JOURNALOF GEOPHYSICALRESEARCH, VOL. 96, NO. Bll, PAGES17,935-17,954, OCTOBER 10, 1991

New PaleomagneticData FromThurston Island' Implications for the Tectonicsof West Antarcticaand Opening

A.M. GRUNOW1 ByrdPolar Research Center, Ohio State University, Columbus

Lamont-DohertyGeological Observatory andthe Department ofGeological Sciences, , New York

I. W. D. D ALZlEL Institutefor ,University of Texasat Austin

Palcomagneticdata from threeWest Antarcticcrustal blocks ( (AP), ThurstonIsland- (TI), and the Ellsworth-Whitmore Mountains (EWM) indicate that there has been motion between the individualblocks and motion relative to EastAntarctica during the Mesozoic. A Triassicpalcomagnetic pole from theTI block(116øE, 61øS, A95 = 19.4ø,N = 3 VGPs)appears toindicate that the block has rotated -90 ørelative to EastAntarctica between 230 Ma and110 Ma. Ourpreviously reported Middle Jurassic palcomagnetic pole from the EWM block indicates that a 90 ø rotation relative to East occurred sometime between the Cambrian and 175 Ma. We believe that the 90 ø counterclockwise EWM rotation occurred between-220 Ma and 175 Ma related to thedevelopment of post-GondwanideOrogeny shear zones. The motion of theAP, TI, andEWM blocksappears to be linked duringthe mid- to late Mesozoicto threemajor events in the evolutionof the southemocean basins. Openingin theMozambique-Somali-Weddell Sea basins may have produced major counterclockwise rotation of the TI blockwith respectto EastAntarctica between the Jurassic and Early Cretaceousbased on new LateJurassic (145øE,64.5øS, A95 = 7ø,N= 5 VGPs)poles. We believe that the TI rotation,as well as deformation inthe southem AP block, wascaused by collisionand shearingof the EWM blockagainst the othertwo asthe EWM block moved southwardwith . An Early Cretaceous palcomagnetic pole (232øE, 49øS, A95 = 7.9ø,N= 5 VGPs) from the TI blockrequires that between the Early andmid- Cretaceousthere was clockwise rotation, with respect to EastAntarctica, of theAP-TI-EWM blocks(an entity we callWeddellia). A changein theopening history of the WeddellSea basin caused by initiationof spreadingin theSouth Atlantic ocean basin at-130 Ma probablystarted Weddellia'sclockwise rotation. Two new - 110and -90 Mapoles from the TI block(210øE, 73 øS, A95 = 7.6ø,N= 7 VGPsand 161øE, 81øS,A95 = 3.9ø,N = 18VGPs, respectively) aresimilartoequivalent agepoles from the APblock andEast Antarctica and indicate that Weddellia was at ornearits present-day position with respect to East Antarctica by -110 Ma. This correspondsto a time of majorplate reorganizaion in the SouthAtlantic and southeast Indian Oceans.Based on both the new TI palcomagneticdata and previously reported data from Marie ByrdLand (MBL), dextralshearing would be expected to haveoccurred between MBL andWeddellia since the mid-Cretaceous. Pine IslandBay, the areabetween the TI andMBL blocks,marks a fundamentaland complex tectonic boundary in West Antarcticathat we proposehas largely been a zoneof transcurrentsheafing.

]_NTRODUCTION GEOLOGIC OVERVIEW OF

The positionof West Antarcticawith respectto cratonicEast The AP and TI blocks are composedprimarily of Mesozoic Antarctica prior to and after breakup of Gondwanaland is and Cenozoic arc intrusive and extrusive rocks associated with critical to the opening history of the Weddell Sea. West the Pacific convergent margin of Gondwanaland [Srnellie, Antarctica,however, is not a singletectonic entity but rather is 1981; Thornson and Pankhurst, 1983; Pankhurst, 1990]. Early composedof four major crustalblocks whose relation to each Paleozoic basementrepresenting the roots of a magmatic arc other and to East Antarctica is not well understood. These indicate that parts of the AP block have been along a crustal blocks or terranes are the Antarctic Peninsula (AP), the convergent margin since the Paleozoic [Pankhu.rst, 1982, Ellsworth-Whitmore Mountains (EWM), the - 1983; Harrison and Loske, 1988; Milne and Millar, 1991]. The EightsCoast (TI), andMarie ByrdLand (MBL) [Jankowskiand EWM block consistsprimarily of a Precambrian basement at Drewry, 1981;Dalziel and Elliot, 1982] (Figure 1). The focus Haag Nunataks [Millar and Pankhurst, 1987] and a Paleozoic of this paperis on new palcomagneticdata from the TI block sedimentarysuccession in the Ellsworth Mountains [Webers combined with existing palcomagnetic, geologic, and and Sporli, 1983] that has often been correlated with the geophysicaldata from Gondwanalandthat links the movement stratigraphic succession in the Pensacola Mountains, East of the West Antarctic crustal blocks to the Mesozoic evolution Antarctica (Figure 1) [Schopf, 1969; Dalziel and Elliot, 1982]. of the basins. MBL consistsof Paleozoic sedimentary and igneous rocks that were intrudedbY mid-Cretaceousgranitoids; it is mostclosely XNowat Departmentof Earth Sciences, University of Oxford,Oxford, associatedwith the Paleozoicrocks of North Victoria Land England. (EastAntarctica) and with New Zealand[Bradshaw et al., 1983]. Copyright1991bythe American Geophysical Union. Variouspositions havebeen suggested forthe West Antarctic Papernumber 91JB01507 crustalblocks ina prebreakup Gondwanaland reconstruction 0148-0227/91/91JB-01507505.00 [Schopf,'1969; Smith and Hallam, 1970; De Wit, 1977;Barron

17,935 17,936 GRUNOWET AL.: NEW PALEOMAGNETICDATA FROM THURSTONISLAND et al., 1978; Norton and $clater, 1979; Dalziel and Elliot, EWM-TI blocks may have acted a tectonic entity which we 1982; Powell et al., 1982; LaBrecque, 1985; Elliot, 1991]. refer to as Weddellia, between the Middle Jurassic and Norton and Sclater [1979] showed that the du Toit [1937] and Cretaceous, insofar as can be resolved paleomagnetically Smith and Hallam [ 1970] fit of the major continentalfragments [Grunow et al., 1987b]. If the AP, EWM, and TI blocks are of Gondwanalandwas essentiallycorrect but that an joined in their present-dayrelative positions(rigid Weddellia), unacceptableoverlap of the AP block with the FalklandPlateau there is an overlap of the AP block with southern South occurs'if West Antarcticais left in its present-dayposition America. Geosat, marine, and airborne gravity data [Sandwell with respectto East Antarctica.Together with the geographic and McAdoo, 1988; Bell et al., 1990] indicate that the isolation and anomalous structural trend of the Ellsworth continentalmargin of the Antarctic Peninsulais much narrower Mountains fold belt, this overlap constituted the primary than necessarily inferred by the imprecise location of the r•asonto/:onsider relative motion among the crustal blocks of shelf-slope break in ice-covered waters, and this lessens the West Antarctica [Schopf, 1969; Dalziel and Elliot, 1982]. pre-breakup overlap shown for the AP block with southern Palcomagneticdata supporta 90ø counterclockwiserotation of South America [Grunow et al., 1987b]. A reconstruction with the EWM block from the Transantarctic margin of East the blocks separated(mosaic Weddellia) fits the geologic and Antarctica(probably near CoatsLand; Figure 1) sincethe Late paleomagneticdata better but requiresa more complex opening cambrian[Watts and Brarnall, 1981; Grunow et al., 1987a]. history for the Weddell Sea basin [Grunow et al., 1987b; Palcomagnetic data obtained from Middle Jurassicrocks in Lawyer and Scotese, 1987; Lawyer et al., 1991]. the EWM block yielded a palcomagneticpole very similar to a Middle Jurassic pole from the AP block [Longshaw and GEOLOGIC SETrING OF TI BLOCK Griffiths, 1983; Grunow et al., 1987a]. An Early CretaceousTI pole [Grunow et al., 1987b] was indistinguishable from the The TI block consists of the area known as Thurston Island Middle JurassicAP and EWM poles and was interpretedby us as and the Eights Coast between ~255øE (Pine Island Bay) and possibly indicating little apparent polar wander for these ~270øE and from 71øS to ~ 75ø S. ThurstonIsland is separated blocks between the Middle Jurassic and Early Cretaceous from the Eights Coast by the , which is found [Grunow et al., 1987b]. This led us to suggestthat the AP- within a graben or half graben structure(Figure 1). The rocks

Paleomagneticlocalities

Otherlocalities mentioned Antarctic Peninsula MaudLand Weddell Land

THURSTON ISLAND Mtns. Thurston_.,., Island• EASTANTARCTIC •,• 1s.,• • CRATON ' Bay MarieByrd • O'o4.,.4•oLand Sea

C ' Sherman•...... '" ':•'•••'••'"'"'! .... 72o30'S--

B¸T ICE SHELF

N LN 0 50kmEIGHTS COAST 100øWI 96oWI '•• 92oWI 73ø30'S Fig. 1. Samplelocations on ThurstonIsland: BN, BelknapNunatak; DI, DustinIsland (ek, EhlersKnob); HG, Hale ,HN, HarrisonNunatak; JM, JonesMountains; LN, LepleyNunatak; LP, LandfallPeak; MB, MountaBtamhall; MB, MountDowling; MI, McNamaraIsland (pb, PelletsBluff); MS, Mount Simpson;PP, ParkerPeak; SH, SheltonHead. Otherlocations: HK, HendersonKnob; MN, MountNoxom. Insetshows the WestAntarctic cmstal blocks (boldly outlined) in theirpresent-day positionswith respectto EastAntarctica. EWM, Ellsworth-WhitmoreMountains; EM, EllsworthMountains; NVL, North Victoria Land;asterisk, Haag Nunataks. Stippled lines show mountain trends; slanted lines show location of ice shelves. GRUNOWET AL.' NEWPALEOMAGNETIC DATA FROM THURSTON ISLAND 17,937 on Thurston Island crop out along two general east-west The Mesozoic rocks in the JonesMountains are unconformably strikingzones, parallel to theAbbot Ice Shelf.The majority of overlain by glacial till and flat-lying Tertiary alkaline rockexposures from the TI blockare of intrusiverocks with volcanics[Craddock et al., 1969;Rutford et al., 1972]. The minoroutcrops of volcanicrocks that are thoughtto represent thermalevent that producedthe Tertiaryalkalic basalts may evolutionalong a convergentmargin during the Mesozoic.A havealso created the grabenlikestructure in whichthe Abbot morecomplete description of therocks is givenby Craddocket Ice Shelf is found (Figure 1). al. [1964a, 1969],Lopatin and Orlenko[1972], Wade and SAMPLING AND LABORATORY PROCEDURES Wilbanks [1972], and Storeyet al. [1991]. The oldestrocks known from TI are Carboniferousgneisses thought to represent The paleomagneticsamples (along with geochronologic, theroots of a magmaticarc [Craddock et al., 1964b;Pankhurst, geochemical,and structuralstudies collected by other members 1990]. Triassicgranodiorites (-230 Ma) at Mount Btamhall of the project) were collectedas part of the joint U.S.-U.K. [Grunowet al., 1990] andLower Jurassic granites (-200 Ma) in West Antarctic tectonics project [see Dalziel and Pankhurst, the JonesMountains [Craddock et al., 1969; Pankhurst, 1990] 1987] in the hope of better constrainingthe relative motion of indicatemagmatic activity on ThurstonIsland in the early the West Antarctic blocks and the opening history of the Mesozoic. A seriesof-140 to 155 Ma granites,granodiorites Weddell Sea. Paleomagnetic samples in the TI block were and dioritesrepresent another phase of magmaticactivity on collectedby A.M. Grunowusing a portable,gasoline-powered, Thurston Island [Storey et al., in press]. Mid- to Late diamond-bit coring drill with glycol as a drill coolant. Over Cretaceous volcanism and felsic dike intrusion occurred in the 700 oriented drill samplesand 10 oriented hand sampleswere [Craddock et al., 1963a], while granitic collected from 123 sites at 28 locations in the TI block. magmaswere being emplaced at LepleyNunatak [Grunow et al., Generally six to seven cores were sampled at each site and 1990]. These magmaticrocks indicate that TI was along a oriented using a Brunton compass.Sun compassmeasurements convergentmargin until the latestCretaceous-earliest Tertiary were made at each locality to allow correction for local when subductionprobably ceasedas a result of eastward magneticvariation. The Sun compassmeasurements from the propogationof thePacific-Antarctic ridge [Watts et al., 1988]. intrusiveand volcaniclocalities (Table 1) were alwayswithin

a. LepleyNunatak b. JonesMountains W,Up NRM W,Up W,Up [ /5.265mT •'UP?NRMNRM• 5mT I• [RM [ I •40.0 4I DirectionsX• FelsiteDike'S9Ma ItI• 15• MancDike [•20'0 a=Mafic dikes I • 60.0 N:6 ß • 30.0 0 12'ø I 3½o

