¹ See e.g. Sir Tarn, W. W., Alexander the Great i (Cambridge 1948) 72Google Scholar [hereafter Tarn]; Schachermeyr, F., Alexander der Grosse, SÖAW Wien cclxxxv (1973) 348–54Google Scholar [hereafter Schachermeyr]; Fox, R. Lane, Alexander the Great (London 1973) 308, 314–16Google Scholar [hereafter Lane Fox]; Engels, D. W., Alexander the Great and the Logistics of the Macedonian Army (Berkeley 1978 104–6Google Scholar [hereafter Engels]. The fullest modern account remains that of Droysen, J. G., Geschichte des Hellenismus i² (Gotha 1877) 2.68–70, 73–80,Google Scholar which is an extended paraphrase of Arrian; there is an interesting insert from Curtius Rufus (75) which details the siege of the rock of Sisimithres, but it does not prevent him from retailing Arrian's version a few pages later as a separate event.

² Arr. iv 7.4–14.4. For the topos see most succinctly Livyix 18.4: referre in tanto rege piget superbam mutationem vestis et desideratas humi iacentium adulationes … et foeda supplicia et inter vinum et epulas caedes amicorum et vanitatem ementiendae stirpis. The same conjunction of the episodes of Cleitus and proskynesis also occurs at great length in Plutarch (Al. 50–6), and it was probably unavoidable for anyone steeped in the rhetorical tradition of the early empire.

³ Cf. Arr. iv 9.1, 12.6–7.

⁴ The most authoritative work for the identification of Alexander's routes has been the survey by von Schwarz, Franz, Alexanders des Grossen Feldzüge in Turkestan ² (Stuttgart 1906)Google Scholar [hereafter von Schwarz]. This work has the merit of detailed acquaintance with the terrain, but it was written long before the advent of systematic archaeology and relied on impressionistic identifications based on the author's intimate knowledge of the country (cf. Engels 99 n.2). In the case of 328, von Schwarz begins on totally the wrong footing, since he denies the traditional equation of Zariaspa and Bactra (cf. P. A. Brunt, Arrian i [Loeb 1976] 503) and makes Alexander's campaign begin at Chardzou on the Oxus, far to the north-west of Bactra (von Schwarz 65 f). The consequence is that he depicts Alexander merely retracing his steps back to Maracanda, and the whole of the work of pacification is deferred to the spring of 327.

⁵ Arr. iv 6.5; Curt, vii 10.1–3. It is clear that the army covered the whole length of the Zeravshan as far as its disappearance into the desert sands some 45 km from the Oxus. Alexander certainly went from there to winter quarters at Bactra/Zariaspa (Arr. iv 7.1; Curt, vii 10.10—for the equation, see below p. 34), but it is not stated whether he retraced his steps to Maracanda or cut across the desert and took the more direct route along the Oxus.

⁶ Arr. iv 7.2; Curt, vii 10.10–12. Curtius adds troop numbers and lists fewer commanders, but the measure of agreement with Arrian is impressive (see further, Bosworth, , CQ xxiv [1974] 61)Google Scholar. There is one slip, which may be a textual corruption, in that Asander is made to come from Lycia (Lytia P) not Lydia (cf. Berve, H., Das Alexanderreich [hereafter Berve] ii [Munich 1926] 87Google Scholar no.165), and one difference of nomenclature. Μελαμνίδας (Berve no. 493) in Arrian corresponds to Maenidas in Curtius: this divergence in nomenclature goes back to the original sources and, pace Hamilton, J. R., CQ v (1955) 217,CrossRefGoogle Scholar it should not be emended out of the texts, Melamnidas may be a mistake, but, if so, he is a mistake of Arrian's source, not his copyists.

⁷ Arr. iv 18.1–3; Curt, viii 3.17.

⁸ Droysen i² 2.77 and Schachermeyr 349 n. 416 have argued for a separate expedition; Berve (nos 814, 719), following W. Geiger, Alexanders des Grossen Feldzüge in Sogdiana (Progr. Neustadt 1882/3—non vidi) 32 f., insists on a doublet in Arrian.

⁹ See the brief notes by Jullien, P., Zur Verwaltung der Satrapien unter Alexander dem Grossen (Diss. Leipzig 1914) 37Google Scholar f; Berve i 265 f; Schachermeyr 312 ff. (the best short narrative).

¹⁰ Arr. iii 25.3; Curt, vi 6.13; Metz Epitome [hereafter ME] 3. For the wide-ranging impact of Bessus' challenge, see Bosworth, , JHS c (1980) 6Google Scholar f.

