Volume 91, Issue 10

A renewed interest in mineral spheres that exhibit oscillatory layering is assured with the spectacular discovery of hematite beads in Meridiani Planum, Mars. The Martian .blueberries. are variously interpreted as sedimentary concretions, impact spherules, or accretionary lapilli. We draw attention to a possible alternative that the spheres of oxides are of igneous origin in carbonate-rich upper mantle systems. Two spherical oxide assemblages from cabonatitic kimberlites in the central Namibia (formerly South West Africa) volcanic field show similarities and contrasts to orbicular structures in carbonatites from Uganda, South Africa, Germany, Finland, and Russia. Structures have rapakivi appearances that bear upon the controversial issue of a relation between kimberlites and carbonatites. The first example is from a carbonatitic kimberlite adjacent to the Mukorob kimberlite (sesu stricto). Oblate spheroidal nuggets (2 × 1.5 × 0.5 cm) have olivine (now serpentine + calcite) cores with concentrically layered bands of ilmenite + calcite, and pedestals of ilmenite (Ilm 55 Geik 40 Hem 5 ) with intergranular calcite; the spheres are matrix-supported in massive ferroan calcite + serpentine, with minor apatite (F = 1.5 wt%) and barite. The matrix contains xenocrystic Mg-Al chromite (55 wt% Cr 2 O 3 ) mantled by ilmenite (Ilm 51 Geik 43 Hem 6 ) but the dominant groundmass oxide mineral is Mg-Al-Ti (MAT) magnetite. The second example is from the carbonatite facies of the Hatzium kimberlite, and is also in a dike. MATmagnetite pellets (2-3 cm) have alternating wide (~1 mm), and thin (~0.25 mm) bands of magnetite nucleated on a complex mixture of calcite (FeO = 0.1 wt%) and serpentine (4 wt% FeO, 0.5-1 wt% Al 2 O 3 ). The matrix assemblage of serpentine + calcite + MAT-magnetite has mineral compositions similar to the spheroids, but is distinguished by abundant Ba-phlogopite and lesser apatite (F = 2 wt%). Phlogopite is strongly zoned in patches and bands, is kink-banded and contains inclusions of MAT-magnetite + calcite. The assemblage is typical of phoscorites (also known as camaforites) in other carbonatites, but the ilmenite spheroids appear to be restricted to Mukorob. Our interpretation is that both types result from the nucleation of immiscible liquids (a high-density and high-viscosity ore-carbonate-silicate component in a low-viscosity phosphoric carbonatite melt) on olivine or olivine + magnetite, assisted by frothing and slow degassing on decompression and dike intrusion.

Late Archean and Early Proterozoic continental margin pelites, graywackes, and quartzites in today.s Black Hills were regionally metamorphosed during the collision of the Archean Wyoming and Superior provinces beginning at ~1755 Ma. During east-west regional compression, metamorphism reached incipient garnet-biotite grade conditions at ~400 °C. Unzoned garnet (Sps 44 Alm 44 Pyr 3 Grs 9 ) characterizes the mineral assemblage in graphitic metapelites. The low-temperature onset of the garnet-biotite assemblage is attributed to high Mn concentrations and low a H₂O due to the presence of CH 4 and CO 2 . During late stages of the collision, the rocks were intruded by the Harney Peak leucogranite (HPG). Vigorous fluid flow, evidenced by abundant quartz veins, metasomatism, and consumption of graphite from metapelites, occurred in the granite aureole. The lowest-grade aureole assemblage includes chlorite that overgrows the early regional foliation and new, clear, almandine rims on garnet. Higher-grade facies include staurolite, then sillimanite, and finally second-sillimanite. The prograde mineral assemblages are consistent with calculated pseudosection assemblages for the average metapelite. Beginning with the sillimanite zone, almandine garnet is mostly inclusion-free. Textures suggest that this new garnet grew after dissolution of the old garnet, for which the only remaining evidence are remnant inclusions or coarse quartz-biotite clots. The new garnet grew in response to more elevated temperatures. Garnet compositions projected onto the pseudosection indicate nearly isobaric heating between 3.5.4.5 kbar during prograde metamorphism. Occasional replacements of staurolite by muscovite and biotite are attributed to infiltration of the rocks by K-bearing magmatic fluids. Andalusite occurs as euhedral crystals in quartz veins or as late poikiloblasts along foliations planes, where it appears to have grown while Si-rich fluids passed through the rocks. Typically, it replaces earlier muscovite, suggesting that the fluids had low a H₂O . The andalusite indicates decompression of the still hot, buoyant, fault-bounded block that included the HPG magma, relative to surrounding blocks in the terrane. This decompression contributed to the high-T/low-P conditions that existed in this portion of the thickened orogen and suggests that the various fault-bounded blocks in the Black Hills may have had different P-T histories.

