Volume 86, Issue 10

The garnet-Al silicate-plagioclase (GASP) geobarometer has been recalibrated using four recent garnet activity models, four analogous garnet-biotite temperature models, and two recent plagioclase activity models. A typical sillimanite-bearing sample that formed at about 5.25 kbar, 575 °C shows a possible P range of ~0.7 kbar due to T error, ~1 kbar due to range of garnet activity model, ~0.9 kbar due to range of plagioclase activity model, and ~5.4 kbar due to range of experimental end-member reversals extended by one sigma. Calibrations were further constrained with the kyanite-sillimanite (K-S) phase boundary such that the best fit of 76 pelitic schist samples from 11 localities provides an individual end-member calibration for each of the eight possible combinations of garnet and plagioclase activity models with the appropriate geothermometer. Samples with low grossular or anorthite component were rejected. The end-member calibrations are constrained to pass through the the best-determined portion of the GASP experimental reversals at 1230 °C, 26.6 kbar. These individual end-member calibrations provide self-consistent models that tend to compensate for error in the garnet and plagioclase activity expressions. The models were also tested on a set of 59 samples from the Alps. The recommended calibration is the average garnet activity model and average garnet-biotite T model of Holdaway (2000), the Fuhrman and Lindsley (1988) plagioclase activity model, and H Grs = -6628521, S Grs = 258.76 to combine with the remaining phases in the Berman database to produce the optimum end-member GASP curve. These thermodynamic data are for the GASP geobarometer only. Error is about ±0.8 kbar absolute and about ±0.6 kbar relative. Geological error is the largest component of error in many of these samples. Care should be taken to be sure that analyzed plagioclase and biotite are near analyzed garnet, that the peak-T portions of garnet and plagioclase are selected, that the peak-T Al silicate is determined, and that the T calculated is the most accurate possible. These calibrations represent an improvement over previous published GASP calibrations. These eight models are available for distribution as three programs (T, P, P-T intersection) for the DOS-based personal computer.

Blue-green nickeloan tourmaline from a micaceous enclave of a marble from Samos, Greece, contains unusually high concentrations of Ni (up to 3.5 wt% NiO), Co (up to 1.3 wt% CoO), and Zn (up to 0.8 wt% ZnO). The polymetamorphic karstbauxite sample has an uncommon assemblage of nickeloan tourmaline, calcite, zincian staurolite, gahnite, zincohögbomite, diaspore, muscovite, paragonite, and rutile. The complex geologic history is reflected in multi-staged tourmaline growth, with cores that represent detrital fragments surrounded by two-staged metamorphic overgrowths. Zone-1 metamorphic overgrowths, which nucleated next to detrital cores, are highly asymmetric and exhibit compositional polarity such that narrow overgrowths of brown schorl developed at the (-) c-pole are enriched in Mg, Ti, and F, and depleted in Al, Fe, and X-site vacancies ( x ⃞ ) relative to wider, gray-blue schorl-to-foitite overgrowths developed at the (+) c-pole. Volumetrically dominant Zone-2 overgrowths are strongly zoned nickeloan dravites with a continuous increase in Mg, Co, Ca, and F at the expense of Fe, Zn, Cr, and V from the Zone-1 interface to the outermost rim. Within Zone 2, Ni reaches a maximum of 0.5 apfu before decreasing in the outer 20-40 µm. Zone-2 overgrowths also exhibit compositional polarity such that, at the (-) c-pole, overgrowths are enriched in Mg, F, Na, Ca, and Cr relative to overgrowths at the (+) c-pole that are, in turn, enriched in Al, Fe, Ni, Co, and X ⃞ . Element partitioning involving tourmaline rims and coexisting minerals indicates that relative partitioning of Ni is tourmaline >> staurolite > gahnite; Co is tourmaline > staurolite > gahnite; and Zn is gahnite > staurolite >> tourmaline.

