﻿of the Electron Theory of Matter,' 617 



emission for the direct dynamical escape of the electrons in 

 virtue o£ their kinetic energy exceeding the work of the 

 surface forces. It will not necessarily be possible to dis- 

 tinguish, by statistical or thermodynamical methods, between 

 the photoelectric or thermionic origin of the electrons. We 

 shall therefore try, first of all, to find out what we can 

 about the nature of the sethereal radiation and of the atmo- 

 sphere of electrons in the statistically steady distribution 

 without considering their origins. 



In the first place, the complete sethereal radiation will be 

 identical in constitution at every point of the enclosure. 

 This may be shown by bringing the enclosure into connexion 

 with any other enclosure at the same temperature. The 

 sethereal radiations may be allowed to intermingle by opening- 

 partitions provided with optically transparent diaphragms 

 which prevent the flow of electrons. If there is any net 

 transference of sethereal energy under these conditions, the 

 system may be made to violate the second law of thermo- 

 dynamics. The equality which thus follows holds not only 

 for the energy of the complete radiation, but also for each 

 range of wave-length, for each plane of polarization and for 

 each direction of propagation; as may be shown by inserting 

 known devices for limiting the waves transmitted. As the 

 equivalence holds independently of the shape and contents 

 of the enclosures and of the position of the point of contact, 

 it follows that the transmissivity of the complete sethereal 

 radiation is identical at every point and that its nature is 

 determined solely by the temperature. Thus Stewart and 

 KirchhofPs conclusions about the nature of the sethereal radia- 

 tion are unaffected by the occurrence of thermionic and 

 photoelectric emission. 



For the present I wish to avoid discussion of the vexed 

 question of the nature of the interaction between the material 

 parts of the system and the sethereal radiation and to confine 

 my remarks to the conclusions which may be drawn from 

 the existence of a statistically steady condition of the sethereal 

 and electronic radiations. The immediate problem may be 

 thus stated: — An isolated system contains a certain amount 

 of energy. This is divided between the sethereal vibrations 

 and the material part of the system in a way which is per- 

 fectly definite but which we leave undetermined. There is, 

 therefore, a definite but undetermined amount of energy to 

 divide among the electrons and the other material portions 

 of the system. The problem which we require to solve is, 

 how is the energy divided among the electrons under these 

 circumstances ? 