To study Li isotopic fractionation during granite differentiation and late-stage pegmatite evolution, Li isotopic compositions and concentrations have been measured for the S-type Harney Peak Granite, the spatially associated Tin Mountain pegmatite, and possible metasedimentary source rocks in the Black Hills, South Dakota. The Harney Peak Granite is isotopically heterogeneous, with δ 7 Li varying from -3.1 to +6.6. The δ 7 Li values of Proterozoic metasedimentary rocks that are possible sources of the Harney Peak Granite range from -3.1 to +2.5 and overlap with post- Archean shales and the Harney Peak Granite. For the granite suite, there is no correlation between δ 7 Li and elements indicative of degrees of granite differentiation (SiO 2 , Li, Rb, etc.). The Li isotopic composition of the Harney Peak Granite, therefore, appears to reflect the source composition. Minerals from the zoned Tin Mountain pegmatite have extremely high Li contents and heavier Li isotopic compositions than the granite or surrounding Black Hills metasedimentary rocks. The heavier compositions may reflect Li isotopic fractionation resulting from extensive crystal-melt fractionation. Lithium concentrations decrease in the order: spodumene (~3.7 wt%), muscovite (0.2 to 2.0 wt%), plagioclase (100.1100 ppm), quartz (30.140 ppm). Plagioclase, muscovite, and spodumene in all zones display a relatively narrow range in δ 7 Li of +7.9 to +11.4. In contrast, quartz is isotopically heavier and more variable (+14.7 to +21.3), with δ 7 Li showing an inverse correlation with Li concentration. This correlation reflects the mixing of isotopically heavy Li in quartz and lighter Li in fluid inclusions, as documented by fluid inclusion compositions (δ 7 Li = +8.1 to +13.4 and Li of 280 to 3960 ppm). Extrapolation of this trend to an estimated intrinsic Li concentration in quartz of <30 ppm, yields an inferred δ 7 Li for fluid inclusion-free quartz of >+21. The large difference in δ 7 Li between quartz and other minerals may reflect 7 Li preference for less highly coordinated sites, which have higher bond-energies (i.e., the two- or fourfold site in quartz vs. higher coordination number sites in other minerals). Comparison of the Li isotopic composition of fluid inclusions with that of the wall zone of the Tin Mountain pegmatite suggests ~4‰ isotopic fractionation during fluid exsolution, which agrees with the results derived from studies of hydrothermal alteration of basalts.

We use trace-element concentrations in olivines and glasses from a closed basaltic system to identify the effects of petrologic processes on the trace-element record of that system. The closed basaltic system in question is the Makaopuhi Lava Lake (MLL), which is closed with respect to magma mixing. Detailed studies of this lava lake have provided important information about system variables and petrologic processes that have been measured and observed at the lake. These previous studies show that olivine crystallized from the lava lake at all stages of the lake.s evolution, which means that olivine and residual glasses contain a record of the lake’s petrologic history. We use this information, in conjunction with variations in trace-element concentrations in olivines and glasses, to show that mineral crystallization, gravitational settling, convective flow, filter pressing, and mineral- melt interface kinetics have characteristic effects on the trace element record of a closed basaltic system. These results are pertinent to the field of comparative planetology because they can be used to evaluate petrologic information in small samples from other planetary bodies, where information about system variables and/or petrologic processes is limited.

More than 50% of the igneous blocks from mélanges of the Franciscan subduction complex in northernmost California and from the equivalent Dothan Formation of southwestern Oregon are more silica-rich than basalt, and some contain supra-subduction zone geochemical signatures such as elevated Th/Ta ratios. In contrast, blocks from more southerly parts of the Franciscan mélange belt are more commonly mid-ocean ridge and ocean island/seamount basalts. The data indicate a northsouth variation in the block population, due to differences in source terrane, or in the processes that provided the blocks and transported them into the mélange, or both. We suggest that, in the northern Franciscan, a relatively greater fraction of igneous mélange blocks was emplaced by block slumping from the forearc crust into the trench than was the case to the south.