¹¹ Arr. iii 25.4; Curt, vi 6.20. On the route see Engels 86 f.

¹² Arr. iii 25.5; Curt, vi 6.21; Diod. xvii 78.1. According to Arrian (iii 21.10) Satibarzanes had participated in the murder of Darius, a statement that Badian has queried (CQ viii [1958] 147Google Scholar n. 1) because of the stark contrast of his treatment with that meted out to Bessus and Barsaentes.

¹³ Diod. xvii 78.1–4; Curt, vi 6.20–34; Arr. iii 25.5–7. Pace Berve ii no. 697 the vulgate tradition is far fuller and more detailed than Arrian and must form the basis of any reconstruction (cf. Schachermeyr 313; Engels 87–91).

¹⁴ Arr. iii 25.7; cf. Curt, vii 3.1. Justin xii 4.12 is hopelessly confused.

¹⁵ Curt, vii 3.2 (daved to Alexander's fifth day in Ariaspian territory—C.January 329); Diod. xvii 81.3; Arr. iii 28.2 (imprecisely dated to spring 329).

¹⁶ Curt, vii 4.38: et barbari duce amisso … arma Erigyio tradunt; Diod. xvii 83.6; Arr. iii 28.3 speaks vaguely of a general flight.

¹⁷ τῶν βαρβάρων ἰσόμαχον ποιούντων τὸν κίνδυνον (Diod. xvii 83.5; cf. Curt, vii 4.33; Arr. iii 28.3).

¹⁸ eodem tempore quae in gente Ariorum … gesserant perferuntur (Curt, vii 4.32). Diodorus reports the episode at exactly the same point.

¹⁹ Arr. iii 29.1; Curt, vii 5.1 (Artabazus installed at Bactra); Curt, vii 7.9, 21–4 (Erigyius at the Iaxartes; he dies in camp before the onset of winter 328/7 ([Curt, viii 2.40]); Arr. iv 3.7, 5.7, 6.1 (Caranus at Maracanda).

²⁰ Arr. iii 29.5. Stasanor may have participated in the earlier expedition against Satibarzanes, for Diodorus xvii 81.3 makes him joint commander along with Erigyius. Curtius, however, makes Caranus joint commander (vii 3.2, 4.32) and neither he nor Arrian knows anything of Stasanor in this context. Berve (no. 719) assumed that Diodorus was simply wrong and retrojected his later satrapal appointment; but it is hard to explain how the mistake arose.

²¹ Berve i 266 ‘beide empören sich’. He is later rather more guarded; ii no. 146 ‘Abfallsgelüste zeigte’.

²² Hdt. i 127.3; v 78; vi 15.1; viii 22, 69.2, 85.1; ix 67. Arrian elsewhere uses the expression twice—at iv 18.3 (an exact parallel) and Tact. 12.11 (where the meaning is unambiguous). For his linguistic dependence on Herodotus see Grundmann, H. R., Quid in elocutione Arriani Herodoto debeatur, Berliner Studien ii (1885)Google Scholar and, in brief, Bosworth, , CQ xxiv (1974) 56Google Scholar.

²³ Phrataphernes (Berve no. 814) had been satrap of Hyrcania and Parthyaea under Darius (Arr. iii 8.4, 23.4) and had surrendered to Alexander in his Elburz campaign in 330. Amminapes (Berve no. 55) had originally been established as satrap of Parthyaea/Hyrcania earlier in summer 330 (Arr. iii 22.1; Curt, vi 4.23f). Nothing more is heard of Amminapes and he was replaced by Phrataphernes by the beginning of 329 (Arr. iii 28.2).

²⁴ Arr. iv 7.1, mentioning not only Brazanes but other rebels who had sided with Bessus. There were obviously several centres of insurrection in Parthyaea.

²⁵ For the Median garrison early in 330 see Arr. iii 19.7. Of these forces the cavalry component, mercenaries and Thessalian volunteers, reached Alexander in autumn 330 (Arr. iii 25.4); the infantry (6,000 phalangites and 5,000 mercenaries) arrived while Alexander was in Arachosia (Curt, vii 3.4), about the time of Satibarzanes' second Areian invasion. Of the original forces only Thracians remained. They are not attested gain in Alexander's army, and, since two of the Median generals, Agathon and Sitalces (Berve nos 8, 712) are elsewhere described commanding Thracian detachments (Arr. iii 12.4), it is a fair assumption that they were left in command of their original forces, now the garrison of Media.

