Dehydration-melting of amphibolite at 10 kbar: the effects of temperature and time

We have simulated the dehydration-melting of a natural, low-K, calcic amphibolite (67.4% hornblende, 32.5% anorthite) in piston-cylinder experiments at 10 kbar and 750–1000°C, for 1–9 days. The solidus temperature is lower than 750°C; garnet appears at 850°C. The overall reaction is: Hb+PL+Cpx+Al-Hb+Ca-Hb+Ga+Opx. Three stages of reaction are: (1) melting dominated by the growth of clinopyroxene and garnet, with little change in composition of liquid or garnet, (2) a reversal of this reaction between 875°C and 900°C, with decreases in the amounts of liquid and garnet, and (3) a large increase in liquid along with the loss of hornblende and decrease of plagioclase while clinopyroxene and garnet increase. Garnet is enriched in pyrope and zoned from Fe-cores to Mg-edges (range 3 mol % pyrope); liquid composition is enriched first in An (to 950°C) and then in Ab. The liquids are more calcic and aluminous than natural tonalites, which is attributed to the plagioclase composition (An₉₀). The formation of peraluminous liquid from the metaluminous amphibolite is caused by anorthite — not H₂O-saturated conditions. The results are consistent with an amphibolite phase diagram with relatively high solidus temperatures in the garnet-absent field (900–1000°C), but with a solidus backbend at 7–9 kbar, coincident with the garnet-in boundary. Hornblende breakdown due to garnet formation in a closed system must make H₂O available for H₂O-undersaturated melting right down to the H₂O-saturated solidus, below 700°C, which defines a large low-temperature PT area where hydrous granitoid melts can be generated with residual garnet and hornblende.     

Dehydration-melting of Biotite Gneiss and Quartz Amphibolite from 3 to 15 kbar

We performed vapor-absent melting and crystallization experimentson two bulk compositions that model metamorphic rocks containinga single hydrous phase: a biotite gneiss [37% bio (mg-number55), 34% qtz, 27% plg (An₃₈), 2% ilm] and a quartz amphibolite[54% hbl (mg-number 60), 24% qtz, 20% plg (An₃₈), 2% ilm]. Experimentswere performed at 3 and 5 kbar in internally heated pressurevessels (IHPV), and at 7, 10, 125 and 15 kbar in piston cylinderapparatus (PC), from the vapor-absent solidi to (at least) thetemperature at which the hydrous mineral disappeared. Dehydration-meltingbegins at similar temperatures in both bulk compositions, rangingfrom T850C at P = 3 kbar T930C at P = 15 kbar. The hydrousmineral disappears 50C above the solidus in both systems, exceptin IHPV experiments at f(O₂) above Ni–NiO, in which biotitestability extends up to atleast 80C above the solidus. At theT at which the hydrous minerals disappear the biotite gneissproduces 2–3 times more melt than the quartz amphibolite(50–60 wt% vs 20–30 wt%). In both systems, variationsin melt productivity with P are controlled by three competingfactors: (1) the positive d P/dT slopes of the solidi, (2) decreasingH₂O activity with increasing P at constant H₂O content, and(3) Na₂O activity, which increases with P concomitantly withbreakdown of plagioclase. Melt productivities at T = 920–950Care maximized at intermediate pressures (7 kbar). The biotitegneiss produces strongly peraluminous granitic melts (SiO₂>70wt%) and residual assemblages of quartz norite (P>125 kbar)or garnet pyroxenite (P>125 kbar). The quartz amphiboliteproduces strongly peraluminous granodioritic melts (SiO₂>70wt%) that coexist with clinopyroxene + orthopyroxene + plagioclase+ quartz at P>10 kbar)garnet. The results of coupled meltingand crystallization experiments on the quartz amphibolite suggestthat strongly peraluminous granitoid rocks (e.g. cordierite-bearingand two-mica granites) can be derived from melting of Al-poorprotoliths. KEY WORDS: dehydration-melting; biotite gneiss; amphibolite; felsic magmas ^*Corresponding author     

