account (40 by Bernoulli's); and 50 from Lahore, in the map of the Panjab. Jummoo stands then about N by E from Lahore ; and Sealkote is 10 cofses from it, towards Lahore: and as Sherefeddin reckons 33 cofses between Jummoo and the Behut river; and Kirkpatrick, 4 marches between Sealkote and Jhylum; the position of Jummoo will be found of the greatest use in fixing the angle of the road between Lahore and Attock.

Noorpour is the next point, situated 12 cofses SE from Bifsooly, by Forster's account; 10 from Shahpour, in Bernoulli. This town is the capital of a district of the same name.

Another principal point is Ray Gaut, or Raypour, a remarkable pass on the Beyah river. This place in Col. Polier's map, stands 24 cofses north-eastward of Sultanpour; 25 from Jallindhar.* And by Forster, it is 19 cofses from Noorpour: but as 2 of these cosses were at a great angle with the rest, 17 or 18 may be taken.

Ray Gaut is situated a considerable way within the level country of the Panjab: for Mr. Forster's party, for security, left the great road of Cashmere, which leads from Bullauspour, through Nadone, and Hureepour, to Noopour; and struck to the left, or SW, from the neighbourhood of Nadone; thereby tracing the south bank of the Beyah, from that place to Ray Gaut. To regulate the direction of this line; and of course to fix Nadone, and Hureepour, and eventually Bullauspour, and Nagorkote; we have the following data:

Bejwara, called also Hoshearpour, is 3 journies (perhaps 36 cofses) from Sirhind (Bernoulli); and 16 cofses N E of Jallindhar (Polier); so that it may be about 25 G. miles N of Rahoon. Rajapour is 10 cosses N, or N by E, from Bej wara; 30 from Roopour, towards Jellamooky, the temple built over the subterranean fire near Nagorkote. Hureepour is 11 cofses beyond Rajapour, on the same road; Jellamooky 6 or 7 farther. All these

* Jallindhar is a known position, 48 G. miles from Sirhind, and 13 to the right of Nakooder, on the Lahore road.

particulars are from Bernoulli: and we collect from them, that Hureepour is 21 cofses; or about 30 G. miles from Bejwara; 84 from Sirhind. And by means of Seba, a fort in Mr. Forster's route, and noticed also in Bernoulli, we find that Hureepour must stand 23 or 24 cofses from Ray Gaut; which, with the distance from Sirhind, will fix its position. For Seba is 18 cofses, east northwardly, from Ray Gaut; and 7 south west, or SSW, from Hureepour. Bompal, in Forster, is 15 cofses to the eastward of Seba: and Nadone, a town on the Beyah, about 3 miles to the NW of Bompal.

Seba appears to be the Siba of the old maps, and the capital of a district of the same name; and may have been succeeded in that capacity, by Nadone.

Kote Kangrah, the fortrefs belonging to the famous temple of Nagorkote, is given at 49 royal cofses, equal to 99 G. miles, from Sirhind (northward). Purchas, Vol. I. p. 438, says, that it is 80 (common) cofses from Lahore, which is probable.* Bernoulli says, if meant for royal cofses. Bernoulli reckons Jellamooky 6 or 7 cofses from Hureepour: and Kangrah 7 from Jellamooky; which is probable. And it would appear that Jellamooky should lie NE from Hureepour; SSW from Kangrah.

says, it is 50 cofses from Jummoo; which can only be true,

Mr. Forster estimated the course NW, from Bullauspour to the camp of the Kangrah troops; and the distance 26 cofses: and from thence 10 more cofses, very circuitous, and perhaps only 6 in direct distance, westward, to Bompal; which we have stated to be 3 miles SE of Nadone. This is, however, but poor authority for placing Bullauspour; and Bernoulli only says, that it is about 3 journies from Roopour (a well known point, on the Setlege) through Nalagur; and this latter is 12 cofses of the way, from Roopour. This would place it 51 G. miles, of direct distance,

The Persian MS. map of the Panjab, places Kangrah about N NW W from Nadone; and Jellamooky midway between them.

from Roopour; 671⁄2 from Sirhind: but the country being exceedingly mountainous, the direct distance is probably much lefs. Forster's distance from Bompal might give 46 G. miles of direct distance, in a country less rugged; but here, perhaps, only 40 or 41. The Panjab map allows 2 parts in 5, of the distance between Nadone and Jummoo, for that between Nadone and Bullauspour; which produces 44 miles: and it is probable that the 3 journies from Roopour may only produce as much; and then Bullauspour will stand about N 26° E, 60 G. miles from Sirhind. And in order to this, we must suppose the Setlege to take a SW course within the hills (as all the maps indeed represent); and it is highly probable that it should run off at right angles with the direction of the mountains, which rise above one another, ridge behind ridge.