²⁶ Cf. Curt, vii 2.28–32 for near-mutiny in Media after the murder. For the more widespread discontent in the army, see Diod. xvii 80.4; Curt, vii 2.35–8; Just, xii 5.5–8.

²⁷ Cf. Arr. v 20.7, Phrataphernes brings to India (summer 326) the Thracians left in his command.

²⁸ Arr. iv 18.3; Curt, viii 3.17.

²⁹ Arr. iv 1.4–5; Curt, vii 6.13–15. The communities near the Iaxartes began the revolt by massacring their garrisons and the southern areas of Sogdiana and Bactria followed suit.

³⁰ Arr. iv 7.2 (on which see CQ xxiv [1974] 60Google Scholar f). See also Arr. iii 16.9, where the name of the previous Persian garrison commander is substituted for his Macedonian successor (CQ xxvi [1976] 121Google Scholar f). As for Curtius, the most likely explanation of his error in making Asander governor of Lycia (above n. 6) is that his source mentioned Nearchus in the same context, who actually was satrap of Lycia and Pamphylia (Arr. iii 6.6, 7.1) and brought troops to Bactra.

³¹ Arsami, Drangarum praefecto, substitutes est Stasanor (Curt, viii 3.17).‘Arsames’ has had a chequered history. Blancardus (1668) substituted his name for that of Arsaces as satrap of Areia in all three passages of Arrian, assuming his identity with the son of Artabazus (Berve no. 148), and Arsames appeared in that role in various standard compilations (e.g. Jullien [n. 9] 38). Roos then correctly restored the manuscript reading in all passages of Arrian, and Arsames disappeared from history as satrap of Areia. No one at any period recognised that an Arsames could have been satrap of Drangiana; 'das wurde erst Stasanor' says Berve (ii 81 n. 1), establishing a dogma with no basis in the sources.

³² Arr. iii 8.4, 21.1. The satrap, Barsaentes, was one of the principal regicides, and after Darius' death he had withdrawn to his satrapy (Arr. iii 25.8; Curt, vi 6.36), from which he withdrew to India in the face of Alexander's advance (cf. Berve no. 205).

³³ Menon's satrapy is firmly attested as Arachosia alone (Arr. iii 28.1; Curt, vii 3.5, ix 10.20). Gedrosia was settled at the same time and placed under Tiridates (Diod. xvii 81.2; cf. Berve no. 755). Now Arrian speaks in passing of the settlement of Drangiana and Gedrosia, and implies that they were dealt with individually (iii 28.1). Drangiana should have had a satrap like the Gedrosians.

³⁴ For the etymology see M. Mayrhofer, Zur Namengut des Avesta, SÖAW, Wien cccviii.5 (1977) 17, 43Google Scholar. Two other Persians named Arsames (Berve nos 148 f.) are attested in the Alexander period alone. As for Arsaces, it is well known as the pre-regnal name of Artaxerxes II (Ctesias, FGrH 688 F 15 [55 f.]; Deinon, FGrH 690 F 14). See also Thue. viii 108.4 for the hyparch of Tissaphernes. There is a comparable coincidence in the Triparadeisus settlement when two Cypriots, Stasander and Stasanor, were given neighbouring satrapies (Diod. xviii 39.6; Arr. Succ.fr. 1.36).

³⁵ Arr. iv 17.3; Curt, viii 2.14 (cf. viii 1.19).

³⁶ So Arr. iv 18.3. For his later position as satrap of both Areia and Drangiana see Arr. vi 27.3; Diod. xviii 3.3; Dexippus, FGrH 100 F 8.6; Justin xiii 4.22.

³⁷ Cf. iii 23.7 (occurring twice); iii 24.3.

³⁸ Curt, iv 12.9, vi 4.25, 5.21, viii 3.17, x 1.39 (no textual variants attested).

³⁹ τῶν βαρβάρων ἰσόμαχον ποιούντων τὸν κίνδυνον(Arr. iii 8.4, 11.4); Τάπουροι (iii 23.1—2, 6–7, 24.3, vii 23.1); Ζαράγγαι (iv 18.3).

⁴⁰ For Ζαράγγαι see iii 25.8 (twice), vi 17.3, 27.3, vii 6.3; for Δράγγαι iii 21.1, 28.1, iv 18.3, vi 15.5, vii 10.6. Of these passages vi 15.5 is a doublet of vi 17.3 and iii 28.1 is a resumptive passage, referring back to iii 25.8. See further, Bosworth, CQ xxvi (1976) 128 f.

