the accuracy of the map of Tinevelly, the hiftory of which, I am unacquainted with: but, I think, the near coincidence of Mr. Pringle's measurement, with it (for I reckon 3′ 20′′ but a trifle in general geography) is a prefumptive proof of its general truth. It is understood that there is from 1 to 2 degrees of wefterly variation, between Negapatam and Cape Comorin: if this be allowed, it will remove the Cape 4' 30" further to the weft; and place it, according to Kelly's and Call's lines in 77° 38′ 5′′.

Let us now turn to the other coaft, and observe how Capt. Huddart's and Capt. Dundas's, deductions of longitude, from the weft to Anjenga, accord with the reputed space, contained in the interval between Cape Comorin and Anjenga; which fpace, by Mr. Dalrymple's obfervation of the difference of longitude fhewn by his time-keeper, in 1777, was 52′ 30′′. This, taken from 77° 38′ 5′′, leaves 76° 45′ 35′′ for the longitude of Anjenga. Capt. Huddart's longitude of Anjenga, deduced by time-keeper from Bombay, reckoned in 72° 40′ is

Capt. Dundas's

Mr. Dalrymple's D

76° 39′

76 30

76 38

As Capt. Huddart's feries of longitudes commenced at Bombay in lat. 18° 58′, and were continued to Anjenga in lat. 8° 39′, and then back again to Bombay; by which the error of his time-keeper was ascertained, and which was only as much as amounted to 24 minutes of longitude; we have every reason to be fatisfied with this feries, as far as refpects general pofitions: and indeed, geography is greatly indebted to the labours of this gentleman, who has presented us with the longitudes of 16 places on this coaft, and by that means given the true general figure of it, which exhibits, to those who have been in the habit of contemplating it, a very different form, from what it ever did before.

I am of opinion that more dependance may be placed on Capt. Huddart's longitude of Anjenga, deduced from Bombay, in the manner abovementioned, than on any other account: but at the

fame

I

fame time I have adopted Mr. Howe's obfervation of longitude at Bombay, as it appears the most confiftent with other accounts. do not mean by this, to determine on the merits of the different obfervations (of which indeed, I am incapable) but rather because it accords with the obfervation taken at Goa, and with the routes acrofs from Negapatam to Tanore; and, as far as may be judged, with the deduction from Negapatam to Cape Comorin. It is true, that if Montrefor's pofition of Madura is admitted, it will place Cape Comorin 12' farther eastward; and if the variation be not allowed, there will be 4' 30" more to be added; in all 16′ 30′′, or the full difference between Capt. Howe's, and Capt. Huddart's observations; the one being 72° 38', the other 72° 54'.

It now remains, after this investigation, to be fhewn, in what manner I have compounded the above differences; that no diftortion of the intermediate parts should take place. Anjenga, I have placed in 76° 40', being the medium of all the different accounts, by observation; and by deduction from Negapatam. Mr. Dalrymple's difference of lon. 52′ 30′′ to Cape Comorin, is then adopted, which places the Cape in 77° 32′ 30′′. Madura, is placed in 9° 52', the latitude, given by Call; and its longitude is determined by Col. Kelly's distance from Tritchinopoly, with the addition of 3 miles to it; that is, in lon. 78° 11'; and Palamcotta in lat. 8° 42', and lon. 77° 49′ 15′′; according to the proportions furnished by Call and Pringle, between Madura and Poolytopu.

The form of the coaft between Madras and Cape Comorin, is from various authorities. The furvey of the Company's lands (or Jaghire) extends beyond Alemparvé. From thence to Negapatam, is from a French MS. map, collated with D'Anville's map of pofitions (above spoken of) and several particulars between Pondicherry and Portonovo, from Mr. Pringle's map of the marches. The mouth of the Coleroone is from an English MS. map. From Negapatam to Tondi, is chiefly from Major Stevens's, and my own observations: from Tondi to Good-water Ifland, Major Stevens's alone :

D 2

alone from thence to Tutacorin, Capt. Delafield's curfory furvey: and the remainder to Cape Comorin, is from Col. Call's map of Tinevelly; corrected occafionally by a printed chart, published by Mr. Dalrymple. It is not pretended that It is not pretended that any of these points beyond Cuddalore, are ascertained with precision: but it is highly probable that Point Calymere cannot be out in its longitude, 4 minutes. Tondi has the bearing and difference of latitude from Point Calymere to correct its pofition; and there was alfo a line drawn from it to Tanjore. Ramanad is fixed by the intersection of two lines, from Madura and Tondi; and therefore must partake of the errors incident to Tondi, and Point Calymere. The Point of Ramiferam is alfo dependant on Tondi. When I conftructed the map of India, in 1782, I concluded that the respective distances between Tondi, Tritchinopoly, and Devicotta, had been measured, and that I worked on fure ground: but I have fince been convinced of the contrary.

