Olivine (ol) leucitite lavas from Northeast China contain hyalophanes with contrasting Ba-zoning patterns. The hyalophanes occur: (1) in a magmatic inclusion (DZ2n), consisting of barian-titanian phlogopite + hyalophane + leucite + sodalite, and (2) as a mantle surrounding a sanidine megacryst (DZ19). Hyalophanes contain 4.6-20.2 mol.% celsian (Cn). The DZ2n hyalophane is normally zoned with respect to Ba (decreasing Ba toward rim), whereas the hyalophane mantle of megacryst DZ19 is reversely zoned. DZ2n probably crystallised from an extensively evolved (>80 wt.% crystallisation) potassic melt derived from a primitive magma chemically similar to the host, at relatively low temperatures and pressures (e.g. T-720°C P∼2 kbar). Competition with barian-titanian phlogopite for Ba, and limited Ba supply from the residual melt are the main cause for the normal Ba zonation. Sanidine megacryst DZ19 originated as a high P-T product (e.g. T > 950°C P > 15 kbar) of an evolved leucite (lc) basanite/trachybasalt genetically related to ol-leucitites. Subsequently, it was incorporated into the host ol-leucititic magma at which point it was partially resorbed and then mantled by hyalophane. An increase in KBaaf/liq values with decreasing temperature and pressure and cocrystallisation of the hyalophane mantle with Ba-free phases may have caused the unusual reverse zonation.

Ultramafic to marie plutons in the Olyutor Range, North Kamchatka, represent the magmatic roots of a late Eocene arc, related to the westward subduction of the Komandorsky Basin beneath the Asian continental margin. Olyutor Range plutons are concentrically zoned with cumulate dunite cores mantled by a wehrlite-pyroxenite transitional zone and, in turn, by a narrow gabbroic rim.

Spinel is a common accessory mineral in these arc plutonics, and we present analyses of spinels from a range of lithologies. A continuous compositional trend is observed from Cr-spinel in the ultramafics to Cr-rich magnetite in marginal gabbros. Complex chemical zoning patterns within individual spinel grains suggest an interplay between fO2, fractionation, volatile content and subsequent sub-solidus reequilibration of spinel with co-existing silicates (mainly olivine).

In general, the spinels from magmatic arc environments are characterised by high total Fe and high Fe3+ contents compared to MORB and boninitic spinels and higher Cr-values relative to oceanic basin spinels. These differences imply a high oxygen fugacity during arc petrogenesis. Differences are also observed between plutonic spinels from arcs and low-Ti supra-subduction zone ophiolites. Low-Ti ophiolitic spinels are generally poorer in iron and richer in Cr, and hence are similar in composition and perhaps tectonic setting to fore-arc boninitic spinels.

Ammonium analyses of representative rocks from the Rosses granite complex (Ireland) and their constituent mineral show that any variation of magmatic origin has been completely obscured by pervasive hydrothermal alteration. The altered rocks are strongly enriched in the ammonium ion, most of which is held by chloritized biotite and sericitized plagioclase. The present ammonium content of the Rosses granites averages 13 ppm, but the original ammonium content of the magmas is estimated to have been less than 5 ppm. Potassium feldspars may preserve a better record of the original magmatic ammonium content than the whole rocks or the biotites.

Sulphide minerals and their analogues yield gases as a result of oxidation reactions. Even where sulphide minerals are in contact with mildly reducing groundwaters, S2- ions pass into solution and their dispersion patterns can be detected in soil as acid-released H2S. In more oxidising conditions, the metastable gases COS and CS2 are generated. Anomalous dispersion patterns of COS have been reported in soils above more than ten sulphide ore deposits, many of them concealed beneath transported exotic overburden. High concentrations of CS2 occur in the soils over several of the same deposits and uniquely reflect others. Anomalies of SO2 over sulphide deposits are confined to arid terrains. Certain anomalous dispersion patterns of arsenic and tellurium in soils are attributed to the generation and migration of unspecified gases from the oxidation of arsenide and telluride minerals.

Carbonado and yakutite are both porous aggregates of polycrystalline micrometre-size diamond, with very different characters from those of monocrystalline diamond. The genesis of carbonado is very controversial, whereas yakutite is thought to have been formed by meteorite impact. Neutron activation analyses of trace elements in carbonado and yakutite indicate that their rare earth element (REE) abundance patterns have common characteristics: heavy REEs are not much depleted and a negative Eu anomaly is observed. These patterns are quite different from those of kimberlite and monocrystalline diamond and are similar to those of crustal materials such as shale, supporting the hypothesis of a crustal origin for carbonado and yakutite.

