I NEXT proceed to the chain of islands that extend from CapeNegrais to Sumatra; and are known by the names of the Preparis, Cocos, Andaman, and Nicobar iflands..

Capt. Ritchie, after leaving Negrais, proceeded agreeably to his inftructions, to defcribe the fituation and extent of the islands that compose this chain.

None of them are more than 84 G. miles distant from each other ;. fo that he needed never to be more than 42 miles from land: and that but once during the voyage; that is to fay, between the Little Andaman and the Nicobar islands. In other places, the distance between the lands is commonly much less: fo that the meridional direction of the course, and other circumstances, render this line of much use in correcting the longitudes, not only of the islands themfelves, but of Sumatra alfo ; and, had it been continued as was intended, to Acheen, would have answered the purpose completely.

Paffing the Preparis and Cocos islands, Capt. Ritchie proceeded to Narcondam, to fix its pofition: then back again to Cocos; down the east fide of the Great Andaman, (which he found to be almost a degree of latitude longer than was before fuppofed) then up the west fide of it, almoft to the latitude of 12°: when finding the attempt to circumnavigate the island, might prove fatal to the remainder of his work, he proceeded fouthward; defcribing the extent, figure and pofitions of the Little Andaman and the Nicobars, till he came to the fouth point of the great (or fouthmoft) Nicobar. Here the wind fuddenly changed to the south, and prevented him from determining the refpective pofitions of the fouthern Nicobar and Acheen which is the more mortifying, as one day's fair wind would have enabled him to accomplish it.

The refult of this line of bearing is, that the fouth end of the Great Nicobar, is in longitude 94° 23′ 30′′; that is, only 3′ 30′′ west from Cape Negrais..

The pofition of Acheen Head, or King's Point (the NW point of Sumatra) has hitherto been deduced from its bearing and distance from Malacca, the neareft place of obfervation; and its longitude according to this deduction, is 95° 30' according to M. D'Apres. Now the bearing of Acheen from Malacca, being in a direction of more than 60 degrees from the meridian, and the distance 450 G. miles; little reliance could be placed on the result of it, if it did not happen that the refpective pofitions of the fouthern Nicobar, and of Pouloo Ronde (an ifland near Acheen) the one deduced from Negrais Point, and the other from Malacca, agreed nearly with their reputed bearing and diftance from each other. For, of two MS. charts which I have examined, the one makes 1°1', the other 1° 2' difference of longitude between them; and thefe being laid down apparently without any attempt to support a system, may be fuppofed to be agreeable to experiment. The bearings and distances in thefe MS. charts are

In one S 56° E72 G. miles.

And in the other S 56 E 75

And according to the deduced longitudes abovementioned, the bearing and diftance is S 56 E-76

So that there cannot be any great error in the longitude of Acheen, as laid down in M. D'Apres', and in my map, if this fort of coincidence can be reckoned a proof of accuracy: a difference of a few miles, in the distance of 8 degrees, being much less than could be expected in fuch a feries of deductions. M. D'Apres makes the bearing and distance between the fouth Nicobar and Pouloo Ronde S 57° 30' E 97 G. miles; or difference of longitude 1° 22′, that is, 22′ more than the MS. charts. It must be obferved, that he reckons the fouth end of this Nicobar, 9 miles farther to the north than the truth; occafioned by his making the island so much too little in extent: for the north end is in its true latitude. Had the fouth point of the island been in its true latitude, the bearing of Pouloo Ronde would have been more eastwardly, and the distance

only

only 93, inftead of 97: and if, on the contrary, he has enlarged his distance on the original bearing, to make it anfwer to the latitude, the original diftance could have been only 85 miles.

I have faid before that Capt. Ritchie went no higher up the west fide of the Great Andaman, than about the latitude of 12°. The remainder of that coaft, as well as the paffage through the islands at the north end of it, is from a MS. chart lent me by Mr. Dalrymple; and which carries with it the greatest appearance of truth, on a comparison of the fouth and fouth-weft parts of the Great Andaman in this chart, with the fame parts in the chart of Capt. Ritchie.

Barren Island, and the rock on the east of Duncan's Paffage, are from the remarks of Capt. Juftice in 1771.

ISLAND OF CEYLON.

IT happens that the ordinary tracks of British fhips, to and from Ceylon, and the coast of Coromandel, are not calculated for determining the relative pofitions of Point Pedro and Point Calymere, the approximating points of Ceylon and the continent of India. Hence it is, that we are fo ill informed, not only of their true fituations with refpect to each other, but also with refpect to the parallel of latitude under which Point Pedro is fituated.

By my obfervations, Point Calymere (the fouthern extreme of Coromandel) lies in 10° 20' latitude, and by inference from Madras, in longitude 79° 54′ 30′′. M. D'Apres places it 6 minutes more northwardly; and D'Anville 7 further fouth. The latitude of Point Pedro, is alfo varioufly reprefented by the above geographers :: I have taken it at 9° 52'.

In M. D'Apres I find the bearing and distance from Point Caly

I had an opportunity in 1764, of determining the position of Cow Island from Tondi, nery nearly: I made it 10 G. miles weft of Point Calymere, and 39 east of Tondi. Between Cow Island and Point Pedro, Van Keulen reckons 41, and D'Apres and D'Anville, 42 miles, of eafting. This will place Point Pedro 31 east of Point Calymere; or in longitude 80° 27′, and in bearing from Point Calymere E 43° 20' S; 42 G. miles diftant. I have placed it in this fituation accordingly; as thinking it not liable to any great objection.

The figure of the island of Ceylon is variously represented by different geographers. Van Keulen makes it too narrow, in the fwelling part: that is, between Batacola and Columbo; as is pretty evident from the longitudes fhewn by the time-keepers of some of the commanders of the Eaft India fhips, and others. D'Anville and D'Apres in their maps of the island, agree much better with thefe obfervations.

Between the meridians of Calitoor and the caft fide of Ceylon in lat. 7° 40', Van Keulen reckons the difference of longitude

However, until a regular feries of obfervations by time-keepers are made by the fame perfon, all round the island, we must despair of feeing the true figure of it, unless its coafts were furveyed. The casual observations which we are in poffeffion of, from different hands, will only affift us in fixing certain points of it; which being done, the general figure of the island must be collected in the best manner it can be done, from the charts exifting.

The

The following are the obfervations of longitude taken on the

Dundrahead by Mr. Topping's obfervation (worked to Pondicherry in lon. 80°) is in lon. 80° 39': I reckon Point de Galle 28' weft from Dundrahead, therefore it fhould be by this account in 80° II'.

Mr. Topping observed the longitude of the Great Baflas also; so did Capt. Dundas: but as we are not well informed concerning the exact difference of meridians between them, and they being at least 1° 22′ diftant, nothing in these observations will apply to Point de Galle.

There is certainly too much discordance between the three longitudes of Point de Galle given above; the medium of which is 80° 8′ 30′′. As Anjenga and Cape Comorin were placed 3' farther to the eastward, than Capt. Huddart's obfervation warranted, in order to accommodate the differences between the two calculations, Point de Galle should be reckoned in 80° 11' 30". Mr. Dalrymple's time-keeper gave the difference of longitude between Anjenga and Point de Galle 3° 29′ 30′′, which added to 76° 40′ (see page 19) gives 80° 9′ 30′′.

If we confider the respective pofitions of Point Pedro and Point de Galle by the different geographers, we shall have the following refult:

Deduced from Bombay, which is reckoned in 72° 40'.

+ Capt. Weft reckoned from Sadras, which I place in 80° 24'. He reckoned it 80° 19′ and of course, Point de Galle in 80° 22.