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



on the external insulators. Taking the mean of all these 

 four sets, the final value for the capacity of the gold-leaf 

 svstem by this method may be taken as 0*75 ± *015 E.S.TJ. 

 at about 100 volts. 



Capacity of the Electroscope by Barratfs Method. 



The method depends on obtaining the relation between the 

 divergence of the leaf and the potential as applied to it, and 

 the observation of the fall of potential of a standard condenser 

 when it shares its charge a convenient number of times with 

 the electroscope. If V is the original potential of the con- 

 denser and C its capacity, v n the final potential of the system 

 after n alternate earthings and re-chargings, and c' the 

 required capacity of the electroscope, then 



■-M>P- 



V and v n being read off the calibration curve of the electro- 

 scope after observing the corresponding deflexions. 



To compare the accuracy of this method with the previous 

 ones, the probable errors may be discussed. 



If E c ', E e , E v , and E V)! be the probable errors of the corre- 

 sponding quantities, 



(E,) 





i 



m 





It will appear from Barratt's paper that, as he used a 

 travelling microscope with vernier reading only to -^ mm. 

 to read the deflexions of the gold-leaf, E y or E^ could not 

 have been less than 0'8 volt*. E c may be taken as 1: then, 

 taking the values given by him in Table III (a) of his paper, 

 it will be seen that 



(E c ) 2 = (0-021) 2 + (0-039) 2 4- (0'04D) 2 

 = (0-0G4) 2 . 



The probable error of the method as used by him may 

 therefore be taken as +0*06. 



* The ordinates on his calibration curve for the electroscope should 

 probably be cm. instead of mm. In the latter case the probable error 

 would be 8 volts, which seems unusual. 



