of the Plane of Polarization of Light. 323 



Another steel mirror, similar to the one used for the series 

 of experiments, gave at an angle of incidence of 65°'3, 



= -L 



-8'-6 +9'-0 



This mirror was then electroplated with silver. Upon estab- 

 lishing or reversing the current there was no perceptible 

 brightening of the field. 



The pole-pieces of the magnet were then put into contact, 

 and the mirror which was to be examined was placed at the 

 side of the magnet, but without touching it. For this pur- 

 pose mirrors of non-magnetic metal and platinized glass 

 were employed which had been electrolytically coated with 

 a thin film of iron. An accurate examination of the same by 

 means of the scale and telescope showed not the smallest dis- 

 placement or distortion when the magnet was excited. A 

 rotation was observed in every case, as was to be expected 

 from the results given in the above tables. Thus, for ex- 

 ample, a glass mirror coated with iron gave for an angle of 

 incidence of 65°, when the iron-coated side was about 2 millim. 

 from the side of the magnet, the following rotations : — 



= -L. 



-8'-l + 13'-0 



A rotation could still be observed when the mirror was at a 

 distance of 45 millim. from the side surface of the magnet. 



I have made also a number of experiments on the rota- 

 tion produced upon oblique incidence on the end faces of a 

 magnet. The arrangement was the same as that employed 

 by Kerr. Platinized glass coated with iron, cobalt, or nickel 

 was used in every case. The mirror was fastened in front of 

 the plane-pole of an electromagnet without touching it, and a 

 pointed pole (submagnet of Kerr) brought near to the mirror. 

 If the plane of polarization coincided with the plane of inci- 

 dences, then for all incidences the direction of rotation is 

 opposed to that of the magnetizing current if we imagine the 

 reflected ray of light rotated in the line joining the poles. 



If the plane of polarization is at right angles to the plane 

 of incidence, then at an incidence of about 80° the rotation 

 is also oppositely, but from 80° to 90° similarly, directed with 

 the magnetizing current. This reversal in the direction of 

 rotation, as already remarked, escaped Kerr's notice. Since 

 I have not yet been able to make a complete series of obser- 

 vations with one and the same mirror on the variation of 

 rotation with the angle of incidence, I defer giving numbers 

 for the present. 



