Presented on October 6. 1989. Bransfield Strait. Antarctica Abstract

Home , Bransfield Basin, Bransfield Strait, Deception Island, Gledhill (crater), Livingston Island, South Shetland Islands

AN ABSTRACT OF THE THESIS OF

Randall A. Keller for the degree of Master of Science in Oceanography presented on October 6. 1989. Title:Geochemistry of Ouaternary. Rift-Related Volcanism in the Bransfield Strait. Antarctica Redacted for privacy Abstract approved: Martin R. Fisk

The Bransfield Strait is the narrow, late Tertiary to Quaternary marginal basin separating the South Shetland Islands from the northern end of the Antarctic Peninsula magmatic arc. Quaternary volcanism in the strait is tholeiitic to mildly alkaline, and contrasts chemically with the pre-Quaternary caic-alkaline arc volcanism. Geochemical evidence presented here shows that the Quaternary volcanism is related to active rifting in the strait.All of the rift- related volcanoes are chemically related by different extents of partial melting of garnet peridotite variably enriched in alkali and alkali earth elements relative to rare earth elements.This enrichment is characteristic of subduction zones, where fluids from the dehydrating subducted slab concentrate alkalies and alkali earths that can then be mixed into, and partially melted with, the surrounding mantle.The lavas that erupted during the formation of the Bransfield Strait provide evidence that subduction zone processes influence the chemistry of marginal basin volcanism. The Bransfield Strait lavas are chemically similar to published analyses from other marginal basins, especially the Cretaceous marginal basin that is now preserved as the Sarmiento ophiolite of southern Chile.This confirms the interpretation that, prior to obduction, the Sarmiento was a narrow, immature marginal basin analogous to the present day Bransfield Strait. Geochemistry of Quaterilary, Rift-Related Volcanism in the Bransfield Strait, Antarctica

Randall A. Keller

A THESIS submittedto Oregon State University

in partial fulfillment of the requirements for the degree of

Master of Science

Completed October 6, 1989 Commencement June 1990 V1'J J

Redacted for privacy Associate Professor of Oceanography in charge of major

Redacted for privacy

Dean ol/ College of Oceanography

Redacted for privacy

Dean of Sch

Date thesis is presented October 6. 1989

Typed by Randall A. Keller ACKNOWLEDGEMENTS

This thesis would not have been possible without the support and inspiration of my advisor, Martin Fisk.J owe its completion to his helpful ideas and encouragement.I could not have asked for a better boss, nor a more enjoyable traveling companion. Bob Duncan and Anita Grunder made numerous helpful comments on improving the content and style of this thesis. Krzysztof Birkenmajer and Szczepan Porebski provided precious samples and expertise from their fieldwork in Antarctica. Many generous people made the drudgery of laboratory work possible, and occasionally even enjoyable.Thanks to the crews at OSU (Andy, Jim, and Greg) and Cornell (Bill, Mike, and Bill S.). The best part of graduate school is the treasured friendships that carry one through.Jay and Jan kept me sane between battles with Lucifer-the-anti-christ-mass spec.I cherish the good times in Oregon on bikes, skis, and barstools with Kelly, Mark, Susan, Sallie, Dana, Joe, Catherine, Margaret, Dan, Doug, Kevin, Mike, Jim, and John. Brigitte merits special thanks for sustaining me during the final stages of this work. This research was supported by grants DPP85-12395 and DPP86-14022 from the Division of Polar Programs of the National Science Foundation. TABLE OF CONTENTS

INTRODUCTION 1 Tectonic Setting 1 Geological Background 6 Geochronology 1 2

PRESENTwox 16 Introduction 1 6 Methods 42 Results 8

DISCUSSION 54 Chemical variation within the Bransfield Strait samples 5 8 Comparisons with other locations 66

CONCLUSIONS 77

BIBLIOGRAPHY 78

APPENDICES I. Keller and Fisk, l989b 8 6 II. Birkenmajer and Keller, submitted 92 III. Complete Petrographic Descriptions 1 07 IV. Results of Normative Calculations 1 2 1 LIST OF FIGURES

