Stott, L.D., B.C. McKelvey, D.M. Harwood, and P.N. Webb. 1983. Arevision of the ages of Cenozoic erratics at Mount Discovery andMinna Bluff, McMurdo Sound. Antarctic Journal of the U.S., 18(5),36-38.

Wilckens, 0. 1911. Mollusks of the antarctic Tertiary formation. Wis-senschaftliche Ergebnissen der Schwedische Sudpolarexpedition,3(13),1-62. (In German)

Wilson, G.J. 1967. Some new species of Lower Tertiary dinoflagellatesfrom McMurdo Sound, Antarctica. New Zealand Journal ofBotany, 5(1), 57-83.

Wilson, G.J., and C.D. Clowes. 1982. Arachnodinium, a new dinofla-

gellate genus from the Lower Tertiary of Antarctica. Palynology, 6,97-103.

Zinsmeister, W.J. 1984. Late Eocene bivalves (Mollusca) from the LaMeseta Formation, collected during the 1974-75 jointArgentine-American expedition to Seymour Island, AntarcticPeninsula. Journal of Paleontology, 58(6), 1497-1527.

Zinsmeister, W.J., and H.H. Camacho. 1980. Late Eocene Struthiolari-idae (Mollusca: Gastropoda) from Seymour Island, AntarcticPeninsula, and their significance to the biogeography of early Ter-tiary shallow-water faunas of the Southern Hemisphere. Journal ofPaleontology, 54(1), 1-14.

Bimodal differentiation: Silicic segregationsin the Ferrar Dolerites

MAYA M. WHEELOCK and BRUCE D. MARSH, Department of Earth and Planetary Sciences, Johns Hopkins University,Baltimore, Maryland 21218

The upper portions of mafic sills commonly contain podsor lenses of much more silicic rock, resulting from in situ

differentiation (figure 1). This strong compositional contrast,without a continuum of intermediate compositions, is a clearand pervasive example of bimodal differentiation. Some ofthe best exposures of such bimodally differentiated sills are ofthe Ferrar Dolerites of Victoria Land, which crop out exten-sively in the McMurdo Dry Valley region (Hamilton 1965;Gunn 1966). Observing and documenting the size, distribu-tion, and geometry of the Ferrar silicic segregations was theaim of our January 1993 field season; applying alternativefractionation mechanisms to understand their origin and evo-lution is our present goal. Understanding the details of this

process may go a long way toward explaining bimodal differ-entiation on a planetary scale, for example, the earliest stagesof continental crust formation.

For nearly a century, igneous petrologists have realizedthat serious chemical differentiation probably proceeds by acombination of mechanisms, including crystal settling andfilter pressing. Recent work on the mechanical behavior ofpartially crystalline magma shows that these mechanisms astraditionally perceived cannot alone explain the formation ofsilicic segregations in bimodally differentiated sills. Chemicalmass-balance calculations, for example, show rhyolitic liquidsto be the products of fractionally crystallizing 80 percent of abasaltic magma. Once the solid fraction exceeds about 55 per-

cent, however, magma be-haves rheologically as a solid(Shaw et al. 1968; Marsh 1981);the interstitial liquid is insepa-rable from the crystal mushusing traditional mechanisms.We have collected two com-plete suites of samples fromthe Basement sill at SolitaryRocks and Pearse Valley andhave made transects of theincomplete Peneplain sill at

Figure 1. A sample from the top ofa Si02-rich lens, collected 23 mbelow the upper contact of theBasement sill. Note the sharpcontact between the coarse-grained, lighter colored silicic lens(61 weight percent Si02) and theoverlying gray, medium-grainedgabbro (53 weight percent Si02).This particular lens is 12 m longand 15 to 30 centimeters wide(vertical thickness).

ANTARCTIC JOURNAL - REVIEW 199319

In

As

llp

PEGMATOI 0ZONE

BASEMENT SILL

Pandora Spire and the partially covered Finger Mountain sillat Maya Mountain. We report here preliminary results fromour observations and analysis of the Basement sill rocks.

