¹ Tarn, W. W., Alexander the Great (Cambridge, 1948) ii 135 ff.Google Scholar; Berve, H., Das Alexanderreich (München, 1926) i 103 ff.Google Scholar All references to these authors, unless otherwise stated, are to pages of these volumes.

² I am indebted for critical and constructive comments on an earlier version of this paper to E. Badian and G. T. Griffith, who need not of course be supposed to agree with all that it now contains.

³ Plutarch 16 agrees with Arrian (a) that Alexander crossed the river in the face of Persian opposition, against Parmenio's advice (contra Diod. xvii 19); (b) about the combat between Alexander and the Persians, Rhoesaces and Spithridates (contra Diod. 20; Curt, viii 1.20); (c) in mentioning the attack on the Greek mercenaries (whose presence Diodorus ignores); he rounds off their casualties at 20,000, whereas Arrian i 14.4 puts the strength of the mercenaries at just under 20,000 and states (16.2) that almost all were killed, except for 2000 prisoners. Plutarch used Aristobulus, whom he cites as stating that on Alexander's side only 34 men were killed, of whom 9 were foot-soldiers. Now Arrian i 16.4 says that 25 hetairoi were killed in the first shock (none later?), that the other cavalry lost 60 and the foot about 30. I take it that the official version of casualties gave Macedonian losses as 25 hetairoi and 9 foot, and that Aristobulus ignored other losses, or else that Plutarch overlooked his statement of them.

⁴ The Paeonians were commanded at Issus and Gaugamela by Ariston, the other prodromoi by Protomachus at Issus and Aretas at Gaugamela (Arr. ii 9.2; iii 12.3 cf. Plut. 39.1; Curt, iv 9.24), perhaps earlier by Hegelochus (i 13.1). At the Granicus Amyntas had a general command over both, and over other units (i 14.6; cf. 12.7).

⁵ Berve, 129; Tarn, 157: see Arr. i 14.1 with 14.7; iii 12.3; Curt, iv 15.13. The equation is certain, but as Mr Griffith has pointed out to me, puzzling. The sarissa used by the infantry required the use of both hands and must have been hard for horsemen to handle; perhaps we should assume that their sarissa, though longer than spears of the Companions, was shorter than that of the foot.

⁶ He argued partly from Arr. ii 8.9 which is corrupt and cannot be used.

⁷ I agree with Berve, 134 against Tarn, 158 (but cf. 160 n. 1), that the Thracian cavalry probably remained in Media under Agathon (cf. Curt. x 1.1) and were not sent home in 330. Allied contingents were then sent home, but the Thracians were subjects, and Alexander actually received reinforcements from Thrace in 331 and 326 (Diod. 65.1; Curt, v 1. 40–1; ix 3.21). The distinction Tarn draws (e.g. 160 n. 1) between Thracian and Odrysian cavalry in Alexander's army is invalid (cf. Arr. i 14.3 with iii 12.4).

⁸ E.g. hippakontitai (Arr. iii 24.1) and hippotoxotai (v 12.2; 16.4).

⁹ Berve, 134; Beloch, K. J., Gr. Gesch. iii ² 2.325.Google Scholar It is indeed in my view not necessary to amend the figures in Diodorus' text (cf. section IV), and Dr Badian has pointed out to me that he need not be anachronistic in making Erigyius, not Philippus, commander of the allied horse; Philippus may have been only in temporary command of them at the Granicus (contra Berve, ii nos. 302 and 779, with evidence). Cassander cannot indeed have been in command of the Paeonians and Thracians (cf. Berve, ii no. 414), but Badian thinks that Beloch's correction to Asander (Berve, ii no. 165) is probable ‘in view of the context of Parmenio's stranglehold on senior appointments’ (cf. TAPA xli (1960) 327–8). This, then, is at least one manuscript error, and the transposition of ‘prodromoi’ suggested is not hard.

