632 Miss J. M. W. Slater on the 



Therefore 1 div. per min. electroscope corresponds to about 

 88 per min. electrometer. On another occasion, 1 div. per min. 

 electroscope corresponded to 68 per min. electrometer (cor- 

 rected to uranium leak = 27*75); therefore mean value of 1 div. 

 per min. electroscope = 78 per min. electrometer. 



On January 27th the disk was exposed to the discharge 

 for J hour, and on removal from the tube the activity was 

 found to have fallen to 85 per cent, of its original value 

 (allowing for the regular decay) ; C now gave an electro- 

 meter leak of 429 div. per min. (uranium leak = 20*6). The 

 15 per cent, of activity lost should therefore give an electro- 

 meter leak of 75*8 per min. (Ur = 20"6), or 102 per min. 

 (Ur = 27*75) ; this corresponds to about 1*3 per min. electro- 

 scope leak. The observed rate of fall of the gold-leaf in the 

 tube was '17 div. per min. before the exposure of disk to 

 the rays, *92 per min. after ; the leak is therefore increased 

 by *75 div, per min. A large proportion of the activity lost 

 by the disk is thus evidently present on the walls of the 

 tube, and in view of the rough nature of the experiment 

 it is extremely probable that all the activity lost is volatilized 

 unchanged. Again, on January 29th, disk A was exposed 

 for one hour to the discharge, and lost 52 per cent, of its 

 activity. The ratio of the electroscope leak with A in the 

 tube to that without A, after exposure, was found to be 2*3 ; 

 the greater part of the activity lost is thus accounted for. 

 The walls of the tube retained their activity for several hours, 

 and drawing a current of air through did not diminish it. 



It is therefore evident that the active matter is volatilized, 

 not destroyed, during the passage of the discharge ; and later 

 experiments made with a rather different arrangement, by 

 means of which the activity driven off could be separately 

 examined, confirmed this conclusion. 



Cause of Loss of Activity. 



It was not easy to determine the immediate cause of the 

 effect observed, as during the cathode-ray discharge a number 

 of secondary actions are possible. With the ordinary ar- 

 rangement of the tube, the disk being at D, C cathode and 

 A anode, a loss of activity was always obtained, whether the 

 disk was insulated, earthed, or connected to the anode. The 

 discharge was sometimes obtained from an induction-coil and 

 sometimes from a large Wimshurst machine, the latter being 

 much steadier in its action ; the results, however, were very 

 similar. The loss of activity varied in amount from 20 per 

 cent, to 60 per cent, for exposures of ^ hour to 1 hour; but 

 the variation seemed to depend more on the air-pressure in 