⁴¹ For instances see Phoenix xxix (1975) 31Google Scholar with n. 24. A good instance where a general's report of recent successes is given in extenso on his arrival in camp is Arr.ii 2.3–7 (Hegelochus in Egypt); Curtius iv 5.13–22 records the same events but places them before the siege of Gaza, rather nearer the time they occurred.

⁴² Arr. iv 15.7–8; Curt, vii 10.13–14. See also Plut. Al. 57.5 (with Athen. ii 42f); Strabo xi 11.5 (518). The variant traditions of this event make an interesting study in their own right.

⁴³ ME ₁₄: deinde post diem undecimum ad flumen Ochum pervenit, id transit, inde ad Oxumflmnen devenit.

⁴⁴ Pliny NH vi 47. The information is unique to Pliny.

⁴⁵ Engels 104 f. His restoration involves Alexander crossing the Oxus and campaigning in Sogdiana before returning to Bactra for the journey to Sogdiana; and he is forced to identify the R. Ochus with the Kashka Darya in southern Sogdiana, which, as we shall sec, is an impossibility.

⁴⁶ Cf. Brunt (n. 4) 506. Von Schwarz, who knew the terrain well, dismissed the idea of a desert march as an impossibility, even though he argued for Alexander wintering at Chardzou, far closer to Merv than Balkh (68 ff.).

⁴⁷ Cf. Arr. iv 5.2–6.2; Curt, vii 7.31–9; ME 13.

⁴⁸ Kaerst, J., Geschichte des Hellenismus i³ (Leipzig 1927) 439Google Scholar n. 3; von Schwarz and Brunt (n. 4); V. Tscherikower, Die hellenistische Städtegründungen, Philologus Suppl. xix.i (1927) 105Google Scholar.

⁴⁹ Meyer, E., Blute und Niedergang des Hellenismus in Asien (Berlin 1925) 17Google Scholar f; Berve i 294; Tarn ii 234 f.

⁵⁰ Schachermeyr 349 n. 416, followed by Hamilton, J. R., Alexander the Great (London 1973) 100Google Scholar and Lane Fox 308.

⁵¹ Arr. iv 18.1. Schachermeyr suggests that the expedition to Margiana was the mysterious mission which occupied Phrataphernes and Stasanor during 328 (see above, p. 19); they were given their instructions at Bactra and led troops north from their satrapies. One wonders how the complicated logistics of this three-pronged campaign were arranged.

⁵² Arr. iv 17.1: οὶ δὲ ὼς έπύθοντο πλήσιον ἐπελαύνοντά σφισι Κρατερόν, ἔφευγον . . . ὡς εἰς τὴν ἐρήμην. καὶ Κρατερὸς ἐχόμενος αὐτῶν αὐτοῖς τε ἐκείνοις περιπίπτει ού πόρρω τῆς έρήμης . . . Cf. also Curt. viii1.6.

⁵³ Isidorus, FGrH 781 F 2 (14); Strabo xi 10.2 (516). Very much later Ptolemy vi 10.4 and Ammianus Marcellinus xxiii 6.54 mention a few other settlements, but they are most obscure, and it is doubtful whether they were all located in the oasis itself.

⁵⁴ There is a possibility that Alexandria Eschate in Sogdiana was destroyed in a Saca attack at roughly the same time; cf. Tarn, . JHS lx (1940) 90–4;Google ScholarWolski, J., Klio xxxviii (1960) 113–15Google Scholar.

⁵⁵ Pliny, NH vi 46 (followed by Ammianus xxii 6.54), talks of Margiana being surrounded by mountains with a circuit of 1,500 stades. Strabo xi 10.2 rightly describes the oasis as surrounded by deserts only. I suspect that Pliny has combined a description of the circumference of the oasis with a reference to the Kopet Dag massif to the south, so creating a wholly fictitious girdle of mountains.

⁵⁶ Diod. index xvii κδ ὡς Βακτριανοὺς ἐκόλασε καὶ Σογδιανοὺς τὸ δεύτερον ἐχειρώσατο καὶ πόλεις ἔκτισεν εὐκαίρως πρὸς τὰς τῶν ἀφισταμένων κολάσεις

⁵⁷ Diod. index xvii κε:ἅλωσις τῶν εἰς τὴν Πέτραν καταφυγόντων Cf. Curt, vii 11.1.

