, the probability is strong that the appeal is from the Aetolians to the Romans. Cf. Holleaux, , CAH VIII.127Google Scholar; Lehmann 23-4. Against Schweighaeuser's identification of the appeal with that from Acarnania to Philip see Walbank, , Commentary II.13Google Scholar.

66. Plb.9.42.1-4.

67. Holleaux, , Rome 239Google Scholar, CAH VIII.127Google Scholar; Thiel 115.

68. His advance along the coast road was probably rapid, and the siegeworks at Echinus were completed (Plb.9.41.10).

69. The function of the commonplaces in Plb.9.40.2-3 about the importance of allies' help being sent promptly may have been simply to stress the urgency of the need. The passage does not imply that the Romans had already been tardy or in other respects unsatisfactory as allies (contra Holleaux, , CAH VIII.127Google Scholar).

70. Plb.9.42.4: . The sentence purports to explain the decision of the people of Echinus to surrender. As it stands, it makes poor sense: if the allies failed to stop supplies reaching Philip by sea, it was not because they could not do so (as the sentence implies), but because the Roman fleet did not act (as Holleaux supposed, following Niese 484). Stylistic considerations show that the fragment 9.41.1-4, which comes from the Anon. de obsid. tol., is no more than a paraphrase of Polybius (see Büttner-Wobst's apparatus and Walbank, , Commentary 11.12-4, 183Google Scholar). It thus seems likely that the illogicality in our sentence is the fault of the epitomator. Polybius' meaning must remain a matter for conjecture. Possibly Dorimachus arrived before Sulpicius and before the siegeworks where complete and it was at this point that Polybius remarked on his inability to disrupt supplies.

71. Philip's use of Phalara in 209 (Livy 27.30.3) does not necessarily imply prior possession (as generally supposed, since Niese 484). It seems likely that Phalara should be identified with the modern Stylida (Béquignon, Y., La Vallée du Spercheios des origines au IV^e siècle (1937) 295–9Google Scholar; Walbank, , Commentary III.82Google Scholar) rather than Emir Bey (Stählin, F., Das hellenische Thessalien (1924) 217–8Google Scholar). If so, Philip probably passed by it, if he did not capture it, since he appears to have returned to Thessaly by way of Xyniae. (This explanation of Polybius' reference to Xyniae (9.45.3) seems preferable to Walbank's, suggestion (Philip 88Google Scholar n. 5, Commentary II.183Google Scholar) that Philip began his campaign by advancing by this route and then transferred to the coast road.) A good deal of the summer will have remained after the fall of Echinus. Philip may have moved against Lamia itself, but been thwarted by the Aetolians and their allies.

72. For the capture of Aegina and its subsequent sale by the Aetolians to Attalus, see above n. 41. Holleaux, (Rome 239 n. 6Google Scholar; CAH VIII.128 n. 1Google Scholar) thought it possible that Sulpicius took Aegina before attempting to relieve Echinus. It is true that the sequence cannot be determined from the fragments of Polybius, since they come from different collections of excerpts, but, if I am right that Philip's campaign began in March and proceeded rapidly (above, nn. 64, 68), Sulpicius must have gone to Echinus first.

73. Above n. 37.

74. Livy 27.30.2. This is the only occasion in the war in which Roman troops are attested as operating in Greece away from the fleet. Pergamene forces are known to have supplied at least one garrison, for Lilaea in Phocis, (Fouilles de Delphes III.4.132-5Google Scholar; Moretti, , ISE 81Google Scholar).

75. Livy 27.30.11. It is clear from Livy's narrative that they arrived during the month preceding the Nemea, which were celebrated in the June or July (Derow, , Phoenix 30 (1976) 276–7 n. 36Google Scholar). Livy does not mean that the Aetolians received the news of the fleets' arrival while the Aegium peace conference was actually in session (so Walbank, , Philip 90Google Scholar; Commentary II.229Google Scholar). There is no reason to suppose that Sulpicius knew of the peace negotiations before his arrival at Naupactus (contra Gruen 379).

76. Livy 27.29.9-30.8.

77. Holleaux, , Rome 240–1Google Scholar; Thiel 123-4. The Carthaginian fleet had crossed to Coreyra cum Philippus oppugnare Aetolos pararet (Livy 27.15.7), and Philip hoped to use it to challenge the Roman fleet to a naval battle (Livy 27.30.16, unnecessarily doubted by Walbank, , Philip 90Google Scholar). We do not know the size of the fleet, who commanded it or what became of it. Thiel, 124 may well be right that it was about the same size as Sulpicius' fleet. Thiel thought that the fleet may have remained in Illyrian waters after Sulpicius' departure for Greece, but it is unlikely that he would have run that risk. Holleaux, 's suggestion (Rome 240 n. 2Google Scholar) that the fleet returned to Tarentum at the news of Fabius' attack is more likely, although he spoilt his case by adducing Plb.9.9.11, which certainly belongs in 211 (Walbank, , Commentary II.9,133Google Scholar). Holleaux's view that Bomilcar was the commander rested on his misuse of Plb.9.9.11; Thiel retained it for no reason (cf. 121 n. 97).

78. Livy 27.31.1-2, 32.1-7, 33.2-3.

79. Livy 27.33.4-5.

80. Plb.10.41.1-5; Livy 28.5.3-8.

81. Plb.10.42.1; Livy 28.5.1,10.

82. Livy 28.5.18-6.12,7.9. On the strategic significance of Chalcis, Oreus and Cynus see above n. 45.

83. Livy 28.7.1-7; Frontinus, , Strat. 1.4.6Google Scholar (cited below, n. 194).

84. Livy 28.7.10. According to Livy 27.30.16, ships were expected from Prusias in 209, although it is doubtful whether he yet possessed a fleet (cf. Habicht, C., RE XXIII. 1092Google Scholar). By 202 Philip and Prusias were related by marriage (Plb. 15.22.1). For the controversy over the nature of the connection see Habicht, , RE XXIII 1086–7Google Scholar; Walbank, , Commentary II.475–6Google Scholar. It has sometimes been dated before the First Macedonian War (e.g. Vitucci, G., Il regno di Bitinia (1953) 48Google Scholar), but it is perhaps more likely that the connection was formed during the war, being offered by Philip as an inducement to secure Prusias' help.

85. Livy 28.7.11.

86. Livy 28.7.11-17. Livy records Philip's capture of Thronium in Epicnemidian Locris and Tithronium and Drymaea in Phocis (Livy's Doris is probably an error: Flacelière 302 n. 2). It was probably now that Lilaea, its Pergamene garrison (above n. 74) withdrawn, was taken and garrisoned by Philip, only to be liberated soon after by the inhabitants led by one Patron (Paus. 10.33.3; Fouilles de Delphes III.1.523Google Scholar = SEG XVI.28Google Scholar = Moretti, , ISE II pp. 30–1Google Scholar). Philip's campaign probably deprived the Aetolians of all their remaining holdings in Phocis north of Parnassus, and they may have lost the whole of Epicnemidian Locris (Klaffenbach, G., Klio 20 (1926) 82CrossRefGoogle Scholar; Walbank, , Philip 96 n. 3Google Scholar; Briscoe, , Commentary I.311Google Scholar; contra, Flacelière 308 n. 1).

87. Livy 28.7.17-8, 8.7-13. Anticyra had already been recovered by the time Philip joined his fleet there (8.7); the recovery can only have been effected by this fleet.

88. As suggested by Thiel 129.

89. Livy 28.8.11; inter medias prope hostium classes is an error, since Attalus had already withdrawn.

90. Holleaux, , Rome 241 n. 3, 243Google Scholar. Livy does not state that Philip sailed back with his full fleet (as Holleaux supposed), and, since he hoped to avoid detection by the Roman fleet, it is unlikely that he did so. It would be less readily excusable if Sulpicius had permitted the Macedonian fleet to sail past Aegina on its way to the Corinthian Gulf, but it may be that the fleet had been sent there while he and Attalus were still conducting their offensive.