N N I •%-1 • 10e-lA• J1C4A • l•-lA/m J2F5A I I l•-2•mJ2Fll

c. McNamara Island d. 0OE •)øE OOE W,Up W,Up NRM W 10 mT .5mT .0 Sample Mean Directions 2.5 mT .] n=17• 7.5 n= 27 20.0 rite 5.0 •530.0 ß Direction]_ 40.00oE 7.5 Gabbro r•O N:4 015.0 ~ 111 Ma 80.0 00.0 • 27.5 [] = Gabbro 37.5 45.0• 60'0N 50.0 60.0 • 80.0'

Diorite ~110 Ma Directions• I 10e-1 A/m TI1B 3A I 10e-1 A/mTI16D3A I•l 10e-1 A/mTI1D 6A

Fig. 2. Orthogonalprojection of vectorendpoint diagrams showing demagnetization behavior of samplesfrom ThurstonIsland. Opencircles (solid circles) are projections on the vertical(horizontal) planes at indicatedlevels of AF (mT, millitesla)or TH (øC) cleaning. Equal-angleprojections of characteristiccleaned sample and ske meandirections. Open (solid) symbolsare on the upper (lower) hemisphere. Intensity plots are shown for localities where TH demagnetizationwas used to determinethe directions. 17,938 GRLTNOWET AL.: NEW PALEOMAGNETICDATA FROM THURSTONISLAND

e. Harrison Nunatak f. Hale Glacier ^ wUp g. W,Upt:•NRM 100.0øC•I ' W,,UpthNR M I • •1 I J NRMW'Up 0'E I1o• I•l I ? 1øø'øøc 5.0mT I•o••'•I •5mT•o.o • 2•ø•••1•,..,,,I I •20• o.•.ok •,• I 10'0••ø'ø•Volc.I n,c•$••••••t• ...... • I 40'0• •144Ma ?• •. i, •go.o.o,,uffs•'::• •,, ......

I • / •530 o I • • •56 oø I • •1•-1 • TI4A5A 4•ø•5••N I • 0'0• '160 '2•' 3•0'4• '5•0 '•0 •N

h.Shelton Head •o ' . i.Mt. Simpson W.•5ø 0.8k TI21D3A I 0øE • 0.6• • • = e I

100.0 0.2 I • Granite • / - -• = •ø'øo,oo:oo'•6o'4•'•6o •o• •" '•" I...... -. t

•.o.• •oø• 0.0 N 10e-2MmTI21D3A 180ø 0 ' 1•' 2•' 3•0' 4•0' 5•0' 6•øC • I 1•-2• TI22A1A

j. Landfall Peak !. Mt. Bramhall OøE

Di n=24 ~ 152 Ma Indusiøn•l• N •o__...... W,Up NRM Directions N L•?1 Diorite• I • 5•t SampleMean--W TI6C5A[ ,Granodiorite•230 Ma 63 ...... 180 ø

. .

600 ø N--4 •l• 0.8 :: : : : : : ; 1.0Mt. Bra•all 590ø o \ SiteMean 0.6 T114B 6A o ' 56o• • 0.4 5705300500ø 40Oø • Directions • "• 1•-2•mTI14B 6A• ] • • • • J Directi•s I----110e-2A/mT!6C5A • 0.00.2 180 ø • 10e-2• TI14C4A Fig. 2. (continued) t8oø

3 ø of the predicted magnetic variation and therefore no The sampleswere analyzedin the paleomagneticlaboratory at correctionswere made. Becausemost of the rocks exposedin Lamont-Doherty Geological Observatory. Measurements of the TI block are intrusions,we attemptedto sampleover as the natural remanentmagnetization (NRM) were made using a many widely scattered sites and different cooling units as cryogenicmagnetometer or a computerizedflux gate spinner possible within any given intrusion in order to average out magnetometeron one or more specimenscut from each core. secularvariation and assessthe likelihood of tectonic tilting. Pilot sampleswere progressivelydemagnetized using stepwise We were sometimes unable to collect more than one or two alternatingfield (AF) or thermal techniques(TH) to determine sites at a given location due to weather conditions. the best demagnetization procedure for that site. Steps of GRUNOWET AL.: NEW PALEOMAGNETICDATA FROM THURSTONISLAND 17,939

W,Up mT W,Up Volcaniclasticrocks k. Mt. Dowling > 164 Ma 0øE 10.0 NRM SampleMean 0OE .0 W,Up10.O 45.0 mT Directions 20.0 20.0[]= Intrusions n=12 Intrusive Baked contact 30.0O = Volcaniclastic rocksn=12 < 164Ma 30.0withintrusion O = Tilt-corrected -[ 30.0 Without tilt Volcaniclasticrocksn=12 35.0 correction 40.0 A= Baked Volcaniclastic rocksn=3.[• \ 40.0 lk = Tilt-correctedBaked 45.0 •60.050.0 Volcaniclasticrocksn=3 50.0 ,50.0 70.0 60.0 55.0 .0, 65.0 80.0 70.0 • 100øC, 200 ø 9O.O 300ø-500 ø N _ /

59Oo, 600 • 585 ø 610 ø N H 10e-4A/m geoCoords TI12A 5A 1410e-4A/m geoCoords I--110e-3 ABn bed Coords TI12D 5A T112A 7A N=2Intru•Volcanic rocks

Fig. 2. (continued) generally 10 mT up to a peak of 100 mT were used during AF mafic dikes which in turn are cut by east-northeast trending demagnetization, while steps of 100øC up to 500øC beyond felsic dikes. The Upper Cretaceousvolcanic rocks are in which smaller steps up to a maximum of 660øC were used in contactwith the granite and felsic dikes. The Mesozoic igneous thermal demagnetization. The demagnetization data were rocks are overlain unconformably by Oligocene alkaline- plotted on vector endpoint diagrams and characteristic olivine basalts [Craddock et al., 1969; Rutford et al., 1972]. magnetization directions were calculated for linear segments Results were obtained from-1/3 of the sites by using AF converging toward the origin using principal component demagnetization on three granite, six felsic dike, and three analysis/Kirschrink, 1980]. Directions were consideredstable mafic dike sites. Pilot samples using thermal demagnetization if definedby three or more colinearpoints. Mean directionsfor producedresults similar to the AF demagnetization. each locality are basedon combining(specimen) sample means A baked contacttest carried out on the Lower Jurassicgranite to site means and then site meansto a unit mean using standard intruded by a mafic dike showed the granite yielding the same Fisher [1953] statistics.Location mean poles were calculated directions as the mafic dike (Figure 2b). Although no age by using site-mean virtual geomagneticpoles (VGPs). determinationshave been made on the mafic dikes, they yield directionsvery similar to the 89 Ma felsic dikes (Table 1 and Figure 2b). The unit mean direction for the three granite sites is PALEOMAGNETIC RESULTS similar to the directionsobtained by Scharnberger and Sharon [1972] from the granite. The site mean directionsfrom both the New paleomagneticdata are presentedhere for the TI block granite and mafic dike sites are indistinguishable from the starting with the results from the youngest radiometrically felsic dikes (Table 1 and Figure 2b). Therefore we believe that dated rocks (see Figure I for locations). Radiometric age the granite has been remagnetized by the thermal event that controlis basedon Rb-Sr andK-At analysesby R.J. Pankhurst produced the mid- to Upper Cretaceous volcanics and felsic andI.L. Millar of the BritishAntarctic Survey and the 40Ar/39Ar dikes. step-heating dating method by A.M. Grunow with T.M. We assignan age of -90 Ma for the age of magnetizationof Harrisonand M.T. Heizler. The paleomagneticresults from the the mafic and felsic dike sites and the overprintedgranite. The diorite/gabbrocomplex (-124 Ma) at Belknap Nunatak have mean direction for all 12 sites is D=31.6 ø, 1=-80.1 ø (Table 2). alreadybeen published[Grunow et al., 1987b].

McNamara Island Lepley Nunatak A medium- to coarse-grained, hornblende + biotite-bearing A biotite-bearinggranite (40Ar/39Ar biotite isochronage of diorite body at Peeler'sBluff on McNamara Island is cut by 89 + 1 Ma) cutby northeasttrending mafic and feldpathic dikes aplitic dikes and granitic sheets.The diorite locally contains forms low rounded hills at this isolated nunatak. A total of 40 what appearedto be a magmaticfoliation (site TI16A, Table 1). coresfrom sevensites were drilled. A single componentof Dating of the rocks at McNamara Island has proved difficult. magnetizationpointing steeply upward to the north-northeast Radiometricages range between 97 Ma (K-Ar biotite) [Drake et was consistentlyfound in six of the granite and dike sites al., 1964] to 123 Ma (Rb-Sr whole rock on the granitesheets) (Table 1 and Figure2a). The seventhsite (a mafic dike) yielded (R.J. Pankhurst,personal communication, 1990). We consider directionsup andto the northwestand was excludedfrom further analysis.The unit meandirection for the six sitesis D= 22.4ø; a 109 Ma K-At hornblende age to approximatethe age of magnetizationbecause it was obtainedfrom the palcomagnetic I= -79.2 ø (Table 2 and Figure 2a). samples whereas the other radiometric ages are based on Jones Mountains samples collected elsewhere at Peeler's Bluff. Also, petrographically,the palcomagneticsamples show significant Two hundredand forty-five sampleswere drilled in the Jones secondarychloritization and alterationwhich may have locally Mountainsfrom 34 sitesin pink, coarse-grainedgranite (198 causedremagnetization and radiometricresetting. + 2 Ma), volcanicrocks, mafic dikes,and pink and greenfelsic Four sites (23 cores) were drilled, three of the sites within the dikes (89 + 3 Ma). The graniteis cut by two generationsof diorite and one within an aplitic dike. The aplitic dike proved 17,940 GRUNOWET AL.: NEWPALEOMAGNETIC DATA FROM 7ItURSTON ISLAND

