1032 Proceedings of Royal Society of Edinburgh. [sess. 
for the component of induction parallel to the field, and points out 
that it gives the constancy in all directions perpendicular to the 
ternary axes which Weiss found. He points out the diamagnetic 
quality which would ensue in certain directions, if k did not lie 
between - 1 and + 3, and remarks that k is probably most often 
less than unity in absolute value. “One sees,” he says, '“that 
these formulae reproduce all the peculiarities observed experi- 
mentally.” 
His formulae certainly give the general nature of observed' 
peculiarities. It will be considered later whether or not they give 
their detailed nature. His primary remark that the components 
of induction are functions of the components of the field might be 
reversed. On the molecular theory, the field acting upon any 
molecule has an internal as well as an external constituent, and 
the former is determined by the magnetisation. Further, it is the 
internal constituent which must exhibit relations to the crystal- 
line symmetry. 
6. Components of Field due to an Ideal Magnet. — Angles being 
measured from the direction proceeding from the south pole to the 
north pole of a magnet, and r being distance from the centre of 
the magnet, whose moment and semi-length are M and a, it is easy 
to verify that, to the second order in a/r, the components of the 
total force, F, along and perpendicular to r, respectively, are 
F cos <f> = 2~ cos 0^1 + ~ g (2 + 5 sin 2 0)J , 
Fsin<£= ^ sin #[~1 + ^ ^-(5 cos 2 0—1)1. 
r d L 2 r 2 J 
From these we obtain, for the components parallel and perpen- 
dicular to the magnet, the expressions 
F cos (6 + <f>) = *[ ( , cos 2 6 - 1) + L?f(35 cos 4 0-15 cos 2 0+3)J, 
F sin (0 + <#>) = ^ sin 6 cos 6 |^3 + 1 ~(35 cos 2 6 - 15) J . 
7. Parallel Component of Force due to an Infinite Homogeneous 