Interaction of granitic and basic magmas: experimental observations on contamination processes at 10 kbar with H2O

Experiments designed to simulate the interaction of juxtaposed rhyolitic and basic magmas were conducted at 10 kbar with H₂O, using reaction-couples consisting of Westerly granite (WG) against basalt (DW-1) and WG against a synthetic mafic glass (SMG, enriched in MgO and Na₂O relative to DW-1). Each couple was run with 5 and 10 wt% H₂O corresponding respectively to H₂O-undersaturated and H₂O-oversaturated conditions. Experiments were run for 42–44 h at 920° C, above the liquidus of WG and within the melting intervals of DW-1 and SMG. WG was run above the basic material in all but one experiment. The composition of the granitic melt was altered only through material exchange with the adjacent basic melts, whereas that of the basic melts also changed (relative to the bulk basic composition) due to partial crystallization. Some crystallization also occurred within the zone of interaction. For control, the basic compositions were also run alone under the same conditions as the reaction-couple experiments. The crystalline phase assemblages in the basic ends of the coupled experiments differed from those produced from the basic materials alone, demonstrating interaction with the granite melt. Moreover, compositional gradients within the basic ends of coupled experiments are indicated by changes in phase assemblage and compositions with distance from the interface with WG. Microprobe analyses of glass collected along the length of the capsules confirm published observations that alkali diffusion is very fast: K₂O and Na₂O homogenized throughout the capsules in less than the two-day run times. This, coupled with the fact that introduction of K₂O into SMG stabilized biotite, produced the result that after interaction the bulk basic material (melt+crystals) contained more K₂O than the coexisting felsic melt. Only very gentle gradients for CaO, FeO, and MgO are preserved in our experiments, in contrast with published anhydrous results, suggesting that the difference in activity coefficients for these components between basic and felsic melts is reduced by the introduction of H₂O. Gradients for SiO₂ and Al₂O₃ are of comparable length to those of the divalent cations, confirming earlier results that the diffusivities of the network-formers limit the rate of diffusion of Ca, Fe, and Mg.     

Anatexis of interlayered amphibolite and pelite at 10 kbar: effect of diffusion of major components on phase relations and melt fraction

 We have conducted H₂O-undersaturated melting experiments on charges consisting of a layer of a sillimanite-bearing metapelite and a layer of garnet-bearing amphibolite, with the goal of studying how layered metamorphic rocks interact during anatexis. Experiments with the layered charges were performed at 10 kbar (1 GPa) at 900 °C (334 h), 925 °C (341 h), and 950 °C (50, 161, 332 h). Dehydration-melting of the amphibolite alone and of the pelite alone were studied at the same P-T conditions. When run alone at 900 °C, the pelite and the amphibolite yielded ca. 40 vol.% and 15 vol.% melt, respectively. Corresponding melt fractions at 950 °C were ca. 50 vol.% and 30 vol.%. When run side by side, melt abundance in the pelite is fairly uniform and ca. 20 vol.% higher than in the solo runs at all temperatures studied. In contrast, complex zoning patterns parallel to the interface develop in the amphibolite layer. Far from the interface the melt fraction is 5–10 vol.% lower than in the amphibolite-solo runs. Closer to the interface the melt fraction in the amphibolite layers are comparable to, or somewhat higher than, in the amphibolite-solo runs, and both clinopyroxene and plagioclase react-out and are replaced by orthopyroxene and quartz. Melt fraction closest to the interface reaches 65–70 vol.%, and the only crystalline phase present is garnet with hollow cores. The increase in melt fraction in the pelite layers of the layered runs is caused by transfer of Na from the amphibolite to the pelite, which forces biotite to be eliminated at lower temperature along the biotite-garnet reaction boundary. K is transferred from the pelite to the amphibolite, and the lower melt fraction in the distal end of the amphibolite layers is caused by faster diffusion of Na compared to K, which results in a net depletion of alkalies. Ca is transferred from the amphibolite to the pelite, and the resulting changes in CaO concentrations affect the stabilities of plagioclase, pyroxenes, garnet and quartz and cause the observed zoning in the amphibolite layer. Ca entering the pelite layer induces crystallization of large plagioclase neoblasts radiating from the interface into the pelite layer. These results show that diffusion of components between contiguous lithologies undergoing anatexis can cause profound changes in phase relations, melt fractions and restite compositions. Received: 21 December 1994 / Accepted: 26 June 1995     

