102 MR. A. E. OXLEY ON THE INFLUENCE OF MOLECULAR 



16'5 x 10~ 22 x 6'8 x 10 23 x 11* 

 the saturation intensity of magnetization of iron - - ' - = 1760, 



which is nearly equal to the experimental value. Therefore the increase of x on passing 

 from the paramagnetic to the ferro-magnetic state can be represented by a purely 

 magnetic molecular field due to the magnetons contained in the iron molecules. This 

 increase of x is represented on WEISS'S theory by the molecular field Nl = 10 7 gauss, 

 and therefore 10 7 gauss is the magnetic field due to the magnetons when the iron is 

 saturated. The continuity of magnetic induction demands that this field is to a large 

 extent localized (see p. 89 supra). 



Now reverting to diamagnetic substances we are led, in the light of RITZ'S 

 theory, to identify the local molecular field (10 7 gauss) in their case as due to 

 the existence of molecular magnets, so arranged that for any particular molecule 

 the moment is zero. It seems that we could readily account on such a view for 

 the fact that the chemical combination of two diamagnetic substances can give 

 rise to a paramagnetic substance possessing magnetons. This is the case for 

 instance with the union of copper (diamagnetic) and sulphuric acid (diamagnetic) 

 resulting in the formation of cupric sulphate (which is paramagnetic and contains 

 10 magnetons to the molecule). The forces which come into play during the 

 chemical combination we may regard as upsetting the magnetic equilibrium of each 

 component. 



If these molecular magnets do exist in diamagnetic substances then the local 

 molecular field of the diamagnetic substance will be comparable with the molecular 

 field in ferro-magnetic substances, and will be represented as far as they are concerned 

 by a true magnetic field of intensity 10 7 gauss (approximately). In this case the 

 localized nature of the field is a necessary consequence of zero moment whicli has 

 been assigned to the diamagnetic molecule. 



In both ferro-magnetic and diamagnetic cases we have represented the molecular 

 fields as due to magnetic force. Now electrostatic forces, or indeed forces of any other 

 nature, may distort the configuration of a molecule in the crystalline structure, and 

 their effect could be represented by a magnetic field for both diamagnetic and 

 ferro-magnetic substances. Hence part of the molecular field for diamagnetic and 

 ferro-magnetic substances may be a true magnetic field and part a magnetic 

 representation of the distortion produced by forces of a different nature. At any 

 rate we cannot deny that the parallelism which has been found between ferro- 

 magnetic and diamagnetic phenomena, taken in conjunction with the work of WEISS 

 on the magneton and that of EITZ on spectral series, points to the conclusion that 

 the molecular field in diamagnetics and ferro-magnetics is represented in part (at 

 least) by a true magnetic field. In this connection the introductory remarks of 



* An atom of iron has 11 magnetons, and the gramme-atom of iron (56 gr.) contains 6'8 x 10 23 atoms, 

 therefore 1 c.c. of iron (whose density is taken as 8) wUl contain 6 ' 8xl 23 atoms. 



