(153) 
A CONTEIBUTION TO THE STUDY OF THE DIAMOND 
MACLE. 
With a Note on the Internal Structure of Diamond. 
By J. E. Sutton, M.A., Sc.D., F.E.S.S.A., Hon. Memb. E.Met.S., 
Hon. Memb. S.A.S.C.E. 
** Composite crystals often occur, in which the several portions have- 
different orientations governed by regular and definite laws. When the 
crystallisation of a substance held in solution is hurried by rapid evapora- 
tion of the solvent, the crystals usually grow together in groups, in which 
the arrangement of the several members is purely accidental. But it was 
observed at a very early date that crystals of certain minerals, in particular 
those of cassiterite and spinel, are joined together in a regular and constant 
manner to form a well defined individual. . 
" Eome de I'lsle was the first to attempt an explanation of the composite 
character of the crystals of spinel and cassiterite, and he introduced the 
word made to denote a kind of composite crystal which we now call a twin. 
Werner employed the word zwilling (= a twin), at present used by Grerman 
crystallographers, and later on Haiiy introduced the word hemitrope (from 
vifii — half, and rpoiroq = a. turn) , for he perceived that the orientation of the 
two portions of every well-defined twin known to him is given by the 
following law : A complete crystal, bounded by the forms observable on the 
twin, is divided along a central plane which is parallel to a possible face ; 
and the half on one side of the plane is then turned through 180° about the 
normal, the two halves remaining in contact to form the twin. This law 
gives in very many cases the relative orientation of the two portions united 
together in a twin crystal ; it offers no suggestion as to the cause of 
twinning, and supplies no explanation of the growth of the twin " (Lewis,. 
* Crystallography,' 1899, p. 461). 
It is rarely, however, that a diamond made is equivalent to two halves 
of a complete crystal, or that the length of an edge is 1*225 times the 
thickness between two opposite triangular faces, or that its " central plane 
is a plane at all. Mostly it is of tabular habit, and its aspect is pretty 
