circumstance that I have just mentioned, that Khojend is considered in all my authorities, as being situated to the east or north-east of Samarkand, proves I think sufficiently, the incorrectness of the position given, almost universally, to Khojend in all preceding maps, which is due north, or nearly so, of Samarkand. I have one more argument in favour of the position I have given to Khojend. The town of Jizzakh, a place of considerable note, is well fixed by numerous routes from Bokhâra and Samarkand. The whole of my routes make it project somewhat to the westward of a line drawn from Samarkand to Khojend, which it also does in the map, as now laid down; whereas, should Khojend be removed more to the north, Jizzakh would lie to the east, instead of the west of this line.

Having thus settled the position of Khojend, I proceeded to determine that of Kokân, and here I was necessitated to trust entirely to the two routes, which alone reached beyond Khojend, and which both agreed in placing it, as nearly as possible, in a line with that city and Samarkand. I had, I think, only one observation on Kokân, and that not much to be depended on; however, on account of its short distance from Khojend, it cannot be much misplaced.

The grand route from Samarkand to Kâshghar, which has hitherto preserved nearly a north-easterly direction, now takes a sudden turn to the eastward, and, passing through Ferghâna, crosses the lofty mountains which lie to the east of that country, and reaches Kâshghar; its general direction being a little to the south of east, though, from the mountainous nature of the country, it makes occasionally considerable deviations from that line.

As it would have been folly to expect any considerable degree of correctness, in protracting so long a route from a point so uncertainly laid down as Kokân, I was obliged to assume a position for Kâshghar from some of the oest authenticated maps; and then having two fixed points, at the extremities of the routes, I easily inserted them, and had the satisfaction of finding, that their length did not materially differ from the distance which I had already given in the map, between the two towns. One of the routes, written by Syed Izzet Ulla, a most intelligent traveller, enabled me, by the information it afforded respecting the surrounding country, to insert many towns and villages of Ferghâna, besides those actually passed through in the journey. The other gave little more than the length of the stages and the names of the places through which it passed.

I had now completed an outline of the country to the north of Samarkand and Bokhara, and it will be sufficient to add, that it has been filled in from the information afforded by Baber's and Mr Elphinstone's description of the country, and from such particulars as could be gleaned from the accounts of Ebn Haukal, and other writers who have touched on the geography of these countries.

As I found, after availing myself of every piece of information which I could at all consider as correct, that the map was still so meagre and imperfect as, in many places, not to answer my chief object, the illustration of Baber's expeditions, it became necessary to insert many towns and some small rivers, as well as to complete the ranges of mountains, from very imperfect and doubtful authorities. The former I

have distinguished by affixing an asterisk to their names; with respect to the latter, I shall distinguish what is doubtful, and what may be depended on, in a short account of the mountains and rivers contained in the map.

But before I proceed to this part of my Memoir, it will be proper to give some account of the method adopted in drawing the countries to the south of Samarkand and Bokhâra. For the situation of all the grand points, I am indebted to a MS. map of Lieut. Macartney, corrected by the Honorable Mr Elphinstone, and, generally speaking, the whole of the intermediate towns, rivers, &c. have been inserted from the same authority. However, from having some routes which were not in existence at the time Mr Macartney constructed his map, I was enabled to make many corrections and additions. Particularly, in the journey between Bokhâra and Balkh, I have inserted some villages, and a small river which runs into the Kohik river; I have plotted another route along the Amu river, which extends as far as Eljîk, the western extremity of Bokhâra; I have laid down two cross routes through the desert, one, from Karshi to Bushîr on the Amu river; another, from Karshi through Kirki to Andkho. I also carefully compared such routes, as I believe must have been in Lieut. Macartney's possession, with his map, and had occasion sometimes to make slight alterations, though I never did so without a most careful examination, being well aware of the general accuracy of his works. One of the most considerable alterations which I have made, is placing Hezret Imâm, the Karatigîn river, and the route from Killa-Barat-Beg to Wiskirni, considerably more to the west than they stand in his map. My authority for so doing, was, on one side, a route which, coming up the Amu river, passes through Hezret Imâm, and proceeds to Kûndez. On the other side, this arrangement agrees remarkably well with the journey along the Amu through Badakshân, which joins the route between Killa-Barat-Beg and Wiskirni, at a place called Yokatût. I have inserted a few additional towns or villages in Badakhshân on the banks of the Amu, as well as the streams which flow into that river from the southward. Amongst the former will be seen Shehr Derwâz the capital of Derwâz, which is inhabited by a fair and handsome race of people, calling themselves descendants of Alexander the Great. The limits and provinces of Bokhâra, I was enabled to describe from the MS. accounts of that kingdom by Mr Elphinstone.

A SHORT ACCOUNT

OF THE

MOUNTAINS AND RIVERS OF BOKHARA AND

FERGHANA.

THE principal range which connects the lofty mountains of Hindûkûsh and Mûztagh, and which gives rise to the two most considerable rivers in the map, is the Belut-tagh, whose highest point appears to be the mountain of Pûshtikhar, the source of the river Amu. From this spot till its junction with the Hindûkûsh, the range is well laid down and described in Mr Elphinstone's Caubul, and it is quite unnecessary for me to say anything about it in this Memoir, a very small part only appearing in the map. That portion of the range which lies to the north of Pûshtikhar, is what must now engage our attention; and, of this little seems to be known, except that it joins the Mûz-tagh. It appears exceedingly probable, and has already been conjectured by Mr Elphinstone, that the mountains crossed by the route between Ferghâna and Kâshghar, are a continuation of this range. These mountains, when they reach the lat. of 42°, throw out a branch called the range of Mingbûlâk, reaching to Tâshkend, and, shortly after, either terminate or become so inconsiderable, as to form no obstacle to a free communication between Tâshkend and Kâshghar to the north of the Mingbulâk mountains. In long. about 71° and lat. about 41° 31' in the Belût-tagh, lies the real source of the Sirr or Seihûn river; though what is usually considered as its source, is situated in the Mingbûlâk mountains, considerably to the north-east, in long. about 70° and lat. 42° 31'. The Belût-tagh, in its progress from Pûshtikhar to Mûztagh, probably throws out many branches to the west, as the whole of the country in that direction is described as mountainous in the extreme. The only branch of the Belût-tagh, to the south of Pûshtikhar, which is contained in the map, is the Badakhshân mountains, which have the effect of giving a north-westerly direction to the river Amu, during part of its course. The rivers which rise from the west of the Belûttagh, are the Sirr river, the Shiber, the Penj or Amu, and the Badakhshân river. Those to the east are the Kâshghar river and the Kâmeh river.