91. The sack is attested by Livy (P) 32.21.28, 22.10; Paus. 7.17.5. For this dating see Walbank, , Philip 98 n. 1Google Scholar. Errington, , Philopoemen 59 n. 1Google Scholar prefers late 209, but it is hard to see why Livy should have failed to mention the sack at 27.33.4; his silence in his narrative of 208 is less surprising, since he makes no reference to Sulpicius' return to the west. The sack was no doubt intended as a reprisal for the part which Dy me had played in the invasion of Elis in 209; probably no attempt was made to hold on to the city for the alliance.

92. Cf. the remarks of Larsen, , GFS 371Google Scholar on the aim of the 208 offensive.

93. The Eleans were involved in the hostilities of 209, but only with the assistance of an Aetolian garrison (Livy 28.31.9-32.4, 33.5); the only other reference to them is the obscure Livy 28.7.14, on which see n. 94. Nothing is recorded in the literary sources about the Messenians, but see below n. 166.

94. Machanidas' activities led to Achaean appeals to Philip both in spring 209 (Livy 27.29.9) and in spring 208, when his army took up position on the Argive border (Plb. 10.41.2; Livy 28.5.5). Later in 208 Machanidas advanced as far as Heraea, but withdrew at Philip's approach (Livy 28.7.14,17); Livy can hardly be right that his intention was to attack the Eleans while they were preparing the Olympic Games, and it may be that the Achaeans had usurped Olympia (so Niese 492 n. 1; Walbank, , Philip 304 n. 5Google Scholar; Errington, , Philopoemen 60Google Scholar defends Livy unconvincingly). The date of the capture of Tegea (attested by Plb. 11.11.2, 18.8) is unknown: early 208 is a plausible context (Errington, , Philopoemen 60Google Scholar).

95. See above nn. 59-61 for Oeniadae, Acarnania and Anticyra, and against the supposed raid on Thessaly, and n. 93 for the garrison in Elis and its activities. The raid on Achaea: Livy 27.29.9.

96. Plb. 10.41.1-3; Livy 28.5.4-6. Flacelière 301 assumed that the attacks took place; his misinterpretation of Polybius was corrected by Feyel (n. 26) 171 n. 2.

97. They raided Dium and Dodona in 219 (Plb.4.62, 67.1-5), Thessaly in 218 (Plb.5.1,17.5-7) and Acarnania and Epirus in 217 (Plb.5.96.1-2).

98. Contra, Holleaux, , Rome 238, 242–3Google Scholar, CAH VIII.127Google Scholar; Errington, , Dawn 115–6Google Scholar; Gruen 379. On Plb.9.40.2-3 see above nn. 65, 69. It has been held that Aetolian dissatisfaction led them to give serious consideration to a separate peace in 209-8: against this see pp. 145-6.

99. Plb. 10.25:

For the Aegium conference as the context cf. Livy 27.30.11; Walbank, , Commentary II.15Google Scholar. The speaker could be Philip or one of his allies, but is perhaps more likely to be a neutral ambassador (Niese 486; Schmitt, , RR 196Google Scholar); Holleaux, (Rome 35 n. 4Google Scholar) attributed it to a Macedonian ambassador, overlooking the fact that Philip was present in person.

100. Plb.9.37.7-10 (with the different analogy of a friendly garrison taking over a city); 11.5.9-6.8. See above n. 52.

101. Rome 220-36.

102. Contra Holleaux, , especially at Rome 230–1Google Scholar.

103. Plb.9.28-30.4, 11.5.1. To cities on the Aetolian side Macedon might seem a threat to their freedom, as the case of Lilaea shows: see above nn. 74 and 86, and note the words of the inscription honouring Patron, .

104. This seems to me the best interpretation of the notoriously difficult lines 15-21 of the inscription: see above n. 17.

105. Plb.9.42.5-8, correctly interpreted by Lehmann 144. Invitations to come over voluntarily may have been made to some of the other captured cities. Roman concern for Greek opinion was perhaps also shown by the senate's despatch of the legatus L. Manlius in 208 with instructions to proclaim at the Olympic Games that Sicilian and Tarentine exiles might return home (Livy (A) 27.35.3-4).

106. For the sources see Holleaux, , Rome 232 n. 1Google Scholar.

107. That the inhabitants of Zacynthus, Oeniadae and Nasus were not enslaved can be inferred from Plb.9.39.2; that Oeniadae was plundered is confirmed by numismatic evidence (trientes overstruck on coins of Acarnania and Oeniadae: Crawford, M. H., Roman Republican Coinage (1914) nos. 21, 32Google Scholar). At Zacynthus the inhabitants may have escaped to the citadel (cf. n. 58); at Oeniadae and Nasus speedy submission may have earned milder treatment. It is clear from Livy's account that Opus had to submit to plundering (by Attalus), but not enslavement; no doubt this was because the city had made a prope voluntaria deditio (Livy 28.7.4-5.9).

108. Xen. Cyrop. 7.72; Plb.2.58.10, 5.11.3; Livy (P) 31.30.2-3; Volkmann, H. Die Massenversklavungen der Einwohner eroberter Slädte in der hellenistisch-römischen Zeit (1961) 7ffGoogle Scholar. The attempt of Kiechle, F., Historia 7 (1958) 150–1Google Scholar to interpret Polybius' ‘laws of war’ as merely descriptions of conduct in war, not ethical norms, is misguided: Finley, M. I., Historia 8 (1959) 163 n. 61Google Scholar; Volkmann 75 n. 3.

109. Volkmann, 73ff.; Ducrey, P., Le traitement des prisonniers de guerre dans la Grèce antique (1968) 313–32Google Scholar; Ilari, V., Guerra e diritto nel mondo antico I (1980) 364–6Google Scholar. In addition to the argument that Greeks should not treat each other in this way, regularly voiced since Plato, (Rep. 469B–471BGoogle Scholar), such conduct was stigmatized by words like ὠμός, and (Ducrey 330 n. 1). That could be used of an action which was in conformity with reflects the confusion of Greek thinking on the subject.

110. There were no mass enslavements of the inhabitants of conquered cities in mainland Greece between Thebes in 336 and Mantinea in 223, chiefly because the Macedonian kings regarded it as in their interest to spare the cities which were (Plb. 18.3.3-7; Volkmann (n. 108) 76-7). Doson made an exception of Mantinea in reprisal for the killing of an Achaean garrison (Plb.2.58.4-15). There is no reason to think that there had been a significant change in attitudes (as supposed by Tarn, W. W., CAH VII.211Google Scholar; Rostovtzeff, M., Social and Economic History of the Hellenistic World (1941) I.192ff.Google Scholar; Kiechle, F., Historia (1958) 149ff.Google Scholar). It may be that Greek public opinion was shocked by the treatment of Mantinea, as Phylarchus claimed (Plb.2.56.6; Plut., Arat. 45.4–6Google Scholar), but there had been a similar reaction in the previous century when such a fate befell ancient and famous cities (notably Thebes: Arr. Anab. 1.9.1). Philip departed from his predecessors' practice, enslaving the inhabitants of Phthiotic Thebes in 217 and of numerous captured cities from 202 on. It is usually supposed that he abstained from such conduct in the First Macedonian War (e.g. Volkmann (n. 18) 20), but we do not know the fate of Echinus, and Larisa Cremaste (taken in 210) is included in a list of cities which had suffered atrocities at his hands (Livy (P) 31.31.4). The Aetolians had no scruples about mass enslavements (despite the rhetoric of Plb. 11.5.6, which Polybius surely did not intend his readers to take seriously): they carried them out in Laconia in 240 (Plb.4.34.9; Plut. Cleom. 18.3) and at Cynaetha in 220 (Plb.4.18.7-8, 19.6,29.2; 9.38.8), and intended to do so at Pellene in 241 (Plut., Arat. 31–2Google Scholar) and probably Medion in 231 (Plb.2.2.6ff.).

