1889.] 



on Optical Torque. 



495 



glass, or still better sodium liglit, to observe the changes in form on 

 reversing the sign of the charges. Figs. 20 and 21, 22 and 23, show 

 the changes of form, but these sketches grossly exaggerate the effects. 

 As you see upon the screen, when the charges imparted by this fine 

 Wimshurst machine are rapidly reversed, there is a decided distortion 



of the rings, but it is small in amount. 



Fig. 20. 



Fig. 21. 



Fig. 22. 



Eeturning to the phenomena of the rotation impressed by magnet- 

 ism on polarised light, I may point out that the torque which a 

 magnetic field exerts on the light waves appears to be really an action 

 upon the matter through which the light- waves are passing. It is as 

 though the magnetic field were really a portion of space rotating 

 rapidly on itself, or perhaps as though the ether were there rotating, 

 and that this rotation in some way dragged the particles of matter 

 along with it. It has long been supposed necessary, in order to 

 account for the refractive and dispersive properties of transparent 

 bodies, to consider that their particles are in some way concerned in 

 and partake of the vibrations going on in the ether within them or 

 between their molecules. It is impossible to explain the phenomena 

 of magneto-optic rotation by the supposition that any skew structure 

 is imparted to the medium ; for these phenomena unlike those of 

 quartz, do not exhibit skew symmetry. There seems to be no other 

 way of explaining the magneto-optic torsion of light than by suppos- 

 ing that the molecules of matter in the magnetic field are actually 

 subjected to rotatory actions ; as indeed was suggested long ago by 

 Sir William Thomson. 



However, there is room here not only for speculation but for ex- 

 periment. Some day, when facts enough have been collected, we shall 



