Theory and Measurement of Residual Charges. 



441 



could be detected having a greater value than the possible 

 experimental error, which was about one in a hundred. It 

 was also thought that there might be a lower limit to the 

 electrostatic pressure capable o£ producing a residual charge, 



Fis. 2. 





—^ 



























— ^4 





S 























■Uj 



























•02 

 'CI 





































VOLTy 



J 

















i 



2 



3 



4- 



5 



6 



7 



8 



9 



IC 



10 





































































































































but no such limit could be found, and it seems, therefore, 

 that the smallest electric force if applied long enough to the 

 dielectric will produce some displacement of the free electrons. 

 Though the curve in fig. 2 is, no doubt, an exponential curve, 

 it approximates very closely to an equilateral hyperbola, the 

 various readings, of which a few are indicated in the figure, 

 falling into position with surprising accuracy. This fact, in 

 itself, might be an indication that the residual charge is not 

 mainly due to any haphazard space variation of K and r, 

 but to something quite simple in its nature, and wholly dis- 

 tinct from the ordinary mechanical properties of the dielectric, 

 and to something, too, produced by the impressed electric 

 force. 



The curve approximates closely to that whose equation is 



y(« + 10) = -05*, 



y being the ratio q/Q, and x the pressure in volts applied to 

 the condenser. Thus we have, roughly, 



? '05V 

 Q-V + 10' 



V being the difference of potential of the two plates in volts. 



