passes over the Indus, and the nature of the countries they lead through, coming from the pen of the Emperor Baber; whose local knowledge in those parts was very considerable, as being obtained by long experience, at the head of armies: and as the snowy ridge of mountains lay between Timur's route to Bunnoo, and Attock; and being unnoticed by the historian, was certainly never crossed by Timur (nor could any ordinary reason have induced him to take so circuitous a route, through exceeding difficult roads): I say, taking all these circumstances into consideration, we can have no doubt that Timur crossed the Indus on the south of this ridge. That Gelaleddin crossed it there, is clearly inferred from the course of his march from Bamian to Ghizni, and thence to Gurdaiz;* all of which places, point towards the Bungush road, and not to that by Attock and it is unquestionable that Gelaleddin and Timur crossed the Indus at one and the same pass, for Sherefeddin positively says so; and this pass can be no other than Deenkote, or Reishi.

The arrival of Timur in the district of Joud, when he had crossed the Indus, is, according to Baber's account of the geography of Joud, an additional proof that the pafs lay far below Attock. And now that the geography of the Panjab is so far improved by Capt. Kirkpatrick's communications, we may with certainty conclude, that Timur returned from Hindoostan by a route below Attock. We trace him from Delhi to Jummoo; from whence, after 33 cofses of road, he came to the river Behut, or Jhylum, near Jubhan, or Gebban; and pursuing the course of the river 20 cofses, he came to Sambaste, situated within the mountains of Joud.‡ This is a clear indication of his having turned to the south, from Jummoo: for had he gone 20 cofses upwards on the Behut, after

Capt. Kirkpatrick informs us that Gurdaiz (probably the Kerdiz of Sherefeddin) was a point in Gelaleddin's retreat from Ghizni: also that the name of the rock in the Indus, nearly opposite to Attock, has no reference to Gelali, or Gelaleddin.

+ Sherefeddin,

+ Ibid.

falling in with it, at a point 33 cofses from Jummoo, he would have been advanced into the northern part of the GHICKER territory, instead of being in that of JoUD. The circumstance of his crofsing the desert of Gelali, both in coming, and in returning; (that desert commencing near the place where Gelali, or Gelaleddin, crossed the Indus), is also in proof of my supposition. The positions of Jubhan, Sambaste, and Berouja, cannot be recognized in the present geography; but that of Berouja must fall very near to Pirhala, if it be not meant for the same place.

Timur's first object after crossing the Indus,* being to effect a junction with the army of his grandson Peer Mahmud, which was then besieging Moultan, he directed his course that way, instead of taking the common road to Delhi, by Rotas and Lahore. The neighbourhood of a navigable river being a desirable object to an army which was marching through a dry sterile country, he pushed for the nearest part of the Behut, or Jhylum river, where he attacked and took the fortrefs and island of Sheabadin Mobarick.

Sherefeddin allows only two marches between the Indus and the Jamad, or Behut; but it cannot be less than 6 marches from the Indus to the Behut, and therefore there must be an error in the statement of the time. Between the time of passing the Indus, and the arrival at the conflux of the Behut and Chunaub, 12 days elapsed; that is, from the 12th, to the 24th of Moharrem; and as it may be about 10 marches from Deenkote to the conflux, by Timur's route, here is an interval of 2 days more than the march required; which might have been employed in the reduction of Sheabadin's fortress, &c. Perhaps the date of Timur's arrival at the fortress, instead of the 14th of Moharrem, should have been the 18th.

A fortrefs, not named, stood on the west side of the conflux of the

* October 11, 1398. The chronology of this event is differently stated: I have followed M. de la Croix's translation.

Sherefeddin, in his description of Cashmere (Vol. II. chap. xxxi, of his History of Timur), applies the names of Jamad and Dindana, to the Behut.

Behut and Chunaub rivers; and just below it, Timur threw a bridge of boats over the confluent waters. This junction, we have observed elsewhere, is accomplished with great violence and noise; and Sherefeddin says, that the waves dashing against one another, make it appear like a troubled ocean: for such large rivers, pent up within strait rocky beds, form a rapid and troubled stream at their meeting. A like description of it is given by the historians of Alexander; which, together with the concomitant circumstances of general position and connexion, render this conflux a most unequivocal point of reference, in the comparison between the ancient and modern geography.

