IO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



because the magnetic forces set up by electrons moving in orbits with 

 about one-hundredth the velocity of light, as his are, would be much 

 too small to be of significance in interatomic actions. However, a 

 still greater objection, to bring out which was the chief purpose of 

 the above quotation, is that such systems, as it seems, would not 

 attract but repel one another magnetically. 



Suppose that two electrons are constrained to move in parallel 

 orbits, and in the same sense. If they can move synchronously, 

 keeping always on the same side of their orbits, they will attract one 

 another magnetically ; but it can be shown that this is not a stable 

 configuration, at least for velocities small compared with that of 

 light. For, since the electric repulsion between them is greater than 

 the magnetic attraction, the resultant force between them is one of 

 repulsion; and thus if by some chance one of them is slightly dis- 

 placed relatively to the other, the action of the tangential component 

 of the repulsion between them will increase the separation until they 

 are on opposite sides of their orbits, in which positions they will repel 

 one another magnetically, as well as electrically. 



When, therefore, it is remembered that the whole of the explana- 

 tion of chemical phenomena given by the present theory depends 

 upon the possibility of magnetic attraction taking place between two 

 magnetons, it is evident that the substitution of the magneton for 

 Langevin's electronic orbit is imperative. 



Thus the magneton not only provides in a simple way the orbital 

 motion which must otherwise be secured by making inconsistent 

 assumptions about the behavior of classical electrons, but, what is 

 equally important, it supplies a foundation for a detailed explanation 

 of specific interatomic attractions of all kinds by providing an orbit 

 which is equivalent to a current circuit at every instant and not only 

 as an average effect in time. 



This theory was first worked out in connection with the phenomena 

 of valence ; and probably that was necessary, for chemical phe- 

 nomena are, from their nature, very much more detailed and distinct- 

 ive than magnetic phenomena; but the groupings of magnetons 

 about to be discussed from a primarily chemical standpoint must 

 also bear the test of criticism from a magnetic standpoint. This test 

 I will apply in detail at the end of this paper, but enough will be said 

 here to show why the atoms of the inert gases should be the most 

 diamagnetic of all atoms — as they are. In the same place the empiri- 

 cal magneton of P. Weiss will be considered : that is not a mechan- 

 istic conception and so could not have been developed in connection 

 with the topics dealt with here. 



