80 ME. A. E. OXLEY ON THE INFLUENCE OF MOLECULAR 



field. In the light of the experimental results on aromatic substances there described 

 and of those given by DU Bois, HONDA and OWEN, which are concerned with 

 elementary substances, this suggestion has been found to lead to interesting results. 

 It is clear that we can easily test whether such an intense local molecular field exists ; 

 for if so it must be of import in connection with the transition from the liquid to the 

 crystalline state in general, and the changes of other physical properties which 

 accompany this change of state must also depend directly or indirectly upon the 

 molecular field. The object of the present communication is to test the experimental 

 and theoretical work which has already been completed by applying the results 

 obtained to a wider range of physical phenomena. Evidence of a change of specific 

 diamagnetic susceptibility owing to crystallization has been obtained with about forty 

 substances, but with some substances the change, if it exists, is exceedingly small. 

 Ten additional substances have been investigated, and the results, in conjunction with 

 the other physical properties, are used to extend the ideas concerning the molecular 

 field to diamagnetic crystalline media in general. 



PART III. 



(2) ON THE LOCAL MOLECULAR POLARIZATION AND THE LOCAL MOLECULAR 

 FORCES WITHIN DIAMAGNETIC FLUID AND CRYSTALLINE MEDIA. 



The passage from a purely mechanical theory of the properties of material media to 

 the actual discrete molecular theory, as interpreted electrodynamically, involves an 

 examination of the local polarizations which act in the immediate neighbourhood of 

 the point considered within the medium. It has been shown by Sir JOSEPH LARMOR* 

 that the forces exerted at such a point by the surrounding polarized medium can be 

 separated into two parts, a purely local part and a part due to the rest of the 

 medium. If we neglect the forces due to the former then those due to the latter are 

 derivable from a potential due to the combined volume and surface density 

 distributions of POISSON. With fluid media we can go further and determine, at least 

 approximately, the forces due to the purely local part. This is attained by imagining 

 a cavity scooped out round the point and determining the force due to the surface 

 distribution of polarization over its walls. The removed molecules are now put back 

 in the cavity and the additional force which they contribute is found by applying the 

 method of averages. The molecules of a fluid are in rapid motion and the force 

 at a point in the interior of the cavity due to these replaced molecules averages out 

 to a small corrective polarization. It can be shown that the force due to the 

 immediately neighbouring molecules is represented by (i+s) P where P is the 



* 'Roy. Soc. Phil. Trans.,' A, vol. 190, p. 233. 1897; and LORENTZ, ' Theory of Electrons,' pp. 137 

 and 303. 



