Kaolinite, montmorillonite and mixed-layer clays, iron oxy-hydroxides, bauxite minerals, and various silica minerals com-prise the usual insoluble phases that form during weathering ofsilicate material. Vermiculite forms as an intermediate productfrom a variety of silicates, but nonmixed layer illite has not beenobserved to form as a surface-weathering product directly fromsolution, only as an intermediate product derived from initialmuscovite. The activities of potassium ion and silica are almostnever high enough in weathering environments to form dis-crete illite, and the stability of montmorillonite/mixed-layerclays is favored by the presence of exchangeable cations (so-dium, calcium, and magnesium) and tetrahedral aluminum-silicon polymerized chains remaining from feldspar break-down. Therefore, conditions within the pit solutions duringevaporation must be fundamentally different from the moretemperate weathering environment, not only with respect totemperature, but also solution chemistry.

The thinness of the material underlying the pits indicates thatpit growth is keeping pace with infiltration and alteration. This,together with the indication of partial-equilibrium, allows apreliminary estimate for the formation rate. Mass transfer ofmaterial during hydrolysis of plagioclase can be roughly esti-mated based on work by Helgeson, Carrels, and Mackenzie(1969). To a first approximation, hydrolysis of plagioclase in theunderlying material results in the destruction of 0.3 gram ofplagioclase and precipitation of 0.002 gram of illite for every1,000 grams of water. A 5-cubic-centimeter pit represents theremoval of approximately 10 grams of plagioclase. If the average

water flux experienced during growth of the pit is the present-day average of about I gram per year, then at least 30,000 yearsare required to form the pit.

Refinement of this first approximation requires recalculationof the rates from Helgeson, Brown, and Leeper (1969) for thedolerite under antarctic conditions, and direct measurements ofthe water flux at these surfaces as well as the solution chemis-tries of the water. It is expected that such refinement will length-en the time required to form a 5-cubic-centimeter pit to morethan 100,000 years.

This work was supported by National Science Foundationgrant DPP 82-06391.

References

Helgeson, H. C., T. H. Brown, and R. H. Leeper. 1969. Handbook of the-oretical activity diagrams depicting chemical equilibrium in geologic systemsI nvolving all phase at one ATM and 0 degrees to 300 degrees C, SanFrancisco: Freeman, Cooper and Co.

Helgeson, H.C., R.M. Carrels, and F.T. Mackenzie. 1969. Evaluation ofirreversible reactions in geochemical processes involving mineralsand aqueous solutions-11. Applications. Geochimica CosmochimicaActa, 33, 455-481.

Helgeson, 1-IC., W.M. Murphy, and P. Aagaard. 1984. Thermodynamicand kinetic constraints on reaction rates among minerals and aque-ous solutions. Geocinnuca Cosmochimica Ada, 48, 2405-2432.

Thermobarometry of two-pyroxene-granulite

inclusions in Cenozoic volcanic rocksof the McMurdo Sound region

J.H. BERG and D.L. HERZ

Department of GeologyNorthern Illinois Un iversiti

DeKalh, Illinois 60115

Two-pyroxene granulites are very abundant as inclusions inthe Cenozoic volcanic rocks of the McMurdo Sound region. Inaddition to plagioclase, these granulites contain orthopyrox-ene, clinopyroxene, and either olivine, spinel, or garnet.Hornblende, quartz, ilmenite, apatite, biotite, sanidine, orscapolite are present in some of these inclusions, and in somecases one of the two pyroxenes is absent. Garnet is rare and isfound only as corroded cores surrounded by fine- grained sym-plectites of olivine, plagioclase, spinel, and orthopyroxene.However, pseudomorphs of garnet, consisting of the abovesymplectites, are more common. Locally, spinel shows evi-dence of having reacted with pyroxenes to form symplectites ofolivine and plagioclase. Not uncommonly the mafics occur in