TABLE1. Site Mean Directions and Virtual Geomagnetic Poles From Thurston Island-Eights Coast Site Lithology NRMx Maximum SamplesD, deg I, deg ix95 kappa VGP 10-1A/m DemagnetizationmT nXN ' LongitudeøE LatitudeøS LepleyNunatak (269.6 øE,73.1 øS) TI15A Granite 1.4 90 4X6 2.8 -80.8 1.9 2350 129.5 88.7 TI15B Mafic Dike (050/70NW) 1.2 90 5X6 40.1 -83.7 2.4 1060 221.9 79.2 TI15C Granite 5.3 100 5x.6 21.3 -77.9 2.8 739 149.0 80.4 TI15E Mafic Dike (020/90) 2.2 100 5x.6 14.7 -79.5 1.6 2421 150.8 84.2 TI15F FelsicDike 1.1 90 3X6 49.1 -78.5 5.8 449 186.2 73.3 TI15G Granite 4.1 100 4X6 14.6 -72.8 6.7 191 118.5 74.0 JonesMountains (265.6 øE,73.5 øS) J1B Felsi½Dike (068/70SE) 6 90 5X7 25.9 -79.6 8.0 93 163.8 81.1 J1C FelsicDike (083/75NW) 12 80 6'x6 35.8 -82.7 2.2 892 145.6 80.4 J1E Granite 0.03 100 6'x6 61.7 -78.5 5.4 158 160.1 69.9 JiG FelsicDike (098/40SE) 13 80 6'x6 34.5 -82.9 6.0 128 144.1 80.7 J1N FelsicDike (088/90) 14 45 6'x6 52.1 -77.3 1.8 1412 172.6 71.1 J2B Mafic Dike (108/65SE) 6 80 7X7 35.0 -80.4 4.4 187 170.2 79.4 J2D Felsi½Dike (079/68SE) 2 100 6'x6 35.1 -79.6 5.0 148 174.2 78.5 J2F(A) Mafic Dike (143/78SW) 8 80 5X5 45.9 -80.2 1.4 3080 187.9 76.1 J2F(B) Granite 0.4 80 6'x6 30.4 -78.4 3.0 497 160.6 78.5 J2H Granite 0.3 70 656 21.9 -80.6 4.4 228 170.9 83.3 J2I Mafic Dike (186/78SW) 3 70 6'x6 9.7 -81.6 5.8 110 170.4 87.2 J2J FelsicDike (083/90) 2 70 5X5 358.4 -74.0 11.9 42 82.1 76.6 TI4A Gabbro 8 80 656 50.3 -70.9 3.5 364 161.1 63.3 McNamaraIsland (266.7 øE, 72.6 •) TI16A Diorite 4 100 6'x6 48.7 -81.4 5.1 174 202.1 76.0 TI16B Diorite 17 90 5X6 48.5 -78.4 4.0 361 184.3 73.4 TI16D Diorite 8 100 6X6 68.2 -76.1 5.2 167 191.9 64.8 Dustin Island (264.9øE, 72.6 '•) TI1A Diorite 15 70 656 52.9 -82.9 4.9 190 213.4 75.8 TI1B Diorite 7 50 7X7 59.9 -76.6 6.5 87 184.7 67.8 TI 1C Gabbro 23 50 7X7 80.7 -86.3 2.1 841 240.5 72.3 TI1D Gabbro 10 80 7X7 145.4 -87.3 2.1 838 256.7 67.9 HarrisonNunatak (263.9 øE,72.5 '•) TI4A Gabbro 8 80 6'x6 50.3 -70.9 3.5 364 161.1 63.3

Hale Glacier (259 2 øE,72.2 •) TI8 COM 1 Granite 0.6 560øC 8M3 244.7 -80.6 7.6 54 293.3 59.7 TI8 COM 2 Granite 560øC 5X13 57.0 -73.2 10.4 55 168.5 64.3

ParkerPeak (261.9øE, 72.45'•) VolcanicLayering 335/49NE TI10A Volcanic Tuff 20.0 100 6'x6 96.7 -76.4 2.7 632.0 208.3 57.4 [233.7 -52.2 2.7 624.0 308.6 21.5] TI10B Volcanic Tuff 32.0 100 6'x6 134.3 -77.6 8.9 58.0 234.0 52.1 [229.1 -44.3 8.9 58.0 306.3 13.9] SheltonHead (26225 øE,72.5øS) TI21B COMA Granite 1 580øC 6X7 128.4 -81.0 15.2 20 235.3 58.5 COMB 610øC 7X7 88.0 80.6 9.7 40 318.3 64.4 TI21D COMA Granite 3.5 500øC 3X6 116.2 -59.5 20.6 37 229.3 31.0 COMB 620øC 6'x6 114.0 75.4 3.1 482 340.6 64.4 Mount Simpson(259.25 øE, 72.1 •) TI22A Granite 0.5 620øC 6'x6 264.6 74.8 6.9 95 143.7 58.2 LandfallPeak (257.9øE,72.0øS) TI6B Granitic Dike 0.5 620øC 7X7 218.8 68.1 2.9 444 138.2 63.2 TI6C Diorite 1.4 620øC 5'X6 230.5 67.0 5.5 192 150.6 58.5 TI6E Granodiorite 0.6 640øC 6X6 209.0 74.8 4.9 192 139.5 74.8 TI6F Granodiorite 0.4 640øC 6X6 223.0 72.2 3.3 402 150.5 67.3 MountDowling (261.9 øE, 72.45 øS) Sedimentary Layering 069/15SE TI12A SilicicIntmsion 0.03 100 659 103.7 -79.4 2.5 744 218.7 60.3 TI12B Felsi½Dike (049/90) 0.1 100 6'x6 100.0 -78.3 1.7 1607 214.0 59.4 TI12C Volcaniclastic 0.09 100 6'x6 66.1 -76.9 3.2 432 187.8 66.3 [19.7 -69.7 3.2 427 116.4 69.3] TI12D Volcaniclastic 0.2 100 6'x6 46.9 -75.3 2.6 691 165.9 70.0 [13.4 -65.5 2.6 686 103.2 64.4] Mount Brainhall (261.6 øE,72 2 •) TI14A Granodiorite 3.7 660øC 6X6 8.3 -66.2 3.3 405 95.2 66.1 TI14B Diorite 2.3 660øC 6X6 221.8 63.4 1.8 1352 140.3 56.5 TI14C Granodiorite 1.6 660øC 6x6 19.0 -59.5 4.2 259 108.5 56.8 Maximumdemagnetization usingeither AF or TH techniques; nXN,number ofsamples used for site mean calculation/total number ofsamples; D, declinationin degrees; I, inclination in degrees; brackets indicate tilt-corrected data; 0t9s , radius of error circle at the 95% confidence level for site mean direction;kappa, estimate of precisionparameter, VGP, virtual geomagnetic pole. GRUNOWET AL.: NEW PALEOMAGNETICDATA FROM THURSTONISLAND 17,941 to be magnetically unstable. A single component of rocks are older than 144 Ma, but given the alterationwithin the magnetization, defined by an upward pointing, northeast rocks,it seemspossible that they havebeen remagnetized. directedvector was found (Table 1 and Figure 2c). The unit mean direction for the three sites is D= 56.5 ø, I= -78.8 ø (Table 2 and Shelton Head Figure 2c). This locality consistsof a pink, medium-grainedgranite with grey-green xenolithsof volcanic rocks and/ormafic dikes. The Dustin Island granite is cut by a set of east-northeasttrending mafic dikes. Two small exposuresseparated by -30 m of ice were visited at Four sites were drilled, two in the mafic dikes and two in the Ehlers Knob on Dustin Island (Figure 1). One was a coarse- granite. Age dating of the granite has been difficult; Rb/Sr grained diorite and the other a coarse-grainedgabbro, the whole rock dating yields an age of 137 + 27 Ma, while K-Ar contact between which was not exposed. K-Ar and Rb-Sr analysis yields on biotite an age of-152 + 2 Ma and on biotite ages of-110 Ma were obtainedfrom the gabbro. Four hornblende an age of 142 + 5 Ma. The discrepancyin ages sites (26 cores) were drilled, two sites in the gabbro,two sites suggestsdisturbance in the isotopic system. The crosscutting in the diorite. dikesyield an age of 123 + 20 Ma. A single component of magnetization was observed in the The two granite sites produced two-component diorite after AF demagnetizationthat was upwarddirected to the demagnetizationtrajectories (Table 1 and Figure 2h). The first northeast(Table 1 and Figure 2d). These resultsare very similar component (A) is upward pointing to the east-southeast and to onesobtained by Scharnbergerand Sharon [1972] from three occurs in the lower temperature part (<530øC) of the sites in the diorite. A single, very steep component of demagnetization plot (D= 120.8ø, I=-73.8 ø) (Table 2 and magnetization was observed in the gabbro that was upward Figure 2h). The secondcomponent, B, is downwardpointing to pointing to the east-southeast(Table 1 and Figure 2d). the east (D= 103ø, I= 78.5 ø) and occurs between 530øC and The diorite and gabbro may be of different age, but the 620øC (Table 2 and Figure 2h). variation in direction between the two units is not significant with so few sites'. The unit mean direction for the four sites Mount Simpson from the diorite and gabbro is D= 067.4ø, I= -83.9 ø (Table 2 We were able to sample only one site (six samples)from this and Figure2d). coarse-grained,pink granite becauseof weather conditions.By

Harrison Nunatak combining samples from Mount Simpson with the nearby Mount Noxom granite,an Rb/Sr wholerock age of 145 _+2 Ma This locality consists of a coarse-grained hornblende- was obtained[R.J. Pankhurst,personal communication, 1990]. bearing gabbro. Difficulty of access and weather conditions A two-componentmagnetization was observedin theserocks preventedus• from samplingmore than one site (six samples). after thermaldemagnetization (Tables 1 and 2 and Figure 2i). A A K-Ar age of-121 + 6 Ma was obtained from a very low- lower unblockingtemperature component below -500øC was potassium bearing hornblende separate. A single, stable generallyupward pointing but variablein orientationand may upward directed to the northeastcomponent of magnetization represent a recent overprinting. The higher unblocking was observedin thesespecimens after AF demagnetization(D= temperaturecomponent, isolated between-500 ø to 620øC, 50.3ø, I=-70.9 ø) (Tables 1 and 2and Figure 2e). yielded consistentdownward pointing west-southwest oriented directionsof D= 264.6ø, I= 74.8ø (Tables 1 and 2 Figure 2i).

