The application of classical nucleation theory to the initiation of crystal growth in glasses is discussed. Its application to experimental results obtained for the rate of nucleation in three types of glass, one showing nucleation separately from crystal growth, another showing simultaneous nucleation and growth, and the third showing crystal growth after metastable liquid/liquid phase separation, is then described. Recent work on the kinetics of unmixing in glasses is outlined. The influence of so-called ‘nucleating agents’ in the glasses described appears to be exerted through changes in the anionic structure.

Quantitative data are presented for the kinetics of crystal growth in an aluminosilicate glass to which a few wt % of a transition-metal oxide had been added. The composition of the base glass was CaO 3o, MgO 2, Al2O3 15, SiO2 53 wt %, to which the oxides of iron, vanadium, chromium, or zinc were added. Glasses were prepared with a range of high to low valency ratios in cases where the transition metal oxide could exist in two or more oxidation states.

Small quantities of Fe2+, Fe3+, Zn2+, V3+, V4+, and V5+ were all found to increase the kinetics of crystal growth from the glass phase. The magnitude of the effect is compared with additions of Mg2+ to the glass. Cr2+, Cr3+, and Cr6+ were found to decrease the rates of crystal growth. Values for the apparent activation energy for crystal growth ranged from ∼50 kcal/mole to ∼150 kcal/mole. The high values were found to be of the same order of magnitude as that for the activation energy for viscous flow in the glass at the crystallization temperature.

The results are interpreted in terms of the structure of the glasses and current theories on the rate of crystallization of glasses.

Glass-ceramics based on the mineral cordierite have good mechanical and thermal properties. They can be easily prepared from glasses whose compositions lie in the primary phase field of cordierite on the 5 % CaO plane of the CaO-MgO-SiO2-Al2O3 system. For uniform crystallization in the bulk, addition of a suitable nucleating agent is essential. The addition of 10–12 wt % of TiO2 is considered to be suitable for both ternary and quaternary glasses. During controlled heat treatment of these glasses extensive phase separation occurs. The nucleation and crystallization processes that follow have been studied by differential thermal analysis, X-ray diffraction, and light and electron microscopy. Some of the physical and mechanical properties of these glass-ceramics have been measured.

The X-ray powder diffraction pattern of synthetic anorthite crystallized from a CaO-Al2O3-SiO2 glass having the anorthite (1:1:2) molar ratio is identical with that reported in the literature, and also with that of a natural Japanese anorthite specimen. Increasing the CaO or SiO2 content of the parent glasses used for crystallization studies has no measurable effect on that portion of the powder pattern attributable to anorthite. However, glasses containing an excess of Al2O3 ranging from 5 to 10 mol % gave, after crystallization at temperatures below c. 1150 °C distinctively different powder pattern. Several powder reflections that are normally strong, such as , etc., were found to be virtually absent. Moreover, the chemical excess of Al2O3 did not appear as a separate alumina-bearing phase. Upon reheating these anorthites to temperatures above 1200–50 °C or upon crystallizing a fresh portion of alumina-rich glass above 1200-50 °C, only the normal anorthite powder X-ray pattern was obtained; the pattern also contained some reflections due to corundum (α-Al2O3). Examination of the anomalous low-temperature anorthite by electron diffraction shows that the apparent absence of some powder lines is caused by both albite and Carlsbad twinning, which occur on an intimate scale not exceeding a few unit cell repeats. From the unique manner of its occurrence, the twinning is believed to be associated with the inclusion of an excess of Al3+ in the anorthite.

Iron-rich olivines have been oxidized in air in the laboratory and the mechanism of their breakdown has been elucidated using X-ray diffraction and electron microscopy. Low-temperature oxidation (500–800 °C) produces well-oriented hematite- and magnetite-like precipitates together with amorphous silica. The reaction is a cellular one in which thin needles of oxide about 50–100 Å apart grow into the matrix separated by regions of amorphous silica. Nucleation of spherical colonies of the iron oxide and silica occurs on dislocations.

Although the hematite or magnetite always shows the same topotactic relationship with the matrix, the direction in which the needle-like precipitates grow is determined by the orientation of the nucleating dislocation. The small size and highly distorted nature of these precipitates accounts for the diffuseness of their X-ray reflections.

Oxidation at 1000 °C produces undistorted equiaxed grains of the oxides about 0·2 μm in size. They are surrounded by silica, which produces a disordered electron diffraction pattern. As the temperature is raised, the silica achieves more structural order and the oxide grains increase in size.

