504 



time the importance of the experiment with tourmaline as a test of 

 the theory. When a suitable piece of tourmaline, with its faces cut 

 parallel to the axis, is used to transmit ordinary light, the light which 

 it transmits is nearly completely polarized, the plane of polarization 

 depending on the position of the axis. The reason of this is, that if 

 we resolve the incident light into two portions, one of which consists 

 of light polarized in a plane perpendicular to the axis of the crystal, 

 and the other of light polarized in a plane parallel to the same axis, 

 nearly all the latter is absorbed, while a notable proportion of the 

 former is allowed to pass. 



Suppose now that such a piece of tourmaline is placed in a red-hot 

 enclosure ; the theory of exchanges, when fully carried out, demands 

 that the light transmitted by the tourmaline, say in a direction per- 

 pendicular to its surface, plus the light radiated by the tourmaline 

 in that direction, plus the small quantity of light reflected by the 

 surface of the tourmaline in that direction, shall together equal in 

 quantity and quality that which would have proceeded in the same 

 direction from the wall of the enclosure alone, supposing the tour- 

 maline to have been removed. Let us neglect the small quantity of 

 light which is reflected from the surface of the tourmaline, and, 

 standing in front of it, analyse with our polariscope the light which 

 proceeds from it. This light consists of two portions, the trans- 

 mitted and the radiated, both of which together ought to be equal 

 in quality and intensity to that which would reach our polariscope 

 from the enclosure alone were the tourmaline taken away. But the 

 light which would fall on our polariscope from the enclosure alone 

 would not be polarized ; hence the whole body of light which falls 

 upon it from the tourmaline, and which is similar in quality to the 

 former, ought not to be polarized. Now part of this light, or that 

 which is transmitted by the tourmaline, is polarized ; hence it fol- 

 lows, in order that the whole be without polarization, that the light 

 which is radiated should be partially polarized in a direction at right 

 angles to that which is transmitted. 



Another way of stating this conclusion is this. The light which 

 the tourmaline radiates is equal to that which it absorbs, and this 

 equality holds separately for light polarized in a plane parallel to the 

 axis of the crystal, and for light polarized in a plane perpendicular 

 to the same. 



