¹ P. H. L. Eggermont, ‘Alexander′s Campaign in Gandhara and Ptolemy′s List of Indo- Scythian Towns’, Orientalia Lovanensia Periodica 1 (1970), 63–123, followed by an excellent map. [Cited ‘Eggermont 1’.] All geographical identifications in the present article follow his work, and the map accompanying this article is based on his map. The only exception is in the case of the ‘Aornus’ rock, where he has himself since changed his view: see OLP 15 (1984), 191–233 [cited ‘Eggermont 2’], where his map will again be found (following p. 230), but where he rejects Stein′s identification of Aornus with Pir-sar, which Stein was specially proud of as one of his principal topographical achievements, and instead accepts the identification with Mt Ham revived by Sir Olaf Caroe. (It is not yet in Caroe′s book, The Pathans 550 BC - AD 1957 [1965].) On geographical and historical grounds, Pir-Sar seems too far out of the way (as Eggermont now points out) to be the nerpa 17 iv rrj x"PZ (A. 4.28.1) of the tribesmen around Bazira, whereas Mt Ham (′dividing Swat from Buner": Caroe, op. cit. p. 499) seems suitable in location.Google Scholar

² To illustrate his lack of familiarity with modern scholarship on Arrian, and on the Alexander sources in general, see (e.g.) Eggermont 2, 194: ‘Arrianus is regarded by classic scholars as a primary source [sic] because he based his story on Ptolemy′s Royal Diaries.’ I doubt if anyone apart from N. G. L. Hammond would nowadays maintain this. And compare his dating of Cleitarchus after 258 B.C., and probably in the second century, on the strength of a complex and confused argument involving Pliny′s Natural History and the Asoka Edicts (Alexander′s Campaigns in Sind and Baluchistan [1975], p. 67 n. 262; repeated Eggermont 2, 193, as a basis for further argument about the Alexander sources). I hope that this traditional gap between experts on India and Alexander scholars can now be overcome.Google Scholar

³ For the various spellings of the name, see Roos ad Arrian vol. 1, p. 216, with Wirth′s additional note p. 444; and cf. RE s.v. FFevKeXawTts. No consistent differences (e.g. between town and region, or between different periods) can be disengaged. Arrian is the main source for events here discussed. Curtius and Diodorus add one or two items, which will be noted in their placeGoogle Scholar

⁴ Eggermont (2, 205 n. 53, with a reference to a Dutch dissertation of 1930) seems unaware of A.′s carelessness and inconsistency regarding technical terms and is puzzled by the term: ‘It is unknown what power the hyparchos has.’Google Scholar

⁵ A. 4.22.8:

⁶

⁷ Thus already A. E. Anspach, De Alexandri Magni Expeditione Indica (1903), p. 13, not known to Eggermont. More recent scholars, where they mention Astis, also tend to be unaware of Anspach and get the time of the revolt wrong; thus N. G. L. Hammond, Alexander the Great (1980), p. 200 (also ignorant of Eggermont′s work). The basic facts therefore need restating.

⁸ Contrast the operation at Orobatis (4.28.5), where A. mentions both the commanders and where, after completing its task, the army departs for its appointed duties on the Indus bank (cf. below).Google Scholar

⁹ Although Eggermont 2 gives a slightly different account of Omphis‘ accession and early relations with Alexander, the author maintains his contention (originally argued 1, 102ff.) that Omphis ruled on the west bank of the Indus as well as east of it (2, 204f.).Google Scholar

¹⁰ Eggermont 1, 74. Cf. (e.g.) P. A. Brunt in vol. 1 of his Loeb edition of Arrian, pp. 507f.; also 2, 455. The anecdote related by Chares (FGrHist 125 F 16), which Brunt cites (p. 439 n. 1) as perhaps confirming April 326, is irrelevant. If it refers to this operation at all, it merely shows that it took place in winter, which we already know from Aristobulus. There is no trace of any more specific information.Google Scholar

¹¹ Caroe, op. cit. (n. 1), pp. 51–3, also showing the importance of the site, which controls the entrance to the whole plain; cf.Google Scholar

¹² It is possible that he has, since Bazira was more centrally placed for this particular campaign. On the other hand, Massaga was strategically much the more important place, both when Alexander arrived (as A. makes clear) and for the future protection of his conquests (see last note). It is therefore possible that A. has got it wrong, misinterpreting his source as he demonstrably does in some other cases.Google Scholar

