The Permian magmatic province of the Oslo rift, south-east Norway, includes large volumes of felsic and silicic rocks. Based on their geochemical character, these rocks may be divided into two main groups. The Larvik larvikites (monzonites) are highly enriched in large ion lithophile elements (LILE) (e.g. 10–32 ppm Th. 8–15 ppm Ta), and have an initial 87Sr/86Sr of 0.70391 ± 5. The syenites and granites have moderate to high concentrations of LILE (e.g. 7–88 ppm Th, 4–25 ppm Ta), and initial 87Sr/86Sr ratios between 0.705 and 0.707. The Larvik larvikites and extrusive equivalents (rhomb porphyry lavas) have similar initial Sr isotope ratios to uncontaminated basalts and gabbros in the rift, and are believed to have a mantle origin. The higher initial 87Sr/86Sr ratios in the silicic than in the felsic rocks reflect a crustal component representing the intermediate or low crust. After intrusion into the upper crust, the major and traceelement concentrations of the silicic magmas were modified through fractional crystallization dominated by removal of alkali feldspar, and transport of elements with a fluid phase. The silicic magmas appear not to have interacted significantly with the side rock at this stage.

The Middle Zone (MZ) of the Insch intrusion lies in a geographically intermediate position between the Lower Zone (LZ) and Upper Zone (UZ) cumulate sequences, but is more complicated than either in comprising two intricately associated, but texturally distinct, components, the MZ cumulates and the fine-grained granular gabbros (FGG). In addition, there is a minor group of porphyritic granular gabbros (PGG), containing abundant plagioclase phenocrysts. A further variety of gabbroic rock, namely quartz-biotite norite (QBN) occupies a similarly intermediate position in the Boganclogh extension of the main Insch mass. These principal components (MZ cumulates, FGG, and QBN) show considerable mineralogical overlap with each other and with the lower part of the UZ succession (UZa). Unlike the UZ rocks, they are essentially olivine-free (apart from the most basic granular gabbros) and consist mainly of plagioclase, orthopyroxene, and Ca-rich clinopyroxene, with accessory Fe-Tioxides and minor interstitial biotite and/or hornblende. Mineral compositions are in the range An70–55 (plagioclase), En71–44 (opx), and Ca46Mg42Fe12 to Ca45Mg30Fe25 (cpx), but in detail each rock group can be characterized mineralogically as well as texturally. It is concluded that the MZ cumulates and FGG (including PGG) are essentially complementary, formed in different locations, and under slightly different conditions, but in the same magma chamber. The complex relationships between them, and the apparently haphazard geographical variations in mineral compositions, may be the combined results of the magmatic events, possibly including the movement of large xenolithic fragments, and later block faulting. The substantial compositional overlap of the more evolved MZ cumulates by UZa is explained in terms of replenishment by magma of slightly more primitive (and potentially olivine-bearing) composition after the deposition of the MZ. The Boganclogh QBN is believed to represent a more hydrous fraction of the Insch MZ magma.

Roxbyite (Cu1.74–1.82S), a new mineral, was isolated from drill core from the Olympic Dam deposit, from copper flotation concentrates from the Olympic Dam pilot plant, and from reaction residues obtained by treating certain concentrates with sulphuric acid. Powder X-ray diffraction data and the composition of the mineral are similar to results reported previously for the unnamed mineral of composition Cu1.83S (with a trace of iron) from the El Teniente deposit in Chile, and to copper sulphides of composition CuxS, 1.75 ⩽ x ⩽ 1.86, prepared electrochemically by several workers. A Weissenberg single-crystal study shows roxbyite to be monoclinic with a = 53.79, b = 30.90, c = 13.36Å, β = 90.0° and space group alternatives C2/m, Cm or C2. Some iron in the structure may stabilize the mineral.

Ion microprobe analysis ofperovskite from kimberlites and alnöites permits the accurate determination of La, Ce, Pr, Nd, Sm, Eu, Gd, Dy, Er and Yb at the part per million level. Other rare earth elements (REE) are subject to interferences in the mass spectrum caused by matrix-derived molecular ions and cannot easily be determined with comparable precision. Chondrite-normalized plots of the ion probe REE data are smooth curves, confirming the superiority of this technique over electron microprobe methods at the levels of REE abundance found in these perovskites. The perovskites analysed contain between 2.8 and 7.1 wt. % REE oxides and are highly enriched in the light REE, having La/Yb ratios of 577–3229. These La/Yb ratios are not representative of the parental magmas but result from REE fractionation during crystallization of the perovskite. Parental magma La/Yb ratios are estimated to be of the order of 120–650. The lower La/Yb ratios (80–200) found for whole-rock kimberlites are considered to result from contamination by relatively heavy REE-enriched crustal material.

Margarite occurs in Ordovician black slate within the contact aureole of the Wyangala Batholith, in the Lachlan Fold Belt in New South Wales. This occurrence is the first described from New South Wales. It is a regional metamorphic mineral replacing chiastolitic andalusite, and contains up to 1.07% V2O3 and up to 0.37% Cr2O3. Vanadium and chromium here substitute for octahedral aluminium. Margarite is produced by a local reaction in which Ca and H2O are introduced into andalusite grains. There is a significant paragonite component in the margarite but negligible muscovite solid solution.

