in his Antiquité de L'Inde reckons it 50 leagues from Delhi, which accords well with my idea of the distance between them.

The MS. maps which I have fo repeatedly mentioned as being communicated by Mr. Haftings and Col. Popham, together with others formerly in the poffeffion of Col. Muir, and since his death *, obligingly communicated by my friend Mr. Benfley, of the East India Direction, are corrected in bearing and scale by the primary points of Agra, Delhi, and Agimere; but only a very trifling alteration was required. It may be observed, in respect of the new matter contained in these communications, that a great number of places appear, that were familiar to us, as well in the hiftory of former times, as in the account of recent tranfactions; but which we could not, heretofore, refer even to any general fituation in a map. Still however, much is wanting, to render in any degree perfect, the geography of the tract in queftion, both as to mathematical exactness, and to relative defcription: in particular the latitudes and longitudes of Lahore, Sirhind, Attock, Moultan, Batnir, Agimere, Cashmere, Jummoo, Sehaurunpour, &c.; together with the intermediate roads and particulars of the face of the country, and the course of the river Indus. Until these are procured, we muft be content to remain in ignorance concerning many curious particulars of Indian geography; and fatisfy ourselves with having the fituations of places that are the most interesting, either from having been the subjects of history, or as being connected with the politics of the present times.

The next place in point of confequence to the construction of this part of the map, as it regulates many of the northern pofitions, is Attock, a city and fortrefs on the eaft bank of the Indus, and built by Acbar in the year 1581. We shall have occasion to speak on the subject of its historical importance, hereafter. The position

It is probable that the fevere indifpofition to which Col. Muir was conftantly fubject, from the moment of his arrival, to the time of his death, left him no leifure to recollect that fuch materials were in his poffeffion. This excellent officer, and moft worthy character, died in 1785.

of

of this place, geographically, can only be regulated by the apparent bearings from Lahore and Moultan, in a Perfian map of the Panjab, together with the distances collected from the different accounts in the fame map; in itineraries; and in the Ayin Acbaree. The latter gives for the breadth of the Panjab, from Ludhana to Attock, reckoning from river to river, on the line of the roads between them 185 coffes: and as two of these roads make a confiderable angle with each other, I allow the distance on a straight line to be only 180 coffes; or 259 G. miles. Some accounts that I have feen of the number of coffes, are too much exaggerated, to be depended on, or noticed; and I rely on the Ayin Acbaree, in preference to them. The MS. map communicated by Col. Murray, gives 181 coffes, or 260 G. miles, for the whole distance: but although it comes so very near to my calculation in the general account, it differs in the measure of each particular Doabah, or space, between two adjoining rivers. Attock is placed accordingly, 259 G. miles from Ludhana, on the bearing from Lahore, as nearly as it could be collected from the Persian map: and these data give its latitude at 32° 27'*; lon. 70° 36'. Col. Murray's map places it in lat. 32° 25'.

90

Moultan, fuppofed, with great appearance of reafon, to be the modern capital of the country, which was defigned by the hiftorians of Alexander, under the name of Malli, is go coffes from Lahore (fouth-weftward) according to the Perfian map; 120 according to Thevenot's account; and 110 in Col. Murray's map. The Ayin Acbaree takes no notice of the distance between them; but gives the latitude and longitude of both. The latitude is alfo given by Thevenot, and by the Miffionary's itinerary; and when allowance is made for the latter, in the fame proportion as it differed from the true latitude at Agra, the three obfervations differ among themselves 22 minutes. That is, the Ayin Acbaree gives 29° 52', Thevenot 29° 40', and the itinerary 29° 32', which with the addi

* Ptolemy's latitude of Taxila, which I fuppofe to be nearly on the fite of Attock, is 32° 20'.

coffes

tion of 30' (fee note, page 68) gives 30° 2'. I have placed it in 29° 52'; which, on the aforefaid diftance of 90 coffes from Lahore, gives 70° 40′ for its longitude; or 7° weft from Delhi. The Ayin Acbaree makes it 7° 3': which agrees very nearly with the 90 in the Persian map. This distance alfo accords with the bearings in the fame map; where Moultan ftands S a very little E from Attock, and about 60° to the weftward of S from Lahore. And Col. Murray's map has nearly the same bearings. These three primary points of Lahore, Attock, and Moultan, vaguely as they may appear to be ascertained, are the foundation on which the scale, and relative parts of the whole Panjab country depend.