We have no materials with which we can fill up the lower parts of the Panjab, in the intervals between the rivers. From the Setlege to the Jhylum, and to the hills that bound the Panjab, on the N and E, it appears to be chiefly a tract of flat and very fertile land: indeed, the courses of the four eastern rivers very plainly indicate a regularity of surface, and sameness of soil. But between the Jhylum, or Behut, and the Indus, there is a large proportion of hilly country, which may be supposed to influence the course of the Jhylum, and give it so different a form from the others; the river conforming to the border of the hilly tract.

The geography of this hilly tract, of which Capt. Kirkpatrick has transmitted a particular account, I had misconceived, when I published my first edition; for the Joud mountains, or rather hills, which compose a principal part of it, I had supposed to lie to the north of the upper road to Attock; whereas they evidently lie to the south of it. And this occasioned me to mistake the pass over the Indus, by which Tamerlane entered and returned from Hindoostan; as I shall make appear to the reader, when I take up the subject of that conqueror's progress.

Capt. Kirkpatrick quotes the commentaries of the Emperor Ba

ber, to prove that the Joud hills are those which extend from a point on the Behut, 7 cofses N of Bheerah (which we have placed at 60 cosses W, a little north from Lahore, page 87), to the territory of the Ghickers, a little N of Rotas; that is, near 70 G. miles along the west side of the Behut; for Baber, who crossed the Behut, a little above Jhylum, is said to have passed beyond (or to the NE of) the Joud hills, in his way from Pirhala to Sealkote. Baber also distinguishes a second ridge of hills, 7 cofses west of the Joud ridge, which were occupied by the Jajoohah tribe (descended from the same stock as the Joudis): and this ridge, he observes, joins, or approximates on the one side, to the Cashmere hills; and on the other extends to Wesunkote, on the Indus. Both these ridges are marked on a Persian route in the Panjab, as extending from the Behut to Diga Serai; and both are also designed in the Panjab map. The Jajoohahs extended westward to the river Sohaun,* in Baber's time; and this emperor observes, that although the hills that lie between Bheerah and the city of Nilab, were occupied by various tribes; yet," that mountainous tract "had obtained the general name of Joud." He remarks also, that the continuation of the hilly tract in question, which joined to Cashmere, was held by the Ghickers: and as Pirhala, 3 marches N, or N W of Bheerah, was the residence of one of their princes, it is certain that the Ghicker territory must have extended from Cashmere to the second ridge of hills mentioned by Baber; which would carry this territory, in some places at least, considerably to the south of the upper road to Attock..

It may be expected that the boundaries of the Joud and Ghicker territories fluctuated very considerably, not only during the invasions of Hindoostan, from the north (which these tribes often favoured; they being by nature and by situation, particularly the Joudis, capable either of favouring, or of obstructing such inva

* This is one of the smaller rivers of the Panjab, and takes its course between the Behut and Sinde. It will be found in the map, together with several others, which were deemed too unimportant to merit a particular description in the Memoir.

sions), but even at ordinary times; and therefore it would be vain to attempt drawing a line of boundary between them. It is, however, clear, that the Ghickers inhabited the northern and largest division of the hilly tract, situated between the Behut, the Sinde, and Cashmere; and the Joudis the southern and least division: and we find also, that the latter division terminated at a few cofses N of Bheerah.

What the nature of the country may be on the S and SW of Bheerah, we are not told. The Panjab map marks no hills on the S of Kooshab, which is 10 or 12 cofses below Bheerah: but by the description of the conflux of the Behut and Chunaub, which is a great way to the S of Kooshab, the soil in that part is rocky: which implies that the new-made land does not extend far into the tract between the Behut and Sinde.

We may remark, that the inhabitants of this hilly tract (meaning both Joudis and Ghickers, or their predecessors), have generally sided with the invaders of Hindoostan at large; considering their own rugged and sterile country only a passage to the other. They invited Alexander, and they appear to have invited Humaioon: and they rather favoured the progress of Tamerlane, and others. Probably, from the nature of their country, they have ever maintained a sort of independency of the Hindoostan emperors, excepting in the most vigorous periods of their government. How they stand with respect to the Seiks, and Abdalli, at present, I know not: but in Col. Browne's List of Residencies of the Chiefs of the Seiks, I perceive only a single one, that is Rotas, in this

tract.

By the help of the Persian and other MS. maps, particularly a map of the Panjab,* obligingly communicated by Colonel John Murray, Commissary General to the army in Bengal, I have been enabled to give the road from Vizierabad to Yehungsaul (or Je

The map in question, was drawn by Lieut. Rind, of the Bengal establishment; as well as that of the roads in the Dooab, mentioned in page 66.