From Cape Comorin to Anjenga the particulars of the coaft, appear to be either little known to us, or very ill defcribed; as the reports of it are various and contradictory. Between Cape Comorin and Ruttera Point, I took the particulars from a French MS. chart, the fcale of which appears to be faulty; for it gives only 35 G. miles of distance between Cape Comorin and Point Ruttera; and the difference of latitude only 13′ 48′′, by which the latter would be in 8° 14'; whereas, it cannot well be under 8° 20′, according to Mr. Howe's obfervation: and Mr. Dalrymple observes that Ruttera Point is nearly 29 G. miles from that of Cadiapatam, which is about 19 from Cape Comorin. On these ideas, I have extended the distance to 46 miles; and ftill Ruttera is only 28' from Cadiapatam. M. D'Apres' account of the bearings and distances between Cape Comorin and Anjenga, gives 42′ difference of longitude; but then his chart contradicts that account, and gives 44 besides an abfolute difcordance in particulars. Mr. Pringle meafured only 49 B. miles of road distance, between Poolytopu and

Anjenga,.

Anjenga, which cannot be more than 38 G. miles of horizontal distance; and falls fhort of my expectations, more than 9 miles. All that I could do, was to give the coaft such a form, as my mind had conceived of it, by perufing thefe different accounts: at the fame time, I confefs, none of them appear conclufive: and until we know the exact pofition of Poolytopu, in respect of Cape Comorin, we cannot allow Mr. Pringle's measurement, to difcredit in the least, Mr. Dalrymple's difference of longitude. We may here observe, by the way, that coafts of fuch rotundity of figure, as the termination of this great peninfula, are seldom fo well determined, as those that embay, and where the fame point remains long in view, and is of eafy difcrimination. Here the projecting points fucceed each other too rapidly to allow a fufficient degree of precifion in calculating either the bearing, or the distance. The latitude of Cape Comorin I have taken at 8 degrees.

Coylan or Quilon, a Dutch factory about 14 G. miles to the NNW or NW of Anjenga, is the next place, whofe longitude is noted by Capt. Huddart; but as it cannot be expected that places differing only a few minutes of longitude from each other, can be determined with precision by this means; I fhall país on to Porca, another Dutch factory, in lat. 9° 15', and longitude by Capt. Huddart 76° 10'. I can by no means reconcile this longitude with the Dutch MS. maps of this coaft; for as the difference of longitude between Cochin and Porca by Capt. Huddart is only 8', the bearing ought to be S 10° E; whereas in the map, it is S 25 E. In order not to do too much violence to either report, I have allowed. 16' difference of longitude, instead of the 8' of Capt. Huddart's, and the 10' of Capt. Dundas. The Dutch MS. in question contains the whole coaft from Coylan in lat. 8° 51′, to Cranganore in lat. 10° 23′; together with that vaft affemblage of lakes, that extend in fome places 30 miles inland; and are the repofitories of the waters that spring from the weft fide of the Gauts; the whole country hereabouts being very flat, marshy, and unwholfome. This

MS.

MS. map, which is also in Mr. Dalrymple's collection, contains a most valuable addition to the geography of this part of the peninfula.

Cochin, the principal fettlement of the Dutch, on this coast, is the next place in Capt. Huddart's table of longitudes: and is reckoned by him, in 76° 2′; lat. 9° 58'. Capt. Dundas makes it 75° 58'; and M. D'Apres, in his new Neptune Orientale, 76° 3′.

Capt. Huddart has not noted the longitude of any place between Cochin and Tellicherry, in lat. 11° 48': and there being only a single observation at the latter, I am inclined to pass over the confideration of it; and proceed to the next point of observation, Mount Dilla (or Delly) where 3 obfervations were taken. Mount Dilla is a remarkable promontory fituated in lat. 12° 1′; lon. 75° 2′; or 1° weft of Cochin.

We have two accounts of the bearings of the coast between these places; the one from M. D'Apres, the other has its particulars from different authorities. Neither of the two, differ effentially; and it is neceflary to examine them carefully, as the result is to be used in comparing the longitudes of Paniany and Tanore, deduced from Capt. Huddart's obfervations; with that deduced from the marches of the Colonels Fullarton and Humberstone, across the peninfula.

ose

The whole, added together, gives a bearing of N 25° 30' W, diftance 144,5 G. miles; on which, the difference of latitude is 2° 10' 30", and wefting 10,62. But as the true difference of lati

tude