in computed miles from one pafs to another, and oftentimes from some distant capital place alfo; but without bearings. This MS. gives the diftance of Dalmacherry, at 75 British miles of road diftance, or about 56 G. miles of horizontal diftance, from Arcot. Montrefor's map gives 64, and Mr. Sulivan's 61. Montrefor, also, places it 47, in a NNE direction from Cudapanattum; and Mr. Sulivan 47. I have placed it 56 from Arcot, in a NNW direction; which makes the interval between it, and Cudapanattum, 46; and its latitude is 13° 43′ 30′′. There are three important paffes leading from this place, into the Myfore and Cuddapah countries: and here it was that Doaft Ally, the Nabob of Arcot, was surprised and defeated by the Mahrattas, in 1740.

Gooty or Gutti, is a strong fortrefs on a hill, beyond the river Pennar*, and towards Adoni; and formerly the feat of government of Morari Row, a Mahratta Prince. This place, together with the course of the Pennar, is found in Montrefor's map; which contains more particulars in this part of it, than any other map I have feen. But a difficulty arose in adjusting the position of Gooty, in my map; because the distance between Dalmacherry and ChinnaBalabaram is much less in it (13 miles) than in Montrefor's; and Gooty appears to be afcertained by two lines, drawn from those. places: fo that either the bearings, or the distances, must be rejected. I thought it the fafeft way to adhere to the diftances: as it is probable they might have more weight, than the bearings had, with Mr. Montrefor, who adjusted this circuitous route, between Arcot and the head of the Pennar. His fcale gives 112,2 G. miles, on a bearing of N 3 E, from Chinna-Balabaram to Gooty; and 118,5, N 43° 45' W, from Dalmacherry to the fame place: The interfection of the bearings (which make an angle of about 40°) would place Gooty in lat. 14° 58': and that of the distances, in

* Or Pen-aur. I believe the term aur, for river, which prevails generally throughout the Carnatic, is not found any further to the north than Nellore. D d

15° 15'; and nearly in the meridian of Chinna-Balabaram. I have preferred the latter, for the reason abovementioned.

The inferior branches of the Pennar are taken from D'Anville's map of 1752: but Tademeri, Anantpour, &c. are from the Univerfal History. Gandicotta, on the fouth bank of the Pennar, is remarkable both as a strong fortrefs, and for having a diamond mine near it a particular account of it, will be found in Tavernier. Penuconda a confiderable place near the Pennar river, is faid to be 20 leagues NE of Sirpy, and 20 NNW of Chinna-Balabaram. This account, alfo, is in the Modern Univerfal Hiftory. Cæfar Frederick mentions it as the retreat of the King of Bifnagar (or Narfinga) 8 days journey from Bifnagar.

The Pennar river, after fpringing from the neighbourhood of the Balabarams, runs directly northward, until it approaches Gooty; and then takes a SE courfe by Gandicotta and Cuddapah : after which it changes to the east, and reaches the fea at Gangapatnam, after paffing the fort of Nellore. The MS. account of the Passes, remarks that this river is 300 yards wide at Sami-Iffuram, about 70 miles from the fea; although it is confined in its courfe, by ༡༠ hills, on both fides.

It has been obferved in the firft fection, that Capt. Ritchie's chart of the coast of Coromandel, made the point at the Pennar river, project too far out. I find by a reference to 6 different MS. and printed maps of this part, that the distance of the fea from Nellore, is not represented in any of them, at more than 13 G. miles, and most of them allow only 12. And although I have allowed 16, it comes confiderably within Mr. Ritchie's account.

I

Cuddapah is determined by the map of the Pennar river: and the conftruction agrees with its reputed distance from Arcot in a Malabar map; or rather a map drawn by a native of the Carnatic. It is there ftated at 60 coffes; which on the fcale adopted for the Carnatic (in page 5) and which allows only 37 coffes to a degree, will correfpond with the 96 G. miles, arifing on the construction.

Tripetty

Tripetty and Chandeghere (or Kandegheri) the first a famous place of Hindoo worship; and the latter, the fite of the capital of the ancient kingdom of Narfinga, are placed with reference to Dalmacherry, by Montrefor's map; and by the MS. account of the Paffes. Kandegheri is there faid to be 22 B. miles (road measure) from Dalmacherry, bearing about ENE; and Tripetty is 3 miles ŞE from Kandegheri. I have placed Tripetty accordingly and it ftands in the map 53 G. miles, nearly north, from Arcot; and about 66, nearly N W by W, from Madras. Mr. Orme supposed it to be 50 miles NE of Arcot: and the Univerfal History fays it is 22 leagues WNW from Madras.

