1911-12.] The Molecular Theory of Magnetism in Solids. 237 
elucidation of Weiss’s results on plates cut perpendicular to a quaternary 
and a binary axis respectively. In that paper it was presumed that the 
crystalline arrangement in the cubic space-lattice was the one which gave 
to each molecular magnet twelve nearest neighbours, and the general nature 
of the experimental results was shown to be in agreement with the theory. 
[In that paper the following errata have to be noted : § 6, second last 
formula, for 15 read 30; § 9, for 44 read 49, and for A>B read A<B. 
The force parallel to a quaternary axis is negative, which agrees with 
Weiss’s theoretical postulate made in explanation of the peculiar form of 
the curve of magnetisation along a quaternary axis.] In the second paper 
(. Proc . RS.E., 1907), a similar investigation was given in respect to the 
arrangement in which each magnet has six nearest neighbours, and the 
results were found to be opposed to the experimental data ; so that this 
arrangement cannot exist in magnetite. Nor can it occur in crystals of 
metallic iron, for these were found by Weiss to have magnetic properties 
similar to those of magnetite. The only other cubic crystalline arrange- 
ment (not yet investigated) is that in which each magnet has eight nearest 
neighbours. 
Quittner has carried out more extended observations on plates cut 
parallel to the cubic faces and the octahedral faces in fields of various 
strengths, and certain peculiarities are brought out. For the general 
elucidation of these, the existence of internal fields of the types above 
discussed are sufficient. The normal results are due to the cubic crystalline 
structure, and the abnormalities are due to modifications of that structure 
by crystalline flaws — incipient foliation in definite planes. This explana- 
tion of the abnormalities was adopted by Weiss and strengthened by 
Quittner. The molecular theory bears it out fully. The abnormalities are 
sometimes present, sometimes absent, even in different parts of the same 
crystal. 
Fig. 6 is reproduced from Quittner’s paper. The ordinates represent 
intensity of magnetisation in the plane of a cubic face ; and various curves, 
corresponding to different strengths of field, are given. In weak fields, the 
regular plate shows no directional variation of magnetisation. This is 
probably due to the existence of thermal motions in the molecules which 
can temporarily free some molecules from the control of an internal field 
and so leave them subject to the external field alone. At higher fields 
maxima appear in the directions of the diagonals and minima in the 
directions of the axes. Now the value of y L' 2 given at the commencement 
of this section, with the condition G>D, shows that the internal force 
along a quaternary axis is negative, while it is positive along a diagonal. 
