Scarce any difcuffion of the fort could be attended with a more fatisfactory refult: and I think it affords the strongest presumptive proofs of the truth of Capt. Howe's observations at Bombay; from which, the longitudes fhewn by Capt. Huddart's time-keeper, are deduced.

With respect to my former ideas of the breadth of the peninsula, although the extent in longitude between Bombay and Madras, remains nearly as before; yet by the fwelling out of the coaft, on the south of Bombay, I reckoned it too narrow by about 30 G. miles in the parallel of Madras; and 27 in that of Pondicherry.

I have now concluded the difcuffion of the longitudes across, and round the southern part of the peninfula; and also an account of their application to the map: for a rigid adherence to difference of longitude even by observations of the above kind, between places not far removed from each other, would in fome cafes, diftort the relative parts of the map beyond probability; and therefore, it was neceffary, in fome measure, to accommodate the differences, when the exifting authorities appeared to carry more weight than the obfervations which, as we have obferved before, are fubject to er ror, even in the application: and they are no less fo, from a casual variation in the rate of the time-keeper. A feries of observations, fuch as we have been confidering, must in a general view, be regarded as decifive; but it would be hazarding too much to adopt each particular longitude, when it was contrary to every other authority. Much less can any abfolute dependance be placed on lines of bearing and distance taken from maps, whofe history and conftruction is not before us. And where more authorities than one may occur, and thofe not agreeing; in fuch cafes, it must be left for the judgment to determine, which appears the most probable. Now, although there are ftrong prefumptive proofs of the general truth of the relative pofitions of the principal points between Cuddalore and Anjenga, yet they do not reft on the fame folid foundations, as the pofitions in the north part of the peninfula: and there5 fore,

fore, Cape Comorin is placed more with relation to Anjenga, than to the eastern coaft. Again, the refpective differences of longitude between Anjenga, Porca, and Cochin, do not well accord with other authorities; and therefore as these differences of longitude are very small, I thought the Dutch MS. map, might be better authority for them, than the differences fewn by the timekeeper. Another particular is to be noted, concerning the longitudes on the south of the parallels of Cuddalore and Mahé: (viz.) that these will be found fomewhat different in the map from the above account; for when the map was constructed, I was not in poffeffion of fome papers which throw an additional light on the subject: but thefe differences are very trifling. Some few errors also crept into the conftruction; so that upon the whole, this account contains rather what the map ought to be, than what it is though, poffibly, the errors I am pointing out may be fo fmall, that they would have escaped the notice of all but profeffed geographers.

Tanore and Cochin are both placed 3' to the eastward of the affigned longitude (page 23); and Negapatam ' to the weftward of what is given in page 14; by which double error the peninsula is made to be 4 narrower than was intended, in that parallel. I was ignorant at that time, that there was a plan of Humberstone's march, which had the author's name to it; and therefore had taken the medium of all the others. We will now resume the subject of the conftruction.

A

The latitude of Calicut I have taken at 11° 18'. This city is remarkable for being the firft Indian port vifited by European fhipping: that is, by the Portuguese, who landed there under Vafco de Gama in 1498. It was then the most flourishing place on the Malabar coaft, the Zamorin or Ernperor making it the capital of a very extensive state. It appears to have fallen in its confequence foon after; the new power of the Portuguese occafioning a revolution throughout the maritime parts of the peninfula.

The form of the coaft between Calicut and Mahé, is taken from a fketch of Major Abingdon's. Chitwa is faid to be in lat. 10° 38',

by Capt. Drummond; but I cannot reconcile its fituation to that parallel; as it cannot well be fo near to Paniany. I have placed it in 10° 33′ 15′′.

Mangalore is the next place to Mount Dilla, in Capt. Huddart's table, and its longitude given is 74° 44′; lat. 12° 50'. For about 5 leagues to the NW of Mount Dilla, the form of the coaft is taken from a French map in Mr. Dalrymple's collection; the rest of the coaft, to Mangalore, and from thence to Coondapour (or Baffelore) in lat. 13° 36′, is little known as to particulars. A large river named Cangerecora, whofe course is from the north-east, falls in about 4 miles to the north of Mount Dilla; previous to which, its course is parallel with the fea coast for about 11 miles, being separated only by a spit of fand. The forts of Nelifuram, Ramdilly, and Matteloy are fituated on this river, which is joined by several other rivers, or ftreams, that defcend from the Gaut Mountains; which, in this part, approach within 22 miles of the coaft. I cannot help confidering this Nelifuram, which is fituated about 12: miles up the river, as the place meant by Nelcynda and Melcynda, by Pliny and Ptolemy; a place vifited by the Egyptian and Roman fhips.

