Hackett — The Ionic Theories of Magiieto-optic Rotation. 15 



values of the susceptibility. The results stated in (5) and (6) then 

 show that such rotating ions could not produce one-thousandth part 

 of the observed rotation. The tentative assumption that all the 

 magnetic rotation in diamagnetic bodies is due to rotating ions must 

 therefore be dropped. It is well known that the theory did not 

 represent the facts for sodium vapour. But this investigation goes 

 still further ; it is now seen that though the presence of rotating ions 

 does cause magnetic rotation, yet the hypothesis of rotati^ig ions must he 

 abandoned as the cause of the effect in diamagnetic suhstances, since it 

 only explains a negligible 'part of the actual magnetic rotation. 



It is more than probable then that the Hall-effect is the true 

 cause of the magnetic rotation ; and it will now be shown that an 

 analysis similar to the above justifies this view. 



In the second explanation of magnetic rotation, this phenomenon 

 is assigned to the distortion of the paths of the ions in the magnetic 

 field due to the electromagnetic force acting on a charged body at 

 right angles to its direction of motion in the magnetic field. This 

 distortion of the path is expressed in Maxwell's equations, by 

 additional terms in the expression for the dielectric current. These 

 terms give the rotation of plane-polarised light travelling in the 

 direction of the field. The form given below is taken from Drude, 

 expressed in the uniform notation of this paper. The transformation 

 is given in note A. The magnetic rotation is found to be — 



The Hypothesis of the Hall-Effect. 



-vHz 



(7) 



where 



1 + 2 



V 



2 



n = index gf refraction, 



c = velocity of light in cm., 



e = charge on the ion in electro-magnetic units. 



