* The following abbreviations are employed

AG² = Hammond, N. G. L., Alexander the Great: King, Commander and Statesman (New Jersey, 1980Google Scholar; second ed., Bristol, 1989).

Berve = Berve, H., Das Alexanderreich auf prosopographischer Grundlage (Munich, 1926).Google Scholar

Borza = Borza, E. N., ‘Fire from Heaven: Alexander at Persepolis’, CPh 67 (1972), 233f.Google Scholar

Bosworth C = Bosworth, A. B., A Historical Commentary on Arrian's History of Alexander, i (Oxford, 1980).Google Scholar

Bosworth CE = idem, Conquest and Empire: the Reign of Alexander the Great (Cambridge, 1988).Google Scholar

Goukowsky = Goukowsky, P., Diodore de Sicile Livre 17 (Paris, 1976).Google Scholar

Hamilton C = Hamilton, J. R., Plutarch, Alexander: a Commentary (Oxford, 1969).Google Scholar

Schmidt = Schmidt, E. F., Persepolis i-iii (Chicago, 1953–1970).Google Scholar

THA = Hammond, N. G. L., Three Historians of Alexander the Great: the So-called Vulgate Authors, Diodorus, Justin and Curtius (Cambridge, 1983).Google Scholar

¹ Karamitrou-Mentessidi, G., Aiani of Kozani (Thessaloniki, 1989), especially p. 56 and Fig. 23.Google Scholar

² Despini, A. in AAA 13 (1980), 206Google Scholar; illustrated in Ancient Macedonia (Athens, 1988), Figs. 226–9.Google Scholar

³ Ergon, 1984 [1985], 32 and 1985 [1986], 19f.Google Scholar

⁴ Andronikos, M. in BSA 82 (1987), 11.Google Scholar

⁵ Schmidt i.75.

⁶ Ibid.

⁷ Schmidt i. 179.

⁸ Ibid.

⁹ Schmidt i.76

¹⁰ Schmidt i.188.

¹¹ Schmidt ii.99; cf. ii.97.

¹² Schmidt i.81, 131, 179 and 185. Also, of course, much furniture and fine cloth.

¹³ Bosworth C, p. 332.

¹⁴ Schmidt ii.3.

¹⁵ Ibid.

¹⁶ At any rate before Arrian wrote; for I am not persuaded by the arguments of Buraselis, K. in favour of a later date in Ariadne 4 (1988), 244f.Google Scholar

¹⁷ So Goukowsky, p. 98 n. 3, Bosworth CE, p. 93 and THA, pp. 56f. and 131 f.; Borza 234 expressed doubts.

¹⁸ I am following my chronology in AG ², p. 313. This date is usually accepted; see Borza 237.

¹⁹ They had borne the brunt of the heavy fighting at the Persian Gates. Alexander had rewarded good service after the battle of Issus by sending the Thessalian cavalry under Parmenio's command to capture the booty at Damascus (Curt. 3.12.27; Arr. 2.11.10; Polyaen. 4.5).

²⁰ Goukowsky claimed that the plundering was to be not of the city itself but of the ‘ville royale … y compris les Palais et la Trésorerie’ (p. 100 n. 2). That is not what was described by Diodorus and Curtius. Diodorus called Persepolis the capital city of the Persian kingdom (17.70.1, µητρ⋯πολιν … τ⋯ς Περσ⋯ν βασιλε⋯ας) and a city surpassing the other cities in misery when it was sacked (17.70.6 τ⋯ν ἄλλων π ⋯λεων), and Curtius wrote similarly of Persepolis as ‘urbs’ and ‘oppidum’ (5.6.1–2). In their accounts it was the city which was plundered. Diodorus commented on the wealth of the private houses (τ⋯ν ἰδιωτικ⋯ν οἴκων) even of the masses (τοῖςπλ⋯θεσιν), which were of course not on the acropolis but in the city. Both authors described the massacre of the men, the suicides of entire families and the outrages committed on the women by the troops in the sack of the city. Whether their accounts included a looting of the acropolis area will be considered later.

²¹ The digression was derived probably from the work of Deinon, perhaps via his son Cleitarchus; see Goukowsky, p. 222 on Diod. 17.71.7 and THA, p. 57.

²² This campaign corresponded at least in part with that which Diodorus appended to the Thaïs episode at 17.73.1 and with the campaign in Paraetacene which Arrian mentioned at 3.19.2. The Mardi whom Curtius mentioned at 5.6.17 were a different tribe from the Mardi of Hyrcania.

²³ See Hamilton C, p. 98 for a discussion of the months involved; Borza 241 had Alexander leave Persepolis in mid-May.