In this paper, we report analysis of various basaltic lunar samples including 14053 and 14072, KREEP basalt 15386, thirty basalt clasts from Apollo 14 breccia 14321, as well as impact-generated samples (matrix from breccia 14168, olivine vitrophyres 14321,1180 and 14321,1539, and impact melt 14310) using a combination of solution and laser ablation inductively coupled plasma mass spectrometry (ICP-MS). The basalt clast samples were previously analyzed by instrumental neutron activation. On plots of incompatible trace elements (ITEs) vs. compatible trace elements, the Apollo 14 high-Al basalts form three approximately subparallel trends that, on the basis of current data, are also separated by age. Plots of ITE ratios (i.e., Nb/Ce vs. Zr/Y) can be used to indicate source composition, and also divide the basalts into three groups: Group A (~4.3 Ga); Group B (~4.1 Ga); and Group C (~3.9 Ga). New data for 14072 suggest the sample does not fit with any of the three groups defined here, and may indicate the presence of a fourth group of high-Al basalts in the proximity of the Apollo 14 site. The Apollo 14 high-Al basalts are compositionally distinct from known Apollo 14 impact melts and impact-generated lithologies. The three groups cannot be related by varying degrees of partial melting of a single, KREEP-contaminated source and, therefore, require three separate source regions. The new data indicate that Group A basalts evolved through closed-system crystal fractionation. However, the new data from basalts forming Groups B and C require open-system evolution that involves combined assimilation and fractional crystallization (AFC). Unlike previous AFC modeling of the Apollo 14 high-Al basalts, an assimilant composed of KREEP is not sufficient to generate the compositional ranges of each basalt group. The modeling of both groups requires a mixture of KREEP and granite as the assimilant, which supports the notion of a genetic relationship between these two lunar components.

High-temperature experiments were performed from 3.0 to 7.0 GPa to investigate the effect of composition on near-liquidus garnet-melt trace-element partition coefficients (D-values) in a Ti- and Fe-rich lunar bulk composition. Starting compositions were doped with Sc, Sr, Y, Zr, Ba, Nd, Sm, Dy, Yb, Hf, and Th. D-values were measured by ion-microprobe analysis. The lattice strain model of Blundy and Wood (1994) was applied to measured D-values, and then compared with the predictive garnet-melt trace-element partitioning model of van Westrenen et al. (2001b). Although the latticestrain model describes our data adequately, there is a substantial mismatch between the prediction of the van Westrenen et al. (2001b) model and the actual measured D-values. We suggest crystal-chemical effects associated with the high Ti content of our experimental garnets may be responsible for this mismatch. Titanium alone does not appear to control partitioning; we infer that Fe 2+ and/or Mg 2+ are required to partition into octahedral coordination on the Y-site, which may further necessitate Si 2+ to also partition on the Y-site for charge balance. This combination of effects influences trace-element partitioning in the garnet. Observed changes in D-values correlate with the integration of Ti on the Y-site in a majorite-like exchange, yet our garnets incorporate little or no majorite. Garnets grown have some of the lowest apparent Young.s modulus values (274.528 GPa) yet documented for the garnet X-site indicating higher compressiblities than previously inferred.

The atomic arrangement of lunar merrillite has been refined to R = 0.0452 in R3c using X-ray diffraction data recorded on a CCD detector; previous attempts at structure solution using a point detector were not successful because of the poorly crystallized nature of the lunar material. The atomic arrangement of merrillite has a structural unit of [(Mg,Fe)(PO 4 ) 6 ] 2 16- that forms a “bracelet-and-pinwheel„ unit that is common in hexagonal-closest-packed layers. The individual structural units are not polymerized and exist in layers at z = 1/6, 1/3, 1/2, 2/3, and 5/6. In lunar merrillite, the [(Mg,Fe)(PO 4 ) 6 ] 2 16- structural units are linked by a [(Ca,REE) 18 Na 2 (PO 4 ) 2 ] 32+ interstitial complex, formed of Ca1O 8 , Ca2O 8 , Ca3O 8 , NaO 6 , and P1O 4 polyhedra. There has long been speculation regarding the relationship between merrillite and terrestrial whitlockite, and the solution of the Fra Mauro merrillite atomic arrangement allows the characterization of the lunar phase. Lunar merrillite and terrestrial whitlockite have largely similar atomic arrangements, but the phases differ due to the presence or absence of hydrogen. In whitlockite, H is an essential element and allows the charge balance. Hydrogen is incorporated into the whitlockite atomic arrangement by disordering one of the phosphate tetrahedra and forming a PO 3 (OH) group. Lunar merrillite is devoid of hydrogen, and thus no disordered tetrahedral groups exist. Charge balance for substituents Y and REE (for Ca) is maintained by Si ↔ P tetrahedral substitution and □ ↔ Na at the Na site. The structure solution demonstrates the effectiveness of the CCD detector in unraveling previously intractable diffraction data and urges that previously analyzed lunar material be reexamined using this instrumentation.

