Recovery of Residual Charge in Electric Condensers. 579 



explanation is on somewhat the same general lines as and 

 leads to analogous results to that which has been advanced in 

 the case of recovery from elastic overstrain, and which is 

 based on the assumption o£ the material consisting in part of 

 viscous and in part of purely elastic constitute nts. This theory 

 of elastic overstrain leads to an exponential expression for the 

 recovery from the strain in terms of the time. Now no such 

 expression fits experimental evidence, but on the other hand 

 the recovery can be well represented by the logarithm of a 

 quantity proportional to the time of recovery. Thus in 

 Rankine's experiments where a constant strain is maintained, 

 the stress while decreasing is given by 



S=S (l-Klog( P t+l», 



where S represents the stress at any moment required to 

 preserve a constant strain in the stretched substance. 



Our work was undertaken in the first place to examine 

 whether the exponential law of the recovery of the Residual 

 Charge was justified by experimental data, and failing this, 

 to attempt to find a law of the above type agreeing with 

 observation. 



The examination of the rate of recovery was carried out 

 in two ways. 



Where the Residual Charges were great enough the current 

 as it came out of the recuperating dielectric was simply 

 passed through a sensitive galvanometer and the rate of 

 recovery thus noted. 



This method was found quite feasible in the case of a 

 large condenser, the dielectric of which was celluloid, but 

 was unsuited for mica condensers of the ordinary standard 

 type. 



For such, a method was devised in which the plates of the 

 condenser were allowed to charge up to a certain difference 

 of potential (due to the Residual Charge coining out of the 

 dielectric) which was measured by an electrometer, and kept 

 at that difference of potential by inserting a variable resistance 

 in parallel with the condenser. 



The difficulty of getting a resistance as high as that re- 

 quired, and one which at the same time could be easily adjusted, 

 was surmounted by adopting ionized air as the material of 

 which the resistance was made. The cross section of this 

 resistance was arranged so as to be easily and quickly varied 

 as alteration in the resistance was required. 



2R 2 



