that the plagioclase point (figure 3) is a little toohigh with respect to strontium-87 and the K-feld-spar is a little too low, thereby: (1) lowering thecalculated age, (2) raising the initial strontium-87 /

strontium-86 ratio, and (3) increasing the apparenterror beyond the limits of analytical precision. Wetherefore conclude that the age of 486 million yearsshould be interpreted as a minimum age.

The other analyzed samples are from paragneissthat contains large amounts of biotite, twinnedand untwinned plagioclase, and small amounts ofquartz and hornblende. Abundant clay or sericiteand minor chlorite occur as alteration products.Rubidium-strontium data for the paragneiss arepresented in table 2, and the isochronal relation-ship is shown in figure 4. As seen in figure 4, theage of 242 ±31 million years is a biotite age. Likethe quartz monzonite, the paragneiss has a highinitial ratio of 0.7082 ±0.0027, but a much largeruncertainty. Unlike the quartz monionite, the para-gnciss yields an apparent age that is less than halfof that reported for similar rocks from VictoriaLand (for example see Goldich et al., 1958; Angino(,--i al., 1962; McDougall and Ghent, 1970). It is un-likely that the age of 242 million years dates anyreal geologic event, but rather reflects a partial lossof strontium-87 from the biotite. This may haveoccurred during the intrusion of the Ferrardolerites or (luring the movement of hydrothermalsolutions or groundwater along a near-verticalfault that cuts the core at a depth of approximately38 meters. We conclude that the true age of theparagneiss is greater than 50() million years.

This research was supported by National ScienceFoundation grant c;v-3695 1.

References

Angino, F. Fl., M. D. Turner, and D. J. feller. 1962. Reon-naissance geology of lower 1 avfor Valley. Victoria Land, Ant-arctica. Geological Societ y of America Bulletin, 73: 1553-1562.

Barrett, P. j.. U. W. Grindley, and P. N. Webb. 1972. TheBeacon Supergroup of East Antarctica. In: Antarcti( Geologyand Gphs.nc.s (.\cfie, R. J ., editor). Oslo., Universitetsforlagtt.

319-332.Brooks, C. 1968. Relationship bet ween feldspar alteration and

the precise post-crystallization ilioveiTletit of itibidi urn andstrontium isotopes in a granite. /ournal of (;eol)h y.u( al Research, 73: 4751-4757.

loinpston. W .. and J. F. 1a)vermg. I 969. The strontiuni so-topn geochemistry of gra mime and eclogitic inclusions fromthe basic pipes at Delegate, eastern Australia. Gea/uinua 0

Coc,noch,mua 411(1. 33: 671-699.Deutsch, S., and P. N. Webb. 1961. Sr/Rb dating on basement

rocks from Vu. toria Land: eviclemice for a 1000 million yearold event. lii: A utarct/( (ealugy, (Adie. R. 1, editor). New York,Wiley. 557-569.

rnh . r/T)tctrnher 1975

Flanagan, F.]. 1973. 1972 values for international geochemicalreference samples. Geochimica et Cosnwchzinica Ada, 37: 1189-

1200.Goldich, S. S., A. 0. Nier, and A. L. Washburn. 1958. A40/K4°

age of gneiss from McMurdo Sound, Antarctica. Washing-ton, D.C., American Geophysical Union. Transactions, 39:

956-958.Gunn, B., and G. Warren. 1962. Geology of Victoria Land be-

tween the Mawson and Mullock glaciers, Antarctica. N.Z.Geological Survey Bulletin, 71. 157p.

Hamilton, W. B. 1965. 1)iabase sheets of the Taylor Glacierregion, Victoria Land, Antarctica. Washington, D.C. U. S.Geological Survey. Professional paper, 456-B. 71 p.

Harrington. H. J. 1958. Nomencla jure of rock units in the RossSea region, Antarctica. Nature, 182: 290.

Kurasawa, H., Y. Yoshida, and M. G. Mudrey, Jr. 1974. Geo-logical log of the Lake Vida core-DVDP 6. In: Thy ValleyDrilling Project Bulletin 3. DeKaIb, Northern Illinois Univer-sity. 94.

