252 Mrs A. T. Mukerjee on a Method of Measuring 



Mean value of c + c' = 2*264 E.S.U. 

 Probable error of a single observation = ±0*004. 

 c, the gold-leaf system, therefore = 0*78 E.S.U. 

 Probable error of a single observation= ±0*006 E.S.U. 

 A second set of twelve observations by an M.Sc. student 

 gave the following statement : — 



c + c' = 2*28, c' = l*49, 

 c=0*79, with the same probable error. 



Final value of the capacity of the gold-leaf system maybe 

 taken as 0*785 ± 0*006 E.S.U. or for all practical purposes 

 0-78 ± 0*01 E.S.U. 



Determination of the Capacity of the Electroscope by 

 Lester Cooke's Method (Phil. Mag. vol. vi. p. 410, 1903). 



The method used is best described in his own words : 

 " To determine the capacity of the gold-leaves the electro- 

 scope was set on an insulating block of paraffin, and the 

 outside cylinder connected to the quadrants of a very delicate 

 Dolezalek electrometer, which was connected in parallel 

 with a standard capacity of *002 microfarad. The leaves of 

 the electroscope were charged, and the reading observed. 

 The leaves were then discharged, and the charge which bad 

 been attracted to the case of the instrument thus released 

 and allowed to charge up the electrometer and standard 

 capacity. The deflexion of the electrometer was then read, 

 and was a measure of the charge which had been on the 

 leaves of the electroscope."" 



Our experience with paraffin as an insulator in delicate 

 electrostatic measurements, especially when in large blocks, 

 warned us against it in 1913. Hence the outer case of the 

 electroscope was insulated by four very small pieces of 

 freshly cast sulphur and the whole case was completely 

 shielded from electrostatic disturbances by an earthed metal 

 cov r er, leaving a small window for the connecter on the 

 quadrant system to work through. A small mercury cup 

 was mounted on the electroscope case in front of the connecter. 

 The capacity of the case with the mercury cup was deter- 

 mined by a set of measurements with the method of mixtures, • 

 the mean value being 37*7 E.S.U. The gold-leaf system 

 was charged to 85 volts, discharged, and the charge attracted 

 to the insulated case thus set free was shared with the 

 quadrant system of capacity 39*2 E.S.U. The resulting ' 

 potential was measured in the usual wav. 



