Weights of Radium and Thorium Emanations. 73 



starting an experiment and subtracted from the leaks sub- 

 sequently obtained. 



The vessels A and B were made as nearly as possible 

 alike, so that the ionization currents through them were not 

 Aery different when they contained equal quantities of emana- 

 tion. Before proceeding to the diffusion experiments, how- 

 ever, a comparison of the leaks produced in A and B by a 

 given ionizing agent was necessary. This was made by 

 detaching the vessel V and replacing it by a glass tube so 

 that no emanation could escape during its passage from A to 

 B. A current of emanation was then allowed to pass until 

 the leaks through the two vessels had become constant, when 

 the series of observations given in Table VI. were taken. 

 The numbers show considerable variations owing, probably, 

 to irregularities in the air-current flowing. The error pro- 

 duced by these fluctuations is, however, greatly reduced by 

 taking the mean of several observations extending over a 

 long time. It will be noticed that for the same ionizing 

 agent the leak in A is greater than the leak in B, in the ratio 

 100 : 96*2. In all subsequent experiments, therefore, the 

 observed leak in B was increased in this ratio (see Table X.). 



Table VI. — Comparison of the Two Leak Vessels. 

 The Current of Emanation was started at 11.11 A.M. 





Leak for 30 seconds, in scale-divisions 







of the electrometer. 





Time. 









Vessel A. 



Mean. 



Vessel B. 



Mean. 



12.27 p.m. 



169 









12.30 „ 







159 





12.33 „ 



167 









12.36 „ 







173 





12.:;v 



177 





... 





12.42 „ 







168 





12.45 „ 



165 









12.46 „ 





171 



164 



104-5 



12.51 ., 



168 









12.54 „ 







164 





12.57 „ 



175 









1.00 „ 







165 





1.03 „ 



175 









106 „ 







158 





1.09 „ 



170 









1.12 „ 







165 





