﻿176 Mr. J. C. Pomeroy on the Charges on Thermions 



Hence the method gives a method of determining the value 

 of <£(N.*.Z). 



It was not necessary for the purpose of this investigation 

 to determine the form of this function, for in these experi- 

 ments the method was to compare : 



I. The value of Q c ./Q r given hy ions from various 

 sources, and 

 II. The value of Z necessary to give Q c /Q r a constant 

 value for these different ions. 

 Thus Townsend found that for hard X-rays, 



+ 

 for all values of Z, and so concluded that e, the charge on a 



+ 

 positive ion, is equal to twice e, the charge on a negative ion. 



In the early stages of the experiment there was con- 

 siderable difficulty in interpreting the results, as the average 

 charge on the positively charged particles given off by the 

 heated platinum seemed to vary from that given by radium 

 to nearly twice this value, while the value given by the 

 negatively charged particles seemed more constant and 

 approximated to twice that given by the ions produced by 

 the radium particles. 



This seemed contrary to what one might naturally expect 

 from a consideration of the fact that Townsend hud found 

 that the positive ions produced by Rontgen rays carried 

 double charges under certain conditions but that negative 

 ions from the same source did not. 



Then the work of Richardson and Brown * with hot metals 

 gave- no indications of any ions from this source with double 

 charges. Their experiments, however, did not determine the 

 charge directly and all their measurements were made at low 

 pressures, and this of: course might make some difference. 



Still an effort was made to see if there might not be some 

 cause, such as a change in the rate of diffusion, for the 

 variation in the ratio Q r /Q c other than an actual change in 

 the charge. 



I. It was thought that a variation of the size and position 

 of the strip, causing unequal heating, might account for some 

 part of the variation. However, on investigation this did 

 not seem to give a solution, for the variations were as great 

 when care was taken to keep these as constant as possible as 

 when they were varied widely, and removing the heated 

 strip to a greater distance did not decrease the variations. 

 * Hiclnmlsou & F. C. Brown, he. eit. 



