James A. Whitney

The origin of granite: The role and source of water in the evolution of granitic magmas

Temperature and water content are the two most important parameters in the formation of granitic magmas. Evidence from volcanic and plutonic lithologies suggests that water contents of 2 to 4 wt. % are present in most silicic magmas. Calculations based on the stability of biotite yield water fugacities within the melt phase from about 500 to 2,000 bars, although these calculations determine the log f H 2 O and the effective uncertainty of ±0.5 log units yields a large absolute uncertainty. Comparison with crystallization experiments demonstrates that less than 2% water would require significant percentages of crystallization at temperatures above 900 °C with liquidus temperatures about 1000 °C. Water contents greatly in excess of 4 wt. % would mean that the magma would become vapor saturated at high pressures and would tend to crystallize during ascent to a fine-grained granite before reaching shallow depths. The main sources of water for magma generation are the dehydration of hydrous silicates within the crust and volatiles transported into the crust from subducted oceanic crust and upper mantle in the form of hydrous basalts and andesites. Dehydration reactions of muscovite, biotite, and horn-blende are of particular significance. Anatectic granites may be partially classified in terms of the probable dehydration reaction responsible for their generation. Melts generated from muscovite dehydration are relatively cool, peraluminous in composition, high in K/Na ratio, and generally high in initial 87 Sr/ 86 Sr and delta 18 O. Biotite-generated melts tend to be higher in temperatures, peraluminous to meta-aluminous in composition, moderately high in K/Na ratio, and relatively high in strontium and oxygen ratios. Both types may contain metasedimentary enclaves. Granitoids generated by hornblende dehydration would be much higher in initial temperature, peralka-line to meta-aluminous in composition, and lower in K/Na ratio and in general would have lower strontium and oxygen ratios. Volatiles deposited in hydrothermally altered oceanic crust and upper mantle will be released during subduction in the form of hydrous basalts and andesites. These melts are important energy-transfer mechanisms to drive anatexis within the crust. If these melts encounter a silicic magma chamber when intruding the crust, they will be trapped below the lower-density granitic melt. While partially quenching against the cooler melt, they may release volatiles and heat, which will aid in the melting process. Granites generated in this way will vary in their geochemical properties, depending on the relative importance of crustal versus mantle magma systems. In general, they will have moderate to high initial temperatures, be meta-aluminous in composition, be variable but somewhat low in K/Na ratio, and have lower initial strontium and delta 18 O ratios than will melts derived completely from the crust. The fact that most terrestrial spreading centers are subaqueous thus may aid in the formation of a thick granitic continental crust.

Age and origin of the Stone Mountain Granite, Lithonia district, Georgia

Petrographic, geochemical, and strontium isotopic studies have been conducted on samples of the Stone Mountain Granite (leucocratic quartz monzonite) from the Lithonia district, within the Inner Piedmont province of Georgia. A Rb-Sr isochron obtained from 10 whole-rock and 3 mineral samples yielded an age of 291 ± 7 m.y. and an initial Sr 87 Sr 86 ratio of 0.7250 ± 0.0005. Petrologic and geochemical data, combined with the high initial strontium isotopic ratio, suggest that the origin of this peraluminous quartz monzonite can best be explained by the anatexis of an older peraluminous, granitic crustal material during Late Pennsylvanian time. The depth of intrusion was probably around 12 km, with initial magma temperatures of 700°C or less. Magma generation could have occurred at depths of 22 to 28 km, depending on the regional geothermal gradient at the time. The most likely source material is the Lithonia Gneiss, which has a peraluminous, granitic composition and underlies the area. The Stone Mountain pluton extends the 300-m.y.-old intrusive belt in the southern Appalachians across the Piedmont province to just south of the Brevard lineament. Its implied depth of intrusion fits well with certain models for the development of the Inner Piedmont and Blue Ridge provinces in the Carolinas because it requires rapid removal of considerable overburden. The origin and magmatic history of this intrusive unit may find use as a model for other peraluminous granitic plutons having similar petrologic, geochemical, and isotopic characteristics.

Igneous sulfides in the Fish Canyon Tuff and the role of sulfur in calc-alkaline magmas

Pyrrhotite is a common inclusion in several phenocrystal phases from the Fish Canyon Tuff, a voluminous, homogeneous ash flow erupted from the Central San Juan field. The fugacities of various gaseous species have been calculated using the composition of the pyrrhotite, iron-titanium oxides, and biotite. Sulfur and oxygen fugacities fall very near the sulfur condensation curve, implying that liquid sulfur may have been present in the parent magma. High sulfur fugacities during melt formation can cause high oxygen fugacity in calc-alkaline magmas due to the breakdown of pyrite at high temperatures. During cooling of such magmas, separation of SO 2 -rich gases causes a reduction in oxygen fugacity. Significant amounts of sulfur are available during degassing and represent a major source of sulfur for related ore deposits. The ultimate source of the high sulfur and oxygen activities is probably pyrite and magnetite deposited during the waning stages of hydrothermal circulation at active oceanic ridges. Other ash flows and calc-alkalic volcanic rocks should be carefully examined for similar sulfide phases.

