334 



Prof. A. Anderson 



on 



The following method, which is but a slight modification of 

 Maxwell's, does away with much of its impracticability. 



In Maxwell's method the armatures of the condenser are 

 connected to A and B (fig. 3), and S is the coil whose coeffi- 

 cient of self-induction is to be found. The only difference in 



F,g. 3. 

 B 



the method now given is that an adjustable resistance r is 

 placed between B and N. Suppose a balance has been ob- 

 tained so that no permanent current flows through the galva- 

 nometer. This balance will not be disturbed by changing r. 

 Let x be the whole quantity of electricity which has passed 

 through AB at the time t, z the whole quantity which has 

 passed through AM, and y that through AD. Then, if K be 

 the capacity of the condenser, and if we suppose the potential 

 of N to be always equal to that of D, we have 





dx 

 dt 



z dz 





z 

 K 



-■& dy 





dz 



~ fdx dz\ c dv , T d s y 



and, 



therefore, 



0-+Q) 



dz Q/ z dz\ S z 

 dt + F\K + ''dt) ~RK + 



Hence we have the conditions 



and 





QR=PS, 



-un+Q'* L 

 ' + Q+ T r= RK- 



Jifk 

 BKdt' 