111. Plb.9.39.1 ff, 11.5.1 ff (speeches of Lyciscus and Thrasycrates). The complaints relate closely to the traditional criticisms of such conduct by Greeks (above, n. 107): the emotive language used is the same (11.5.6 7 ), and it was natural for those who held that Greeks should not behave in this way to each other to draw the corollary that such behaviour was to be expected of barbarians. On the view that the Romans were , asserted in these speeches, but never by Polybius in his own person, see Schmitt, H. H., Hellenen, Römer und Barbaren (1958)Google Scholar; Deininger (n. 52) 23-37. Resentment of the Romans' conduct lasted long, for example in Achaea, where it was probably an important factor in the debate in 198 about whether to join the Roman side. (Livy (P) 32.22 admits it only for the people of Dyme, but it plays a much larger part in the accounts of App, Mac. 7Google Scholar and Paus.7.8.2. These versions are distorted but it is not unlikely that Livy played down what Polybius said on the subject: cf. Aymard, A., Les premiers rapports de Rome et de la confédération achaienne (1938) 71-2, 94–7Google Scholar; Deininger, 44-5.)

112. Plb.16.27.2, 34.3.

113. In my view (contra Badian, , FC 71–2Google Scholar; Seager, R., CQ 31 (1981) 106–12CrossRefGoogle Scholar), this was implied by Flamininus' declaration at the Aous (Livy (P) 32.10; Diod.28.11; App, . Mac. 5Google Scholar).

114. Chalcis, Antipatrea and Acanthus were taken by storm and sacked; heavy loss of life is attested at the first two (Livy (P) 31.23.7-8,27.4,45.16). At Andros the citizens secured their liberty by surrendering on terms (Livy (P) 31.45.6-8); the citizens of Oreus, who surrendered unconditionally, were enslaved (Livy(P) 31.46.16; cf. Paus.7.7.9). Pelion was treated more generously merely for reasons of policy (Livy (P) 31.40.4-5).

115. Phaloria was stormed and sacked, and Daulis presumably suffered the same fate (Livy (P) 32.15.3, 18.8). Eretria surrended and was plundered (Livy (P) 32.16.16; Paus.7.8.1); it cannot be assumed that the inhabitants retained their liberty (contra Volkmann 23). The inhabitants of Carystus (Livy (P) 32.17.2) and Elatea (Livy (P) 32.24.7; Paus. 10.34.4) surrendered on a promise of liberty; an inscription (SEG XI.1107Google Scholar = Moretti, , ISE 55Google Scholar) shows that the Elateans were not allowed to remain in their city, and the same may have happened at Carystus. (For bibliography on the inscription see Briscoe, , Commentary 1.214, 345Google Scholar, rightly rejecting the view of Passerini that the Elateans were expelled not by Flamininus, but by the Aetolians in 196). On the similarity between Sulpicius' and Flamininus' treatment of conquered cities see Eckstein, A. M., Phoenix 30 (1976) 134ffCrossRefGoogle Scholar.

116. Plb.11.8-18; Plut., Philop. 9–10Google Scholar; Paus.8.50.1-2. For 208/7 as the date of Philopoemen's first see Walbank, , Commentary II.279Google Scholar. On the date of the battle of Mantinea see Lehmann 144; Errington, , Philopoemen 249–50Google Scholar. Plb. 11.10.9 says that Philopoemen led his troops out after preparations lasting . If this is reckoned from his entry into office, the battle took place in June 207 (see De Sanctis 444; Walbank, , Commentary II.280–2Google Scholar), but, as Errington observes, Plut., Philop. 11Google Scholar makes it unlikely that the battle had taken place by the time of the Nemea of 207 (held in June or July: n. 75), and, if the period of nearly eight months refers simply to the time spent on drilling, it implies a later date, possibly (as Errington suggests) as late as September. The Achaeans recaptured Tegea immediately after the battle: Plb.11.18.8.

117. Livy 28.8.14. The project was probably soon discontinued (cf. Holleaux, , Rome 244 n. 2Google Scholar); the senate did not reckon with a Macedonian fleet of this scale when it sent out Sempronius with 35 quinqueremes. Holleaux supposed that the Carthaginian fleet suffered a major defeat on its way back from Greece in 208 and that this put an end to Carthaginian naval activity for several years (Rome 244 n. 2; followed by Walbank, , Philip 77Google Scholar; Ferrary (n. 2)). This hypothesis is rightly rejected by Thiel 130-2. There is no good reason to postulate a defeat for this fleet, and, if one had occurred, its effects would have been less serious than Holleaux supposed, since the fleet was probably quite small (n. 77). Whatever be made of Livy's suspiciously similar notices of Carthaginian naval defeats off the African coast in 208 and 207 (Livy (A) 27.29.7-8; 28.4.5-7), there is no reason to associate them with the fleet returning from Greece; in any case, Livy, who is unlikely to have minimized Carthaginian losses, says that most of their ships escaped from these battles.

118. The Roman-Aetolian treaty had provided that neither party should make peace with Philip, unless he undertook not to conduct hostilities against the other party and its allies (Livy 26.24.12-3). For Roman resentment of the Aetolians' action, which they regarded as dissolving the alliance see Livy 29.12.4; 31.1.8, 29.4, 31.18; Plb.18.38.8 = Livy 33.13.11; App., Mac. 3Google Scholar. Badian, , Titus Quinctius Flamininus, Philhellenism and Realpolitik (1970) 50–2Google Scholar argues ingeniously but unconvincingly that the Aetolians complied with their obligations to their own satisfaction by adscripting the Romans to their treaty with Philip.

119. The despatch of Sempronius was interpreted as a shift to a more aggressive policy by Larsen, , CP 32 (1937) 31 n. 48Google Scholar and GFS 377; Oost, S. I., Roman policy in Epirus and Acarnania (1954) 36Google Scholar. Cf. Cassola, F., I gruppi politici romani nel III secolo a.C. (1968) 406 n. 4Google Scholar.

120. The order of the excerpts shows that 11.8-18 on Achaean affairs came after 11.4-7 (Walbank, , Commentary II.16Google Scholar). It follows that Polybius must have given Achaean affairs in this year separate treatment, since 11.8-18 deals not just with the battle of Mantinea but with the preparations made by Philopoemen since assuming office in autumn 208. This was a sensible arrangement of material, since events in the Peloponnese in this year were quite independent of the war in northern Greece. For a similar arrangement cf. 28.12-7, 29.23-5, where Polybius narrated the year's events in Achaea and other parts of Greece after his account of Rome's war with Perseus. The matter has been obscured in earlier discussions, although the possibility that Polybius arranged his material in this way is admitted by De Sanctis 444; Walbank, , Commentary II.277Google Scholar; Lehmann 155.

121. Plb.11.4-6.

122. The speaker's name is given in the margin of the manuscript. Schweighaeuser's guess that he was a Rhodian has been widely accepted, but in fact we have no means of knowing which state he represented (Errington, , CR 19 (1969) 167Google Scholar).

123. Plb. 11.7.2-5 (it is clear from the order of the gnomic excerpts that this episode came after the Aetolian assembly: Walbank, , Commentary II.16Google Scholar).

124. Plb. 11.6.1. It has been thought to be a difficulty that Polybius narrated the peace conference before the battle of Mantinea, but in fact Mantinea was probably fought after the Metaurus (n. 114) and in any case Polybius' account of Achaean events was a separate narrative (n. 118). On the date of the battle of the Metaurus see De Sanctis 574-6; Walbank, , Commentary II.270–1Google Scholar; Lehmann 151-5; Derow, , Phoenix 30 (1976) 280–1Google Scholar. A date in May or June seems to me to fit best with what is known of Hasdrubal's activities before the battle; I am unconvinced by Derow's attempt to extract greater precision from Livy's complex and dubious narrative of events in southern Italy. Ovid, Fasti 6.770 says that on 22 June cecidit telis Hasdrubal ipse suis, but in my view this more probably refers to the death of Hasdrubal, son of Gisgo, than to that of Hannibal's brother at the Metaurus (Hannibal's brother did not commit suicide, and Ovid is unlikely to have commemorated the anniversary of a major Roman victory in so oblique a fashion). In any case, if the line does give us the Roman date of the battle, the relation of the Roman to the Julian calendar at this time remains obscure (cf. n. 143).

125. Plb. 13.1-8 deals with Greek affairs; it is difficult to determine which fragments should be assigned to O1.143.3 (206/5) and which to O1.143.4(205/4) (Walbank, , Commentary II.20–1Google Scholar). There is no reference to Rome's war with Philip in these fragments; 1-2 deal with the consequences of peace for the Aetolians.

