248 W. A. Norton on Molecular Physics. 
air. As a consequence, a certain portion of the electric fluid on 
the nearerside of b, will be urged away and pass into the grou 
This side will thus. be electrised negatively by induction, and 
react upon the air particles between the two balls, increasing 
their polarization, and upon the nearer parts of a, drawing more 
electricity to them. All these changes will be attended within. 
creasing wave movements, and increasing discharges from one 
particle of air to another along the line between the nearest 
points of the balls. The electric ether in transitu between the air 
particles on this line, may thus come to have sufficient density to 
the two bodies, and so to convey a sudden conductive discharge 
along this line. This result will be partially due to the lateralex: 
pansion, which the free electricity received by the air particleson 
the line will occasion in their electric atmospheres. ere 
be two causes in operation to produce this effect, the pressure of 
the stream of ether passing from one particle to the next, agaist 
the atmosphere upon which it falls, and the mutual repulsion of 
the particles that will thus become momentarily overcharged. 
* We have seen (p. 247) that such lateral expansion would give 
rise to a diminution in the resisting force of the polarized ag 
spark. There is also a sudden diminution in the tension of the 
electricity received by a from the prime conductor, which 1s an 
other cause of this interruption. Apparently another cause esi 
spiring with these two is a reaction to the sudden lateral expal® 
sion above mentioned. If a were previously charged pee 
lated, and 6 brought continually nearer to it, the mutual u rl 
tive action of the two balls upon each other would initiate ¥ 
electric movement above alluded to through the intervening af 
park. | : 
the van 
movements in the atmospheres of the air molecules, meg 
apon the discharge. (See p. 219.) Experiments by wee 
son, and other physicists, have conclusively esta bias 
action of the electrical shell that surrounds it, and the 1 7 
conveyed by the electric current should tend to detach the p#* 