Whole-rock chemical analyses (XRF) of selected samples,along with a schematic cross-section of the Basement sill, areshown in figure 2. In the pegmatoid zone (Gunn 1966), wefind the horizontal pegmatoid segregations to begin about 12meters (m) beneath the upper contact and increase dramati-cally downward in thickness to a maximum at 30 in theupper contact. At this level, typical segregations are 30-45 inin length and 0.2-1.5 in the undulating, subhorizontallayers interdigitate to form a horizon of silicic rock that isnearly continuous throughout the sill. They decrease in sizeand frequency down to 70 in the upper contact, wherethey vanish altogether. The chemical profiles (figure 2) of thisregion are marked by spikes of high silicon and iron and lowcalcium and magnesium. A thick zone of orthopyroxene-richrock (norite zone of Gunn 1966) occurs in the center of thesill, marked by low calcium and strontium and excess chromi-um and magnesium (figure 2). Orthopyroxene crystals up to 7millimeters (mm) long constitute up to 50-60 percent of thisrock, whereas orthopyroxene itself is sparse above and belowthis zone. This norite horizon apparently represents the lastportion of the injection sequence of a flow-differentiated,originally phenocryst-rich magma, rather than the product ofin situ differentiation.

Silicic segregations in the Basement sill contain from 58to 61 weight percent of silica (Si0 2). To test the idea that these5i02 -rich interstitial liquidswere extracted during crystal-lization to form the siicic seg-regations (Wheelock andMarsh 1993), we have per-250m

formed computer simulationsusing SILMIN (Ghiorso 1985,now known as MELTS).SILMIN is a remarkable com-puter program that simulatescrystallization of silicate melts.Under specified conditions ofbulk chemical composition,pressure, temperature, andoxygen fugacity, the programutilizes thermodynamic data-bases and an iterative opti-mization procedure to mini-mize the Gibbs free energy of amagmatic system and todetermine the prevailing sta-ble mineral-melt assemblage.

Figure 2. A schematic cross-section of rock types in theBasement sill is shown at left.Chemical profiles (abundance vs.height above base in the sill)illustrate their compositionalrange.

Beginning with a "local" bulk composition (for the upperthird of the sill, the norite zone being a late arrival), we ranSILMIN until the Si0 2 content of the calculated interstitial liq-uid matched that of the observed silicic segregations (58-61weight percent). The crystallinity (volume fraction of solid)was then calculated from the output. Crystallinity vs. Si02content in figure 3 shows that the magma attains a crystallini-ty of 70 to 85 percent at the time of late-stage liquid extractionto form the silicic lenses.

Once the solid fraction of the magma exceeds about 55percent, the crystals form a rigid, almost brittle, network (onthe time scale of sill solidification); beyond this critical crys-tallinity, the interstitial liquid increases strongly in SO 2 con-tent. The interdigitating subhorizontal segregations display ageometry reminiscent of fracture sets. Coupling this funda-mental obsrvation with the results of the SILMIN simulationsuggests a mechanism involving tearing or fracturing of thecrystalline network. One such mechanism, termed solidifica-tion front instability (Marsh 1991) occurs with progressivedownward solidification from the roof of the sill. A thickeningzone of cool, dense crystal mush overlying the hot, lower-density magma becomes gravitationally unstable and sags orpartially tears away from the roof. Tearing in a region contain-ing between 15 and 30 percent liquid draws interstitial liquid,essentially identical to that observed, into the resulting gash.A theoretical model is being developed to further investigatethe plausibility of this mechanism.

FeO+ F%03S110MgO

' I

240. .9 S ••SS S S•

200

160 .

120a S

80

40

0

581114505560650 4 8 12 16 20[Q3 (*1%) (wt%)

CaO

Sr C

240S#5

200

160

120•

80

40

0 .

58111450100150

400800

(wt%) ppm ppm

ANTARCTIC JOURNAL - REVIEW 1993

20

-D

Cl)()

0C,)

65

60

55

5000.20.40.60.81

Basement sillResults from SILMIN

Crystallinity

Figure 3. Crystallization of the Basement sill was simulated usingSILMIN (Ghiorso 1985). Results show that observed silicicsegregations with 58-61 percent Si02 are similar to calculatedresidual liquids when the cooling magma is 70-85 percent solid.