Liquidus phase relationships on the join anorthite-forsterite-quartz at 10 kbar with applications to basalt petrogenesis

Liquidus phase relationships determined on the join CaAl₂Si₂O₈ (anorthite)-Mg₂SiO₄ (forsterite)-SiO₂ (quartz) at 10 kbar show that increasing pressure causes the forsterite and anorthite primary phase fields to shrink and the spinel, enstatite and silica fields to expand. The boundary line between the enstatite and forsterite fields and that between the enstatite and quartz fields both move away from the SiO₂ apex as pressure increases. Therefore, simplified source peridotite would yield simplified basaltic partial melts with decreasing silica as pressure increases, as has been found in other studies. Also, increasing pressure decreases the amount of silica enrichment in residual liquids produced by fractional crystallization. Although anorthite is unstable in simplified peridotite above 9 kbar in the system CaO-MgO-Al₂O₃-SiO₂, it is an important phase in the fractional crystallization of simplified basalts at 10 kbar and probably also in natural basalts.Contribution no. 419, Department of Geosciences, University of Texas at Dallas     

Partial fusion of basic granulites at 5 to 15 kbar: implications for the origin of TTG magmas

Partial fusion experiments with basic granulites (S6, S37) believed to represent the lower crust beneath the Eifel region (Germany) were performed at pressures from 5 to 15 kbar. Water-undersaturated experiments were carried out in the presence of 1 wt% H₂O plus 2.44 or 0.81 wt% CO₂ equivalent to mole fractions of H₂O/(H₂O + CO₂) of 0.5 and 0.75, respectively, of the volatile components added. At temperatures from 850 to 1100 °C the weight proportions of melt range from 7 to 30 %. Melt compositions change from trondhjemitic over tonalitic to dioritic with increasing degree of partial melting. Crystalline residua are plagioclase/pyroxene dominated at 5 kbar to garnet/pyroxene dominated at 15␣kbar. Dehydration melting was studied in granulite S35 similar in composition to S6. The magmatic precursors of the granulite xenoliths used in this study had geochemical characteristics of cumulate gabbro (metagabbro S37) and evolved melts (metabasalts S6, S35), respectively. Melts from granulite S37 match the major element compositions of natural trondhjemites and tonalites. At 5 kbar, their Al₂O₃ is relatively low, similar to tonalites from ophiolites. At 15 kbar, Al₂O₃ in the melts is high due to the near absence of plagioclase in the crystalline residua. The Al₂O₃ concentrations in 15 kbar melts from S6 (˜20 wt%) are higher than in natural tonalites. Depth constraints on the formation of tonalitic magmas in the continental crust are provided by REE (rare earth element) patterns of the synthetic melts calculated from the known REE abundances in metagabbro S37 and metabasalt S6 assuming batch melting and using partition coefficients from the literature. The REE patterns of tonalites from active continental margins and Archean trondhjemite-tonalite-granodiorite␣associations low in REE with La_N (chondrite normalised) from 10 to 30 and Yb_N from 1 to 2 are reproduced at pressures of 10 and 12.5 kbar from metagabbro S37 which displays a slightly L(light)REE enriched pattern with La_N = 8 and Yb_N = 3. Natural tonalites with La_N from 30 to 100 require a source richer in REE than granulite S37. At 15 kbar, H(heavy)REE_N in melts from granulite S37 are depressed below the level observed in natural tonalites due to the high proportion of garnet (>30 wt%) in the residue. Melts from metabasalt S6 (enriched in REE with La_N = 38 and Yb_N = 16) do not match the REE characteristics of natural tonalites under any conditions. Received: 1 July 1994 / Accepted: 11 September 1996     

The system granite-peridotite-H2O at 30 kbar,with applications to hybridization in subduction zone magmatism