¹⁰ Curt. viii 1.19; 2.14 places here the appointments recorded by Arr. iv 17.3 and interposes between them the death of Clitus; Arr. iv 8 gives no clear date.

¹¹ Berve, 107 ff., put the reorganisation in 329 on the strength of a mistranslation of Arr. iii 30.6 (cf. Tarn, 163 n. 4); and thought that the hipparchies numbered 4. For the agema as a hipparchy cf. Tarn, 164 n. 1.

¹² Schachermeyr, F., Alexander der Grosse (1949) 293 Google Scholar, cf. n. 192, still follows Berve in holding that former commanders of foot-battalions who became hipparchs also retained their previous commands; I believe Tarn refuted this (142 ff.), but it is to be regretted that Schachermeyr has not yet fulfilled his promise to develop his views on Alexander's army.

¹³ Cf. Section IV.

¹⁴ CQ (1962) 143 ff.

¹⁵ The Lost Histories of Alexander the Great (Amer. Philol. Association, 1960) ch. viii. Tarn reached the same conclusion on Clitarchus' date, but for the wrong reasons.

¹⁶ See Callisthenes (Jacoby, no. 124) F. 35 = Polyb. xii 19.1; Plut. 327DE for Anaximenes (Jacoby, no. 72) F.29 (cf. Plut., Alex. 15.i), Aristobulus (Jacoby, no. 139) F.4 and Ptolemy (Jacoby, no. 138) F.4. Cf. Arr. i 11.3 (Ptolemy?); Justin xi 6.2 (Clitarchus?); Liv. ix 19.5; Frontin., Strat. iv 2.4.

¹⁷ Cf. Tarn, 159: ‘probably … the official figure, but in any case it is not likely to be too high’. Contra Beloch, op. cit. (n. 9) 33–5.

¹⁸ See next section and table in Berve, 179.

¹⁹ Berve, 179.

²⁰ Curt. iv 1.34–5; 5.13; Memnon (Jacoby, no. 434) F.1, ch. 12.4; Diod. xviii 22.4.

²¹ E.g. by Beloch, op. cit. (n. 9) 323; see contra J. Droysen, G., Hermes xii (1877) 227 ff.Google Scholar; Griffith, G. T., Mercenaries of the Hellenistic World (Cambridge, 1935) ch. i.Google Scholar

²² Diod. xvii 16.2; 17.3; Arr. i 11.6.

²³ Justin xi 5.8 names Amyntas, Diod. xvi 91.2; xvii 2.4 ff.; 5.1–2 only Parmenio and Attalus, xvii 7.10 Calas.

²⁴ Griffith loc. cit. (n. 21). It is not certain that the 200 horse left at Halicarnassus (Arr. i 23.6) were mercenaries.

²⁵ Tarn argues (i) that Alexander could not have given Antipater more than half the phalanx; (ii) that Antipater could not have had any hypaspists, of whom Alexander had 3000; therefore his foot could not have been equal in numbers to Alexander's. But (i) is assumed, and falsified by the evidence on Alexander's reinforcements and the reserves of Macedonian manpower. The common assumption that there were 12 territorial regiments of which Alexander had 6 is mere conjecture; in any case the strength of these regiments must have varied as reinforcements were received, and it is ludicrous to suppose that in 334 or later Alexander completely denuded of young men the territories from which his regiments were drawn.

²⁶ Bevölkerung der gr.-röm. Welt (1886) 209. His estimates of Macedonian, population, somewhat greater in Griech. Gesch. iii ².1 294 ff.Google Scholar, do not take adequate account of the levies in and just after Alexander's reign, described in the text.

²⁷ Isocr. viii 118 estimated Thessalian cavalry at 3000. In the Lamian war, a life and death struggle fought in their own country, the Thessalians, most of whom were involved, put 2000 into the field (Diod. xviii 15). Jason allegedly claimed that Thessaly could provide 6000 (Xen., Hell. vi 1.8) and is said to have raised over 8000 from Thessalians and allies (ibid. 19), but these figures must surely be exaggerated, or include mercenaries (cf. ibid. 5); and in other fourth-century data we cannot distinguish Thessalian horse from that of their allies or mercenaries.

