Nickelblödite, the nickel analogue of blödite, has been discovered in nickel mines at Kambalda and Carr Boyd Rocks in Western Australia. The Kambalda sample, found in an underground opening, has a composition corresponding to Na2·02(Ni0·79Mg0·14Fc0·05)(SO4)2·00·3·17H2O. The sample from Carr Boyd Rocks, collected from an open pit, is a more magnesian variety, with a composition corresponding to Na1·93 (Ni0·55Mg0·46Cu0·02Co0·02)(SO4)2·01·4·39H2O. The mineral occurs as a surface efflorescence on nickel-rich sulphide ore in both cases. Nickelblödite is light green and transparent, and occurs as tabular crystallites up to 150 μm in diameter. The Kambalda material is biaxially negative with 2V = 60–70° α = 1·513, β(calc) = 1·518 and γ = 1·520. D 2·43, indentation hardness VHN 139. The Carr Boyd material has lower refractive indices and hardness. Strongest lines of the indexed powder pattern (Kambalda sample) are 4·466(9), 4·193(7), 3·720(6), 3·223(10), 3·190(8), 2·589(6). These measurements conform to a monoclinic unit cell with α = 10·87, b = 8·07, c = 5·46Å, and β = 100·72°. The possibility of H3O+ substitution is discussed.

Electron-microprobe analyses of chromites from the Helgeland area, Norway, show that they are all low in zinc. Many chromites show physical properties identical with donathite (anisotropism; polar magnetism) but have cubic lattices. One exception, from Rødøya, shows an unusual deviation from cubic symmetry and that there may exist complex intergrowths of two different tetragonal lattices and a cubic lattice. The similarity between the chromites studied and donathite suggests that the chemistry and crystallography of that mineral may be more complex than is currently accepted.

The decomposition of chrysotile fibres heated in air has been studied in the range 100–1400°C by electron microscopy and infra-red absorption. The first observable change in the structure occurred at 580°C, where cavities started to open up between the (001) layers of chrysotile as the fibres were dehydrated, giving rise to strong low-angle diffraction. There was no evidence of any structure in the remaining material but some degree of the original atomic arrangement was preserved for the magnesium silicates, forsterite, and enstatite, later developed in certain preferred orientations. The manner of this crystallization was determined by the thermal treatment, for in samples held between 600°C and 800°C forsterite developed slowly with little further disruption of the fibre while above 800°C the remaining amorphous areas rapidly recrystallized to a mixture of forsterite and enstatite. It is suggested that the mechanisms described by other investigators to explain the development of forsterite in preferred orientations may serve simply to nucleate the crystallization and a similar mechanism to account for the nucleation of the enstatite crystallization is considered. At high temperatures a possible doubling of some of the lattice parameters of the silicates was observed.

Fragments of a hydrothermal deposit, partly moulded on pieces of basalt lava, were dredged from the edge of the median valley in the centre of the Gulf of Aden. The deposit consists of two main components, manganese oxide in the form of spongy brown to hard black lumps and coatings, and friable green massive smectite. There are smaller amounts of an iron-oxide component, and mixed manganese-oxide-smectite material. Contents of the less mobile trace elements in the smectite are too low for it to have formed by alteration of basalt glass, and it is interpreted as a direct precipitate from hydrothermal solution. None of the components is enriched in Cu or Zn, or contains appreciable amounts of sulphur. If there are sulphides in this deposit, they must lie beneath the parts sampled.

