The Freetown Intrusion, Sierra Leone, evolved by recurrent injection of olivine-bearing magma producing an aggregate thickness greater than 6000 m of rhythmically layered rocks. Sedimentation and constrained viscous flow of crystallizing magma concentrated olivine crystals during initial formation whilst concentration of volatiles locally lowered residual melt temperatures. This affected the crystallization of pyroxenes especially in pyroxene-troctolites where abundance of olivine caused deferred pyroxene growth. Three examples are described: (1) Flattened and elongated schlieren of augite enclosing plagioclase in ophitic intergrowths show preferred orientation cutting across primary layering defined by olivine, proving that pyroxene growth was secondary and influenced by stress directions in a relatively rigid rock. (2) In some pyroxene-troctolites, augite is segregated into coarse ophitic nodules which are widely and evenly spaced in a troctolitic host rock. Successful pyroxene nucleation and extensive diffusion caused development of domains of equilibration within a ‘cotectic’ pyroxene-plagioclase-volatile system. Crystallization within this was deferred compared with the surrounding olivine-plagioclase assemblage in which effects of granulitization are conspicuous. (3) Occasionally the pyroxenes are segregated in dispersed lenticular bodies of noritic pegmatite in troctolite. As in the previous cases, growth of pyroxene involved resorption and regrowth of olivine. This reflects the highest temperatures of crystallization in equilibrium with volatile-enriched pegmatite fluids. Temperatures decreased inwards giving a zonal structure and late growth of biotite, orthoclase, and apatite.

The formation of late crystalline pyroxenes and pegmatites thus involved extensive diffusion of volatile and other components accompanying the establishment of domains of reduced pressure in hot and slowly cooled rocks subjected to gradual deformation due to progressive subsidence of the central parts of the intrusion. The final fabric is therefore metamorphic resulting from annealing rather than crystal accumulation.

The Driggith vein is representative of the lead-zinc mineralization in the Caldbeck Fells area, with a primary mineral assemblage consisting of quartz, baryte, calcite, chalcedony, pyrite, arsenopyrite, galena, sphalerire, bournonite, argentian tetrahedrite, and native antimony. Broadly similar assemblages are found in Pb-Zn mineralization of Upper Devonian to Permian age else-where in the Lake District. The paragenetic sequence appears to have been early pyrite and minor arsenopyrite associated with quartz; with later baryte, calcite, sphalerite, and chalcopyrite enclosed by galena. Within the galena occur inclusions of tetrahedrite, bournonite, and native antimony. A later stage of baryte mineralization was followed by the alteration of primary sulphides to secondary minerals. Elsewhere in the Lake District, galena occurring in this association also contains inclusions of native antimony, bournonite, and other sulphosalts, and in this respect differs from galena associated with earlier veins composed dominantly of chalcopyrite, pyrite, and arsenopyrite. The inclusions make a major contribution to the high antimony values reported in trace element analyses of galena from the Lake District. The occurrence of these inclusions, which in some cases display preferred orientation within the host galena, is discussed in terms of phase relations involving PbS and AgSbS2.

Strontium-rich (up to 16 wt% SrO) melilite occurs as microphenocrysts in a nosean-leucite-nephelinite lava from Etinde, Cameroon. Electron microprobe analyses show that the melilite is very strongly zoned with Sr, Fe, and Na enrichment towards the crystal rims. Sodium melilite and sodium ferrimelilite are the dominant end-member molecules and together account for up to 76 mole%. The crystals are optically negative and strongly birefringent with birefringence ranging from 0.014 in the cores to 0.032 at the crystal rims. Refractive indices measured at the rim of one crystal gave ω = 1.656, ɛ = 1.628. The cell dimensions of strontian melilite are a = 7.765Å, c = 5.054Å.

Examination of cotype cosmochlore from the Toluca meteorite confirms Laspeyres's observations (1897) in every respect, except that what he determined as iron was largely titanium. His data are completed by an electron-probe analysis and by full optical and X-ray data. Accepting the identity of cosmochlore and ureyite, the optical data of Frondel and Klein (1965) for the latter are partly in error or misprinted.

Cesanite occurs both as a solid vein (1 cm thick) and as a cavity-filling of an explosive breccia in core samples of the Cesano 1 geothermal well (Cesano area, Latium, Italy).

