wien: recent theories of heat and radiation 283 



of positive rays is not proportional to the pressure but is very much 

 slower. Thus we see that even under these simple conditions 

 the behavior of the atoms is very complicated. 



Now it is certain that the emission of light takes place thru 

 collisions of the moving atoms with atoms or molecules at rest 

 and the question is whether or not the collisions effecting the emis- 

 sion of light are the same as the collisions which cause the atoms 

 to lose or receive their charges. If we assume that the two kinds 

 of collisions are the same, we must perforce apply the theory 

 of quanta, because one atom cannot send out less radiation than 

 one quantum. From the free path we know the number of col- 

 lisions per cm of path, and, having found the mean energy emitted 

 by one atom, we may calculate how many collisions are neces- 

 sary to effect the emission of one quantum of a spectral line, that 

 is, how many collisions must occur for each one which excites one 

 quantum emission. But the emission of light by the positive 

 rays depends very greatly on the velocity of the particles. If 

 the velocity is very small no emission at all can come from the 

 particles. The emission increases rapidly with the velocity but 

 after reaching a maximum value it decreases so that for great velo- 

 cities it again disappears. The emission of light is therefore associ- 

 ated with a small range of velocities. 



In the light emitted by the positive rays we always have two 

 spectral lines, one coming from the molecules at rest and the 

 other from the moving molecules. One might think it possible 

 to find a relation between those two by applying the principle of 

 relativity. If we make the whole system move with the velocity 

 of the moving atoms and in the opposite direction, then the atoms 

 at rest become moving atoms, the moving atoms are now at rest 

 and nothing is changed. One might therefore conclude that the 

 light emitted by the moving atoms differs from the light coming 

 from the atoms at rest only by the amount of the change de- 

 manded by Doppler's principle. But we have here a complica- 

 tion in that not- only atoms or molecules but also electrons are 

 emitted by atoms in collision and these secondary electrons like- 

 wise cause an emission of light if they are absorbed. 



All these considerations show that the emission of light by the 



