Force of the Electric Discharge through Gases. 451 



being open, while the discharge passed between the elec- 

 trodes, the effect of the current of gas across the electrodes 

 produced in this way was observed for all pressures of the 

 gas in the discharge-tube : of course the blast through the 

 discharge diminished with the pressure of the gas and finally 

 vanished. 



When air or carbonic acid gas was used, the deflexion pro- 

 duced in this way was at first nothing, but gradually increased, 

 attained a maximum, and then died away again as the vessel 

 A became exhausted. In hydrogen the deflexion produced 

 was always very small, and the maximum deflexion was pro- 

 duced after exhaustion had gone much further than in the 

 case of air or carbonic acid gas, and consequently when the 

 blast was weaker. 



The velocity of the blast in air through the discharge-tube 

 produced by this pumping method was about 2 centim. per 

 second, when the pressure of the gas was about 14 centim. of 

 mercury, and the discharge was bent as shown in fig. 21, 

 though hardly so much. 



The velocity of the blast in hydrogen, for about the same 

 pressure of the gas, was 3 centim. per second, but the 

 deflexion produced in the path of the discharge was hardly 

 perceptible ; blasts of smaller velocities for different pressures 

 were also used, in addition to the strong blasts obtained by 

 the first method, i. e. by exhausting vessel B, and then opening 

 the stop-cock k until the pressures in A and B became equal 

 by a flow of gas through the discharge-tube, but in no case 

 was the discharge in hydrogen much deflected. 



Fig. 20. Fig. 21. 



Fig. 20 shows the deflexion in hydrogen, about the maxi- 

 mum observed. Fig. 21 shows the deflexion in air, the 

 actual deflexion was even greater in some cases. 



When the blast through the discharge-tube was very 

 strong, the deflexion in either air or hydrogen diminished. 