Small irregular to lens-like occurrences (maximum 0.5 × 1 m) of apatite associated with magnetite and silicates are present in the Agoriani area of the Othrys ophiolite complex, central Greece. The Ano Agoriani area is dominated by peridotite (plagioclase lherzolite) of the mantle sequence, intruded by irregular bodies, dikes, and veins of gabbro, as well as by dikes of pyroxenite and pegmatitic gabbro. Apatite occurs as large (up to 3 cm long) well-formed crystals associated with magnetite. The aggregates consist predominantly of apatite, or massive magnetite with subordinate amounts of apatite. Apatite may also be accompanied by silicate minerals, mainly chlorite and lesser amounts of serpentine, tremolite, and Ni-silicates (nepouite, pimelite), and by Ni-sulfides (pentlandite, violarite, heazlewoodite). The apatite-magnetite association from the Agoriani area differs from nelsonite hosted by anorthosite suites with respect to: (1) the host rock type (ophiolites); (2) the highly variable proportion between apatite-magnetite; (3) the large size (up to 3 cm) of the apatite crystals; (4) the lack of fluorine (<20 ppm F) in the apatite; (5) the presence of abundant liquid-rich, fluid inclusions in apatite, and (6) the lack of ilmenite. The high V content (700-1000 ppm) of the magnetite from Agoriani differs from that of disseminated Fe-Ti mineralization (Ti-magnetite, ilmenite) in the magmatic sequence of ophiolite complexes (mainly hosted in gabbronorites), as well as the pure, massive magnetite associated with Fe-Ni-Cu-Co sulfides found in shear zones in ophiolite complexes. The composition of the apatite (chlor-hydroxylapatite), the presence of abundant primary two-phase aqueous fluid inclusions in the apatite, and the composition of the associated magnetite and sulfides, suggest that a hydrothermal system played an essential role in the formation of these deposits.

Absrtract More than 120 coesite and few polycrystalline quartz (coesite pseudomorph) inclusions were identified by Raman microspectroscopy in zircons of various kinds from gneissic rocks located in the Dabieshan and Sulu ultrahigh-pressure (UHP) metamorphic terranes in eastern China. The coesite inclusions have undergone, to various degrees, the coesite-quartz transition. Raman spectra of coesite and subsidiary quartz inclusions show various shifts, which are closely correlated to the extent of the transition. The coesite-pseudomorph inclusions in zircon show the highest Raman shift. Calibrations based on the experimental work of Hemley (1987) yielded inconsistent present-day overpressures (0-24 kbar) in coesite inclusions inside zircons. The availability of fluids, which resulted in the pervasive regional retrogression of the UHP gneisses during later exhumation stages, is regarded as the main factor controlling the extent of the coesite-quartz transition. This study underscores the need for the elastic model applied by previous researchers to explain the preservation of the coesite inclusions in rigid host minerals.

Mineralogy and geochemistry of a sulfuric acid spring water with a pH of 3.37 to 2.89 were investigated to verify the formation processes of iron minerals and the effects of bacteria on their formation. To estimate the solubility of schwertmannite, experimental dissolution in 10.0 mM H 2 SO 4 was conducted and this solubility data was used for geochemical modeling. Experimental incubation of the spring water containing bacteria was also performed and compared with a simulated abiotic system to evaluate the role of bacteria in the mineral formation. The spring water seeps through cracks of hydrothermally altered andesitic rocks containing pyrite, and precipitates schwertmannite and jarosite. Schwertmannite appears as a film-like thin layer floating on the water surface and composed of aggregates of spherical particles with diameters of 1 to 5 µm. Jarosite is produced as a precipitate on submerged rock surfaces. The precipitate contains well crystallized jarosite spheres 5 to 10 µm in diameter. Some ellipsoidal to rod shaped bacteria covered or decorated by poorly ordered iron minerals are also present in close association with the schwertmannite spheres. Results of the experimental incubation demonstrate that the oxidation rates of Fe 2+ are 5.3 × 10 3 to 7.2 × 10 3 times greater than those of the simulated abiotic system, suggesting that the formation of the iron minerals is promoted by bacterial oxidation of Fe 2+ . The dissolution experiment indicates that the solubility product of the schwertmannite having an average chemical composition of Fe 8 O 8 (OH) 5.9 (SO 4 )1.05 is approximately log K S = 7.06 ± 0.09. Using this data, geochemical modeling reveals that the spring water is supersaturated with respect to schwertmannite and also goethite and jarosite, but undersaturated with respect to ferrihydrite. Additionally, it is confirmed that the bulk solution chemistry deviates slightly into the stability field of goethite rather than jarosite. This suggests that the aquatic environments in contact with the rock surfaces may be more acidic and/or enriched in SO 4 2- relative to the bulk solution, which may eventually lead to the formation of jarosite instead of goethite.

