Lord Kelvin on Magnetism and Molecular Rotation. 239 



differences. These differences will be different for different 

 molecules, because of the different values of 6 for their different 

 orientations. Hence each bright line is not split into two 

 sharp lines, but is broadened to an extreme breadth corre- 

 sponding to the value = 0. No simplifying suppositions as 

 to the character of the molecule, such as symmetry of forces 

 and moments of inertia round the axis of the ring, cart 

 possibly give Zeeman's normal results of the splitting of a 

 bright line into two sharp lines circularly polarized in opposite 

 directions, when the light is viewed from a direction parallel 

 to the lines of magnetic force; and the dividing of each bright 

 line into three, each plane-polarized, when the light is viewed 

 from a direction perpendicular to the lines of force. Hence, 

 although from 1856 till quite lately I felt satisfied in knowing 

 that it sufficed to explain Faraday's magneto-optic discovery, 

 I now, in the light of Zeeman's recent discovery, discard my 

 old tempting gyrostatic hypothesis for an irrefragable reason, 

 which is virtually the same as that stated by Larmor* in the 

 following words : — " Hence a principal oscillator magnetically 

 tx*ipled mast be capable of being excited with reference to any 

 axis in the molecule ; otherwise there would be merely hazy 

 broadening or duplication instead of definite triplication." 



§ 8. It now seems to me that the theory of H. A. Lorentz 

 (of Leyden), as expressed by equations (1) in Zeeman's first 

 paper " On the Influence of Magnetism on the Nature of the 

 Light emitted by a Substance " |, is essentially true. 



§ 9. Though it cannot explain Zeeman's discovery, the 

 molecular rotation caused by the institution of a magnetic 

 field, which is the subject of the present communication, 

 may, however, be considered as interesting not only because 

 the idea of it seems to be new in electromagnetic theory, but 

 also because it may conceivably constitute the explanation of 

 Faraday's diamagnetism. Gro back to §§ 2, 3 above, and 

 remark that if a body containing a vast number of the molecules 

 there described is situated between the poles of a steel magnet, 

 the total energy will be greater than if there were nothing but 

 ether between the poles, by a difference equal to the kinetic 

 energy of the motion of the resinuously electrified fluid. 

 Hence if a body containing the supposed congregation of 

 molecules is movable, it must be repelled from the place of 

 strong magnetic force between the poles to places of weaker 

 force further from them. 



* Phil. Mag. vol. xliv. 1897, " On the Theory of the Magnetic Influence 

 on Spectra," p. 507. 

 f Phil. Mag. vol. xliii. (1897) p. 226. 



