1893.] Experiments on the Reflection of Light. 79 



constants R and a which are in fair agreement with one another, 

 but as there is some uncertainty as regards their precise values, 

 perhaps a very close agreement cannot be expected between the 

 theoretical and the experimental value of the aiigle of incidence 

 at which the direction of rotation vanishes and changes sign. 



Although Kerr states that the principal incidence of the piece of 

 soft iron which formed the reflector in his experiments was about 

 76°, he does not state the value of the principal azimuth. As 

 so much depends upon these quantities, I should suggest that 

 if any farther experiments are made, the values of the principal 

 incidence and azimuth should be accurately determined for the 

 particular piece of metal experimented upon. It would also be 

 desirable to obtain quantitative results giving the magnitude of 

 the angle of rotation at different angles of incidence, and for light 

 of certain selected refrangibilities. 



5. The formulae which have been discussed in the present 

 paper furnish results which are in such substantial agreement 

 with Kerr's experiments, that I feel strongly persuaded that 

 they are a close approximation to the truth ; but as the complete 

 theory has not yet been discovered, I shall conclude this paper 

 by making some observations upon the difficulties which lie in 

 the way of its realization. It is hopeless to attempt to construct 

 a perfectly satisfactory theory of Kerr's experiments until a satis- 

 factory theory of metallic reflection has been discovered, and 

 from the results of the present paper I incline to the opinion 

 that the difficulty does not consist in explaining the magnetic 

 effect, but in explaining metallic reflection. The formulae in 

 question have been obtained by assuming in the first place that 

 Hall's effect is capable of existing in transparent media, and by 

 the aid of this hypothesis the formulae giving the amplitudes 

 of the reflected waves were rigorously deduced two years ago 

 by means of Maxwell's theory of the electromagnetic field. The 

 next step was to transform these formulae by assuming that 

 the pseudo-refractive index is a complex quantity in the case 

 of a metal ; and the fact that the resulting formulae agree so 

 closely with Kerr's experiments justifies, I think, the conclusion 

 that such an intimate connection exists between Hall's effect 

 and the discoveries of Faraday, Kerr and Kundt upon the action 

 of a magnetic field upon light, that the two classes of phenomena 

 are in great measure due to the same ultimate cause. 



The difficulties, which lie in the way of constructing a satis- 

 factory electromagnetic theory of metallic reflection by taking 

 into account the conductivity of the metal, arise from the fact 

 that the physical conditions imposed by such a theory impera- 

 tively require that a. should be less than 45°; whereas experiment 