The mineral lawsonite, CaAl 2 Si 2 O 7 (OH) 2 ⋅H 2 O, was deuterated to about 93% in a thermobalance by H 2 O-D 2 O gas-phase exchange at temperatures between 375 and 425 °C. The kinetics of this reaction are reported and diffusion coefficients for the H 2 O-D 2 O exchange are calculated from thermogravimetric measurements.

The non-convergent ordering of Mg and Ni over the M1 and M2 sites of synthetic olivines has been studied using “time of night” neutron powder diffraction and X-ray absorption spectroscopy (EXAFS). The compositional dependence of order/disorder at room temperature was established for solid solutions of general formula (Mg 1-x Ni x ) 2 SiO 4 . where X= 0.15, 0.2, 0.25,0.3, 0.5, and 0.8 atoms Ni (X Ni ; i.e.. mole fraction of Ni-olivine end-member). Ni orders into M1 with K D = (Ni/Mg in M1)/ (Ni/Mg in M2) reaching a maximum of 9.5 at a composition of Mg 1.0 Ni 0.4 SiO 4 . The temperature dependence of order/disorder at up to 1100 °C was determined for two samples (X Ni = 0.2 and 0.5). Between about 600 and 750 °C the samples show an increase in order due to kinetic effects, while above 750 °C the samples show a progressive decrease in order and describe an equilibrium disor­dering path. Equilibrium data define a Ni-Mg, M1-M2 intersite exchange energy of 21.5 ± 1.9 kJ/ mol. On cooling, the blocking temperature for cation exchange is about 800 °C. The kinetics of disordering behavior were analyzed using a Ginzburg-Landau model giving acti­vation energies for Mg-Ni exchange between M1 and M2 for samples of composition Mg 1.6 Ni 0.4 SiO 4 and Mg 1.0 Ni 1.0 SiO 4 of 145 ± 5 and 160 ± 5 kJ/mol. respectively. The model also shows that the characteristic time scale for re-equilibration of M1-M2 order decreases from around 2.5 s at 1000° to 0.03 s at 1300 °C. This points to the inapplicability of intracrystalline Ni-Mg partitioning for obtain­ing geothermometry and geospeedometry information for magmatic conditions. Ni X-edge EXAFS data show that samples with X Ni = 0.15.0.2,0.25 and 0.3 all show Ni clustering on adjacent M1 sites, indicating the presence of domains of Ni-rich and Mg-rich regions on a nanolength scale of < 10 A. These “precipitates” are at least an order of magnitude too small to be detectable by neutron powder diffraction. We suggest that the elastic strain at the interfaces between the Ni-rich precipitates and the Mg-rich matrix is responsible for the plateau or possible maximum in the b unit-cell parameter as a function of composition across the solid solution, which is observed at a composition of Mg 1.6 Ni 0.4 SiO 4 at room temperature. Comparison of our data with earlier studies at high P and T on Mg 1.0 Ni 1.0 SiO 4 olivine suggests that the effect of P is to increase the degree of order of Ni into M1 and to slow down the kinetics of intersite exchange with a ΔV disorder of 0.039 J/bar.