Pontlyfni and Mount Morris (Wisconsin) are briefly described. Chondrule structure is absent from both, the latter is coarsely crystalline and both are sulphide-rich and contain forsterite and enstatite. Kakangari has well-defined chondrules; its silicate and sulphide minerals are unequilibrated, but similar to those of Pont-lyfni and Mount Morris (Wisconsin). New bulk chemical analyses are presented; Pontlyfni has 33·84 % total Fe and 7·04 % S; Mount Morris (Wisconsin) has 19·88 % total Fe and 4·72 % S; and Kakangari has 22·79 % total Fe and 5·30 % S. These three meteorites, together with Winona, have ordinary chondritic Mg/Si ratios, which result in their having abundant forsterite; this distinguishes them from E-chondrites. The four meteorites have Mg/Si ratios lower than those of C-chondrites, and they are more reduced than C or ordinary chondrites. These four unusual stones, therefore, have some chemical similarities, for example all have Mg/Si (atomic) of about 0·95, mean olivine composition ranging from Fa1 to Fa5, and significant Cr in the sulphide. Cumberland Fails chondritic xenoliths and other exotic fragments in polymict meteorites may be related to the four stones, which may ultimately prove to belong to a distinct chemical group of chondrites related to the irons of Group IAB.

These meteorites are described and discussed together because they have similarities in mineralogy, bulk composition, and oxidation state. All have more than 10 % forsterite (i.e. an olivine in the range Fa0·10). In our study we investigated the mineral chemistry of, first, Kakangari (Graham and Hutchison, 1974), and then Mount Morris (Wisconsin), with cursory examination of the previously described Winona (Mason and Jarosewich, 1967) and Cumberland Falls chondritic fraction (Binns, 1969; Jarosewich, 1967). By chance, the British Museum (Natural History) acquired the Pontlyfni stone during the course of this work. Pontlyfni fell in Wales in 1931, but is so far undescribed; it proved to be chemically and mineralogically similar to Kakangari. We thank Mr. J. R. Owen, the original owner, for its timely addition to the Museum's Collection.

Three-dimensional counter-diffractometer data and a full-matrix least-squares method have been used to refine the crystal structure of a sub-silicic hastingsite in the space group C2/m. The chemical composition of the amphibole is with ceil parameters a = 9·8659(4), b = 18·0139(8), c = 5·3545(2)Å, and β = 105·082(1)°. Unit weights were used throughout the refinement and the final R-factor for 1263 observed non-equivalent reflections was 4.1%. The mean tetrahcdral bond length for aluminous amphiboles varies linearly with the total amount of tetrahedral A1. Curves are derived relating individual mean bond lengths with A1 occupancy. The positional disorder on the A-site of the aluminous amphiboles appears to be partly related to the amount of substitution of A1 into the T(2) tetrahedron.

Chloroxiphite, Pb3CuO2(OH)2Cl2, is monoclinic, space group P21/m, with a = 10·458, b = 5·750, c = 6·693 Å, β = 97·79°, Z = 2. The structure has been refined by the method of least squares from three-dimensional Mo-Kα intensity data to a conventional R value of 0·115. The structure consists of sheets of composition [Pb3 CuO2(OH)2]2−, in themselves made up of the layer sequence Pb-(O, OH, Cu)-Pb, lying parallel to (01). The structure contains eight-fold PbO4Cl4 and seven-fold PbO5Cl2 polyhedra common in other oxychlorides, Pb6 clusters and copper atoms in square planar, Cu(OH)4, coordination.

After approximately 50% of K from the 1–2 μm fraction of a naturally occurring microcrystalline muscovite was extracted with BaCl2 solution at 98°C electron-diffraction patterns showed a honeycomb-shaped diffuse scattering as well as intense diffuse streaks. The diffuse scattering became pronounced with increasing K-depletion. It may be interpreted in terms of a short-range order distribution of interlayer Ba ions, using an analogy to the work by Alcover et al. (1972, 1976). Since the diffuse streaks still remained after re-entry of K to the K-depleted specimen they might be related to structural disturbance, particularly in the basal oxygen arrangements, resulting from the K-extraction. Subsequent conversions of Ba-saturated vermiculitized muscovite into Na-, Mg-, and Ca-forms revealed distinct extra diffraction spots midway between two ordinary diffraction spots. They were broader than the ordinary spots and elongated along the three directions of [01]*, [11]*, and [11]*. The diffraction pattern containing the extra spots was, therefore, interpreted as a composite one made up by superimposing three independent patterns, each of which was composed of ordinary Bragg diffraction spots and extra diffraction maxima elongated in one of the three directions. This situation does not require a superlattice but modified the original face-centred lattice symmetry to a primitive lattice symmetry. However, the presence of a real superlattice should also be possible. The elongation and broadness of the extra diffraction maxima may be related to the orientation and size of coherent domains defined by a certain ordering of interlayer cationic distribution, which took place during the hydration processes including extraction and conversion.

