246 MEASUREMENT OF CURRENTS. 



the method gives results which are independent of the intensity of 

 the external magnetic field, or at any rate in which terms of cor- 

 rection only come in, it is particularly suitable for absolute measure- 

 ments. It is then necessary to calculate the coefficients G and S' for 

 the two coils. 



In Weber's apparatus the two coils are concentric, and the 

 movable coil, consisting of a large number of windings, occupies a 

 considerable part of the space left by the fixed coil ; with this mode 

 of construction the calculation of G in particular presents great 

 difficulties. 



In the dynamometer constructed by Latimer Clark* for the 

 British Association, the two coils are still concentric ; but the 

 diameter of the movable coil is small compared with that of the 

 fixed one, and each of them is closed, as was Helmholtz's coil, by / 

 V two equal coils, the distance of which is equal to their mean radius. *^ 

 The calculation of G is then easily deduced from the dimensions of 

 the coils by formula (70) of (793). 



The two coils, being always at right angles to each other in 

 the original condition of equilibrium, which facilitates calculation, 

 the coils can be displaced from each other, so that, for instance, the 

 movable coil occupies a principal position that is to say, that its \^ 

 V centre is in the mean plane, or on the axis of the fixed coil. These 

 arrangements were adopted in Weber's experiments on the elementary 

 action of currents on currents. The formulas in (793) and (794). 

 enable us to calculate the couple due to the reciprocal action of 

 the two coils. 



866. BALANCE ELECTRODYNAMOMETERS. Another mode of 

 using the electrodynamometer consists in weighing, by means of 

 a balance, the attraction or repulsion between circuits traversed by 

 currents. 



We have seen (786) that the reciprocal action exerted between 

 two coils A and A', having the same axis, passes through a maximum 

 value when the distance of the centres is suitably chosen. It is 

 evidently advantageous to take this position of maximum for that of 

 equilibrium of the movable coil, for the action is then virtually 

 constant at least, within certain limits of oscillation and that 

 therefore the sensitiveness of the balance is not modified. 



In strict accuracy, equilibrium would be unstable from the elec- 

 trical point of view ; but the stability which the balance itself gives 

 is sufficient to allow of observations. 



* MAXWELL. Electricity and Magnetism, Vol. II., p. 339. 



