TRANSVEESE EFFECT AND ON SOME RELATED ACTIONS IN BISMUTH. 237 



To find how this effect varied with the primary current strength, the field was kept 

 constant, while the primary was varied : — 



Primary Current 

 Proportional to. 



Effect in Scale Parts. 



Effect. 

 Primary. 



60-0 

 124-1 

 187-0 

 373-0 



90-0 

 177-6 

 269-4 

 515-4 



1-5 

 1-43 

 1-44 

 1-39 



or the effect is, for the currents used, practically directly proportional to the primary 

 current. 



This effect had the same direction in Plate II., for which the following results were 

 obtained : — 



Field, . 

 Effect, . 



36-0 

 



•61-0 

 0-0045 



120-0 

 0-0509 



137-0 

 0-07175 



146-0 

 0-084 



In Plate XII. again the direction of the effect was the opposite. In the other plates 

 it had sometimes the one direction, sometimes the other ; indeed, after a plate had been 

 planed or hammered, it sometimes had the opposite direction to that in the original plate. 

 So long as it was less in magnitude than the transverse effect proper, its disturbing 

 influence was eliminated by making four different experiments, according as the direction 

 of the field or the primary current was varied. Should it, however, have a greater 

 value than the transverse current, this was no longer the case ; and when the transverse 

 current vanished, it alone was observable ; it increased in all cases with the field strength, 

 and in no case did it change sign, unless the plate was modified. 



The existence of an effect whose direction can in no case be predicted follows from 

 the general equations for a non-isotropic body. For, suppose we have an isotropic body 

 which is brought into a magnetic field and carries a current, it is acted on by mechanical 

 forces. The body becomes anisotropic, and the transverse coefficients of resistance are 

 brought into play ; a transverse current flows, whose direction is determined by the 

 structure of the body. Or, suppose the body to be originally non-isotropic, a transverse 

 current will be observed with no magnetic field present ; this, however, can be eliminated 

 by the insertion of a proper resistance. When the magnet is excited, the transverse resist- 

 ance is modified, so that the inserted resistance no longer balances it. Result is, the 

 transverse current again appears. The fact that it does not depend on the direction of 

 the field shows that the resistance concerned is proportional to some even function of the 

 magnetisation. 



