452 On the Deflexion of the Electric Discharge through Gases. 



and the discharge became unsteady, as shown in figs. 17 and 

 18. The blast required to produce this effect for a given 

 pressure of the gas was much less in hydrogen than in air. 



From these experiments it is quite clear that the ionized 

 path produced by a discharge is dissipated much more 

 rapidly in hydrogen than in air or carbonic acid gas, other- 

 wise the path of the discharge should be bent as the ionized 

 gas is blown aside. 



Conclusion. 



The experiments described in this paper show that the 

 deflexion of the electric discharge by magnetic force, and the 

 deflexion of the discharge by a blast across the electrodes, are 

 affected by the gas through which the discharge passes in 

 exactly the same way ; and that the properties of a gas on 

 which the magnitude of the deflexion of the discharge in it 

 depends are as stated at length on pages 443 and 444: — 



(1) The rate of diffusion of the gas. 



(2) The rate of recombination of the gas ionized by a dis- 

 charge. 



(3) The velocity of the ions in the gas under an electro- 



motive force. 



(4) The duration of the discharge. 



(5) The relative resistance of the gas to the passage of 



electricity through it before, and after, a discharge 

 has passed. 



Also for a given gas, the greater the frequency of the dis- 

 charge the more readily is the discharge deflected, the 

 deflexion vanishing for very slow rates of discharge. 



For discharges following one another with very great 

 rapidity we should still expect to get the magnitude of the 

 deflexion produced by magnetic force, or by a blast, different 

 in different gases unless (5) above is the same in all gases. 



In conclusion I desire to acknowledge my indebtedness to 

 Professor Thomson for the valuable assistance his suggestions 

 have afforded me. 



Cavendish Laboratory, 

 20th June, 1898. 



