Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

This webpage reproduces an article in the
American Journal of Philology
Vol. 51, No. 1 (1930), pp22‑31.

The text is in the public domain.

This page has been carefully proofread
and I believe it to be free of errors.
If you find a mistake though,
please let me know!

 p22  When Did Alexander Reach the Hindu Kush?

[In order to avoid, among other things, difficulties in Arrian and Strabo, Plutarch's statement of a four months' halt by Alexander in Persis should be cut to seven weeks, so that Alexander may spend the winter of 330‑29 at the Hindu Kush, and the next two winters at Zariaspa and Nautaca respectively.]

There are many problems, of course, connected with the campaigns of Alexander in Asia, but to a large degree they concern solely the sources. One might expect several perplexing chronological difficulties in an expedition so long and so varied, but happily this is not the case.

The only serious problem of chronology from the crossing of the Hellespont to Alexander's death at Babylon in 323 concerns the winters 330‑29 and 329‑28. It seems to be clear from Arrian​1 and Strabo​2 that the winter of 330‑329 was spent on the south side of the Hindu Kush (i.e. before the first crossing). But the meaning of the Greek, we are told, is not absolutely certain and can therefore be challenged; furthermore, if the statements of Arrian and Strabo are accepted, the almost insuperable difficulty at once arises of explaining how Alexander traversed the 1300 miles from the Caspian to the Hindu Kush in the few months before winter set in. It is unfortunate that the language Arrian uses​3 to describe the winter quarters of the following year (329‑28) is not more explicit. The apparent meaning is that Alexander took winter quarters at Zariaspa, yet it has been maintained that Arrian is referring​4 to the first part of a winter which was spent chiefly at Nautaca (328‑27); in which case the previous winter (and not that of 330‑29) must have been spent at the south foot of the Hindu Kush and it then becomes necessary to invent some place for the winter quarters of 330‑29. That briefly is the problem, but before we proceed to examine it more closely, it will be well to put down a few of the almost certain dates in Alexander's life:

Born 356
Succeeds Philip 336
Victorious at Gaugamela 331
Takes up winter quarters in Persis 331‑30
p23 Takes up winter quarters at Nautaca 328‑27
Recrosses the Hindu Kush 327
Crosses the Indus 326
Reaches the Indian Ocean 325
Returns to Ecbatana 324
Dies at Babylon 323

It will be seen from the table that there is no difficulty up to Alexander's arrival in Persis and none after the recrossing of the Hindu Kush; the problem is to account for the intervening time. The evidence for this period is as follows: Alexander, setting out from Persepolis some time in 330, resumed the pursuit of Darius, whose murdered body he found near Shahrud. Then, after a delay in the region of the Caspian, he turned south into Seistan and, according to Arrian,​5 marched in deep snow through the land of the Arachotai. This brings us fairly to the Hindu Kush. Arrian says nothing of winter quarters (although he gives the season); but we learn from him that Alexander stopped long enough at the foot of the Hindu Kush to found a city. Strabo says​6 that Alexander passed through the land of the Paropamisadae after the setting of the Pleiades, established winter quarters below the Hindu Kush, where he built a city, and thence crossed the range into Bactria in fifteen days. We learn from Arrian​7 that Alexander crossed before the snow was yet out of the passes.

The Hindu Kush crossed, Alexander went to Bactra, crossed the Oxus in pursuit of Bessus, and halted at Maracanda. Thence he went to the Jaxartes, where he carried on a vigorous campaign in the surrounding neighborhood. He then proceeded to Maracanda and Zariaspa. Arrian says that Alexander remained at Zariaspa "until the depth of the winter passed."​8 He does not mention the coming of spring, but says that Alexander next recrossed the Oxus, swept the country as far as Maracanda and carried on a campaign for some time in Sogdiana, chiefly against Spitamenes. He then had his army rest at Nautaca "for what was about the depth of the winter."​9 On the approach of spring Alexander left Nautaca and resumed operations against the "rocks," went to Bactra and in the early summer of 327 recrossed the Hindu Kush.

