mountains, Alexander returned to Ecbatana, (Isfahan,) and there rejoining his main army, employed the winter in reorganizing his troops, and dismissing homeward those Greeks whose time of service was expired. This effected, early in the spring of 330, B. c. Alexander crossed the Elburz mountains* at the pass near Tehran, called that of Dumavund, and formed his army in two divisions, employing one of them in reducing the Mardi, a poor and semi-barbarous race, who occupied a tract of country between the Elboorz range and the Caspian, while the other was destined to operate northward up the eastern shore of the Caspian against Hyrcania. With this latter went Alexander, his reason for reducing this wild country being that a body of Greek mercenaries had retired thither. Hearing, however, while his army was crossing the Elburz, that Darius was in force at no great distance, he countermarched and formed a light division with which he went in person to attack him.

The Persian king, assisted by Bessus and the chiefs of Bactria and Darangia, (Seestan) had appeared with an army towards Mushhud,† but refused to abide another battle, and fled as the Greek force approached. In this flight Darius was first deposed and made prisoner, and then slain, and Bessus assuming the royal title, fled towards Bulkh in Bactria. This was in the month of July 330 B. C. and, if Quintus Curtius is right in naming Tabas as the place of the assassination, the flight must have taken a southerly direction from near Mushhud

The pass is called in Arrian, the Caspian gates, and Ragoa is placed near it, D'Anville's map of the ancient world gives precisely the locality of Dumavund for this pass. Darius's flight cannot have taken the line of the west shore of the Caspian, so as to pass the Durband Caspian gates.

†There is much confusion in this part of Arrian's narrative. He mentions the reorganization of the army and many arrangements made at Ecbatana, but leaves it to be supposed that these were operations of a day or two, and that the pursuit of Darius to Rage and the Caspian gates was immediately taken up. But there is a winter intervening between Alexander's march to Ecbatana and the campaign, in the course of which Darius was deposed and assassinated: this season therefore was evidently devoted to the reorganization of the army, and if Alexander did follow Darius to Raga in 331 B. C., it was a mere excursion at the end of the season, not a continuance or renewal of the campaign. Dr. Thirlwall has been misled by not allowing for a winter here. He supposes that season to have been occupied in the operations near Persepolis. The date given by Arrian for Darius's death, compared with that of the battle of Arbela, and the stated military, and civil arrangements made at Ecbatana, prove the manner of the campaign.

or Abbasabad, which, as the Seestan Satrap was the ally of Bessus, is not impossible. Alexander returned, according to Arrian, with the body of Darius, and crossing the Elburz range to the river Atruk, finished the conquest of Mardia and Hyrcania, (Mazenderan and Gheelan.) This effected, he took the direct route to the country of the Arii. The capital of Aria at that time is called by Arrian Susia, probably the Hellenism of Subza or Subzawar,* Herat was not in existence, but is supposed to be on the site of the city or fort erected by Alexander afterwards to control the Arians. Alexander established a Persian governor at Susia, and returned northwards to pass into Bactria after Bessus, by the routes probably of Merv or Mymuna. The Grecian king, however, had no sooner turned his back on the Arian country, than the Persian governor revolted, and having overpowered the detachment left with him in Subzawar, retired to make head at Artakaona amongst the mountains east of Herat. This brought the Grecian army back in haste. Artakaona is a place written six ways, but which probably will be the Greek version of Oordoo Khan, a common name. Sakhir, the capital of the Ghorians at the head of the Kashk river, is a site well suited for a stronghold of refuge, and the Oordoo Khan or Artakhan intended, will probably have been near it.

Alexander followed thither with a light force, making a rapid march of 600 stadia in two days, while the bulk of his army returned southward more leisurely, and moved down to the Pontus, or inland sea, into which the Helmund discharges itself. Artakhan was evacuated on his approach, whereupon Alexander turned southward also, and the Persian governor of the southern districts, called Zarangai or Dranga, (Seestan,) having fled eastward to the Indus, Alexander returned again into the mountains and remained some time there, while he built the fort before-mentioned on the site of Herat to check the Arrians. Here he received the submission of the tribes of the southern

* Dr. Thirlwall supposes this Susia to be Toos, the ruins of which have been traced about seventeen miles NNW. of Mushhud, but Toos would be in Parthia, and not in Aria, as thus situated

† All the Persian poems and traditions mention Astakhar, as the place whence Alexander marched towards India, but the Astakhar of the Shahnama is the capital of Persia. The great Roostum was a native of Seestan.

districts as far as Kandahar in Arachotia,* but it does not appear that he went thither; on the contrary at a late period of the year 330, B. C. he made the passage over the high ridges between Herat and the Kabool valley, suffering much from cold on the march, and then at the junction of the Punjshuhur and Koh-damun rivers, in the plain of Beghram, near Charikar, he founded the city of Alexandria apud Caucasum, about which there has been so much dispute. Its identity with the Beghram ruins has been established in a late† essay of Major Rawlinson, now at Kandahar, and the whole story of Arrian confirms the site. Here Alexander wintered, and at the first opening of spring in the following year, 329 B. C., crossed the Hindoo Koosh to attack Bessus. I consider it most probable that the passage was made from Charikar by the Gorebund or Purwandura Passes, for Drapsacus, which was attacked immediately after the traverse, was evidently the present Indrab, the fortress which gave so much trouble to Chungeez Khan.

