﻿318 ox EXTENDiyO A GEOGRAPHICAL SURVEY 



values of the angles made by the chords, by doing 

 Mhicl), Me can come nearer the truth, than by sup- 

 posing them to be spherical ; and though these arcs 

 mav not be precisely correct, yet it has been found 

 that a triHing deviation from the truth will not sen- 

 sibly affect tlje angles. 



It m'ay be further observed, that we are not cer- 

 tain, either of the ratio of the earth's diameters, or 

 of its being an ellipsoid. We have assumed that fi- 

 gure, and have drawn our results froni the average of 

 different measurements, made in different parallels, 

 though among themselves they appear contradictory : 

 but we must adopt them, until better measurements 

 can be made, to enable us to come nearer the truth. 

 Should the figure of the earth prove to be the ellip- 

 soid, and tlie ratio of the equatorial diameter to the 

 polar axis become known, a celestial arc would afford 

 a datum in any assigned latitude, by which, and the 

 observed angles corrected, the direct distances might 

 be computed, and also the distance of any object 

 from a known meridian and its perpendicular, and 

 consequently its longitude and latitude. 



But should the earth prove to be neither an ellip- 

 soid, nor a figure generated by any particular curve, 

 of known properties,- but a iigure whose meridional 

 section is bounded by no law of curvature, then we 

 can obtain nothing until v/ehave an actual measure- 

 ment, to be applied as has been already mentioned. 



Thus much I have thought necessary to premise, 

 that the general principles of the work 1 have before 

 r!;e may be understood ; — principles, which I believe 

 have never been applied in Indian geography, though 

 in England sufficient lias been done to manifest their 

 perfection, and to give tliose gentlemen, M'ho have 

 «p;)lied them, a disUnguished reputation in the an- 

 nals of science : and 1 own, that it was from reading 

 the details of their operations I was first led to con- 

 sider the subjeet. Tlie publications of the late 