From this evidence it would seem that Alexander passed  p24 through the land of the Paropamisadae (roughly, the Cabul valley) in November,​10 330, spent the winter of 330‑29 at the south foot of the Hindu Kush, the winter of 329‑28 at Zariaspa, and that of 328‑27 at Nautaca; but, as I have already said, the objection may be raised that Alexander, leaving the Caspian in October, 330, as is generally assumed, could not possibly have covered the 1300 miles to the Hindu Kush, including several stops on the way, by that winter, and therefore the disposition of the subsequent events is also in doubt. Two notable attempts have been made to get around these difficulties.

Hogarth​11 first focused attention on the problem by pointing out that it was physically impossible for Alexander, starting from the Caspian in October, to reach the Hindu Kush by the winter of 330‑29. Hogarth therefore places Alexander in Seistan for the winter of 330‑29, in Cabul (the land of the Paropamisadae) in November, 329, and at the foot of the Hindu Kush that winter. This leaves the winter of 328‑27 to be accounted for, with both Zariaspa and Nautaca mentioned by Arrian as winter quarters. Hogarth maintains that but one winter is referred to, and therefore divides this winter between the two towns, further supporting his argument by the statement that, since there would be so little for Alexander to do in the summer of 328, two separate winters cannot be meant.

The latest statement of the case is by Tarn.​12 While his account of Alexander is, on the whole, excellent, on this particular point Tarn neither states the difficulties nor offers a satisfactory solution. He says that Alexander apparently never took winter quarters at all in 330‑29, but in the spring of 329 Alexander had reached the Cabul valley. This meets Hogarth's objection of distance, for Tarn, too, assumes that Darius died in midsummer, 330.​13 Accepting the obvious meaning of Arrian, Tarn assigns the winters of 329‑28 and 328‑27 to Zariaspa and Nautaca respectively.

The arguments against Hogarth's theory are these: "There is not a shred of evidence in Arrian or elsewhere that Alexander  p25 spent a winter in Seistan. Furthermore, the distance from Seistan to the Hindu Kush, where Alexander next had his winter quarters according to Hogarth, is by no means great enough to require a year to traverse. Then, too, there is more than one difficulty in assuming that only one winter was spent at Zariaspa-Nautaca. Would Alexander undertake operations against so formidable an opponent as Spitamenes in the heart of winter, and, if he did, would he then change his winter quarters? The question, however, is easily decided by the Greek, for the meaning is clear. After campaigning in Bactria and Sogdiana, Alexander went to Zariaspa, where he remained "until the depth of the winter passed."​14 Then followed an expedition against Spitamenes, after which Alexander had his army rest at Nautaca "for what was about the depth of the winter."​15 Certainly two distinct winters are here referred to. Hogarth, then, in not reading Arrian correctly, has failed to offer a satisfactory solution of the problem.

Hogarth's argument that there was not much for Alexander to do in 328, if we allow a winter each at Zariaspa and Nautaca, is not a strong one. Arrian's account, as a matter of fact, is quite long enough, for we must bear in mind that Alexander was dealing with a national revolt of serious proportions in eastern Iran, not so easily crushed as the sporadic outbreaks elsewhere in Asia.​16 Indeed, Alexander had this summer to contend with the best opponent he had yet met. Leaving Zariaspa, he crossed the Oxus and proceeded to Maracanda. Here he divided his forces, and, by building fortified posts, tried to subdue Spitamenes not only by direct encounter, but by leaving him no place in Sogdiana where he might stay. Hogarth, however, has made an important contribution to the subject by pointing out that it is at least 1300 miles from Zadracarta, on the Caspian, to the Hindu Kush, and that Alexander, leaving Zadracarta in October,  p26 could not possibly have reached the Hindu Kush, as had been previously believed, by December.