Late-stage rims of magmatic pyroxenes from some martian basalts show decreases in Li and B contents relative to earlier-formed pyroxene cores. This behavior is different than expected from their documented incompatible element behavior. Previous workers interpreted such depletions to reflect the loss of several wt% magmatic water during basalt crystallization. This interpretation has profound implications for the nature of the martian mantle and recent exchange of volatiles between the martian mantle and atmosphere. To assess alternative mechanisms that may influence the behavior of Li and B in the absence of aqueous fluid activity, the effects of changing pyroxene composition during crystallization, shock pressure, and shock-associated thermal metamorphism were studied. Lithium and B depletions are documented in late-stage rims of pyroxenes from anhydrous lunar basalts indicating that mechanisms other than aqueous fluid activity must have influenced Li and B partitioning in these pyroxenes. Depletions of Li and B are most likely associated with changing pyroxene composition during crystallization, and occur in lunar and martian pyroxenes with late-stage Fe-enrichment. It is interesting that pyroxenes without late-stage Fe-enrichment show no concomitant Li and B increases. Lithium loss may occur during breakdown of metastable pyroxferroite. Additionally, changes in Cr content may influence the substitution mechanism involved for incorporating Li. Shock does not redistribute Li or B but may facilitate subsequent thermally driven diffusion by the introduction of mechanical defects in grains. Thermally metamorphosed pyroxenes exhibit higher Li and lower B contents relative to unheated pyroxenes. It is likely, therefore, that Li and B are redistributed through interactions between pyroxenes and surrounding zones of mesostasis during thermal metamorphism.

The behavior of multivalent elements such as Fe, Cr, V, and Eu in magmatic systems reflects the f O₂ of the environment. In particular, Eu behavior in pyroxene from basaltic systems has been demonstrated to be an effective measure of f O₂ . We selected two nearly isochemical lunar pigeonite basalts (15058, 15499), a lunar high-Ti basalt (75035), and Pasamonte (representing asteroid 4 Vesta, an unequilibrated eucrite), to explore other potential variables that may affect this indicator of f O₂ . All of these basalts crystallized at an f O₂ of approximately IW-1, yet they experienced different cooling and crystallization histories and their pyroxenes exhibit a wide range of compositional trajectories within the pyroxene quadrilateral. There are several variables that influence the Eu/Eu* recorded in pyroxene that may compromise the determination of f O₂ . Previous experimental studies show that pyroxene composition influences the ability of pyroxene to accommodate REE and fractionate Eu 2+ from Eu 3+ . We demonstrate that in addition to the influence of Ca in the M2 site, the Al content in the pyroxene and its influence on coupled substitutions will also influence the fractionation of Eu 2+ from Eu 3+ . For example, the coupled substitution T Si 4+ + M2 R 2+ → T Al 3+ + M2 REE 3+ may accommodate REE 3+ in preference to Eu 2+ , which is too large. Different pyroxene growth surfaces will incorporate Eu 2+ and Eu 3+ differently due to differences in growth rate, Al content, and site configuration. In consort with the pyroxene composition, fractionation of Eu 2+ from Eu 3+ will be aided by the Al content of the basaltic melt, which increases the activities of network-forming components such as CaAl 2 O 4 and FeAl 2 O 4 in the melt during pyroxene crystallization. The Al content will result in changing the partitioning behavior of Eu 3+ while having very little effect on Eu 2+ . Melt composition, the appearance of plagioclase on the liquidus, and the kinetics of plagioclase crystallization are influenced by cooling rate. Data from the four basalts selected also suggest that Eu 2+ /Eu 3+ in the melt remains buffered even with extreme differences in cooling rate and plagioclase crystallization kinetics. Unexpectedly, many of these same variables affect the substitution of multivalent V. If the f O₂ determined from Eu behavior in pyroxene is not placed within a petrologic and crystal-chemical context, errors of 1 to 2 log units may result. The influence of these variables may be reduced by using multiple, co-crystallizing phases (i.e., plagioclase and pyroxene) and ratioing D Eu to adjacent REE (D Sm, Gd ).

Pyroxene grains from 14 basalt suites from the Earth, Moon, Mars, and Vesta were studied by electron- and ion-microprobe techniques. The results show that several elemental trends can be related to planetary parentage and crystallization conditions including paragenetic sequence and kinetics. Ferric iron (Fe 3+ ) systematics show that terrestrial pyroxene is enriched in Fe 3+ compared with pyroxene from Mars due to higher oxygen fugacity (f O₂ ) conditions on the Earth that produce more Fe 3+ in basaltic melts. Low f O₂ conditions on the Moon and Vesta result in very little or no Fe 3+ in pyroxene from these bodies. Terrestrial pyroxenes contain more Na than those from Mars, yet martian plagioclase contains more Na than terrestrial. This difference is because terrestrial pyroxene contains more Fe3+ and thus the acmite component (NaFe 3+ Si 2 O 6 ) is more important on Earth than Mars. Pyroxenes from the Moon and Vesta have very little Na, which can be attributed to the overall volatile-depleted nature of these bodies. All planetary pyroxenes show that Cr decreases with increasing fractionation because it is compatible in pyroxene, and in many basalts the crystallization of chromite depletes the melt in Cr 3+ . The Mn/Fe 2+ systematics in pyroxene show distinct trends for the planetary bodies in the order Vesta > Mars > Earth > Moon. These Mn/Fe 2+ trends are most likely affected by the volatility of Mn relative to Fe, and thus there is an increase in the Mn/Fe ratio with increasing distance from the Sun, except for the Moon, which likely lost Mn during its giant impact origin. These same trends have been documented for olivine from the different planets, and this parameter as measured in basaltic silicates is a robust fingerprint of planetary parentage. Vanadium partitioning into planetary pyroxene grains is affected by oxygen fugacity, the availability of charge-balancing elements, basaltic crystallization sequences, and kinetics. Partitioning of V into pyroxene at low f O₂ conditions (i.e., Moon and Vesta) is seen to increase as the charge-balancing cation IVAl increases. Partitioning of V into pyroxene at relatively high f O₂ conditions (i.e., Earth and Mars) increases with increasing Na and IVAl, which provide charge balance for incorporation of V 4+ . Because of the above complexities, a V-valence oxybarometer as measured in planetary pyroxene grains is not likely to be robust.