Figure Page 1. Map of interpreted ocean floor magnetic anomalies in the southeastern Pacific. 2 2.Tectonic sketch map of the Bransfield Strait region. 4 3.Location and bathymetry map of South Shetland Islands and Bransfield Strait. 7 4.Silica versus total alkalies plot of published analyses from Deception, Bridgeman, and Penguin Islands. 11 5.Bathymetry and dredge locations in the eastern end of Bransfield Strait. 1 9 6. Geologic sketch map of Deception Island. 20 7.Geologic cross-section of Penguin Island. 2 1 8. Silica versus total alkalies plot of samples analyzed in this study. 22 9.Geologic cross-section of Low Head, King George Island. 43 10.Plot of MgO versus Ti02, Fe203, and Na20/K20. 49 11.Plot of 87Sr/86Sr versus 143Nd/144Nd of samples analyzed in this study. 50 12.Plot of 2O7Pb/2O4Pb versus 208Pb/204Pb of samples analyzed in this study. 5 1 13.Plot of trace element data from Sun (1980). 5 6 14.Plot of trace element data for Andean and Patagonian rocks. 57 15.Chondritenormalized rare earthelement patterns for two MelvillePeak samples. 5 9 16.Chondritenormalized rare earthelement patterns for two PenguinIsland samples. 60 17.Chondritenormalized rare earthelement patterns for two dredgedseamount samples. 6 1 18.Results ofbatch melting modelsof garnet lherzolite and spinel lherzolite (from Lin et al.,1989). 6 2 1 9.Chondrite normalized rareearth element patterns for representative samples from Melville Peak, Penguin Island, and a dredged seamount. 63 20.Chondrite normalized rareearth element patterns for three Deception Island samplesand a dredged sample. 64 21.Plot of trace element datafor Scotia Rise spreading center and South Sandwich Islands volcanic arc. 67 22.Plot of trace element datafor dredged samples. 6 9 23.Plot of trace element datafor Deception Island samples. 7 1 24.Plot of trace element datafor Melville Peak samples. 7 2 25.Plot of trace element datafor Penguin Island samples. 7 3 26.Plot of rare earth elementdata for a dike in the Sarmiento ophiolite. 7 6 27.Sketch map of BransfieldStrait area. 90 28.Silica versus alkalies plotof Bransfield Strait samples. 9 1 29.Volcanoes in the BransfieldStrait area. 100 30.Map of the Melville Peakvolcano. 10 1 3 1.Cross-section of Cape Melville. 1 02 32.Exposure in Melville Peakabove Sherratt Bay. 103 33.Exposure in eastern ridgeof Melville Peak. 1 04 LIST OF TABLES

Table Page 1. Representative analyses of South Shetland Island Arc rocks (data from Smellie, 1983). 9 2.Coordinates and results of dredge stations. 1 7 3.Major and trace element concentrations and loss on ignition data. 23 4.Trace and rare earth element concentrations by isotope dilution. 27 5.Isotopic compositions of samples from the Bransfield Strait.28 6.Microprobe analyses of glasses. 29 7. Microprobe analyses of olivines. 3 0 8.Microprobe analyses of oxides. 33 9.Microprobe analyses of plagioclases. 3 7 10.Microprobe analyses of pyroxenes. 3 9 11.Bransfield Strait dredge stations and results. 8 9 12.Summary of evolution of Melville Peak volcano. 105 13.Potassium-argon ages of Melville Peak samples. 106 Geochemistry of Quaternary, Rift-Related Volcanism in the Bransfield Strait, Antarctica

INTRODUCFION

The Quaternary volcanic rocks of the Bransfield Straitand South Shetland Islands are chemically very different fromthe pre- Quaternary volcanic rocks.Jurassic to late Tertiary volcanic rocks intermediate between caic-alkaline and tholeiitic types make upthe bulk of the South Shetland Islands arc (Smellie et al., 1984).In contrast, the Quaternary volcanic rocks aretholeiitic to mildly alkaline, and are associated with the rifting that has createdthe Bransfield Strait marginal basin (Weaver et al., 1979).There was clearly a major transformation in the style and source ofvolcanism on the South Shetland Islands duringthe late Tertiary, from subduction-related arc volcanism to rift-related marginal basin (back-arc?) volcanism.This thesis presents geochemical evidence that the source of the Quaternary, rift-related volcanism waspartial melting of a garnet bearing mantle source that has beenenriched in alkali and alkali earth elements by subduction zone processes.

