The geometrical concept of twinning in the diamond is inadequate because a reflection across either (111) or (112) will give identical relationships. The twinning plane can be identified by observing the contact or by demonstrating structurally that (111) lends itself to the growth phenomena involved in twinning. In either normal or twinned growth across (111) the first and second coordinations are identical and differences appear only in the third or higher coordinations. The statistical opportunity for twinned structures to be initiated is therefore high.
The probability of atoms falling into the normal or twinned position is not wholly a function of the structure. The surface energy distribution on a growing crystal face may be modified by the type of impurity that forms an adsorbed film on the surface. The resultant energy pattern may increase or decrease the statistical probability of the atoms attaching themselves in the normal or twinned position. By slowing the rate of migration of atoms to the surface the impurity film will promote the orderly accretion of atoms. Structural conditions indicate that diamonds do not grow by the accretion of single atoms.
Twinning is highly characteristic and what are apparently random intergrowths are more often multiple twins. The classical examples of contact, penetration, polysynthetic, and cyclic twins, by focussing attention on the geometrical symmetry of highly unusual and non-characteristic growth forms, obscure the true structural basis of twinning. Tables useful in the analysis of complex twins of non-symmetrical external form are given.
Crystallographic continuity exists across a twinning plane. All other boundaries between twinned individuals are suture contacts which are very often the loci of voids and inclusions.