²⁸ Droysen, H.'s suggestion, Untersuchungen über Al. d. Gr. Heerwesen u. Kriegführung (1885) 39 Google Scholar, that he included the neogamoi presupposes gross carelessness by Callisthenes or Polybius or both. Mr Griffith points out to me that Polybius is unlikely to have misreported Callisthenes here, when he is picking holes in his account. He suggests that Alexander was now able to increase the size of his force, because phoros was now coming in, and would increase with further conquests; Macedonian revenue was inadequate to support the large expeditionary force of 334. And in 334 Alexander did not expect to meet the Persian Grand Army and so had at first less need of men. Tarn, 158 n. 1, dismissed Callisthenes' evidence on the ground that he was no ‘authority on military matters’; but ancient and modern criticisms of Callisthenes as a military historian seem to me largely misconceived, and in any case no military expertise was needed to state numbers correctly.

²⁹ Amyntas (Berve, ii no. 57) was sent for them in autumn 332, after the capture of Gaza (Diod. xvii 49.1; Curt, iv 6.30); as they reached Susa by late 331 (supra), they must have been sent off long before Agis' defeat in October 331.

³⁰ Cf. n. 26. Berve, 's a priori argument (p. 117)Google Scholar that Alexander did not want reinforcements of politically unreliable Macedonians cannot stand.

³¹ See Arr. iii 20–1; 28.1; 28.8 (cf. Str. xv 2.10; Curt. vii 4. 22 ff.); iv 21 and 23 ff.; Str. xv 1.17 and 27 with Arr. v 27.5–6.

³² Hermes lxxx (1952) 456 ff. Ptolemy of course concealed the extent of the disaster and extenuated the motives for the march; Tarn follows him uncritically. Strasburger guesses that Alexander may have lost some 50,000 men in Gedrosia but concedes that the Macedonians may have come off with least loss.

³³ Arr. iv 4.1, cf. perhaps 22.5; 24.7. Generally the types of settlers are not specified.

³⁴ Arr. vi 14.4 I am now convinced that the partitive genitive implies that there were more Companions at the time (contra Tarn, 162, discussed in Section VI). Berve wrongly says (p. 110) that Arrian here designates the Companions as a part of the cavalry and then oddly goes on to infer that in a wide sense Oriental cavalry could be so described. To fit his view, Arrian should have written

^34a On Craterus' slow progress home see Badian, E., JHS lxxxi (1961) 34 f.Google Scholar

³⁵ I doubt if half the 12,000 foot with which Alexander crossed to Asia would have survived (but we must allow for some others with the original expeditionary force); perhaps Diodorus' source distinguished between (i) men who had fought at the main battles, Issus and Gaugamela, and (ii) later reinforcements.

³⁶ See table on p. 39.

³⁷ Some 12,000 Balkan troops and 80,000 Greeks (cf. Berve, 182, for table).

³⁸ Berve, ii nos. 794, 501; the statement that the troops came ‘from the sea’ (Arr. vii 24.1) does not exclude my view.

³⁹ The statement in Diod. xix 41 that the 3000 argyraspides in Eumenes' army in 316 were veterans of all Philip's and Alexander's wars and men of 60 to 70 is (as Tarn says, p. 151) silly; but what doubtless underlies it is that some were veterans of some campaigns conducted by Philip or Alexander; for instance, men who had fought at Chaeronea (and then in Asia) need not have been more than 40 in 316.

⁴⁰ We do not know how many Macedonian horse or foot were in Asia before 334, or whether Macedonian prodromoi are included in Diodorus' 1800 horse, cf. supra.