Calaftri appears alfo in the alfo in the map of Montrefor. There is alfo a route of General Caillaud's from Polypet to Udegheri, and Nellore, paffing through Calaftri. I have endeavoured to fix the pofition of Calaftri, by these joint authorities; and have placed it 15 G. miles ENE from Tripetty; and 61 from Arcot: but I have my doubts concerning the accurary of its pofition.

Sami-Iffuram pafs, on the Pennar river, is reckoned in the MS. of the Paffes, 55 B. miles, or 44 G. miles horizontal diftance, west from Nellore and this pofition is corroborated by Montrefor's. map (as far as the apparent rudeness of his materials for this part, may be faid to confirm any pofition) and I have placed it accordingly. It comes within about 15 G. miles, or 9 coffes of Cuddapah; which bears from it W sw.

Udegherri and Sangam, two places in General Caillaud's route, are corrected by Col. Pearfe's pofition of Nellore; and by the relative pofitions of Sydaporum and Nellore, in a French copy of M. Buffy's marches: Sydaporum being alfo a pofition in Mr. Cail. laud's route. I am conscious how incomplete the northern part of the Carnatic is, in comparison with the fouthern part: but all my enquiries have produced nothing fatisfactory, on the NW of Polipet. Had the route of General Caillaud been measured, it would

have produced several primary stations: but as it is, the scale appears to be ill proportioned in the different parts of it.

Narnaveram and Bomrauzepollam, are both placed on the authority of Mr. Pringle's obfervations: and the Pullicate lake is from the Jaghire map; Col. Pearfe's route; and other authorities. This lake, called by D'Anville, Ericans, feems to owe its exiftence to the fame caufe as the Chilka lake; that is, to the fea's breaking through a low fandy beach, and overflowing the lands within; for its communications with the fea, are extremely narrow, like the embouchures of fmall rivers. This lake is in extent 33 B. miles from N to S, and II over, in the broadest part; and contains fome large islands within it. One of these is named Ircum, in Mr. Barnard's map of the Jaghire, published by Mr. Dalrymple: and as M. D'Anville names this ifland, as well as the lake, Ericans, I conclude it to be a corruption, or misconception of Ircum.

I have not found it an easy task to fix the pofitions either of Innaconda, Combam, Adoni, or Canoul. On these four places, many others depend, in the conftruction of the map; and they are neither of them afcertained to my fatisfaction. There is, in particular, a degree of obfcurity in the accounts of Canoul, that I cannot clear up. My local information fails me entirely, in this place: and this kind of knowledge is fo requifite to a geographer, that no degree of study, or investigation, can compenfate for the want of it. It not only enables him to reconcile names and fituations; but oftentimes furnishes him with a criterion to diftinguish the value of his materials. Few Europeans, vagrant ones excepted, have vifited these places fince the time of M. Buffy (1751) and it is a misfortune to geography, that his marches between Arcot, Hydrabad, Adoni, Canoul, and Seringapatam, have not been recorded, in the fame intelligent manner, as the rest of his marches have been; and from whence we have drawn fo much information. But, however I may repine, as a geographer; I ought, perhaps, as philofopher, to be fatisfied, that fo much has been preserved.

Innaconda (called alfo Viniconda, and Huiniconda) is a fortress on a hill; within, or bordering on, the Guntoor circar. It is undetermined, as to its precife bearing, from any known place: therefore I have been reduced to take it on the authority of fome vague maps, and by a reference to circumftances: and have placed it about NW by N from Ongole (a point in Col. Pearfe's route). Mr. Pringle measured the road, and found the diftance to be 46 B. miles; for which I allow 36 G. miles, in horizontal distance. By the Malabar map, it is 28 coffes, which may be reckoned about 45 G. miles. It is fomewhat more wefterly in bearing, and also more distant, from Medipilli, than from Ongole.

Combam is reckoned 25 coffes from Innaconda; and 32 from Ongole; or about 51 G. miles from the latter. It is placed in the map, at 48 miles diftant, and nearly weft, from Ongole: but as its parallel is regulated by the affumed pofition of Innaconda, it is fubject, of course, to the fame errors. Tavernier's route from Gandicotta to Mafherlaw, paffed through Combam, (which he calls Kaman) and its pofition accords very well, with the proportion of distance affigned it. More will be faid on this subject, hereafter.

Adoni is reckoned to be 66 coffes from Combam, by the Malabar map; and 67 by a route tranfmitted by Col. Harper to the Madras Government; and which was collected from the information of his guides, while at Innaconda, in 1781. These coffes on the Carnatic fcale (37 to a degree) are equal to 106 G. miles; and this is the distance allowed in the conftruction, between Combam and Adoni, weftward. And for its parallel, no better authority can be found, than its diftance from Gooty, which is fituated to the SE of it, two days journey, or 36 G. miles according to my calculation; which is founded on fome confiderable degree of experience in these matters. Adoni, thus placed, is 63 G. miles. N NE from Roydroog, and about 44 fouth of the Kistna river.