236 FRESNEL. 



We may, indeed, faii-ly deduce from the absence of all 

 rotation in tlie lever suspended by tiie spider's thread, 

 under the action of an enormous quantity of hght, that 

 the elementary particles of the luminous rays have not 

 dimensions comparable to the millionth part of the finest 

 molecules possessing any weight. But as there is nothing 

 to shovF any absurdity in supposing them a million, or a 

 myriad, times less than this, this kind of experiment and 

 argument (the first idea of which is due to Franklin) 

 cannot furnish any decisive conclusion. 



Among the objections whicli Euler has presented in 

 his works against the theory of emission, I will point out 

 two, on which he has particularly insisted, and which 

 seem to him irresistible. "• If tlie sun," (said this great 

 geometer,) "continually darts out particles of his own 

 substance in every direction, and with enormous velocity, 

 he must end by exhausting himself: and during the many 

 ages which elapsed since the historical period, some dimi- 

 nution ought already to have become sensible." 



But is it not evident that this diminution depends on 

 the magnitude of the particulars ? Now there is nothing 

 to hinder our supposing them of such small diameters 

 that, after millions of years' continual emission, the mass 

 of the sun should not be sensibly altered. And, besides, 

 there is no accurate observation to prove that this lumi- 

 nary does not waste, or that its diameter is really as great 

 as it was even in the time of Hipparchus. 



No one is ignorant of the fact, that millions of rays can 

 penetrate together into a dark room through a pin-hole, 

 and there form distinct images of external objects. In 

 crossing each other in that minute space, the material 

 elements of which we suppose this multitude of rays to 

 consist ought, nevertheless, to encounter and clash against 



