128 DR. HAROLD A. WILSON ON THE ELECTRIC EFFECT OF 



small enough to make the electrometer needle " dead-beat " when at the maximum 

 sensibility for quantity. The damping of the needle was done by the air inside the 

 quadrants. 



One pair of the quadrants was always connected to the case and to earth, while 

 the other pair was connected, by a wire about 2 feet long, to the brush which made 

 contact with the outside coating of the ebonite cylinder. This wire was completely 

 enclosed in metal tubes, which, together with the electrometer case, the metal bobbin 

 of the solenoid and the inner coating of the ebonite cylinder, formed a complete 

 metallic screen, all connected together and to earth, surrounding the insulated parts 

 of the apparatus. Outside electric disturbances had no effect whatever on the 

 electrometer needle. For example, reversing the current through the solenoid was 

 repeatedly proved to produce no deflection of the electrometer needle when the 

 cylinder was not rotating. Also running the motor had no effect on the electro- 

 meter. 



The wire WW (fig. 4) leading from the brush to the electrometer was supported 

 by a sealing-wax rod and by an arrangement which was used to determine the 

 electrometer sensibility for quantity. This consisted o? a small circular parallel plate 

 guard-ring condenser, ZZ. The plates of this condenser were held 2 '00 centims. 

 apart by three ebonite pillars, and the lower plate and guard ring were each 13 - centims. 

 in diameter. The hole in the guard ring was 3 '05 centims. in diameter and the 

 condenser plate 2'95 centims. in diameter. The plates were all made of brass 

 0'5 centim. thick, and were turned up truly plane. The condenser plate was sup- 

 ported by a disk of ebonite screwed down on the guard ring, and it carried a rod, on 

 the top of which was a mercury cup, Y (fig. 4). The wire WW was supported 

 by this rod. A brass rod could be let down into the mercury cup when it was 

 desired to connect the insulated parts of the apparatus to the metallic screen 

 surrounding them. This rod was soldered to a wire spring, the other end of which 

 was soldered to the tube containing the wire WW. 



The larger plate of the condenser was connected with a commutator, by means of 

 which it could be either connected to earth or charged up by means of a battery of 

 small secondary cells. 



To determine the sensibility of the electrometer for quantity the insulated parts 

 were first earthed and the larger plate of the condenser charged to a potential 

 of V volts. A quantity of electricity , q(g = C V/300 electrostatic units, where C is the 

 capacity of the guard-ring condenser), was thus induced on the small condenser plate. 

 The rod at Y was then raised and the electrometer scale reading noted. The larger 

 plate of the condenser was then earthed, so setting free the quantity of electricity q 

 on the insulated parts of the apparatus. The resulting deflection was then read oft' 

 and was taken to be the deflection due to the quantity of electricity </. It was 

 verified carefully that the deflection was proportional to q. 



The sensibility of the electrometer for potential difference was measured when 