⁵⁸ Justin xii 5.13. The city foundations are the only aspect of the military history of Alexander in Sogdiana that Justin cares to stress. His information on Alexandria Eschate in the previous sentence, however, is reliable.

⁵⁹ Arr. iv 16.3. For the use of συνοικίζειν in the sense of adding settlers to newly founded cities, compare vi 17.4, where settlers are to be provided for newly fortified cities (cf. vi 17.1).

⁶⁰ Curt, viii 1.1 (Hephaestion and Coenus); cf. vii 11.29, multitudo deditorum incolis novarum urbium cum pecunia capta dono data est. In the summer of 329 Alexander had first enslaved the rebels of Cyropolis but then liberated them to become incolae of the new foundation (Just, xii 5.12; Curt, vii 6.27; Arrian iv 4.1 speaks only of barbarian ‘volunteers’). For the procedure see Briant, P., Klio lx (1978) 74–7Google Scholar.

⁶¹ Kiessling, RE ix (1914) 470 f., 483, 492 f; Sturm, RE xvii (1937) 1768–70; Tarn ii 8 n. 1, 310 n. 4; The Greeks in Bactria and India 113 n. 4.

⁶² Strabo xi 11.5 (518); Arr. iv 6.6 ( = FGrH 139 F 28). See also Strabo xi 10.1 (515), differentiating the Areius from the Margus; Ptolemy Geog. vi 17.2; Amm. Marc. xxiii 6.69.

⁶³ Herrmann, A., Alte Geographie des unteren Oxusgebiets, Abh. kgl. Ges. Wiss. Göttingen NF xv.4 (1914) 30–5;Google Scholar so RE ii.A (1921) 29Google Scholar.

⁶⁴ All but two of the attested fragments of this author come from Strabo (cf. FGrH 779). See, in general, Altheim, F. & Stiehl, R., Geschichte Mittelasiens im Altertum (Berlin 1970) 359–79CrossRefGoogle Scholar.

⁶⁵ This river, which enters the Caspian through the desert country north of Hyrcania, fits well with passage (c). The Aparnians, the tribe of Arsaces the conqueror of Parthia, are explicitly located by the Caspian immediately north of Hyrcania: cf. Strabo xi 7.1 (508), xi 8.2 (511) (the readings vary between Πάρνοι and Απαρνοι but the same people are concerned in all cases). The objections of Altheim and Stiehl (see n. 64) 449 f. rest on the mistaken orthodoxy that the modern Tedzhen is the ancient Ochus.

⁶⁶ The excursus is garbled textually in Strabo, but the parallel passage of Arrian (iv 6.6) proves Aristobulus' authorship. Both authors begin with the Polytimetus (Zeravshan) and follow with the Areius (Hari Rud); they emphasise different aspects of the excursus and make different selections, but there is clearly a common source.

⁶⁷ Strabo is unique in placing it near the Ochus; Arrian (iv 15.7–8) places it by the Oxus, as does Plutarch (Al. 57.5). Curtius records only a spring of fresh water near the Oxus (vii 10.13 f).

⁶⁸ Pliny NH vi 49; cf. xxxi 75.

⁶⁹ Herrmann (n. 63) 31, following the identification K. J. Neumann. He refers to the river as the Sangalak, whereas I give the nomenclature of The Times Atlas.

⁷⁰ Ptolemy Geog. vi 11.2–4. He makes the Ochus form a confluence with the Dargamanes (so Amm. Marc, xxiii 6.57) and join the Oxus west of the Zariaspa and the Artamis.

⁷¹ Ptolemy vi 11.9 (Maracanda), 20.2 (Helmand). Cf. Thomson, J. A., History of Ancient Geography (Cambridge 1948) 294:Google Scholar ‘He draws the Oxus badly, and joins to it several rivers which were really lost in deserts then as now … some (towns) are false doublets like Zariaspa–Bactra and others like Samarcand are so grossly misplaced that the text seems hardly credible.’

⁷² Curt, vii 5.1–16, esp. 5.2: per CCCC stadia ne modicus quidem humor existit; Diod. xvii index ιθ. For the modern conditions see von Schwarz 30 ff.

⁷³ For full details, see L. W. Adamec, Historical and Political Gazetteer of Afghanistan i (Graz 1962) 169 ff. The only other choice is the Kokcha, the river next to the east; but the Kokcha has no tributary that can be identified as Ptolemy's Dargamanes (above n. 70).

⁷⁴ Bernard, P., Rev. Num. xvii (1975) 58–69,Google Scholar promising a fuller study in the future. For the Achaemenid evidence see 68 n. 19.