Hale Glacier Landfall Peak This outcrop consists of a pink, coarse-grained, biotite- Located at the westernmost end of Thurston Island, 50 km east bearing granite cut by north-south trending mafic dikes. The of Mount Simpson, this medium-grained, hornblende + exposure occurs on the eastern side of Hale glacier at the biotite-bearingdiorite/granodiorite body is locallynet veined glacier's entrance onto the Abbot Ice Shelf. K-Ar dating on by a grey, biotite-bearinggranite which in turn is cut by biotite yielded an age of 144 _+5 Ma for the granite.While the graniticdikes of similarcomposition. The upperpart of the granite sites were stable, the directions yielded were single nunatakappears to be an intrusionbreccia with raftsof granitic component but in two orientations unrelated to the site or and mafic material within a diroitic matrix. The main demagnetizationhistory. The directions for component 1 are granodioritebody from the lower part of thenunatak yielded an D= 244.7ø, I= -80.6 ø, while for component2 the directions are 40Ar/39Ar biotite isochronage of 152 _+2 Ma. Six sites were D= 57ø, I= -73.2ø (Tables1 and2 andFigure 2f). drilled(40 cores).A sitein thenet-veined granite and a sitein the dioriteproved to be magneticallyunstable. The remaining Parker Peak two siteswithin the granodiorite,one site in a diorite block Two sites, -5 m apart, were sampled within tilted, well- surroundedby graniticnet veining,and one site in a granite layered, andesitic crystal tuffs. A dolerite sill within the dike cuttingthe dioriteyielded stable directions. volcanicswas datedat 144 + 3 Ma usingK-Ar whole rock (R.J. Like the Mount Simpson samples, the Landfall Peak Pankhurst, personal commmunication,1990). The volcanic granodioritecontains a two-componentmagnetization (Tables rocks are fractured with epidotization occurring along the 1 and 2 and Figure 2j). A lower unblocking temperature fracture surfaces. A single componentof magnetization was componentwas generallyupward pointing but varied in observedin theserocks after AF demagnetization(Table 1 and directionand did not producewell-grouped results. The higher Figure 2g). The mean directionsfor the 12 samplesfrom the temperaturecomponent yielded consistent downward pointing two sites are D= 114.4 ø, I= -77.7 ø before tilt correction and D= to the southwestdirections (Tables 1 and 2 and Figure2j). This 231.2 ø, I= -48.3 ø (Table 2 and Figure 2g). The age of high coercivity component(30% of the magnetization magnetization for this location is unknown. The volcanic remainedafter 100 mT) wasdefined between 570øC and 620øC 17,942 GRUNOW ET AL.' NEW PALEOMAGNETICDATA FROM THURSTONISLAND GRUNOWETAL.: NEW PALEOMAGNETIC DATAFROM THURSTON ISLAND 17,943 and given this high unblockingtemperature (Tub > 585øC), we localitiesto producetime averagedpoles at ~90 Ma, ~110 Ma, believe the carrier to be a titanohematite. The mean directions and ~150 Ma age for the TI block. The similarityin agesand for Landfall Peak are D= 221.3ø, I= 70.7ø (Table 2). paleopolesfrom the JonesMountains and Lepley Nunatak(125 km to the east) allowed us to combine these two localities to Mount Dowling give a ~90 Ma TI paleopoleof 161.4øE,80.8øS, A95= 3.9 ø, N= Gently dipping, thinly bedded, dark green to gray 18 VGPs (Table 3 and Figure 3b). This pole is significantly volcaniclasticrocks are cut by a northeasttrending, 5-m-wide differentfrom the pole obtainedfrom Tertiary volcanicrocks in vertical feldspathic dike and a silicic intrusion at Mount the JonesMountains and HudsonMountains by Scharnberger Dowling.Rb/Sr wholerock datasuggest an ageof at least~164 and Sharon [1972]. + 9 Ma for the volcaniclastic rocks (R.J. Pankhurst, personal We have chosen to combine the Dustin Island and McNamara communication, 1990). Another outcrop visited by one Island VGPs becausethe radiometricages overlap to yield a memberof the expeditionproved to be a felsic intrusionwhich ~110 Ma pole of 209.8øE, 73.1øS,A95=7.6 ø, K=64 for N=7 yielded an Rb/Sr age of 182 + 2 Ma and may be cogeneticwith VGPs (Table 3 and Figures 3a and 3b). The Harrison Nunatak the volcaniclastic rocks (R.J. Pankhurst, personal (121 Ma) VGP is differentfrom both the ~110 Ma pole and the communication, 1990). The dike and silicic intrusion have so ~125 Ma pole (from Belknap Nunatak, Table 3) and has not far proved to be undatable. Three sites were drilled in the been used because we believe that either secular variation has volcaniclasticrocks of which one was not magnetically stable, not beenaveraged out the rockshave been tilted or the age does one site in a feldspathic dike and one site in the silicic not reflect the time of magnetization. Argon loss in intrusion for a total of 32 cores. The remaining two hornblendeis not uncommonand combinedwith the very low volcaniclastic sites were separatedby ~30 m of section and potassiumcontent of the mineral separateupon which the age were located~20 m from the dike. Three volcaniclasticsamples is based (0.1%; R.J. Pankhurst (personal communication, were collected adjacentto the intrusionfor a baked contacttest. 1990)), we feel that the radiometricage may not be reflecting A single, upward pointing east-southeast directed the age of magnetization. magnetization was observed from the feldspathic dike and TheVGPs from Landfail Peak and Mount Simpson produced a intrusion (D= 101.8ø, I=-78.9 ø) (Tables 1 and 2 and Figure ~150 Ma pole of 144.9øE, 64.5øS,A95=7.0 ø, K=122, N=5 2k). The contact between the dike and the intrusion was not VGPs (Table 3 and Figures 3a and 3b). We believe that the observed but it is likely that the dike originated from the reversedpolarity magnetizationwas acquiredprior to the intrusion and probably accountsfor the high precision value. Cretaceous normal polarity interval since the data give The volcaniclastic rocks yielded upward pointing north to paleolatitudes too low for a Late Cretaceous to Tertiary northeast directions (D= 56 ø, I=-76.3) (Tables 1 and 2and position and are not similar to the Late Cretaceousto Tertiary Figure 2k). The samplescollected adjacent to the intrusionfor a AP or East Antarcticreference poles [Watts, 1982; Watts et al., baked contact test did not yield directions similar to the 1984]. We have not used the Hale Glacier (145 Ma) results intrusion but did yield directions similar to the other sites becauseof the lack of within site precisionand the uncertainty within the volcaniclastic rocks (Figure 2k). Therefore we in the origin of two different single-component believe that the volcaniclastic rocks have not been magnetizationswithin a site. It is interestingto note that remagnetizedby the adjacentintrusion and were probably tilted component 2 from Hale Glacier is not disimilar to the ~150 Ma during emplacementof the intrusion.We considerthe age of polesand very similarto HarrisonNunatak (Table 2 andFigure magnetizationto be ~175 Ma basedon the 164 Ma minimum 3a). age for the volcaniclasticrocks and the 182 Ma age for the The bakedcontact test at MountDowling suggests that the nearby felsic intrusion. volcaniclasticrocks were not remagnetizedby the silicic Mount Bramhall intrusions.We considerthe single-componentdirections to be primaryand have applied the tilt correctionbecause it is likely Mount Btamhall consistsof a medium-grained,hornblende + that thesefine-grained volcaniclastic rocks were deposited biotite-bearing granodioritebody with pods of finer-grained nearthe horizontal datum. The coincidenceof K-Ar (andAr/Ar) diorite schlieren.The 40Ar/39Ar analysisyielded a biotite agesin hornblendeand biotiteat Mount Btamhall(230 Ma) isochron age of ~228 _+ 5 Ma, while K-At analysis on suggeststhat theserocks did not experiencea prolonged hornblendeyielded ages of ~230 Ma. Two sites were drilled in coolinghistory and have not beenreheated beyond the closure the granodioriteand one in a diorite schlierenpod for a total of temperatureof biotite(~300øC). This combined with the high 18 cores. A single high-temperature component of unblockingtemperature of the magneticminerals, the single- magnetization (unblocking temperature to 670øC) was componentnature of the magnetization, and the normal and observedin theserocks and is most likely deutericallyformed reversedpolarity directions suggest that the magnetizationis hematite. The granodiorite sites yielded upward pointing, primary.The similarityof the 230 Ma and 175 Ma TI poles north-northeastdirected vectors (D= 23.3ø, I=-63.7 ø) (Tables (Table 3 and Figures3a and3b) suggeststhat the TI block 1 and 2 and Figure 2/). The diorite schlieren site yielded experiencedlitfie apparentpolar wander. downward pointing, south-southwest directed vectors The SheltonHead component B and Hale Glacier component 1 approximately antipodal to the granodiorite (Tables 1 and 2 resultsare not like any of our otherresults from ThurstonIsland and Figures 2/). andcould be reflecting remagnetization (Table 2 andFigure 3a). The granitessampled at HendersonKnob, Mount Noxom, and The closesimilarity of SheltonHead component A with our did not yield internallyconsistent results (Figure results from Belknap Nunatak (125 Ma) may reflect 1). remagnetizationin the EarlyCretaceous caused by the igneous INTERPRETATION OF mI BLOCK DATA event that produced the -123 Ma Shelton Head dikes. The Parker Peak paleopole without tilt correction is similar to All mean poles of the localities from the TI block are shown thosefrom BelknapNunatak, Shelton Head component A, and in Table3 andFigure 3a. We havecombined poles from several theMount Dowling dikes (Figure 3a). TheMount Dowling dike 17,944 GRUNOWET AL.: NEWPALEOMAGNETIC DATA FROM TItURSTON ISLAND

0øE ar>•k4erPeak-tilt-corrected•

270øE 90øE

o

b.

Fig. 3. (a) Equal-angleprojection of TI poleswith theirrespective A63s (except for Hale Glacier,Harrison Nunatak, Mount Simpson,and Shelton Head where the data are too few). D andG, DustinIsland diorite and gabbro. Poles are from Table 2. (b) TI apparentpolar wander path (with A63s);see Table 3.

TABLE 3. TI Paleopolesfor TI ApparentPolar Wander Path Age,Ma LongitudeøE LatitudeøS A95• deg A63,deg Basis for Poles -230 116 61 19.4 8.4 N=3 site mean VGPs from Mount Bramhall -175 109 67 .... N=2 sitemean VGPs from Mount Dowling - 150 145 64 7 3.5 N=5 sitemean VGPs from LandfallPark (4); Mount Simpson(1) -125 232 49 7.9 4.1 N=5 sitemean VGPs from BelknapNunatak -110 210 73 7.6 4.0 N=7 sitemean VGPs from DustinIsland (4); McNamaraIsland (3) -90 161 80 3.9 1.7 N=18site mean VGPs from Leple 7 Nunatak(6); Jones Mountains (12) and intrusion pole lie between the 150-125 Ma and 125-100 However, these data are not used in the reconstructionsgiven Ma TI poles; therefore we considerit possiblethat the age of the uncertaintyin the age of magnetization. magnetizationfor these rocks is between 150 Ma and 100 Ma Althoughwe cannoteliminate the possibilityof tilting for (Figure 3a). The similarity of the SheltonHead, Parker Peak, these data, we believe that the similarity of paleomagnetic and Mount Dowling intrusion poles with Belknap Nunatak polesobtained from variousrock types of equivalentage over a indicates that our previous results were not anomalous. large regionargues against significant tilting of smallblocks. GRUNOW ET AL: NEW PALEOMAGNETICDATA FROM THURSTON ISLAND 17,945

TABLE 4. MeanPalcopoles (With TheirA9s ) usedin thetectonic reconstructions

, , Aide,Ma EastAntarctica* , TI AP EWM MBL References -230 225ø, 54 ø (11 ø) Af 1i6ø, 61 ø (19.4 ø) 1,2 -175 220ø, 55ø (3.9ø) 109ø, 67ø (--) 237ø, 46ø (6.4ø) 235ø, 41ø (5.3ø) 2, 3, 4 -150 201o, 69ø (13o) S.Am 145ø, 64ø (7ø) 1, 2 -125 202ø, 72 ø (11ø) Af 232ø, 49ø (7.9ø) 5 -110 194ø, 72ø (13ø) Aus 210ø, 73ø (7.6ø) 195ø, 65ø (9ø) 1, 2, 6 -90-95 204ø, 76 ø (12ø) Aus 161ø, 80ø (3.9ø ) 132ø, 77 ø (10ø) 242ø, 65ø (9ø) 1, 2, 7, 8 References:1,Irving and Irving [1982]; 2,'this paper, see Table 3; 3, Longshaw andGriffiths [1983]; 4, Grunowetal. [1987a;] 5,Grunow etal. [1987b]'6, Kelloggand Rowly [1989]; 7, Wattset al. [1984]' 8, Grindleyand Oliver [1983]. *The Gondwanalandreference poles (Af, Africa;S.Am, SouthAmerica; Aus, Australia) have been rotated to EastAntarctica using the Norton and Sclater [1979] rotationpoles.