Timur made only one march between the conflux and Toulumba: and as he did not bring the army with him, but left it crossing the river, it may be inferred that the march was not a short one; and therefore I allow 20 road miles for it, in a SE by S direction; that being the position of Toulumba from the conflux, by the above construction. According to Sherefeddin's manner of writing, one could hardly be certain whether Toulumba was on the side of the Rauvee, or the Chunaub. Ferishta has placed . it on the latter, erroneously; for the missionary above quoted, came down the Rauvee, and passed by it; moreover, giving its latitude at 15′ north of Moultan, though it is probably 25'. This is a considerable town, and a pafs of consequence, on the Rauvee; and often occurs in Ferishta's History of Hindoostan. It was in the neighbourhood of this place, that Alexander made war on the Malli, or people of ancient Moultan; as will be taken notice of in its place. Timur staid here 6 days, and then proceeded with the whole army across the Baree Doabah, to Shahnawaz (or Shanavas), a large and populous town near the north bank of the Beyah, after its separation from the Setlege. Here he found more grain than his army could consume; whence we may infer the fertility of the country, which is low and flat, and subject to periodical inundations, like Bengal. Sherefeddin describes at this

place, a deep lake, fortified round with a wall, and defended by 2000 men; which reminds one strongly of something similar at Sangala, which Alexander attacked, before he reached the Hyphasis; but the hill, which was fortified round with carriages, is wanting and Shahnawaz is 90 G. miles from Lahore, whilst Sangala was only 3 marches, that is, lefs than 40 G. miles from the place where the Hydraotes (Rauvee) was crossed; which is supposed to be at the place where Lahore now stands.

It was something more than 2 marches from Toulumba to Jenjian, a town on the south bank of the Beyah, opposite to, and not far from Shahnawaz. As Timur's army was 3 days in passing the Beyah, some in barks, and others by swimming, it may be reckoned a considerable river.

At Jenjian, Timur staid 4 days, and was joined by Peer Mahmud, who had by this time taken Moultan. Timur's next station is Jehaul, 3 marches from Jenjian, on the road to Delhi: and here he separated from his grand army, which he directed to proceed by Debalpour, and to rendezvous at Samanah, a town 85 cofses on the west of Delhi; whilst he proceeded with 10,000 horse to Batnir, a strong fortrefs about 70 cofses from Jehaul, and far to the right of the Delhi road; being beyond the desert which stretches along the south side of the Setlege. He was led to this place, from resentment, at its giving protection to the people of Debalpour, who had massacred a garrison placed in it by Peer Mahmud: and possibly the great reputation it had for strength, might be one inducement to undertake the siege of it; as Aornos, in like manner invited Alexander.

Timur, after leaving Jehaul, proceeded the first day to Adjodin, or Pauk-putton (of which we have spoken before), a town included in one of the large islands formed by the branches of the Setlege: here he visited, and spent some time in devotion, at the tomb of Sheik Furrid (see page 104), and then set forward for Batnir ; which is stated, by Sherefeddin, at 60 cofses from Adjodin: 50 of

which, equal to about 95 B. miles, Timur is said to have marched in one day and night only, with a view to surprise the place.

On the south-east of Adjodin, a few cofses, he crossed the river Dena; which I take to be one of the four branches of the Setlege (pofsibly the Dond), and perhaps the only principal one among them, except the Beyah, already noticed.

It must not be omitted, that Timur crossed an extensive desert in his way to Batnir: I mention this particular to shew that Alexander was not misinformed, when he was told that there was a desert beyond the Hyphasis.* After taking and destroying Batnir,. which employed only a few days, he marched on the goth of November, taking a very circuitous road to Samanah; where he joined his grand army on the 8th of December, 1998. It is not easy to collect the distance, from Sherefeddin's account of the marches; and no rule can be applied to the marches of a party of horse, as to those of an army in general. For certain points in the march of Timur, between these places, see pages 76 and 77.

His march from Samanah to Delhi, about 85 cofses, appears to have taken up 12 days; whence we may collect, that the common marches of his grand army, were about 7 common cofses each day; or about 13 British miles by the road.

Quintus Curtius.

+ Batnir is represented as a very strong place, and yet Timur is said to have taken only a body of horse with him (and indeed the extraordinary length of one of his marches seems to prove it).

The mean length of the marches of great armies, is more susceptible of rule; and, of course more useful in its application to geographical purposes; than may be at first imagined. I shall here mention the result of my inquiries concerning Indian armies only; that being the point to my purpose. The mean of 95 marches, where the measured distances are given, is equal to 14,6, or about 142 B. miles (that of European armies in India is considerably lefs). These 14 miles of road distance, may be allowed to produce 11 G. miles of direct distance, when applied to geography, on lines of considerable extent: although each individual march might produce half a mile more, by the compound winding being laid out of the question (see page 6 and 7).

The march and retreat of the Greeks from Natolia to Trebizonde, give, at a mean, only 15 British miles for each march; reckoning Xenophon's parasangas at 3 Roman miles (as M. D'Anville does, in his Mes. Itin); and, it is pretty evident, both from the space marched through, and from other circumstances, that his number of parasanges is tiken too high.