Cesanite is coloudess, medium to coarse grained, soft (H = 2 to 3) and light (ρmeas 2.786±0.002 g cm−3). It is uniaxial negative, ε = 1.564, ω = 1.570, with space group P63/m and cell parameters a = 9.442 (4), c = 6.903 (3) Å, for c/a = 0.730. Identifying spacings are 8.161, 2.822, 2.727,1.844 Å, in X-ray powder patterns strikingly similar to those of apatite. The chemical formula (microprobe analyses on two grains) is Ca1.53Sr0.03Na3.42K0.02[(Cl0.06F0.06OH0.44)(S2.99O12)]·0.44H2O, while the theoretical formula, derived from considerations on structural identity with apatite, is Ca2Na3[(OH)(SO4)3]. Cesanite is the end member of the apatite-wilkeite ellestadite series where [PO4]3− is entirely substituted by [SO4]2−, the charge balance being made up by Na+ substituting for Ca2+.

Gahnite and columbite with compositions (Zn5.96Fe1.55Mg0.21Mn0.13)Σ7.85Al16.09O32 and (Nb7.36Ta0.60Mn2.95Fe0.98Ti0.09)Σ11.98O24 occur as accessory phases in a highly evolved garnetiferous muscovite alkalifeldspar granite. Both phases have developed as a result of a concentration of Zn, Nb, and Ta during the fmal stages of the fractional crystallization which led to the formation of the pluton, and the lack of a major mafic mineral such as biotite, in which these elements are normally camouflaged in granitoid rocks.

The Kenyan trachyte volcanoes Kilombe and Londiani show some unusual chemical features. Analyses of elements normally regarded as residual in salic peralkaline rocks do not show the linear covariance characteristic of these elements in other East African volcanoes. The scatter from expected trends is greatest for Y and the lanthanides, less for Nb and Rb, and not apparent for Zr. The lavas also show variable loss of Na, P, and Fe. It is suggested that this chemical deviation is the result of a deuteric event during the later stages of crystallization of the lavas. The flows showing the chemical deviations, in particular negative Ce anomalies, also tend to have their ferroaugite phenocrysts deuterically altered to a mineraloid closely resembling iddingsite.

During the crystallization of these trachyte lavas an H2O-, Cl-, and F-rich vapour separates from the liquid. This absorbs some elements, particularly Na, P, Fe, Y, and the lanthanides from the silicate liquid as water-soluble or volatile halides. Some of the absorbed material is redeposited in other parts of the flow and the rest is lost altogether as the vapour escapes from the surface.

Sections perpendicular to [001] of ion-thinned specimens of fibrous grunerite (amosite) have been examined by high-resolution transmission electron microscopy. In this orientation, two kinds of dislocation have been observed with about equal frequency. One lies on [001] and has a Burgers vector a. The other is on [001] and has a Burgers vector ½a+½b Interpretation of features associated with these dislocations has been assisted by the use of two-dimensional models of I-beam cross-sections which can be interlocked to simulate the possible modes of stacking.

In Moinian metasediments regionally metamorphosed to sillimanite grade, garnet has subsequently been re-equilibrated in the inner (cordierite-potash feldspar) zone of the aureole of the Strontian Granodiorite. Zoning profiles of garnet from the aureole, and from surrounding regional assemblages, show retrograde Mnrich rims. Some regional garnets are also internally zoned. In the aureole as a whole, the partition measure for Fe and Mg between garnet and biotite shows a correlation with that for Ca between garnet and plagioclase. If interpreted in terms of a simple equilibrium model, this might suggest that pressure decreased with increasing temperature towards the granodiorite, but the trend can more plausibly be attributed to deviation from ideal solid-solution behaviour of Ca in garnet and plagioclase, with incomplete homogenization of garnet as a complicating factor in the outer parts of the aureole. From the difference in the temperature thresholds for garnet core equilibration, the timescale t2 of the thermal event relative to the regional t1 is estimated as t2/t1 ≅ 10−1.1±0.7, consistent with emplacement of the intrusion at an early stage of regional cooling.

Zoning in garnets in Barrovian garnet zone rocks of central Perthshire is shown to be dependent on the continuous prograde reaction histories in which the garnet was involved. In highly aluminous rocks, garnet grows at the expense of chloritoid and staurolite whereas in less aluminous rocks garnet grows at the expense of chlorite. Minerals included in garnet such as ilmenite, plagioclase, chlorite, and chloritoid show compositional changes which with the garnet zoning elucidate the garnet-forming reactions. Where garnet is resorbed to produce staurolite the former loses Fe at its rim to staurolite so that ‘reverse’ zoning of decreasing Fe and increasing Mn and Ca towards garnet rim develops.

