Pyroxenes and amphiboles from the under-saturated to over-saturated syenites of the Kangerdlugssuaq intrusion have been examined to see what light they throw on the two contrasting petrogenetic models for the intrusion. Pyroxene crystals are strongly zoned outwards from augitic to more acmitic compositions, with the most calcic cores present in the foyaite, contrary to the expected pattern if the foyaite is the most evolved rock type as previously postulated. Amphiboles, which are absent in the foyaites, show an exceptionally wide compositional field varying from actinolite through richterite and katophorite to arfvedsonite. Many are manganoan and potassian varieties. However, there is no consistent variation throughout the intrusion as previous work has suggested. These results are not favourable to the idea of a crystal fractionation model for the intrusion and we suggest that the foyaite is closest to the original magma which has hybridized with the enclosing gneisses and basalts to produce the over-saturated rocks. Such a model is consistent with the existing isotopic data.

Iron-rich (Fs:En ∼0.8) calcic pyroxenes that have been subjected to granulite-facies metamorphism contain up to seven generations of exsolution lamellae. They can be grouped into four stages. In stage 1 pigeonite exsolved parallel to ‘001’ and ‘100’ (where ‘hkl’ signifies ∼ (hkl)) and mostly inverted later to orthopyroxene. During stage 2 orthopyroxene exsolved parallel to (100), while during stage 3 orthopyroxene was quickly followed by metastable ‘001’ pigeonite. The stage 3 precipitates clearly grew and thickened together for some time. During stage 4 a ‘100’ pigeonite was exsolved. The stage 3 and 4 precipitates show evidence of reheating, dissolution and later, renewed growth. Sometimes orthopyroxenes of stage 3 have crossed a ‘001’ pigeonite lamella and caused it to invert by a shear mechanism.

Chemical analysis shows no rotation of the tie lines between Ca-rich and Ca-poor phases, in contrast to previous studies of Skaergaard and Bushveld pyroxenes. The geothermometers of Wood and Banno (1973) and Wells (1977) indicate solidus temperatures of about 850°C and 900°C respectively, but the geothermometers were found to be unsuitable for subsolidus conditions. We estimate the pressure to have been about 9 kbar during solidification. Estimates of nucleation temperatures obtained from the orientations of the exsolved lamellae (Robinson et al., 1977) were 850–700°C for stage 1, and 600–400°C for stage 3. We believe this geothermometer to be unreliable for the low temperatures involved in stage 4.

The low-symmetry forms of the clinopyroxene structure fit the rules proposed previously (Chisholm, 1981) but the low-symmetry forms of ortho- and proto-pyroxene do not. The rule governing the packing of the tetrahedral chains into layers is therefore revised to take into account the different chain configurations. Starting with simple sequences of. and × chains (Thompson, Hallimond Lecture 1979, unpubl.) a scheme is proposed for deriving most of the known pyribole structure types and their space groups.

Textural and chemical data are reviewed for amphiboles occurring in Cr-diopside series and Al-augite series xenoliths; as megacrysts in a variety of igneous environments; and also as minor phases in tectonic slices of upper-mantle peridotite. New data are given for pargasite and associated phases in a lherzolite from Tanzania and in kelyphite replacing garnet in a S. African kimberlite xenolith. It is concluded that only pargasites and titaniferous pargasites occurring in Cr-diopside series blocks have formed under upper-mantle conditions. Although amphibole is present in the upper mantle, as suggested by Oxburgh (1964), its relative paucity suggests it is not a major alkali and water storage site; phlogopite is a more likely candidate, particularly in the deeper parts of the upper mantle.

Amphibole in the lower parts of the Lilloise layered intrusion occurs interstitially and as a replacement of pyroxene; in the upper rocks it is a major cumulus phase. There is an overall trend of increasing Fe/(Fe + Mg) with height. Coupled substitutions which effect the variation in composition of the amphiboles are chiefly Na,K(A)+Al(T) for □A+SiT) and Ti+Al(T) for Fe3+(C)+ Si(T). There is considerable variation in composition both on the specimen scale and within individual grains. This variation, plus scatter found in plots of the coupled substitutions, is partly attributed to many of the amphiboles having replaced pyroxene and also to the effects of magmatic-hydrothermal fluids.

