332 Mr. W. Grant on the Curves of 



outer circumferences are 2 8' 2 centim. The depth of the sec- 

 tion of the coils perpendicular to the axis is 0'9947 centim. ; 

 the width of the groove parallel to the axis is 1*0 centim. : 

 the section of the coils is therefore approximately one square 

 centimetre. The mean radius is 3*99, or approximately 4*0 cen- 

 tim. The resistances do not agree quite so closely ; but their 

 mean gives 11 = 2*4 ohms at 11°*5 0. 



The coils are mounted on straight guides which have 

 grooves in their upper surfaces, along which the coils can be 

 made to slide ; and their positions are indicated by scales at- 

 tached to the guides. The readings are obtained to centi- 

 metres and millimetres direct, and tenths of a millimetre by 

 estimation. The coils A and B are mounted with their axes 

 coincident on a straight guide E F, of sufficient length to allow 

 their centres to be separated to a distance of 70 centimetres. 

 The coils C and D are mounted on a pair of guides P and 

 M N, P being about half the length of M N, and attached 

 to it at right angles midway between its ends. The axes of 

 the coils are parallel to P and perpendicular to M N : hence 

 C can be moved parallel and D perpendicular to the axes with- 

 out destroying the parallelism. 



The coils A and were connected with the battery L and 

 with the key K to form the primary circuit ; the coils B and 

 D were connected with the galvanometer G, or sometimes 

 with a telephone, to form the secondary circuit. Now it is 

 evident that with this arrangement, in order that there may 

 be no current in the secondary circuit on making and breaking 

 the primary one, the coils C and D must be so placed that 

 their mutual induction is equal to that between A and B ; and, 

 with reference to the secondary circuit, that the electromotive 

 forces induced in the coils B and D must oppose each other. 

 Hence, when A and B were clamped at a fixed distance apart, 

 every pair of positions in which the coils C and D could be 

 placed, so as not to cause a deflection of the galvanometer on 

 making and breaking the primary circuit, were positions in 

 which their mutual induction was constant, and determined 

 one point of a curve of constant induction. Beginning with 

 the coil D at P, the middle point of the guide M N, the coil 

 C was moved until the induction between C and D balanced 

 that between A and B. C was then moved nearer and 

 nearer to P, being shifted a centimetre or so at a time; and 

 the balance was restored in each case by moving D towards 

 M or N. Now, as the arrangement is symmetrical with re- 

 spect to the axis of C, it is evident that when a balance was 

 obtained by displacing the coil D on one side of P, it could 

 also be obtained by moving it to the corresponding position 



