ON THE ACTION OF MAGNETISM ON LIGHT. 357 



ie 



•where [p] and ? ^ represent vectorial coefficients. For the simpl 

 harmonic oscillations of period r that are here contemplated 



so that for the very small periods of light-vibrations multiplication of a 

 coefficient by djdt increases its importance enormously. When the 

 oscillations are very slow the coefficient [p] has still in a magnetic field a 

 rotational part which reveals itself as the Hall effect ; the presence of a 



•coefficient hz ;5^ could hardly be detected. On the other hand, with 



greater i-apidity of vibrations, the importance of the rotational part in 



2 j^ increases steadily, and finally absolutely overshadows any possible 



effect of the \_-p'\ terms, unless the latter should contain a part whose 



origin was of the form ^ J72 • ^^ ^^^^ were so, at a still higher rapidity 



of vibrations the [p] terms would again become the important ones : but 

 the wave-lengths would then be too small for such vibrations to have any 

 physical reality. 



17. The question occurs whether to secure complete generality a 

 corresponding rotational quality should be imparted to the linear relation 

 connecting the magnetic flux (i.e., magnetic induction) with the magnetic 

 force. It is, however, usual to assume, on various grounds, that the 

 vibrations of light are too rapid to allow of their being accompanied by 

 an oscillating magnetisation of the material medium. The phenomena of 

 magnetisation of iron leave possibly no room for doubt that the magnetic 

 movement is an afiair of loosely associated groups of molecules, not of 

 individual molecules themselves, the free periods corresponding to these 

 groups being much too slow to follow the light-vibrations. These groups 

 are broken up at the temperature of recalescence without the occurrence 

 of any very striking effisct : nor is there any striking difference in kind 

 between the behaviour of iron to light and the behaviour of non-magnetic 

 metals. 



The effect of strong magnetisation on light- waves would be on this 

 view a secondary effect due to a change of structure of the medium. 

 Soon after the experimental discovery of the Hall effect, and the attention 

 which was concentrated on it owing chiefly to the influence of Lord 

 Kelvin, it was pointed out by J. Hopkinson that the existence of an 

 «ffect of that character had been anticipated by Maxwell in his ' Treatise,' 

 vol. i. § 303, where in discussing the possibility of the occurrence of a 

 rotational term in the equations expressing the genei'al form of Ohm's 

 law of conduction, he remarks that such a coefficient ' we have reason to 

 believe does not exist in any known substance. It should be found if 

 anywhere in magnets, which have a polarisation in one direction, probably 

 due to a rotational phenomenon in the substance.' The theory of such 

 rotatory coefficients had also been worked out long before by Lord Kelvin,* 

 in a thermo-electric connexion. 



18. It seems worth while to examine how much in the way of magnetic 



' Cf. Lord Kelvin (Sir W. Thomson), Collected Papers, vol. ii. 



