1889.] 



on Optical Torque. 



491 



but nearly so. The bar lies along the line of light from our lantern, 

 but the polariser P (the Ahxens reflector, Fig. 7), and analyser A 

 (the Ahrens triple spar prism, Fig. 6), are crossed, so that here is the 

 (^ark field. On turning on the current, light is at once restored, 

 being twisted to the right when the current circulates right-handedly. 

 To measure the rotation, I must turn the analyser ; and now I find 

 that, owing to the greater rotation of blue waves than of red, complete 

 extinction does not occur. Introducing a half-shadow plate, and 

 using coloured glasses, it is very easy to verify the greater amount of 

 rotation for blue light, and to show that reversing the current reverses 

 the rotation. You will perhaps better understand it if I use (as in 

 Fig. 16) the 24-ray star S, which I have previously employed. It is 

 now obvious to you that there is a large rotation — over 50^ in fact — 

 v\ hich is reversed when I reverse the magnetising current. We have 

 here the fundamental experiment of magneto-optics. But now we meet 

 with another consideratioD. Eeflect that the circulation of current, if 

 it be taken as right-handed when regarded from one end of the coil, 

 will be left-handed when regarded from the other end of the coil. 



Fig. 16. 



I 



'jection of magnetic rotation of light. C, condensing lenses ; P, reflecting polariser ; 

 lagnetising coil surrounding bar of heavy glass ; S, mica star of twenty-four rays ; 

 A, analyser (Ahrens's triple prism). 



his is, therefore, no case of skew symmetry ; it clearly indicates that 

 )mething is going on in the glass which tends to twist the light quite 

 respective of which way the light enters. 

 The next magneto-optic phenomenon is that discovered by Dr. Kerr 



the rotation of the plane of polarisation by reflection at the surface 

 'a magnet. To observe this at all requires good apparatus and a 

 3en eye. So far as I am aware, it has never been projected on the 

 reen. If I can succeed in doing so, it will only be because I have 

 lecial means of the most favourable character for so doing. We 

 ithdraw the bar of heavy glass from the coil, and replace it 

 'ig. 17) by an iron core polished at its coned end. This will be 

 : tensely magnetised when the current is turned on. 



Now we must throw the beam of light obliquely down the hollow 

 < the coil, polarising it by one of my improved Nicol prisms P, as 

 Vol. XII. (No. 83.) 2 l 