Sixteen fluorapatite specimens from regional skarns in granulite terrain were associated with Al-zoned diopside ± scapolite ± actinolite ± calcite(+ rare phlogopite). Apatite was low in Ce (ave. 0.19% Ce2O3) and enriched in LREE relative to HREE (La/Yb = 31 to 74 in 4 specimens). Some specimens showed small negative Eu anomalies and some crystals were zoned in REE. SrO averaged 0.36%. The mineral contained some carbonate (ave. 0.5% CO2 in 5 specimens), appreciable silica (ave. 0.5%), and variable sulphate (0.1 to 1.2% SO3). Excess charge due to S6+ was largely compensated by Si4+. Chlorine was minor and F accounted for 75–98% of the F, Cl and OH ions. Apatite from marble lacking amphiboles and pyroxenes has a similar chemical composition, but apatite from later carbonatite and fenite contains more Ce and Sr. Apatite from Gatineau fenite, Gatineau carbonatite and world-wide siliceous igneous rock generally contains less S. Apatite from Gatineau skarns normally contains more Cl and less S than that from phosphorite. Magnesian marble was silicated to skarn by reaction with siliceous gneiss. Phosphorus, REE, and Sr were removed from nearby rocks and transported in aqueous, carbonated solutions containing minor amounts of F, Cl and S at granulite-facies conditions. Apatite and calcite precipitation took place in skarns and marble during the Grenville (Proterozoic) orogeny.

Eight samples (seven volcanic rocks and one quartz sandstone) have been dredged from the Funk Seamount, 60 km NW of Marion Island in the southwest Indian Ocean (lat. 46° 15′S, long. 37° 20′ E). The volcanic rocks are fine-grained vesicular basanitoids and glass-rich volcanic breccias geochemically similar to the Marion Island lavas. Lavas and breccias contain a suite of megacryst minerals and of small polymineralic xenoliths, in both of which kaersutite is a prominent constituent.

The megacryst suite comprises large unzoned single grains of kaersutite, plagioclase, pyroxene, magnetite and ilmenite, all showing textural evidence of resorption/reaction with the basanitoid host. The megacrysts have a limited range of compositions except for the plagioclase which ranges from oligoclase to labradorite.

The small (2 mm to ∼ 3 cm) xenoliths are mostly two-pyroxene amphibole assemblages with or without olivine, magnetite, ilmenite, plagioclase and apatite. The xenoliths show some evidence of reaction with the basanitoid host and most have undergone recrystallization and/or localised decompression melting.

Xenolith and megacryst assemblages are interpreted as being associated with the formation and partial crystallization of a hydrous basanitoid melt at depth.

Ganophyllite and [Na+ + Al3+]-substituted tobermorite exhibited similar cation exchange properties and, in particular, selective cesium exchange. [Na+ + Al3+]-substituted tobermorite and ganophyllite showed a selective Cs exchange of 10.77 and 8.71 meq/100 g from 0.02N NaCl and of 11.08 and 9.04 meq/100 g from 0.02N CaCl2, respectively. The [Na+ + Al3+]-substituted tobermorite is structurally similar to ganophyllite: both are roughly analogous to 2:1 layer silicates, both have cross-linking tetrahedra across the interlayer region, and both have exchangeable cations located in zeolite-like sites in the interlayer region. The similarities in cation exchange properties imply that the residual charge configuration of the analogue 2 : 1 layer must be approximately equivalent also.

The crystal structure of gyrolite from Qarusait, Greenland, was solved and refined with the space group P and cell parameters a = 9.74(1), b = 9.74(1), c = 22.40(2)Å, α = 95.7(1)°, β = 91.5(1)°, γ = 120.0(1)°. The structure is built up by the stacking of the structural units already found in the crystal structure of reyerite (Merlino, 1972, 1988), namely tetrahedral sheets S1 and S2 and octahedral sheets O. The tetrahedral and octahedral sheets are connected by corner sharing to give rise to the complex layer which can be schematically described as 1OS2, where S2 and , as well as O and , are symmetry-related units. Successive complex layers with composition [Ca14Si23AlO60(OH)8]-5 are connected through an interlayer sheet made up by calcium and sodium cations and water molecules.

The unit cell content NaCa16Si23AlO60(OH)8·14H2O, determined by the structural study, was confirmed by a chemical analysis, apart from the indication of a somewhat larger water content. The crystal chemistry of gyrolite is discussed on the basis of the present structural results and the chemical data given in the literature for gyrolite from different localities: the crystal chemical formula which accounts for most gyrolite samples is Ca16Si24O60(OH)8·(14+x)H2O, with 0 ⩽ x ⩽ 3.

Stacking disorder, twinning and polytypic variants in gyrolite, as well as the structural relationships of gyrolite with truscottite, reyerite, fedorite and the synthetic phases K and Z are described and discussed.

Pottsite occurs in the oxide zone of a tungsten mine northwest of Potts, Lander County, Nevada, and is named for the locality. It occurs with scheelite, clinobisvanite, bismutite, and vanadinite. The colour is yellow, luster adamantine. Dobs. is 7.0, Dcalc. 7.31. Uniax (-) ω = 2.49, ɛ = 2.37. Microprobe analysis gave PbO 32.4, Bi2O3 34.0, V2O5 26.6; H2O by Penfield method 6.71% (total 99.71%), leading to HPbBi(VO4)2·2H2O. Crystals are tetragonal 14122, a = 11.084, c = 12.634Å. Strongest powder diffraction lines are 4.618 (9), 4.168 (3), 3.210 (4), 3.062 (10), 2.476 (4), 2.252 (3).