The damped harmonic oscillator model for thermal conductivity of insulators is improved, lead­ing to a formula that predicts thermal conductivity at ambient conditions (k 0 ) from various physical properties, most of which are commonly measured. Specifically, k 0 = [ρ/(3ZM)] C v [(u P + u s )/2] 2 / <Γ>, where ρ is density, Z is the number of formula units in the primitive unit cell, M is the molar weight, C v is heat capacity, u is the sound speed (P denotes compression; S denotes shear), and <Γ> is the average of the damping coefficients determined from peak widths in infrared reflectivity spec­tra, or from suitable Raman and Brillouin spectra. The classical physics and quantum-mechanical basis for this model is discussed, with emphasis on the effect of phonon-phonon interactions on mode properties. The calculated values of k 0 all lie within the experimental uncertainty of the mea­surements for all samples with the spinel or olivine structure examined by Horai (1971) with known or approximately correct chemical compositions. Other divergent measurements of k for MgAl 2 O 4 are discounted for various reasons. Early studies of Fe-bearing spinels are not generally reliable, but rough estimates from the above equation are consistent with all data, and good agreement is ob­tained for samples such as Mg 0.5 Fe 0.5 Al 2 O 4 and γ-Fe 2 SiO 4 for which the previous authors obtained chemical data, and for which IR reflectivity data exist. The theory reproduces the measured depen­dence of k 0 on composition and structure. Anisotropy in k 0 results mainly from differences in lattice constants (j): the equation for olivine is kj/k 0 =(V ⅓ /j) 0.73 which predicts the ratios within 3%. For solid solutions between Fe and Mg, the model provides a non-linear dependence of k 0 on mol% Fe, with the damping coefficient being the key factor producing non-linearity. Predicted ambient values are 11.3 ± 0.4 W/m-K for γ-Mg 2 SiO 4 , 6.5 ± 0.7 W/m-K for γ-Mg 12 Fe 0.8 SiO 4 , and 6.9 ± 0.3 W/m-K for β-Mg 2 SiO 4 . The high k 0 for ringwoodite suggests that heat in Earth’s transition zone should be conducted twice as efficiently as in the adjacent upper and lower mantles: this discontinuous depth dependence of k could impact thermal models of conduction in subducting slabs and of mantle convection

The elasticities of six polycrystalline silicate garnets (almandine, grossular, pyrope, uvarovite, andradite, and Prp 25 Alm 56 Spe 19 ) have been experimentally studied at pressures up to 3.0 GPa using a phase comparison method with an ultrasonic interferometer in a liquid cell piston-cylinder apparatus. Complete elasticity data sets (P- and S-wave velocities, bulk moduli K s , shear moduli G, and their first pressure derivatives K S ′ and G′) have been obtained for all six garnets, and are used together with an up-to-date compilation of garnet elasticity data to examine composition-elasticity systematics of garnets. Our results suggest that pyralspite and ugrandite have different relationships between bulk sound velocity (V φ ) and mean atomic weight (M̅ 0 ), between Poisson’s ratio (σ) and density (ρ), and between G/K s and K S /ρ ratios. A large error may occur when the systematics are applied across different garnet groups.

