236 



MR J. C. BEATTIE ON THE 



the transverse effect. From this we may draw three conclusions : — (l) The two assump- 

 tions are both wrong ; (2) one is wrong ; (3) or we may still suppose both true, and 

 assume that in bismuth two constants with opposite signs are involved in the transverse 

 effect. That is, instead of assuming that it is proportional to the vector product of the 

 primary current and the magnetisation, we assume that it is the vector product of the 

 primary current and (cj + c 2 I 3 ). 



In the first case we may write the electromotive force 



where Cj is negative for bismuth and those metals which have a negative transverse 

 effect ; positive for those which have a positive effect. 

 In the second case 



e = Yu(c 1 l + c 2 l s ) 



where c x is the same constant as before, c 2 is another constant positive in the first class of 

 substances negative in the second. In those substances in which the transverse effect is 

 proportional to the magnetisation, c 2 is infinitesimally small in comparison with c x ; in 

 bismuth and any other substances where this is not the case, c 2 has such a value that for 

 sufficiently high fields the transverse effect may vanish, and for still higher reverse its 

 direction. Similarly, c 2 might be of such magnitude that the transverse effect did not 

 vanish, but still reached a maximum value, and then began to decrease as in Plate Ia, 

 fig.l. 



The validity of this assumption could be tested by determining the magnetisation 

 directly, and thus determining c x and c 2 for different field strengths. 



Section IV. — On Effects other than the Transverse Effect proper. 



Two other such effects were observed. The first was evident in the whole of the 

 plates experimented upon. In the plate of pure bismuth, III., it was such, that when 

 the apparatus was arranged, as in diagram (D), in passing from the entrance of the 

 primary current at B to that of the effect at D, the motion was counter-clockwise. It 

 changed with the change in direction of the primary current, but not with the 

 reversal of the magnet. Thus, with one arrangement of the magnet, it acted against the 

 transverse current ; in the other with it. In Plate III. it acted against the transverse 

 current when the north pole of the magnet was in front of the diagram, with it when the 

 south pole was in front. 



The following results were obtained with Plate III. : — 



Field, 

 Effect, 



55-0 

 003613 



100-0 

 0-08908 



138-2 

 0-14475 



153 

 0-16332 



