MAGNETISM AND LIGHT 335 



to account for the effect by supposing, not that the molecules alter 

 their distances apart, but that they are oriented with axes of some 

 kind along the lines of force. They consider that Voigt's theory will 

 not account for the double refraction which they have discovered, 

 for his theory should apply to all bodies, whereas their double 

 refraction is so far confined to a particular class. 



The Kerr magnetic effect. Reflection from magnets. 

 The reflection of light at a surface is not merely a return of the waves 

 at an impenetrable surface, for the nature of the reflection depends 

 on the constitution of the medium within the surface. For instance, 

 the polarising angle differs with different reflecting media, showing 

 that the medium near the surface takes part in the process. Con- 

 sidering this, and further considering that, as the Faraday effect in 

 some salts of iron is already great, it might be expected to be 

 vastly greater in the steel of a magnet, Dr. Kerr* was led to 

 expect some kind of rotation in a plane-polarised ray when 

 reflected from the pole of a magnet. 



To test his expectation he polished a flat as perfect as 

 possible on one pole of an electro-magnet and then brought the 

 other pole near to it, just leaving between them a chink through 

 which a ray of plane-polarised light could be directed on to the 

 surface at an angle of incidence between 60 and 80. When the 

 plane of polarisation was either parallel to or perpendicular to the 

 plane of incidence and the current was off, the plane was unaltered 

 by reflection. These two planes were selected because, in accordance 

 with the ordinary result of reflection from a metal, the light in any 

 other plane would after reflection be elliptic-ally polarised. After 

 reflection the ray passed through a Nicol to the observer's eye. 

 The Nicol was turned to extinction. On putting the current on, 

 the light reappeared faintly, the plane of polarisation always 

 turning in the direction opposite to that of the magnetising 

 current, and the polarisation was slightly elliptical, as the principle 

 of reversibility would lead us to expect. 



In further experiments Kerr worked with normal incidence 

 through a perforation in the opposing pole and found again a 

 rotation, though less in value, and again in the opposite direction 

 to the magnetising current. According to the later work of 

 Kundt, the light entering the metal and finally absorbed by it 

 would have its vibration turned round in the direction of the 

 magnetising current so that we might expect the reflected vibration 

 to be turned in the opposite direction. 



Afterwards Kerr obtained similar effects with reflection at 

 surfaces worked on the equatorial parts of a magnet. He summed 

 up his results in the statement that a new small component was 

 introduced in the reflected ray with vibration perpendicular to that 

 of the ray reflected before the iron was magnetised. The vibration 



* P/til. Mag. [5], vol. iii. p. 321 (1877), for reflection from a pole ; vol. v. p. 161 

 (1878), for reflection from the side. 



