119 
1921-22.] On Models of Ferromagnetic Induction. 
be susceptible to influences of temperature, vibration, strain, and so forth, 
to which the nucleus itself would be impervious. 
26. We have next to consider what circumstances affect the stability 
of the Weber element when controlled in the manner here suggested. 
Suppose, as an ideal case, that the opposing actions within each atom 
(taken separately) are exactly balanced ; in other words, that the opposed 
poles of the fixed magnets are exactly equal, and that the clearances are 
exactly equal. Suppose also that there is complete magnetic rigidity, so 
that the presence of the Weber element between the opposed fixed poles does 
not disturb the symmetry of their action. Imagine atoms in which there is 
this complete balance of forces to have their centres arranged in rows, as they 
would be in a crystal. Then there is still some stability on the part of the 
Weber elements, for they exert magnetic forces on one another from atom 
to atom as in my old model, and their mutual magnetic forces tend to make 
them form rows which can be broken up by the application of a feeble 
external field. Thus in fig. 17 let there be two adjoining atoms with (in 
one line) the respective fixed poles and A 2 B 2 , and the respective Weber 
W2 62 
Fig. 17 . 
elements W^ and W 2 . The mutual action between W^ and Wg will give 
them some stability even if the action on them of the fixed poles is exactly 
balanced. When a transverse field is applied there will be, as there should 
be, a small stable deflection preceding rupture, then a violent swinging over 
into new rows more favourably oriented to the direction of the field ; and 
when the field is removed they will remain in rows that are so oriented 
as to give the piece residual magnetism. It is not impossible that in 
a crystal of pure unstrained iron the resultant controlling force which 
determines the stability of the Weber element in each atom may be 
entirely due to the mutual forces which the Weber elements exert on one 
another from atom to atom. The coercive force in pure annealed iron is 
so weak as to favour the idea that it may be explained in this way. 
On the other hand, the structure of the atom may be such that the 
attracting poles always act more strongly on one another than the repelling 
poles, so that in every atom the Weber element has an independent stability. 
It is at least possible that in the “ fixed ” parts of the atom there is not 
complete magnetic rigidity, and that the Weber element is made stable by 
internal displacements which are due to the forces between it and the shell, 
namely, by the drawing together of attracting poles, and the pushing apart 