This paper presents the results of microprobe investigations of the Platinum-Group Elements (PGE) of the Selukwe Subchamber, Great Dyke, Zimbabwe. The PGE are associated with base metal sulphides in the uppermost pyroxenites of the Ultramafic Sequence of the Great Dyke. The following minerals have been indentified: bismuthotellurides (moncheite, maslovite, michenerite, kotulskite and polarite); arsenides (sperrylite); and sulphides and sulpharsenides (cooperite, laurite, braggite and hollingworthite). Platinum Group Minerals (PGM) occur in three distinct textural environments: (1) at the boundary of sulphides and silicates/hydrosilicates, (2) entirely enclosed within sulphides, and (3) entirely enclosed within silicate or hydrosilicate minerals. The stratigraphic distribution, environments and textures of the PGM have important genetic implications, and cannot be explained by a single process. A multi-process model for the petrogenesis of the PGE mineralisation in terms of complexation and intermediate compound formation is proposed. The primary mineralising events were due to orthomagmatic processes, but the observed textures are the result of microscale remobilisation of PGM components by trapped interstitial fluids (bydromagmatic processes).

Reversed brackets on As and Sb partitioning between tetrahedrite-tennantite fahlores {∼CU10(Fe,Zn)2(Sb,As)4S13} and bournonite-seligmannite solid solutions {CuPb(Sb,As)S3} (400°C evacuated silica tubes, NH4Cl flux) indicate that the maximum nonidealities associated with the As-Sb substitution are about 165 (bournonite) and 250 (fahlore) cal/gfw on a per atom exchange basis. The experimental constraints are consistent with the following calibration of the As-Sb exchange reaction between these phases:

where kcal/gfw, X2 = Zn/(Zn + Fe) (or one-half the number of Zn atoms/formula unit) and X3 = As/(As + Sb) refer to atomic ratios in fahlore, and ΔG°23 = ΔH°23 = 2.59 ± 0.14 kcal/gfw (Raabe and Sack, 1984; Sack and Loucks, 1975; O'Leary and Sack, 1987; Sack, 1992).

Watanabeite Cu4(As,Sb)2S5, in which As>Sb, is a new copper sulphosalt that occurs with quartz in a hydrothermal vein at the Teine mine, Sapporo, Hokkaido, Japan. It is silvery lead-grey in colour with lead-grey streak. VHN100 = 253-306 kg/mm2, brittle. It has no cleavage and the fracture is uneven. The measured density = 4.66(2) g/cm3. The mean of six microprobe analyses gave Cu 41.1, Ag 0.1, Mn 0.3, As 15.4, Sb 14.3, Bi 2.4, S 26.2, a total of 99.8 wt.%, corresponding to: (Cu3.94Mn0.03− Ag0.01)∑3.98(As1.25Sb0.72Bi0.07)∑2.04S4.98 (basis: total atoms = 11), or ideally, Cu4(As,Sb)2S5, with As > Sb. The X-ray powder pattern resembles that of tetrahedrite but has subsidiary diffractions and is similar to that of synthetic Cu24As12S31 (Maske and Skinner, 1971). It is indexed on an orthorhombic cell with a = 14.51 Å, b = 13.30 Å, c = 17.96 Å (all ± 0.01 Å), and Z = 16. Calculated density is 4.66 g/ cm3. It is optically similar to tetrahedrite but is grey and weakly bireflectant. No internal reflections were observed. The maximum and minimum reflectance values in air and in oil (nD = 1.515) for the COM wavelengths are: 470 nm −32.5, 31.5; 17.7, 17.0, 546 nm −32.0, 31.1; 17.0, 16.3,589 nm −31.1, 30.3; 16.1 m 15.5, 650 nm −30.0, 29.3; 15.0, 14.5%, respectively. Watanabeite forms masses composed of aggregates of minute grains up to 50 μm in diameter. Apart from some minute inclusions of emplectite, native bismuth and tennantite, it is almost monominerallic.