Experiments with mixtures of granite, peridotite and H₂O at 30 kbar were designed as a first step to test the hypothesis that the calc-alkaline igneous rocks of subduction zones are formed by differentiation of magmas derived by partial melting of hybrid rocks generated in the mantle wedge, by reaction between hydrous siliceous magma rising from subducted oceanic crust, and the overlying mantle peridotite. Experiments were conducted in gold capsules in half-inch diameter piston-cylinder apparatus. Results are presented in a 900° C isotherm, and in a projection of vapor-present phase fields onto T-granite-peridotite. Isobaric solution of peridotite in hydrous, H₂O-undersaturated granite liquid at 900° C causes only small changes in liquid composition, followed by precipitation of orthopyroxene until about half of the liquid has solidified; then orthopyroxene is joined by jadeitic clinopyroxene, garnet, and phlogopite. Phlogopite-garnet-websterite continues to be precipitated, with evolution of aqueous vapor, until all of the liquid is used up. The product of hybridization is a pyroxenite without olivine. The products of partial melting of this material would differ from products derived from peridotite because there is no olivine control, and the clinopyroxenes contain up to 7% Na₂O, compared with less than 1% Na₂O in peridotite clinopyroxenes. The reaction products are directly analogous to those in the model system KAlSiO₄-Mg₂SiO₄-SiO₂-H₂O, where, with decreasing SiO₂ in the hydrous siliceous liquid, the field for phlogopite expands, and phlogopite instead of orthopyroxene becomes the primary mineral. If this occurs with less siliceous magmas from the subducted oceanic crust, there is a prospect for separation of discrete bodies of phlogopite-rock as well as phlogopite-garnet-websterite. We need to know the products of hybridization, and the products of partial melting of the hybrid rocks through a range of conditions.     

The system tonalite-peridotite-H2O at 30 kbar,with applications to hybridization in subduction zone magmatism

We present data on the phase relationships of mixtures between natural tonalite and peridotite compositions with excess H₂O at 30 kbar, and on the composition of the piercing point where the peridotite-tonalite mixing line intersects the L(Ga,Opx) reaction boundary. These data, in conjunction with earlier analogous data along peridotite-granite and basalt-granite mixing lines, permit construction of a pseudoternary liquidus projection that is relevant to interaction of peridotite with slab-derived magmas. Knowledge of the liquidus phase and temperature for a range of compositions within this projection enables us to map primary crystallization fields for quartz, garnet, orthopyroxene, clinopyroxene, and olivine, and to estimate the distribution of isotherms across the projection. Using this projection, we explore the consequences of peridotite assimilation by mafic to intermediate (basalt to dacite) hydrous slab-derived melts. Progressive assimilation under isothermal conditions results in garnet precipitation as the melt composition traverses the garnet liquidus surface and then garnet+orthopyroxene crystallization once the melt reaches the L(Ga,Opx) field boundary. The melt is constrained to remain on this field boundary and further assimilation of peridotite simply results in continued precipitation of garnet+orthopyroxene until the melt is consumed. The product is a hybrid solid assemblage consisting of Ga+ Opx. It is noteworthy that this process drives the melt composition in a direction nearly perpendicular to the mixing line between peridotite and the initial melt. If assimilation occurs with increasing temperature (as might occur if a slab-derived magma rises into the hotter mantle wedge), intermediate magmas (e.g. andesites) will again precipitate garnet until they reach the L(Ga,Opx) reaction boundary at which point Ga re-dissolves and orthopyroxene precipitates as the melt composition moves up-temperature along this boundary. The product of this process is a hybrid solid assemblage with garnet subordinate to orthopyroxene. For more mafic initial compositions (e.g. basalts) originally plotting in the Cpx field, it appears possible to avoid field boundaries involving garnet and shift in composition more directly toward peridotite, if assimilation is accompanied by a sharp increase in temperature. Considering published REE evidence (arguing against garnet playing a significant role in the genesis of many subduction-related magmas) in light of our results, it appears unlikely that peridotite assimilation by intermediate magmas under conditions of constant or increasing temperature is an important process in subduction zones. However, if assimilation is accompanied by an increase in temperature, our data do permit the derivation of high-Mg basalts from less refractory precursors (e.g. high-Al basalts) by peridotite assimilation.     