⁷⁵ Arr. iv 2.1–3.5; Curt, vii 6.16–24.

⁷⁶ Arr. iv 6.3–5 (ἐπῆλθε πᾶσαν τὴν χώραν ὅσην ὁ ποταμός . . . ἐπέρχεται); Curt, vii 9.21–10.9; Diod. xvii index κγ.

⁷⁷ Arr. iv 7.1; Curt, vii 10.10.

⁷⁸ Curt, vii 10.13; cf. Arr. iv 15.7; ME 14. The only hardship recorded (Curt, vii 10.14) was caused by the muddiness of the Oxus, a phenomenon well attested in other ages: cf. Polyb. x 48.4; de Clavijo, R. Gonzales, Embassy to the Court of Timour, ed. Markham, C. R., Hakluyt Soc. 1st ser. xxvi (1859) 118Google Scholar:21 Aug. 1404.

⁷⁹ Wood, J., A Journey to the Source of the River Oxus ² (London 1872, repr. Karachi 1976) 268Google Scholar. Wood was able to ford the Oxus with relative comfort at Jan-Kila, a little upstream from Ai-Khanum (260 f.).

⁸⁰ For bibliography see J. Seibert, Alexander der Grosse (Darmstadt 1972) 145, to which add the successive reports by P. Bernard in CRAI 1974–6 and, for the significant coin-hoard unearthed in 1973, Petitot-Bichler, C.-Y., Rev. Num. xvii (1975) 23CrossRefGoogle Scholar ff.

⁸¹ Cf. Frumkin, G., Archaeology in Soviet Central Asia Leiden/Köln 1970) 62Google Scholar f., 66–8.

⁸² Arr. iv 18.4 (ἅμα τῷ ἦρι ὑποφαίνοντι) 18.5 χιων πολλή) 19.1–2 (omnipresent snow).

⁸³ Arr. iv 22.3.

⁸⁴ Tarn i 72–6 (cf. 72 n. 1: ‘On this scheme it is impossible to get in all that happened at Bactra before he finally quitted it; he must have taken the two strong-holds by mid-winter’); Fraenkel, A., Die Quellen der Alexanderhistoriker (Breslau 1883) 186Google Scholar.

⁸⁵ Strabo xv 1.17 (691) = FGrH 139 F 35. The evening setting of the Pleiades must be at issue; the morning setting, in April, occurred while Alexander (on any chronology) was still north of the Hindu Kush.

⁸⁶ Anspach, A., De Alexandri Magni Expeditione Indica (Leipzig 1903) 8Google Scholar n. 18, suggested that Alexander needed to reconquer Parapamisadae and was reluctant to move before his envoys returned from India (Arr. iv 22.6). See also Brunt (n. 4) 507.

⁸⁷ For the interpretation of the passage see Appendix below.

⁸⁸ Curt, vii 11.1–29; ME 15 ff.; Diod. xvii index κε.

⁸⁹ Curt, viii 1.1–9; cf. Arr. iv 15.1–6.

⁹⁰ Curt, viii 1.11–19 (cf. 19:inde ad Maracanda reditum est); Diod. 17 index κς.

⁹¹ Curt, viii 2.13–19; ME 19; Diod. xvii index κθ.

⁹² Curt, viii 2.19–33; cf. ME 19.

⁹³ Curt, viii 2.33–3.16; ME 20–3.

⁹⁴ Curt, viii 4.1–20; ME 24-7; Diod. xvii index κθ.

⁹⁵ Curt, viii 4.21–30; ME 28–31; Diod. xvii index λ. Cohortandus (Curt, viii 4.21) is traditionally emended to Oxyartes, on the grounds that Rhoxane is immediately presented as his daughter (filia ipsius, viii 4.23). The Metz Epitome, however, explicitly mentions Chorienes as the giver of the feast and adds that he introduced his own daughters together with the daughters of his friends, including Rhoxane, Oxyartis filia. Oxyartes is then named as present as the feast (ME 29). It is evident that Curtius has erroneously conflated Rhoxane with the daughters of Chorienes, and that the corrupt name Cohortandus should be emended to the palaeographically similar Chorienes. This was immediately recognised in Wagner's, O. edition of the Metz Epitome, Jb.klass. Phil. Suppl. xxvi (1901)Google Scholar, and the emendation has been largely accepted in German scholarship (cf. Berve 355 n. 2; Schachermeyr 353 n. 423). Tarn, however, refused to guess ‘what weird error in transmission lies behind “Cohortandus” ’ (ii 103; cf. 341 n. 5).