McDougall, 1. 1963. Potassium-argon age measurements ondolerites from Antarctica and South Africa. Journal of Geo-physical Research, 68: 1535-1545.

McDougall, I., and E. D. Ghent. 1970. Potassium-argon dateson minerals from the Mt. Falconer area, lower Taylor Valley,south Victoria Land, Antarctica. N.Z. Journal of Geology andGeophysics, 13: 1026-1029.

McIntyre, U. A., C. Brooks, W. Compston, and A. Turek. 1966.The statistical assessment of Rh-Sr icohrons. Journal of Geo-physical Research, 71: 5459-5468.

McKelvey, B. C., and P. N. Webb. 1962. Geological investiga-tions in southern Victoria Land, Antarctica; part 3, geologyof Wright Valley. N.Z. Journal of Geology and Geophysics. 5:

143-162.Smithson, S. B., P. R. Fikkan, and D. J. Toogood. 1970. Early

geological events in the ice-free valleys, Antarctica. Geological

Sociee.i oJ America Bulletin, 81: 207-210.Stuckless. J S., and J . R. O'Neil. 1973. Petrogenesis of the

Superstition-Superior volcanic area as inferred from stron-tium- and oxygen-isotope studies. Geological Society of AineruaBulletin, 84: 1987-1998.

Webb, P. N., and B. C. McKelvey. 1959. Geological investiga-ions in south Victoria Land, Antarctica; part 1, geology of

Victoria Dry Valley. N.Z. Journal of Geology and Geophysics, 2:

120-136.

Structure and petrology of theScotia Arc and the PatagonianAndes: R/V Hero cruise 75-4

W. D. DAI.zIEL, MAARTFIN J . or WIT, andW. IAN R1DI.FlY

Lamont-Doherty Geologi (a! ObservatoryColumbia University

Palisades, New York 10964

The purpose of RIV Hero cruise 75-4 was to studythe geologic structure and history of the region be-tween Punta Arenas and Puerto Montt, Chile, as

307

5C

ndex map of southernSouth America.

part of a continuing U.S. effort to study the tec-tonic history of the Antarctic Peninsula, the sub-antarctic islands of the Scotia Arc, and the south-ernmost part of the Andean Cordillera. We sailedfrom Punta Arenas on 14 june 1975 and arrived atPuerto Monti on 14 July 1975,

The specific objective of the cruise was to relatethe geology of Fierra del Fuego and of the ChileanAndes, the Scotia Arc, and the Antarctic Peninsulasouth of:' 00S. (all have been extensively studied by1)r. I)alziel and field associates on previous cruisessince the 1968-1969 field season) to the geology ofthe Central Andes. The latter, of course, are rela-tively more accessible and have been studied byChilean and Argentine geologists for many years.The geology of the Central Andes is in many ways

more like that of the Antarctic Peninsula than thatof the southernmost Andes. This has importanteconomic implications, since major copper depositsare known to exist in the Central but not in theSouthern Andes.

Randall Forsythe, Lamont-Doherty GeologicalObservatory of Columbia University, Sergio Ces-pedes, Empresa Nacional del Petróleo, Jorge Skar-rneta and Hans Niemeyer, lnstitut() de investiga-clones Geologicas, and Estanislao Godo y , Univer-sity of Chile, Santiago, were also interested in theregion and participated in cruise 75-4, thus con-tinning a cooperative program begun in 1969 withRJ' Hero cruise 69-6. William Ta y lor, Universityof Liverpool, England, was also aboard particularlyto study the Patagonian batholith.

308tM'T'A D('m1i TI TT

Our primary goal was achieved. It was possibleto recognize that the so-called basement (i.e., preto Late Jurassic) metamorphic complex of the Ant-arctic Peninsula, the South Shetland Islands, andthe South Orkney Islands continues north into theCentral Andes at least as far as 4405 Diagnosticred cherts and associated rocks on both sides ofthe batholith confirmed the identical nature of theSouth American basement in these tectonic posi-tions. The Patagonian batholith was thus emplacedwithin, and is entirely surrounded by, SouthAmerican continental basement as indicated byDalziel et al. (1974). Suggestions based on airphotographs of basement rocks on Isla Desolación,which lies immediately south of the west end of theStrait of Magellan (see map), were confirmed.These basement rocks and those forming islandsfarther north (around Isla Madre de Dios) containmafic pillow lavas underlying white, green, and red(probably manganiferous) chert deposits. They aretherefore comparable to the basement rocks of theSouth Orkney and South Shetland islands pre-viously studied in this project (see summary byDalziel, 1975). We interpret them to representPaleozoic oceanic deposits deformed in the arc-trench gap of an early Mesozoic calc-alkaline vol-canic chain that lay to the east (Dalziel and de Wit,in preparation).