On the role of diapirism in the, segregation, ascent and final emplacement of granitoid magmas

Only bodies of magma with a high crystal content and partially molten (crustal) country rocks can ascend as diapirs; once such an envelope is pierced, the diapiric ascent of the pluton is arrested by the high viscosity of a solid aureole. Deformation by shortening of the carapace of these bodies may lead to the expulsion of a magma with a relatively low crystal content, which may then continue ascent via fractures and dykes.

The details of the mechanisms of granitoid magma segregation are still unknown, but it appears that many magmas hegin their ascent through the crust as mushes with at least 50% melt, and that such magmas are rheologically able to ascend through a thickness of crust. This ascent mechanism explains the dearth of structures attributable to the ascent of granitoids, in contrast to the abundance of structures that developed during their final emplacement.

When a magma becomes too crystalline (melt < 25%) to continue its ascent via dykes, it is immobilised. At approximately this stage, a hydrous magma may become saturated with water and release fluids into the aureole, making it particularly susceptible to deformation. Magma that continues to arrive at this level is also immobilised, and the pluton grows as a ballooning diapir. These characteristically deform themselves and their aureoles by bulk shortening.

Magmas that are able to ascend to shallow depths, largely by virtue of lower water contents and higher initial temperatures, tend to become finally accommodated by such brittle processes as stoping and cauldron subsidence. High level intrusions lend to be tabular, are also fed by dykes or conduits, and assemble in tabular batholiths.     

Liquidus phase relationships on the join anorthiteforsterite-quartz at 20 kbar with applications to basalt petrogenesis and igneous sapphirine

Liquidus phase relationships determined on the join anorthite-forsterite-quartz at 20 kbar show primary phase fields for quartz (q), forsterite (fo), enstatite (en), spinel (sp), anorthite (an), sapphirine (sa), and corundum (cor). Increasing pressure causes (1) thefo andan primary phase fields to contract, (2) theen, q, andcor fields to expand, (3) thefo-en boundary line to move away from the Q apex, (4) theen-q boundary line to move also away from the Q apex but by a smaller amount, and (5) a primary phase field forsa to appear at a pressure between 10 and 20 kbar. Seven liquidus piercing points at 20 kbar have been located as follows:

Empirical equations to predict the sulfur content of mafic magmas at sulfide saturation and applications to magmatic sulfide deposits

Empirical equations to predict the sulfur content of a mafic magma at the time of sulfide saturation have been developed based on several sets of published experimental data. The S content at sulfide saturation (SCSS) can be expressed as:

Composition of liquids coexisting with spinel lherzolite at 10 kbar and the genesis of MORBs

Because of the controversy over the nature of the parental magma for MORBs, experiments have been performed at 10 kbar in order to assess the effect of modal variations in the source peridotite and the effect of temperature (degree of partial melting) on the composition of partial melts. A peridotite-basalt sandwich method was used and a run duration of 72 h was found to be necessary to equilibrate basalt and peridotite. A range of melt compositions, coexisting with olivine, orthopyroxene, clinopyroxene and spinel, was produced at 10 kbar, indicating that partial melting of peridotite cannot be regarded as isobarically pseudoinvariant. On projections in the normative tetrahedron OL-PL-CPX-SIL, the liquids obtained in this study define an area, rather than a point or narrow band. The compositions of some liquids in this study are similar to magnesian MORBs (MgO>9.5 wt%), providing evidence in support of the derivation of magnesian MORBs by partial melting of mantle lherzolite at about 10 kbar.     