Six garnet crystals with compositions close to the uvarovite-grossular binary from three different localities (Saranov, Veselovsk, and Saranka, Ural Mountains, Russia) were investigated by optical methods, electron microprobe analysis, and polarized UV-VIS-IR micro-spectroscopy. Under crossed polarizers the crystals display “strong” birefringence in the order of 0.006 and formation of domains, aligned parallel to the crystal faces of the form {110}. Referring to the cubic cell, Y is always parallel to a principal lattice direction, e.g., [010]cub, whereas X and Z are then parallel to [101]cub and [101̅ ]cub, respectively. The garnets are uvarovite-grossular solid solutions with 48 to 71 mol% uvarovite component and traces of Ti, Fe, Mn, and Mg, and exhibit no compositional zoning. Polarized single crystal infrared absorption spectra in the region of the OH stretching vibration display 14 discernible, partly superimposing, and strongly anisotropic absorption bands between 3470 and 3680 cm -1 , which are caused by structurally incorporated hydroxyl groups. The polarization behavior observed for individual samples complies with orthorhombic, monoclinic, and triclinic crystal symmetry, respectively. The water content calculated using an integral extinction coefficient, ε = 43800 L/cm 2 ·mol, ranges from 0.07 to 0.34 wt%. Annealing experiments indicate that the incorporation of hydroxyl groups is not the primary cause for the anisotropic behavior. Instead, cation ordering on octahedral positions may play an important role. This assumption is substantiated by single crystal X-ray diffraction investigations reported in Part II of the present study (Wildner and Andrut 2001). Contrary to the IR spectroscopic data, single-crystal electronic absorption spectra show no significant polarization behavior in the UV-VIS-NIR spectral region. The observed absorption bands are caused by d-d transitions of Cr 3+ cations in octahedral coordination. The linear extinction coefficients of the investigated uvarovites strongly indicate that the Cr 3+ cations occupy centrosymmetric sites, regardless of the exact symmetry reduction from Ia3̅ d.

The crystal structures of six birefringent uvarovite-grossular garnets from three localities (Saranov. Veselovsk, and Saranka. Ural Mountains. Russia) were investigated using single-crystal X-ray CCD diffraction data. The intensity and lattice parameter data attest to the violation of the cubic garnet space group Ia3̅d and the symmetry reduction to subgroups with triclinic (I1̅), monoclinic (I2/a), or at most orthorhombic symmetry (Fddd). Careful structure refinements starting in space group /I reveal that partial long-range Cr 3+ /Al ordering on the octahedral sites is the most prominent non- cubic feature. For each crystal, a nearly perfect linear correlation of the individual octahedral size with its Cr occupancy is observed. Considering the dependence on the bulk Cr mole fraction, the individual octahedral size in non-cubic uvarovite-grossular solid solutions is represented by <Cr/Al- 0> (Å) = 1.9247 + 0.0147 XCr (bulk) + 0.0534 Xc r(individual) . These uvarovites also structurally incorporate traces of hydrous component (<1 wt% H 2 O) as O 4 H 4 “hydrogarnet” substitution in a non-cubic way. thus leading to further subtle deviations from cubic symmetry. Within the range of these low water contents, the refined Si-O bond length, averaged over crystallographically different tetrahedra. cor­relates with the total integral OH absorption coefficient α i through the equation ≪ Si-0≫ (Å) = 1.6455 + 1.0074·10 -7 α i (cm -2 ). Although the Si and Ca atoms occupy general positions in I1̅. they deviate little from their respective special positions in la3̅d. Consistent with the respective angular lattice distortions, the refined Cr 3+ /Al site distribution pattern is definitely triclinic in the Saranka sample, somewhat less pronounced triclinic (pseudomonoclinic) in the Veselovsk sample, and distinctively pseudo-orthorhombic or orthorhom­bic in all four samples from the Saranov locality. Considering crystal chemical evidence, three of the pseudo-orthorhombic Saranov samples with elevated water content appear to have triclinic symme­try as well, while one low-water uvarovite is classified as orthorhombic. These results are in good agreement with those of optical and UV-VIS-IR spectroscopic investigations reported in Part I of the present study (Andrut and Wildner 2001). Previous structure data for birefringent garnets reported in the literature as non-cubic are com­pared and discussed. Special emphasis is also drawn to structural and crystal chemical details exhib­iting a non-ideal mixing behavior along the uvarovite-grossular join. Evidence for and against the cation ordering models postulated in the literature are discussed.

A transmission electron microscopy study reveals the microstructure of polysynthetically twinned, near end-member grunerite from Rockport, Massachusetts. A small percentage of Pnma ferroanthophyllite of similar composition is present as thin slabs intergrown parallel to (100) of grunerite and as individual µ-size crystals. HRTEM imaging shows that grunerite is free of chainmultiplicity faults. The observed microstructures are interpreted as a result of partial transformation of ferroanthophyllite to grunerite, a mechanism that is also supported by microstructures observed in the Pnmn protoferroanthophyllite from Cheyenne, Colorado, with similar composition. This study supports the possibility that grunerite, ferroanthophyllite, and protoferroanthophyllite may all possess true stability fields.

