49: 



Professor SUvanus P. Thompson 



[May 17, 



it goes down. After reflection it is focussed by a lens which sends it 

 throucyh the analysing prism A. You see the dim spot of reflected 



Fig. 17. 



Apparatus for projecting rotation of plane of polarisation by reflection at pole of 

 magnet. P, polariser ; M, magnetising coil with coned iron core ; A, analyser. 



light upon the screen. Xow for the current : " on," " off," " on," 

 " off," Reversing its direction ought to double the aniount of torsion. 

 Whilst Mr. Thomas is making the needful arrangements for the 

 next experiment, I may mention that it was found by Kerr that the 

 effect was approximately proportional to the magnetic induction 

 through the iron. I have myself tried some further experiments : for 

 example, using a bar of lodestone instead of an iron core. The light 

 reflected from lodestone is also twisted. I should expect the ferro- 

 aluminium alloy which Sir H. Roscoe showed us a fortnight ago to do 

 the same thing, because that alloy is, as I have found, susceptible of 

 macrnetisation. But I should not expect manganese steel to rotate the 

 light, because of its singularly non-magnetisable nature. 



The experiment of Kundt, transmitting polarised light through a 

 thin transparent film of iron, magnetised normally whilst the light is 

 passing through it, is another difficult of repetition before an 

 audience. The small disks here are covered with films of iron, kindly 

 prepared for me by ^Ir. Crookes, by squirting them electrically in a 

 hif^h vacuum. But the thin ones barely transmit enough light to 

 make the observation of the effect possible, even to the solitary 

 observer. I have observed the efiVct projected on the screen, using 

 this very coil and these transparent mirrors. It requires, however, 

 an absolutely dark room, and is at best so faint that it would be 

 hopeless to attempt to show it to a large audience. Professor Kundt 

 has not only observed similar rotations in other magnetic films of nickel 

 and cobalt but has even shown that the degree of rotation of the light 

 is proportional not to the magnetising force, but to the resulting mag- 

 netic induction. This is a result of utmost importance in considering 

 the theorv of the phenomenon. He has further shown that whereas the 

 mafynetic rotations in elementary bodies, whether magnetic or dia- 

 macmetic, are in the same sense as that in which the current circulates, 

 the ma^metic rotations in compound magnetic bodies, such as a 

 solution of sulphate of iron in water, are in the opposite sense. 



