hung-salan), through the Retchna Doabah, with many other positions in and about the Panjab. The road from Jummoo to Berouja, &c. is from Sherefeddin's Timur. Debalpour is known to be on the great road from Delhi to Moultan: and the divisions of the country in the Ayin Acbaree, point out its situation to be far down the Setlege, in the Jallindhar Doabah.

Although we have the dimensions of the Panjab country, in a tolerably satisfactory manner, from N W to S E, both in the aggregate, and in particulars; yet we have not the means of determining its breadth from north to south; or rather, from Lahore to the Setlege. The first junction of the Beyah and Setlege, is stated by the Persian map, at 63 cofses below Ludhana; but we are left to guess the cross distance from Lahore, unless what is said in the same map be true, that it is only 18 cofses from Lahore to Kufsoor, and also that Kufsoor is on the banks of the Setlege: but this is highly improbable from other circumstances; in particular, that the same Persian map allows a greater space between the Rauvee and Setlege, than between the Rauvee and Chunaub; and that Ferosepour, at the conflux of the Beyah and Setlege, is reported by Capt. Kirkpatrick, to be go cofses from Lahore, Col. Murray's map places Kufsoor on the Beyah, and not within 25 cofses of the Setlege. The marches of Timur across the lower parts of the Panjab, afford but a faint light to guide us, such as days' marches; and those in an oblique direction.

THE geography of the Panjab country, as being, comparatively, of little consideration in a map of such extent, has been detailed much beyond its seeming importance. The reason is, that we are not likely, as far as I can judge, for a great length of time, to be possessed of any better materials than those I have exhibited; indifferent as they may be, in many instances: and therefore I con

sider it as the finishing stroke to the whole matter, for some time to come. And if any good materials do cast up, such as the latitudes and longitudes of some principal points, or some measured routes; I shall, I flatter myself, have prepared the ground for the erection of a fabric of a better construction. The Panjab country being the frontier province towards Tartary, and the northern parts of Persia, from whence have sprung the conquerors of Hindoostan in every age, Alexander alone excepted; it follows, that their route to the interior parts of the country, must have led through it. Of all these conquerors, as far as I can learn, the routes of Alexander, Timur (or Tamerlane), and Nadir Shah, are the only ones that have their particulars on record.* Timur's route I have already in part given, as it was interwoven so closely with the geographical construction, and towards which it furnished a considerable proportion of materials; and the remainder will be given. Nadir Shah's route was the ordinary one, by Attock and Lahore; and he returned, as appears by Abdul Kurrim, and M. Otter, by nearly the same route; save that instead of crossing the Indus at Attock, he went higher up, and passed the borders of Sewad, in his way to Jalalabad and Cabul. Unfortunately for geography, no detail of this part of Nadir Shah's route has yet appeared: for was it particularized, it might throw much light on the geography of Sewad, and Bijore; and might also explain some circumstances respecting the courses of the rivers in that quarter, which now remain in uncertainty.

The particulars of the marches of the late Acmet Abdalla (King of Candahar), during his frequent visits to Delhi, in the present age, have not come to my knowledge. Alexander's route then is the only one that remains to be mentioned; and although last

* Turmeshirin Cawn, a descendant of Gengis, or Zingis Cawn, made an irruption into Hindoostan about the year 1240; but the particulars of his route are wanting. Sherefeddin mentions, in one place, that he crossed the Chunaub at Toulumba; and in another, that he besieged the city of Merat in the Dooab. But Ferishta confines the exploits of this descendant of Gengis, to the Panjab country.

in point of order, here, is to be considered as the first, as it respects history, and the gratification of popular curiosity: and therefore, of course, merits a particular and pointed discussion.

I shall now proceed to trace the route of Timur, or Tamerlane, chiefly from his history, written by Sherefeddin; and illustrated by the observations of Capt. Kirkpatrick.

In order to fix the situation of the pass over the river Indus, by which this conqueror entered Hindoostan, and returned from it (for it is evident by the history, that he came and returned by the same pass), it will be necessary for me to anticipate the description of the roads leading from Cabul, and other places in that quarter, to Hindoostan; a subject which properly belongs to the description of the countries on the west of the Indus; but which can never be introduced with more effect, or more to the reader's advantage, than in this place; where it will serve to illustrate at once, the account of the progrefs of Timur, and of Alexander.