126. For the earlier passages and their relation to Polybius see above n. 56.

127. Livy 29.12.1-4 neglectae eo biennio res in Graecia erant. itaque Philippus Aetolos desertos ab Romanis, cui uni fidebant auxilio, quibus uoluit condicionibus ad petendam et paciscendam subegit pacem. quod nisi omni ui perficere maturasset, bellantem eum cum Aetolis P. Sempronius proconsul, successor imperii missus Sulpicio cum decem milibus peditum et mille equitibus et triginta quinque rostratis nauibus, haud paruum momentum ad opem ferendam sociis, oppressisset. uixdum pace facta nuntius regi uenit Romanos Dyrrachium uenisse Parthinosque et propinquas gentes alias motas esse ad spem nouandi res Dimallumque oppugnari. eo se auerterant Romani ab Aetolorum quo missi erant auxilio, irati quod sine auctoritate sua aduersus foedus cum rege pacem fecissent.

128. Livy 29.12.3-7. The number of Laetorius' ships is uncertain: the OCT apparatus implies that quindecim is the reading of P but XXV of the other tradition (Σ) (on the textual tradition of Livy 26-30 see Walsh's Teubner edition of 26-27 (1982) v ff).

129. Livy 29.12.8-15.

130. Livy 29.12.16: iusseruntque (pacem) omnes tribus, quia verso in Africam bello omnibus aliis in praesentia levari bellis volebant. P. Sempronius pace facta ad consulatum Romam decessit. Sempronius' election, Livy (A) 29.11.10: consules facti M. Cornelius Cethegus P. Sempronius Tuditanus absens cum provinciam Graeciam haberet.

131. In the list of Roman adscripti to the treaty (14) Ilium must be an insertion from an annalistic source and the same may be true of Athens (below nn. 209-10). Derow, , JRS 69 (1979) 7Google Scholar holds that 16 is entirely Polybian, but omnes tribus must be an annalistic touch, and I argue below that this section is entirely annalistic (on the difficulty of determining the source of transition passages between Polybian and annalistic sections see Briscoe, , Commentary I.10, II.2Google Scholar). It is usually supposed that there is annalistic distortion in 2-4, but I argue below that Polybius has been reproduced accurately here.

132. Livy 31.29.3 quibus enim de causis experta inutili societate Romana pacem cum Philippo fecissent, compositam semel pacem servare eos debere; 31.19 et forsitan dicatis bello Punico occupatis nobis coactos metu vos leges pacis ab eo qui tum plus poterat accepisse. The meeting place: 29.16 (with 29.8). These speeches are elaborations of speeches in Polybius; it is likely, but not certain, that the points made in these passages were to be found in the Polybian originals. Cf. Briscoe, , Commentary I.17ff., 129Google Scholar and Tränkle 86 n 66, 128 n. 70, rightly rejecting the view of Burck, E., Wege zu Livius (1967) 452–63Google Scholar that the speeches are entirely Livy's work.

133. Livy 32.21.17: Aetolos tum classe adiuverunt; nee duce consulari nee exercitu bellum gesserunt; sociorum Philippi maritimae tum urbes in terrore ac tumultu erant; mediterranea adeo tuta ab armis Romanis fuerunt ut Philippus Aetolos nequiquam opem Romanorum implorantes depopularetur. The question of relationship to Polybius is the same for this passage as for those cited in n. 132.

134. Livy (P) 36.31.11: Philippi Macedonum regis Zacynthus fuerat; eam mercedem Amynandro dederal, ut per Athamaniam ducere exercitum in superiorem partem Aetoliae liceret, qua expeditione fractis animis Aetolos compulit ad petendam pacem.

135. Tränkle 220-1 (leaving the choice between the two dates open).

136. Above, n. 56.

137. A less radical departure from the orthodox chronology was briefly proposed by Balsdon, , JRS (1954) 31Google Scholar (followed by Errington, , Philopoemen 69Google Scholar; criticized by Walbank, , Commentary II.278Google Scholar). Balsdon did not question that Livy 29.12 derived from Polybius' Res Graeciae for 206/5, but held that Sempronius was appointed in the consular year 206, arriving to take up his command at the end of the year. This suggestion is, in my view, an improvement on the orthodoxy but still unsatisfactory: see below nn. 153, 163.

138. Livy's misdating of the campaigns of 209 and 208 is attributed to a false annalistic chronology by Walek-Czernecki, T., Rev. Phil.^e 11 (1928) 14Google Scholar; cf. Lehmann 32-4. The suggestion that Livy went astray converting Polybius' Olympiad years into consular years (Schmitt, , RR 198 n. 1Google Scholar, following Weissenborn) is unattractive: there might conceivably be uncertainty about whether a campaign recounted by Polybius under (e.g.) the Olympiad year 209/8 belonged in consular 209 or 208, but it is hard to see why it should be converted to consular 207. The explanation offered by Costanzi, V., Studi Storici 2 (1909) 227–9Google Scholar and De Sanctis 443 is preposterous.

139. Livy 25.11.20; 27.7.5.

140. For Livy's error see De Sanctis 446 n. 4, 454 n. 18, 468 n. 38; Walbank, , Commentary II.8-9, 14–5Google Scholar. They attribute the error to confusion in converting Polybius' Olympiad into consular years, but Livy 27.7.5 shows that the error was present in the non-Polybian tradition.

141. Livy reports elections in absence at 4.42.1, 48.1; 7.26.12; 10.5.14, 22.9; 22.35.7; 23.24.3; 24.9.9, 43.5; 26.22.13; 29.11.10, 12; 31.50.6. The two earliest notices can hardly be credited, and the statement that Marcellus was elected in absence to the consulship of 210 (26.22.13) conflicts with the earlier account of his return (26.21) and should probably be rejected (Lehmann 28-9; but in defence of Laevinus' election in absence see below Appendix II.) None of these notices need derive from an archival record. In some cases Livy's annalist source seems to have had further information about the election (e.g. 29.11.12, 31.50.6), but many of the notices may be based just on the annalist's knowledge that the man elected had been on campaign in the previous year.

142. Cf. Niese 502 n. 5; Holleaux, , Rome 280 n. 2Google Scholar.

143. The consuls entered office in this period on the Ides of March, and the elections normally took place a month or two beforehand (cf. Mommsen, , Röm. Staatsrecht I.583Google Scholar). On recent hypotheses, the Ides of March in 204 fell on 24 or 25 February (cf. Briscoe, , Commentary II.25Google Scholar, slightly modifying the calculations of Derow and Marchetti), but in my view 17 March is also possible (cf. JRS 73 (1983) 240Google Scholar). Livy 29.12.16-13.1 implies that Sempronius was back at Rome by the time he entered office. Triennio prius in Livy (A) 31.1.8 does not imply that the peace was not ratified at Rome until the consular year 204 (as supposed by e.g. Weissenborn on 29.12.1; De Sanctis 444); for the correct explanation of the phrase see Briscoe's note.

144. Two or three months would have been adequate for his activities up to the Phoenice agreement, which allowed two months for receiving approval from Rome.

145. The biennium is correctly identified by Tränkle 220, who compares Livy 4.53.1.

146. A further argument is acutely adduced by Tränkle 221 n. 106. If the Peace of Phoenice took place in 205, Polybius' account will have been in book 13, separated from his narrative of the war which it ended by the polemic against Timaeus which occupied the whole of Book 12. Is it likely that Polybius would have chosen so infelicitous a place for his excursus?

147. For the identification: Woodhouse, W., Aetolia (1897) 261Google Scholar; De Sanctis 430 n. 87, 444; Holleaux, , Rome 253 n. 4Google Scholar; Klaffenbach (n. 58) xxxi-ii; Flacelière 304; Walbank, , Philip 99, 305–6Google Scholar, Commentary II.278Google Scholar. Against the identification: Oberhummer, E., Akarnanien, Ambrakia, Amphilochien, Leukas im Altertum 170Google Scholar; Niese 495, 500; Walek-Czernecki, , Rev. Phil. (1928) 20–3Google Scholar; Geyer, F., RE XIV. 749, XIX.2309Google Scholar; Schmitt, , RR 211 n. 1Google Scholar. Oost, , CP 52 (1957) 13 n. 13Google Scholar overlooks the problem.

