EXPERIMENTS ON ELECTRICAL DISTRIBUTION. 591 



a point ; in one case the point disperses the electricity 

 with which the body is charged originally, in the other 

 it disperses one of the two electricities which are 

 separated by induction in the neutral body; in either 

 case points, whether attached to an electrified body, or 

 only near it, diminish the original charge, or render it 

 impossible or difficult to charge it properly. 



A few precautions are necessary for success in these experiments. 

 The metallic conductor must be perfectly round and smooth; a 

 rough surface and projecting corners act like points. The lower 

 edges of the vertical opening of the conductor are therefore rounded 

 off by being turned inwards, so as to avoid sharp edges, which would 

 disperse electricity towards the support and the table. The glass 

 rod which carries the conductor must be well rubbed down with a 

 dry cloth, if it should not insulate properly. 



For proving the equality of distribution on a sphere, its in- 

 equality when the sphere is connected with the wire, and the absence 

 of electricity in the interior, it is quite sufficient to draw the rubbed 

 glass once along the conductor ; a stronger charge would cause so 

 great a divergence of the gold leaves that it could not be further 

 increased, and the difference between the quantity of electricity 

 distributed over the sphere and that distributed over the ring could 

 not be demonstrated by the difference in the divergence when elec- 

 tricity from both parts of the conductor is transferred to the electro- 

 scope by the proof- plane. Too strong a charge might also cause the 

 upper edge of the vertical opening of the conductor to communicate 

 electricity to the proof-plane while it is being withdrawn from the 

 interior, and this charge might then be mistaken for one carried 

 away from the interior of the conductor. 



For the experiments on the action of points, the glass rod should 

 be drawn over the conductor 5 or 10 times, having each time been 

 rubbed again. The spark given by the conductor on bringing the 

 finger near should be l cm long. 



The experiments which prove that electricity diffuses itself only 

 over the surface of a conductor, may also be performed by means of a 

 conical gauze bag, the opening being formed by a wire ring, fig. 314. 

 The ring, which is 10 cm in diameter, is formed of brass- wire, 2 or 3 mm 

 thick, and has a handle 10 cm long, the end of which is heated and 

 fused into a stick of sealing-wax so as to insulate it. During the 

 experiment the sealing-wax is clamped in the retort-stand, the foot 



