128 DISTRIBUTION OP THE ACTIVE DEPOSIT OF RADIUM 



in this way were calculated from the length of the spark gap, 

 by use of the formula: 



V = 1500 + 30000d 

 where V is the potential difference in volts and d is the 

 length of the spark gap in cms. When the bell-jar was 

 removed the emanation was allowed to escape before the 

 potential was withdrawn. The activities of the plates 

 were then measured in an a-ray electroscope, at from 10 

 to 25 minutes after the escape of the emanation. The 

 procedure was much the same as that described fqr thorium 

 in the previous paper, except that correction was, of course, 

 made for decay during the interval of time between removal 

 of the emanation and measurement of the activity. 



III. Experimental Results. 



While carrying out experiments in air and other gases 

 it was found that a much higher potential-gradient was 

 required to bring over to the cathode a certain percentage 

 of radium rest-atoms, than was needed to bring over the 

 same percentage of thorium rest-atoms. A series of experi- 

 ments was then made in dry air using high potential-gradients. 

 The results obtained are given in Table I. The first column 

 gives the potential-gradient in volts per cm. The second 

 gives the percentage cathode activity, i. e. the percentage 

 of the total activity collected by the cathode. The per- 

 centage of the total positively charged is obtained by sub- 

 tracting from 100 twice the percentage found on the anode, 

 as there is also deposited on the cathode by diffusion an 

 amount of activity equal to that deposited on the anode. 

 The error in the percentage cathode activity, which depends 

 largely on the error made in measuring the anode activity, 

 is probably somewhat less than 1%. The values of the 

 percentage cathode activity given in Table I are each the 

 mean of several values calculated from different observations. 



1 



