248 Mr. A. T. Mukerjee on a Method of Measuring 



a fixed capacity enclosed under the same conditions as the- 

 electrometer itself, can be measured with a probable error 

 not greater than 0'05 E.S.TJ. When the comparison is made 

 with a condenser placed ouside the case, the probable error 

 is much greater owing to insulation troubles which cannot 

 be avoided, but is still within 0*6 E.S.U. 



In all the experimental arrangements dealt with, the needle 

 of the quadrant electrometer was suspended by a 6/jl quartz 

 fibre and charged to 20 volts only. The sensitiveness was 

 such that when the quadrants were connected to the poles 

 of a standard cadmium cell deflexions between 350 and 

 300 mm. were obtained at about one metre. The average 

 capacity of the quadrant system for needle charges between 



these limits was determined. The variation of the capacity 

 between these limits is not more than one E.S.U., as will 

 appear from the table on page 239 of our previous paper. 

 The form of the electroscope used in the following work is 