lenses or clusters. The red-brown to orange-red hornblende istypically concentrated near the edges of these clusters but mayoccur throughout as well. Several of the minerals in the two-pyroxene granulites have a regionally restricted distribution.Garnet and fine-grained garnet pseudomorphs occur only ininclusions of two-pyroxene granulites which have been eruptedthrough the Transantarctic Mountains, whereas primaryolivine in inclusions of two-pyroxene granulites is restricted tovolcanics erupted in the Ross Embayment. Primary spinel islargely absent from two-pyroxene granulites of the Ross Em-bayment; however, spinel is abundant in inclusions from BlackIsland in the Ross Embayment. The presence of nonprimaryspinel or olivine in coronas, symplectites, or intergranular glassis not regionally restricted. Hornblende occurs in inclusionsfrom both the Ross Embayment and Transantarctic Mountains,but is much more common in inclusions from the former.Quartz occurs in a few Ross Embayment inclusions but is muchmore common in Transantarctic Mountain inclusions.

Based on the experimental studies of a two-pyroxene-gran -ulite inclusion from Australia by Irving (1974) and the miner-alogy of the compositionally similar inclusions from Antarctica,it is possible to define broadly the pressure-temperature condi-tions under which the inclusions equilibrated. The presence ofgarnet or spinel and the absence of coexisting primary olivineand plagioclase in inclusions from the Transantarctic Moun-tains indicate that these inclusions originated between 5 and 14kilobars pressure (17-45 kilometers depth). If one assumes thatthe volcanics were able to sample all sections of the crust (andthis appears to be a reasonable assumption based on the wide

1986 REVIEW 19

0iA-'-CPX+PL

3A+PL+Px +oP

+cPx+ *'X(±SP)

oL+PL+-cPx+oPx

CID-o10

a)

a-

40

E

30 aW

20

10

variety of inclusions which have been erupted), the absence ofcoexisting primary olivine and plagioclase suggests that thebasic compositions are largely absent above a depth of about16-17 kilometers. In other words, upper crustal granitic andmetasedimentary rocks probably extend down to a depth of aleast 16-17 kilometers. The predominance of coexisting primaryolivine and plagioclase and the rarity of spinel in inclusionsfrom most of the sites in the Ross Embayment indicate thatthese inclusions originated at depths only slightly above 20kilometers at most and probably below that for the most part.

The figure shows the results of more specific thermobarome-try on the lower crustal inclusions using mineral chemistry. Thevarious methods show remarkable agreement, and there is alsogood agreement between these data and the phase equilibria ofIrving (1974) mentioned in the previous paragraph. These dataprovide a thermal profile of the lower crust that is as continuousand precisely documented as any known to us. The geothermshown is intended to be representative of the TransantarcticMountain data only; the Ross Embayment data may actually liealong a slightly higher geotherm. Whether the resultant geo-therm should be considered relevant to the present-day orrecent geothermal gradients or, instead, should pertain to anancient or fossil geotherm is an important question (Harte,Jackson, and Macintyre 1981) and is critical to the conclusionswhich follow below. Ancient deep-seated cooling, as opposedto recent rapid transport to the surface and cooling, should have

. Transantarctic Mts. (Foster Crater)6,0 Garnet Granulites

I-* Spinet GranulitesI Ross Embayment

15 -• Spinet Granulites 50

IIIIIIII

0 50010001500Temperature, °C

Thermobarometry of garnet- and spinel-granulite inclusions fromthe McMurdo Sound region. The garnet-granulite data are based ontwo- pyroxene thermometry (Wells 1977) and garnet-pyroxene-fel-dspar barometry (open circles used Harley 1984; open trianglesused Bohlen, Wall, and Boettcher 1983). The spinel granulite dataare based on Gasparik (1984, 1985) adjusted to two-pyroxene tem-peratures. Phase equilibria are from a two-pyroxene granulite stud-ied by Irving (1974). All Transantarctic samples are from FosterCrater. ("kbar" denotes "kilobar." "Km" denotes "kilometer.")

resulted in compositional zoning in minerals, but in fact, theminerals are largely unzoned. Also indicating that the geothermis a present-day one is the fact that where potassium feldspar ispresent in the granulites, it occurs as sanidine.