It was also possible to explain the essential struc-tural difference between the Southern and CentralAndes and conversely the similarity between thestructural style of the Antarctic Peninsula and theNorthern Andes. The ophiolitic rocks representingthe floor ofaJapan Sea-like marginal basin of EarlyCretaceous age in the southernmost Andes (Dalzielet al., 1974) clearly pinch out northward at about51°S. The deformation involving basement reacti-vation in southernmost Andes is less intense andpenetrative north of about 4905• This strengthensthe suggestion by Dalzicl et al. (1974) that the de-formation occurred when the marginal basin"closed" and was uplifted in the mid-Cretaceous,and that large penetrative strains at high crustallevels behind a volcanic arc are not generated bysubduction alone, but may he related to a "colli-sion" between volcanic arcs and continental fore-land during the closing of marginal basins.

Slightly deformed Tertiary sediments on TresMontes Peninsula contain vast amounts of coarseandesitic-basic volcanic detritus presumably de-rived from the volcanic equivalents of the batho-lith to the east.

Considerable rock areas previously mapped asbasement were found to consist of the graniticbatholith. Fjords extending east of Golfo de Penaswere investigated for the first time and were foundto be cut wholly in granitic rocks.

East-west traverses across the batholith indicate

.h-/Iernher 1975

tonalite as the major plutonic rock type; it is as-sociated with lesser amounts of adamellite, minordiorite, and very rare gabbro. Textures are ex-tremely variable: equigranular, plagioclase-phYflcand strongly fabricated varieties were all observed.The main mineralogic variations involve absoluteabundances of ferromagnesiafl minerals and pres-ence or absence of hornblende and biotite.

Associated with the batholith are several periodsof minor intrusions, commonly as thin, highly in-clined dikes. Synpiutonic diabase dikes were verycommon, displaying various stages of disruptionand assimilation. Undoubtedly some shadowy in-clusions observed in some tonalite bodies are partlyresorbed synplutonic dikes, whereas elsewhere,especially where highly concentrated, the inclu-sions probably represent roof pendants. Laterdikes, striking rather randomly between north andeast, are dominantly diabase; local concentrationsof quartz-felsite dikes, however, are also observed.These dikes postdate irregular veins and stringersof aplite and pegmatite that presumably repre-sented the late-stage mobile components of thetonalitic magma.

Relationships between the various plutonic rocksremain poorly determined. Wherever observed,tonalite veins earlier diorite, and probably therare gabbros predate the diorite.

In summary, the batholith appears remarkablyuniform in composition, with an estimated averagecomposition of granodiorite. North of Golfo dePenas, exposure is uniformly poor; future attemptsat detailed study would best be focused on moresoutherly parts. There is a remarkable lack of con-tact metamorphism associated with the batholith,which may be attributed to the present deep levelof exposure and the relatively dry nature of themagma. No clear field evidence exists for a west-east change in rock type or rock abundances.This is probably present, but involves subtle varia-tions in K-feldspar abundance that cannot be easilyobserved in the field, particularly during a recon-naissance.

We are grateful to our pilot, Lieutenant Com-mander M. Alejandro Sepulveda Mattus of theChilean army, Captain Norman Deniston, masterof Hero, and to the Hero crew for their willingnessto operate the vessel at night to help us accomplisha full program. This research was supported byNational Science Foundation grants oPT' 74-21415

and DES 75-04076. The Empresa Nacional delPetrOleo, through its general manager Generaldon Orlando Orbina Herrera, through its direc-tors, and through Edwardo Gonzalez P. (Jefe deGeologia), made a generous donation of fuel forthe cruise.