- 90 - 95 Ma - 110Ma Australia at-90 Ma [Irving and Irving, 1982] rotated into an East Antarctic reference frame using the Norton and Sclater [1979] rotationparameters (Table 4 and Figure 4). I MBL[I EAT'•'•_•-¾ \ • • ?:'•::• ', We did not use the mean 150 Ma pole from Africa [Irving and Irving, 1982] because it is primarily based on poles from Morocco where the Cretaceous mobile zone associated with creation of the Atlas Mountains may have disturbedthem. The 120 Ma pole from Grunow et al. [1987b] is based on the 180 ø E 180 ø E African poles listed by Irving and Irving but without the 125 Ma 1•0 Ma Morrocan pole for the reasonstated above. The -120 Ma pole I I I I -t- I I listed by Besse and Courtillot [1988] is very similar to the Grunow et al. pole when rotatedto East Antarctica. t \ I EAT-- • / \ Palcomagneticpoles from the other West Antarctic crustal •,••: • • EAT : blocks used in the tectonic models include the 175 Ma pole \ / •: ....- from the AP and EWM blocks [Longshawand Griffiths, 1983; Grunow et al., 1987a], the -110 Ma pole from the AP block [Kellogg and Rowley, 1989], the-90-110 Ma pole from the 180øE 180 ø E MBL block [GrindIcy and Oliver, 1983], and the 90-95 Ma

175 Ma 230 Ma - -- - pole from the AP block [Watts et al., 1984] (Table 4 and Figure ''' ' + ' ' I'' ' ' + ' 4). RECONSTRUCTIONS \ /" '. • // The pre- and post-Gondwanalandbreakup rotation poles of Norton and Sclater [1979] have been used in the reconstructions for the major continents. The rotation poles used to restore the West Antarctic crustal blocks and the 180ø E 180ø E Falkland Islands (FI) [Taylor and Shaw, 1989] are listed in Fig. 4. Equal-angleprojection of palcomagneticpoles (with their A95s) Table 5. For simplicity, the present continentaloutlines and from the West Antarctic crustal blocks (AP, TI, EWM, MBL) and East subice topographyhave been used to define the crustal blocks Antarctica(EAT) for eachtime period;see Table 4. in the reconstructions,the exact configurationof theseblocks prior to Cenozoic time being unknown. The East Antarctic The flat-lying Tertiary volcanic rocks in the JonesMountains reference pole is used as the geographic in the indicatethat there has been'no tilting here since the Oligocene. figures. The Precambriangneisses from the Haag Nunataks The similarity of poles of mid-Cretaceous and younger age [Millar and Pankhurst, 1987] are consideredto be part of the from TI with those of equivalent age from the AP and East EWM block. For lack of other data, the MBL block is left in its Antarctica suggeststo us that little tectonic tilting on any 90-95 Ma positionbased on the data of Grindley and Oliver scale has occurred since at least mid-Cretaceous time (Table 4 [1983] for the older reconstructions(Table 5). The major and Figure 4). Last, few faults or shear zones were observedat events in the evolution of the West Antarctic blocks, East the locations discussedin this paper, although it should be Antarctica, and the southern ocean basins between -230 Ma emphasizedthat many exposureswere of limited extent. and -90 Ma are outlined in Table 6.

COMPARISON OF PALEOMAGNETIC REFERENCE POLES TECTONIC MODELS

The palcomagneticpoles from the West Antarctic crustal 230 Ma blocks and East Antarctica are shown in Table 4 and Figure 4. Apart from the 175 Ma EastAntarctic reference pole [Grunowet The 230 Ma reconstruction is based on the Mount Brainhall al., 1987a], there are no other well-established Mesozoic pole (Figure 5a). The position shown for the TI block in this referencepoles from East Antarctica.Therefore, for the other reconstructioncreates an opening along the Pacific convergent relevanttime periodswe have usedreference poles from Africa margin between TI and MBL. No data exist for the AP block at at 230 Ma [Irving and Irving, 1982], SouthAmerica at 150 Ma 230 Ma, and therefore we leave it in its 175 Ma position. We [irving and Irving, 1982], Africa at-120 Ma [Grunow et al., believe that the 90 ø counterclockwise rotation of the EWM 1987b], Australia at 110 Ma [Irving and Irving, 1982], and block predicted from Cambrian palcomagneticdata [Watts and 17,946 GRUNOWET AL.: NEW PALEOMAGNETICDATA FROMTHURSTON ISLAND

TABLE 5. Rotation Poles Relative to East Antarctica

Plate Time,Ma Lon[•itudeo Latitudeo An[lie,de[• Thurston Island 230 88.22 78.16 102.47 175 86.14 74.23 79.96 150 91.64 79.72 35.68 125 -72.94 -61.10 29.64 110 -179.54 85.33 -5.40 90 0.00 0.00 0.00 Antarctic Peninsula 230 -79.50 -70.02 51.89 175 -93.20 -72.30 35.48 150 -94.73 -67.32 27.23 125 -72.94 -61.01 29.64 110 0.00 90.00 -1.81 90 0.00 0.00 0.00 Ellsworth-Whitmore Mountains 230 -27.54 -84.05 -79.07 175 -88.10 -62.14 35.37 150 -88.86 -62.17 33.69 125 -81.65 -62.59 34.14 90 0.00 0.00 0.00 230 18.55 61.91 22.28 175 34.84 53.63 12.54 90 34.84 53.63 12.54 40 0.00 0.00 0.00 Falkland/Malvinas Islands 230 169.20 45.30 -156.25 (fixed to Africa) 175 -33.40 47.40 55.77 Positiverotations arecounterclockwise whenviewed from above the p01e. øWest and south are shown with (-) sign.

Bramall, 1981; Grunow et al., 1987a] and geologic data The FI block is shown after a 155 ø clockwise rotation in two [Schopf, 1969; Dalziel and Elliot, 1982] occurredbetween the positions'its presentday positionwith respectto South Permo-TriassicGondwanide Orogeny and the emplacementof Americaand in a positionnear Africa (Figure 5b). An estimated the 175 Ma granitesin the EWM block [Millar and Pankhurst, -400 km of Middle to Late Jurassic east-west directed 1987] from which the EWM 175 Ma pole is based[Grunow et extensionin the Falkland Plateauarea [Lorenzo and Mutter, al., 1987a, b] (Table 6). A proposed120 ø clockwiserotation of 1988]may have moved the FI blockwestward to itspresent day the FalklandIslands (FI) occurredbetween Early Jurassicand positionwith respectto SouthAmerica. the opening of the South Atlantic [Taylor and Shaw, 1989]. -150 Ma Therefore, in the reconstructions, we have shown the EWM and FI in a prerotationposition at 230 Ma and in a postrotation Based on comparisonof the 150 Ma East Antarcticand TI position at 175 Ma. pole means, counterclockwise rotation of the TI block is The EWM and FI rotationsmay have been causedby early predictedbetween >150 Ma and-125 Ma (Tables4 and 5 and extensionand associatedtranscurrent motion during the early Figure5c). The TI rotationprobably started with creationof the breakupof the supercontinent[Dalziel et al., 1987;Taylor and first sea-floor (M25 which is 156 Ma using the Decadeof Shaw, 1989; Dalziel and Grunow, in press]. Early Jurassic NorthAmerican (DNAG) timescale [Palmer, 1983]) in mafic lavas found in DronningMaud Land [L•vlie, 1979] could the Mozambique-Somali basins [Simpsonet al., 1979; be indicating early extension along this margin. Early Segoufinand Patriat, 1980;Rabinowitz et al., 1983] andceased Mesozoic strike-slip shear zones have been documentedin with opening of the South Atlantic Ocean basin at-130 Ma Patagonia[Rapela, et al., 1989] providing additionalevidence (M10) [Rabinowitz and LaBrecque,1979] (Table 6). This that major transcurrent motion occurred just after the spreading system was connected to the southeasternWeddell Gondwanide Orogeny and prior to the main phase of Sea alongthe ExploraEscarpment [Lawver et al., 1991] and Gondwanalandbreakup (Table 6) [Rapelaet al., 1989]. presumably to the southwesternWeddell Sea basin and Rocas 175 Ma Verdesbasin in SouthAmerica [Dalziel, 1981;Mukasa et al., 1988] (Figure5c). Upper Jurassicto Lower Cretaceousback are The 175 Ma reconstructionis basedon the AP and EWM 175 basinsedimentary deposits found along the easternside of the Ma poles[Longshaw and Griffiths,1983; Grunow et al., AntarcticPeninsula provide geologic evidence for existenceof 1987a] (Figure5b). Minor counterclockwiserotation of the AP a submarinebasin in the westernWeddell Sea area [Suarez, andEWM blocksis requiredto restorethese poles to the175 1976; Thomson,1982; Kellogg and Rowley, 1989]. Geosat Ma EastAntarctic pole. We havespeculatively shown an data reveal northeast directed lineaments in the southwestern extensionalbasin between the AP-TI and EWM blocks and Weddell Sea Basinrecord opening behind the AP block [Bell, betweenEWM and East Antarcticawhich coincideswith the 1989] (Figure 5c). The Geosat data also reveal a series of presentlocation of the Byrd subglacialbasin (BSB) and presently east-northeast trending lineaments in the ExploraWedge (EW), respectively (Table 6 andFigure 5b). southeasternWeddell Sea basin interpretedas fracturezones Extensionof suchbasins along the TransantarcticMountains that record east-northeastdirected Late Jurassic-Early is consistentwith structuraldata from this region [Wilson, Cretaceousextension [Bell, 1989; Bell et al., 1990] related to 1991] and the location of Middle Jurassicextension related the Mozambiquebasin opening. igneousrocks in southernSouth America and the AP block The southward motion of an EWM block connected to East [Dalzielet al., 1987].This reconstruction would also suggest Antarctica during opening of the Mozambiquebasin may have thatthe TI coastline parallel to thepresent trend of PineIsland caused the counterclockwise rotation of the TI block and Bay (Figure1, inset)was once along the Pacific convergent transpressional deformation, known as the Palmer Land margin(Figures 5a and5b). deformationalevent, in the southernAP block [Kelloggand GRUNOWET AL.: NEW PALEOMAGNETICDATA FROM THURSTON ISLAND 17,947 17,948 GRUNOWET AL.: NEW PALEOMAGNETICDATA FROMTHURSTON ISLAND

SOUTH 2 3 0 M a AMERICA

AFRICA

/

// / / / / " / / /• eat ANTARCTIC•

• / /

South Pole/ "' J

Fig.5a. Reconstructionfor230 Ma. In allof the reconstructions, thegeographic south pole (+) is centeredonthe East Antarctic (eat)reference pole; the West Antarctic paleopoles arelocated with asterisks. For simplicity, the present continental outlines and sub-icetopography have been used to define the crustal blocks. The FI blockis pointed on one side to better display its rotation andis notmeant to reflectits configuration in the Mesozoic. Abbreviations areas follow: AP, Antarctic Peninsula; ap, AP block A95circle; BSB, Byrd Subglacial Basin; EW, Explora Wedge; EWM, Ellsworth-Whitmore Mountains; ewm, EWM block A95 circle;MAD, Madagascar;, MB, MozambiqueBasin; MBL, Marie Byrd Land; mbl, MBL block paleopole; SB, Somali Basin; SNZ,South Island New Zealand; PIB, Pine Island Bay; PL, Palmer Land; RVB, Rocas Verdes Basin; TI, ThurstonIsland; ti, TI blockA95 circle. At 230Ma, approximately -90ø of TI clockwiserotation isneeded torestore the TI poleto the East Antarctic one.The EWM and FI blocksare shown in theirprerotation positions. The AP and MBL blocks are left in their175 Ma and100 Mapositions, respectively. The bold lines mark the location ofthe Gondwanide foldbelts in South America, Africa, the FI block, the EWM block and East Antarctica.

