Plotting these values it is found that R = 4*57, I (less 

 copper leak) = 635. Hence, RI, corrected for want of satu- 

 ration =298. 

 Hence — 



Total ionisation for C._,H, 298 , .^^ 



Total ionisation for air 231 



Also stopping power — 



^ 483 ^ 760 ^ 3105^ -^.-^^ 

 ~ 457 763 2930 ~ 



In the paper by Mr. Kleeman and myself, to which T 

 have already referred, it was pointed out that Rutherford 

 had found it more easy to obtain the saturation current from 

 a gas when it was removed from the influence of the ionising 

 agent. We observed that this could be easily explained by 

 supposing initial recombination to be completed before the 

 gas was subjected to the electric field. Nevertheless, as I 

 now see, it is otherwise no essential feature of the initial re- 

 combination hypothesis that the act of recombination should 

 take place within any set time. The one important point is 

 that the recombination takes place between two ions origi- 

 nally forming parts of one molecule. It is quite conceiv- 

 able that for a certain time the positive and negative may 

 remain "semi-detached," their recombination in suspense un- 

 til precipitated by some change of conditions. It is curious 

 that Mr. Madsen, working in this laboratory, has not been 

 able to confirm Professor Rutherford's experiment, and his 

 results point to a prolonged existence of these pairs. He 

 finds it hard to saturate a mixture of air and ether vapour 

 which has been ionised by radium and then drawn away into 

 a separate ionisation chamber. It is not easy to reconcile this 

 result with that of Professor Rutherford : and it will be 

 necessary to repeat the experiment under varying conditions. 



