horizontal distance, by three-fourths of a mile only; in 26 miles. Thefe, I apprehend, are mistakes, occafioned by hafte.

Velore is a poft of great importance, commanding the great road leading into the Carnatic, from the valley of Vaniambaddy; and the directest route from the Myfore country. It confifts of three strong forts, on as many hills: and is juftly deemed impregnable to an Indian army. It is faid to have been originally fortified by the Mahrattas, more than 200 years ago. Among other acts of generalfhip exhibited by Sir Eye Coote, during the late war; the relief of this place, in the face of Hyder's whole army, may be reckoned a capital one. It is about 90 miles to the weftward of

Madras.

Paliconda, had its bearing taken from Velore, and the distance between them, was measured. Amboor is determined by the distance from Paliconda, and the bearing of Coulafgur; which latter is determined by bearings from Velore and Arcot. The distances are obtained by means of Mr. Pringle's table of roads, with allowance for winding. Laftly, Sautgud is placed by angles taken at

Amboor and Velore.

Thefe, together with Col. Pearfe's line, along the eastern coaft of the peninsula, and thofe before taken notice of in the southern provinces (in fection I.) are all the measured lines that occur in these parts together with one, and only one, obfervation of latitude; that is, at Chinna-Balabaram, about the middle between the two feas. This being the cafe, it will follow, that the various materials of which the geography of the tract in question is compofed, can be no otherwife arranged, than by establishing as primary Stations, fuch places as we have the beft data for; and which have the greatest number of positions dependant on them: and this being done, to adapt the matter to the respective intervals, between these primary stations: which intervals muft of courfe determine the scales and bearing lines of the feveral MS. maps, of which the materials are composed.

I fhall attempt only to give an account of the manner of determining the principal of thefe ftations, or points of connexion; and that for the use of future geographers. To defcribe the whole, would lead me into unneceffary prolixity. The primary points. which it became neceffary to defcribe, in the account of the sea coast, were chiefly from actual measurement; as Tritchinopoly, Tanjore, Madura, Palicaudcherry and I may now add, Coimbettore and Carroór; which are placed by measurement alfo. Arcot and Amboor, I have juft mentioned; and Poliput and Chittoor, are the only remaining ones, that are established by furvey.

Of those that remain to be determined, by tiresome difcuffion, and comparison; and in fome cafes by mere judgment, founded perhaps, on doubtful teftimony; are Bangalore, Trinomalee, Darampoury, Dalmacherry, Gooty, Calaftri, Sami-Iffuram, Innaconda, Combam, Adoni, and Timerycotta. And of those furnished by Col. Pearfe's march, are Nellore, Ongole, and Siccacollum.

Bangalore, as the first mentioned, is also the most important; as being in the centre of the peninfula, and having routes paffing through it, in every direction. It is in itself, a place of great political importance, being a fortress of strength; and from fituation, the bulwark of the Myfore country, towards Arcot. A variety of MS. maps of the country lying on the weft of the Carnatic, and between it, and Seringapatam, have appeared: most of them, I believe, the offspring of the war of 1767-8, with Hyder Ally. One of a much later date, communicated by Mr. John Sulivan, contains the whole peninfula, south of the parallel of 15°; and is particularly valuable, on account of its having many routes and fituations in Myfore, and Bednore; as well as in Tanjore and Madura. By its comprehenfive nature, it furnifhes an opportunity of proportioning the respective distances between Amboor, Bangalore, Seringapatam, Mangalore, and Bednore; as thefe places all appear in the fame map, together with the routes from one to the other. It would require whole sheets to give an analysis of this, and the rest of the MSS. which have

Bb 2

have been confulted on this fubject, and therefore, I shall only ftate generally, that by the medium of the distance from Amboor to Bangalore, in 4 maps, it comes out to be 73,6 G. miles; (the variations between them, was 6 miles) and its parallel, according to the fame method of proceeding, was 4′ 10′′ fouth of Madras, or 13° o' 50". To this may be added, that Chinna-Balabaram, is by the medium of the fame 4 maps, 23′ 40′′ to the north of the parallel of Bangalore: and the latitude of the former, which is fortunately preferved in the Lettres Edifiantes, being 13° 23', confirms the general accuracy of the former refult: this being only 1' 10" different from it; or 12° 59′ 20′′. Laftly, the interval on Mr. Sulivan's map, between Bangalore and Bednore, is 176 G. miles; and in mine, when Bangalore is placed, as above (73,6 from Amboor; and in lat. 13°) 172,5. And again, the distance between Bangalore and Mangalore in Mr. Sulivan's map, is 176,5; and in mine, 169: that is, 74 different in one cafe, and 3 in the other. And this difference is to be accounted for, by our giving a different degree of width to the peninfula: Mr. Sulivan's map making it 15′ 15′′ wider than mine does, in the parallel we are speaking of. I have placed Bangalore in lat. 13o, and lon. 77° 37′ 10′′, according to the above data: and this capital point, or primary station, being fixed with fo much fuccefs, both in latitude and longitude, gives fome degree of confidence to all the pofitions round it: for almost every pofition between Col. Kelly's line on the south, and the Tungebadra river on the north; has, in its construction, a reference to Bangalore. It is the common point of union, in the centre of the peninfula, as Coimbettore is in the S W, and Tritchinopoly in the S E.

