CONSTITUTION AND TEMPEEATURE ON MAGNETIC SUSCEPTIBILITY. 99 



and MOTJTON and myself. From the table it is clear that there is some connection 

 between these two effects. In each case the benzene nucleus plays an important role. 

 According to COTTON and MOUTON all liquids whose molecules contain such nuclei have 

 a magnetic double refraction which is easy to detect experimentally. These liquids 

 also show a large value of BX on crystallization. If the nucleus be modified by 

 substitution of a monovalent atom or group for one of the hydrogen atoms the 

 magnetic double refraction may increase or diminish but always retains the same order 

 of magnitude. The parallelism in this respect is well represented by the benzene 

 derivatives investigated in Part I. All these show a value of 2x/X f the same order 

 except nitrobenzene. In this substance the substituent -NOX, which is so active in 

 augmenting the induced double refraction of the liquid, also contributes an abnormally 

 large part to BX as would be expected. 



Both effects appear to depend also upon the degree of unsaturation of the substance, 

 and if by any process we can lessen the number of double linkages associated with 

 the nuclear carbon atoms there is a sudden diminution both of double refraction and 

 3x- Hexamethylene is particularly interesting from this point of view, for here all six 

 double linkages are destroyed by the additional six hydrogen atoms. In their later 

 experiments COTTON and MOUTON have found that most organic liquids and some 

 inorganic ones show a feeble double refraction in the stongest magnetic fields. These 

 would be accounted for if the molecules possess a slight degree of dissymmetry, for then 

 they would become partially orientated by the external field. As regards the change 

 of x with such substances on crystallization, we should expect that the large molecular 

 field which comes into play would produce a polarization or distortion in any molecule 

 characterized by the molecule's own dissymmetry, while the distorting force would 

 depend upon the degree of unsaturation of the molecule. Although the extent of this 

 distortion is small, yet the forces which produce it must be large, for even with the 

 most unsaturated and unsymmetrical molecules 3x amounts to a few per cent, only 

 when the molecular field is of the order 10 7 gauss. 



With more delicate means of detection and a more intense magnetic field, COTTON 

 and MOUTON conclude that a double refraction would be shown by all substances. 

 But this is exceedingly small compared with the natural double refraction of the 

 crystalline medium, as far as experimental data on this question are available up to 

 the present,* and this increase of double refraction in passing from the liquid to the 

 crystalline state demands a value for the local molecular field large in comparison 

 with the strongest magnetic field which we can produce in the laboratory (4). 



It is expected too, that if we could use a more delicate means of measuring x we 



* A large number of ordinary and extraordinary refractive indices are given in ' Recueil des Constantes 

 Physiques,' published by the French Physical Society. These, however, refer to minerals which occur in 

 sufficiently large crystals for optical experiments to be readily carried out. Among the few organic 

 crystals which have been investigated are cane sugar, % t 1'570, MO = I'USTj benzilc, n f = 1'563, 

 o = 1-659. 



o 2 



