Short-Period Electrometer. 249 



thrown into the circuit. The deflexions caused by reversing 

 the battery were observed, and from these was subtracted 

 the deflexion due to the capacity of the connecting wires 

 leading to the fork. The capacities were taken as propor- 

 tional to the deflexions. The capacity of one-half of the 

 coil A on the other was found to be '000662 microfarad. 



This is of course rather greater than the capacity between 

 two consecutive layers, but the difference was thought to be 

 negligible. Hence, with w=100, 



^^0' = -0000065 microfarad. 



n 2 



This is to be added to the capacity of the ley den- jar in 

 calculating the period. As the capacity of the smallest jar 

 was '001041 the correction is not large. 



In addition the outermost windings and the end layers 

 will have some capacity which will depend upon the position 

 of the coil with respect to other bodies. It was thought that 

 a sufficiently close approximation to the effect of this would 

 be obtained by regarding this capacity as uniformly distri- 

 buted along the wire, and, that being so, it can be shown 

 that if C" is the capacity of the whole coil, the correction in 



calculating the period of oscillation is made by adding ^ C" 



to the capacity of the condenser. In order to measure Q" 

 the two parts of the coil were re-connected and the coil was 

 set up in the position which it occupied during the oscillation 

 experiments. Its capacity was then compared with that of 

 condenser II. by the method just described. ,r was found 

 to be -000026 microfarad. 



The capacity C" was increased by inserting the coil B in 

 A, and was also affected by connecting the terminals of B to 

 one of the paraffin-paper condensers. In the subsequent 

 calculation of periods of oscillation the value appropriate to 

 each particular case was used. 



The capacity of the electrometer, i. e. of the attracting 

 plate with the other earthed, was by the same method found 

 to be -0000138 microfarad. This is to be added to the 

 capacity of the condenser. 



The same method was employed to determine the capacity 

 of the connecting wires used in Maxwell's method, and the 

 capacity of each of the paraffin- paper condensers connected 

 to coil B in some of the oscillation experiments. The latter 

 were compared with a Nalder J-microfarad condenser, and 



Phil. Mag. IS. 6. Vol. 14. No. 80. Aug. 1907. S 