148. On the Aetolian assemblies see Holleaux, , Études, I.219–27Google Scholar; Larsen, , TAPA 83 (1952) 1–33Google Scholar. There is no reason to suppose that Agelaus of Naupactus, who had argued for peace in 217, was elected for 207/6: see Klaffenbach, (n. 58) 6; Deininger (n. 52) 35 n. 6.

149. Schmitt (n. 147). This is the only argument of any substance which has been produced against the identification.

150. The 218 invasion is fully described in Plb.5.5-14. Polybius' account of Philip's advance (7-8) stresses the importance of speed and surprise and (with some exaggeration: cf. Walbank, , Commentary I.542–5Google Scholar) the difficulty of the route.

151. On Philip's route see Woodhouse (n. 147); Walbank, , Commentary II.278Google Scholar. Philip will have crossed into Athamania by the pass from Gomphi (Oost, , CP (1957) 2–3Google Scholar) and then descended the Inachus to the Achelous. Plb. 11.7.2,4–5 shows that he approached Thermum along the south side of Lake Trichonis, as in 218.

152. Walbank, , Philip 101Google Scholar, Commentary II.278Google Scholar favours autumn 206: ‘disheartened (by the invasion), the Aetolians continued to appeal to Rome but waited to see what help would arrive the next summer…; when none came they made peace that autumn’. It seems unlikely that Philip would have let the summer pass without pressing home his advantage. Cf. Schmitt, , RR 211 n. 1Google Scholar.

153. An interpretation on these lines was first proposed by Niese 501 n. 2, who has been followed by many subsequent writers: e.g. Costanzi, , Studi Storici 1 (1908) 420 n. 1, 2 (1909), 227Google Scholar; De Sanctis 443; Holleaux, , Rome 255 n. 2Google Scholar; Thiel 153 n. 1; Walbank, , Philip 102 n. 2Google Scholar, Commentary II.278Google Scholar; Briscoe, , Commentary I.52Google Scholar; Lazen by (n. 2) 167. Of the scholars who reject the identification of Philip's invasions (above n. 147), only Walek dates the Aetolian separate peace in 205 and upholds the accuracy of Livy 29.12.2–4. Balsdon (n. 137) defends Livy, but the chronological crux is not resolved by his dating of Sempronius' arrival to the end of 206.

154. Some older scholars hold that Sulpicius was recalled to Italy with all his forces in 207: e.g. Schorn, W., Geschichte Griechenlands von der Entstehung des ätol. und ach. Bundes bis auf die Zerstörung Korinths (1833) 194Google Scholar: Weissenborn on Livy 29.12.2; Clementi, G., Studidi storia antka 1 (1891) 75Google Scholar: cf. Scullard, H. H., Roman Politics 220-150 B.C. (1951) 77 n. 1Google Scholar. However, the natural interpretation of successor imperii missus Sulpicio in Livy 29.12.2 is that Sulpicius remained until Sempronius arrived to replace him, and his continued presence is attested by App., Mac. 3Google Scholar (discussed below). The omission of Greece from the information on commands for 207 and 206 in Livy 27.36-8, 28.10 proves nothing, since these lists are in other respects incomplete.

155. E.g. Niese 494; De Sanctis 429; Holleaux, , Rome 245–51Google Scholar, CAH VIII. 132Google Scholar; Walbank, , Philip 98–9Google Scholar; Thiel 136-8. Holleaux, , Rome 248Google Scholar n. 1 suggests that Sulpicius may have retained 10 quinqueremes and about 1,200 soldiers. There is no ancient evidence for the withdrawal of Roman forces at this time; the doctrine is merely a modern conjecture to account for Sulpicius' inactivity. It is rejected by De Regibus, L., La repubblica romana e gli ultimi re di Macedonia (1951) 74–5Google Scholar; Balsdon (n. 137).

156. Before Holleaux Roman inactivity was generally attributed to their other preoccupations: see Niese and De Sanctis, cited in n. 155, and the earlier writers cited by Holleaux, , Rome 247 n. 4Google Scholar. More recent writers who consider that this factor may have played a part include Thiel 137-8; Will, II.91; Errington, , Dawn 115Google Scholar.

157. Holleaux, , Rome 248–9Google Scholar.

158. Holleaux, , Rome 252Google Scholar, refuting Niese 500-1. In 206 no new initiatives were launched elsewhere, 70 of the 100 ships in the Sicilian fleet were recalled to Rome, and those who accept Livy's legion lists calculate that the legions in service were reduced from 23 to 20 (Livy 28.10; Thiel 139; De Sanctis 506). 205, on the other hand, saw Scipio's departure for Sicily and Africa (although according to the tradition no new forces were levied for this expedition and the number of legions may have dropped to 18).

159. This is apparently the view of Will, II.94 and Gruen 300. Sulpicius, who had witnessed the indifferent Aetolian performance in the war so far and will have been well aware how much the advantage had now swung in Philip's favour, cannot have supposed that the Aetolians would fight on unaided, and it is hard to accept that he could have failed to make this clear to the senate. For the Romans to withdraw from the war indefinitely was, of course, a clear breach of their treaty obligation.

160. Holleaux, , Rome 250–1Google Scholar; CAH VIII. 132Google Scholar. Cf. Will, II.91; Ferrary (n. 2).

161. It is not true as Holleaux supposed, that the Carthaginian navy had been put out of action in 208 (above, n. 117), but fears of a Macedonian landing at Tarentum will have been ended by the recovery of the city in 209, and Scerdilaidas and Pleuratus may have recovered Lissus ca. 208 (May, J. M. F., JRS 36 (1946) 48–52Google Scholar; Cabanes 262-3).

162. Holleaux, , Rome 254–5Google Scholar; CAH VIII. 135Google Scholar: ‘this peace, for which she was responsible, roused Rome from her lethargy’.

163. This sequence of events is also implied by the view of Balsdon (n. 137). However, Balsdon supposed that Sempronius did not arrive until the end of 206, and made no attempt to account for this delay.

164. Livy says nothing about whether Sempronius took over the force under Sulpicius' command. This silence does not warrant the general, tacit assumption that Sulpicius took what was left of his force home.

165. It is not a serious difficulty that there is no reference to this undertaking in Thrasycrates' speech: the speech is a general argument against the war without reference to the Aetolians' immediate situation and looks ahead to the time when the Romans might be free to cross to Greece in full force (Plb. 11.6.1-2). No inference about the precise timing of the Aetolian appeal can be drawn from Livy 32.21.17 (cited n. 133).

166. Fouilles de Delphes III.4.3–6Google Scholar; cf. Flacelière 304-5, 491. This context, first proposed by Niese, has been generally accepted, but the fact that the decrees relate to two separate garrisons constitutes a difficulty. The usual view that the first garrison was replaced or reinforced shortly after its arrival seems unlikely and the assumption that decrees 5-6 were passed under the same archon and bouleutai as 3-4 rests on weak evidence (Flacelière, 491 n. 1). It is possible that the first garrison may have been installed earlier: Philip's successes in 208 had posed a more immediate threat to Delphi.

167. Cf. Rich, J. W., Declaring war in the Roman Republic in the period of transmarine expansion (1976) 18ffGoogle Scholar.

168. Polybius failed to grasp that the senate's response to Hannibal's attack on Saguntum was postponed ad novos consules: Rich, , Declaring War 41Google Scholar. On Polybius' attitude to Aetolia see Sacks, K.S., JHS 95 (1975) 92–106CrossRefGoogle Scholar.

169. On Polybius' belief that this was the Romans' aim see above n. 53.

170. Cf. Walsh, P. G., Livy (1961) 67–8Google Scholar; Merten, M., Fides Romana bei Livius (1965) 12ff.Google Scholar The most notorious case of an ally whom Rome had failed to help was Saguntum; Livy made much of that (21. 16.2, 19.9; 31.7.3; 34.10.8), and may well have had the parallel in mind when he wrote of the Roman desertion of the Aetolians.

171. Meloni, P., Il valore storico e le fonti del libro Macedonico di Appiano (1955)Google Scholar argues that in addition to these sources the work draws on another second century Greek writer, but this thesis is convincingly refuted by Gelzer, M., Kl. Schr. III.280–5Google Scholar; Gabba, E., RAL 12 (1957) 339–51Google Scholar.

