Electricity by Radium and Thorium Emanations. 465 



get as good a vacuum as possible in the apparatus, was 

 sealed off at D, in order to prevent any variation in the 

 amount of emanation in the bulb owing to diffusion, and 

 absorption by the charcoal. The apparatus was then left for 

 some days, during which the positive leak gradually increased 

 as the emanation in the bulb accumulated. The negative 

 leak also increased slightly, and this was probably in part an 

 ionization leak, due to the evolution of small quantities of 

 gas from the radium; the maximum reading obtained for the 

 negative leak was, however, 7*5, the maximum positive leak 

 beino- 65, so that the greater part of the latter must still be 

 ascribed to the slowly-moving negative rays given off by the 

 emanation. 



The radium tube was then sealed off, and the rate of dimi- 

 nution of the positive leak was measured over a considerable 

 time. No such rapid fall was observed as had been previously 

 obtained ; but the rate of leak fell off approximately according 

 to an exponential law, going to half value in four davs 

 (fig. 3). 



Fi<r. 3. 



I 1 



\ 































ft: 

 f 



felO 



\ 



i 



K . 































s 



































^^v 





x 















































































— , 



Time 



\l. 13 



It is thus evident that the radium emanation gives off a 

 quantity of slowly-moving negative rays, along with the 

 a particle, during its change to the excited activity. 



The action of thorium was similarly examined. A small 

 bulb half full of thorium hydroxide was sealed onto the 

 bottom of the testing vessel, the apparatus exhausted, and the 

 leak determined as before. The mean values obtained were, 

 with a positive charge 11*5 divisions per hour, with a negative 

 charge 2*1 divisions per hour. Here again, therefore, the 

 same slowly-moving negative rays are evidently present. 



Later on, the experiment was repeated with a larger quantity 

 Phil. Mag. S. 6. Vol. 10. No. 58. Oct. 1905. 2 K 