Rowley,1989] (Table 6 andFigure 5c). We showa RFF triple -130 Ma junctionin the southwesternWeddell Sea basin with a faster relativerate of openingin the easternversus western Weddell The two-plate rifting system between East and West Sea basin to account for dextral shearing (Figure 5c). The Gondwanaland(the AP block constitutesa third small plate presumedtectonic boundary between the AP andTI blocksmay prior to 130 Ma) becamea three-platesystem with the startof have been located at the base of the AP block in what is called SouthAtlantic opening [Lawver et al., 1991]. We proposethat the EllsworthFault System (EFS) [Kelloggand Rowley, 1989]. this changeto a three-platesystem affected the movement Movementalong the EFS is thoughtto havebeen in the Late historyof the AP, TI, and EWM blockssuch that the EWM Cretaceousor Tertiary [Kellogg and Rowley, 1989] but based block becameconnected to the AP-TI blocks creatinga fourth on the new TI paleomagneticdata, the EFS may have alsobeen minorplate that we referto asWeddellia. A --30ø clockwise active in the Jurassic and Early Cretaceous.The TI block rotationof Weddelliawith respectto EastAntarctica is required rotationmay also have been causedby strike-slip associated between ~125 Ma and -100 Ma based on our results from with the Pacific convergentmargin, like that observedin the BelknapNunatak (Tables 4 and5 andFigure 5d) [Grunowet al., Andes [Beck, 1989; Cunninghamet al., 1991]. The EWM 1987b].This wouldproduce sinistral shear of-750 km between block may not have escaped deformation along this the EWM block and East Antarctica which is significantly less transpressionalzone becausethe marginsof the EWM block thanwhat we proposedearlier [Grunow et al., 1987b]because record an unusualsecond structural fabric [Storeyand Dalziel, in this model the EWM block remained attached to East 1987] that we have related to shearingof the EWM against Antarctica between -150 Ma and 130 Ma. Inferred Cenozoic both the Antarctic Peninsula and East Antarctica [Grunow et al., extension within the EWM block [Garrett et al., 1987] may 1987b] (Table 6). explainthe overlapof the EWM blockwith the AP andTI GRUNOW ET AL.: NEW PALEOMAGNETICDATA FROM THURSTON ISLAND 17,949

175 Ma SOUTH AMERICA

AFRICA

AUSTRALIA

Fig. 5b. Reconstructionfor 175 Ma; captionsame as Figure 5a. The TI, AP, and EWM blockshave been shownusing the individualAP, TI, and EWM poles. The EWM blockhas rotated -90 ø between230 Ma and 175 Ma. The FI block is shownafter a -155 ø rotationin both a present-dayposition with respectto SouthAmerica (FI 1) and in a positionadjacent to Africa (FI 2). The locationof the T1-175 Ma? palcopole(ti) is alsoshown. Speculativelyshown are possiblerift basins: the Byrd Subglacial Basin (BSB); the Explora Wedge (EW) and rifting along the TransantarcticMountains and extensionwithin southernSouth America and on the eastern(backare) side of the AP block. blocks.A RFF triple junctioncould explain this shearmotion Island Bay area (Figure 5d). If MBL was near the Transantarctic (Figure5d) whilethe RRR triplejunction connects the opening Mountains during this time period, then sinistral motion is in the Weddell Sea to that in the South Ariantic and southwest requiredbetween MBL andWeddellia as Weddelliarotated into Indian oceans [Lawyer et al., 1991]. The relative rate of its presentday positionwith respectto East Antarctica. openingin theWeddell Sea would have to havebeen faster than A model proposedby Kristofferson and Haugland [1986] in the SE Indian •Ocean to account for sinistral shearing suggestedthat Weddellia was lockedto East Antarcticaby tho between the EWM and East Antarctica. The NE trending Middle to Late Jurassicon the basisof interpretingthe Andenes lineamentsin the GEOSAT data [Bell, 1989; Bell et al., 1990] and ExploraEscarpments as the samerift structure(Figure 1). foundadjacent to the AntarcticPeninsula and acrossmost of the Lawyer et al. [1991] consider the Andenes Escarpmentto be Weddell Sea reflect northeast directed extension that provided relatedto openingof the centralWeddell Sea basinthat began the mechanism for the clockwise rotation of Weddellia. with the openingof the South Atlantic basin -130 Ma; its Upper Jurassicto Lower Cretaceoussedimentary rocks in the alignmentwith the Explora Escarpmentis coincidental.They northern AP block experiencedEarly Cretaceousstrike-slip alsosuggest that Weddellia was locked to EastAntarctica when deformation [Whitham and Storey, 1989]. Initiation of theopening in the SouthAtlantic ocean basin began, but clear spreadingin the SouthAtlantic oceanbasin may have resulted identificationof magnetic anomaliesthat definitively tie East in a cessationof spreadingin the Rocas Verdes basin and Antarctica to Weddellia have not been documented; initiation of a strike-slip setting in southernmost South Controversystill occursover the identificationof magnetic America and the northernmostAntarctic Peninsula (Figure 5d). anomalies in the Weddell Sea; the M29 (-160 Ma) anomaly There could also have been continued dextral strike-slip identifiedby LaBrecqueand Barker [1981] is thoughtto be M10 motion between the EWM block and the base of the AP block (-130 Ma) by Martin and Hartnady [1986], coincidentwith between -125 Ma and ~100 Ma given that deformation is openingin the SouthAtlantic oceanbasin. thoughtto havecontinued in thesoutheastem AP blockuntil at leastthe Albian [Meneilly et al., 1987].The WeddelliaEarly -110-90 Ma Cretaceousclockwise rotation may have been transferredto the Weddellia appearsto have been in or near its present-day Pacific convergentmargin along a strike-slip zone in the Pine positionwith respectto EastAntarctica by ~110 Ma (Figure 17,950 GRUNOWET AL.: NEW PALEOMAGNETICDATA FROM THURSTONISLAND

150 Ma SOUTH AMERICA

AFRICA

INDIA

AUSTRALIA • t.!J

Fig. 5c. Reconstructionfor 150 Ma; captionsame as Figure 5a. Between>150 Ma and-130 Ma, theTI blockunderwent counterclockwiserotation with respect to EastAntarctica. This rotation may have been caused by collisionand sheafing of the EWM blockwith the AP andTI blocksin a dextraltranspressional zone as the EWM block moved southward during opening of theMozambique (MZB) and Somali (SB) basins. We relate the Palmer Land (PL) deformational event to thesheafing of the EWM blockagainst the AP blockand closure of thebackarc basin in thisregion,

5e). The paleomagnetic poles from Lepley Nunatak-Jones Weddellia and East Antarctica became locked together. The Mountains and McNamara Island-Dustin Island are very similar geologichistories of southernSouth America and the Antarctic to the ~95 Ma and ~110 Ma poles from the Antarctic Peninsula Peninsulaalso divergedin the mid-Cretaceouswhen the Rocas [Watts et al., 1984; Kellogg and Rowley, 1989] and East Verdes basin began to close [Dalziel, 1981] while the Weddell Antarctica (Table 4 and Figure 4). The McNamara Island and Sea basin continuedto open (Table 6). Dustin Island diorite poles are virtually identical to that The mid-Cretaceous (90-110 Ma) paleomagneticpole from obtainedfrom the base of the AP block by Kellogg and Rowley MBL, assumingthat the rocks have not been regionally tilted [1989]. Although oroclinal bending at the base of the AP block [Grindley and Oliver, 1983], does not fall on the common has been proposedto explain the difference in the ~110 Ma apparent polar wander path for either Weddellia or East versus~95 Ma AP poles [Kellogg, 1980; Kellogg and Rowley, Antarctica suggestiveof some relative motion between not 1989], we find the agreementof our TI poles with the AP poles only MBL and East Antarctica[Grindley and Oliver, 1983] but to favor rapid apparentpolar wander [Wattset al., 1984]. also MBL and Weddellia (Figure 5e). We have shownthe MBL The marine magneticanomaly data show that by C34 time (84 paleomagneticpole on the 90 Ma reconstructionalthough the Ma), East Antarctica and Weddellia were locked together ages of the MBL rocks range up to 110 Ma and comparison [LaBrecque and Barker, 1981] (Table 6). The paleomagnetic with an older referencepole may be necessaryupon refinement data suggestthat this locking of the plates occurredearlier of the MBL ages.The MBL data suggestpossible displacement (~110 Ma) within the Cretaceous Normal Polarity interval of 200-500 km and dextral rotation of 10ø-45ø with respectto between M0 and C34 (118-84 Ma). At ~110-115 Ma, a major East Antarctica [Grindley and Oliver, 1983]. MBL may have changetook place in the SouthAtlantic andMozambique basin moved away from East Antarcticaduring the Late Cretaceousto openinghistories [Rabinowitz and LaBrecque,1979; Besseand Tertiary stretchingof continentalcrust in the (Figure Courtillot, 1988]. The rotation pole of South Atlantic opening 5e) [Cooper et al., 1991; Wilson, 1991], coinciding with changedat ~ 115 Ma [Rabinowitzand LaBrecque,1979] (using rifting betweenMBL and New Zealand [Mayes et al., 1990]. the DNAG time scale) and at ~110 Ma rifting in the During this time, dextral transtensional motion would be Mozambique basin transfered to the eastern Indian Ocean expected between MBL and East Antarctica, while dextal [Besse and Courtillot, 1988]. This reorganization event may motion with a compressionalcomponent of convergencewould have also altered the Weddell Sea opening history such that be expectedbetween MBL and Weddellia. Antarcticabroadly GRUNOWET AL.: NEWPALEOMAGNETIC DATA FROM THURSTON ISLAND 17,951