The country of Cabul, in respect of its natural geography, is divided into two parts, separated by a ridge of very high mountains, usually covered with snow, which runs from west to east, from the neighbourhood of Ghizni, to that of Deenkote on the Indus, below Attock. The tract lying to the north of this ridge, is named Lumghan (or the Lumghanat); and that on the south Bungush (or the Bungushat): each having one or more considerable rivers intersecting their whole length, and disemboguing themselves into the Indus. That of the Lumghanat, is the river of Cabul, called also the Kameh, and in its lower part the Attock: and that of the Bungushat is the river that passes by Nughz, and Bunnoo, and whose proper name I am as yet uninformed of. Each tract has also its proper roads, and its países over the Indus, from the districts of Cabul, Candahar, and Ghizni, respectively: but the northern, or Lumghanat road, is the one in common use, whether to Persia, or Samarcand, notwithstanding its circuitous nature, as it respects the former of these countries; for it is both the easiest and the safest.

Q

Not to enter too much into detail, I shall only mention the principal of these communications.

1st. The upper, or Lumghanat road, between Cabul and Lahore, passes through Paishawur, and crosses the Indus at Attock. This, which is named the KING'S ROAD, is the most practicable for armies, and the safest for caravans and travellers: and is accordingly, as we have said before, the most in use.

2d. Kherpeh, which appears to be only a branch of the former, separating from it at Jalalabad. It is not sufficiently particularized, to enable us to understand its entire course.

3d. The upper Bungushat road, which crofses the Indus either at Deenkote, or Reishi: and leading to Cabul by Bunnoo and Nughz and to Lahore by Bheerah. The positions of Deenkote, and Reishi, I am not enabled to fix with precision; but am led to suppose that Deenkote is from 30 to 35 G. miles below Attock; and Reishi, a very few miles above Deenkote. The great ridge of mountains just mentioned, lies immediately between these passes and Attock; and shuts up the communication between the Bun→ gush and Lumghan roads, as far as respects the commodious, or safe passage of an army; for the heights of SINDIA BUSTEH, which are described as steep, craggy, and tremendous, are to be crossed between Deenkote and Paishawur, by those who would go from the former, to Cabul. A better way is by following the course of the Bungush road, through Kohaut, Bunnoo, Nughz, and Irjab; and this way Timur came. Kohaut I find in Bernoulli (Vol. I. p. 70) at 12 cofses from Paishawur; and it is also four marches from Reishi (Kirkp. MSS.).

How far the snowy ridge extends westwards, I know not; probably it sinks in its progress that way; as, although mention is made of it in Timur's march from Cabul to Nughz, yet it is not represented as forming an obstacle to the progress of his army. Its position in that quarter, is marked by the march of Timur, and by the circumstance of Irjab being in the Lumghan division; Shenuzan, in that of Bungush. What direction the ridge afterwards

takes, we know not; but it does not cross the line of the road.between Ghizni and Candahar. Probably it passes on the south of Candahar, and forms the height named Suhmandroog; which occurs in the march of Dara Sheko, at 45 G. miles to the S E of Candahar: in which case, it answers to the chain of mountains described by Ptolemy under the name of Parveti; and as the southern boundary of the province of Paro-pamisus (Asia, Tab. IX.).

The 4th, is the lower Bungush road, which crosses the Indus at Chowpareh pass, 25 to 30 G. miles below Deenkote; or perhaps 60 below Attock. This route also leads to Lahore through Bheerah; to Ghizni and Cabul, by Fermul (which lies SE from Ghizni, S from Cabul); and to Candahar, by the DESERT. When we consider the existence of the desert in question, and the mountains, wilds, and dreadful torrents, that occur in the Bungush roads, we are no longer at a loss to know why the road by Attock and Cabul, to Candahar, and Persia, in general; though circuitous, obtains the preference. It was in use even in Tavernier's time, when the Mogul government was in its vigour. To the above obstacles, we may add, that arising from the presence of certain fierce Afghan tribes, which occupy the country in the line between the Panjab and Candahar: and which have never continued long enough under a course of subjection and civilization, at any one time, to enable them to retain any traces of the latter, after the subjugation had ceased.

These are all the great communications, which apply to our present purpose of describing the route of Timur from Cabul towards Hindoostan. According to Sherefeddin, he went by way of Irjab, Shenuzan, Nughz, Banou (or Bunnoo), and thence to the Indus, at the very place where Gelaleddin, King of Karasm, fought with Gengis Cawn; and so heroically swam the river, after his defeat, in 1221.

This account of his progress admitting of no doubt, as to the general facts; and the description, bearing, and tendency, of the