172. The coincidence in troop numbers puts it beyond doubt that Appian's reinforcements should be identified with those brought by Sempronius: so e.g. De Sanctis 429 n. 83; Holleaux, , Rome 255 n. 2Google Scholar; Meloni (n. 171) 14-7; Schmitt, , RR 207Google Scholar. For unsuccessful attempts to vindicate Appian see Niese 493; Hammond, N. G. L., Epirus (1967) 611–2Google Scholar.

173. Livy 29.12.5, with n. 128 above.

174. The inscription adduced by Hammond (n. 172) in support of his view that Philip twice held Ambracia in the period 209-6 is irrelevant: see Cabanes 263-4, 518.

175. As suggested by Salvetti, C., Studi di storia antica 2 (1893) 120Google Scholar; Niese (n. 158). The Epirots had not previously been affected by the hostilities; an attack by Laetorius on their territory would help explain the part they played in the peace negotiations. The common view (Cabanes 254-61, with bibliography) that Epirus obtained some form of neutrality in the war should be rejected. The Epirots are counted as allies of Philip in Plb. 9.38.5; 11.5.4, and feared attack in 208 (Plb. 10.41.4). It was merely the accident of war that they were spared more active involvement. They could provide a refuge for the Acarnanians' families in winter 211/0 (above, n. 60) because Acarnania was the immediate object of the Aetolians' attack.

176. E.g. Holleaux, , Rome 35–6Google Scholar; Walbank, , Philip 99, 305Google Scholar, Commentary II, 274–5Google Scholar; Meloni 10-20.

177. Ferro, B., Le origini della seconda guerra macedonica (1960) 7Google Scholar n. 6, 139-42; Huss (n.40) 110-3; cf. Niese 490 n. 1. Ferro holds that Appian implies three mediations and identifies the first two with those of 209 and 208; this rests on a misinterpretation of δίς in Mac. 3.1. Huss identifies Appian's first mediation with that of 209.

178. Schmitt, , RR 204–10Google Scholar, followed by Lehmann 136.

179. Schmitt, , RR 207–8Google Scholar. Schmitt, also (RR 205Google Scholar) seeks support from the fact that Amynander is listed by Appian among those participating in the first invasion, but not mentioned in Plb. 11.4.1; he suggests that Amynander's change to the Macedonian side (implied by Livy 36.31.11) had taken place in the interim. This argument has no force (cf. Walbank, , Commentary II.274Google Scholar; Huss (n. 40) 112 n. 42): Polybius' silence does not show that Amynander was absent, since Thrasycrates was speaking on behalf of the Greeks of Asia and the islands (Plb. 11.4.6), and Appian implies that the same powers took part in both mediations. Philip's agreement with Amynander probably took place after the Aetolians had rejected peace.

180. Walbank, , Commentary II.274–5Google Scholar.

181. Holleaux, , Rome 36–8Google Scholar; CAH VIII. 128–9Google Scholar.

182. Schmitt, , RR 25–6, 57–8, 193–211Google Scholar, arguing that the mediators were not acting against Rome.

183. For Holleaux, e.g. Bleicken, J., Gnomon 31 (1959) 441Google Scholar; Walbank, , Commentary II.229Google Scholar; Larsen, , GFS 371Google Scholar; Habicht, C., Studien zur Geschichte Athens in hellenistischer Zeit (1982) 136CrossRefGoogle Scholar. For Schmitt, e.g. Ferro (n. 172) 7 n. 6; Will, II.93; Heinen, H., ANRW I.1.641Google Scholar.

184. This was Thrasycrates' thesis, and he claimed that the mediators had often expressed these views before (Plb. 11.4.1ff.). The claim is confirmed for 209 by Livy 27.30.10; Plb.10.25 (above, n. 99).

185. The passages of Livy cited by Holleaux, , Rome 36Google Scholar n. 4 do not show this.

186. Livy 27.30.5, on which see Schmitt, , RR 195Google Scholar. Other factors will have included opposition to Attalus' intrusion (Livy 27.30.10) and perhaps the concern of commercial powers at the war's effects on their trade. Cf. Huss (n. 40) 129-31; Habicht (n. 183) 137.

187. The Ptolemaic court evidently coordinated the mediation: Rönne, T. and Fraser, P. M., JEA 39 (1953) 86–94Google Scholar (embassy from Chios in Alexandria in March 209). On the friendship between Rome and Egypt see Heinen, , ANRW I.1.631ff.Google Scholar; Huss (n. 40) 163-72. Egypt was not Rome's only friend among the mediators; amicitia had also been established with Athens (Plb.2.12.8; Zon.8.19.7), and possibly with Rhodes (Plb.30.5.6 with Walbank's note) and Chios (Derow, P. S. and Forrest, W. G., ABSA 77 (1982) 79–92Google Scholar).

188. Livy 27.30.14; 28.7.15; Plb. 11.6.9-10.

189. For the negotiations see Livy 27.30.4-6, 10-5; Plb.10.25.

190. Cf. Errington, , Philopoemen 56Google Scholar.

191. The council of war: Plb. 10.42.4 = Livy 28.5.13 (Polybius' evidently includes the Romans, as at 42.1). The mediators' presence: Livy 28.7.14 adfuerant enim legati nuper Heracleae concilio Romanorum Aetolorumque.

192. Plb. 10.42.4: (on the meaning of here see Walbank's note). Livy's language is misleading (particularly 28.5.13 concilium Aetolis Heracleam indicium): cf. Larsen, , TAPA 83 (1952) 19–22Google Scholar. Both Holleaux, (Rome 36 n. 3Google Scholar) and Schmitt, (RR 197–8Google Scholar) take the meeting to be a session of the Aetolian assembly.

193. Livy 28.7.13-6. Evidently they went first to Demetrias, but Philip did not have time to grant them an interview until he reached Heraclea.

194. Frontinus, , Strat. I.4.6Google Scholar: Philippus Macedonum rex Graeciam petens, cum Thermopylas occupatas audiret et ad eum legati Aetolorum venissent acturi de pace, retentis eis ipse magnis itineribus ad angustias pertendit securisque custodibus et legatorum reditus exspectantibus inopinatus Thermopylas traiecit.

195. Contra Walbank, , Philip 95 n. 1Google Scholar.

196. Contra Walbank, , Philip 102Google Scholar.

197. On the Panaetolica see the works cited above n. 148.

198. Philip could, however, hardly have countenanced Spartan retention of Tegea; he may have stipulated that the Spartans should not be included in the peace unless they returned Tegea to the Achaeans.

199. Plb. 11.6.9; App., Mac. 3.3Google Scholar.

200. On his return Philip may have deprived the Aetolians of Dolopia: Klaffenbach, , IG IX.1².1, 31Google Scholar; Flacelière 304. It used to be thought that Philip recovered much of Thessaly from the Aetolians in 207, but, if the Aetolians ever held this territory, they must have lost it to Doson (above, n. 26).

201. The terms of the peace are discussed by Holleaux, , CAH VIII. 134–5Google Scholar; Klaffenbach, (n. 58) xxxii-iii; Flacelière 306-9; Walbank, , Philip 100–1Google Scholar, Commentary II.555–6Google Scholar; Briscoe, , Commentary I.53–4Google Scholar. There is no justification for the view that Echinus, Larisa Cremaste, Phthiotic Thebes and Pharsalus were either ceded (Klaffenbach, Flacelière) or promised (Holleaux and Walbank, following Stählin, F., Philol. 77 (1921) 199–206CrossRefGoogle Scholar) to the Aetolians now. The Aetolians' later claim to these places is sufficiently explained by the fact that they had once been members of the League (Plb. 18.36.6); the fact that the Aetolians had been obliged to cede these cities when Philip had the advantage did not preclude them from reasserting their claim at the time of Philip's defeat.

