marched away from the rock, performed some military feats, and "then came to the river Indus." The course of the narrative shows that it is simply impossible for Aornos to have been washed by the waters of the Indus; for as soon as Alexander had blockaded Aornos, he sent a division of his troops to the Indus, which, on its way, captured and garrisoned a place called Orobatis. It is perfectly clear from this that Aornos was not on the bank of the Indus, or troops would not have been sent from that place in order to reach the Indus. There must, also, have been sufficient interval to allow of marching some miles, and the garrisoning of a town as a link of communication. Modern Bazdira is at the pass of Koh-i-tanga, and as Alexander therefore held possession of that pass, what could be more natural for him to do than to use it, and thereby reach the river-valley of the Barhind, the banks of which would lead his troops to the Indus in thirty-five miles. Curtius distinctly says that Alexander cleared the defile of enemies for the passage of his troops, thereby showing that some mountain pass was actually availed of at this juncture.

Along this route, and at about twenty miles from Bazdira, there is a place called Narbat-awal, and this might represent the Orobat-is of the ancients. The detachment of troops marched on beyond this place Orobatis towards the Indus, and they probably effected a junction with the detachment previously sent down the Kâbul river, and began to prepare a bridge for crossing the Indus. Alexander then gave

over command of the district "on this side of the Indus" (including the rock Aornos) to Nikanor, and marched himself towards the Indus. This is again conclusive evidence that Aornos was not on the banks of that river. Furthermore, in the course of this march, Alexander received the submission of the city of Peukelaôtis, "which lay not far from the Indus." Then comes the statement, "He was accompanied on this occasion by Kôphaies and Assagêtes, the local chiefs. On reaching Embolima, a city close adjoining the rock of Aornos, he there left Krateros, with a part of the army, to gather into the city as much

corn as possible," etc. It is perfectly clear that Aornos could not have been on the bank of the Indus. There must have been sufficient space between the rock and the Indus for the operations of an army in the district of Peukelaôtis. It does not seem to have attracted notice that the expression, "he was accompanied on this occasion," almost implies that Alexander paid a flying visit to the Indus to see how his detachments were progressing with the bridge; and he might also have wished to clear the country around them. This would also account for the blockade of Aornos. It was his ambition to capture the place himself; he therefore simply shut it in, leaving the command of the district to Nikanor, while he proceeded to traverse the ground between Bazdira and the Indus. When he had completed his tour, Arrian recorded the fact that "he was accompanied on this occasion" by the local chiefs; thereby apparently closing the episode. The phrase "on reaching Embolima" would, therefore, mean that Alexander had returned to the siege of Aornos, which he had temporarily abandoned. Krateros is then posted where he could collect stores, and the siege of Aornos began in real earnest. Embolima may possibly be some place on the Landai river not far from Mora, in what is now called the Mûsâ-khail district. It deserves notice, however, that Curtius makes the advance upon Embolima follow the capture of the rock Aornos. As soon as the rock was secured the bulk of Alexander's troops passed over into the valley of the Barhind, and reached the Indus not far from Amb. Arrian says that after the capture of the rock Alexander himself marched towards the Indus, "and the army, going on before, made a road for him, without which there would have been no means of passing through that part of the country." On this General Abbott remarks: "This road was probably the path leading among the precipices above and along the torrent of the Burindu" -in exact conformity with my deduction.

The foregoing route agrees in every detail with the careful statements of Arrian, and traces the course of

Alexander step by step, along a natural and practicable path, finding places along the route which punctually correspond with the Greek narrative in distance, position, geographical character, and name. Furthermore the track is in no place disconnected, nor does it cross impossible ground; but it leads up and down traversable valleys, and round the bases of mountains, and across well-known passes. It shows that Alexander entered India about eighty miles above the Khaibar Pass, and crossed the Indus somewhere near Amb, and not at Attock.

J.R..S. 1894.

45

ART. XXI.-The Cuneiform Inscriptions of Van. Part V. By Prof. A. H. SAYCE.

SINCE the publication of my last paper on the Cuneiform Inscriptions of Van (J.R.A.S. Jan. 1893) a discovery of considerable importance has taken place. In 1890 Mr. de Morgan, now Director of the Service of Antiquities in Egypt, succeeded, at the risk of his life, in taking a squeeze of the inscription on the pillar of Keli-shin Ushnei, which I have described in the first part of this Memoir (J.R.A.S. XIV. p. 66, No. LVI.). He found that there was engraved upon the stele, not only the Vannic text which I have published, but also an Assyrian text which had never been noticed before.

Mr. de Morgan's squeezes will consequently form an epoch in the history of Vannic decipherment. They have been carefully copied and published by the Rev. V. Fr. Scheil in the Recueil de Travaux relatifs à la Philologie et à l'Archéologie égyptiennes et assyriennes, xiv. 3, 4 (1893), pp. 154-160. Dr. Scheil has compared the new Vannic text with that published by myself, with the result that they agree very closely together. He has, however, made the curious mistake of supposing, in spite of Mr. de Morgan's description, that the squeeze taken by the latter represents, not the inscription of Keli-shin Ushnei, but that of Keli-shin Sidek, of which Sir H. Rawlinson had heard a report.1

Along with the Vannic text the squeezes furnish us with a second text, which, as I have said, is in Assyrian. Owing to the mutilated condition of both the Assyrian and the Vannic version, Dr. Scheil did not perceive that

1 See Dr. Lehmann in the Verhandlungen der Berliner Anthropologischen Gesellschaft, October, 1893.