1904 - 5 .] Magnetic Quality in Molecular Assemblages. 1047 
iron gauze ; and both curves of the longitudinal effects simulate 
very well his curves for magnetite (figs. 2 and 3) when a constant 
amount, corresponding to I 0 , is added to the radii. The points of 
difference requiring special consideration are (1) the acuteness of 
the angle at which the curves of the parallel component cut the 
axes of minimum magnetisation ; (2) the great relative size of the 
smaller loop of the transverse effect in the diagonal plane. 
With regard to (1), there seems to be little evidence in Weiss’s 
curves to make it certain that they should cut the axes at right 
angles. They might quite readily, so far as the observational points 
in these diagrams — with possibly one exception — indicate, be drawn 
so as to cut at an acute angle. In the case of the iron gauze, the 
angle is acute. 
With regard to (2), the difference is readily explainable by the 
existence of crystalline flaws perpendicular to the quaternary axes. 
The peculiar form of curve Q, fig. 1, practically necessitates this 
assumption. Weiss essentially made it in his theoretical discussion 
of the question ; he also ascribes, as we have already seen (§ 3), 
the great differences in the sizes of the loops in fig. 2 to the 
existence of flaws. Since, in the region between the binary and 
ternary axes in the diagonal plate, the transverse magnetisation 
does not depart much, in the neighbourhood of its maximum, from 
the quaternary axis, its magnitude must be greatly reduced by the 
existence of fissures perpendicular to that axis. 
17. Magnetisation in Principal Planes. — The only remaining 
principal plane is one perpendicular to a ternary axis. Under the 
experimental conditions which he employed, Weiss found (§ 3) 
that the parallel component of magnetisation was constant, and that 
the transverse effect was practically zero. Yariations of the former 
component amounted to 0T6 per cent, of its value ; the magnitude 
of the latter did not lead to an obliquity of half a degree between 
force and magnetisation. 
This plane differs from the other principal planes in that the 
force and the magnetisation do not in general both lie in it. Two 
distinct cases arise, therefore, according as we regard the plane as 
the locus of the force or as the locus of the magnetisation. The 
latter case has been already considered (§ 14). When the external 