We have been lately brought acquainted with the particulars of the coaft, between Barcelore and Meerzaw (or Merjee) by means of a map drawn by Lieut. Reynolds, during the war which terminated fo unfortunately for the British arms in 1783, in the Bednore country; to which this part of the coaft is oppofite. This map is. drawn in a most masterly style, and contains near 60 G. miles of the coaft; and extends inland to the foot of the Gauts, which here, approach in fome places within 6 miles of the fea, and are never more than 20 from it. It includes the pofitions of Bednore and Bilghey within the Gauts; and alfo, Onore, Batcole, and Coomtah on the coaft. We are furnished with the means of joining this. portion of geography to the reft, by having the longitude of Pigeon Island determined by Capt. Huddart; and by the position of it in.

respect

refpect of Fortified Island, near Onore. Pigeon ifland is very finall, and lies about 8 G. miles from the coaft, and 15 from Onore *, and is in lat. 14° 1. Its longitude is 74° 6' 30".

Be

From Meerzaw, to Cape Ramas in the neighbourhood of Goa, we are but ill informed concerning the particulars of the coast. tween these, are fituated the port of Carwar, and the islands of Angedive, both of them better known to the English in the early period of their India trade, and before they were in poffeffion of Bombay. Capt. Huddart fixed the longitude of Oifter Rock in the mouth of Carwar Bay, and also that of the Aguada Point and castle, on the north fide of the entrance of Goa Bay, or road. This he makes to be in 72° 34′ 30′′; and it is worthy of observation, that the city of Goa, which is 11′ 15′′ more to the eastward, and confequently in 72° 45′ 45′′ by the fame account, was placed in the fame pofition within a fraction of a minute, by the obfervation in the Con. de Temps; and which, for want of being better informed, I formerly difregarded. The pofitions of Cape Ramas, Angedive, and Carwar Points, are corrected in refpect of Aguada, by a set of obfervations and bearings of the late Capt. Howe, whose attention to marine science, was equal to his gallantry, and knowledge of the practical part of his profeflion. I have had occafion repeatedly to acknowledge the aids I have been furnished with, by means of his collection of Obfervations and Remarks, in the poffeffion of Mr. Dalrymple.

The figure of the island of Goa, and its environs, to the foot of the Gauts, is taken from a Portuguefe MS. map of Mr. Dalrymple's. The latitude of Goa, and of the Aguada Caftle, is 15° 28' 20".

Goa, the capital of the Portuguese fettlements in India, and the feat of a Viceroy, is a very confiderable city. It was first taken poffeffion of by Albuquerque in 1510, and from a Prince of Saracen extraction, according to Jarric.

When Fortified Island bore EN distant 2 miles, Pigeon Ifland bore S4 W, and Hog Hand SS E. Hog Island bears from Pigeon Ifland ES diftant about 7 miles.

The

The coaft between Goa and Bombay, near 220 G. miles, is little known to us in detail, nor indeed was the general bearing of it known, until Capt. Huddart furnished us with his series of longitudes. By his account it appears, that this coaft, although in the neighbourhood of our principal marine station and arsenal in India, was defcribed in the charts, with an error of very near a whole point of the compass, in bearing. Indeed the whole western coaft of India has far too great an obliquity from the north towards the west, in all the former charts; my own not excepted. Mr. Dalrymple accounted very rationally for it, by bringing to our recollection the great quantity of wefterly variation of the needle, that prevailed here, during the time of our first voyagers; and which is now reduced to less than 2 degrees. It is well known that it was a long time before the true north was discriminated in charts; and the original idea of the direction of this coaft, was transmitted down to our days.

Perhaps there are few coafts fo much broken into small bays and harbours, and that at the fame time have fo ftraight a general outline. This multitude of fmall ports, uninterrupted view along fhore, and elevated coaft, favourable to distant vifion, have fitted this coaft for the feat of piracy; and the alternate land and fea breezes that prevail during a great part of the year, oblige veffels to navigate very near the fhore. No wonder then, that Pliny should notice them in his time as committing depredations on the Roman Eaft India trade; and although a temporary check has been given them, in the deftruction of Angria's fleets, &c. yet we may expect that they will continue the practice while commerce lasts. They are protected by the shallowness of their ports, and the strength of the country within. As pirates, they have greater natural advantages than those of Barbary, who being compelled to roam far from their coafts, have expenfive outfets; here the prizes come to their own doors; and the cruifers may lie fecure in port, until the prey is discovered.

The