ment in determining the general bearing of it, further than that we may conclude it to be to the westward of Sanore-Bancapour, because the road from Hydrabad, leads through it; and as the Sanore river (the Toom) was croffed 9 coffes beyond Bancapour, it may probably bear to the northward of weft from it, as the rivers in that part, run to the fouthward of eaft (see the map). The distance between Hydrabad and Bancapour (133 coffes) determines the scale of coffes to be at the rate of 39 to a degree; fo that Chinnaputtun is about 56 G. miles from Sanore-Bancapour; most probably, in a W NW direction. We learn one interesting particular, if true, from this route; which is, that the Nizam's territories extend 31 coffes beyond Bancapour.

Bifnagur, or Bijinagur, the capital of the ancient kingdom of Narfinga, is fituated near the western bank of the Tungebadra river, and about 3༠ miles SE or SSE from Bancapour. It was vifited by Cæfar Frederick in 1567; and was then a very large city. He. reckons it 8 days journey from Goa, which, by the calculation in page 207, should be 144 G. miles; but it is only 130 by conftruction. We are told by Ferishta, that Bijinagur was founded by Belaldeo, King of the Carnatic, in 1344. The Carnatic then, included the whole peninfula; or at least, all that lay to the east of the Gauts. Our hiftories of the Deccan and Carnatic are very imperfect; and at this day we can hardly distinguish between the kingdoms of Bifnagur and Narfinga; and whether they were two fucceffive, or two coexisting kingdoms. It appears probable, however, that in the 16th century, the kingdom of Bisnagur included the greatest part of the peninfula; and that on the invasion of the King of Vifiapour, and other northern Princes of the Deccan, the King of Bifnagur retired, firft to Penuconda and then to Kandighery (or Chandegheri) but ftill preferved his ancient title of Bifnagur. In 1599, Kandegheri was the refidence of a Hindoo King, whofe dominion extended over Tanjore and Madura; and

in 1640, a descendant of this Prince reigned there: and permitted the English to fettle at Madras.

Ranni-Bedalore, as well as the heads of the Tongebadra river, are from M. D'Anville. We know generally, that this river is formed out of several finaller ones, that iffue from the eastern fide of the Gauts, in and about the Bednore country. Further down, it paffes Bifnagar (as is faid above, although Cæfar Frederick calls the river of Bifnagar, Nigonden) and between that and the Kistna, it receives the Hindenny, or Endri river, which passes by Adoni; as well as feveral finaller rivers. The general course of the Tungebadra is represented in the map of M. Buffy's northern marches: and that of the Hindenny is marked more particularly, in the map of his fouthern marches, by D'Anville. It is also described in the map communicated by Mr. Sulivan, as paffing under Chitteldroog, Rydroog, Chitrigally, &c.

M. Buffy's route from Seringapatam to Adoni and Rachore is not to be found in the map of his other marches. Those who caft their eyes on that naked part of the map, will regret its being wanting.

The route from Goa to Galgala is from Mr. Dalrymple's collection; and appears to have been travelled by fome Portuguese, who vifited Aurungzebe's camp on the Kistna, in the latter part of the laft century; or early in the prefent.

The environs of Goa and the country to the foot of the Gauts, are from a Portuguese MS. It is from Goa only, if from any quarter, that we are to expect the geography of the tract between the Gauts, Vifiapour, and Adoni; and which yet remains almost a perfect void, in the map.

The general courses of the rivers in the peninsula, indicate that a ridge of high land runs directly across it, from Calaftri to Mangalore: but if we are to truft report, the country has not a hilly appearance between the Gauts and Bangalore; but that rifing fud

denly

denly from the west, at the Gauts, it declines gradually eastward: fo that the Gauts form a fort of a terrace on an immense fcale.