Thevenot defcribes Moultan as a city of fmall extent for the capital of a viceroyalty, but ftrongly fortified, and having a Hindoo pagoda of great celebrity. The Ayin Acbaree represents it, as one of the most ancient cities of Hindooftan. It has, or had, a great manufacture of cotton cloths; the province itself producing the cotton; as well as fugar, opium, galls, brimftone, &c. Thevenot defcribes the river that led to Moultan, as being partly choaked up, or fpoiled, in its channel, in his time (about 1665) and this had greatly leffened its trade. He also takes notice of a particular fsect of Hindoos there, called Catry; and fays, that this is their proper country. In another place, he explains the Catry tribe, to mean Rajpoots, or warriors; that is, the Kuttry tribe, properly. We fhall take notice hereafter, that these Catries were the Catheri of Diodorus, and the Cathei of Arrian; with whom Alexander warred, on the borders of the Malli. Moultan belongs now to the Seiks, though the poffeffion of it, as well as Lahore, has been often difputed by the Abdalli.

I have not extended the large map of India further to the north, than Attock and Jummoo, because it would have added confiderably to the width of it, without furnishing any fubject, particularly interefting to modern enquiries: and the materials are no of a quality or quantity proper to correct the geography of that part, on an extended scale. I have therefore added a map on a smaller fcale,

in which the tract between the Panjab, Bochara, &c. is described; and a feparate account of it will be given in the course of the Memoir.

The river called by Europeans Indus, and by the natives generally Sinde* (or. Sindeh) is formed of about 10 principal ftreams which defcend from the Perfian and Tartarian mountains, on the north-eaft, and north-weft. The Ayin Acbaree defcribes its fource as being in Cashgur and Cashmere; by which it appears that the people of Hindooftan confider the north-eaft branch as the true Sindet. From the city of Attock, in about lat. 32° 27′ downwards to Moultan, or to the conflux of the Jenaub, or Chunaub, it is commonly named the river of Attock, which in the Hindooftan language, imports forbidden: probably from the circumftance. of its being the original boundary of Hindooftan on the north-weft ; and which it was unlawful for the fubjects of Hindooftan to passover, without special permiffion. Below the city of Moultan, it is often named Soor, or Shoor, until it divides itself into a number of channels near Tatta; where the principal branch takes the name of Mehran. The river, however, when spoken of generally, is called Sinde, although particular parts of it are known by different names. The courfe of the Indus below Moultan, has its particulars from M. D'Anville; but the general direction of its courfe, is confiderably more to the weft, than he defcribes it. This is occafioned by my placing its embouchure so much farther weft than usual, in refpect of Bombay (fee page36) while the pofition of Moultan remains nearly as it formerly was. I obferve that most of the old maps of India give the Indus much the fame courfe as I have done.

* The name Sinde was not unknown to the Romans: Indus incolis Sindus appellatus, Pliny, Book VI. + The ancients reckoned otherwife: the fame Pliny continues to fay, in jugo Caucafi montis, quod vocatur Paropamifus, ADVERSUS SOLIS ORTUM EFFUSUS.

Superftition gave birth to this law, among the Hindoos: a precept nearly allied to that, which forbids their eating any food dreifed on board a boat or veffel. Ferifhta calls the river on which Attock is built, NILAB ; anglice, the blue river. There is fo much confusion in the Indian hiftories, refpecting the names of the branches of the Indus, that I cannot refer the name Nilab to any particular river, unless it be another name for the Indus or Sinde.

The

The Miffionary's itinerary beforementioned, gives the names of many places, and fome latitudes, on the Indus. It places the fortress and city of Bhakor, which the Ayin Acbaree fays, is the ancient Manfurah (though D'Anville fays the contrary) in latitude 27° 12′; Tatta in 24° 20′; and Bunder Lawry (called also Bunder Laheri) in 24° 10. All thefe, I take to be from 20 to 30 minutes too far fouth.

Moultan is about the fame distance from the fea, as Allahabad; that is, about 800 B. miles by the courfe of the river; and our author was 21 days in dropping down with the ftream, in the months of October and November: when the strength of the land floods were abated.

The boundaries of the provinces of Moultan and Sindy on the west, extend a confiderable way beyond the bank of the river; that is to fay, from 50 to 100 miles. The country is in general flat and open from Moultan to the sea; and the province of Tatta itself (the Patale or Patala of Alexander) is said to resemble Bengal, not only in the flatness of its surface, richness of soil, and periodical inundations; but also in the food of its inhabitants, which is chiefly rice and fish. The fite of the ancient capital, Braminabad, is near Tatta; and, in the time of Acbar, fome confiderable ruins of it were remaining: particularly the fort, which is faid to have had an astonishing number of bastions to it. Tatta is made fynonymous to Daibul, in the Perfian tables (which were obligingly lent me by Sir William Jones, and are those mentioned in his face to Nadir Shah) where it is placed in 24° 10′. The itinerary fays 24° 20', and D'Anville 24° 40'. I have placed it according to its reputed distance from the mouth of the Sinde, which brings it to 24° 45' *.

pre

The country known by the name of Panjab, or that watered by the five eastern branches of the Indus, has been very little known to

Pliny reckons the length of the Patale, or Delta of the Indus, at 220 Roman miles; in which he was very near the truth, it being about 210.

5

us

1