r 



I 39^ ] 



To underftand it more perfedly, let us fuppofe the fun for the 

 firft time in the meridian, and to communicate its heat every in- 

 ftant fifteen degrees all around. If it were to remain in this fitua- 

 tion the furrounding air could have no other motion but upwards, 

 for the lateral dilatations being equal, would neceffarily check each 

 other, but in the fecond, and all the fucceeding inftants, the fun 

 moves weflwards ; therefore, of the originally equidiftant eaftern 

 and weftern points, the weftern, to which the fun approaches 

 nearer, is more heated than the eaftern, from which the fun re- 

 cedes; therefore, in this, and all the fucceeding inftants, the 

 eaftern, being more cooled, will prefs on the weftern, and thus 

 an eaftern wind will be eftabliftied. 



It is true, that, in the northern hemifphere, the northern air alfo 

 prefles upon the more heated fpaces, but as this alfo follows the 

 fun's path to the weftward, it becomes- alfo eafterly-, preferving 

 •only a few points of its primitive diredion. D'Alembert adds alfo 

 the folar attradion, which, according to him, elevates the air in 

 the points over which the fun is vertical, and, confequentiy, pro- 

 duces a dilatation advancing from eaft to weft. But Mr. De la 

 Place, not denying this caufe, confiders it too weak to produce 

 fingly any confiderable effect*. 



* Mem. Paris, 1776. 



■: C 2 About 