TectonicSetting

From Jurassic to Tertiary time the western margin ofthe Antarctic Peninsula was the site of continuous subduction ofoceanic crust of the Aluk Plate (Barker, 1982; Barkerand Daiziel, 1983).At approximately 50 Ma, a segment of the Aluk Ridge spreading center entered the trench at the base (southern end) of theAntarctic Peninsula, causing a cessation of spreading on that ridge segmentand a cessation of subduction in that segmentof the trench (Figure 1). 2

( .J2 S I SOUTH1 /43 I8// /'7/ / .,..- ,.-I8 / \ ' / (27 /20'/J7 \ '!2N c'b\ / "J8/ 9 3 / - --...,/ IS-... II 26 ..20 i

24 -._.___ Ii; " I, / ° 60°S / 8 . 1/! N7 r i'cMyr 2( 2f',,

20 25

,i /5 pf' ?

\ ZO'S c

Figure 1. Map of interpreted ocean floor magnetic anomalies in the southeastern Pacific (from Barker and Dalziel,1983).Numbers are magnetic anomalies, with lessernumbers (younger crust) closer to the margin of the Antarctic Peninsula,showing ridge crest-trench collisions of progressively younger ages northward upthe coast of the Antarctic Peninsula. 3 Because the oceanic crust on the northwest side of the subducted Aluk Ridge was part of the same plate as the Antarctic Peninsula, that part of the western margin of the Antarctic Peninsula became a passive margin (Barker and Daiziel, 1983).Ridge-trench collisions then occurred progressively northward along the margin of the peninsula as other segments of the Aluk Ridge arrived at their respective segments of the trench.This process ended when the ridge segment directly south of the Hero Fracture Zone arrived at the trench. The dates of each of these ridge-trench collisions can be estimated from the age of the youngest crust produced by each consumed ridge segment, assuming that that ridge segment stopped spreading when it reached the trench.Crustal ages interpreted from marine magnetic anomalies suggest that the first ridge-trench collision occurred at 50 Ma, and the northernmost and final ridge- trench collision occurred at approximately 4 Ma (Barker, 1982). Marine magnetic profiles also indicate that the only remaining segment of the Aluk Ridge, the one offshore of the South Shetland Islands between the Hero and Shackleton Fracture Zones (Figure 2), has spread a maximum of 20 km in the past 4 Ma(0.5mm/yr), and appears to now be extinct (Barker and Daiziel, 1983).Cessation of spreading on the surviving segment of the Aluk Ridge may have been synchronous with the ridge-trench collision that occurred just south of the Hero Fracture Zone at 4 Ma (Barker, 1982).These changes in the regional tectonics at 4 Ma resulted in the present plate tectonic configuration. The present plate boundaries in the region of the South Shetland Islands and Bransfield Strait place the South Shetland Island Arc on a separate (Shetland) microplate (Figure 2), bounded on the northwest by the South ShetlandTrench and on the southeast by the Bransfield Rift.The southwest and northeast boundaries of the Shetland microplate are indistinct, but appear to be strike-slip faults corresponding to onshore extensions of the Hero Fracture Zone and the Shackleton Fracture Zone. The South Shetland Trench has been the locus of subduction of Drake plate oceanic crust for at least the past 200 million years. ow / 6dw SOUTH AMERICAN