The mountains which I shall next speak of, and which, from their magnitude, ought perhaps first to have engaged our attention, are the Asfera mountains, which I have also seen denominated the Pamere mountains. This range forms the southern boundary of Ferghâna, and runs in a direction almost due east and west. From its latitude and

lxviii

ACCOUNT OF THE MOUNTAINS AND RIVERS

its peculiar nature, which is that of a very broad chain of mountains rising from an elevated table land, there is little reason to doubt that it is a continuation of the Mûztagh. The only considerable river to which this range gives rise on the south is, I believe, the Surkhâb or Karatigîn, which, according to Lieutenant Macartney, has a course of 180 miles to its junction with the river Amu. A large river runs into the Sirr near Kokân, dividing into two branches as it approaches that city, which, according to Lieutenant Macartney, comes from the Asfera mountains, and has a course of 70 miles before it joins the Sirr. It appears probable, that the river rising in the mountains between Kokân and Kâshghar, which is said by Izzet Ullah to join the Sirr near Khojend, is the same as this Kokân river. The Asfera mountains extend from a long. of 71° to their termination near Khojend, and it is a strong evidence of their magnitude and impracticability, that the only communication between Bokhâra and Ferghâna, is by the pass of Khojend, between the extremity of this range and the river Sirr. All the mountains which now remain to be described proceed from this range.

I shall first notice the Ak-tagh, or White Mountains, which leave the Asfera mountains in long. 67° 30'. On approaching Uratippa they separate into two parts, the most westerly, which I can only trace as far as 63° 30′ long., forming the northern boundary of the Valley of Soghd; whilst the other, which is the proper Ak-tagh, separates Bokhara from Yâr-Ailâk, and terminates in two divisions at Jizzakh and Jopâr. This range has been laid down from the concurrent testimony of all my routes, as well as of Baber's accounts, and I feel, myself, great confidence in its correctness. The next branch which proceeds from the Asfera mountains, and which is much more considerable than the former, is the Kara-tagh or Black mountains. Of this range I have scarcely any information. All that appears to be certainly known of it is, that it proceeds from the Asfera mountains, from which it holds nearly a southerly direction, and that it is lofty and exceedingly rugged and precipitous. The celebrated pass of Derbend is situated in this range, which is the usual communication between Shehr Sebz and Hissâr. Mr Erskine has suggested to me, since the construction of the map, that this range probably leaves the Asfera mountains near Khojend, and that the hills running from the north of Khojend to the Mingbûlâk mountains, are a continuation of the range, through which the Sirr forces its way at the pass of Khojend. I am rather inclined, however, to prefer the position I have given to the northern part of this range in the map, as I can find, in my routes, no account of any such lofty mountains near Khojend, and, were I to insert them, they would leave no space for the length assigned to the Ak-Sû and Khojend rivers, and would not agree with Baber's marches over the mountains from Asfera to Samarkand. The Kara-tagh gives rise to several rivers. The Kohîk, the Shîrâbâd river, the Hissâr river, and the Cheghâniân, are amongst the number. It will not be unseasonable here, to mention the great uncertainty which there is concerning the positions of Hissâr and Cheghâniân, and indeed respecting the whole country of Hissar; a circumstance peculiarly unfortunate, as it is the scene of many of Baber's exploits. The two above-mentioned cities have had many observations, but they differ so widely that no confidence can be placed in them. I have laid them down in the map from some routes in Ebn Haukal.

The Samarkand mountains, which form the southern boundary of the Valley of Soghd, though I have not traced them beyond 66° of long., I have every reason to suppose, are a branch of the Kara-tagh. Were I to turn them to the north, they would intercept the Kohîk river, the source of which is universally agreed to be situated at a great distance to the eastward in the mountains which lie towards Sirkul. The only river which rises from the Samarkand hills, is the Kârshi river.

Much more might be said respecting the geography of these countries, but I have studied brevity as much as possible in this Memoir. For the same reason I have avoided giving any account of the political boundaries and divisions of Bokhâra and Ferghâna, a general idea of which may be formed from an inspection of the map. From what little I have said, it will be seen that the geography of these countries is still in a most imperfect state. I trust that the attempt I have made to give a tolerably correct delineation of them, though it must contain many and considerable errors, will be received with indulgence; particularly when it is considered that, of its northern portion, Ferghâna, little more has appeared in preceding maps than the name. I have the satisfaction at least of knowing that its principal object, the illustration of the first part of Baber's Memoirs, has been in a great measure attained, and that whatever faults may hereafter be discovered in it, have not arisen from want of diligence, in the use and comparison of such materials as could be procured. The public already know what Mr Elphinstone has done for geography in his excellent map lately published with his description of Caubul. The greater part of the materials used in the construction of this map, have been supplied by his kindness. The only merit I can claim, is that of comparing these modern accounts with the particulars of the country already known, and committing the result to paper.

TANNAH, DECEMBER 29, 1816.