195 



Now, since the radiation is supposed to be the same for the different 

 particles A, B, C, D, it follows that the absorption of the stream of 

 heat by these particles must also be the same for each ; and in order 

 that this may be the case, it is necessary that the stream which im- 

 pinges on one particle be the same in quantity and in quality as that 

 which impinges upon another. This consideration leads us to a me- 

 thod of viewing internal radiation, which is wholly independent of the 

 diathermanous or athermanous character of the body. For whatever 

 be the absorption of a particle for any description of heat, its radia- 

 tion must necessarily be precisely the same in order that the stream 

 of heat in passing the particle may be just as much recruited by its 

 radiation as it is reduced by its absorption ; in other words, we may 

 regard the substance through which the heat passes as perfectly 

 diathermanous. 



We gain another advantage by this method of viewing the subject : 

 for, in the law which is expressed by saying that the absorption of a 

 particle is equal to its radiation, and that for every description of 

 heat, the word description is used to define and separate those rays 

 of heat which are absorbed in different proportions by the same sub- 

 stance. Therefore in any problem connected with this subject we 

 may suppose that a separate equilibrium holds for every such ray. 



Now it is well known that rays of different wave-lengths are ab- 

 sorbed in different proportions by the same substance. We are 

 therefore entitled to suppose that a separate equilibrium holds for 

 each wave-length. The advantage of this is obvious in problems 

 which admit of the application of optical principles. But we may 

 go even further. For we know that in tourmaline, and in some other 

 crystals cut parallel to the optic axis, the ordinary ray is more ab- 

 sorbed than the extraordinary ; and the experiments of Prof, Kirch- 

 hoff and the author have shown that in tourmaline the ordinary ray 

 is also radiated in excess. It thus appears that, in the case of crystals, 

 we have not only a separate equilibrium for each wave-length, but 

 for each of the two rays into which the incident ray is divided. 



The following method of comparing together two streams of radiant 

 heat has been adopted : Consider a square unit of surface to be 

 placed in the midst of a solid of indefinite thickness on all sides, and 

 find the amount of radiant heat which passes across this square unit 

 of surface in unit of time in directions very nearly perpendicular to 



