With the aim of testing further the conclusion (Wilman 1950
a
) that a new deformation process, rotational slip, occurs prominently when crystal surfaces are abraded unidirectionally, an approximately {110} face of a copper crystal was smoothed by electro-polishing, abraded along <11¯0> and examined by electron diffraction after various stages of etching. The immediate surface regions were heavily disorientated, but those slightly below the surface had mostly become rotated by about 35°, though a decreasing proportion was rotated by larger angles up to about 90°, about the cube edge lying in the surface and normal to the abrasion direction. When abrasion was along <001> or <112>, rotation occurred about <11¯0> and <11¯1> respectively. In iron, also, large rotation of parts of the crystal surface occurred when crystals were abraded parallel to either {100}, {110} or {111} planes which were normal or steeply inclined to the surface. The interpretation in terms of rotational slip on {001}, {110} or {111} planes is fully supported by previous metallographic observations of such rotated lamellae formed parallel to these planes in copper (and aluminium ) and iron, and by electron-diffraction evidence of the structure of beaten metal foils. It is pointed out that the deformation caused by the abrasion is analogous to the development of ‘deformation bands’.