146 MR A. E. OXLEY ON THE INFLUENCE OF MOLECULAR CONSTITUTION, ETC. 
consideration of the continuity of magnetic flux has led Weiss to abandon the idea 
that the magnetic field is actually a uniform magnetic field. The difficulty may be 
surmounted by assuming the molecular field to be localized. This localized nature is 
a necessary property of the molecular field in diamagnetic crystalline substances. 
But, as Weiss points out, it is rather difficult to see how the molecular magnets, or 
the electron orbits, in adjacent molecules, can approach one another so closely that, 
in spite of thermal agitation, the local field between the molecules should amount to 
the enormous value of 10 7 gauss. 
In any case, whatever may be the nature of the local forces which hold the 
molecules of a crystalline structure together, we may interpret the accompanying 
distortion by means of a magnetic field of such magnitude that it produces in the 
molecules the same change of moment as the actual forces within the crystalline 
structure will produce. This statement holds whether the substance be ferro¬ 
magnetic or diamagnetic. 
The large value of this local magnetic field in diamagnetic substances has been 
briefly mentioned at the end of Part II. It can be shown that a local field, 
comparable in intensity with the ferro-magnetic field of Weiss, will produce, accord¬ 
ing to Langevin’s theory, a modification of the electron orbits in the molecules of a 
diamagnetic substance such as would account for the change of x observed on 
crystallization in the experiments of Part I. In a subsequent paper I hope to 
describe in detail several extensions of the work contained in Parts I. and II., and I 
should like to make a brief reference here to one or two points because they furnish 
additional support to the conclusions which have already been mentioned. Further 
experiments have been made, similar to those described in Part I., and the change of 
susceptibility on crystallization has been correlated with the magnetic double 
refraction of the same substance observed by MM. Cotton and Mouton. The large 
double refraction of a crystalline structure, as compared with that of a liquid 
subjected to the strongest magnetic field which we can produce in the laboratory, is 
compatible with the relatively large local magnetic field by which we may express 
the forces which hold the molecules in position in the crystal. 
Perhaps these large local fields between the molecules of a crystalline structure 
may be identified with the intense intra-atomic fields of Ritz (‘ Ann. der Phys.,’ IV., 
vol. 25, p. 660, 1908) which give us a direct interpretation of Balmer’s and 
Rydberg's expressions for spectral series, or with the magneton of Weiss. This 
would imply that the molecules in the crystal structure are exceedingly close 
together and that their oscillations are of extremely small amplitude. It is possible 
that the magneton will prove to be a constituent of diamagnetic as well as of para¬ 
magnetic and ferro-magnetic matter .—Note added March , 1914d\ 
PRESENTED 
12 fcWtf. 1314 