The Gauts are marked only in certain places where the different roads cross them, or where they have been viewed from the coaft. This famous Appenine, which marks, with more preci- | fion, perhaps, than any other boundary whatever, the line of fummer and winter, or rather of dry and wet; extends 13 degrees of latitude; that is, from Cape Comorin to Surat (with the exception of the gap mentioned in page 196) at unequal diftances from the coast: feldom more than 70 miles, and commonly about 40: and within one short space only, it approaches within fix miles. Although the altitude of these mountains is unknown, yet it is fufficiently great to prevent the great body of clouds from paffing over them; and accordingly, the alternate NE and SW winds (called the monfoons) occafion a rainy feafon on one fide of the mountains only; that is on the windward fide. It would appear, though, that clouds enough do pass over, to occafion a rainy feafon, at a confiderable diftance to leeward, where those clouds defcend as we may suppose them to do, although at the time they paffed over the Gauts, they must neceffarily have been too high, and of course too light, to condense and fall in rain, there. This, I am led to confider, by Lieut. Ewart's account of the weather at Nagpour, in the very centre of India; where the seasons differ but little from their ufual course in Bengal, and on the western fide of India: that is, the SW monfoon occafions a rainy feafon but the rains are not fo violent, nor of fuch long continuance, as in thofe places. At the mouth of the Godavery river and its neighbourhood, the SW monfoon occafions a rainy feafon alfo; and the Godavery is then fwoln and overflows and this part is about as far to leeward of the Gauts, as Nagpour is. It is poflible, however, that the clouds

6

which

which occafion a rainy feafon at the mouth of the Godavery, may come from the east of Cape Comorin: though I rather believe the contrary, as the Cape bears SSW from it, and the reigning winds are much more wefterly. The Nagpour clouds, however, must pass over the Gauts. We may, I think, conclude then, that the ridge of the Gauts fhelter a particular tract only; beyond which, the light and elevated clouds that pafs over it, defcend in rain. Madras is within the limits of the fheltered tract, though at least 300 miles to leeward of the Gauts: Rajamundry (near the mouth of the Godavery) and Nagpour, may be about 500. It would be curious to know the exact limit of wet and dry. If I mistake not, until lately it was a general opinion, that the Gauts extended from the northern (or Bootan) mountains to Cape Comorin; and occafioned a diversity of seasons, at one and the fame time, throughout all India. But the truth is, that different seasons exist at the fame moment, only in a part of the peninsula: for the cause ceases in the parallel of Surat where the SW wind, no longer oppofed by a wall of mountains, carries its fupplies of moisture uninterruptedly, both far and near, over the whole face of the country. For fome particulars refpecting the northern extremity of the Gauts, fee page 179.

As the peninfula, or tract difcuffed in this fection, contains more interesting matter than could well be comprised within the space furnished by fuch a fcale, as could conveniently be applied to a general map, of fo large a tract as India; it was judged neceffary to form another map of the peninsula, on a much larger scale: but an accident has retarded the publication. Thofe who may hereafter become poffeffed of it, will find the account of its construction in this Memoir: which is common to both maps, throughout this. whole fection; and alfo in the firft fection, as far as the map of the peninfula is concerned. The fcale of this map, is just double that of the general one.

SECTION

IT

SECTION VI.

The Countries between HINDOOSTAN and CHINA.

T has been faid before (page 48) that the firft ridge of moun tains towards Thibet and Bootan, form the limits of the furvey to the north to which I may now add, that the furveys extend no farther eastward, than to the frontiers of Affam and Meckley.

The Jefuit's map of China, as given in Du Halde, places the western boundary of Yunan (the westmost of the provinces of China) between the 97th and 98th degrees of east longitude, in the parallel of 24° so that the eastern frontier of Bengal (Silhet) is within 350 British miles of the western part of China; or to speak comparatively, the fame distance as Silhet is from Calcutta. Here one is apt to wonder, that confidering their proximity to each other, there should be no communication between the two countries. The reasons probably are, that Yunan does not produce fuch manufactures as are in request among foreigners; and that the courses of the great navigable rivers in those parts, are unfavourable to a communication by water. The space between Bengal and China, is occupied by the province of Meckley, and other districts, subject to the King of Burmah, or Ava.

1

The river Nou-Kian, little, if at all, inferior to the Ganges, runs to the fouth, through that angle of Yunan which approaches nearest to Bengal; where the Jefuits, who formed the map of China, left it, in its course to the fouth-weft. This river, M. D'Anville conceived to be the fame with that of Pegu; in like manner

as