125 Ma AMERICA

AFRICA

EAST ANTARCTIC

AUSTRALIA

Fig. 5d. Reconstructionfor 125 Ma; captionsame as Figure5a. At -130 Ma, the SouthAtlantic began to openand may have changedthe openinghistory in the WeddellSea basin such that the EWM blockbecame attached to the AP and TI blocks. A -30 ø dockwiserotation of thesethree blocks, called Weddellia, translated the EWM block-750 km to its present-dayposition with respectto EastAntarctica. A RFF triplejunction is shownin the easternWeddell Sea to providethe mechanismfor Weddellia's rotation. We have tentativelyconnected the RFF triple junctionto a RRR triple junctionin the northeasternWeddell Sea. Openingof the RocasVetdes basin may have ceasedwith initiationof rifting in the SouthAtlantic and sheafingbetween the northemAP block andsouthern South America may haveoccurred. attainedits presentconfiguration by the mid-Tertiary [Lawver southernocean basins. Weddellia, consistingof the AP-EWM- et al., 1991] although there are some younger extensional TI blocks, was in its present-dayposition with respectto East basinsfound in the western Ross Sea [Cooper et al., 1991] Antarctica by 100-110 Ma. We associate the suturing of (Figure 5e, inset). Weddellia to East Antarcticawith the mid-Cretaceouschange in The offset of TI from the MBL coastline in the Pine Island plate motions in the South Atlantic [Rabinowitz and Bay area marks a major geologicboundary in West Antarctica. LaBrecque, 1979] and Indian Ocean basins [Besse and Rotation of Weddellia and MBL in the Early and Late Courtillot, 1988]. Paleomagneticdata from Marie Byrd Land Cretaceous,respectively, is likely to have causedshearing in [Grindley and Oliver, 1983] appearto indicate relative motion the Pine Island Bay region. In addition, Pine Island Bay of MBL with respect to both East Antarctica and Weddellia separatesthe Cretaceousconvergent margin turned passive since the mid-Cretaceous. The relative motion of MBL with margin of the TI block from the solely passive margin respectto Weddelliamay have occurredalong a dextralstrike- evolution of the MBL block. The onshore trace of the slip zone(s) associated with extension in the Ross Sea and UndintsevFracture Zone (dextral offset) appearsto enter the beneaththe RossIce Shelf.The Early Cretaceouspole from TI Pine Island Bay area [Sandwelland McAdoo, 1988] and may indicates that-30 ø clockwise rotation of Weddellia with reflect the oceanic continuationof the major shear zone respect to East Antarctica occurredbetween-125 Ma and-110 betweenWeddellia and MBL. Thereforethe PineIsland Bay area Ma. We relatethis to a majoropening phase in the WeddellSea marks a very fundamentaland long-lived, complex tectonic basinfollowing initiation of SouthAtlantic basin opening at boundarywithin West Antarctica. -130 Ma. If the EWM blockwere rigidly attachedto Weddellia at 125 Ma, sinistral shear motion of-750 km would be TECTONIC CONCLUSIONS expected between the EWM block and East Antarctica. However,some of thismotion may havebeen taken up between The Mesozoic relative motion of the three West Antarctic the AP-TI blocks [Grunow et al., 1987b]. crustal blocks of the Antarctic Peninsula, Ellsworth-Whitmore Results from the Jurassic rocks indicate that the TI block Mountains, and ThurstonIsland-Eights Coast appearsto be underwentmajor counterclockwiserotation with respectto East linked to three major events during the evolution of the Antarctica between-164 Ma and 125 Ma. We relate the TI 17,952 GRUNOWET AL.: NEW PALEOMAGNETICDATA FROMTHURSTON ISLAND

90 Ma SOUTH AMERICA

AFRICA

EAST ANTARCTICA

AUSTRALIA 25 Ma

Fig.5e. Reconstructionfor 90 Ma; captionsame as Figure 5a. TheAP andTI polesare indistinguishable from the East Antarctic pole,therefore we leaveWeddellia in its present-dayposition with respect to EastAntarctica. Locking of Weddelliawith East Antarcticaoccurred by around-110 Ma and may have beena consequenceof a changein the SouthAtlantic rotation pole. Overalldextral motion is predictedbetween MBL andWeddellia between the mid-Cretaceous and the the present. Although the MBL poleis shownon the 90 Ma reconstruction,the rocks actually range between 90 Ma and110 Ma. It is likelythat Pine Island Baymarks a majortectonic boundary in WestAntarctica. Inset is 25 Ma. Eastand West Antarctica are broadly in theirpresent- day configuration. block rotation and the AP block Palmer Land deformational REFERENCF• event to dextral shearingand collision of the EWM block Barron,E.J., G.G.A. Hardson,and W.W. Hay, A revisedreconstruction ofthe againstthe southeasternAP and TI blocksduring this time southemcontinents, Eos Trans.AGU, 59, 436-450, 1978. interval. The tectonicdriving mechanismfor the TI rotation Beck,M.E., Jr., Analysisof Late Jurassic-Recentpaleomagnefic data from activeplate margins of SouthAmerica, Jour. S. Am.Earth Sci., I, 39-52, was providedby Weddell Sea basin rifting connectedwith 1989. openingin the Mozambiqueand Rocas Verdes basins. Triassic Bell, R.E., High resolutionmarine and airbornegravity surveys: Applica- to EarlyJurassic rotation of theEWM blockand perhaps the FI tionsto riftedmargins, Ph.D thesis,Columbia Univ., New York, 1989. could be associated with accommodation of the continental Bell, R.E., J.M. Brozena,W.F. Haxby, and J.L. LaBrecque,Continental crustalong major strike-slip shear zones after the Gondwanide marginsof the WestemWeddell Sea: Insightsfrom airbomegravity and Geosat-derivedgravity, in Contributionsto Antarctic Research I, Antarct. Orogenyand preceding the main breakup of Gondwanaland. Res.Ser., vol. 50, pp. 91-102,AGU, Washington,D.C.,1990. Besse,J., andV. Courtillot,Paleogeographic maps of the continentsborder- Acknowledgments.This work was supportedby the Divisionof Polar ingthe Indian Ocean since the Early Jurassic,J. Geophys. Res., 93,11,791- Programs,National Science Foundation through grants DPP 82-13798and 11,808, 1988. DPP 86-43441 to I.W.D.D. and by the British AntarcticSurvey (BAS), Bradshaw, J.D., P.B. Andrews, and B.D. Field, Swanson Formation and National EnvironmentalResearch Council. We are gratefulto the VXE6 relatedrocks of Marie ByrdLand and a comparisonwith the Robertson Bay Squadronof theU.S. Navy and the BAS air unit for theirlogistical support. Groupof northemVictoria Land, in AntarcticEarth Science, edited by R.L. Veryspecial thanks to Chuck Kroger, our mountaineering expert, and to Chris Oliver, P.R. James,and J.B. Jago,pp. 274-279, AustralianAcademy of Maronefor theirhelp in collectingthe paleomagnetic samples. The radiomet- Science, Canberra, 1983. ric constraintswere essentialfor the interpretationof the dataand we are Cooper,A.K., F.J. Davey, and K. Hinz, Crustalextension and originof gratefulto RobertPankhurst and Ian Millar at BAS andMark Harrisonand sedimentarybasins beneath the Ross Sea and Ross Ice Shelf,Antarctica, in Matt Heizlerat UCLA for theirhelp. The tectonic reconstructions were made GeologicalEvolution of Antarctica,edited by M.R.A. Thomson,J.A. with the assistanceof Larry Lawvetand Lisa Gahaganusing an Evansand Crame,and J.W. 'Ilaomson, pp. 285-292, Cambridge University Press, New Sutherlandterminal at the Institutefor Geophysics,University of Texasat York, 1991. Austin. Discussionswith Robin Bell, Terry Wilson, andespecially Larry Craddock, C., T.W. Bastien, and R.H. Rutford, Geology of the Jones Lawvetas well ascomments by reviewerJohn Geissman and an anonymous Mountainsarea, in AntarcticGeology, SCAR Proceedings, edited by R.J. reviewersignificantly improved this paper. University of TexasInstitute for Adie,pp. 171-187, North-Holland Publishing Company, Amsterdam, 1964a. Geophysicscontribution 888. Craddock,C., P.W. Gast, G.N. Hanson,and H. Linder, Rubidium-Strontium GRUNOWET AL.: NEW PALEOMAGNETICDATA FROMTHURSTON ISLAND 17,953

