﻿480 Dr. Norman Campbell on 



that which corresponds to Y' , it would make little difference 

 to the theory; Y' would be an average value for all possible 

 directions of projection. Bat if the initial velocity were 

 equal to or greater than V the theory would break down 

 completely ; the effect of an electric field would not be to 

 increase ionization, for all the electrons projected in the 

 direction of the field have already the power of making ions 

 at collision, but to decrease it by preventing from ionizing 

 by collision the electrons projected against the field. We 

 may conclude from the agreement between theory and ex- 

 periment that the delta rays from a gas have not an initial 

 velocity of the order of 20 volts. 



Secondly, if the delta rays had an initial velocity of the 

 order of that required to ionize, the rays from the metal 

 would produce more ionization per electron than the rays 

 from the gas ; for the rays from the metal only enter the 

 gas if their initial velocity is in the direction in which they 

 fall under the field. Accordingly, when we calculated the 

 number of electrons sent out from the metal from the amount 

 of ionization produced by them on the assumption that they 

 behave like the electrons sent out from the ga*, we should 

 obtain too high a value. As a matter of fact we obtain too 

 low a value ; the delta rays from the metal appear to produce 

 slightly less, not more, than their fair share of ionization. 

 Again we conclude that the initial velocity of the rays is not 

 of the order of that required to ionize *'. 



General Conclusions. 



19. From all the evidence which has been brought forward 

 in this and the two previous papers we may draw the following 

 conclusions :— 



(A) The delta rays have not, as has been supposed, an 

 initial velocity of the order of 20 volts. 



(B) If the initial velocity is more than a small fraction of 

 a volt, it is independent of the speed of the exciting alpha 

 rays, and almost, if not quite , independent of the nature of 

 the material in which the deita rays are excited. 



(0) The number of delta rays excited by a given beam of 

 alpha rays at a metal surface is independent of the nature 

 of that surface. 



* It may be noted that in Townsend's experiments the ions from 

 the electrode, being those due to the photoelectric effect, probably had 

 a small initial velocity. This velocity would only affect the results so 

 far as the first collision is concernad, and, since the electrons made many 

 collisions, its influence is likely to be small. But it ma} r account partly 

 for the fact that Townsend's value for V is less than my own. 