The crystal structure of killalaite has been approximately determined, using a crystal from the original locality. The ionic constitution is Ca3+x (H1–2xSi2O7) (OH), with x ≈ 0·2 for the crystal studied. Crystal data are: monoclinic pseudo cell with a 6·807, b 15·459, c 6·811 Å, β 97·76°, space group P21/m, Z = 4, Dx-ray = 2·94 gcm-3. Additional, very weak reflections indicate a B-centred monoclinic true cell with doubled a and c. The structure was determined using only the pseudo cell reflections, and includes one fractionally occupied Ca site; the larger true cell possibly arises from ordering of Ca atoms in these sites, together with associated small shifts of other atoms. There are weak indications that the degree of occupancy of this site may be variable. The X-ray powder pattern has been indexed and a calculated pattern is also given. A synthetic compound called ‘Phase F’ by Aitken and Taylor (1960) appears identical with or closely similar to killalaite.

Paragonite has been found in two chloritoid-chlorite-mica schists from the Dalradian, north of Stonehaven. The principal bulk-chemical constraints on the occurrence of this mineral within the chloritoid zone appear to be the ratios of soda to potash and the Al2O3 contents of the rock. The composition of paragonite coexisting with muscovite indicates a crystallization temperature of 550 °C for rocks some way above the chloritoid isograd. Although this result is supported by other mineral equilibria, the muscovite compositions suggest a temperature of 690 °C which is too high because the mineral assemblage chloritoid-paragonite is unstable above about 600 °C By combining data on the breakdown of the chloritoid-paragonite assemblage with a mineral equilibrium for the assemblage garnet-plagioclase-muscovite-biotite, it has been deduced that rocks close to the staurolite isograd crystallized at about 7·5 kb and 600 °C.

New X-ray powderdata are presented for vandendriesscheite, fourmarierite, masuyite, and wölsendorfite. Literature data on the known U-Pb hydrated oxides is reviewed, and suitable criteria for their distinction are set forth.

Compositions of An61 and An69·5 have been determined for two domains in plagioclase from a gneiss. Optical properties indicate that the same compositions apply respectively for an intergrowth of ‘matrix’ and lamellae found in zones between the domains. The intergrowth appears to span the composition range between the Bøggild and Huttenlocher intergrowths.

An unusual tennantite, containing 4·64% Pb, has been found on a mineral specimen from Sark's Hope Mine collected in 1843. An electron-probe analysis is given, together with visible spectrum reflectance data, quantitative colour values, VHN, and cell dimensions determined from X-ray powder diffraction photographs.

Sea-water from Deception Island was found to contain 0·55–1·48 mg Mn 1−1 and 10·2-64·3 mg Si 1−1. Reaction of Deception Island basalt and sea-water at 190 °C and 500 bars simulated the measured water composition but leaching experiments suggest that much of the Mn etc. may derive from local volcanic ash. Mn and Mg in thermal waters at Reykjanes have concentrations compatible with equilibrium with components of montmorillonite present as an alteration mineral of average composition . However, Fe is controlled by equilibrium with sulphides. Flux calculations based on these and other data place upper limits on hydrothermal Mn input to the oceans of 5–36×1011 g yr−1 and Mn accumulation in metalliferous sediments of 2–9×1011 g yr−1.

