504 CONSTANTS OF COILS. 



or, if /3 is the ratio , 

 a 



Q i+/?' 



This, however, would in general be only a very insufficient 

 approximation, owing to the difficulty of measuring the swings, 

 especially with moVable frames, as the times of rotation should be 

 very small in comparison with the time of oscillation of the bal- 

 listic galvanometer. 



1093. By means of the differential galvanometer, the observa- 

 tion can be brought to that of a state of equilibrium. The coils 

 B and B' communicate separately with the two coils of a differential 

 galvanometer, and the total resistances R and R' of the two circuits 

 are regulated by trial, so that the needle is stationary when induction 

 is produced, either by the rotation of coils or by the inversion of an 

 inducing circuit. As the quantities of induced electricity m and m' 

 are then equal, it follows that the flows of induction are proportional 

 to the corresponding resistances (515), and we have 



It is sufficient then to determine the ratio of the resistances as 

 soon as equilibrium is obtained, so as to avoid the influence of 

 change of temperature. 



This method would require an absolutely symmetrical differential 

 galvanometer; but if the defect of adjustment is very slight, it is 

 sufficiently eliminated provided the connection of the coils with 

 the frames can be interchanged. The resistances which re-establish 

 equilibrium being R x and R' I} we have sensibly 



Q AR R; 



A series of coils of graduated surfaces would enable us to act by 

 opposition. The coils B and B' being placed in the same circuit in 

 such a manner that their flows of induction are in opposite directions, 

 the auxiliary coils of increasing surfaces are successively interposed 

 until the deflection is zero or changes its sign. If Q x and Q a are 