Tausonites occurring in aegirine potassium feldspar syenite from the Little Murun potassic alkaline complex are characterised by complex growth and resorption features. These are attributed to nonequilibrium crystallisation in a dynamic environment characterised by magma mixing and/or volatile degassing. The crystals represent a transported assemblage which has no simple relationship to the magma from which their host rock crystallised. Tausonites exhibit significant normal, reverse and oscillatory compositional zoning with respect to strontium and the rare earth elements. The compositions are unique, ranging from tausonite to strontian cerian loparite, and are unlike those found in strontian perovskites from lamproites and fenites associated with carbonatites.

Compositional data are presented for Ba-rich titanates belonging to the hollandite group, Ba(Ti,Fe)5O11 and K2Ti13O27. The titanates, on the basis of textural evidence, are interpreted to have formed by the reaction of K-Ba-rich residual liquids with pre-existing Ti-magnetite, ilmenite and tausonite. The compositions of the titanates are similar to those of primary titanates found in kimberlites and carbonatites.

Brianyoungite, which is chemically and structurally related to hydrozincite, occurs as white rosettes (<100 μm) with gypsum on rubbly limestone within the oxidised zone at Brownley Hill Mine, Nenthead, Cumbria. The mineral contains (wt.%) 71.47 ZnO, 9.90 CO2, 6.62 SO3 and 10.70 H2O+. Based on 29 oxygen atoms, the empirical formula is Zn11.73[(CO3)3.00,(SO4)1.10]4.10(OH)15.88 or ideally Zn3(CO3,SO4)(OH)4. Brianyoungite is either orthorhombic or monoclinic with β very close to 90° Cell parameters determined by electron diffraction and refined from X-ray powder diffraction data are a = 15.724, b = 6.256 and c = 5.427 Å. Density is > 3.93, < 4.09 g/cm3 (meas.) and 4.11 g/cm3 (calc.); Z = 4. Thermogravimetric analysis, IR and XRD powder data (23 lines) are presented.

A Mössbauer investigation has been carried out on garnets from the Ivrea-Verbano zone and the results are compared with those obtained on the same samples by X-ray absorption spectroscopy (XAS). The problem addressed is the precise structural characterisation of the local environment of iron in garnets with Fe/Ca ratio variable between ∼3.0 and ∼18.0. Ferric iron is octahedrally coordinated and ferrous iron is in the dodecahedral site in all the samples. Mössbauer results are in agreement with those obtained by XAS and show that, at least in the compositional range of the garnets examined (0.1-0.5 calcium atoms p.f.u.), the iron environment is not significantly modified by the larger calcium cations sharing the same dodecahedral site. It is confirmed that the Mössbauer technique is more sensitive than XAS in detecting low percentages of iron, especially when the cation is present in more than one oxidation state and coordination number.

The structure refinement of a hastingsite (Cl 0.91 p.f .u.) from West Ongul Island, Ltitzow-Holm Bay, East Antarctica, has been carried out in order to characterise the effect of CI on the crystal structure. The composition of the Cl-rich hastingsite is K0.69Na0.26Ca1.99Mg1.09Fe2+2.71Fe3+0.92Mn0.04Ti0.10−Al2.31Si5.83O22OH0.99Cl0.91F0.10 with unit cell parameters a 9.962(3), b 18.283(7), c 5.572(1) Å and β 104.87(3)°.

In the Cl-rich hastingsite, A1, Fe3+ and Ti occupy only the M(2) site, and Fe2+ strongly prefers the M(1) and M(3) sites to the M(2) site with respect to Mg. The Mg-Fe2+ distribution suggests that the Mg-CI avoidance is realised in the intracrystalline exchange reaction among the octahedral sites in the Cl-rich hastingsite. The individual position of CI and OH was refined and the effective ionic radius of Cl in fourfold coordination is determined as 1.79 Å. Substitution of Cl for OH expands the sizes of the M(1) and M(3) sites and extends the double chains in comparison with hydroxy calcic amphiboles. Cl especially, is close to the 0(6) and 0(7) oxygens beyond the minimum distance calculated from effective ionic radii. Therefore, the Cl-rich hastingsite is considered to be deformed locally by Cl.

The Or-rich part of optically blebby to lamellar mesoperthite crystals from an Adirondack granulite has been shown by TEM to be a lamellar cryptoperthite, composed dominantly of tweed orthoclase. A fluid-absent, two-stage thermal history is proposed to explain the coarse and fine textures, with the cryptoperthite forming by coherent exsolution below ∼350°C probably during uplift. The mechanism was most probably homogeneous coherent nucleation rather than spinodal decomposition. However, cutting the orthoclase cryptoperthite are thin (<1 μm) seams of tartan microcline with sharp boundaries, often defined locally by {110} planes, and micropores. The microcline has replaced orthoclase by solution-redeposition along narrow planes infiltrated by fluid during minor retrogression at T < 350°C Solution-redeposition is a common process in feldspars at T < 500°C potentially accompanied by 18O exchange, because release of elastic strain energy in coherent perthite lamellar boundaries and twin-domain walls, followed by Si, Al ordering, provide driving forces for dissolution and reprecipitation of unstrained, more ordered phases.

