Mr. Louis Schwendler on Differential Galvanometers, 269 



Farther, we have ^ = 11—, where b is the length of an ave- 

 rage convolution, and X the absolute conductivity of the wire 

 material, supposed to be a constant for the coil. 



Now, for brevity's sake, we will suppose that S, the cross 

 section of the insulating covering, can be neglected against q 

 the metallic cross section of the wire. 



Consequently we have 



A 



— :=U (approximately) 



and 



or 



similarly, 



But using wire of the same conductivity in both the differential 

 coils, which should be as high as is possible to procure it, and 

 further supposing the manner of coiling to be identical in both 

 coils, we have 



A. — A, , 



. »'_ /A[ b 



" n V A ' b'' 



Further, we know that if the shape and dimensions of each 

 coil are given, and in addition also their distance from the mag- 

 net acted upon, it will always be possible to calculate m and ?n', 

 though it may often present mathematical difficulties, especially 

 if the forms of the two coils differ from each other and are also 

 not circular. This latter condition is generally necessitated in 

 order to obtain the greatest absolute magnetic action of each 

 coil in as small a space as possible. 



However, it is clear that we may assume generally that the 

 two coils have each an average convolution of identical shape 

 and of the same length, placed at an equal distance from the 