⁹⁶ Curt, viii 5.2f; cf. Arr. iv 22. if.

⁹⁷ Strabo xi 11.4 (517). The note is preceded by details on cities founded and destroyed by Alexander, including details not found elsewhere in the tradition (e.g. Callisthenes arrested at Caryatae in Bactria) and it is continued by a report of the massacre of the Branchidae. The source most recently quoted is Onesicritus (xi 11.3 (517] = FGrH 134 F 5 ).

⁹⁸ Arrian (iv 19.5 ff.) states that Rhoxane was captured on the Sogdian rock and that Alexander fell in love at first sight. He goes on to report the wedding, but gives no indication how long after the capture it took place. Arrian also claims that Oxyartes surrendered to Alexander at the news of the favourable reception of his daughter (iv 20.4). Now all traditions mention Oxyartes’ presence at the second great siege (Arr. iv 21.6 f.; Curt, viii 2.25 ff.; Plut. Al. 58.3), and it is reasonable to assume that his daughter had already come into Alexander's power, even if she were not yet married to him. It is a possibility at least that Rhoxane was captured at the first rock (Arrian) and married at a subsequent banquet (Curt, viii 4.23; ME 28; Plut. Al.47.7). Strabo can be related to neither tradition; Rhoxane was neither captured nor married at the second rock: cf. Hamilton, J. R., Plutarch Alexander (Oxford 1969) 129Google Scholar.

⁹⁹ τήν τε ἐν τῇ Βακτριανῇ, τὴν Σισιμίθρου (Strabo): cf. Curt, viii 2.13–15; ME 19 (in Bactros).

¹⁰⁰ The Rock of Sisimithres is 15 stadia in height and 80 in circuit, that of Ariamazes is twice as high. Cf. Curt, vii 11.2: the Rock of Ariamazes is 30 stadia in height and 150 in circuit (cf. ME 15). Arrian makes the second rock 20 stades high and 60 in circuit (iv 21.2).

¹⁰¹ Plut. Al. 58.3: Sisimithres, according to Oxyartes, is the most cowardly of men (cf. Curt, viii.27–8, 30).

¹⁰² Polyaenus iv 3.29.

¹⁰³ Arrian makes a feature of the mountain forest in his account of the second siege (iv 21.3; cf. Curt, viii 2.24).

¹⁰⁴ οἱ δὲ Μακεδόνες . . . ἠλάλαξαν ὁ δὲ ᾿Αριο- μάζης ἐκπλαγείς κτλ (Cf. Curt, vii 11.25).

¹⁰⁵ Cf. i 17.5, 27.5 f., 29.1, ii 23.5, iii 30.10, iv 21.2, v 22.4.

¹⁰⁶ Von Schwarz 75–7, 85 f. In both cases he paraphrases references to caves in Curtius as mountain clefts open to the sky (cf. Engcls 106 n. 34).

¹⁰⁷ See the comparative analysis in Appendix 2.

¹⁰⁸ Arr. iv 21.4 f; cf. Curt, viii 2.23 f.

¹⁰⁹ Arr. iv 19.3 (πέμψας δὴ κήρυκα) may reflect Curt, vii 11.23 ff. (cf. Berve no. 459).

¹¹⁰ Von Schwarz 83 f.; Berve no. 708; Tarn ii 96; Hamilton (n. 98) 129.

¹¹¹ Curt, viii 4.21; ME 28 (see above, n. 95).

¹¹² For Alexander's policy of repression, see Arr. iv 1.4; Curt, vii 6.16, 9.22. Curtius' story of the death of Ariamazes is accepted by von Schwarz 78 n. 1; Berve no. 112; Briant, P., Klio lx (1978) 72Google Scholar f., 76.

¹¹³ ME 18: Ariomazen interfecerunt. deinde ipsi se dediderunt.

¹¹⁴ Arr. iv 8.1 (εἰ καὶ ὀλίγον ὕστερον ἐπράχθη) 14.4, 22.2.

¹¹⁵ Curt, viii 1.1: Hephaestionem uni, Coenon alteri duces dederat; viii 1.10: Hephaestionem et Artabazum opperiens stativa habuit.

¹¹⁶ Curt, viii 2.13: cum parte exercitus Hephaestionem in regionem Bactrianam misit commeatus in hiemem paraturum.

¹¹⁷ Curt, viii 1.19 (dated to the second stay at Maracanda after the hunt at Bazaira).