⁴¹ Droysen, op. cit. (n. 21) 237 held that the ilai must have varied in strength; his arguments are unsound. Though they were still basically territorial in recruitment in 331 (Arr. iii 16.11, contra Curt. v 2.6), it does not follow that they must have differed in strength with the differing populations of the recruiting areas, as (a) everywhere there were great reserves of manpower; (b) deficiencies in one area could have been supplied from another, without destroying the territorial character of the ilai.

⁴² In the Roman army, when the nominal strength of a legion was 6200 men (Festus 453 L), but the real strength often much less, multiples of 6000 are often unrealistically employed to designate the number of legions (e.g. Plut., Cic. 36.1; App. BC iii 58; Bell. Mith. 72). Berve's view of course implies that a nominal strength was given, below the real strength.

⁴³ Anaximenes (Jacoby, no. 72) F.4, by ascribing the institution of lochoi and decads, and also the establishment of both hetairoi and pezetairoi, to a king Alexander implies at least that all were earlier than Philip, though in my view nothing else can be based on this much-discussed statement. I do not doubt that Anaximenes was fathering on Alexander I the whole organisation of the Macedonian army, as Philip found it, but that this is quite unhistorical; Macedon had no good infantry till Archelaus at least (cf. Thuc. ii 100.1–2 and 5; iv 124 ff.); on the other hand we cannot amend to ‘Archelaus’ reven if that were palaeographically likely), nor assume that Alexander II, who ruled only for one disturbed year, is meant, since the Macedonian cavalry were noted even before Archelaus (n. 47). No one king was responsible for all these measures, and none ‘accustomed the most distinguished men to ride horses and called them hetairoi’; horse-riding and the name, hetairoi, were surely of immemorial antiquity.

⁴⁵ Cf. Theopompus (Jacoby, 115) 225 (b) at end; Diod. xvi 34.5; Ps-Dem. vii 41; SIG ³ 332; Arr. Ind. 18.4 and 10 (Nearchus at Amphipolis). Hampl, F., Der König d. Maked. (1934) 22 ff.Google Scholar, adduces some other texts which may be relevant. There is much evidence and exaggeration on the expropriation or enslavement of the citizens of towns taken by Philip; doubtless many pro-Macedonian inhabitants remained (and did service for their lands), but much was clearly available for distribution to Macedonians and foreigners enrolled in the Companions.

⁴⁶ Dem. vi 20; Liban. Hyp. to Dem. i, section 2.

⁴⁷ Thuc. i 61.4; 62.3; iv 124; especially ii 100.5. One Upper Macedonian principality, Elymia, could put 400 horse into the field in 382 (Xen. Hell. v 2.40).

⁴⁸ Plut. 66. His figure of 120,000 foot roughly agrees with Nearchus' total of over 120,000 for horse and foot (Arr., Ind. 19.5).

⁴⁹

⁵⁰ P. 111.

⁵¹ See section V.

⁵² This objection applies against Tarn's assumption (p. 165) that the fifth hipparchy had a much smaller Macedonian squadron than the rest. If it was largely barbarian, why are we not told how it was composed, as with the other hipparchies ?

⁵³ Perhaps read ἐς αὐτήν for ἐς αὐτό, and delete as a gloss; τε is curiously unemphatic, in contrast to καί, joining each other grievance. Read in that case: The partitive genitive τῶν βαρβάρων now reads better; with the manuscript text οἱ βάρβαροι (vel sim.) might be expected.

⁵⁴ Oxyathres is attested as a hetairos (in the narrower sense) by Plut. 43 and Curt. vi 2.10. But the title cannot have been withheld from other Iranians high in Alexander's service. And Iranians in the agema must have been hetairoi, at least in the broader sense that all Macedonian cavalry were.

⁵⁵ But cf. Arr. vii 11.3, for Alexander's action in pique when the mutiny broke out; we do not know if he persisted in it later.

⁵⁶ Berve, 112.

⁵⁷ I do not understand Tarn's statement (p. 167) that after Alexander's death Macedonian troopers are not found; they are well attested in the civil war of 321.

⁵⁸