Subsolidus and melting relations for the join CaCO3-MgCO3 at 10 kbar

Mixtures of pure dry CaCO₃ and MgCO₃ were reacted at 10 kbar in a piston-cylinder apparatus. Solidus and liquidus boundaries were delineated by interpretation of quenched textures. X-ray determined compositions of quenched carbonates are not a reliable guide to the phase relations. The binary melting loop for CaCO₃-MgCO₃ extends from CaCO₃ at 1460°C through a liquidus minimum near 30 wt% MgCO₃ and 1075°C, and it is terminated at the incongruent melting reaction for dolomite solid solution at 1125° C (liquid with 32 wt% MgCO₃) Magnesite solid solution dissociates at 1090°C to produce dolomite + periclase + CO₂, truncating the dolomite-magnesite solvus. The 10 kb liquidus minimum at 1075°C and 30 wt% MgCO₃ occurs at lower temperature and higher $\text{Ca}\text{Mg}$ ratio than the 27 kbar liquidus minimum at 1290°C and 38 wt% MgCO₃. This relationship suggests that the first liquid produced by melting of a carbonate-bearing peridotite has increasing $\text{Mg}\text{Ca}$ ratio with increasing pressure. These phase relations provide part of the framework required to trace paths of crystallization of kimberlite and carbonatite magmas.     

Some trace element relationships among liquid and solid phases in the course of the fractional crystallization of magmas

The classical equations relating the trace element concentrations of the liquid and solid phases coexisting in the simple fractional crystallization of a parental magma have been put in a simple graphical form, which allows rapid analysis of the possible genetic relationships in a given rock suite. The effects of an incomplete separation between the two phases are taken into account. The approach does not require the use of otherwise estimated partition coefficients. Trace element data concerning the minerals of cumulates, where available, may provide an independent estimation of the effective mineral-liquid partition coefficients. With reasonable assumptions, this approach may even be applied to plutonic rocks. Interpretation of the published rare earth element data from the Southern California Batholith by this procedure suggests that a tonalitic parental magma could generate a granodioritic liquid by crystallizing 40–50 wt % of a solid residue of gabbroic composition, in agreement with Larsen's (Mem. Geol. Soc. Amer. 29, 1948) calculations. The calculated mineral-liquid partition coefficients for the REE fall in the range of published phenocryst-groundmass values for acidic volcanic rocks.     

Experiments and models of anhydrous,basaltic olivine-plagioclase-augite saturated melts from 0.001 to 10 kbar

 A new method for modeling fractional crystallization processes that involve olivine (ol), plagioclase (plag) and augite (aug) is presented. This crystallization assemblage is the major control on the chemical variations in mid-ocean ridge basalts. The compositional and temperature variations in ol-plag-aug saturated basalts over a range of pressures are described using empirical expressions. A data base of 190 experiments in natural and basalt-analog chemical systems is used to describe temperature, Al, Ca and Mg molar fractions as functions of Si, Fe, Na, Ti and K molar fractions and pressure. Increases in the abundances of Na and K cause Ca and Mg abundances to decrease and Al abundance to increase in ol-plag-aug saturated melts. The equations can be used to predict pressure and temperature and thus provide a useful thermobarometer. A model is described to calculate ol-plag-aug fractional crystallization as a function of pressure and melt composition, using melt and augite models developed here, combined with existing models for olivine-melt and plagioclase-melt equilibria. We compare the fractional crystallization sequence of ALV-2004-3-1 predicted from the models presented in this paper, Langmuir et al. (1992) modified by Reynolds (1995), Ghiorso and Sack (1995) and Ariskin et al. (1993) at 0.001 and 4 kbar. As an example the model is applied to estimate pressure of crystallization of glasses from the east flank of the East Pacific Rise at 11°45′N. Received: 24 July 1995 / Accepted: 12 January 1996     

Temperature dependence of the compressional wave velocity in an anisotropic dunite: Measurements to 500°C at 10 kbar

Measurements of compressional wave velocity V_p were made in a gas apparatus to 500°C at 10 kbar in three cores of an anisotropic dunite specimen from Twin Sisters Mountain. The axial directions of the three chosen cores coincide with the preferred directions and concentration of olivine crystallographic axes (a [100], b [010], andc [001]).Measured (δV_p/δT)p values at 10 kbar in the three cores (−6.7, −5.4 and −6.2 · 10⁻⁴ km/sec · deg, respectively), and the mean value for the dunite (−6.1 · 10⁻⁴ km/sec · deg) are larger than the Voigt-Reuss-Hill values calculated from single-crystal data. This discrepancy is explained by the presence of internal thermal stresses, due to anisotropic expansion of olivine grains, causing grain boundary cracks to widen.It is concluded that high negative values of (δV_p/δT)p for rocks reported in the literature should be carefully evaluated in terms of the formation of new cracks or widening of cracks already present under high pressure-temperature environments.     