Tarn's theory is harder to attack, since it is not so detailed as Hogarth's. Tarn does not represent Alexander as being in any particular place in the winter of 330‑29, but he does say that in the spring of 329 Alexander was already in the Cabul valley. In this case, he goes against the evidence of Arrian that Alexander traversed all this district in snow, and against the evidence of Strabo that Alexander passed through the land of the Paropamisadae after the setting of the Pleiades (November) and established winter quarters at the foot of the Hindu Kush. It may be objected that the snow in Arrian was due to a late winter, but why does not Arrian mention this and why does he omit all reference to the real snows of winter? Arrian, to be sure, does not mention winter quarters, but the halt at the foot of the Hindu Kush to found a city probably represents accurately enough the main stop of the winter. This is the obvious meaning of Arrian and it is substantiated by the important passage in Strabo. Strabo specifically says that Alexander passed through the land of the Paropamisadae (the Cabul valley) after the setting of the Pleiades (ὑπὸ πλειάδος δύσιν). The Pleiades, of course, set every day, but it was the custom of the ancients to refer only to the morning and evening setting of stars. In antiquity the apparent evening setting of the Pleiades occurred in April, the morning one about the middle of November. Strabo does not state which setting he has in mind, but it was the general practice of ancient writers (particularly the later ones) to mean the morning setting, and this is obviously Strabo's meaning, for a few lines later he speaks of winter quarters at the Hindu Kush and of Alexander founding a city there. It is manifest that Tarn, in bringing Alexander here some months later, after the conclusion of winter, has failed to solve our difficulty.

The solution I offer is this: I agree with Hogarth that the obvious meaning of Arrian and Strabo is that Alexander was in the Cabul valley in November and took up winter quarters at the south foot of the Hindu Kush; and with Tarn, that Arrian's later account simply means that Alexander spent one winter at Zariaspa and the next at Nautaca. In other words, I do not agree with Hogarth that Alexander spent a winter in Seistan,  p27 before the one at the Hindu Kush, and divided a winter between Zariaspa and Nautaca; nor with Tarn, that Alexander did not reach the Hindu Kush until spring of 329. But in this latter case I lay myself open to the very serious objection raised by Hogarth that Alexander leaving Zadracarta in October could hardly reach the Hindu Kush that winter. The question then is, Did Alexander leave Zadracarta in October? This view is generally accepted, for chronologists have agreed that Darius died in midsummer, and several weeks must elapse between his death and Alexander's departure from Zadracarta. But acceptance of this will forever leave the winters of 330‑29 and 329‑28 in doubt. The evidence that Alexander spent the winter of 330‑29 at the Hindu Kush and that of 329‑28 at Zariaspa rests on such good authority that we must make sure that the evidence for Darius' death in midsummer of 330 is unimpeachable.

Hogarth gives a convenient table.​17 After the battle of Gaugamela, dated to October 1st, 331, by a lunar eclipse eleven days previously, we must allow for —

March to Babylon at least 40 days 
Halt in Babylon 34  "  

(Curt., V.1.39;

 Just., XI.14)

March to Susa 20  "  

(Arr., III.16.7)

Stay in Susa x  "  
March to Persepolis 30  "  
Stay in Persis 120  "  

(Plut., Alex. 37.3)

March to Ecbatana 12 + x  "  

(Arr., III.19.3)

Stay in Ecbatana x  "  
March to Rhagae 11  "  

(Arr., III.20.2)

Stay in Rhagae 5  "  

(Arr., III.20.3)

Last stages of the pursuit 5  "  

(Arr., III.21)

    Total 277 + x days.

"The death of Darius, therefore," concludes Hogarth, "took place near Shahrud, about the three hundredth day after Arbela, i.e. at the very end of July or beginning of August, 330. This, as it happens, coincides, according to received computation, with Arrian's statement (III.22) that the month of the murder was the Attic Hecatombaeon." But, two pages previously, Hogarth warns against accepting the month-dates of Arrian, and I should urge a like reserve here.