A (21)-clinopyribole with the composition K 1.10 Na 2.32 Ca 1.52 Mg 5.85 Al 1.23 Si 12.04 O 34 (OH) 2 has been syn­thesized at 10 GPa and 1250 °C in a multi-anvil apparatus. The unit-cell parameters are a = 9.8390(9), b = 26.6471(6), c = 5.2665(5) Å, β = 106.25(5)°, and V = 1325.6(4) Å 3 . The structure (space group A2/m) consists of an alternating arrangement of single- and double-chain silicate slabs along the b axis, with a + + + + configuration. This phase possesses all the features predicted by Veblen and Burnham (1978b) for a mixed-chain silicate intermediate between pyroxenes and amphiboles. The single-chain portion of the structure corresponds to a clinopyroxene with the omphacite composition Di 55 Jd 45 , whereas the double chain portion is essentially a potassic richterite. The MS2 cation in the single-chain portion occupies a coordination environment that is similar to the M4 site in richterite, whereas the MD4 cation coordination in the double-chain portion is comparable to the M2 site in C2/c omphacite. The observed unit-cell volume is 1.5% smaller than the equivalent mixture of Di 55 Jd 45 + richterite, accounting, in part, for its high-pressure stability relative to its pyroxene and amphibole components.

The phase boundaries limiting the stability fields of the three different Mg(Fe)SiO 3 phases [orthoenstatite (Oen), low-temperature clinoenstatite (LCen), and high-pressure clinoenstatite (HCen)] that occur at the pressure-temperature conditions prevailing in the Earth’s upper mantle are inferred from Raman spectroscopy on quenched samples from high P-T experiments. There are subtle but significant differences between the spectra of Oen and the spectra of samples quenched within the stability field of HCen or LCen. The most prominent differences are additional peaks at 369 and 431 cm -1 in the spectrum of Cen and a systematic shift of the peak at 236 cm -1 in the Oen spectrum to 243 cm -1 in the Cen spectrum. However, no distinction can be made between samples quenched from the HCen and LCen stability fields. This is consistent with the fact that the HCen phase is non-quench- able and that the pyroxene phase observed in the experimental products is LCen as verified by pow­der X-ray diffraction. Experiments performed in the pressure-temperature range 1.2-14 GPa and 750-1900 K have been used to constrain the Mg(Fe)SiO 3 phase diagram. Pyroxene compositions cover the range from pure MgSiO 3 to Mg 0.9 Fe 0.1 SiO 3 with minor amounts of Al, Ca, Na, and Cr. The results are similar to previous determinations from X-ray and optical studies and tightly constrain the HCen-Oen phase boundary, which can be expressed by the equation P (GPa) = 0.00454 T (K) + 1.673. The LCen-Oen boundary is not as well constrained, but the data are sufficient to locate the invariant point where all three MgSiO 3 phases coexist at 6.6 GPa and 820 °C.

Raman spectra of phase E have been measured between 100 cm -1 and 4000 cm -1 in a diamondanvil cell to 19 GPa at 300 K. We observe several broad bands in the region below 1200 cm -1 and three bands in the OH stretching region. All modes shift continuously with increasing pressures to 19 GPa and there is no indication for a major change in the crystal structure of phase E. First constraints on hydrogen sites and hydrogen bond strengths are derived from the OH frequencies together with their pressure dependencies and the correlation of OH stretching frequencies with O···O bond distances. The 3429 cm -1 band decreases linearly with pressure at -7.8 cm -1 /GPa and is clearly involved in hydrogen bonding. The corresponding O···O distance at ambient conditions is estimated to be around 2.82 Å. The 3617 cm -1 band shows a positive pressure trend allowing only very weak or absent hydrogen bonds. Corresponding O···O distances would be greater than 3 Å.