202. The meeting place: above, n. 132. Naupactus is the only known location for the Panaetolica; Holleaux, 's conjecture (Études I.227Google Scholar) that it was held in various cities by rotation may be ill-founded. The Panaetolica seems to have been held before or about the spring equinox (cf. Holleaux, , Études, I.204–6Google Scholar). If the inscription republished by Papazoglou, F., Ziva Antika 20 (1970) 99–113Google Scholar belongs to the reign of Philip V, it may show that he had time for a Dardanian campaign in spring 206, but the king named may (as Papazoglou believes) be Philip II.

203. Livy 29.12.2: proconsul; cf. Larsen (n. 119).

204. It is usually supposed that Philip was ceding territory which he still held (e.g. Hammond, , JRS 58 (1968) 19Google Scholar; Cabanes 265-6), but Livy 29.12.3-5 does not make it clear what the outcome of Sempronius' campaign was, and it seems unlikely that Sempronius would have withdrawn to Apollonia before Philip's arrival unless his attack on Dimallum had been successful and the Parthini had been confirmed in their renewed allegiance to Rome.

205. For Lissus see above n. 161.

206. Atintania assigned to the fourth region under Paullus' settlement: Livy (P) 45.30.6. Holleaux, 's view (Rome 278 n. 1Google Scholar) that Flamininus demanded its return in 198 rests on an untenable interpretation of Plb. 18.1.14 (see Walbank's note). Cabanes 276, 386 assumes without discussion that Rome recovered it in 196.

207. Holleaux, , Rome 284–305Google Scholar; CAH. VIII.135-7. He is followed by, e.g., Walbank, , Philip 105Google Scholar; Gruen 381. However, not all those who have accepted Holleaux's interpretation of the Roman conduct during the war have been convinced by his view of the peace settlement: e.g. Thiel 153-4; Will, II.96-7; Ferrary (n. 2) 737.

208. Livy 29.12.14. For the enormous bibliography on the authenticity of the adscripti see Schmitt, Staatsverträge, no. 543, adding Ferrary (n. 2) 736; Habicht (n. 183) 138-41. Debate is now confined to Athens and Ilium, the authenticity of the other names being generally accepted. The description of Nabis as Lacedaemoniorum tyrannus must be Livy's gloss.

209. The conclusive argument has been pointed out by Habicht (n. 183) 140. The other names in both the Roman and Macedonian lists are grouped by the well-established categories of Hellenistic diplomatic usage (for which see Walbank, , Commentary II.117Google Scholar): personal rulers (kings and dynasts) followe by peoples (ἔθνη) and cities). The only comparable list of adscripted states which has survived, in the treaty of 179 between Eumenes and Pharnaces (Plb.25.2.12-3), is arranged in the same way. The position of Ilienses at the head of the Roman list is thus anomalous and can only be explained by supposing that they did not figure in Polybius' list but were an addition from an annalistic source. Not all such additions are unhistorical, and in 38.39.10 Livy was able to add an authentic detail relating to Ilium from the annalistic tradition. However, it is much more plausible to suppose that an annalist keen to stress Roman pietas invented the inclusion of Ilium in the Phoenice treaty than that Polybius accidentally omitted Ilium from the list and Livy was able to restore it from an annalist.

210. The widely held view that the inclusion of Athens is an annalistic falsification, linked to the exaggerated annalistic account of Athens' role in the preliminaries of the Second Macedonian War, may well be correct, but no conclusive argument has been adduced to show that Athens did not stand in Polybius' list. Neutrals could be adscripted in treaties, like Lampsacus in Rome's treaty with Philip in 196 SIG ³ 591.64-8) and some of the states adscripted in the Eumenes-Pharnaces treaty (Burstein, S. M., AJAH 5 (1980) 4 n. 19Google Scholar). There was sufficient reason for Sempronius to nominate Athens in her friendship with Rome (above n. 187) and delicate relations with Macedon (cf. Habicht (n. 183) 129ff., arguing that Athenian policy in this period was dominated by the fear that Macedon might seek to deprive them of the liberty won in 229). The delicacy of those relations would no doubt have deterred Athens from soliciting inclusion, but there is no reason to suppose that any of the states were consulted about their adscription.

211. The view that the Romans were only postponing war with Macedon until Carthage had been defeated recurs in the later historiographical tradition (Sall., Hist. 4.69.5Google Scholar; App., Mac. 3.3Google Scholar; Justin 29.4.11; Zon.9.5.1) and may go back to Polybius (Bickerman, E., CP 40 (1945) 147Google Scholar). Livy, in annalistic passages or in speeches, speaks of the Romans as making peace with Macedon to concentrate on war with Carthage and resuming war with Macedon once the Carthaginian war was over: 29.12.16; 31.1.8-9, 31.19-20; 32.21.18 (the connection between the peace and the invasion of Africa made in 29.12.16 and perhaps 31.31.19 rests on the false annalistic chronology: above, p. 138). Revenge for the previous war is given as a motive for the Second Macedonian War in Livy (A) 31.11.19-20.

212. Livy (P) 31.29.4; App. Mac.4.2. For bibliography see Briscoe, , Commentary I.130Google Scholar, adding Derow, , JRS 69 (1979) 7–8Google Scholar. The doubts about the authenticity of the episode raised by Badian and others are almost certainly misplaced. Derow defends Appian's dating of the appeal at the same time as those from Attalus and the Rhodians, but Appian's evidence on this point can carry no weight since he says the same of the Athenian appeal, which was certainly later. Holleaux, 's argument (Rome 293 n. 1Google Scholar) that the Romans would not at the same time have rebuffed the Aetolians and made a positive response to Attalus and the Rhodians retains its force. The Romans later showed more eagerness to secure Aetolian help than Derow allows: Sulpicius initiated overtures to them shortly after arriving in his province (Livy (P) 31.28.3, not cited by Derow). The precise date of the appeal is uncertain, although in view of Livy (P) 31.46.4, autumn 202 (Holleaux) is perhaps preferable to summer 201 (Briscoe).

213. Holleaux, , Rome 289–97Google Scholar took the episode as confirmation of his view that the Romans had no interest in the war with Philip. For the interpretation adopted here cf. Harris 208; Derow (n. 212).

Notes to Appendix I

214. Discussion has centred on the Second Punic War legion lists. For the view that these lists are substantially reliable see especially De Sanctis, esp. 417ff.; Toynbee, A. J., Hannibal's legacy (1965) II.36ff.Google Scholar, Brunt, P. A., Italian manpower 225 B.C.-A.D. 14 (1971) 645–60Google Scholar; Marchetti, P., Histoire economique et monetaire de la deuxieme guerre punique (1978) 13ff.Google Scholar For sceptical views see Kahrstedt (n. 21); Gelzer, , Kl. Schr. III.220–55Google Scholar; Gschnitzner, F., Hermes 109 (1981) 59–85Google Scholar. These works are cited below by authors' name alone.

215. Livy 23.32.17, 38.7-9; 24.11.3, 40.2, 5. In 23.38.9 Livy mistakenly gives the number of the enlarged fleet as 55; he or his source had evidently overlooked the fact that the five ships which escorted the captured Macedonian embassy to Rome had been detached from Laevinus' squadron (Thiel 67 n. 100). The number 50 is confirmed by Livy 24.11.5-6; 26.1.12.

216. Livy 24.44.5; 26.1.12; 27.7.15, 22.10.

217. Livy 26.24.10.

218. Livy 28.5.1.

219. Thiel 100 n. 199.

220. Brunt 666, 669.

221. Marchetti 111-2.

222. So Gelzer 243, who supposes that 25 quinqueremes was the full strength.

223. Livy 23.32.16-7, as amended by Madvig; for Varro's army in late 216 see Livy 23.25.6, 11.

224. Livy 23.38.9-11.

225. Livy 24.10.4, 11.3. The troops levied by Varro in Picenum in 215 (Livy 23.32.19) are also now spoken of as a legion.

226. Livy 24.40.5.

227. Livy 24.44.5; 25.2.6-7 (not mentioned explicitly, but implied in the total); 26.1.12.

228. Livy 26.28.2, 9.

229. Milan, A., Critica Storica 2 (1973) 6–11Google Scholar, citing such texts as Livy 26.48.6-12, 29.35.14.

230. Livy 27.7.15, 22.10.

231. First proposed by Schemann, L., De legionum per alterum bellum punicum historia quae investigari posse videantur (1875) 26–7Google Scholar. Accepted by e.g. De Sanctis 232, 249; Holleaux, , Rome 187 n. 2Google Scholar; Thiel 92 n. 169; Brunt 649, 651; Marchetti 53. Rejected by Toynbee, II.527-8.