50S SAI\DWICH SCOTIA

DRAKE

1 9OW SHETLAND . I . 'a / I 60S / ANTA IC / / O°\ c9) / 9 / " 70S 0 708 0W 60W 30W \ '_, /, r ._. / /

Figure 2. Tectonic sketch map (upper figure) of Bransfield Strait region showing names of plates and locations ofspreading centers (double lines) and subduction zones(with solid triangles on overriding plate).Lower figure shows location of Bransfield Strait and major bathymetric and geographic features(bathymetry contoured in km).Small square in the Bransfield Strait (lower figure) shows location of Figure 5. 5 Subduction probably slowed considerably at approximately 4 Ma in response to the cessation of spreading on the Aluk Ridge(Barker and Dalziel, 1983).Waveform inversion analyses of teleseismic data from earthquakes in the South Shetland Islands region yielded two focal mechanism solutions with depths (35 km and 55 km) and principle stress directions compatible with tensional stresses in a downgoing slab (Pelayo and Wiens, 1989).Curiously, there are no moderate sized events in the trench itself that would indicate the expected shallow thrust zone, nor are there the intermediate and deep events normally associated with a downgoing slab (Pelayo and Wiens, 1989). So if subduction is continuing beneath the South Shetland Islands, it is virtually aseismic and probably relatively slow. Four million years ago is also the time proposed for the onset of extension and subsidence in the Bransfield Strait (e.g., Barker and Dalziel, 1983), but geologic evidence for the timing of the onset of extensional tectonics is sparse.The occurrence of Pliocene(?) volcanism along normal faults bounding the northern margin of the Bransfield Strait (i.e., on Penguin Island and King George Island) has been interpreted as evidence for the onset of diffuse extension and faulting in the strait (Barton, 1965).The proposed Pliocene age of these volcanic rocks has not been confirmed by radiometric dating, but it is reasonable to expect a period of diffuse extension prior to the present seafloor spreading in the basin. Roach (1978) used marine magnetic data to model the seafloor spreading history of the Bransfield Strait, and concluded that only small amounts of reversely magnetized crust at the edges of the basin are necessary to match the data.Roach then assumed a constant spreading rate in the strait, and concluded that seafloor spreading replaced diffuse extension at approximately 1.3 Ma (Roach, in Barker and Dalziel, 1983).However, the brevity, and consequent ambiguity, of the marine magnetic record in the Bransfield Strait have led to the general acceptance of a more conservative estimate of 1-2 Ma for the onset of seafloor spreading in the strait. Evidence for active rifting in the Bransfield Strait includes a linear chain of seamounts down the middle of the deepest part of the strait, as well as normal fault focal mechanism solutions for earthquakes located in line with this seamount chain (Forsyth, 1975; Pelayo and Wiens, 1989).Dredging and coring on and around the seamounts show that they are volcanically (Fisk, in press) and geothermally (Suess et al., 1987) active.

GeologicalBackground

B asement

The South Shetland Islands (Figure 2) are part of a continental magmatic arc built upon a basement of schists and gneisses of uncertain age (Smellie et al., 1984).Basement rocks are exposed only on the southwestern and northeastern ends of the South Shetland Island chain (Smith Island and Elephant Island, respectively; Figure 3).Both Smith and Elephant Islands consist entirely of metamorphic basement, so the relationship between the basement and the volcanic rocks that make up the other islands has never been observed.Nevertheless, this same metamorphic basement is assumed to lie beneath the other islands in the South Shetland chain.

Jurassic to Tertiary Arc Volcanism

The bulk of the South Shetland Islands was created by extensive Jurassic to Tertairy arc magmatism associated with subduction of Pacific Ocean crust (Thomson and Pankhurst, 1983). The arc magmatism was generally intermediate between island arc tholeiite and calc-alkaline types (Smellie et al.,1984). 7