Two-feldspar geothermometry in granulite facies metamorphic rocks

A geothermometric technique based on equilibria between coexisting plagioclase and alkali feldspar was applied to quartzo-feldspathic granulites from Salvador, BA, Brazil. The conditions of metamorphism were determined to be in the range 750 ° C–800 ° C, 4–8 Kb, by comparison with experimental data on the stabilities of sapphirine, phlogopite and other minerals occurring in the associated rocks. Selected feldspar data gives temperatures near, but slightly below, this range. Several variants of the Wood and Banno model, as well as an empirical two-pyroxene geothermometer, were also tested and found to give temperatures which were apparently 50 °–100 ° high. The solubility of Al2O3 in orthopyroxene indicates temperatures which are about 200 ° to high, suggesting that Fe in the natural assemblages significantly changes relationships observed experimentally in MgO-Al2O3-SiO2 systems.

Two-feldspar geothermometry, geobarometry in mesozonal granitic intrusions: Three examples from the Piedmont of Georgia

The equilibrium temperatures for coexisting plagioclase and potassium feldspar pairs have been calculated for various textural varieties of feldspar from 3 post-metamorphic granites from the Georgia Piedmont; the Danburg, Siloam, and Stone Mountain plutons. Assuming an intermediate structural state for the feldspars at time of equilibration, crystallization temperatures match those expected from experimental data for quartz monzonite magmas (650 to 780° C). The variations in solidus temperature, recorded in the feldspars, may be used to estimate relative differences in depth of intrusion. Sharp reversals in plagioclase compositional trends may be caused by isothermal decreases in confining pressure associated with upward migration through the crust. In fine grained and slowly cooled intrusions, albite tends to be lost from the alkali feldspar grains, and recrystallizes as separate unzoned grains of oligoclase, thus erasing the previous thermal history. Perthite exsolution and re-equilibration within the alkali feldspar grains appears to continue down to temperatures of 400° C or so, although the zoned plagioclase does not homogenize. The recrystallization associated with changes in structural state may facilitate exsolution within alkali feldspar grains.

Model for the intrusion of batholiths associated with the eruption of large-volume ash-flow tuffs.

Pyroclastic eruption and the intrusion of batholiths associated with large-volume ash-flow tuffs may be driven by a decrease in reservoir pressure caused by the low density of the magma column due to vesiculation. Batholithic intrusion would then be accomplished by the subsidence and settling of kilometer-sized crustal blocks through the magma chamber, resulting in eventual collapse to form large caldera structures at the surface. Such a model does not require the formation of a large, laterally extensive, shallow magma chamber before the onset of large-volume ash-flow eruptions. Eruption could commence directly from a deeper reservoir, with only a small channelway being opened to the surface before the onset of catastrophic ash-flow eruptions of the scale of Yellowstone or Long Valley. Such a model has wide-ranging implications, and explains many of the problems inherent in the simple collapse model involving shallow magna chambers as well as the process and timing of batholith intrusion in such cases.

Mineralogical constraints on the petrogenesis of trachytic inclusions, Carpenter Ridge Tuff, Central San Juan volcanic field, Colorado

Although bulk-rock normative analyses of the trachytic inclusions from the Carpenter Ridge Tuff yield abundant quartz and minor corundum, a portion of the phenocryst assemblage is indicative of an alkaline parentage. Sanidine and biotite contain up to 8 and 5 wt% BaO respectively. In addition, both amphibole and clinopyroxene compositions are compatible with having crystallized from a mildly silica-undersaturated magma. Amphibole is magnesiohastingsite with 3 wt% TiO2 and less than 0.3 mole fraction vacancies in the A site. Clinopyroxene compositions straddle the calcic augite-salite boundary. Chrondite-normalized REE patterns are similar for both inclusions and rhyolites. The inclusions are slightly poorer in REE and have a positive Eu anomaly versus the negative anomaly of the rhyolites. The similarity in REE patterns would seem to indicate that the two rock types are genetically related with the positive Eu anomalies resulting from feldspar accumulation. However, this possibility is denied by the antithetic alkaline and subalkaline phenocryst assemblages of the two rock types. We suggest that the best explanation for these discrepancies is that a mildly silica-undersaturated magma was the parent for the phenocrysts. This magma intruded the Carpenter Ridge chamber, and because the crystallization temperatures of both magma overlapped, the alkaline magma mixed with the ambient rhyolite to form a hybrid. This hybrid consisted of a portion of the phenocryst assemblage from the alkaline magma but the bulk-rock chemistry depended upon the proportions of the endmember liquids. The abundance of normative quartz, the minor normative corundum, and the similarity of REE patterns indicates that the inclusions are mixtures dominated by the rhyolitic component. Additional processes such as liquid-state diffusion, crystal accumulation, and alkali loss may have contributed to obscure the compositions of the initial liquids involved in the inferred mixing process.

Fish Canyon Tuff, Colorado: the problem of two magnetic polarities in a single tuff

Abstract The Fish Canyon Tuff, located in the central San Juan Mountains, southern Colorado, is currently recognized as one of the largest ash-flow tuffs in the world. Paleomagnetic samples from 21 sites have been obtained from a composite vertical section of the tuff. The two uppermost sites, from the otherwise normally magnetized tuff, exhibit stable but anitpodal (reversed) remanent magnetic (RM) directions. Field relationships and geochemical consistencies indicate that the entire section was emplaced relatively rapidly, probably in