Merrillite, also known as “whitlockite„ , is one of the main phosphate minerals, along with apatite, that occur in lunar rocks, martian meteorites, and in many other groups of meteorites. Significant structural differences between terrestrial whitlockite and lunar (and meteoritic) varieties warrant the use of “merrillite„ for the H-free extraterrestrial material, and the systematic enrichment of REE in lunar merrillite warrants the use of “RE-merrillite„. Laser Raman spectroscopy of extraterrestrial merrillite and terrestrial whitlockite confirms the absence of H in the former and presence of H in the latter. Lunar merrillite, ideally (Mg, Fe 2+ , Mn 2+ ) 2 [Ca 18-x (Y,REE) x ](Na 2-x )(P,Si) 14 O 56 , contains high concentrations of Y+REE, reaching just over 3 atoms per 56 O atoms, or up to ~18 wt% as (Y,RE) 2 O 3 . In the absence of extensive Si↔P substitution, the “availability„ of the Na site limits Y+REE substitution to ~2 atoms per 56 O atoms. Higher concentrations of Y+REE, with coupled substitution of Si for P to balance charge are possible, but rare in lunar material. Intrinsically low Na concentrations in lunar rocks, combined with the typical formation of merrillite in late-stage basaltic mesostasis or residual, intercumulus melt pockets, produce these high REE concentrations. Lunar merrillite typically contains 0.1.0.4 Na atoms per 56 O atoms. For comparison, martian merrillite contains significantly higher Na concentrations (up to 1.7 Na atoms per 56 O atoms) and much lower REE concentrations. Meteoritic merrillite has relatively low REE contents, but exists in both Ca-rich and Na-rich varieties. Concentrations of Fe and Mg in all varieties sum to near 2 atoms per 56 O atoms. Merrillite in lunar crustal lithologies typically has Mg >> Fe; however, Fe-rich mare basalts contain up to 1.8 Fe 2+ per 56 O. The structure of merrillite accommodates a variety of substitutions, and the compositional characteristics reflect conditions and processes specific to the parent planet.

Equilibrium and dynamic crystallization experiments were used to determine distribution coefficients (D-values) for the REE and the trace elements Sc, V, Rb, Sr, Y, Zr, Nb, Ba, Hf, Th, and U between Ti-Al-clinopyroxene and melt. Equilibrium values for D correlate well with previous studies where values have been determined. The D-values were also studied as a function of cooling rate. Increased cooling rates produce higher growth rates. At growth rates that exceed the diffusion rates in the crystals, the D-values increase to near unity and the HREE D-values exceed unity only at the most rapid crystal growth rates. The increase in D-values is the result of the inclusion of components into the growing pyroxene from the boundary layer that develops at the crystal-melt interface because of the disparity between the growth and diffusion rates. The origin of sector zoning is best explained as an interplay of primarily crystallographic control with kinetic effects.

Minor- and trace-element sector zoning is nearly ubiquitous in pyroxene grown from basaltic melts using 1 atm gas-mixing furnaces. Mineral/melt partition coefficients can differ between sectors by more than a factor of five. A synthesized hypersthene-normative basalt composition was doped with Al, Cr, and Ti using a matrix of doping experiments, to understand better the mechanisms behind the formation of sector zoning. Isothermal and controlled-cooling experiments were conducted. The {010} sectors contain higher enrichments of Al, Cr, and Ti than {100} sectors, at the expense of Mg and Si. Evaluation of these relative sector enrichments indicates that sector zoning results from several kinetic mechanisms, which exert influence even when attempting laboratory conditions that approximate equilibrium: (1) rapid growth rate, not necessarily synonymous with rapid cooling; (2) crystal-chemical control by means of charge-coupled substitutions; and (3) diffusion limitations of Si tetrahedra in the melt boundary layers respective to the different crystal faces.

