The nature and stability of intercrystalline boundaries are discussed and conclusions as to preferred relative orientations are illustrated by ball-bearing and bubble-raft models.
Electron-diffraction evidence from crystals of layer-lattice, ionic, metallic and van der Waals (organic molecular) types is presented, showing that crystal pairs and symmetrical or unsymmetrical triplet or multiplet crystal groupings occur, having a common lattice row, but in relative azimuthal orientations at intervals agreeing with the predictions. The component crystals were thus in coherent metastable contact of the type discussed, and in some cases strong secondary elastic scattering showed that the crystals were extensively superposed on the plane which was normal to the common row.
The seven types of electron-diffraction pattern observed from such groupings, and the nature of the specimen preparation, suggest that the origin of these groupings is a mechanical deformation of the crystals during or after growth, by a process of `rotational slip' which has not hitherto been explored or defined in nature. Such slip is directly demonstrated macroscopically as a deformation process in potassium ferrocyanide trihydrate and gypsum, the slip being produced on the planes parallel to the highly perfect cleavage.
The relation between rotational slip and the known types of translational slip and dislocation theory is indicated, and in particular the `deformation bands' investigated especially by Barrett and his collaborators are concluded to be essentially rotational slip bands initially at low deformation, though modified by translational types of slip at higher deformation.
A valuable new insight is also provided into many observations on crystal growth and properties, especially the nature, determining conditions, stability and disorientation of crystals growing epitaxially on single crystal substrates, and into the nature of the deformation caused by unidirectional abrasion of crystals