1887.] Discharge of Electricity through Gases, 377 



doubt in each case as to the existence of a current. The whole effect 

 was smaller on that day for reasons which could be traced. 



1 Leclanche gave a current of 0*0010 micro-ampere. 

 5/18 „ „ 0-0011 



1/6 „ „ 0-0003 



The main current in these last experiments was 0'008 ampere. 



An electromotive force of one-sixth Leclanche is about one- quarter 

 of a volt, and a current has thus been obtained in a gas from an 

 electromotive force which could not maintain a current through 

 water. 



An electromotive force of 0*1 volt gave doubtful results, but this 

 was probably due to the experimental difficulty of detecting the 

 current. 



In some previous experiments, which, however, were not quite free 

 from objection on other grounds, the lowest electromotive force for 

 which the currents could be measured was 0*2 volt. 



The experimental arrangement which is the best for the qualitative 

 investigation of the effect is not the best for quantitative measure- 

 ments, and I have therefore not endeavoured to follow out to any 

 great extent the quantitative laws of these currents produced by low 

 electromotive forces. I may give, however, some facts which I have 

 observed. The intensity of the current depends on a great many 

 circumstances. 



1. It increases rapidly with the intensity of the main discharge, and 

 also with a reduction of pressure, as far as I have tried it (that is 

 about J- mm.). 



2. The intensity of the current from the auxiliary battery increases 

 less rapidly than the electromotive force. 



3. In some experiments in which one of the electrodes of the 

 auxiliary battery was a copper wire and the other a copper cylinder, 

 the current was nearly always considerably stronger when the larger 

 surface was the kathode. 



4. Anything that facilitates the diffusion of gas from the main 

 current to the auxiliary electrodes will increase the strength of the 

 current observed. In some experiments, in which the screen sepa- 

 rating the two fields was made of wire gauze instead of tinfoil, the 

 currents were stronger than those given above. 



5. In the arrangement shown in fig. 2 the currents were stronger 

 when the main electrode A was negative than when it was positive. 



Considerable care has to be taken, especially when no screen is used 

 or when it is not conducted to earth, in order to avoid leakage 

 currents. However well the battery may originally have been insu- 

 lated, the insulation always grows worse with time (owing to dust and 

 moisture). If, then, any part of the auxiliary circuit itself is not pro- 