Unusual replacement-type sulphide mineralization occurs in the northern part of the Bulong Complex, about 30 km east of Kalgoorlie, Western Australia. The mineralization is non-economic (up to 0·5% Ni, 0·4% Cu, 0·1% Co, and 8·6% S) and occurs in altered unlayered olivine-rich rocks immediately above a thin sheet-like inclusion of country rock. The supergene-modified primary opaque assemblage pyrrhotine-magnetite-pyrite-chalcopyrite-(pentlandite) is texturally unusual in that the opaques are largely pseudomorphic after primary olivine grains, mainly within irregular fracture networks in a manner similar to ‘early’ serpentine. Textural relations between opaques and silicates indicate that the mineralization was introduced during the early stages of serpentinization prior to the onset of deformation and regional metamorphism. Monoclinic pyrrhotine is the main opaque phase, with some grains containing relict cores of the hexagonal variety. Magnetite associated with the mineralization is Ni-poor (< 0·1% Ni) compared with ‘serpentinization magnetite’ from elsewhere in the Bulong Complex, which contains 0·5–0·8% Ni. As the mineralization was intersected at relatively shallow depth, supergene alteration effects are evident; pyrrhotine is locally altered to pyrite and marcasite (texturally and chemically distinct from the primary pyrite), and pentlandite is largely replaced by cupriferous violarite.

Textural features and consideration of phase relations in the system Cu-Fe-S-O suggest that the mineralization is of low-temperature (350± °C) hydrothermal origin. By contrast, the more commonly developed Fe-Ni-Cu sulphide mineralization of the Kalgoorlie region is generally considered to be of high-temperature (1200± °C) magmatic origin.

Bandy (1938) described as metavoltine? a mineral coming from Chuquicamata (Chile) with chemical composition K2Na6Fe3+(SO4)12(OH)2·20H2O. Owing to some differences between this mineral and that first described by Blaas (1883), doubts persist on the identity of these two compounds. X-ray analyses were performed to clarify the relation between them. An artificial product, closely related to metavoltine as regards lattice and chemical composition, was also studied, and affords a better understanding of the Na and K diadochy already observed in metavoltine.

Ktenasite from Glomsrudkollen zinc mine, Modum, Norway, is menoclinic, space group P21/c. The cell dimensions are a 5·598, b 6·121, c 23·762Å, β 95·55° The chemical formula is (Cu3·5Zn1·5)(SO4)2 (OH)6·6H2O with Z = 2; ρcale 2·96 g/cm3, ρobs 2·94 g/cm3. The mineral is biaxial negative with α (colourless) 1·574, β (bluish green) 1·615, γ (light green) 1·628, 2Vα 59°.

Ferrinatrite crystallizes in space group P, with a = 15·566(5), c = 8·69(1) Å, and Z = 6. The crystal structure was solved by three-dimensional Patterson and Fourier syntheses, and refined by least squares employing 2378 independent reflexions to a final R value of 0·068. The iron ions occupy special positions and are surrounded octahedrally by oxygen atoms. Fe3+O6 octahedra and SO4, tetrahedra are linked together to form infinite chains of Fe-O-S linkages in the [0001] direction. These chains are linked to each other by [NaO5(H2O)2] polyhedra and probably by hydrogen bonds. The topology of the arrangement is the same as that of the hypothetical P312 structure proposed by Moore and Araki (1974).

Spherulitic and felted green to brown minerals are described from various volcanic rocks of the Western Ochils, Stirlingshire. They fill amygdales and replace pyroxenes and appear to be an interlayered series of sheet silicates from celadonite to vermiculite. The minerals are hydrothermal in origin and suggest the presence of a potassium-rich hydrous fluid at a late stage in the volcanic history.

The compositions of apatite crystals from seven Skaergaard rocks were obtained by electron-probe microanalysis. These span the range of occurrence from the base to the top of the exposed zones of the layered series. Apatite occurs as an intercumulus phase in the lower zones, but becomes a cumulus phase at a structural height of 1850 m (c. 98 % crystallized), defining the base of Upper Zone b. All apatite analyses show a high F:Cl ratio. There is a slight but significant increase in F and decrease in Cl when the apatite becomes a cumulus phase, the F/Cl values changing from < 10 to > 30. Variations in F:Cl:OH are attributed to differential volatile migration from trapped intercumulus liquid sites. The apatite data provide new support for ferrodiorite-granophyre liquid immiscibility.