This research is supported by National Science Founda-tion grant OPP 91-17576. We thank Bernard Gunn for suggest-ed locations for field study, and K.A. McCormick for tirelessassistance in the field.

References

Ghiorso, M.S. 1985. Chemical mass transfer in magmatic processes, I.Thermodynamic relations and numerical algorithms. Contribu-tions to Mineralogy and Petrology, 90(2/3), 107-120.

Gunn, B.M. 1966. Modal and element variation in antarctic tholeiites.Geochimica et CosmochimicaActa, 30(9), 881-920.

Hamilton, W. 1965. Diabase sheets of the Taylor Glacier region, Victo-ria Land, Antarctica (U.S. Geological Survey professional paper456-B). Washington, D.C.: U.S. Government Printing Office.

Marsh, B.D. 1981. On the crystallinity, probability of occurrence andrheology of lava and magma. Contributions to Mineralogy andPetrology, 78(1), 85-98.

Marsh, B.D. 1991. Solidification front instability (SF1) and silicic chaosin basaltic magma chambers. Geological Society of AmericaAbstracts with Program, 23(5), A270. [Abstract]

Shaw, H.R., T.L. Wright, D.L. Peck, and R. Okamura. 1968. The viscosi-ty of basaltic magma: An analysis of field measurement inMakaopuhi Lava Lake, Hawaii. American Journal of Science,266(4), 225-264.

Wheelock, M.M. and B.D. Marsh. 1993. Silicic segregations in basalticsills: Bimodal differentiation. EOS, Transactions of the AmericanGeophysical Union, 74(16), 336-337. [Abstract]

Paleomagnetic and geochronologic studies of igneous rocksfrom southern Victoria Land

ANNE GRUNOW, Byrd Polar Research Center, Ohio State University, Columbus, Ohio 43210JOHN ENCARNACION, Department of Earth Sciences, University of Michigan, Ann Arbor, Michigan 48109

During the 1992 field season, we undertook a paleomag-netic and geochronologic sampling program in the base-

ment intrusions of southern Victoria Land (figure 1). The aimof this research is to obtain new late Proterozoic(?) and earlyPaleozoic paleomagnetic reference poles that will establish anapparent polar wander path for East Antarctica and Gond-wana. Paleomagnetic data, along with precise radiometric ageconstraints, will help determine Antarctica's early Paleozoicpaleogeographic position. The radiometric dating programwill better establish the age of the early Paleozoic igneousactivity and provide further constraints on the tectonic settingof this area during and after the Ross deformational event.

The field party consisted of Grunow collecting rock coresfor paleomagnetic study, Encarnacion collecting for uranium-lead (U-Pb) and argon-40/argon-39 (40Ar/ 39Ar) geochronolo-gy, and Charles Kroger, working as a mountaineering guide.The field program was designed to sample Cambrian, Ordovi-cian, and possibly Proterozoic intrusions rocks over a largearea. We visited 13 major locations in southern Victoria Land(figures 1 and 2) over the 2-month field season. The locations

visited were the Brown Hills (figure 1), Briggs Hill along theFerrar Glacier, the head of the Byrd Glacier (figure 1), theCatspaw Glacier, Mount Falconer, Granite Harbor, KillerRidge, Koettlitz Glacier (Miers Ridge and southern WalcottGlacier), Lake Vanda, Lake Vida, Mount Loke, the SkeltonGlacier (Cocks Block and Bareface Bluff), and Sperm Bluff(figure 2).

Helicopters from McMurdo were used in mid-Novemberto establish a Ski-doo-based camp at Cape Geology for 1 weekof fieldwork around Granite Harbor. Ski-doo travel across thesea ice was still feasible at this time of the season, and wewere able to sample many locations between Cape Archer,Cape Roberts, and the Flatiron (figure 2). In late November,we established a Ski-doo-based camp in the Brown Hills(along the south side of Cooper Nunatak) (figure 3) using aTwin Otter for the camp put-in. Extensive blue ice cover inthe Brown Hills made Ski-doo travel difficult (consuming sig-nificantly more mogas than anticipated), and so we were lim-ited to sampling along the northern side of the Brown Hills.We returned to localities in the dry valleys and the Royal Soci-

ANTARCTIC JOURNAL - REVIEW 1993

21