Recent optical and differential scanning calorimetry measurements indicate phase transitions in leonite-type compounds at low temperatures. The crystal structures of these phases, i.e., leonite, K 2 Mg(SO 4 ) 2 ⋅4H 2 O, “Mn-leonite”, K 2 Mn(SO 4 ) 2 ⋅4H 2 O, and mereiterite, K 2 Fe(SO 4 )2⋅4H 2 O, have been determined at low temperatures. The leonite structure (space group C2/m at room temperature) is composed of sulfate tetrahedra and MeO 6 octahedra which are interconnected by K cations and hydrogen bonds of the H 2 O molecules. Previous structure investigations at room temperature have shown that one of the sulfate groups is disordered. Refinements of single-crystal X-ray data at ambient and low temperatures indicate that the dynamic disorder in leonite and “Mn-leonite” is “frozen” in two steps and thus results in two new, ordered structures at low temperatures. In mereiterite only one transition from the dynamically disordered to the ordered structure is observed. The two low-temperature crystal structures of leonite have been refined to R = 0.0236 at 170 K (space group I2/a, Z = 8, a = 11.780(2) Å, b = 9.486(2) Å, c = 19.730(4) Å, β = 95.23(3)°, V = 2195.6 Å 3 ), and to R = 0.0230 at 100 K (space group P21/a, Z = 4, a = 11.778(1) Å, b = 9.469(1) Å, c = 9.851(2) Å, β = 95.26(1)°, V = 1094.01 Å 3 ). The two low-temperature crystal structures of “Mnleonite” have been refined to R = 0.0272 at 185 K (space group I2/a, Z = 8, a = 12.035(2) Å, b = 9.549(2) Å, c = 19.839(4) Å, β = 94.99(3)°, V = 2271.3 Å 3 ), and to R = 0.0237 at 110 K (space group P21/a, Z = 4, a = 12.031(1) Å, b = 9.531(1) Å, c = 9.902(1) Å, β = 95.02(1)°, V = 1131.08 Å 3 ). The low-temperature crystal structure of mereiterite has been refined to R = 0.0219 at 185 K (space group P21/a, Z = 4, a = 11.834(2) Å, b = 9.502(1) Å, c = 9.913(2) Å, β = 94.87(1)°, V = 1110.66 Å 3 ). The different behavior of mereiterite (i.e., stability range, sequence of transitions) in comparison to the Mg and Mn endmembers may be explained by more distorted MeO 6 octahedra and by strongly different hydrogen bond lengths around the disordered sulfate groups.

Two new mineral species, carraraite and zaccagnaite, were found in cavities in calcite veins in marble quarries of the Carrara basin (Apuan Alps, Italy). Carraraite, Ca 3 Ge(OH) 6 (SO 4 ) 1.08 (CO 3 ) 0.92 ⋅12H 2 O, occurs as submillimetric crystals, tabular on {001}. The cell dimensions are a = 11.056 (3), c = 10.629 (6) Å, and the space group is P63/m. Carraraite is optically uniaxial (-), ω = 1.509, ε = 1.479. The strongest lines of the X-ray diffraction pattern are at d-spacings (Å): 9.57 (vs) (100), 5.53 (s) (110), 3.83 (s) (112), 3.56 (ms) (202), 2.74 (ms) (302). Carraraite is a new member of the ettringite-thaumasite group, which is characterized by columns of composition [Ca 3 Ge(OH) 6 ·12H 2 O] 4+ running along c and interconnected through hydrogen bonding to (SO 4 ) 2- and (CO 3 ) 2- groups. Zaccagnaite, Zn 4 Al 2 (OH) 12 (CO 3 )⋅3H 2 O, occurs as minute hexagonal crystals, elongated parallel to [001]. The cell dimensions are a = 3.0725 (3), c = 15.114 (4) Å and the space group is P63/mmc. The crystals are always covered by a thin crust of fraipontite. The strongest lines of the X-ray diffraction pattern are at d-spacings (Å): 7.51 (vs) (002), 3.794 (m) (004), 1.542 (ms) (108), 1.539 (ms) (110). Zaccagnaite is a new member of the hydrotalcite-manasseite family; its structure is characterized by a regular alternation of brucite-like layers with composition (Zn 2/3 Al 1/3 )(OH) 2 and an interlayer composed of carbonate groups and water molecules.

