Electricity through Argon and Helium. 297 



The readings were : — 



262, 270, 265. Mean 266. 



E.M.F. added 4O0 volts. 



The value of the cathode-fa] 1 from the mean of all these 

 experiments is 



226 volts. 



Care was taken in all cases to free the cathode from occluded 

 gases by heating. 



No difficulty was found in keeping the telephone silent in 

 these experiments. 



The cathode-fall for helium is not very different from the 

 value 230 volts found for nitrogen by Warburg and others. 

 This may suggest that my helium was contaminated with 

 nitrogen. But I do not think that there is in reality any 

 ground for suspicion. When traces of nitrogen were present, 

 the cathode-fall in the mixture was very much higher than 

 230 volts ; as soon as the spectroscopic test (which is in this 

 case very delicate) showed the absence of nitrogen, the cathode- 

 fall fell to 226 volts. 



Cathode-Fall in Argon. 



The determination of the cathode-fall in argon presents 

 peculiar difficulties — difficulties which, so far as I know, have 

 not been encountered in the case of any other gas. 



When a wire cathode of the usual form is used, the positive 

 column of the discharge insists on curling round the testing 

 electrode, and going on to the back of the cathode (see fig. 2). 



Fig. 2. 



The positive column is represented by the light shading, the 

 negative glow by the dark. The result is that the whole 

 potential-difference, corresponding to the length AB of the 

 positive column, is added on to the cathode-fall measured 

 between the cathode d and the testing electrode c. The 

 distance to which the extremity B of the positive light 



