It was recalled that the faces of ice crystals which were in equilibrium with the vapor were thought to undergo reconstruction, to a quasi-liquid transition film, at temperatures above about -20C. At much lower temperatures, it appeared that the free energy could be reduced by means of a less drastic surface reconstruction which involved the reorientation of surface dipoles to give a superlattice. Calculations suggested that this superlattice was energetically preferred on (00▪1) basal faces, at temperatures below about 30K, and on (01▪0) prism faces, at temperatures below about 70K. On (00▪1) faces, twin boundaries had a very low energy and could be present in large numbers. The rate of reconstruction was expected to be slow, because of the low temperatures which were involved. However, molecular interchange with the vapor could facilitate it. It was suggested that this structure might be observable by means of electron diffractometry or fine-focus infra-red spectroscopy.

N.H.Fletcher: Philosophical Magazine B, 1992, 66[1], 109-15