² For this see Livy xxxii 9.6-8; 14.7; 15.5-7; 16.4; and 18.2-5.

³ Livy xxxiii 1.1 ‘initio veris’ preceded the mustering of the Macedonian army at Dium. If we allow ten days for the call-up, Philip sent out his orders around 14 March, ‘primo vere’ (xxxiii 3.1).

⁴ At Cape Zelasium (Livy xxxi 46.7).

⁵ This is certain from the timing of the events, to which we shall come later. It is clear from Plb. xviii 20.3 that the crops were standing in the fields. Scotussa was a large city with a circuit-wall of some 3 km, and its territory was almost all to the west of the city.

⁶ If Flamininus had advanced through the gap by Pherae and headed north, Philip would have returned to make contact with Demetrias and Thebes; and he would then have occupied the Pherae gap and cut Flamininus' line of supplies.

⁷ Livy at xxxiii 6.3 distorted Polybius' account by saying that Philip, being near Larissa, had been informed that Flamininus had moved from Phthiotic Thebes to Pherae. For Polybius said at xviii 19.3 that Philip heard of Flamininus being encamped round Thebes.

⁸ Eretria—by the village now called Eretria (previously Tsangli)—has been described by Pritchett 140 and by I. Blum in AAA xv (1982) 159 ff.

⁹ See the discussion by Walbank 576, who summarises earlier views. Kromayer, riding in April from near Pherae to Sotussa, found Pritchett's river Onchestus dry and judged the Platanorrema by Ayia Triadha the only watercourse worthy of the name in the region (his p. 68).

¹⁰ The lack of reconnaissance is surprising, as indeed in the approach of each army to the Pherae gap; but it is incomprehensible if Flamininus really intended to invade Scotussaean territory on the other side of the Karadag range.

¹¹ This territory extended eastwards probably to the narrows by Dasolophos and Ambelia.

¹² Kromayer 67 f. and Stählin, RE s.v. ‘Skotoussa’, 614.

¹³ Kromayer 103 n. 3 and Pritchett 135 accepted Livy's figure of 500 for the Aetolian infantry as correct; they thereby reduced the total of Flamininus' forces, both Italian and Greek, to 27,000. Walbank, F. W. followed Kromayer in his Philip V of Macedon (Cambridge 1940) 167Google Scholar, but he gave a total of 32,000 in his Commentary, p. 572.

¹⁴ Kromayer 114 dated the battle to the end of May or beginning of June. He therefore postulated up to a month's delay (111); it was accepted by Walbank in his Philip 322, but less firmly in his Commentary, p. 579.

¹⁵ Livy's statement is vague. He moves from winter to spring at xxxiii 1.1, and he does not make it clear with which phrase ‘initio autem veris’ is to be associated. As we see from n. 3 above, he thought of spring beginning c. 14th March; he would hardly have asked Attalus to travel any earlier than that.

¹⁶ Kromayer 111 ‘die Kornerernte fällt heutzutage in Thessalien um die Mine des Juni’. If the crops had six weeks or so to ripen for harvesting, they would be good fodder for the cavalry mounts; but within, say, a month of harvesting they would be too rich and cause colic. I owe this information to a horseman, Corbett Capps of the National Humanities Center. Combine harvesting was completed by 18th June in 1986. Walbank 579 gave July as the harvest month.

¹⁷ Kromayer 115 tries to explain the origin of Livy's mistake.

¹⁸ Morgan 23-54; he refers in n. 3 to Y. Béquignon's excavation.

¹⁹ Since Peleus was represented as ruling over both Phthia and Pharsalus (lines 22 and 731), the Thetideum had to be on the edge of their territories' meeting place, that is not in the plains but somewhere overlooking the plains. J. D. Morgan has very kindly drawn my attention to the Scholia on E. Andr. 16-17, which, like Strabo, place the Thetideum between these towns, ‘Phthia and the Pharsalian town.’ For the view that there was also a place or town called Thetideum, see F. Stählin in RE vi.i (1936) 205 f. s.v. ‘Thetideion’ ‘Tempel der Thetis und Ortschaft’. He gave the evidence for Thetis being also a corn-goddess in Thessaly. See also Walbank 579.

²⁰ LSJ⁹ s.v. ‘άντί’ C i gives the meaning of the compound as ‘over against’ which I take to be the opposite of ‘alongside’ or ‘parallel to’. The πρός of πρός τό μεσόν shows in which direction the ridges were pointing. For κατά see Appendix I. We should retain άντιτεινόντων as the lectio difficilior and not emend read άνατεινόντων. Plutarch used άντιτείνω for a precise geographical description at Them. 8.3.

²¹ The Foundation elected me to a Fellowship for a month and the British Academy gave me a grant to travel. I am most grateful to both. Dr M. B. Hatzopoulos and his son Philip accompanied my wife and me on the reconnaissance. Dr Hatzopoulos gave me much excellent advice on matters of topography and tactics.

²² In 1835 Leake 472 described the river as ‘often to very formidable in winter and sometimes quite dry in summer’. Kromayer 118 wrotein 1907 of the only stone bridge in the area, that of Dendrakia. Where the river cut into the alluvium at bends, it created steep banks (Caesar, B.C. iii 88.6).

²³ Peek, W., Griechische Vers-Inschriften aus Thessalien (Heidelberg 1975) 7Google Scholar; and Tziaphalias, A. S. in Arch. Delt. xxx (1975) [1983] B 1 Chron. 196Google Scholar f.

²⁴ That there was a tumulus, said to be the burial-place of Amazons, ‘in Scotussaean territory and by the Cynoscephalae’ is known from Plu. Theseus 27 fin. What is uncertain is whether Philip mounted such a tumulus to address his men.

