ROTATING A DIELECTRIC IN A MAGNETIC FIELD. 



133 



cut into three parts and the capacity of the middle part found in the way just 

 described, using the end parts as guard rings. The electrometer sensibility was 

 0'000150 electrostatic unit per millimetre, and charging the inside coating to 

 0'400 volt gave a deflection of 107 millims. The capacity of the middle part was 

 therefore 12'0 centims. The length between the centres of the two cuts (each 

 inilliin. wide) in the outside coating was 4'95 centims. The capacity of the 

 cylinder per unit length was therefore i? B ' = 2'425 centims. Now the capacity per 



unit length is - - ; hence K = 2 log 7-g/r, X 2'425, or K = 4'85 log * ;j = 3'54. 



2 log r 2 /r, 



Tin- value obtained for K dejrends on the time allowed to elapse between dis- 

 charging the inside coating and reading the electrometer deflection. This is due to 

 the leaking out of residual charge from the elx>nite. In the experiments on the 

 effect due to rotating the cylinder in a magnetic field the deflection of the electro- 

 meter was read as soon as the index came to rest after reversing the field. In 

 measuring the specific inductive capacity, therefore, the same plan was adopted, so 

 that the specific inductive capacity obtained corresponds to a time of discharge equal 

 to the time taken by the electrometer index to move from the zero position to its 

 new position. The electrometer needle was just dead-beat, so that this time was 

 fairly constant and was about 15 seconds. 



Fig. 6. TT, brass tube ; DD, ebonite disk ; CC', coils. 



The variation of the magnetic field along the solenoid was measured by means of 

 two coils of fine wire wound on an ebonite disk which could be slid along the axis of 

 the solenoid. The arrangement is shown in fig. 6. 



DD is the ebonite disk with coils wound on it at C and C". The disk was attached 