Microstructures of crystal surfaces can reveal growth mechanisms. However, original surfaces of phenocrysts are rare because they are commonly covered with glass or rock matrix. In an unusual occurrence, microscopic growth features are clearly visible on original surfaces of 100 Ma melanite garnet phenocrysts of the Crowsnest Formation, Alberta, Canada. Such details as the steps of growth pyramids are clearly resolved at the 100 nm scale (vertically) in reflected- light Nomarski differential interference contrast (NDIC). However, because of the small step size (<0.5 micrometers), these features are not resolved in transverse section using a transmitted-light polarizing microscope. In transmitted-light microscopy, growth features of a thin-section appear indistinguishable from oscillatory zoning. These results suggest that, in general, oscillatory zoning may be surface growth forms seen in transverse section. Growth of the Crowsnest phenocrysts was both lateral and intermittent along the crystal face rather than continuously outward from the face, an observation of considerable significance to the theoretical modeling of natural crystal growth. This empirical study should alert other workers to the possibility of studying crystal growth using natural magmatic crystals. In particular, this unique Crowsnest material would repay study by diverse techniques.

We present 17 O MAS NMR data for crystalline calcium dialuminate (grossite), CaAl 2 O 4 and monoaluminate, CaAl 4 O 7 . The first of these contains an oxygen tricluster site and serves as a model compound for sites of this type in aluminosilicate glasses. Tricluster site NMR parameters are distinct from those of bridging O atoms (Al-O-Al), allowing partial resolution in triple quantum MAS NMR spectra. Such spectra for calcium aluminosilicate glasses are consistent with the presence of a small fraction of tricluster sites. Observed chemical shifts for non-bridging oxygen (NBO) atoms in an impurity phase in the CaAl 2 O 4 sample are distinct from those for NBO in Ca-aluminosilicate glasses, indicating that the latter are primarily bonded to Si, not Al.

The heat capacity of synthetic pentlandite Fe 4.60 Ni 4.54 S 8 was measured over the temperature range 6-306 K. Experimental values of C P (T) indicate that there are no phase transitions. The thermodynamic functions C P (T), H(T) - H(0), and S(T) - S(0) were evaluated: C P (298.15) = 442.7 J/mol·K, S(298.15) - S(0) = 474.9 J/mol·K, and H(298.15) - H(0) = 76280 J/mol (molar mass: 779.877 g). Below 10 K, the heat capacity of pentlandite fits the regression C P (T) = a T + b T 3 . The linear term is typical of metals and known as an electronic contribution to the heat capacity. This agrees with the hypothesis that there is metal bonding among cations in tetrahedral sites.

Over the past 30 years the cooling rates of terrestrial and extraterrestrial basaltic rocks have commonly been inferred from the size of antiphase domains (APDs) in pigeonite. However, the coarsening rate of APDs has been observed to deviate substantially from an ideal rate. It is believed that this deviation is caused by Ca segregation to the antiphase boundaries (APBs) of domains, since Ca is expected to substantially slow boundary migration rates due to solute drag. This letter presents direct experimental evidence of Ca segregation to APBs in pigeonite. The local atomic structure and chemistry of APBs in pigeonite were examined using high-resolution and energy-filtered transmission electron microscopy. High-resolution images show that APBs in pigeonite have a structure similar to that of augite, and energy-filtered compositional images reveal that Ca segregates to APBs. These results have direct ramifications for the use of antiphase domain (APD) size as a marker for the thermal history of the rocks in which pigeonite crystallizes.