agesfrom Antarctica,Geol. Soc.Am. Bull., 75, 237-240, 1964b. 13,Marine Science Intemational, Woods Hole, Mass., 1985. Craddock,C., C.M. White, and R.H. Rutford, The geology of the Eights LaBrecque,J.L., and P. Barker, The age of the Weddell Basin, Nature, 290, Coast, Antarct. J. U.S., 4, 93-94, 1969. 489-492, 1981. Cunningham,D., K. Klepeis,W. Goseand I.W.D. Dalziel, ThePatagonian Lawyer,L.A., and C.R. Scotese, A revised reconstruction ofGondwanaland, Orocline: New palcomagneticdata from the Andeanmagmatic arc in inGondwana Six: Structure, Tectonics andGeophysics, Geophys. Monogr. Tierra del Fuego,J. Geophys.Res., in press,1991. Ser.,vol. 40, edited by G.D. McKenzie, pp. 17-23, AGU, Washington, Dalziel, I.W.D., Back-arc extensionin the southernAndes: A review and D.C. 1987. criticalreappmisal, P tu'los.Trans. R. Soc. Lo ndon , Ser. A , 300, 319-335, Lawyer,L.A., J.Y. Royer, D.T. Sandwell, and C.R. Scotese, Evolution ofthe 1981. Antarctic continentalmargins, in Geological Evolution of Antarctica, Dalziel, I.W.D., and D.H. Elliot, West Antarctica: Problem child of editedby M.R.A. Thomson,J.A. Crame,and J.W. Thomson,pp. 533-540, Gondwanaland, Tectonics, 1, 3-19, 1982. CambridgeUniversity Press, New York, 1991. Dalziel, I.W.D., and A.M. Ginnow, Late Gondwanide tectonic rotations Longshaw,S.K., andD.H. Griffiths,A palcomagneticstudy of Jurassicrocks within Gondwanaland,Tectonics, in press,1991. from the AntarcticPeninsula and its implications,J. Geol. Soc.London, Dalziel, I.W.D., and R.J. PaRkhurst, Joint U.K.-U.S. West Antarctic 140, 945-954, 1983. tectonicsproject: An introduction,in GondwanaSix: Structure,Tecton- Lopatin,B.G., and E.M. Odenko,Outline ofthe geology ofMarie Byr_d Land ics and Geophysics,Geophys. MoRogr. Ser., vol 40, edited by G.D. andthe Eights Coast, in AntarcticGeology and Geophysics,edited by R.J. McKenzie, pp. 107-108, AGU, Washington,D.C., 1987. Adie, pp. 843-847, Universitetesfodaget,Oslo, 1972. Dalziel, I.W.D., B.C. Storey,S.W. Garrett,A.M. Ginnow, L.D.B. Herrod, Lorenzo,J.M., andJ.C. Mutter, Seismicstratigraphy and tectonic evolution and R.J. PaRkhurst,Extensional tectonics and the fragmentationof of the Falkland/Malvinasplateau, Revista Brasileira de Geociencias,18, Gondwanaland,in ContinentalExtensional Tectonics, edited by M.P. 191-200, 1988. Coward,J.F. Dewey, andP.L. Hancock,Geol. Soc.Spec. Publ. London, Lovlie,R., Mesozoicpalcomagnetism in Vestfjella,Dronning Maud Land, 28, 433-441, 1987. EastAntarctica, Geophys. J. R. Astron.Soc., 59, 529-537, 1979. Delisle, G., Resultsof PalcomagneticInvestigations in northernVictoria Martin,A.K., andC.J.H. Hartnady, Plate tectonic development of the south Land, Antarctica,inAntarctic Earth Science,edited by R.L. Oliver, P.R. westIndian Ocean: A revisedreconstruction of EastAntarctica and Africa, James,and J.B. Jago, pp.146-155, AustralianAcademy of Science, J. Geophys.Res., 91, 4767-4786, 1986. Canberra, 1983. Mayes, C.L., L.A. Lawver, and D.T. Sandwell,Tectonic history and new De Wit, M.J., The evolutionof the ScotiaArc asa key to thereconstruction isochronchar of the SouthPacffic, J.Geophys. Res., 95, 8543-8567, 1990. of southwesternGondwanaland, Tectonophysics, 37, 53-81, 1977. Meneilly, A.W., S.M. Harrison,B.A. Piercy, and B.C. Storey,Structural Drake, A.A., Jr.,T.W. Stem, andH.H. Thomas,Radiometric ages ofzircon evolutionof themagmatic arc in northemPalmer Land AntarcticPeninsula, and biotite in quartz diorite, Eights Coast,Antarctica, U.S. Geol. Surv. in GondwanaSix: Structure,Tectonics and Geophysics, Geophys. Monogr. Prof. Pap., 501-D, D50-D53, 1964. Set, vol. 40, editedby G.D. McKenzie, pp. 209-219, AGU, Washington, du Toit, A.L, Our WanderingContinents, 366 pp., Oliver and Boyd, D.C., 1987. Edinburgh, 1937. Millar I., andR.J. Pankhurst,Rb/Sr geochronologyin theregion between the Elliot, D.H., Triassic-EarlyCretaceous evolution of Antarctica,in Geologi- AntarcticPeninsula and the TransantarcticMountains: Haag Nunataks and cal Evolutionof Antarctica,edited by M.R.A. Thomson,J.A. Crame,and Mesozoicgranitoids, in GondwanaSix : Structure,Tectonics and G eophys- J.W. Thomson,pp. 541-548, CambridgeUniversity Press, New York, ics,Geophys. Monogr. Set, vol. 40, editedby G.D. McKenzie,pp. 151-160, 1991. AGU, Washington,D.C., 1987. Fisher,R.A., Dispersionon a sphere,Proc. R. Soc.London, Ser. A, 217, Milne, A.M., and I.L. Millar, Mid-Paleozoic basementin eastern Graham 295-305, 1953. Landand its relationto eastAntarctica, GeologicalEvolution of Antarctica, Garrett,S.W., L.D.B. Herrod,and D.R. Mantripp,Crustal structure of thearea editedby M.R.A. Thomson,J.A. Crame,and J.W. Thomson,pp. 335-340, aroundHaag Nunataks, West Antarctica:New aeromagneticand bedrock CambridgeUniversity Press, New York, 1991. elevationdata, in GondwanaSix: Structure,Tectonics and Geophysics, Mukasa, S.B., I.W.D. Dalziel, and H.K. Brueckner,Zircon U-Pb constraints Geophys.Monogr. Set., vol. 40, editedby G.D. McKenzie,pp. 109-115, on the kinematic evolution of the noRhem Scotia Arc, Geol. Soc.Am. Abstr. AGU, Washington,D.C., 1987. Programs,20, A12, 1988. Norton, I.O., and J.G. Sclater, A model for the evolution of the Indian Ocean GrindIcy, G.W.. and P.J. Oliver, Palcomagnetismof Cretaceousvolcanic rocksfrom Marie Byrd Land, in AntarcticEarth Science,edited by R.L. andthe breakup of Gondwanaland,J. Geophys.Res., 84, 6803-6830, 1979. Oliver, P.R. James,and J.B. Jago,pp. 573 -578, AustralianAcademy of Palmer,A.R., The decadeof North AmericanGeology geologic time scale, Science, Canberra, 1983. Geology,I1, 503-504, 1983. Grunow, A.M., I.W.D. Dalziel, and D.V. Kent, EllswoRh-Whitmore Moun- Pankhurst,R.J., Rb-Sr geochronologyof GrahamLand, Antarctica,J. Geol. tainscrustal block, western Antarctica: New palcomagneticresults and Soc. London, 139,701-712, 1982. their tectonic significance,in Six: Structure, Tectonicsand Pankhurst,R.J., Rb-Sr constraintson the ages of basementrocks of the Geophysics,Geophys. Monogr. Set., vol. 40, editedby G.D. McKenzie,pp. AntarcticPeninsula, in AntarcticEarth Science,edited by R.L. Oliver, P.R. 161-171,AGU, Washington,D.C., 1987a. James,and J.B. Jago, pp. 367-372, AustralianAcademy of Science, Ginnow, A.M., D.V. Kent, andI.W.D. Dalziel, Mesozoic evolutionof West Canberra, 1983. Antarcticaand the WeddellSea Basin: New paleomagneticconstraints, Pankhurst,R.J., The Paleozoicand Andean magmatic arcs of WestAntarctica Earth Planet. Sci. Lett., 86, 16-26, 1987b. and southernSouth America, Spec. Pap. Geol. Soc.Am., 241, 1-7, 1990. Grunow,A.M., D.V. Kent, and I.W.D. Dalziel, New Palcomagneticdata Powell, C.M., B.D. Johnson, and J.J.Y. Veevers, A revised fit of East and bearingon theMesozoic evolution of WestAntarctica and opening of the West Gondwanaland,Tectonophysics, 63, 13-30, 1982. WeddellSea, Geol. Soc.Am. Abstr.Programs, 22, 143-144, 1990. Rabinowitz,P.D., and J. LaBrecque, The Mesozoic South Atlantic Ocean and Harrison,S.M., andW.P. Loske,Early PaleozoicU-Pb isotopicage for an evolutionofits continentalmargins, J. Geophys.Res., 84, 5973-6002,1979. orthogneissfrom north-westernPalmer Land, Antarctic Peninsula,Br. Rabinowitz,P.D., M.F. Coffin,and D. Dalvey,The separationof Madagascar Antarct. Surv. Bull., 81, 11-18, 1988. and Africa, Science,220, 67-69, 1983. Irving, E., and G.A. Irving, Apparentpolar wanderpaths Carboniferous Rapela,C.W., R.J. Pankhurst,and S.M. Harrison,Gondwana plutonism of throughCenozoic and the assemblyof Gondwana,Geophys. Surv., 5, 141- northernPatagonia, in InternationalGeological Congress Abstracts, 2,675 188, 1982. pp., AGU, Washington,D.C., 1989. Jankowski,E.J., and D.J. Drewry, The structureof West Antarcticafrom Rutford, R.H., C. Craddock, C.M. White, and R.L. Armstrong,Tertiary geophysicalstudies, Nature, 291, 17-21, 1981. glaciationin the JonesMountains, in AntarcticGeology and Geophysics, Kellogg,K., Palcomagneticevidence for oroclinalbending of the southern editedby R.J. Adie, pp. 239-243, Universitetesfodaget,Oslo, 1972. Antarctic Peninsula,Geol. Soc. Am. Bull., 91,414-420, 1980. Sandwell,D.T., andD.C. McAdoo, Marine gravityofthe southemocean and Kellogg,K., andP.D. Rowley, Structural geology and tectonics of theOrville Antarcticmargin from Geosat,J. Geophys.Res.,93, 10, 389-10,396, 1988. Coastregion, southern Antarctic Peninsula, Antarctica, U.S. Geol. Surv. Schamberger,C.H., and L. Sharon,Palcomagnetism and plate tectonicsof Prof. Pap., 1498, 25 pp., 1989. Antarctica,in AntarcticGeology and Geophysics,edited by R.J. Adie, pp. Kirschvink,J.L., The least-squareline andplane analysis of palcomagnetic 843-847, Universitetesforlaget,Oslo, 1972. data,Geophys. J. R. Astron.Soc., 62,699-718, 1980. Schopf,J.M., EllsworthMountains: Position in WestAntarctica due to sea- Kristoffersen,Y., and K. Haugland,Geophysical evidence for the East floor spreading,Science, 164, 63-66, 1969. Antarcticplate boundary in theWeddell Sea, Nature, 322,538-541, 1986. Segoufin,J., and P. Patriat, Existenced'anomalies mesozoiques dans le LaBrecque,J.L., Southern Atlantic Ocean and adjacent Antarctic Continental bassinde Somalie,implications pour les relationsAfrique-AntaRique- Margin,in OceanMargin Drilling Project,Reg. Data SynthesisSet., Atlas Madagascar,C.R. Acad. Sci., Ser. B 291, 85-88, 1980. 17,954 GRUNOWET AL.: NEWPALEOMAGNETIC DATA FROM THURSTON ISLAND

Simpson,E.S.W., J.G. Sclater,B. Parsons,I. Norton,and L. Meinke, Watts, A.B., J.K. Weissel,R.A. Duncan,and R.L. Larson,Origin of the Mesozoicmagnetic lineations in theMozambique Basin,Earth Planet. Sci. LouisvilleRidge and its relationshipto theEltanin Fracture Zone System, Lett., 43, 260-264, 1979. J. Geophys.Res., 93, 3051-3077,1988. Smellie,J.L., A completearc-trench system recognized in Gondwanase- Watts,D.R., Potassium-argonages and paleomagnetic results from King quencesofthe Antarctic Peninsula region, Geol. Mag. 118,139-159,1981. GeorgeIsland, South Shetland Islands, in AntarcticGeoscience, edited by Smith,A.G., andA. Hallam,The fit of southemcontinents,Nature,225,139- C. Craddock,pp. 255-261, University of WisconsinPress, Madison, 1982. 149, 1970. Watts, D.R., and A.M. Bramall,Paleomagnetic evidence for a displaced Storey,B.C., and I.W.D. Dalziel,Outline of the structural and tectonic history tewainin westernAntarctica, Nature, 293, 638-641, 1981. of the EllsworthMountains-Thiel Mountains Ridge, West Antarctica,in Watts, D.R., G.C. Watts, and A.M. Brainall,Cretaceous and Early Tertiary GondwanaSix: Structure,Tectonics and Geophysics,Geophys. Monogr. palemagneticresults from the Antarctic Peninsula, Tectonics, 3,333-346, Ser.,vol. 40, editedby G.D. McKenzie,pp. 117-128,AGU, Washington, 1984. D.C., 1987. Webers,G.F., and K.B. Sporli, Palaeontological and stratigraphic investiga- Storey,B.C., R.J. Pankhurst, I. Miller, I.W.D. Dalziel,and A.M. Gnmow,The tions in the Ellsworth Mountains, West Antarctica, in Antarctic Earth geologyof ThurstonIsland, Antarctica, Geological Evolution of Antarc- Science,edited by R.L. Oliver, P.R. James,and LB. Jago,pp. 261-264, tica,edited by M.R.A. Thomson,J.A. Crame, and J.W. Thomson, pp. 339- AustralianAcademy of Science,Canberra, 1983. 404, CambridgeUniversity Press, Cambridge, 1991. Whirham,A.G., andB.C. Storey,Late Jurassic-Early Cretaceous strike-slip Suarez, M., Plate tectonic model for southernAntarctic Peninsulaand its deformationin theNordenskj61d Formation of GrahamLand, Antarct. Sci. relationthe southernAndes, Geology, 4, 211-214, 1976. , 1,269-278, 1989. Taylor,G.K., andJ. Shaw,The FalklandIslands: New palaeomagneticdata Wilson,T.J., Mesozoic and Cenozoic structural patterns in theTransantarctic and their origin as a displaceterrane from southernAfrica, in Deep Mountains,Antarct. J. U.S.,in press,1991. Structureand Past Kinematicsof AccretedTerranes, Geophys. Monogr. Ser.,vol. 50, editedby J.W. Hillhouse, pp. 59-72, AGU, Washington, D.C., I.W.D. Dalziel, Institute for Geophysics,University of Texas at 1989. Austin,8701 Mopac Blvd., Austin,TX 78759. Thomson,M.R.A., Mesozoicpaleogeography of West Antarctica, in Antarc- A.M. Ginnow,Department of EarthSciences, University of Oxford, tic Geoscience,edited by C. Craddock,pp. 331-337, University of Wiscon- ParksRoad, Oxford, OX 1 3PR, England. sin Press,Madison, 1982. D.V. Kent, Lamont-DohertyGeological Observatory, Palisades, NY Thomson,M.R.A., and R.J. Pankhurst,Age of post-Gondwaniancalc- 10964. alkalinevolcanism in the AntarcticPeninsula Region, in AntarcticEarth Science,edited by R.L. Oliver, P.R. James,and J.B. Jago,pp. 328-333, AustralianAcademy of Science,Canberra, 1983. Wade,F.A., andJ.R. Wilbanks, Geology of MarieByrd and Ellsworth Lands, (ReceivedNovember 7, 1990; in AntarcticGeology and Geophysics,edited by R.J. Adie, pp. 207-214, revisedMay 1, 1991; Universitetesforlaget,Oslo, 1972. acceptedJune 5, 1991.)