It was recalled that 3 techniques were used to study dislocations in ice. These were etch-pit replication, transmission electron microscopy and X-ray topography. Because ice was a weak absorber of X-rays, and could be produced with a low dislocation density, the most useful technique was X-ray topography. The introduction of high-intensity synchrotron radiation, with its short exposure times, had shown that the images obtained by conventional X-ray topography were those of dislocations that had undergone recovery. Dynamic synchrotron X-ray topographic in situ observations of single crystals undergoing deformation had shown that slip occurred mainly via the movement of screw and 60° 1/3<11▪0> dislocations on the basal plane. Non-basal slip via edge dislocations could also occur. The operation of Frank-Read, and other dislocation multiplication sources, was clearly demonstrated and dislocation velocities were measured. In polycrystals, dislocation generation occurred at grain boundaries - where stress concentrations existed - before lattice dislocation-generation mechanisms could operate. Faulted dislocation loops were found to be mainly interstitial in both polycrystals and single crystals.

Imaging Dislocations in Ice. I.Baker: Microscopy Research and Technique, 2003, 62[1], 70-82