﻿from the Cathode in Helium and Argon. 485 



excess given off at the start. This excess was inappreciable 

 in the case of argon until a current o£ three rnilliamperes 

 was reached, but was very marked in helium for two rnilli- 

 amperes. This suggests the possible cause. For with argon 



Fig. 3. 



O.Sr 



S /o 



Duration of Current (Minutes) 



it was observed that the current-density was at no time more 

 than twice its "normal" value, while in helium it varied 

 from about twice its u normal " value at one milliampere to 

 six times at three rnilliamperes. The excess may therefore 

 arise from the abnormally high current-density. Another 

 respect in which the two gases were remarkably different, 

 namely in their cathode fall, strengthens the belief that the 

 evolution of gas is not merely incidental. For each current 

 used the cathode fall in helium was at least three times that 

 in argon, hence the energy dissipated at the cathode was three 

 times as great. If the evolution of gas be not controlled 

 directly by the current, we should expect it to increase with 

 this energy dissipated, which evidently is not the case. 



Observations with Magnesium Cathodes. 



Experiments similar to the preceding were performed with 

 magnesium cathodes of the same form as the aluminium and 

 subjected to a similar treatment. The results obtained from 

 these are incorporated in Tables IV. and Y., and represented 

 graphically in tigs. 4 and 5. The designations are the same 



