EillSSIOX THEORY. 235 



of that of light ; if the weight of a luminous molecule 

 were the 640,000th part of that of the cannon ball, it 

 would in like manner overthrow a wall. 



These deductions are certain : but let us look at the 

 facts. A luminous molecule not only cannot overthrow a 

 wall, but it even penetrates into an organ so delicate as 

 the eye without occasioning the least pain, without even 

 producing any sensible dynamic effect. We can say 

 more : in experiments undertaken with the view of ren- 

 dering sensible the impulsions of light, physicists have 

 not been content to use an isolated agent, they have 

 brought to act simultaneously the immense quantity of 

 light which can be condensed at the focus of a large lens ; 

 they have not opposed to the shock of the rays very re- 

 sisting objects, but bodies so delicately suspended that a 

 breath could derange them enormously ; they have ope- 

 rated for example, on the extremity of a very light lever 

 suspended horizontally by a spider's thread. The sole 

 obstacle to the rotatory movement of such an apparatus 

 would be the force of reaction, which the thread would 

 acquire in twisting. But this force might be consid- 

 ered as nothing, since from its nature it always increases 

 rapidly with the degree of torsion ; and, in this instance, 

 one of the observers whose experiments I am analyzing, 

 found no perceptible force of this kind, after having had 

 the patience to give the thread 14,000 turns, by turning 

 the lever round on its centre. It is then well established 

 that, in spite of their excessive velocity, myriads of lumi- 

 nous rays acting simultaneously produce no perceptible 

 force. But we should be <roinr bevond the legitimate 



/ o 



consequences which this interesting experiment author- 

 izes, if we concluded that a ray is not composed of mate- 

 rial elements endowed with a rapid motion of translation. 