¹³ My suggested calculations allow for the ten days specified by Curt. 8.10.17 as devoted to Bacchic revels at Nysa: whatever the facts about the latter, Curtius often gives precise day counts (e.g. for the stay in Babylon, or, in our own context, for the march from Embolima to the Indus crossing: see below), which we ought not to reject. A. adds the story of Nysa from Vulgate sources in Book 5 (1.1; cf. 2.5f.), for compositional reasons, after completing his account of the conquest of the country up to the Indus from his main sources in Book 4. As regards Nysa and Dionysus, he expresses pious unwillingness to disbelieve, no doubt modelled on Herodotus’ attitude to things divine and stories about them, but (2.6) he tones down Curtius‘ revels. I am not convinced by Brunt′s suggestion that the story must come from Aristobulus (op. cit. [n. 10], 2, 435ff.), including even Acuphis’ lecture on the basis and the excellence of aristocratic government (cf. also 2, 531). The very fact that A. adds the story at the beginning of a new book serves to show that his ascription of it to a logos (i.e. not one of his two main sources) must be taken strictly. That Nysa as a town was mentioned by one of his two main sources, we cannot tell which (Berve, Alexanderrekh 2, 17 suggests Ptolemy, while Brunt prefers Aristobulus - neither gives any reason that might help us decide), is shown by the reference to its cavalry force at 6.2.3. But we do not know that that source had mentioned more than the capture of the place and the demand for that force. In any case, A., already planning his long appendix on the logos for the beginning of his next book (for his composition was never hand-to-mouth), decided to omit any reference to Nysa in his main account of the campaign in Book 4, in order to give it self-contained prominence there. Whatever his main source(s) said about the town thus dropped by the wayside.Google Scholar

¹⁴ On the identification of Aornus see n. 1. It is clear that, if Mt Ham is correct, the marches would be much shorter than they would be in the case of Pir-sar.

¹⁵ See E. J. Bickerman, Chronology of the Ancient World1 (1980), p. 112.

¹⁶ Aristobulus is cited as reporting that he saw only snow in the mountains and saw rain for the first time at Taxila. On the extended time-table usually proposed for the whole campaign, this does not seem conceivable.Google Scholar

¹⁷ Eggermont′s identification of Orobatis (1, 71–4) is erudite and entirely convincing. However, he is confused about the chronology, suggesting that Alexander there ‘first joined the army of Hephaestion and Perdiccas’. This is impossible, for according to A. that army, after fortifying Orobatis, had long ago departed to carry out its instructions on the bank of the Indus. Alexander did not rejoin it until he reached the Indus crossing. Brunt appears to miss the fact that the account of Orobatis is a flashback: his translation, at any rate, suggests that the action takes place at the point where it is narrated. A.′s wording, where he describes the task of the army corps at the Indus, which it was ‘now engaged’ (cTrpaaoov ^8rj) in completing when Alexander reached Orobatis, carefully echoes the instructions he had reported them as receiving (22.8)

¹⁸ Eggermont (1, 69) at least notes the puzzle, even though his explanation-that the ‘so-called surrender of that town to Alexander cannot have meant any more than an official act’-seems to me unacceptable, if (as he in fact knows) the town had earlier been taken over by Hephaestion and had received a governor: no such official act would be needed, or could even be reported. We certainly have no parallel for such a situation. However, not all recent historians of Alexander have even commented on the puzzle.

¹⁹ Compare the way in which modern readers can be misled even where A. is at pains to make the sequence clear (cf. n. 17). The aorist is, of course, the normal way of expressing either action at the particular point in the narrative or action at some earlier point (since the pluperfect is rarely used): the ‘past definite’ use and the ‘pluperfective’ (or ‘past anterior’) use are not technically distinguished, and are occasionally difficult for the modern reader to distinguish, whether or not the author has deliberately taken advantage of the ambiguity. Diodorus provides a long list of obvious examples. Where the author wants to establish the ‘past anterior’ use, he can, of course, use various stylistic devices, as A. does in the Orobatis affair.

²⁰ See Eggermont 2, 198. His identification of that town (1, 91–4) is again a masterpiece of scholarly argument, based (like that of Orobatis) on knowledge of both the actual topography and the Indian sourcesGoogle Scholar

²¹ This article is an incidental product of my work in Trinity Term 1985, which I was able to spend in Oxford owing to a grant from the John Simon Guggenheim Foundation. I should also like to acknowledge the kindness of St John′s College, and help received from Dr Ellen Rice.Google Scholar