174 Electromagnetic Capacity of a Condenser, 



which give 



P + R -P 



Q + S + dS -Q 



E -(P + Q) P + Q + B | 



x 



A nearly 



A nearly 



t— rfS, 



assuming the value of the determinant A to be but little 

 affected by the change of S. 

 Therefore, finally, 



x T 2 sin 0/2 ,. ' ' 



- = 7T- • — r x correction tor damping 



2-7T tan a r & 



JL'x 



CS 2 (PB + PR + QR + BR)(PQ + GQ + RQ + PG) 

 " A P.rfS ' 



which on expansion of A gives 



cs; 2 



C = tF-- 



T 2 sin 0/2 rfS 



27T tan a 



• -^j x correction for damping. 



In the practice of the method, in order to avoid error 

 arising from the altered value of A due to the change of S, 

 readings of the deflexion, a may be taken with equal positive 

 and negative values of dS. Thus an error which otherwise 

 might be as great as 1 in 100 is reduced to 1 in 10 4 . 



In order to avoid change of the 

 E.M.F of the battery due to its 

 being for any appreciable time on 

 open circuit, it is desirable to use 

 a reversing-key in the battery 

 branch, and to observe the throw 

 of the galvanometer due to re- 

 versal of the sign of the current in 

 S ; and immediately after observe 

 the deflexion due to the alteration 

 of S. 



The method is necessarily subject to the same disadvantages 

 as that of Fleeming Jenkin, in respect of it not giving the 

 instantaneous capacity of the condenser. But measurements 

 of the same condenser taken with different galvanometers, or 

 with a graded moment of inertia of the galvanometer-needle, 

 have led to nearly constant results, the capacity being appa- 

 rent!}' but little dependent on the period of charging and 



