318 Prof. A. Kundt on the Electromagnetic Rotation 



actually falling upon the metallic surface exactly at right 

 angles. 



Mirrors of iron, cobalt, and nickel were used for these ex- 

 periments, prepared by electrolytic deposition of the metal 

 upon platinized glass. It has already been remarked that in 

 consequence of the small thickness of these plates, the forces 

 exerted upon them by the magnets are so small that no dis- 

 tortion of the glass plates which support the metal films can 

 occur. 



Iron. 

 Four mirrors were employed. Rotation. 



No. 1. Iron film so thick as to be com- 1 KO/ a 



pletely opaque J 



No. 2. Transparent — 45'*4 



No. 3. Transparent -l°5 /! 8 



No. 4. Transparent -48 / '8 



The rotation is opposite to the direction of the magnetizing 

 current ; consequently the above numbers, like the following 

 ones for cobalt and nickel, are marked with the negative sign. 

 It must be mentioned that a dispersion due to rotation was 

 distinctly observed in reflection also, which, like that pro- 

 duced by the transmission of light, is an anomalous disper- 

 sion — i. e. the red rays are rotated more powerfully than the 

 blue ones. It is remarkable that mirror No. 3, which was 

 very transparent, shows so large a rotation. I have noticed 

 several times that other mirrors, also of smaller thickness, 

 have shown a comparatively large rotation. I have not yet 

 been able to make more accurate experiments on the connexion 

 of the rotation with the thickness of the reflecting layer. 



Cobalt. 

 Three cobalt mirrors were examined. 



No. 1. Transparent, gave a rotation of . —1° 7''2 



No. 2. Hardly transparent, „ — - 50'" 9 



No. 3. Opaque, „ - 50'-3 



Here also a feeble dispersion due to rotation could be observed. 



Nickel. 

 Of two mirrors which were examined, 



No. 1. Transparent, gave a rotation of . — 20 /, 5 



No. 2. Opaque „ — 23'-0 



Thus nickel, upon reflection, rotates only about half as 



powerfully as iron and cobalt. Moreover, as was shown 



above, the rotation of transmitted light produced by nickel is 



perceptibly smaller than that due to iron and cobalt. 



