4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



well adapted to the purpose. As for the possible intersection of 

 positive spheres, since any great amount of intersection, or coales- 

 cence, of the model atoms of this or of any other theory must 

 abolish their individuality, and since the positive sphere is little 

 more than a simple mathematical expression of the coherence and 

 individuality of the atom (see also §7), it is consistent, as well as 

 very necessary, to assume that positive spheres cannot intersect. It 

 will also be assumed that the volume of the positive sphere is nor- 

 mally proportional to its charge, that is, to the number of magnetons 

 in the atom, but that it is compressible ; and that the normal radius 

 of the magneton is about half that of the positive sphere of the 

 Hydrogen atom : * that the volume of the positive sphere of an atom 

 is usually very different from the total space occupied by the atom, 

 and a way to account for this, will be made clear later (§15). 



Some reasons for believing that the electron is this magneton may 

 be enumerated now, and discussed more fully afterwards. They 

 are: 



1. It seems to be the only satisfactory way of securing valence 

 electrons which are at rest, or vibrating within narrow limits, near 

 the surface of the atom — a great desideratum from a stereochemical 

 standpoint — without abandoning the very essential idea of orbital 

 motion in the atom. 



2. Even if the orbital motion is abandoned, and we suppose that 

 the atom does contain electrons of the usual type in positions of 

 equilibrium near its surface, the purely electrostatic nature of their 

 action would be altogether inadequate from a chemical point of view. 

 The additional magnetic forces furnished by the magneton are 

 exactly what the phenomena of chemical action require. 



3. It alone can give the atom a structure that accords closely with 

 what is known about the magnetic properties of matter. 



A general discussion of these points is given in §§2, 3, 4, the last 

 being considered first. In §5 there is a brief study of the forces 

 between two magnetons. In §6 it is argued that a number of mag- 

 netons within a sphere of uniform positive electrification must tend 

 to arrange themselves in groups of eight. This suggests structures 

 for the atoms (§7) that are in good accord with the general relations 

 in the Periodic Scheme. A model which partially illustrates the 

 behavior of the group of eight magnetons is also described, and the 

 accompanying plates (1 and 2) show photographs of it. In §8 these 

 results are compared with what is known about the number of 



1 The diagrams in this paper are drawn to scale on this basis. 



