Ox THE ABSOLUTE UNIT OF ELECTRICAL KESISTANCE 159 



6. Variation of the resistance of German silver with the temperature. 



7. Temperature of standard resistance. 



8. Arc of swing when the time of vibration is determined. 



9. Length of needle in tangent and other galvanometer (nearly com- 

 pensated by the method). 



10. The variation of resistance of circuit during the experiment. 

 The following errors are compensated by the method of experiment. 



1. The local and daily variation of the earth's magnetism. 



2. The variation of the magnetism of the needle. 



3. The magnetic and inductive action of the parts of the apparatus 

 on each other. 



4. The correction for length of needle in the tangent galvanometer 

 (nearly). 



5. The axial displacement of the wires in the coils for induction. 



6. The error due to not having the coils of the galvanometer and the 

 circle parallel to the needle. 



7. Scale error (partly). 



8. The zero error of galvanometers. 



CALCULATION OF CONSTANTS 



Circle. For obtaining the ratio of G to G", it is best to calculate 

 them separately and then take their ratio, though it might be found 

 by Maxwell's method ('Electricity,' article 753). But as the ratio is 

 great, the heating of the resistances would produce error in this latter 

 method. 



For the simple circle, 



where A is its radius and B the distance of the plane of the circle to 

 the needle on its axis. 



Galvanometer for Induction Current. For the more sensitive galvano- 

 meter, we must first assume some form which will produce a nearly 

 uniform field in its interior, without impairing its sensitiveness. If we 

 make the galvanometer of two circular coils of rectangular section 

 whose depth is to its width as 108 to 100, and whose centres of sections 

 are at a radius apart from each other, we shall have Maxwell's modifi- 

 cation of Helmholtz's arrangement. The constant can then be found 

 by calculation or comparison with another coil. 