 p28  Certainly the marching time allowed in the above table must be accepted. In fact, the only point that can possibly be questioned is Plutarch's statement that Alexander remained in Persis four months. We have no way of knowing on what authority Plutarch based this statement; we only know that everything else he says about Alexander's stay in Persis, as well as his approach to the country, is worthless anecdote; whereas we do know that Arrian generally drew on very good material, usually Ptolemy and Aristobulus. The whole argument comes down, then, to whether we ought to believe Arrian (incidentally, supported by Strabo) or Plutarch. We must choose between them, for they cannot be reconciled. If we accept the four months of Plutarch, then Darius' death must have occurred in midsummer, 330, in which case we are unable to accept Arrian's statements that Alexander reached the Hindu Kush that winter and spent a winter each at Zariaspa and Nautaca. We are forced either, on the one hand, to interpose a winter between Zadracarta and the Hindu Kush, and to allow but one winter for Zariaspa-Nautaca; or, on the other hand, if we give a winter each to Zariaspa and Nautaca, we must postpone Alexander's arrival at the Hindu Kush until spring of 329. But, if we assume that Plutarch exaggerated the length of Alexander's stay in Persis, then Darius' death will fall early enough in 330 to allow Alexander time to reach the Hindu Kush that winter and to spend the next two winters at Zariaspa and Nautaca respectively.

According to Hogarth's figures, Alexander reached Persepolis 124 + x days after Gaugamela, or about February 5th. Does it seem reasonable that Alexander would delay in Persis four months and not set out against Darius until about June 1st? This hardly seems consistent with Alexander, the man of action. Of course Alexander remained some weeks in Persis, for he had reached the heart of the Persian empire and undoubtedly wished to take stock, and perhaps to undertake a short winter campaign against the neighboring tribes.​18 Let us, then, shorten his stay in Persis from four months to seven weeks, and see what the result will be.

The march from Persepolis to the place of Darius' death,  p29 according to the table, occupied 33 + x days, or, to make a generous allowance, about 40 days. Alexander, then, leaving Persepolis about March 25th, found Darius' body about May 3rd. He then crossed the Elburz mountains to Zadracarta in Hyrcania, overran Tapuria, reduced the Mardi and returned to Zadracarta, where he remained fifteen days. In view of the distance to be covered and the halt at Zadracarta we should allow six weeks to elapse between the finding of Darius' body and the departure from Zadracarta. This brings us to June 13th.

Hogarth gives the stages by the great caravan route from Zadracarta (near Sari, on the Caspian littoral) to the Hindu Kush in round numbers as follows:19

Sari to Shahrud at least 100 miles
Shahrud to Meshed over 300   "  
Meshed to Herat about 220   "  
Herat to Candahar by the great road about 330   "  
Candahar to the Hindu Kush north of Cabul about 350   "  
1300 miles

In addition to the actual marching time, about 75 days should be allowed for the Treason Trials, the halt among the Euergetae and the founding of colonies in the region of Candahar.

The time allotted by Hogarth​20 for this stretch, 14 or 15 months, is certainly too great. Half that would be nearer the truth, for it is not too much to assume that Alexander could average 12 miles a day for four months with 75 days of rest added, they when we remember that he was anxious to catch Bessus before Bessus had time to cause a serious revolt. The following table, I think, fairly represents the facts:

Alexander leaves Zadracarta (east of Sari) June 13
March to Shahrud (8 days) June 20
Shahrud to Meshed (24 days) July 14
Meshed to Herat (17 days) July 31
Herat to Candahar by the great road:
Marching time  25 days
Halts  75 days
100 days
November  8
Candahar to the Hindu Kush north of Cabul (35 days) December 13

This schedule demands of Alexander less than 13 miles a day  p30 for the 74 days from Zadracarta to Candahar, with 75 days of rest added. The march from Candahar to Cabul was more difficult, and therefore an average of 10 miles a day is assigned to this section. Surely the schedule is conservative, for it was a time when speed undoubtedly was called for.