The NWA 1068/1110 martian meteorite contains megacrystic olivine and associated chromite set in a basaltic groundmass of pyroxene, maskelynite (shock-isotropized plagioclase), phosphates, ilmenite, and ulvöspinel. Spinel is compositionally zoned from chromian cores to titanian rims, and lacks the chromite-ulvöspinel miscibility gap. Co-crystallizing olivine + pyroxene + spinel and ulv öspinel + ilmenite assemblages were identified on the basis of compositional and textural criteria, and were used to determine the oxygen fugacity from the earliest to the latest stages of crystallization. Megacrystic olivine with earliest chromite and low-Ca pyroxene yield an oxygen fugacity 2.5 log units below the Quartz-Fayalite-Magnetite (QFM) buffer, whereas titanian-spinel rims, olivine rims, and co-crystallizing pyroxene yield QFM + 0.3 and groundmass ulvöspinel + ilmenite assemblages yield QFM + 0.5. Vanadium in spinel was examined to determine whether it would be sensitive to the changing redox conditions of the NWA 1068/1110 melt. Only small changes in V concentration were observed between spinel cores and rims; however, the relationship between V and Ti in spinel is consistent with an increase in the valence state of V from V 4+ to V 5+ in accordance with the increase in oxygen fugacity. Application of a quantitative version of the “V-in-chromite„ oxybarometer results in an oxygen fugacity of QFM -3, consistent with mineral equilibria results, although assumptions and uncertainties limit its application to circumstances where the parental melt V concentration is known independently of the redox conditions. Oxygen fugacity results and textural observations are strongly indicative of a xenocrystic origin for the olivine megacrysts and associated chromite, and lend support to a model in which martian basalt oxygen fugacity is determined by mantle-source redox conditions as opposed to assimilation of oxidized crustal material.

Salt hydrates have an active role in regolith development on Mars. The Mg-sulfate system, with highly variable values of n in the formula MgSO 4 ·nH 2 O, is particularly subject to transformations among several crystalline and amorphous forms. The Ca-sulfate system, CaSO 4 ·nH 2 O, is likely to be associated with the Mg-sulfates in most occurrences, but is less susceptible to transformations in n. Desiccation of MgSO 4 ·nH 2 O occurs in exposed soils at the martian equator in summer where higher daytime temperatures at low relative humidity prevail against sluggish nighttime rehydration at high relative humidity. Desiccation and rehydration are both accelerated in the finest size fractions, particularly in silt-size aeolian particles subject to global redistribution by dust storms. This redistribution and periodic excursions into long-term episodes of high obliquity work to rehydrate desiccated MgSO 4 ·nH 2 O to form epsomite, MgSO 4 ·7H 2 O, at higher latitudes in the first case and more globally in the latter. Kieserite, a monohydrate form of MgSO 4 ·nH 2 O resistant to desiccation, can survive equatorial summer conditions, but not protracted high relative humidity; preservation of kieserite at the surface may place limits on the equatorial distribution of ice during past episodes of high obliquity. Deeper horizons in equatorial regolith may preserve hydrated phases through repeated obliquity episodes, raising the possibility of an ancient regolith archive of past hydration. At shallower depths in the regolith, in situ determination of the hydration states of the Mg-sulfates, and possibly the Ca-sulfates, may be used to constrain regolith dynamics if rates and modes of transitions in n can be fully characterized.

The distribution of V in magmatic rocks is controlled primarily by spinel stability. Extensive previous experimental work at oxidized conditions on doped (V-rich) compositions has led to the recognition of the importance of temperature, oxygen fugacity, and spinel composition, but also left ambiguity with respect to the relative importance of these variables in controlling D V spinel/melt . One major uncertainty has been the valence of V in the spinel and glass. Spinel-melt pairs were equilibrated at low and variable oxygen fugacities, with a range of V and Ti contents. XANES spectra were measured on the spinel and glass products, and pre-edge peaks measured and calibrated against valence with the use of glass and oxide standards. The valence of V is always greater in the glass than in the spinels. In spinel, V is dominantly 3+ at oxygen fugacities near the FMQ (fayalite magnetite quartz) buffer, but we find evidence for mixed 3+, 4+, and 5+ at oxidized conditions (FMQ to air), and 2+ and 3+ at very reduced conditions [FMQ to IW-1 (1 log f O₂ unit below the iron wüstite buffer)]. Increased V contents in spinels are correlated with increased D V spinel-melt , at constant temperature and oxygen fugacity. However, increased Ti content causes only a slight decrease in D V spinel-melt and a shift to more reduced V (smaller pre-edge peak), which may be related to Fe-V exchange equilibria. Using the new partition coefficients, together with published results and valence information, expressions have been derived to predict D V spinel/melt for basaltic systems. Application of these expressions to natural suites illustrate their utility and also the great range of D V spinel/melt values relevant to natural systems. Calculation of V depletions in planetary mantles from basalt suites must take silicate, oxide, and metal fractionation into account, as is demonstrated using terrestrial, lunar, martian, and eucritic samples.