232. De Sanctis 232 n. 38, expressed some doubt about the exchange of troops, but did not question that all the Cannenses found their way to Sicily. On the difficulties of the tradition on the Cannenses in Sicily see Brunt 652-5. Those who cannot accept that Varro retained any of the Cannenses may fall back on Brunt's suggestion (649, 651) that Varro drew his troops from the garrison which had been sent to Tarentum in 217 (Plb.3.75.4).

233. Thiel 92 n. 169: 10,000 citizens and allies less 2,000 to garrison Brundisium.

234. Niese 468, 474; Kahrstedt 461 n. 1; Holleaux, , Rome 187 n. 2Google Scholar, 248 n. 1; Gelzer 243. Holleaux (accepting that Laevinus had 50 warships) estimates the force at 5-6,000; Gelzer reduces it to 3,000.

235. E.g. De Sanctis 429, 457; Walbank, , Philip 88Google Scholar; Larsen, , GFS 370Google Scholar; Errington, , CR 19 (1969) 167Google Scholar; Brunt 421 n. 2, 653; Marchetti 72-3.

236. Thiel 103 n. 210; Walbank, , Commentary II.183Google Scholar; Lehmann 20-1, 29 n. 58.

237. Livy 26.28.2.

238. Cf. Brunt 420 (likely strength of legions ca. 214), 671-2 (nominal strength), 677-80 (allies).

239. Plb. 1.26.6-7.

240. Philologus 56 (1897) 482–91Google Scholar.

241. Thiel 59, 189-90, 196; Walbank, , Commentary I.86Google Scholar; Brunt 421, 669.

242. Marchetti 125-30.

243. Livy 30.2.1-6, 27.8; 35.20.12; 37.2.10. Marchetti 128 gives the range as 50 to 75, but the last two passages could imply a higher rate. The value of these notices is very doubtful: for the formidable difficulties faced by those who seek to reconcile the annalistic and Polybian data on Roman fleets in the war against Antiochus see Thiel 258ff; Marchetti 96ff.

244. Such a force could probably have been accommodated for the crossing on Laevinus' fleet of 25+ quinqueremes, but some transports may have been taken as well.

245. Holleaux, , Rome 190 n. 5Google Scholar, based on the figure of 50 men to a lembus given by P.II.3.1; but note Walbank's reservations at Adams, W. L. and Borza, E. N. (ed.), Philip II, Alexander the Great and the Macedonian Heritage (1982) 226 n. 60Google Scholar. Hammond, , Epirus 609Google Scholar holds that Philip had also brought a force overland, but against this see Cabanes 253.

246. Explicitly at Livy 28.7.4.

247. Livy 27.30.2 (presumably they rejoined the fleet when it reached Naupactus).

248. Livy 27.32.2. That this force may have been partly drawn from the crews was observed by Kahrstedt 507 n. 2 and Holleaux, , Rome 248 n. 1Google Scholar; failure to recognize this has bedevilled many discussions. For the use of crews for fighting on land see Cato, , ORP³ fr. 66Google Scholar; Livy 9.38.2; 21.61.2; 22.31.3; 23.40.2; 26.17.2 (=Plb. 10.35.5); 34.29.5, 38.1-3; 37.16.11; Milan, , Critica Storica 2 (1973) 11–5Google Scholar.

Notes to Appendix II

249. For bibliography see Schmitt, Staatsverträge no. 536. The best statement of the case for autumn 211 is by Walbank, , Commentary II. 11–3Google Scholar. For recent defences of the dating to 212 see Lehmann 10-50; Crake, J. E. A., Phoenix 23 (1969) 216–20CrossRefGoogle Scholar.

250. McDonald, A. H., JRS 46 (1956) 157Google Scholar argued that Polybian material only begins at 24.7; against this see Badian, , Latomus 17 (1958) 198–200Google Scholar.

251. De Sanctis 440-3, followed by Walbank, , Philip 301–4Google Scholar, held that the treaty was concluded shortly before the Aetolian elections; at the elections Scopas was succeeded by Dorimachus, and at the time of the invasion of Acarnania and the attack on Anticyra Scopas was acting just as a leading Aetolian. But Livy is naturally taken to imply that Scopas was still general, and it is most unlikely that Dorimachus would have effectively delegated the conduct of the war to his predecessor in the first half of his generalship. This view also requires the rejection of hibernanti in 25.1.

252. The usual view that the meeting was a special assembly (cf. Hopital, R. G., RHDFE 42 (1964) 24 n. 15Google Scholar; Lehmann 16 n. 30) rests on Livy's words ad indicium ante ad id ipsum concilium (24.1) and on the fact that he does not state at 24.7 that Scopas had only just been elected. I doubt whether Livy's language at either point can carry this weight. Livy (P) 31.32.3 states that under their laws it was only at the regular Panaetolica and Thermica that the Aetolians were allowed ‘de pace belloque agere’. This statement is usually rejected as a misunderstanding by Livy (e.g. Nissen, H., Kritische Untersuchungen über die Quellen der 4. and 5. Dekade des Livius (1863) 127–8Google Scholar; Swoboda, H., Klio 11 (1911) 456–8CrossRefGoogle Scholar; Larsen, , TAPA 83 (1952) 16Google Scholar), but in my view on inadequate grounds. The motions passed at the assemblies at Plb.4.15.8-9 and 35.43.7ff were not straightforward declarations of war, and Livy, of course, did not mean that peace treaties could only be ratified at the regular assemblies.

253. Niese 476-80 and Costanzi, V., Studi Storici 1 (1908) 31–45Google Scholar, who dated the alliance to 212, supposed that the events described in Livy 26.25-26.3 all took place in the campaigning season of 211; on the conclusions which Holleaux based on this see above n. 63. Against the rather different interpretation of Livy offered by McDonald (n. 245) see Badian, , Latomus (1958) 200–1Google Scholar.

254. Walek-Czernecki, T., Rev. Phil. 11 (1928) 11Google Scholar; Lehmann 26-7.

255. Syracuse fell in either late 212 or spring 211 (Walbank, , Commentary II.7–8Google Scholar). Capua fell in 211.

256. Badian, , Latomus (1958) 202–3Google Scholar; Walbank, , Commentary II. 12–3Google Scholar. This solution was first proposed by Walek-Czernecki (n. 254), but he formulated the case badly: cf. Walbank, , Philip 303Google Scholar; Lehmann 36-7. Hopital, , RHDFE (1964) 23 n. 14Google Scholar notes the difficulty of finding room for Dorimachus' four (attested by IG. IX 1².1.30Google Scholar). If Scopas was in 211/0, Dorimachus may have held one of his tenures in 212/1.

257. Cf. above, at n. 19.

258. Cf. above, n. 22.

259. Livy 26.22.13, 26.4-27.17; Val. Max. 4.1.7.

260. Livy 26.22. As Lehmann 27-30 observes, there are suspicious similarities with the story of the 214 elections, and the statement that Marcellus was elected in absence (Livy 26.22.13) is almost certainly false.

261. Livy 26.22.1, Sulpicius assigned Macedonia immediately before the elections; 26.4, a despatch reaches Laevinus informing him that he has been elected consul in absence and that Sulpicius is on his way to succeed him. Lehmann's view (15, 29) that Sulpicius must have left Italy before Fulvius was summoned to hold the elections is quite unwarranted.

262. Livy 26.26.4. See especially Lehmann 40-2 (arbitrarily transferring the illness to Rome) and Crake, , Phoenix (1969) 218–9Google Scholar.

263. Above, n. 61.

264. On the calculations of Derow, , Phoenix 30 (1976) 265–81CrossRefGoogle Scholar, the Ides of March fell on 20 February (Julian) in 210, but the Roman calendar may have been running about a month later than this (cf. above n. 143).

265. Cf. Walbank, , Commentary II.12Google Scholar: ‘Were there no events of any consequence in Greece related by P. under 211 after the fall of Anticyra? And if there were, why has Livy omitted them?’