CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY. 
145 
In Part II., the extension of Langevin’s theory shows that an exceedingly small and 
reasonable deformation of the molecules in the liquid state is sufficient, owing to the 
large value of the constant a c , to account for the change of susceptibility actually 
observed when the substance crystallizes. 
Moreover, the introduction of the idea of a molecular field of the diamagnetic 
crystals allows us to neglect the mutual influences of the diamagnetic molecules in 
exactly the same way as the molecular field of Weiss allows us to express the 
phenomena of ferro-magnetism, without bringing into consideration the mutual 
influences of the molecules. Both the molecular field of Weiss and the molecular 
field of the diamagnetic crystals are proportional to the intensity of magnetization, in 
the respective cases, and the minimum value of the constant of the latter field, 
deduced from the variation of magnetic rotatory power when crystallization sets in, is 
of the same order as that of the ferro-magnetic field. 
Although on account of a particular structure of the molecules the magnetic 
properties of different substances, observed externally, may vary enormously, yet, 
within the substances, the local intensities of magnetization and the forces between 
the molecules will be large in all crystalline substances. These forces determine the 
rigidity of the crystalline structures. 
These investigations have been made in the Cavendish Laboratory and I wish to 
thank Prof. Sir Joseph Thomson for placing the necessary apparatus at my disposal 
and for the interest he has shown in the work. 
I am indebted to Prof. Pope for supplying me with many substances for investi¬ 
gation and also for a valuable discussion on the subject of crystal structure. 
k/(p 5 )-k/(pj 
OX Cl) 
dt, (which 
[It will have been observed that the theoretical treatment given in Part II. consists 
e 2 A 
of an interpretation of the polarisation term, —— 
^ 4 m 
represents that part of the diamagnetic moment due to the distortion of a molecule 
by the influences of neighbouring molecules of the crystalline structure) by means of a 
magnetic field. This field is superposed upon the applied field and it has been shown 
that we may disregard the mutual influences of the molecules forming the crystalline 
structure if the resultant of these fields is substituted for the applied field. 
This corresponds to the interpretation, given by M. Weiss, of the mutual influences 
of the molecules of ferro-magnetic substances by a uniform magnetic field, Nl, super¬ 
posed upon the applied field. 
Since the work described in Parts I. and II. was communicated to the Royal 
Society, a paper has appeared by Weiss (‘Comptes Rendus,’ vol. 157, No. 25, p. 1405, 
December 22, 1913) which deals more fully with the nature of the molecular field in 
ferro-magnetic substances. It is there shown that the large forces which act between 
the molecules are not necessarily magnetic but may be of any other nature. A 
VOL. CCXIV.-A. U 