Orthoamphiboles from the cordierite-amphibole-bearing rocks at Outokumpu span a large part of the gedrite-anthophyllite solid-solution series. The amphiboles grew at temperatures near to or just above that of the solvus crest with post-metamorphic cooling resulting in sub-solvus exsolution, the distribution of which is compatible with an asymmetric solvus in which the anthophyllite limb is the steeper. Metamorphic conditions in the area imply that the crest of the solvus is unlikely to be at a temperature in excess of about 600 °C Electron microprobe traverses and Al-Kα scans show that the amphibole compositions vary rapidly both along and across grains. These variations can be explained by varying availability of Na controlling the extent of the edenitic substitution in the solid solution series. In this case this variability in Na availability may be explained by the external control of μNa through the fluid phase.

Infra-red spectra of gedrites have been measured in the hydroxyl region (3800–3500 cm−1). The spectra can be related to those of clinoamphiboles and of anthophyllites, and show the normal amphibole pattern of peaks whose intensities reflect the proportions of different cations in the adjacent octahedral sites. The absorption is weak especially in spectra of gedrites high in Al. Hyperfine splitting of peaks has been observed in anthophyllite spectra (Law, 1981) but the extent of hyperfine splitting in gedrite spectra is limited. The underlying baseline absorption, upon which the hydroxyl peaks are superimposed, becomes steeper and more awkward to estimate in gedrites. Some higher frequency peaks, probably arising from unit cells with A sites occupied by Na, are observed.

Chemistry of six white micas from the high-pressure garnet granulites of Jijal is presented, together with trace elements, refractive indices, and 2Vα for one. The micas are rich in the paragonite component (over 93%), with one analysis having the highest 100 Na/(Na + K) ratio (98.5) so far reported. The paragonite is not a stable mineral of the granulites but was overprinted on the granulites at about 505–540°C and 8–9 kbar (suggesting a depth of 30 km).

Green biotite-like material occurring in Haast Schist at Brighton, eastern Otago, consists of biotite and a kaolin-group mineral interlayered on a 1 μm scale. Electron probe analyses of composite grains show K2O contents of 4.4 to 8.5 wt.%, in part reflecting the kaolin content, and in part, leaching of K+ ions. Phengitic muscovite shows similar but less extreme effects. Kaolin, siderite, calcite, and titanium oxides have formed in the rock during hydrothermal alteration of epidote, sphene, and biotite by carbonated waters. Potassium leaching due to surficial weathering processes appears to be widespread in biotites from the schists of eastern Otago, and biotites (hydrobiotites?) with K2O as low as 4 wt. % are reported. Similar material with 0.8 to 2.4% K2O and with Na2O about 1% is thought to be more highly degraded biotite. Microprobe analyses of kaolin and chlorites from associated rocks are given in addition to those of ‘biotite’ and phengitic muscovite. Analyses of sheet biotite and sheet muscovite from the Mataketake Range pegmatites in the highest grade parts of the Haast Schist terrane are presented for comparison. The study shows that biotite is more widespread in eastern Otago than was previously thought, and that almandine-rich garnet is not confirmed down-grade of the first appearance of biotite in eastern Otago.

The Laytonville Quarry exotic block contains a series of deep-ocean sediments of varying Fe: Mn, Fe:Al and Fe2+ : Fe3+ contents, but all enriched in Fe-Mn carbonates. These sediments have a complex metamorphic history that has involved the early development of aegirine and garnet, followed by various phases of ‘secondary’ mineral growth, all of which required the addition of water to the metamorphic assemblages under high-P, low-T conditions. During the course of an early hydration episode deerite formed within most of the ironstones, and zussmanite in certain Fe-rich, Mn-poor pelites. At a later, and apparently lower pressure period in the metamorphic history, howieite, riebeckite, stilpnomelane, and a more manganous variety of zussmanite (ZU2) have overtaken much of the earlier mineralogy. Subsequent to this, and contemporaneous with the development of the tremolite-talc exotic block rind assemblage, there is a minor development of minnesotaite and ekmanite. During the course of the metamorphism the ironstones have become enriched in K and Na, necessary to form the blueschist meta-ironstone mineralogy. Such rocks therefore have the potential to reveal details of fluid expulsion within the subduction zone.

The distribution of various trace elements between the K-feldspar and albite phases of perthites from several pegmatites was determined by ion microprobe. Ranges in distribution coefficients (wt. % in K-feldspar/wt. % in albite) are: Li, 1.2–780; Mg, 0.2–1.1; P, 0.1–17; Ca, 0.02–1.6; Cs, 32–820; Ba, 24–284; Pb, 1.6–30; Fe, 0.3–0.7; Rb, 59–5505; Sr, 1.3–5.1. The trace elements are zoned within the K-feldspar lamellae and the profiles are interpreted as the result of cross coefficients in the diffusion matrix.