²⁵ This suggests that Livy forgot his reduction of the Aetolians from 6,000 to 600 infantrymen, or that his text originally had 6,000 infantrymen.

²⁶ The expression είς τούς έπιπέδους τόπους is not the same as είς τό πεδίον which occurs in Plu. Pel. 32.

²⁷ A picturesque touch but unlikely, when a man was carrying his pike; for Polybius mentioned the burdensomeness of a pike on the march at xviii 18.3.

²⁸ Whether ‘the right-hand part of the phalanx’ meant 8,000 phalangites or fewer is not clear. In addition Philip had his 2,000 Peltasts. If he had 10,000 altogether, he formed them at first 16 men deep and then 32 men deep, which would have given him a front of 300 metres width for the charge downhill. It is probable that the right-hand part of the phalanx meant something like 5,000 phalangites, in which case the front was of some 220 metres. At xviii 30.1 Polybius described what was evidently a normal charge; his phalanx then was 16 men deep. At Cynoscephalae this formation was doubled in depth, i.e., to 32 men deep, This weight of men gave an additional impetus at the moment of impact (xviii 25.2 τῷ βάρει τῆς σuνντάξεως and 30.4 αύτῷ γε μὴν τῷ τοῦ σώματος βάρει). In making his analysis of the Macedonian phalanx Polybius was thinking not of this battle but of the battle of Pydna, when two battle-lines did engage in an orthodox manner. See my article in JHS civ (1984) 42 for ‘the double phalanx’, then of 32 men. Kromayer 81 n. 1 and Walbank 582 put the single phalanx at 8 men and the double phalanx at 16 men; but at xviii 30.1 Polybius was writing of a normal phalanx formation and not of a double phalanx. Sec also Plu. Flam. 8.4-5. Kromayer is inconsistent; for at 323 with n. 1 he made the double phalanx at Pydna one of 32 men.

²⁹ The Romans must have been perfectly aware of this convention after many years of campaigning with and against the Macedonians. As at Pydna, the Roman commander wanted a battle of extermination for as long as there was action.

³⁰ Leake 458 was accurate in his translation.

³¹ Kromayer 72 ‘damit ging der Tag hin.’

³² In RE xii 1.35.

³³ He identified the river Onchestus with the stream by Mikron Perivolakion, which Kromayer found to be dry in April—not much use for thirsty men and horses.

³⁴ The account of Diodorus was ‘certainly founded on Ephorus’ (Westlake, H. D. in CQ xxxiii [1939] 17Google Scholar), a contemporary pro-Pelopidas writer, and ‘the agreement (between Diodorus and Plutarch) is closest in the two accounts of operations in Thessaly’ (ibid. 18). Plutarch, then, also used Ephorus for our battle; it is unlikely that he went beyond Ephorus to consult e.g. Callisthenes.

³⁵ If Alexander had as many as 8,000 hoplites and we allow a metre to each man in a phalanx 8 men deep, his line was about one kilometre in length. That this number of hoplites is about right is shown by his alarm in 368 BC, when 8,000 Boeotian hoplites moved against him (D.S. xv 71.3).

³⁶ See LSJ s.v. ‘συνίστημι’ B I, citing X. An. vi 5.28-30.

³⁷ They saw the cavalry, when the cavalry appeared at the foot of the valley, or on the western side of the ridge opposite Hill 260.

³⁸ See n. 34. Sordi, M., La Lega Tessala (Rome 1958) 193–202Google Scholar maintained that Plutarch followed Callisthenes for the battle, and that Diodorus followed Ephorus; however, Diodorus abbreviated drastically, and the little he does say in xv 80.4-5 seems to me to be in line with Plutarch's much fuller account.

³⁹ Kromayer 116-122 made no attempt at all to relate the detail of the battle to his location of the Cynoscephalae; instead, he indulged in his theory of an offensive wing and a defensive wing, which is not relevant to my understanding of the battle. Pritchett, placing the Thetideum at Ayios Athanasios (114-117), located the battle on the ridge north of Sitochoron (118 with Fig. 12). Pritchett 119 brought the cavalry into action prematurely against ‘the infantry on the hill’, Since Pritchett has put the scene of this battle and the scene of the battle of 197 BC at different locations, he has to plead as follows. ‘No modern scholar can agree with this conclusion’, expressed by Plutarch, who ‘believed that both battles of Cynoscephalae, those of 364 BC and 197 BC, were fought on the same ground’. I beg to differ; for I think that Plutarch was the more informed on this matter.

⁴⁰ See the excellent note of Kromayer 67 n. 1 on περικεκλασμέναι. He compared the passage at Plb. ix 26.7, where hilly, flat and undulating (or uneven) sites for towns are contrasted. Kromayer translated as ‘uneben’.

⁴¹ Y. Yadin, Masada (Jerusalem 1966) 215-19

⁴² F. L. Lucas in BSA xxiv (1921) 52, in his study of the battle of Pharsalus in 48 BC, wrote as follows. ‘It is my regret that, though I must have walked a hundred miles up and down and round the Pharsalia, I found no sign of Roman fortification.’ Many others will have echoed those words.

⁴³ Schulten, A., Numantia iv (Munich 1929) 37Google Scholar with Plans IIIa and IV, his Camp 1. For the plans of a permanent camp see The Cambridge Ancient History viii facing p. 322 and Walbank i 710.

⁴⁴ Morgan placed Caesar's camp in the plain on the north side of the river about opposite Vasile.

⁴⁵ See CAH ix 659 ff.; and for the camp in the plain of Pharsalus Caesar, B.C. iii 94.5 and 95.1-3.

⁴⁶ This possibility was pressed by J. D. Morgan in correspondence for which I am grateful.