president's address — SECTION A. 45 



stance is simply one in which E is greater than in air. 

 When the substance is brouo-ht from A to B the lines from 

 the magnet avoid the substance, as it were,. for displacements 

 in it are difficult of production. A little tap will help the 

 displacements to be made: the displacements become more 

 nearly what they would be if the substance were not there, 

 and so the diamagnetic effect is lessened. Consequently, on 

 the withdrawals of the bismuth from the magnet the force 

 of repulsion is less than on its approach, and there is no 

 violation of the principle of conservation of energy. 



It is interesting to notice that if there is any retentiveness 

 in bismuth the effect of a magnetising coil will be to produce 

 magnetism, so to speak, not diamagnetism. The bismutli 

 will be magnetised in the same way in which a piece of iron 

 would be under similar circumstances, only, of course, to a 

 fai- less degree. For, thinking of the ])roblem in the strain 

 analogy, the particles of the bismuth are not displaced in a 

 direction opposite to that of tiie displacing force ; they are 

 displaced in the same direction, but less than those of the 

 majority of substances would be. 



I have described this strain analogy somewhat at length : 

 it is not the only one of its class, but \ will not say more than 

 a word or two about its fellow members. 



If we replace the displacement of translation of the theory 

 just discussed by displacement of rotation, we have a new 

 analogue of electric and magnetic theories. Here the strain is 

 one of torsion : we must suppose the particles of the medium to 

 be twisted from their natural set, and forces of i-estitation to 

 be called into play proportional to the amount of twist. The 

 proportion between strain and stress will be different in 

 different parts of the medium. There will be lines and tubes 

 of twist, and along any tube the amount of twist is handed 

 on unchanged. This theory represents the phenomena of 

 electricity and magnetism just as well as the displacement 

 theory, but it is more difficult to realise. 



Again, we may rejdace displacement by " spin," and obtain 

 a kinetic analogue : amount of displacement must be replaced 

 by rate of spin, force of restitution per unit volume by 

 momentum per unit volume, elasticity per unit volume by 

 moment of inertia per unit volume. Energy per unit volume 

 will, of course, be ^ momentum x rate of spin. On this 

 analogy the particles of iron possess much less inertia than 

 those of other substances. 



There will be lines and tubes of spin ; and along any tube 

 the amount of spin will be handed on unchanged. 