A model has been constructed that establishes the geometry of inclusion fabrics to be expected in near-spherical crystals that grow continuously while either the crystals rotate in a static matrix or a matrix fabric rotates about them. Garnets in certain metamorphic rocks contain inclusion fabrics whose geometry is very similar to that of the model. Such syntectonic, snowball, or rotational garnets, as they have previously been termed, can be studied in three dimensions. A method is outlined whereby the rotation axes for individual crystals viewed in thick sections can be located in certain planes of cut. The significance of the geometry of included rotational fabrics is discussed in relation to previous work and to possible mechanisms of rotation.

In the aureole of the syn-orogenic Hasvik Gabbro a variety of exsolution phenomena developed during the contact metamorphism. These exsolution phenomena have similar textures to those produced during the age-hardening of metal alloys. In feldspars the exsolution precipitates are closely related to the pattern of excesses over and deficiencies from the formula requirements, particularly in terms of SiO2 and Al2O3. The exsolution products are dependent upon both the temperatures of ageing and the details of the mineral chemistry. Evidence supporting an exsolution theory of origin for myrmekite is provided, and it is shown that precipitation from heterotype feldspar solid solutions during metamorphism may in fact be a complex process involving a number of precipitate phases.

The nucleation of sillimanite as a precipitate phase is discussed and it is shown how sillimanite develops at a number of distinctive sites within feldspar and garnet. From the textural relations of sillimanite precipitates and their host minerals a control exerted by the thermal stress configuration over the nucleation of precipitate phases is suggested. It is argued that the thermal stresses exerted across grain boundaries and within minerals during metamorphism may have considerable relevance to the interpretation of textures in metamorphic rocks.

The close interrelation between sintering and grain growth is discussed critically, and the necessity for control, but not suppression, of grain growth is established. A distinction between discontinuous and exaggerated grain growth is drawn, and the superiority of grain boundary films as grain-growth control agents is demonstrated. Work on the establishment of tentative criteria for the selection of grain-growth control additives for alumina is reviewed and discussed, and the extension of these ideas to the selection of NaF as a successful aid for magnesia is described.

The densification of pure magnesia was examined along with mixtures containing 0·5 and 1·0 mol % of calcium oxide and forsterite. The sintering behaviour at temperatures between 1400 and 1800 °C was studied and measurements made of changes in shrinkage, porosity, bulk density, and grain size for sintering times of up to 8 h at sintering temperature. It was found that both the 0·5 and 1·0 mol % additions enhanced the sintering of pure magnesia in the temperature range 1500–1700 °C Increased grain growth was found for mixtures containing forsterite but little change occurred for mixtures containing lime. Examination of the grain growth data showed that the mixtures obeyed a relationship: Time at temperature × Constant = (grain size)n, where n = 2 for the grain growth of pure magnesia and mixtures containing lime, but n = 3 for mixtures of magnesia with additions of a silicate such as forsterite. A value of 78 kcal/mole was obtained for the grain growth of pure magnesia, which agrees with other researches.

The extent of the solubility of lime in periclase for each lime-magnesia mixture at each temperature was also examined using electron probe analysis and related to the sintering behaviour observed for these mixtures.

It is now well established that naturally—occurring sulphides—particularly galena—are frequently found in the deformed state and that at least part of this deformation can be eliminated by annealing in the laboratory.

The present contribution is concerned principally with galena, for which it is indicated: that tectonic deformation leads to natural work hardening, which may be reduced by simple heating in vacuo; that the rate of softening is related to the annealing temperature; that softening is related to recovery at low temperatures and to both recovery and recrystallization above 250–300 °C that single crystals of galena exhibiting minor slip and kinking recrystallize yielding new, unstrained grains elongated parallel to the old slips and kinks, i.e. a foliation involving unstrained grains may form under a regime devoid of compressive stress; and that where deformed galena and deformed sphalerite coexist, the galena may be made to recrystallize completely (e.g. at 325 °C in 50 days) while leaving the deformation in the sphalerite entirely intact; apparently the coexistence of a strained component A and an unstrained component B in a deformed ore or other metamorphic rock does not necessarily indicate differences in age or deformational history—it may simply indicate that post-deformational temperatures have exceeded the recrystallization temperature of one component but not of the other.