INTRODUCTORY xxvii 



Lord Kelvin) " method, employing a very sensitive Astatic 

 galvanometer and two platinum-silver resistances, arranged 

 so that a difference of one ohm resistance gave me a 

 difference of 001 microfarad capacity. The insulation 

 of the battery and other apparatus was absolutely 

 perfect ; I used a current due to very low electro- 

 motive force, in order to avoid heating, and took all the 

 precautions which are laid down by others and which our 

 own experience suggests. The 10-micro condenser varied 

 in the most inexplicable manner between 8-929 and 9-931 

 micros. In all there might have been a hundred readings 

 taken, each time, or almost each time, with a different 

 result, with a discrepancy of about 0-001 micro, and it was 

 not until I observed a slight galvanometric deflection while 

 the battery circuit was open that the probable cause 

 suggested itself to me. During the course of some experi- 

 ments I afterwards made under different conditions to 

 verify the idea then formed, I stood as closely as possible 

 to the galvanometer circuit, and upon being charged with 

 20 volts produced a slight inverse deflection upon the 

 galvanometer ; when the circuit was opened a slight direct 

 deflection was noticeable. After having connected myself 

 with an earth of low resistance the phenomenon ceased to 

 manifest itself and I succeeded in getting a balance. ' 



My association with Mr. Finlay, short as it was, was 

 fortunate. Had it not been for that association I should, 

 in all probability, have dismissed the vagaries of the 

 galvanometer as being due to leakage, and, so far as I am 

 concerned, the experiments might never have been made. 



Hundreds of other electricians have observed the same 

 phenomena during the last thirty or more years, but have 

 not bothered themselves to do more than attend to the 



