434 Sir J. J. Thomson on the Origin of 



If we calculate c' from the relation 



, 4 h 2 



7r 4 em 



■18. 



5 



we find </ = 3'97xl0- 



this differs by about 10 per cent, from that calculated from 

 the ionizing potential. The determinations of this potential 

 for the hydrogen atom are hardly definite enough to exclude 

 the possibility of an error of 10 per cent., so that the value 

 3'97 x 10" 18 for c' is admissible. This value would make 

 the numerical constants in the series relation agree with 

 those given by Bohr's theory which give values for the 

 frequencies agreeing with observation. 



The ionizing potential is usually measured by the potential 

 difference through which cathode rays must fall to enable 

 them to ionize the gas. 



When, however, the atom is, as we have supposed, the seat 

 of intense magnetic forces, it is possible that the energy 

 which the cathode rays must possess to be able to ionize the 

 gas may not measure the energy which the cathode ra}^ 

 have to give up to an electron to enable it to escape from 

 the atom, but rather the energy which the cathode rays must 

 possess if they are to penetrate far enough into the atom to 

 reach that electron which has to be liberated. 



Again, in many cases of ionization by cathode rays, the 

 electrons are ejected with a finite amount of kinetic energy 

 which is independent of that of the cathode rays ; when this 

 is so, the ionizing potential is greater than the line integral 

 of the electric force from the initial to the final position of 

 the electron. 



When the spectrum of a gas is excited by the impact of 

 cathode rays, the first process which goes on is the ejection of 

 electrons from the atom ; this does not, on our view, give rise 

 to radiation corresponding to the series lines in the spectrum. 

 This is due to a second process, the return of electrons to the 

 ionized gas to replace those which have been ejected ; these 

 electrons are not the high speed cathode rays but slow speed 

 electrons produced by the ionization of the atoms, and the 

 energy which is converted into radiation is that which these 

 electrons acquire when falling into the positions of equi- 

 librium. From some experiments I have made on the 

 radiation produced by the bombardment of substances by 

 cathode rays, I conclude that by far the larger part of the 

 radiant energy is concentrated in radiation of definite wave- 

 length, for I found that when the energy of the cathode rays 

 was gradually measured the character of the radiation did 