The immediate effect of this line of operation was to drive Bessus out of the whole country between the Oxus and Hindoo Koosh, and to

Τε και

Arrian is cited as authority for Alexander's having marched by Kandahar to Kabool, and by Bamian to Bulkh, but Arrian only says the Arachotians submitted, not that Alexander ever went into their country. His words are Ταυτα δε διαπραξαμενος προῄει ως επι Βακτρα τε και Βησσον, Δραγγας Δραγωγους εν τη παροδῳ παρατησαμενος. Παρετησατο δε και τους Αραχωτους και σατραπην κατέτησεν επ' αυτοις Μενωνα. Επήλθε δε και των Ινδών τους προσχωρους Αραχωτοις. Ξυμπαντα δε ταυτα εθνη δια χιονος δε πολλης, και ξυν απορια &c. "Having finished these things, he set off for Bactria and Bessus, in the route having established his authority over the Dranga and Dragogi; he also established his authority over the Arachoti, and appointed Menon their Satrap. He came then into the country of the Indians, bordering on that of the Arachoti, and all these nations he reached through much snow, and in great want of necessary supplies, and with much suffering to the troops." This shews he passed through the Huzara country north of the open plains of Seestan and Kandahar, for in crossing them to the Kabool valley even in October, his army would suffer from extreme cold. He crossed apparently by the route, and in the season, when Babur suffered so much on his return from Herat to Kabool. If the march was made in the season when there is snow at Kandahar, and by that route, the passage to Ghuzni, and especially over the mountains between Ghuzni and Kabool, must have been quite closed.

+ I much regret never having met with this essay, and doubt not that it would have thrown light on many points which are still obscure.

cut him off from retreat into Kashghur. He had fomented another insurrection at Herat, and sent 2,000 horse to support it, while Alexander was making the Huzara passage, prior to wintering in the Kabool valley; but this was defeated by the garrison left in the new city, aided by a detachment sent back, without requiring Alexander's presence. Bessus therefore on the passage of the Hindoo Koosh being effected, retired at once to the mountains of Sogdiana, Nautaka, supposed to be Karshee or Nukhshab, being the position he took up to watch the further course of events. Alexander took Bulkh and all the country south of the Oxus, and established six stations according to Quintus Curtius to guard and command the passes of the mountains. He then crossed the Oxus on skins, at a point where the river was rapid and deep, and had a sandy bottom, which is the character of all the fords about Bulkh. Bessus was betrayed and given up before Alexander reached his position at Karshee, and thereupon Alexander followed up his success by seizing Markanda, (Samarkund), and he thence continued his march, meeting with no serious opposition, to the Sir or Jaxartes, called by Arrian the Eastern Tanais. He crossed this river to punish the Scythian cavalry, who had inflicted on him some loss as they retired before him through Sogdiana. Alexander fought on the other side of the Sir a sharp cavalry action, in which he was wounded severely by an arrow in the leg, his fibula or smaller leg bone being broken. He gained the victory, however, and dislodged the enemy from a mountain supposed to be that opposite to Khojund, with a loss stated at 20,000 men.

Alexander remained sometime on the Jaxartes, and commenced building a city or fort near Khojund. He at the same time summoned all the tribes to a general convention to be held at Zariaspe, (Huzarasp on the Oxus,) in the coming winter; but while he was so occupied in advance, the nomade tribes of the Kizil-koom desert and Lower Jaxartes, rose on the garrisons he had left in his rear, and under Spitamenes, an active and energetic partisan, besieged Markanda. Alexander on the first news of the insurrection retraced his steps towards Markanda, reducing all the cities on his way without difficulty until he came to Cyropolis, which is probably Kesh, or Shuhur Subz, where Persian tradition fixes the birth of the great Cyrus. This siege proved difficult, for the city is described as large and

populous, the walls strong and high, and the inhabitants warlike. He mastered it at length, effecting an entrance by the river bed, during a season of drought, and then returned to secure his posts on the Jaxartes, sending a division to strengthen Markanda. While he was encamped on the banks of this river, seeking to inspire the Scythians with a dread of his power, the division of his army sent for the relief of Markanda, was defeated and utterly destroyed on the banks of the Zurafshan river by Spitamenes. This called Alexander back to Samarkund, and after ravaging the valley of the Zurafshan, he moved to Huzarasp, where he had proposed to winter, in order to hold the convention before proclaimed, and to confirm by policy, the influence his victories had established. It was here that Clitus was slain, and that Scythians from the western Tanais (the Wolga or Don) came, and endeavoured to persuade Alexander to attempt the passage that way back to Europe; but Alexander excused himself, saying, he must first conquer India, and then would come by the route of Europe round that way to the Tanais and Huzarasp.†

Early in the spring of 328, B. c. while the snow was still on the ground, Alexander took the field again, for reduction of the cities of Sogdiana, which still held for Spitamenes. His army marched in five divisions, Alexander heading that which took the mountain road by Samarkund. Soon after the march, Huzarasp was attempted by surprise, but saved by its garrison. Spitamenes then made a gallant attack on the left division led by Conus, which skirted the desert, and had marched apparently for relief of the garrisons of the Jaxartes: it was met by Spitamenes while countermarching for the defence of Huzarasp, consequently on the attack of that post. Being defeated by

Arrian calls the river of Markanda, Polytimetus, the much-valued, Zurafshan is gold scattering. The description of its losing itself in the sands of the Bokhara Desert confirms the identity, if the name and other circumstances had left any doubt on the subject.

The site of Huzarasp, no less than the similarity of name, proves it to be the Zariaspe referred to. It is on the Oxus in the advanced position suited to the convocation, and is exactly the place the Scythians of the Steppes towards the Wolga might be expected to come to, while the river being navigable gave all the desired facilities for forming a depôt. The next year's march back to Samarkund establishes the correctness of this position, and it is further confirmed by the communications held at it with the Choras meni, (Kharizmees.) Kheeva and Orgunj, the capitals of Kharizm, being only a few marches down the stream of the Oxus. The limits of ancient Bactria might well extend down the Oxus as far as Huzarasp.