Measuring the Coefficient of Self-induction. 55 



first the resistance B will have to be altered, as D is generally the 

 resistance of the coil. This necessitates a fresh adjustment of 

 0, and so on. 



The first modification of this method I made was by putting 

 a resistance possessing no self-induction in the arm D in series 

 with the coil, 



fig- 



this 



2, a 



as m .. 

 slider moving over this re- 

 sistance and the condenser 

 being connected as shown. 



An ordinary bridge-balance 

 is first obtained, and the slider 

 is then adjusted until there is 

 no throw on the galvanometer, 

 when its circuit is closed before 

 that of the battery. Then 



L = Kr 2 , 



where r is the resistance be- 

 tween the armatures of the 

 condenser. This method was 

 found to be insensitive, as, unless K were large, r had to be 

 made high. It will not do to make K too large with an 

 ordinary mirror-galvanometer, on account of a sort of double 

 throw being obtained, the effect of self-induction being more 

 rapid than that of the condenser, and the galvanometer not 

 ballistic enough. I then adopted the following modification, 



Fig. 3. 



which answers very well. 



The arm B is a resistance 

 on which two sliders move, 

 one of them being connected 

 to each armature of the con- 

 denser K. 



Obtain a permanent ba- 

 lance ; then AC = BD. 



Now adjust sliders until 

 there is also a balance when 

 galvo. -circuit is closed first. 

 Let x be the current flowing 

 in the arms A and D when 

 it has attained its permanent 

 value, and let y be that in 

 the arms B and C. Let r 

 be the resistance between the sliders when both balances are 

 obtained. Let the battery and galvanometer circuits be both 

 closed, and let the former be broken. The quantity of elec- 

 tricity which passes through the galvanometer due to the 