09

References

Dalziel, I. W. D. 1975. The Scotia Arc tectonics project, 1969-1975. Antarctic Journal of the U.S., X(3): 70-81.

Daiziel, 1. W. D., and M. J . de Wit. In preparation. The evolu-tion of the Scotia Arc and its bearing on the reconstructionof SW Gondwanaland

Dalziel, I. W. D., M. J . de Wit, and K. F. Palmer. 1974. A fossilmarginal basin in the southernmost Andes. Nature, 250: 291-294.

Aerosol observations overthe ice caps

A. W. HOGANAtmospheric Sciences Research Center

State University of New York at AlbanyScotia, New York 12302

D. NELSONGlobal Monitoring for Climatic Change

National Oceanic and Atmospheric AdministrationBoulder, Colorado 80302

Greenland and Antarctica present unique casesfor aerosol studies. While these are "continental"situations, the absence of vegetation and fossil fuelcombustion, and the presence of continuous icecover, result in an absence of surface sources. Theonly source of particulate matter over the ice capsappears to be meteorological transport from othermaritime or continental regions, and chemical orphotochemical reaction of gases to form particles.

Fenn and Weickmann (1958) studied aerosolsover Greenland and in many instances found thatthe concentration was below the detection thres-hold of a Pollak counter. Pollak and Metnieks(1959) changed the illumination of the counter toprovide a converging light beam, and were able torecalibrate the instrument to respond to aerosollevels that approach particle-free gas.

Megaw (1973) and Flyger et al. (1973) recentlyrenewed interest in the study of Greenland aero-sols by hypothesizing that the "background" or"end point" of Northern Hemisphere aerosol oc-curs over the Greenland Ice Cap. I accompaniedone of these expeditions (Hogan et al., 1975) andmade aerosol observations at sea level and atop theice cap near the center of the land mass.

The aerosol at sea level was similar to maritimeaerosol, but a small diurnal variation was detected.

310A xim A 1,'rnT

The concentrations observed on the vegetatedsouthern tip of Greenland were similar to ruralaerosols in North America and Africa. These ob-servations are included only for interest 'and arenot typical of ice-cap observations.

The aerosol measured at Dye II, atop the Green-land Ice Cap at an elevation of over 3,000 meters,was generally of low concentration. The range ofconcentration was from 100 to 250 particles percubic centimeter on clear days. At the station,slightly higher readings were generally observedfrom a platform 20 meters above the ice, althoughthere is a possibility of local contamination. Thehighest values occurred in subsiding air with a coldfront west of the station.

Extremely low values accompanied and followeda period of riming, light snow (graupel), and pre-cipitating fog. Flow at this time was from the DavisStrait. The concentration fell to 30 particles percubic centimeter or less during this period, andslowly returned to about 100 particles per cubiccentimeter in 36 hours.

The low concentration observed is similar to thatfound near maritime cumulus. It would appearthat the fog and liquid water cloud that accom-panied this storm system were efficient aerosolcollectors, thereby reducing the concentration ofaerosol to a value approaching the number ofliquid cloud drops. The aerosol concentration thenremained at a low value because only gas reactionswere available as an aerosol source in the new high-pressure system building over Greenland.

Voskresenskii (1968) first reported on antarcticaerosols. He found that typical maritime concen-trations existed at the coast with northerly winds.The concentrations accompanying strong katabaticwinds from the continent ranged from 60 to 400particles per cubic centimeter. Hogan (1975) foundconcentrations of more than 500 particles per cubiccentimeter in a strong katabatic wind at Siple Sta-tion, under conditions of subsidence over the PolarPlateau.

Experiments at Amundsen-Scott South Pole Sta-tion detected concentrations of 50 to 125 particlesper cubic centimeter at the surface during the aus-tral summer. During subsidence conditions theconcentration rises considerably, sometimes to ashigh as 1,500 particles per cubic centimeter asshown in figure 1. These additional particles arevery small and are near the threshold of detec-tion. These particles apparently form somewhereabove the strong surface inversion usually presentover the Polar Plateau; they cannot be transportedin because coagulation would greatly decrease thenumber concentration of particles of this size with-in a few hours. These particles come down with thesubsiding air and are transported along the surfaceby katabatic winds. This is in agreement since the