1904 - 5 .] Magnetic Quality in Molecular Assemblages. 1029 
cube edge is now vertical only. The horizontal line is parallel to 
the face diagonal. The main difference lies in the relative magni- 
tude of alternate lobes. 
In directions perpendicular to body diagonals the parallel 
component was constant while the transverse component was 
practically zero; so that “the magnetite acts as an isotropic 
substance for all directions contained in the face of the octa- 
hedron.” 
Weiss remarks that the evident irregularities are probably due 
to fissures in the disc. 
4. Weiss’s Experimental Results. — Fig. 4 represents results 
similarly to fig. 2. In this case the observations were made upon 
a disc of iron-wire network. This disc is regarded as being a 
slice cut from a face of a cube constituted of three mutually 
perpendicular sets of sheets of magnetic material, successive sheets 
being separated by layers of non-magnetic material. The actual 
network has nothing in its structure corresponding to the alternate 
layers parallel to its own plane. The effect of these would be to 
add a constant magnetisation to the component of magnetisation 
parallel to the field. The pointed curve inside the outer circle 
r 
T 
