1887.] 



Discharge of Electricity through Gases. 



373 



effects of the ordinary discharges we have at our command, at 

 atmospheric pressure. I took two light balls, and suspended them 

 so that they could be made to diverge by electrification. The 

 electrodes (either spheres or points) of a Voss machine were placed at 

 a distance of 3 inches from each other, and the electrified balls were 

 placed at a distance of 9 inches from the discbarge. The results 

 are contained in the following table, in which the two first 

 columns indicate whether the electrodes of the Voss machine were 

 points or spheres. The third column gives the electrification of the 

 balls, and the fourth column the results. 



Negative electrode. 



Positive electrode.. 



Balls. 



Eesult. 



Sphere 

 Point 

 Sphere 

 Point 



Sphere 



Point 

 >> 



>> 



>> 



Sphere 



Positive 



Negative 



Positive 



Negative 



Positive- 



Negative 



Positive 



Negative 



Balls collapse slowly. 

 „ remain divergent. 

 „ collapse quickly. 

 „ remain divergent. 



>> JJ >> 



„ collapse slowly. 

 „ quickly, 

 remain divergent. 



It will be seen that when the two electrodes are similar, whether 

 spheres or points, the balls collapse when they are electrified posi- 

 tively only ; but that when one electrode is a> sphere and another a 

 point, the balls collapse if their electrification is of the opposite nature 

 to that supplied by the point. 



I shall refer to the theoretical bearing of these experiments at the 

 end of the present paper, but wish at once to point out, that the 

 apparent difference in. the results, for positively and negatively 

 electrified balls, can be one of degree only, and* not one of kind. 

 If the balls are positively electrified, they collapse when the two 

 electrodes are similar ; but in the other case, when, the balls are 

 negatively electrified, an equal and opposite charge will be found on 

 the objects placed in the room, or on the walls. If this positive 

 electricity is neutralised by the discharge, the balls must ultimately 

 collapse in this case, as well as when they were originally positively 

 electrified. To test this argument, I placed the balls in a glass 

 case partially covered on the inside with tinfoil. The inside of the 

 case was nearly a cubical space of sides 37 cm. long. The front of 

 the case was taken out, and the discharge was taken between the two 

 points near the open front. The balls now collapsed, whatever their 

 charge was to begin with, but more quickly when they were positive 

 than when they were negative. 



Confining ourselves to the case of a discharge between points and 