The morphology and structures of intergrowth precipitates in some Norwegian aventurinized plagioclases have been described in previous publications; data collected during an investigation of a wide variety of heat treatments upon the same specimens are now reported. In detail the behaviour of different feldspar crystals is highly individualistic, though there is a common pattern in sub-solidus reaction(s) with the feldspar; even if the identification of the products is uncertain, it is this process which causes the apparent ‘homogenization’ of the feldspar.

Some comparisons can be made between the present studies and the more limited experiments of earlier workers, and suggestions for reconciling experimental data are discussed. On the paragenesis of aventurinized feldspars, these heating experiments confirm that mechanism(s) are complex, but that exsolution from an originally iron-enriched feldspar by cooling in the solid state is unlikely to be of more than peripheral importance.

The structural relations of solid solutions in the series (Sr1−xBax)Al2O4 were studied using room- and high-temperature X-ray methods, infra-red spectroscopy, and DTA. At room temperature, SrAl2O4 and solid solutions with x up to 0.31 are monoclinic, between x = 0.31 and 0.43 monoclinic and hexagonal forms coexist, and between x = 0.43 and 1.0 only hexagonal forms occur. On heating, a member of the monoclinic series of solid solutions transforms to hexagonal symmetry over a range of temperature within which both monoclinic and hexagonal forms coexist. The proportion of the hexagonal form increases instantaneously as the temperature is raised. The transformation temperature decreases with increasing BaAl2O4 in solid solution and, in addition, the temperature width of the region of coexistence is markedly enlarged. SrAl2O4 transforms over the range 665–705 °C and (Sr0.7Ba0.3)Al2O4 over 170–405 °C. The DTA trace for SrAl2O4 shows a peak at 677 °C. On cooling, the transformations show hysteresis of 15 to 25 °C.

The coexisting monoclinic and hexagonal forms are believed to be isochemical, and discontinuities in cell parameters occur within the region of coexistence both in the compositional series at room temperature and in the elevated temperature transformation experiments. The low-to-high transformation is accompanied by a volume change of −0.2 to −0.3 %, and is believed to be first-order displacive with additional characteristics similar to those of martensitic transformations.

The thermal expansion behaviour of structures in the (Sr,Ba)Al2O4, series indicates that two tilt systems are operative: co-operative rotation of tetrahedra about the c-axis, and tilting of tetrahedra relative to the 0001 plane.

The results for the (Sr,Ba)Al2O4 series are shown to be invaluable in reinterpreting the structural behaviour of members of the nepheline and leucite groups of minerals.

Peridotite xenoliths 73–106 and 73–109 have coarse textures. All minerals in 73–106 are very Ti-rich, and include olivine (Fo 87.6), orthopyroxene (En 90.0), garnet (core TiO2 0.8, Cr2O3 7; rim 1.4, 4 wt. %), spinel (TiO2 5, Cr2O3 38), clinopyroxene (A12O3 1–8) secondary mica (TiO2 6, Cr2O3 2.5 near garnet 0.5 away from garnet), and serpentine (FeO 5–12) enclosing perovskite-rimmed ilmenite and minute apatites. All minerals in peridotite 73–109 are Ti-poor, and include olivine (Fo 92.6), orthopyroxene (En 93.6), garnet (Cr2O3 3.1–3.8), spinel (55), ureyitic diopside, primary mica (TiO2 0.06, Cr2O3 1.0, BaO 0.5, Cl 0.04), and serpentine (FeO 2–23). Various thermometers indicate ∼ 1350K (73–106) and 1100K (73–109). The low Al2O3 in the orthopyroxenes gives 39–47 kb for 73–109 from the Wood-Banno barometer. In 73–106, the spinel lies in secondary mica next to the garnet rim, whereas the spinel of 73–109 occurs in grains enclosed by garnet. The former assemblage indicates diffusion-dependent disequilibrium, whereas the latter is attributed to simultaneous growth of spinel and garnet, perhaps consequent upon exsolution from orthopyroxene. Complex behaviour was found for the Cr/Al distribution in published analyses of garnet and spinel. The 73–109 pair lies near the Thaba Putsoa trend, and the 73–106 assemblages are displaced towards the kelyphite region in which garnet and spinel have similar Cr/Al. Spinel and garnet may coexist over a 30 kb pressure interval ranging from ∼ 16 kb for low bulk Cr to ∼ 40 kb for high Cr.