The valence state of vanadium (V 2+ , V 3+ , V 4+ , and V 5+ ) is highly sensitive to variations in redox conditions of basaltic magmas. Differences in valence state will influence its partitioning behavior between minerals and basaltic liquid. Using partitioning behavior of V between olivine and basaltic liquid precisely calibrated for martian basalts, we determined the oxidation state of a primitive (olivine- rich, high Mg no.) martian basalt (Y980459) near its liquidus. The behavior of V in the olivine from other martian olivine-phyric basalts (SaU005, DaG476, and NWA1110) was documented. The combination of oxidation state and incompatible-element characteristics determined from early olivine indicates that correlations among geochemical characteristics such as f O₂ , LREE/HREE, initial 87 Sr/ 86 Sr, and initial ε Nd observed in many martian basalts is also a fundamental characteristic of these primitive magmas. These observations are interpreted as indicating that the mantle sources for these magmas have a limited variation in f O₂ from IW to IW+1 and are incompatible-element depleted. Moreover, these mantle-derived magmas assimilated a more oxidizing (>IW+3), incompatible-element enriched, lower-crustal component as they ponded at the base of the martian crust.

Magnesium-rich forsterite grains (>Fo 98 ) in FeO-poor chondrules show variations in cathodoluminescence (CL) intensity that are associated closely with variations in Al 2 O 3 and TiO 2 contents. CL thus can be used to map the distribution of Al and Ti in individual grains. We describe several categories of CL distribution, including grains that are homogeneous, grains that are zoned continuously, grains that show oscillatory zoning, and grains that show a heterogeneous distribution of CL intensity. Heterogeneous CL intensity is interpreted as the result of disequilibrium growth from the chondrule melt. CL surveys of individual chondrules help to identify forsterite grains that are potentially relics in their host chondrules, i.e., grains that did not crystallize in situ from the chondrule melt. Several of these grains show minor-element variations that differ from the majority of melt-grown grains in their host chondrules: this observation supports the inference that they are relics. Little evidence exists for growth of individual grains in multiple brief heating episodes. Identification of the abundance of relict grains is important for determining the thermal histories of chondrules as well as the nature of the precursor assemblages from which they formed.

We report petrography, mineral chemistry, bulk chemistry, and bulk isotopic compositions of a suite of 40 spinel-rich inclusions from the Murchison (CM2) carbonaceous chondrite. Seven types of inclusions have been identified based on mineral assemblage: spinel-hibonite-perovskite; spinelperovskite- pyroxene; spinel-perovskite-melilite; spinel-hibonite-perovskite-melilite; spinel-hibonite; spinel-pyroxene; and spinel-melilite-anorthite. Hibonite-bearing inclusions have Ti-poor spinel compared to the hibonite-free ones, and spinel-hibonite-perovskite inclusions have the highest average bulk TiO 2 contents (7.8 wt%). The bulk CaO/Al 2 O 3 ratios of the inclusions range from 0.005 to 0.21, well below the solar value of 0.79. Hibonite-, spinel-rich inclusions consist of phases that are not predicted by condensation calculations to coexist; in the equilibrium sequence, hibonite is followed by melilite, which is followed by spinel. Therefore, hibonite-melilite or melilite-spinel inclusions should be dominant instead. One explanation for the .missing melilite. is that it condensed as expected, but was lost due to evaporation of Mg and Ca during heating and melting of spherule precursors. If this theory were correct, melilite-poor spherules would have isotopically heavy Mg and Ca, assuming Rayleigh fractionation accompanied evaporation. Except for one inclusion with F Mg = 4.3 ± 2.6./amu and another with isotopically light Ca (F Ca = -3.4 ± 2.0./amu), however, all the inclusions we analyzed have normal isotopic compositions within their 2σ uncertainties. Thus, we found no evidence for significant mass-dependent fractionation. Conditions necessary for non-Rayleigh evaporation are unlikely if not unrealistic, and our preferred explanation for the general lack of melilite among hibonite-, spinel-bearing inclusions is kinetic inhibition of melilite condensation relative to spinel. Because of similarities between the crystal structures of hibonite and spinel, it should be easier for spinel than for melilite to form from hibonite.

