ELECTRIC DISTRIBUTION AND CONDENSATION. 69 



strongly electrified. The same test, applied to inter- 

 mediate points, shows the charge on them to be in pro- 

 portion to their distance from the central point. 



ELECTRIFIED CYLINDER. Instead of the two spheres, 

 we may substitute an insulated metal cylinder, with 

 hemispherical ends, provided with pith - ball electro- 

 scopes at the ends and center, as represented in Fig. 14. 



A light charge of electricity G^^ 

 on the cylinder will cause the 

 balls at the ends to diverge in 

 opposite directions, while the 

 central ball will remain un- 

 moved, or but slightly affected ; 

 showing that the principal part 

 of the charge is condensed on 

 the ends, and that induction 

 and repulsion are operating to 

 remove electricity to the points Fig ' ^- 

 farthest from the center, as shown by the position of 

 the balls at the ends. 



If a sphere be made to oscillate near one of the balls, 

 at right angles to the length of the cylinder, the effect 

 of induction will be shown by the ball following the 

 movement of the sphere. 



INFLUENCE OF POINTS. If a cylinder having cone- 

 shaped ends be substituted for the one with hemispher- 

 ical ends, dissipation of the charge, instead of condensa- 

 tion, will occur. For, on the hemispherical ends, the 

 charge is retained by the resistance of the air on the 

 surface ; but the cone-shaped ends terminate in points 

 which have no surface, hence there can be no resistance. 



But if resistance is removed, even from a single point, 

 it is evident that the entire charge must pass off through 



