Fatii^iic and Recovery o-f the Photo-Electnc Current 5 



as electrodes with a glow discharge, a tube, R^ with an addi- 

 tional electrode, P, was attached to the chamber as indicated. 

 For this purpose a second spring brush similar to B served as 

 connection. The arrangement was such as to insulate all elec- 

 trodes except the one in action. All fixed joints which could not 

 be fused were sealed with de Khotinsky's laboratory cement. The 

 chamber proved to be perfectly air-tight. The electrodes could 

 be renewed by removing the base plate, 0, which permitted the 

 plate carrying them to be taken out. 



In the vacuum system all stop-cocks and ground joints were 

 lubricated w' ith a mixture recommended by Travers as free of 

 hydrocarbons. Evacuation was produced by a Rapps automatic 



pump. A ^IcLeod gage, multiplying the pressure about ten 

 thousand times, served as pressure indicator. P.,0.- was used as 

 dryer. A battery of small accumulators served as source of 

 potential. 



For an electric arc that found by Varley to be most satisfac- 

 tory, namely steel electrodes in a slow stream of hydrogen, was 

 used. It proved to be much more constant than other sovirces 

 tried. 



The capacity, C, consisted of an air condenser of tin-foil 

 mounted on plane glass plates insulated with quartz. 



129 



