Some Controversial Notes on the Biamond. 
138 
separate diamonds can grow outwards independently from a common 
starting-place. Then there are occasional diamonds whicli seem to have set 
out by first intentions to be cubes, but have developed incipient octahedral 
forms at each corner, thus assuming a distant resemblance to the rare form 
in which the octagon and the cube combine symmetrically — the corners of 
the octahedron projecting through the centres of the cubic faces. Equally 
significant are the -ectahedra with re-entering corners. Three of these have 
been preserved as good specimens — rounded octahedra of different sizes and 
colours. It was not easy to interpret the indented corners, but it seemed 
reasonable to suppose that they must once have contained octahedron 
diamond satellites. Later on another such diamond was found with the 
corner satellites in situ and strongly held. 
Now these cases do seem to indicate a group-forming tendency. The 
corner satellites surely must have formed after the primary diamond had 
reached an advanced stage. For it is not in reason that seven distinct 
centres of crystallisation should start so symmetrically placed that six outer 
ones should be exactly at the places of the final corners of the central 
one. Still less is it in reason that the same process should be repeated 
again and again. 
I would submit that a theory of growth into groups and clusters after 
the manner of calcite and rock crystal affords an explanation of the habit of 
many " cleavages," whose irregular contours are not bounded by broken 
surfaces. It explains, may we say, why so many diamonds appear to have 
been formed in tight corners, and why so many have what look like sur- 
faces of attachment. The so-called cleavages found in situ in the matrix 
are more often such misshapen diamonds from clusters than fractured frag- 
ments derived from single crystals. A flat white diamond in its blue-ground 
matrix, from De Beers Mine, is one of these misshapen stones. Its exposed 
length is 10 mm. ; its thickness varies from 1 to 3 mm. It projects 5 mm. 
from the blue ground, while its outside edge is broken off', accidentally or 
otherwise. It has all the appearance of having once been in a cluster, and 
its exposed faces are much the same as that which prevails when two 
diamonds have crystallised together in easy contact. 
Bultfontein Mine produces ]nore groups and clusters than other mines 
do, perhaps for the same reason that it produces more interpenetrating 
crystals than all the other mines put together. Cross-grained stones in 
general, however, are not uniformly distributed throughout Bultfontein, 
so that, consequently, some market assortments contain great numbers, 
whereas others have much fewer. This same mine is pre-eminently a source 
of rhombic dodecahedra, a good half of its symmetrical stones being of this 
habit ; and the groups also seem to favour diamonds inclining to the 
same class. 