Liquids in the system CaO-Na2O-MgO-Al2O3-SiO2 which crystallize plagioclase and forsterite, often with diopside and sometimes with spinel, were studied. The movement of the plagioclase-forsterite saturation surface to higher plagioclase contents as albite replaces anorthite, and the concurrent expansion of the forsterite and diopside primary fields, are documented. The maximum temperature of the liquid in equilibrium with plagioclase, diopside, and forsterite decreases as the anorthite to albite ratio decreases and as enstatite is added, and was experimentally determined over part of its range from 1278 to 113°C The decrease was not linear, being generally small even when 67 mole% albite was present, and thereafter, it must decrease rapidly to reach the values reported for the system albite-diopside-forsterite.

Clinopyroxenes are dominant in highly potassic, silica undersaturated mafic volcanics occurring on the western rim of the uplifted, rifted East African craton. A kimbcrlite style of eruption provides nodules of alkali clinopyroxenite (clinopyroxene + titaniferous phlogopite+titanomagnetite, apatite, sphene, and rare corroded olivine) which have similar bulk chemistry to the feldspathoid-bearing lavas. Many nodules display metasomatic textures supporting a formation from the alteration of pre-existing material; clinopyroxene growth is characterized by complex, non-oscillatory colour zoning. Comparison of natural clinopyroxene chemistry with published data for elinopyroxenes crystallized from synthetic potassium-rich mafic material, suggests that a significant proportion of the nodules crystallized at upper-mantle pressures. Neither garnet- nor orthopyroxene-bearing nodules have ever been recorded from south-west Uganda, suggesting that metasomatism of the local mantle has proceeded far enough to obliterate all recognizable remnants of four-phase lherzolite.

Sediments from DSDP 424, a core taken in a hydrothermal mound in the Galapagos Mounds Field, comprise biogenic oozes, siliceous ash-rich layers and two varieties of hydrothermal precipitates, an Fe-rich clay (nontronite), and Mn oxides. Compositionally, these sediments differ considerably from one another and vary in proportion down the core to basement. Geochemical partition studies on the sediments have indicated a variable fractionation of the elements that they contain between clays, carbonate, and Mn oxides with the hydrothermal phases being strongly depleted in trace elements. The hydrothermal sediments are thought to have been formed as a result of the fractional precipitation of Fe silicates and Mn oxides from ascending hydrothermal solutions which are depleted in trace elements due to subsurface precipitation of sulphides, and have had insufficient time to scavenge trace metals from sea water due to their rapid precipitation.

Hydrohonessite has a composition that can be expressed by the formula [Ni8-x2+Fex+(OH)16] [x/2SO2-4.yH2O.zNiSO4], where x is approximately 2.6, y is 7, and z is 1. The X-ray powder diffraction pattern has strongest lines at 11.0 (10), 5.56 (5), 3.68 (4), and 2.709 Å (3), and can be indexed on a hexagonal unit cell with a = 3.09 Å and c = 10.80 Å. It is optically uniaxial negative with ɛ = 1.59 and ω = 1.63; bright yellow in hand specimen and transmitted light. Hydrohonessite is the hydrated equivalent of honessite, and is related to carrboydite, motukoreaite, and mountkeithite. These minerals are related to brucite and pyroaurite-type minerals in that they have a layered structure consisting of brucite-like layers separated by about 7 Å of interlayer material consisting predominantly of water, but also containing anions that balance the charge of the brucite-like layer. Hydrohonessite is a secondary mineral resulting from the weathering of Ni-Fe sulphides, and appears to be stable between pH 6 and 7.

Honessite and hydrohonessite from Unst, Shetland, and Linden, Wisconsin, are sulphate-containing members of the pyroaurite group and are analogous to reevesite. The structural formula can be written (Ni5.55Mg0.10Fe2.353+)(OH)16(SO4)1.18.xH2O for the Unst material, with sulphate groups replacing the interlayer carbonate groups in reevesite. Microprobe analyses of the Unst material show amounts of SO3 in excess of that required to balance all trivalent cations and this may be due to the presence of an amorphous nickel sulphate material. We find no evidence for the presence of trivalent nickel in honessite.

The Unst material occurs in both 8.8 (honessite) and 11.1 Å (hydrohonessite) forms, depending on composition, humidity, and temperature, similar to carrboydite, motukoreaite, and SO4-exchanged takovite. The Wisconsin material occurs only with an 8.8 A basal spacing. The increase from the 7.7 Å basal spacing of reevesite is due to the difference in molecular geometry of carbonate and sulphate groups. The presence of the sulphate molecule is unambiguously shown by the infra-red spectra, and the observed bands are consistent with the sulphate groups lying in the interlayer with their trigonal axes parallel to c. The Unst material is found intimately mixed in variable proportions with reevesite, distributed in patches on chromite, and the mixed nature is revealed by X-ray powder diffraction, by variations in total SO3 content, and by the presence of absorptions due to both SO2-4 and CO2-3 in the infra-red spectra.

