Comments on the paper of Hallett and Mason

Abstract

If Dr Mason does not venture an explanation of the phenomena of the different morphological forms of ice formed at different temperatures, I will take the risk of putting forward one which, to be honest, I only thought about this morning. If it is right, and I think there should be fairly simple means of testing it, it would apply in general to all crystallizations from the vapour which occur in conditions not far removed from equilibrium, that is where we do not need to consider the different concentration of vapour near the crystals, for it has been long known especially since the work of Papapetrou (1935) that rapid crystallization involving impoverishment of vapour in the neighbourhood of the growing crystals favours the growth of rapidly extending dendrites reaching out into the denser vapour regions. Now the hypothesis I want to put forward depends on some work of Hale (1957) reported at the International Crystallographic Congress at Montreal in 1957 on the crystallization of a very different substance, namely, quartz in hydrothermal containers. Crystals of quartz grown from a small seed show on sectioning diagonal lines marking out the regions that have crystallized on the c or (0001) face from those that have grown on the prismatic faces (101̅0) (figure 64 A ). This, of course, has been known for a long time, but what was not so well known was that careful analysis showed that the very small degree of impurity present—and the major impurity in the case of quartz-growing is nearly always aluminium—had a very different concentration in the different parts of the crystals being practically absent in the c and present to its full amount in the a . In other words we were dealing not strictly with a homogeneous material, but with two different materials crystallizing together—what might be called c quartz and a quartz. If this is correct it is not sufficient in general to specify crystalline material by its formula, but rather to say this is material crystallized on such and such a face, because strictly every face will have a slightly different composition. This would be true even in the absence of material impurities, because there is one impurity that cannot be avoided and that is simply lattice vacancies. Now the condition of growth of a crystal which is very anisotropic, that is growing in long needles or thin plates, will be quite different from that of one growing approximately equally in different directions, as quartz does. The observations on such growth show that after an initial period when the crystal reaches approximately its full width, subsequent growth is confined to the major direction, and may be many times more than that in the other. The kind of conditions are shown in figures 64 B and C .