Before I proceed to the detail of the remaining primary stations, in the peninfula, I shall obferve at once, that all the roads and pofitions of principal places in the Myfore and Bednore countries, on the weft of Bangalore; and between Roydroog and the parallel of Tellicherry; were furnished by the faid map of Mr. Sulivan's and

of

of which, I believe, there is no other copy in Europe. Chitteldroog, Shevagunga, Bankypour, Chennyroypatam, Ananpour, and many others, are quite new; and Sera, Sirripy, Roydroog, Rettingery, and Cenapatam, appear much more confiftent in their pofitions, than heretofore: and although we cannot expect that either the positive, or relative distances, should be perfectly exact, yet I have every reason to think that they are not far from the truth; and that at all events, the prefent map affords the best materials that can be procured in this country.

Seringapatam is placed nearly in the pofition it occupies in Mr. Sulivan's map, in respect both to Bangalore and Mangalore. Its parallel is very uncertain, as there is no good line of diftance, to check it, either from the north, or fouth. Mr. Sulivan's map places it 99 G. miles to the northward of Coimbettore; and Mr. Montrefor's 87,4. As I have adhered more to the account of its relative position, in respect of Bangalore; it ftands in the map, only 91 from Coimbettore. Befides, a MS. Itinerary gives its diftance from Sera, at 3 days journey, or 54 G. miles horizontal distance; which is perfectly confiftent with the 91 miles from Coimbettore. It is placed in lat. 12° 31′ 45′′, lon. 76° 46′ 45′′*.

Sir George Staunton's journey across the peninfula, from Madras to Mangalore, in 1783, as one of the commiffioners for negociating a treaty of peace with Tippoo Sultan, furnished a lift of stages, and the estimated bearings and distances between them, the whole way. Thefe, Sir George moft obligingly communicated to me, together with his mifcellaneous obfervations and reflections as he went along; and which, being written on the spot, and dispatched as opportunities offered, may be allowed to exhibit a faithful pic

* Other accounts are as follow: a large map, which I confider as the firft that was conftructed from the materials, collected during the war of 1767-8; and which may be ftyled the parent of most of the others, that appear on different fcales; gives 99,3 G. miles, weft from Kiftnagheri; and 2' north of its parallel. This would place Seringapatam 6' weft of its pofition, in the new map. Montrefor's map, gives 66 G. miles from Bangalore, and Mr. Sulivan's, 58. The firit goes 9 beyond it; the other only 2.

ture

ture of the mind that dictated them: and they afford a proof that a long journey in which many things occurred that usually excite disgust, disappointment, and chagrin (not to mention fatigues and want of comforts); may be performed, not only without disturbing the tranquillity of the traveller; but in fuch a state of mind, as to leave him a fufficient portion of good humour, to enable him to amuse others.

The commiffions were conducted by a very circuitous route, as well as impeded in their journey: for after proceeding on the great road from Anicul towards Seringapatam, in a W S W direction, to Malavilly, within about 12 miles of Seringapatam; they were carried to the N, and N W, fo as to leave the latter place, at least 25 miles to the fouthward of them. As far as these estimated bearings and distances enable me to judge, the positions of Anicul and Seringapatam, in the map, are too far to the west, by several miles, in refpect of Caveripatam; the laft point established in this route, with any degree of certainty. But the route is unfortunately, too crooked, to admit of its being applied as a corrective, in fmall errors of distance.

Seringapatam is the capital of Myfore, the dominions of Tippoo Sultan; and it is fituated in an island of the Cauvery river, about 290 or 300 miles from Madras. It has little in it worthy of attention. Myfore, a town and fortified poft, and as I understand, the ancient capital; lies about 8 miles to the fouthward of Seringapatam. Mr. William Townsend, of the East India Company's civil fervice, who travelled from Onore to Bednore, and Seringapatam, was 11 days in travelling between the two latter places; which, however, cannot be more than 180 or 190 miles afunder. He represents the whole country he travelled through, as being open and fruitful: nor did he meet with any mountains between the Gauts and Seringapatam.

Darempoury, Caveripatam, Kiftnagheri, and Changamaħ, are obtained by means of a map of the valley of Vaniambaddy (or the

Barra