Mountkeithite [(Mg,Ni)9(Fe3+,Cr,Al)3(OH)24]3+[(CO{in3,SO4)1.5(Mg,Ni)2(SO4)2(H2O)11]3− occurs with pyrite, magnetite, magnesite, hexahydrite, and morenosite in a low-temperature hydrothermal vein system in the Mount Keith disseminated nickel sulphide deposit, Western Australia. Electron-microprobe analyses gave the following average composition: MgO 31.7, NiO 6.1, CuO 0.1, Cr2O3 6.1, Fe2O3 8.3, Al2O3 2.6, SO3 14.7; microanalysis for C and H gave CO2 3.9, H2O 30.6; total 104.1. The X-ray powder diffraction pattern has the following strongest lines: 11.30(10), 5.63(8), 4.63(2), 3.765(6), 2.645(4), 2.545(4), 1.554(5), and 1.505(3); these can be indexed on a hexagonal cell with a = 10.698 Å, c = 22.545 Å.

Mountkeithite occurs as friable aggregates and rosettes composed of soft, pearly to translucent, pale pink to white flakes with perfect basal cleavage. Optical properties are uniaxial negative with ɛ = 1.51 and ω = 1.52; weakly dichroic from colourless to palest pink. Specific gravity is 2.12 (calc. 1.95). Mountkeithite is soluble, with efferves-cence, in dilute HCl. Mountkeithite changes to a pyroaurite-like phase, with a 7.8.Å basal spacing when immersed in water, but can be changed back to an 11.3 Å phase by immersion in 1M MgSO4; mountkeithite also collapses to a 7.8 Å basal spacing under vacuum. Phases similar to mountkeithite have been synthesized by immersing grains of stichtite in 1M MgSO4, at 20, 50, 105, and 200°C.

The structure of mountkeithite is interpreted as consisting of positively charged brucite-like hydroxyl layers, containing both divalent and trivalent cations, separated by about 7 Å-thick, negatively charged interlayers containing carbonate and sulphate anions, magnesium sulphate, and water. Carrboydite, motukoreaite, and hydrohonessite are minerals with similarly large basal spacings of 10–11 Å, and are also believed to have pyroaurite-related structures with expanded interlayers.

In the extensive literature dealing with the growth twinning in crystals, two approaches have been followed. Friedel (1926) related twinning to a lattice common to both parts of the twin, while others have considered how the atoms at the junction must be packed to give substantially the same interatomic distance and bond angles as in the Z single crystal. The second approach has been adopted in this paper.

This investigation has been carried out by constructing ball-and-spoke models of the various twins. The balls constituting the boundary between the two parts of the twin have been drilled to accord with the full symmetry of the twinned crystal, while the rest of the model has the normal symmetry of the untwinned crystal. The orientation of the two sets of crystallographic axes is determined by the twin law. Only when it was possible to construct a model in which the environment of the boundary balls closely approximated to the normal environment, was the proposed atomic arrangement accepted.

Recent studies on the development of Sn-W mineralization in south-west England (Sheppard, 1977; Simpson et al., 1979; Jackson et al., 1979) indicate that the hydrothermal fluids dominantly consisted of local meteoric waters. These waters were heated up and drawn into the granite in a convective hydrothermal system. It is proposed that the fluids leached Sn and W from minerals such as the dark micas, during their passage through the granite. These metals were then deposited in veins at higher levels in the system.

The advantages of polished ultra-thin sections (PUTS) in the study of very fine-grained materials, such as occur in some meteorites, have been illustrated by Fredriksson et al. (1978) whose technique is based on the earlier work of Beauchamp and WiUiford (1974). An essential feature of such methods for friable and heterogeneous materials is the use of a medium, usually an epoxy resin, to consolidate and partially impregnate them. Normally one polished side of the specimen is bonded to a glass slide during preparation, and the finished PUTS are integral with the slide on completion. PUTS are typically 2-5 microns in thickness.

This report extends the work of Corrigan and Gibb (1979) who determined losses of Fe and Na from a typical basaltic liquid during 1-atmosphere experiments using the wire-loop method (utilizing both Pt and Ag40Pd60 wires) by electronmicroprobe analyses of the quenched glasses. They found that Na losses were much more serious than Fe losses, the former being in direct linear proportion to the length of the experiment and in inverse proportion to the mass of the sample, such that the Na loss was 25 % after 120 hours for samples of 60 mg. Even using samples with weights in excess of 200 mg would mean that Na losses would be unlikely to be less than 5 % except in experiments shorter than a few days or on samples with very low initial Na contents.