²⁴ The usual view has been that the city was looted in January, the acropolis was exempted then (despite Diod. 17.70.3 and Curt. 5.6.5), and the acropolis building – ‘the Palace’ in the texts – was burnt in mid-May. So Bosworth CE, p. 94 ‘Alexander turned over the city to his troops… only the palace complex was exempted… the fate of the palace must have been discussed during the months of occupation… the question was resolved dramatically' by the Thaïs episode. With this scenario the Palace was never looted at all: its store of valuables must have been involved in the conflagration started suddenly by Thaïs. This lack of looting is directly contrary to the archaeological evidence and is therefore unacceptable. Borza 243 had given a different sequence: looting of the city but not the palace complex; ‘Alexander ordered the systematic looting of the royal monuments’ (I suppose he meant the Palace, the Throne Room and the Treasury); and then in mid-May the burning of the Palace à la Thaïs. Here he invented the intermediate stage; it is not in the Cleitarchan version at all, and the archaeological evidence does not support a ‘systematic’ form of looting.

²⁵ Before the excavation of Persepolis Berve ii. 175 accepted the Thaïs episode as historical, Wilcken, U., Alexander the Great (London, 1932), p. 145Google Scholar called the Thaïs episode ‘one of the fables of Cleitarchus’, and Tarn, W. W., Alexander the Great, ii (Cambridge, 1948 and 1979), p. 48Google Scholar ‘I need hardly say that there is not a word of truth in the Thaïs story’. Since Schmidt's publication most scholars have accepted the Thaïs episode without argument and reproduced the sensational description of brutalities and atrocities and then the drunken revel in the Cleitarchan manner, e.g. Lane Fox, R., Alexander the Great (London, 1973), pp. 259–64Google Scholar and Bosworth CE, pp. 92f.

²⁶ It is of course possible that he did return to Persepolis, because Arrian was abbreviating the account of his sources drastically. But if Alexander did return, Arrian's silence indicates that nothing spectacular or important happened at Persepolis.

²⁷ Attempts to import Thaïs into the account given by Arrian, as in Borza 235, are misplaced; for the discussion between Alexander and Parmenio was a sober consideration of policy and not part of a drunken revel with prostitutes and fun and games. Strabo 729/730 also gave Alexander's reasons.

²⁸ I am following the chronology of AG ², pp. 159f. The dating of the war of Agis is disputed (see Brunt, P. A. in the Loeb edition of Arrian, Anabasis, i [London, 1976], pp. 480–5)Google Scholar; but even an earlier dating is compatible with ‘the possibility that the news of the rebel collapse could have greeted Alexander almost any time after mid-December 331 B.C.’ (Borza 242), and so after the burning of the Palace in January.

²⁹ Such lies were the visit of the Amazon Queen to bed down with Alexander (Plut. Alex. 46.1; cf. Diod. 17.77.1–3 and Curt. 6.5.24–32) and the participation of Ptolemy in saving the life of Alexander at the city of the Malli (Curt. 9.5.21). For his fictional battle-narratives see THA, pp. 13–27.

³⁰ It is likely that the work published by Callisthenes did not extend beyond the battle of Gaugamela, and that in any case it did not cover the burning of the Palace; for Alexander's official version of what he did at Persepolis would surely have been cited by the surviving Alexander-historians. The river Araxes which Callisthenes mentioned (Strabo 531) was not the river of that name near Persepolis.

³¹ Arrian described ‘Ptolemy’ as holding a command during the action at the Persian Gates (3.18.9). He is probably our Ptolemy, the son of Lagus; for he was mentioned with his patronymic in command of a much larger force which captured Bessus a year later (3.29.7). Aristobulus may have been in the party which went ahead to Pasargadae; for his description of the contents of the Tomb of Cyrus before it was robbed was surely that of an eye-witness (6.29.4–6).

³² I am including as historical the action by Tiridates and the mention of one day of looting, which figured in Diodorus and Curtius because such details might well have been reported accurately to Cleitarchus.

³³ He was named in Diod. 17.69.1 and Curt. 5.5.2 and 5.6.11: see Berve ii.374f.

³⁴ The fact that the army stayed there so long shows that the city of Persepolis was not the desolation which Diodorus and Curtius had depicted, with the males slaughtered, the women and children enslaved, and some buildings burnt according to Curt. 5.6.7 (cf. 5.7.4–5 urbem, and 5.7.10 urbem… deletam).

³⁵ I have not included in this article the account of Alexander's Persepolis in Plutarch. Alexander 37–8. As regards the Thaïs episode Plutarch added little to what Diodorus and Curtius had already written, and his other stories concerning the statue of Xerxes and the tears of Demaratus are not relevant here. I have discussed them in a book which is forthcoming on the Sources of Plutarch's Alexander and Arriaris Anabasis Alexandrou. Justin 11.14.10 is too brief to be of significance.