In order to obtain this schedule, I have been forced to disregard a passage in Plutarch, but as a result we find no difficulty in Arrian. The death of Darius is placed early enough in the year to enable Alexander to proceed according to Arrian. Arrian tells us, it will be recalled, that Alexander passed through the country of the Arachotai in deep snow. The Arachotai inhabited the district between Candahar and Cabul, much of it mountainous, and, since Alexander did not leave Candahar until about November 9th, the mountain passes would be blocked with snow at this time. Although he does not mention winter quarters, Arrian does speak of a halt at the foot of the Hindu Kush to found a city. Alexander reached this point about the middle of December; obviously, he could not cross the great range until early spring and the only inference to be drawn is that Alexander spent this winter at his new city. This is confirmed by Strabo, who says that Alexander passed through the land of the Paropamisadae (roughly, the Cabul valley) after the setting of the Pleiades. Alexander left Candahar before the setting of the Pleiades (middle of November) and reached the Hindu Kush within less than a month of their setting. Since the land of the Paropamisadae stretched south and west from the Hindu Kush, Alexander arrived sufficiently soon after the setting of the Pleiades so that the designation of the season might be used accurately enough. Finally, Strabo expressly states that Alexander built a city and took up winter quarters at the Hindu Kush. The next spring, while there was yet snow in the passes according to Arrian, Alexander crossed the range. Furthermore, these conclusions permit us to accept Arrian's statement that Alexander spent a winter at Zariaspa and one at Nautaca.

On the other hand, if we accept Plutarch's statement that Alexander remained four months in Persis, then the death of Darius, as I have already said, must fall in midsummer of 330, in which case Alexander could not have reached the Hindu Kush before the winter was over. This necessitates one of two things:  p31 either we must assume, with Hogarth, that Alexander spent the winter of 330‑29 somewhere in Seistan (on no evidence whatever), reached the Hindu Kush the following winter, and, contrary to Arrian, divided the winter of 328‑27 between Zariaspa and Nautaca; or, with Tarn, we must, on account of the distance, postpone Alexander's arrival at the Hindu Kush until spring, 329, which is contrary to the meaning of Arrian and Strabo, although it permits us, correctly, to assign a winter each to Zariaspa and Nautaca.

I have little hesitation, then, in saying that Plutarch must bow to Arrian; that his statement of a four months' halt by Alexander in Persis should be cut to seven weeks, so that Alexander may reach the Cabul valley in November (or early December) and take up winter quarters at the foot of the Hindu Kush (so Arrian and Strabo). At the same time, I am able, with Arrian, to assign a winter each to Zariaspa and Nautaca.

The following table will serve to illustrate the conclusions reached in this discussion:

Alexander reaches Persis early February 330
" leaves Persis end of March 330
" overtakes Darius early May 330
" leaves Zadracarta middle of June 330
" reaches Candahar early November 330
" establishes winter quarters at the south foot of the Hindu Kush middle of December 330
" crosses the Hindu Kush early spring 329
" takes up winter quarters at Zariaspa winter 329‑28
" campaigns in Sogdiana summer 328
" takes up winter quarters at Nautaca winter 328‑27
" recrosses the Hindu Kush early summer 327

C. A. Robinson, Jr.

Brown University.

The Author's Notes:

1 III.28.1 f.

2 Pp724-5.

3 IV.7.1.

4 IV.18.2.

5 III.28.1 f.

6 Loc. cit.

7 III.28.9.

8 IV.7.1.

9 IV.18.2.

10 That is, just after the setting of the Pleiades. For a discussion of the precise season meant, see below.

11 Philip and Alexander of Macedon, appendix.

12 The Cambridge Ancient History, Vol. VI, pp390 ff.

13 P385.

14 Arrian, IV.7.1. ἐστε παρελθεῖν τὸ ἀκμαῖον τοῦ χειμῶνος. Hogarth, p303, says that "full winter-time" (used also in reference to Nautaca, see below) may be applied equally to December or to February, and therefore to two winter quarters; but the word to stress here, rather, is παρελθεῖν, which means passed.

15 Ib.IV.18.2. ὅ τι περ ἀκμαῖον τοῦ χειμῶνος.

16 Furthermore, Curtius' account (VII.10.13 f.) is longer than Arrian's and implies a full year's campaign.

17 P289.

18 Curt., V.6.17‑20. Arr., Ind. 40.

19 P296.

20 P300.

[image ALT: Valid HTML 4.01.]

Page updated: 10 Feb 19