Controls on gold solubility in arc magmas: An experimental study at 1000 °C and 4 kbar

In order to (1) explain the worldwide association between epithermal gold-copper-molybdenum deposits and arc magmas and (2) test the hypothesis that adakitic magmas would be Au-specialized, we have determined the solubility of Au at 4 kbar and 1000 °C for three intermediate magmas (two adakites and one calc-alkaline composition) from the Philippines. The experiments were performed over a fO₂ range corresponding to reducing (∼NNO−1), moderately oxidizing (∼NNO+1.5) and strongly oxidizing (∼NNO+3) conditions as measured by solid Ni-Pd-O sensors. They were carried out in gold containers, the latter serving also as the source of gold, in presence of variable amounts of H₂O and, in a few additional experiments, of S. Concentrations of Au in glasses were determined by LA-ICPMS. Gold solubility in melt is very low (30-240 ppb) but increases with fO₂ in a way consistent with the dissolution of gold as both Au¹⁺ and Au³⁺ species. In the S-bearing experiments performed at ∼NNO−1, gold solubility reaches much higher values, from ∼1200 to 4300 ppb, and seems to correlate with melt S content. No systematic difference in gold solubility is observed between the adakitic and the non-adakitic compositions investigated. Oxygen fugacity and the sulfur concentration in melt are the main parameters controlling the incorporation and concentration of gold in magmas. Certain adakitic and non-adakitic magmas have high fO₂ and magmatic S concentrations favorable to the incorporation and transport of gold. Therefore, the cause of a particular association between some arc magmas and Au-Cu-Mo deposits needs to be searched in the origin of those specialized magmas by involvement of Au- and S-rich protoliths. The subducted slab, which contains metal-rich massive sulfides, may constitute a potentially favorable protolith for the genesis of magmas specialized with respect to gold.     

Water and magmas: clarification of a controversy on applications of the Gibbs-Duhem equation

It is shown mathematically that if the activity coefficient of water in ternary water-magma (aluminosilicate) systems is constant or varies only with the mole fraction of water, it is not necessary that the binary magmas form ideal solutions contrary to the claims by Burnham et al. (1978, Geochim. Cosmochim. Acta42, 275–276). A molecular viewpoint is presented to support this argument. The properties of analytical equations capable of representing the activity coefficients of usual and unusual systems are discussed. The correct form of the Gibbs-Dunhem equation for dissociative dissolution processes is presented to disprove the claims by Burnham (1975, Fortschr. Mineral.52, 101–118; 1975, Geochim. Cosmochim. Acta39, 1077–1084), and by Burnhamet al. (1978, Geochim. Cosmochim. Acta42, 275–276).     

Anhydrous partial melting of an iron-rich mantle I: subsolidus phase assemblages and partial melting phase relations at 10 to 30 kbar

Anhydrous partial melting experiments, at 10 to 30 kbar from solidus to near liquidus temperature, have been performed on an iron-rich martian mantle composition, DW. The DW subsolidus assemblage from 5 kbar to at least 24 kbar is a spinel lherzolite. At 25 kbar garnet is stable at the solidus along with spinel. The clinopyroxene stable on the DW solidus at and above 10 kbar is a pigeonitic clinopyroxene. Pigeonitic clinopyroxene is the first phase to melt out of the spinel lherzolite assemblage at less than 20°C above the solidus. Spinel melts out of the assemblage about 50°C above the solidus followed by a 150° to 200°C temperature interval where melts are in equilibrium with orthopyroxene and olivine. The temperature interval over which pigeonitic clinopyroxene melts out of an iron-rich spinel lherzolite assemblage is smaller than the temperature interval over which augite melts out of an iron-poor spinel lherzolite assemblage. The dominant solidus assemblage in the source regions of the Tharsis plateau, and for a large percentage of the martian mantle, is a spinel lherzolite.     