Black, opaque grains of a spinel whose composition is (Mg2TiO4) 85.8, (MgFe2O4) 0.4 (FeFe2O4) 13.8 (mole %) coexist with a MgAl2O4 spinel and geikielite in a periclase-forsterite marble that has been thermally metamorphosed against an alkalic ultramafic intrusion of Caledonian age in the Kangerdlugssuaq region of East Greenland. The spinel appears to be the closest recorded approach to the end-member Mg2TiO4 among natural rocks, and to be part of a solid-solution series extending across the join Mg2TiO4-MgFe2O4-FeFe2O4, the existence of which has not previously been reported. The composition of the series appears to be controlled by the fO2 that prevails during metamorphism.

Electron probe microanalysis of tourmaline from late-stage aplites and mineralized (Sn-W) pegmatites from the Hub Kapong batholith of peninsular Thailand has revealed considerable chemical variation, and has shown that skeletal tourmaline crystals predominate within the pegmatites. The tourmalines show substitution towards the schorl end-member of the schorl-dravite series, as well as substitution towards alkali-free or proton deficient tourmaline. The extent of both substitutions increases (a) from aplites to pegmatites and (b) from core to rim within a single grain and so is believed to become greater with increasing degrees of differentiation of the aplite-pegmatite system. The development of skeletal crystals within the pegmatites and non-skeletal crystals within the aplites is believed to be due to the different kinetics of crystal growth obtaining within a hydrous, pegmatitic, fluid (viz. heterogeneous nucleation and comparatively rapid crystal growth rates) which in turn arise from differences in the structure of the aplitic magma and pegmatitic fluid. The observed chemical and morphological variation shown by the tourmalines is used to discuss the petrogenesis of the aplites and pegmatites, and associated ore deposits. During differentiation, the volatile content of the aplite magmas increased, leading to water saturation and the separation of a hydrous pegmatitic fluid. After a period of coexistence, possibly in chemical equilibrium, the crystallization of the pegmatites continued once the aplite magma had completely consolidated.

Astrophyllite has been found in alkalic pegmatite dykes on the two sides of Kangerdlugssuaq fjord. At the eastern locality, Kraemers Island, the dyke is silica oversaturated and the astrophyllite occurs in a broad tabular crystal habit. The dyke at the western locality, Bagnaesset, is silica undersaturated and the astrophyllite occurs in elongated prisms of acicular habit. Electron probe analyses of the crystals of the two habits have shown systematic chemical differences which not only relate to Mn and Ti compositional variations, but also reveal departures from ideal stoichiometry in varying degrees. Based on this latter feature a model involving polysomatism has been constructed which may explain the differences in crystal habit.

Svetlozarite, previously described as a new member of the mordenite group of zeolites, is reinterpreted as a multiply twinned and highly faulted dachiardite. X-ray diffraction study and transmission electron microscopy revealed (001) twins and (100) stacking faults. The former are attributed at least mainly to the insertion of a b glide plane between dachiardite sheets at each twin interface, which preserves the 5-ring linkages between the sheets. The latter are associated with loss of the C face-centring relation, and are attributed to replacement of shared 5-rings by linked 4-rings in the dachiardite sheets. Further study is needed to determine whether the structural faults in the dachiardite structures are chemically controlled.

Stanleyite, a cobalt-blue mineral, occurs as efflorescent outgrowths on patronite from Cerro de Pasco, Minasragra, Peru. The determined composition is very close to the ideal formula VOSO4.6H2O and yields an empirical formula of V1.01O1.01S0.99O3.99 5.78H2O on the basis of O = 5 in the anhydrous part. X-ray powder data (31 lines) indexed on an orthorhombic cell give a 12.12, b 9.71, and c 14.92Å, cell volume 1755.8Å3 and Z = 8. Refractive indices are α 1.505, β 1.519, and γ 1.533, D is 1.95 (measured) and 2.01 (calculated), H = 1−1½ (Mohs). Thermal analysis and X-ray powder data show stanleyite to be quite different from minasragrite (the pentahydrate) which it clearly mimics. Associated minerals are potassium alum, pyrite, and quartz.

A new kind of chromitite texture—a porphyritic nodular texture, found in an alpine-type complex in Pakistan, is described. The chromite from phenocrystic nodules and that from their disseminated chromitite groundmass show comparable trivalent elements and postcumulus to subsolidus Mg−Fe2+ variations, which are also recorded from ‘normal’ nodular ore samples. The nodular ores differ from all other textural types of chromitite from this complex in an unusual, though slight, increase in Cr/(Cr+Al) with Mg/(Mg+Fe2+) between stratigraphically different desposits, which is caused by magmatic differences. The implications of major element variations in chromites and olivines of nodular chromirites are discussed and a cumulus origin is supported. Temperatures of final chromite-olivine equilibration ranged from 750 to 900 °C the individual nominal values being higher for chromitite with coarser chromite crystals and, or, higher chromite:silicate ratio.