Disappearance of Gas in (he Electric Discharge. 589 



difference lies in the manner in which the chemical changes 

 are brought about. If this view is accepted, the study of 

 the thermal actions may well provide the clue to the 

 explanation of the electrical actions. 



3. It is the purpose of the experiments described in this 

 paper to extend to the electrical actions the investigation 

 which Langmuir lias made of the thermal actions. There 

 appears to be much evidence for believing that there is an 

 electrical action which is as independent of the thermal 

 action as the thermal action is of the electrical. In studying 

 1 his action the incandescent cathode is a great experimental 

 convenience, because it permits the electrical discharge to 

 take place under potential differences too low to cause the 

 discharge to pass between cold electrodes in the same gas ; 

 it gives much greater control over the electrical conditions. 

 Further, it is the discharge with an incandescent cathode 

 that is of primary technical importance. For these reasons 

 all the experiments have been made with such a cathode. 

 But it must be remembered that the conditions sought are 

 those in which the incandescent cathode has no direct action 

 on the gas causing it to disappear, but acts only by abolishing 

 the cathode fall of potential and allowing the discharge to 

 pass under small potential differences. If these conditions 

 can be obtained, they will be much nearer to those of the 

 older experiments, in which the discharge was passed between 

 cold electrodes by high-potential differences, than to those of 

 most of Langmuir's experiments, where the high temperature 

 of the filament is the main agent in determining the dis- 

 appearance of the gas. 



Apparatus. 



4. The apparatus employed is shown diagrammatically in 

 fig. 1. L is the vessel in which the discharge took place ; it 

 is drawn approximately to scale and will be termed for 

 brevity the lamp. It is connected through the liquid-air 

 trap T, which served to exclude vapours, (1) to the McLeod 

 gauge, (2) through the mercury cut-off C : to the pump, 

 (3) to the gas-filling apparatus. The pump was a mercury 

 diffusion pump backed by a rotary oil-pump, and gave 

 pressures too low to measure. The gas was introduced fr< m 

 a generator or reservoir through the porous plug P, which 

 could be covered by a mercury column ; it then passed into 

 the liquid-air trap T 2 to remove condensable gases, and could 

 be admitted to the lamp by lowering the mercury cut-oft' C 2 . 

 The quantity admitted could be regulated roughly by the 

 time for which the porous plug was uncovered. 