Anhydrous partial melting of MORB pyrolite and other peridotite compositions at 10 kbar: Implications for the origin of primitive MORB glasses

Summary Anhydrous partial melting experiments on four peridotite compositions have been conducted at 10 kbar providing a relatively internally consistent set of data on the character of primary melts expected from the oceanic upper mantle in the mid-ocean ridge setting. The four peridotite compositions are: MORB pyrolite (considered to be suitable for the production of primitive (Mg#0.68) MORB glasses at 10 kbar), Hawaiian pyrolite (representative of enriched upper mantle), Tinaquillo lherzolite (representative of more depleted upper mantle), and the spinel lherzolite KLB-1 which is a suitable composition for the production of primitive MORB glasses. The equilibrium liquids were determined by sandwich experiments. The primitive MORB glass DSDP 3-18-7-1 was used in experiments using MORB pyrolite and KLB-1, while a calculated 10 kbar liquid composition fromJaques andGreen (1980) was used in experiments with Hawaiian pyrolite and Tinaquillo lherzolite. The results of the experiments are used to test a 10 kbar melt model for the generation of primitive MORB glasses, which are parental magmas to typical MORB compositions. The melt compositions from the four peridotites studied are significantly different from primitive MORB glasses in major element chemistry and plot away from the field of primitive MORB glasses in the CIPW molecular normative Basalt tetrahedron. The results indicate that primitive MORB glasses are derivative compositions lying on olivine fractionation lines from picritic parents, which themselves are primary magmas at pressures greater than 10 kbar. The results of this study are integrated with previous 10 kbar experimental studies.

---

Wasserfreie partielle aufschmelzung von MORB pyrolit und andere peridotit-zusammensetzungen bei 10 kbar: bedeutung für die entstehung primitiver MORB gläser

Zusammenfassung Vier Peridotit-Zusammensetzungen wurden bei 10 kbar unter wasserfreien Bedingungen partiell aufgeschmolzen, und die Ergebnisse mit möglichen primitiven Schmelzen Mittel-Ozeanischer Rücken verglichen.Die folgenden perioditischen Zusammensetzungen wurden untersucht: MORB pyrolite [mögliche Ausgangszusammensetzung für primitive (Mg# > 0.68) MORB-Glaszusammensetzungen bei 10 kbar], Hawaiian pyrolite (representativ für angereicherten Oberen Mantel); Tinaquillo lherzolite (representativ für verarmten' Oberen Mantel) und spinel lherzolite, KLB-1 (im Gleichgewicht mit primitiver MORB-Glaszusammensetzung). Die Schmelzen im Gleichgewicht mit diesen Ausgangszusammensetzungen wurden mittels Sandwich-Experimenten ermittelt.Die primitive MORB-Glaszusammensetzung DSDP 3-18-7-1 wurde mit MORB pyrolite und KLB-1 equilibriert, während eine Modell-Zusammensetzung vonJaques and Green (1980) in Verbindung mit Hawaiian pyrolite und Tinaquillo lherzolite vermischt wurde. Die Resultate der Experimente werden mit einem 10 kbar Aufschmelzungsmodell zur Entstehung primitiver MORB-Gläser verglichen. Die Schmelzen im Gleichgewicht mit den vier Peridotit-Ausgangszusammensetzungen unterscheiden sich wesentlich von primitiven MORB-Gläsern, sowohl hinsichtlich ihrer Hauptelemente als auch ihrer Plot-Parameter im Basalttetraeder. Primitive MORB-Glaszusammensetzungen stellen keine primären Schmelzen dar, sondern sind durch Olivinfraktionierung von primitiven Magmen abzuleiten. Die Resultate dieser Untersuchungen werden mit früheren 10 kbar Experimenten verglichen.

---

---

With 10 Figures