62 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



If it had been assumed that v for the magneton was about .01 Xc 

 (which is the sort of velocity usually attributed to the electron in 

 orbital motion), the calculated heat of dissociation would have been 

 10,000 times less, and even if the radius of the magneton were sup- 

 posed to be as great as that of the whole atom, the value would only 

 be increased 10 times. This is strong support for the assumption 

 v = c, which was made originally on more general grounds. 



In the foregoing calculation the electric work required to draw the 

 magnetons from the centers of their atoms, and the elastic work 

 required to compress the envelopes of the atoms, have been neglected. 

 These would obviously be smaller quantities, as they are associated 

 with secondary effects, and could not affect the order of magnitude 

 of the result. Their effect would, in fact, be to diminish the work 

 required to separate the two atoms, and so, probably, to bring the 

 calculated value into even better agreement with the actual value. 

 It should also be borne in mind that the calculation has been made 

 for the more improbable of the two conceivable configurations of 

 the molecule (see §§ 5, 20) ; but it is certain that the resultant forces 

 would be very much the same in the two cases, although the mathe- 

 matics involved would be more intricate for the other configuration. 



§19. The Magnetic Properties of Matter 



A brief survey of the present state of our understanding of 

 magnetism may be included at this point. 



A theory giving a complete account of magnetic phenomena must, 

 in its final deductive aspect, proceed in certain logical steps. First 

 it must provide a sub-atomic mechanism that would actually be 

 expected to produce, in a general way, the external magnetic phe- 

 nomena that are observed for gross matter. Secondly, it must 

 include a fairly detailed view of the structures of the different atoms, 

 so as to explain their different magnetic properties. And lastly, the 

 combining properties which it gives to the atoms must be such that 

 the explanation can be extended to the magnetic properties of all 

 kinds of molecules and solid aggregates of atoms and molecules. 



Hitherto no theory of magnetism has attempted to go further than 

 the first step, and I have shown in §2 how incomplete even that has 

 been. The present theory, on the one hand, is of so definite and 

 far-reaching a character that it is able, in a certain sense, to cover 

 the whole ground. It is true that very little is known as yet about 

 the magnetic properties of the atoms themselves, but what is known 

 is explained by it (see §2, and below, for Helium and Argon). 



