Ultra-Violet Light from the Mercury Arc. 397 



With quartz interposed in the path of the light, the leak 

 was zero for a potential between — 1"42 volts and — 1'44 

 volts. 



The results of Table I. are plotted in fig, 2. 





















Fig. 



2. 





































cLu, 



X 











































c 



i 









— ^ 



























































/ 





i 





















































ji) 



f 



i 

































trM 



JL 



' J> 























' < 



I 





























Vcl 







' 





t - 



7 - 



» -« 





I - 



j 



i 



i 



» + 





t + 



j + 



*■ 









*-U 





To make sure that B did not charge up positively as a 

 result of positive ions coming up from the arc, the beam of 

 light was passed through an electric field at F. Two brass 

 plates 10 cm. long and '7 cm. apart were maintained at a 



difference of potential of 200 volts. Assuming — = 10 3 for 



a positive ion, the field would stop all ions moving more 

 slowly than about 4 x 10 7 cm./sec. Ions faster than this 

 would get through the field and communicate a charge to the 

 plate B. However, it was found that the leak was entirely 

 stopped when B was in a magnetic field of 200 gauss parallel 

 to the surface. In such a field, positive ions of velocity 

 4xl0 7 cm./sec. would move in a circular path of 50 cms. 

 radius. Consequently, the point of impact of the positive 

 ions on B would be shifted very slightly and the change in 

 the leak would be inappreciable. Hence the charge acquired 

 by B must be due, not to positive ions coming from the arc, 

 but entirely to the loss of electrons under the action of the 

 ultra-violet light. 