¹¹⁸ Arr. iv 17.3. The reference to the location could not be vaguer (αὐτοῦ!) but Arrian is resuming his main narrative after the interlude of Spitamenes' raid on Zariaspa, and the place previously mentioned in the narrative of Alexander's actions was Maracanda (iv 16.2). Curtius viii 2.14 mentions that Amyntas succeeded to his satrapy just before Alexander left Maracanda.

¹¹⁹ E.g. ii 5.7; iii 2.3 ff. (above n. 41); vi 29.3.

¹²⁰ Strabo xi 11.2 (516), 8.9 (514) with Pliny NH vi 45 ( = FGrH 119 F2). Contrast Arr. iii 30.5 and iv 7.1–3; Arr. iv 7.1 and Curt, vii 10.10.

¹²¹ Arr. iii 30.5 = FGrH 138 F 14; cf. iii 25.3 f., iv 22.1, 3.

¹²² Cf. Arr. iv 1.5, 16.5–6. At iv 7.1 the reference immediately follows a demonstrable extract from Aristobulus (FGrH 139 F 28b); it appears that Arrian digested his report on the Zeravshan campaign from Aristobulus and after taking his account to the winter's pause at Bactra/Zariaspa he turned to Ptolemy for the reports of reinforcements (above p. 23).

¹²³ Arr. iv 16.2–3. Ptolemy may have given an account of his own adventures which Arrian ignored, as he sometimes does. He says nothing, for instance, about Ptolemy's mission in India at the time of the siege of the Malli town, even though it involved several engagements with the enemy (cf. Arr. vi 11.8, ἅλλας μάχεσθαι μάχας καὶ πρὸς ἅλλοις βαρβάρους Curt, ix 5.21).

¹²⁴ Arr. iv 8.9 (FGrH 139 F 29); Curt, viii 1.45–8. The details vary, but both traditions confirm Ptolemy's presence at Maracanda.

¹²⁵ Arr. iv 19.1, κατὰ τὸ ἀποτομώτατον; Curt, vi 11.14, qua minime asper ac praeruptus aditus videbatur.

¹²⁶ Arr. iv 19.2, ὥστε οὐδὲ τὰ σώματα . . . εὑρέθη; cf. ii 11.8 (FGrH 138 F 6) on which see Entr. sur l'ant.class, xxii (Fondation Hardt 1976) 27Google Scholar.

¹²⁷ The peaks in the Hissar range retain their snow cover until summer, and snow has been known as late as July (cf. Engels 107).

¹²⁸ Cf. Arr. iv 21.10 (rock of ‘Chorienes’); the snow at Aornus is not attested in the campaign narrative, but it is amply attested by Aristobulus (cf. Appendix 1).

¹²⁹ The transition is at iv 18.4. On my analysis the reference to spring is the last detail from Aristobulus, after which Arrian turns to Ptolemy.

¹³⁰ Curt, viii 4.1–20; ME 24–7; Diod. xvii index κθ.

¹³¹ Strasburger, H., Hermes lxxx (1952) 470–3Google Scholar. As a result, recent histories have tended to accept the story of the snowstorm (cf. Schachcrmeyr 353; Lane Fox 314), but there has been no attempt to reconcile it with the general chronology of Arrian.

¹³² Arr. iv 21.10 refers to hardships in the siege of the second rock citadel, which was relieved by provisions supplied by ‘Chorienes’. It is possible that this is a confused reference to the provisions which the vulgate radition records were sent by Sisimithres after the snowstorm (Curt, viii 4.19; so Strasburger 472). More probably the Macedonians were twice relieved by the Sogdian dynast.

¹³³ Curt, viii 2.32 f.; ME 19; Arr. iv 21.9 (Sisimithres); Curt, viii 4.21; ME 28 (Chorienes).

¹³⁴ See the lengthy but inconclusive article by Treidler, H., RE ix.A (1967) 1480–4Google Scholar.

¹³⁵ Arr. iv 18.1; Curt, viii 2.19; ME 19. Von Schwarz 83 (so Brunt [n. 4] 507) considered Curtius' Nauta to be a corruption of an otherwise unknown place name, but Nauta is unquestionably the same as Nautace of the Metz Epitome and Ναύτακας in the Diodorus index κθ. What is more, Alexander's march route south from Maracanda took him inevitably to the vicinity of Nautaca proper.

¹³⁶ Arr. iii 28.9; cf. von Schwarz 74 f.; Sturm, , RE xvi (1935) 2033Google Scholar.