Vapors of Al-Fe-SiO-O 2 -H 2 having two different compositions produced ferroaluminosilica grains as a function of agglomeration and fusion along mixing lines in the Al 2 O 3 -FeO-SiO 2 system that are defined by the predictable, deep metastable eutectic (DME) compositions of the smallest condensate grains. Disorder of these amorphous grains is higher than in quenched glass of identical composition, which is the very property of dissipative structures (Prigogine 1978, 1979) that are states of organization of matter where disequilibrium becomes a source of order. Iron-oxidation states control ferrosilica condensate compositions. We present the first magnetic measurements showing a high Fe 3+ content in condensed ferrosilica grains. The Fe-cordierite grain composition is primarily the result of predictable non-equilibrium condensation, not the bulk gas phase composition. Natural terrestrial and anthropogenic (e.g., smelters, coal fly ash) Fe-cordierite might well be a metastable phase due to kinetically controlled processes. Amorphous Mg,Fe-bearing aluminosilica dust in chondritic interplanetary dust aggregates and (rare) Mg,Fe-aluminosilicates in meteorites might have condensed via similar processes.

In a recent paper, Wu et al. (2005) claimed that they had found nanoscale TiO 2 with the α-PbO 2 - structure in omphacite from a coesite-bearing eclogite at Shima in the eastern part of the Dabie Mountains, China. However, some results in their paper are questionable. First, the defects that were regarded as intercalated α-PbO 2 -type TiO 2 lamellas by Wu et al. may be induced during the preparation process of rutile. Second, the SAED pattern in Figures 1b, 2b and 3 from Wu et al. is not due to the diffraction of intercalated α-PbO 2 -type TiO 2 . Third, the lattice parameters reported by Wu et al. are too accurate to be derived only from HRTEM and SAED analyses. In this comment we show evidence for our doubts about their analysis.

Our efforts to introduce defects in rutile by conventional mechanical grinding as used for TEM study did not produce any defects or twinning. Experimental SAED patterns of rutile with intercalated α-PbO 2 -type TiO 2 lamellae are consistent with those expected from calculated patterns. The values and precision of lattice parameters of α-PbO 2 -type TiO 2 derived from such SAED patterns are also consistent with crystal data from X-ray diffraction studies.

Neighborite (NaMgF 3 ) with the perovskite structure, transforms to a post-perovskite (ppv) phase between 27 and 30 GPa. The ppv phase is observed to the highest pressures achieved (56 GPa) at room temperature and transforms to an as yet unknown phase upon heating. Rietveld structure refinement using monochromatic synchrotron X-ray diffraction data provide models for the perovskite and postperovskite structures at high pressure. The refined models at 27(1) GPa indicate some inter-octahedral F-F distances rival the average intra-octahedral distance, which may cause instability in the perovskite structure and drive the transformation to the post-perovskite phase. The ratio of A-site to B-site volume (V A /V B ) in perovskite structured NaMgF 3 (ABX 3 ), spans from 5 in the zero-pressure high-temperature cubic perovskite phase to 4 in this high-pressure perovskite phase at 27(1) GPa, matching the V A /V B value in post-perovskite NaMgF 3 . Using Rietveld refinement on post-perovskite structure models, we observe discrepancies in pattern fitting, which may be described in terms of development of sample texture in the diamond-anvil cell, recrystallization, or a change of space group to Cmc2 1 , a non-isomorphic subgroup of Cmcm-the space group describing the structure of CaIrO 3 .

29 Si MAS NMR spectroscopy on samples of 10-Å phase synthesized from oxides (6.0 GPa/600 °C/400 h) and from partial transformation of talc (6.5 GPa/650 °C/12.5 h) reveals that this phase contains Q 2 -type Si sites in a ratio Q 3 :Q 2 of 5.33:1. It is proposed that the Q 2 arise from adjacent vacancies in the tetrahedral sheets for which charge balance is most likely achieved by hydroxylation via a hydrogarnet-like substitution involving the formation of Q 2 silanol groups. Variable-contacttime 29 Si { 1 H} CP/MAS NMR spectra of the talc/10-Å phase product support the assignment of Q 2 Si to the proposed SiO 3 (OH) groups. Electron microprobe analysis, including oxygen, gives the following empirical formula normalized to three Mg apfu and inferring a hydrogarnet component Si → 4H associated with Si vacancies: Mg 3 Si 3.83(8) O 9.32 (OH) 2.68 .1.1(4)H 2 O. The observed Mg:Si indicates a significant Si deficiency relative to talc. Comparison of the 29 Si MAS NMR and microprobe data indicates that Si vacancies likely occur as single isolated entities, rather than as pairs or clusters, and that between 1 in 18 and 1 in 23 Si sites is vacant. The results suggest new and intriguing possibilities for the incorporation of excess H into the 10-Å phase and, potentially, other phyllosilicates under upper-mantle conditions.