Short-Period Electrometer. 247 



The capacity of the condenser (including the connecting 

 wires) was calculated from the equation 



which was sufficiently accurate for the purpose, the omitted 

 factor (depending on the battery and galvanometer resistance) 

 only causing a difference of about 1 in 8000. 



One of the four condensers measured (No. III.) showed a 

 much greater residual charge than the other three, and in 

 this case the capacity was found to depend largely upon 

 previous charging. After having been kept rather highly 

 charged for a few minutes, the capacity of this jar was less 

 by about 6 per cent, than before. Its coats were therefore 

 always kept connected for two or three days before a 

 determination of its capacity. The same precaution was 

 taken with the three other condensers in which, however, the 

 effect of previous charging on the capacity was much smaller. 

 The capacity of condenser III. would probably depend con- 

 siderably upon the frequency of the alternations of charge, and 

 as this was about 48 per second in the measurements by 

 Maxwell's method and about 400 per second in the oscillations, 

 no very close agreement was to be expected between the 

 observed and calculated values of the period. 



The correction for the capacity of the connecting wires was 

 found by a method of comparison described below r *. 



When the connexions to the jar are interchanged so that 

 the outer coat is connected to the platinum points and the 

 inner to the point C of the bridge, a value for the capacity 

 greater by one or two p^r cent, is found. It can be shown 

 that in such cases the capacity upon which the period of 

 electrical oscillation depends is that of the inner plate. 



(6) The Inductances of the ColU. 



The self-inductance of the coil A was determined by 

 Rayleigh's method. The galvanometer used was a suspended 

 coil instrument of period 11*07 seconds and logarithmic 

 decrement "0220. The swings and deflexions were taken 

 alternately, and the process for allowing for previous swing 

 and change of zero, as described by Gray f , strictly followed. 



* The capacity of an air-leydeu, by Kelvin and James White, was 

 also measured. Some leakage was found to take place at the glass 

 supports of the inner plate, and when this was prevented the capacity 

 was found to be (corrected for the connecting wires) '002444 microfarad. 



+ A. Gray, ' Absolute Measurements in Electricity and Magnetism,' 

 vol. ii. part ii. p. 483. 



