Overgrowths on Diamond. 
101 
together in their original positions by means of a transparent cement slightly- 
harder (or slightly less solvent) than the glass ; and lastly the edges of the 
octahedron to be abraded (or dissolved) away. The layers of cement being 
more resistant than the glass will stand out in slight relief as the attack 
proceeds. 
Next let the same octahedron of glass be split again into thin laminae 
parallel to another pair of opposite faces and the cementing and abrading 
(or dissolving) process repeated. We shall obtain another set of layers 
protruding in relief and intersecting the first at a constant angle. 
Finally, let the series of operations be repeated for laminae parallel to 
the two other pairs of opposite faces. The whole octahedron will now be 
spaced out into a device of small octahedra, marking its faces with a network 
of flush interlacing equilateral triangles, showing in relief on the reduced 
edges. 
Laminated diamonds are exactly like this. The laminae are sometimes 
as thin as paper ; more often they are in the vicinity of half a millimetre 
thick. They are not hemitropic. They alter the overall contour of a crystal 
very little, differing essentially in this respect from the made, but on the 
other hand they modify the surface detail to some extent. The layers corre- 
sponding to the cement layers of the hypothetical glass model always 
protrude where they show themselves, and, as a rule, they may be easily felt 
with the finger-nail or with the edge of a knife. A formal analogy will be 
found in the octahedral structure of iron meteorites, wherein kamacite stands 
for the glass, and the less easily dissolved taenite for the cement. 
The laminae scarcely ever manifest their existence save on the well- 
developed somewhat coarsely rilled faces of the rhombic dodecahedron or 
tetrahexahedron. On octahedron faces they are rarely seen, as also on the 
finely grained rounded edges of the octahedron (which are embryo dodeca- 
hedron and tetrahexahedron faces). On octahedron-dodecahedron combina- 
tions they may be traced all round a stone excepting where the octahedron 
faces interrupt. The intersections of the laminae with the octahedron faces, 
however, are often indicated by an array of indented triangles. 
In the majority of cases only one set of parallel laminae appear on any 
one diamond. When two sets appear, intersecting on a solution face, one 
set is nearly always much more prominent than the other. An examination 
of a great number of laminated specimens has only determined a very few — 
and some of these doubtful — with more than two sets of laminae in one 
stone. 
In Fig. 3 I have attempted to depict a laminated octahedron-dodeca- 
hedron combination from Wesselton, showing two sets of laminae. It is 
one of the rare cases in which the edges visibly cut the octahedron face. 
Though some relationship between the laminae and the indented triangles 
is shown, it is not so marked as in many other specimens, wherein, while 