Oriented inclusions of monazite occur in the dark core of apatite porphyroblasts in a muscovite schist from the Archaean Hemlo gold deposit, Ontario, Canada. The monazite inclusions are elongated along the b-axis and parallel to the c-axis of the apatite host; the complete orientation relationship of the monazite/apatite intergrowth is bMnz//cAp, cMnz//aAp. From analysis by SIMS and EMP, the dark core of the apatite porphyroblasts is depleted in LREE (LaN/YbN = 0.56). The monazite inclusions are correspondingly enriched in LREE, but markedly depleted in HREE, compared with monazite grains in the rock matrix and cross-cutting veins. The monazite inclusions precipitated by oriented reaction through rock-fluid interactions during a late hydrothermal alteration. Their unusual REE composition is probably related to both a preferential leaching of LREE from the dark core and a selective transfer of HREE out of the apatite porphyroblasts.

Twenty-four apatite (Ap) samples mainly from carbonatite and alkaline rocks were studied by electron microprobe, IR spectroscopy and X-ray powder diffraction. The crystal structures of six were refined using single crystal X-ray diffraction data to R = 1.7-2.5%. The generally high Si content of Ap from carbonatite and alkaline rock has been related to the presence of characteristic Si-O absorptions in IR spectra. Bands, whose intensities change with Si content, were observed at 520, 650, 930 and 1160 cm-1. The IR absorption features of v3 CO3 mode of Ap from carbonatite are different from those of v3 CO3 mode of Ap from sedimentary rock. This phenomenon is probably due to the different effects of F and OH on the CO3 substitution for PO4. The structural refinements yield more information on the CO3=PO4 substitution, which is now supported also by the geometrical evolution of the tetrahedron with increasing CO3 content: the tetrahedral size decreases and the angle distortion increases with C-content. It is likely that the triangular planar CO3 group is disordered on the four faces of PO43-tetrahedron. It was observed also that Ap from early-stage carbonatite is OH-dominant with considerable LREE, Si, CO3 and negligible Mn, Fe, Mg, K, S and C1 contents. They have high Sr/Mn, Si/S and C/S ratios.

Caryinite from Långban, Sweden is associated with barite, richterite, manganberzeliite, hedyphane, sarkinite or eveite, and an exsolved fermorite-like mineral. The crystal structure of caryinite, space group I2/a, a 6.855(2), b 13.147(3), c 11.479 (4) Å, β 98.97°, V 1022.0(5) Å3, has been refined to a conventional R = 3.5% using 1295 observed [F > 3α(F)] reflections. Caryinite is isostructural with alluaudite and shows the following key features: (1) Pb is ordered at X2, Mg is ordered at M2, Mn is disordered over M1 and M2, Ca is disordered over M1 and X1 and Na is disordered over X1 and X2, and (2) the X1 and 02 atoms of the X1 polyhedron show positional disorder. With the above site preferences taken into consideration, the assignment rules of Moore and Ito (1979) can now be used to accurately predict site occupancies for caryinites and arseniopleites using chemical data only. On the basis of the general structural formula X2X1M1M22[AsO4]3 (Z = 4), caryinite is (Na, Pb) (Ca, Na) Ca (Mn, Mg)2 [ASO4]3; the analogous formula for arseniopleite is Na(Ca,Na)Mn(Mn,Mg)2[AsO4]3. Carinite has Ca dominant at M1; by analogy, arseniopleite should have Mn dominant at M1.

The structure of chrysotile, an important asbestos mineral, consists of layers curled concentrically or spirally into tubes. Published transmission electron microscope (TEM) images suggest that successive layers are generally stacked out of register with one another. However, a lattice model can be constructed in such a way that each layer can be stacked in register with the next, five times around the circumference, so that the structure exhibits a global 5-fold symmetry. We report here TEM observations confirming that chrysotile asbestos can form crystals with this structure, the first observation of five-fold symmetry in a natural crystalline material.