CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY. 277 



expected, under the influence of an external magnetic field, to form a perfect junction, 

 in other words to become welded together. 



TAYLOR JONES* has obtained an induction as high as 74,200 Maxwells in soft iron 

 under strong fields. The tension necessary to pull the surfaces apart in this case will 



TD3 7'4. 2 X 10 8 



be - - = -,or2x 10* dynes/square centimetre, which is about a twenty-fifth 



OTT O7T 



of the tensile strength of the material. 



(6) ON A MAGNETIC THEORY OF CHEMICAL COMBINATION. 



On the theory of chemical action developed by Sir J. J. THOMSON,! the determining 

 feature of an atom from the point of view of chemical combination is the number of 

 positive valency electrons it possesses. These electrons are dragged from their loose 

 attachment to the nucleus, during chemical combination, and pass from one atom to 

 another. The two originally neutral atoms thus become oppositely charged and so 

 attract one another and form, as it were, an electric doublet. Let us look at this 

 problem from the magnetic standpoint. Each electron orbit is equivalent to a small 

 magnetic doublet and it is interesting to enquire how far the magnetic forces of such 

 doublets may represent the force of chemical affinity. Recent work on radio-activity, 

 the wide deflections of /8-rays, and the diffraction of X-rays, all point to a localisation 

 of the electrostatic charges in a minute core or nucleus. Round this micleus, and 

 under its control, the valency electrons (in part responsible for the magnetic properties) 

 rotate. It is conceivable, therefore, that the magnetic forces, in addition to the 

 important role they play in crystallization, may also in part be responsible for the 

 forces of chemical affinity.^ 



If, during chemical combination, there is a definite transfer of valency electrons from 

 one atom to another, we should expect to find an abrupt change in the magnetic 

 behaviour of an atom before and after chemical combination. If, on the other hand, 

 there is no such electron transference, we might expect that the atoms would preserve 

 their magnetic properties, which would be more or less of an additive nature. In a 

 remarkable series of investigations, PASCAL has shown that in a very large number 

 of organic compounds, the molecular susceptibility is the sum of the atomic 

 susceptibilities of the component atoms, provided the molecule contains no peculiarity 

 of molecular configuration such, for instance, as the ethylene linkage, unsaturated 

 atom, or complex nucleus. Thus if XM is the molecular susceptibility, and XA the 

 atomic susceptibility of a component atom, we have 



XM = SXA + X 



where the summation extends to all the atoms in the molecule and X is a positive or 



* ' Phil. Mag.,' vol. xli., p. 165, 1896. 

 t 'The Corpuscular Theory of Matter,' 1907, Chap. VI. 



I See footnote p. 274. Also W. M. HICKS, ' Roy. Soc. Proc.,' A, vol. 90, p. 356. 

 Ann. de Chim. et de Physique,' se"r. 8, vol. 19, p. 5, 1910. 



