190 REMARKS on the 



blance, the methods of this aftronomy are as much diverfified 

 as we can fuppofe the fame fyftem to be, by pafling through the 

 hands of a fucceflion of ingenious men, fertile in refources, and 

 acquainted with the variety and extent of the fcience which they 

 cultivated. A fyftem of knowledge, which is thus aflimilated 

 to the genius of the people, that is diffufed fo widely among 

 them, and diverfified fo much, has a right to be regarded, 

 either as a native, or a very ancient inhabitant of the country 

 where it is found. 



IV. The conftruction of thefe tables implies a great know- 

 ledge of geometry, arithmetic, and even of the theoretical part 

 of aftronomy. 



In proof of this, it is unnecefTary to recapitulate the re- 

 marks that have been already made. It may be proper, how- 

 ever, to add, that the method of calculating eclipfes, to which 

 thefe tables are fubfervient, is, in no refpecl, an empirical one> 

 founded on the mere obfervation of the intervals at which 

 eclipfes return, one after another, in the fame order. It is 

 indeed remarkable, that we find no trace here of the pe- 

 riod of 6585 days and 8 hours, or 223 lunations, the Saros of 

 the Chaldean aftronomers, which they employed for the pre- 

 diction of eclipfes, and which (obferved with more or lefs ac- 

 curacy) the firft aftronomers every where muft have employed, 

 before they were able to analyfe eclipfes,, and to find out the 

 laws of every caufe contributing to them. That empirical me- 

 thod, if it once exifted in India, is now forgotten, and has 

 long fince given place to the more fcientific and accurate one, 

 which offers a complete analyfis of the phenomena, and calcu- 

 lates, one by one, the motions of the fun, of the moon, and of 

 the node. 



But what, without doubt, is to be accounted the greateft 

 refinement in this fyftem, is the hypoihefis employed in calcu- 

 lating the equations of the centre for the fun, moon and pla- 

 nets, that, v/?». of a circular orbit having a double eccentricity, 



or 



