Force oj the Electric Discharge through Gases. 44^ 



The spark- discharge of the induction-coil is not visibly ben^ 

 by a blast. The blowing-out of the discharge was frequently 

 observed during the experiments on the deflexion by mag- 

 netic force (p. 433 above). 



An interesting case, showing the similarity between thP 

 action of magnetic force and of convection-currents in thP 

 gas, in deflecting the discharge, is shown in figs. 17 and 18. 



Fur. 17. 



The large discharge-tube was filled with air, and exhausted 

 until the coil was just able to send the discharge — the directic* 11 

 of the discharge was horizontal — when it was found th, a t 

 instead of remaining permanently deflected, the path of tl le 

 discharge rose and fell ; the first discharge was near jj 

 straight, the second was more bent, and so on for four or fi^ e 

 discharges, as shown in fig. 18, the discharge then fell to tl 3e 

 straight line as if it had been blown out, and was blown asi(^ e 

 again, the convection-currents in the gas being too strong to 

 allow the discharge to preserve a permanent path. 



Fig. 18. 



+ 



In fig. 17 the pressure of the gas was reduced until the dis- 

 charge passed freely, and maintained a permanent path deflected 

 by convection-currents, as b, fig. 17, only turned upward^ 

 the convection -currents being always vertical. 



Now when the magnetic field was made so as to bend t ne 

 discharge upwards still more, it presented the appearan ce 

 shown in fig. 17, a. The discharge rises and falls as befoi e ? 

 the increase in the blast across the electrodes due to the mag" 

 netic field being sufficient to blow out the ionized pat n * 

 When the magnetic force deflected the discharge downward' 



against the convection-currents, the discharge passed steadily? 

 as shown at b } fig. 17. 