Although the geothermal gradient in the lower crust itself isnot extremely high, the data require a very high gradient of60-100°C per kilometer in the upper crust. This gradient shouldresult in heat flow at the surface of at least 3-4 heat-flow units,and Decker and Bucher (1982) measured heat flow in the regionranging up to 3.4 heat- flow units. The high geothermal gra-dient may cast some doubt on the uplift rates calculated for theTransantarctic Mountains based on apatite fission track dating(Gleadow, McKelvey, and Ferguson 1984). The uplift rates of55-135 meters per million years assume a depth of apatite-ageresetting of about 4 kilometers, whereas the preliminary gra-dient determined by us would suggest a resetting depth ofcloser to 2 kilometers. If correct, this might reduce the calcu-lated uplift rate by roughly 50 percent.

The geothermal gradient is one of the highest ever docu-mented (Jones et al. 1983) and is consistent with a virtuallyrequires a continental-rift setting (Wickham and Oxburgh1985). Not only is this important to our understanding of ant-arctic tectonics but the well- documented geothermal gradientwill provide important constraints to geophysicists modelingcontinental rifts anywhere.

This work was supported in part by National Science Founda-tion grant DPP 82-13943.

References

Bohlen, S.R., V.J. Wall, and A.L. Boettcher. 1983. Experimental inves-tigation and application of garnet granulite equilibria. Contributions toMineralogy and Petrology, 83, 52-61.

Decker, E.R., and G.J. Bucher. 1982. Geothermal studies in the RossIsland—Dry Valley region. In C. Craddock (Ed.), Antarctic geoscience.Madison: University of Wisconsin Press.

Gasparik, T. 1984. Two-pyroxene thermobarometry with new experi-mental data in the system CaO-MgO-ALO 2-Si02 . Contributions toMineralogy and Petrology, 87, 87-97.

Gasparik, T. 1985. The effect of Fe 21 on the Al-content of orthopyroxenein equilibrium with spinel and olivine. LOS Transactions of the Ainer-ican Geophysical Union, 66, 413.

Gleadow, A.J.W., B.C. McKelvey, and K.U. Ferguson. 1984. Uplifthistory of the Transantarctic Mountains in the Dry Valleys area,southern Victoria Land, Antarctica, from apatite fission track ages.New Zealand Journal of Geology and Geophysics, 27, 457-464.

Harley, S.L. 1984. The solubility of alumina in orthopyroxene coexistingwith garnet in Fe0-Mg0-Al 201-Si07 and Ca0-Fe0-Mg0-Al201-Si02.Journal of Petrology, 25, 665-696.

Harte, B., P.M. Jackson, and R.M. Macintyre. 1981. Age of mineralequilibria in granulite facies nodules from kimberlites. Nature, 291,147-148.

Irving, A.J. 1974. Geochemical and high pressure experimental studiesof garnet pyroxenite and pyroxene granulite xenoliths from the Dele-gate basaltic pipes, Australia. Journal of Petrology, 15, 1-40.

Jones, A.P., J.V. Smith, J.B. Dawson, and E.C. Hansen. 1983. Meta-morphism, partial melting, and K-metasomatism of garnet- scapolite-kyanite granulite xenoliths from Lashaine, Tanzania. Journal ofGeology, 91, 143-165.

Wells, P.R.A. 1977. Pyroxene thermometry in simple and complexsystems. Contributions to Mineralogy and Petrology, 62, 129-139.

Wickham, S.M., and E.R. Oxburgh. 1985. Continental rifts as a settingfor rcgional metamorphism. Nature, 318, 330-333.

20

ANTARCTIC JOURNAL