Figure 3. Location and bathymetry map of South Shetland Islands and Bransfield Strait (modified from Ashcroft, 1972).Depths are contoured in meters below sealevel.Dashed lines with question marks are 'major faults' of Ashcroft (1972). Extrusive rocks are ubiquitous and mostly low-K, high-alumina basalts, basaltic andesites, and low-silica andesites (e.g., analyses 1- 7 in Table 1).Dacites and rhyolites are rare.Intrusive rocks are less common than the extrusive rocks, and are mainly gabbros, tonalites, and granodiorites (e.g., analyses 8-12 in Table 1) that arechemically very similar to the extrusive rocks (Smellie, 1983). Many of the South Shetland arc magmas are subalkaline, with Fe/Mg and Na20/K20 similar to island arc tholeiites (Smellie, 1983). All of the magmas have slight iron enrichment trends in the basaltic compositions, but the more siliceous compositions have alkaline enrichment trends characteristic of caic-alkaline magmas. Overall, chemical variation within the Jurassic to late Tertiary South Shetland magmas appears attributable to slight differences in extent of partial melting and/or fractionation of olivine, plagioclase, clinopyroxene, titaniferous magnetite, and small amounts of apatite, all of which exist as phenocrysts in the samples (Smellie, 1983). There is little variation in ratios of any element compared to Zr, and Smellie (1983) suggested that this is evidence for a relatively homogeneous source for the South Shetland arc magmas.

Quaternary Volcanism

Quaternary volcanic rocks are found only on Bridgeman Island, King George Island, Penguin Island, Deception Island, Livingston Island, and Greenwich Island (Weaver et al., 1979; Smellie, 1983; Smellie et al., 1984; Godoy et al., 1987). Bridgeman, Penguin, and Deception Islands are volcanoes in the Bransfield Strait built entirely of Quaternary pyroclastic flows and lavas transitional between caic-alkaline and mid-ocean ridge types (Weaver et al.1979).The geochemistry of these islands, along with their locations at the margins of a 2 km deep rift basin, suggest that they are associated with the rifting that is creating the Bransfield Strait marginal basin (Weaver et al.,1979). Table1. RepresentativeanalysesofSouthShetlandIslandslavas andinirusions

Sample 1 2 3 4 5 f 7 8 9 10 11 12 Si02 48.3 78.0 49.8 50.7 53.1 67.5 53.7 48.2 54.1 58.2 64.0 71.4 Ti02 1.07 0.11 0.67 0.87 0.97 0.64 0.86 0.39 0.75 0.77 0.54 0.39 A1203 15.5 12.4 17.1 19.9 17.2 16.2 24.0 18.7 17.3 16.9 15.2 14.4 FeO3 2.36 0.06 2.46 2.74 2.46 0.74 0.82 2.08 2.32 2.24 1.62 0.86 FeO 7.85 0.14 6.14 6.85 6.15 1.86 2.74 5.19 5.80 5.61 4.05 2.15 0.09 MaO 0.20 0.01 - 0.17 0.16 0.17 0.13 0.13 0.14 0.21 0.12 MgO 9.6 0.2 9.8 3.9 5.7 0.7 1.8 8.7 6.4 4.1 1.9 1.1 CaO 10.5 03 1139 1137 8.00 2.11 10.35 13.9 8.39 7.72 4.44 2.40 Na20 2.50 2.26 2.24 3.38 3.36 5.12 4.16 8.7 3.30 3.33 4.12 4.60 K20 0.24 5.48 0.34 0.32 1.00 3.81 0.66 0.20 1.41 2.18 1.82 2.52 P205 0.22 0.01 0.12 0.12 0.22 0.14 0.20 0.04 0.17 0.21 0.12 0.05 Total 98.33 98.97100.06100.52 98.32 98.99 99.42 99.72100.08101.47 97.93 99.96

71 665 157 149 333 860 246 262 407 319 318 & 256 - 360 - 20 3 19 150 20 - - - NI 84 6 142 - 16 - 12 '10 11 - - - Rb 2 121 - - 14 87 10 - 34 38 54 61 Sr 350 37 670 609 563 315 10 480 597 719 200 200

Zn - - 54 6? 66 94 9) 5) ) 86 30 46 Y 18 15 10 13 23 25 23 7 19 24 36 27 Zr (1) 102 58 57 93 39(1 129 21 127 253 140 168 - Nb 3 9 - - 9 1 - - 3 5 La 4 19 - - 16 48 10 - 19 29 19 18 Ce 16 35 16 17 20 75 21 - 33 52 37 34 Pb 2 9 - - - 10 3 - - - - 35 1 13 - - - 18 2 - - - - 10 - - 20 24 21 19 24 21 24 25 19 15

All data are from Smellie (1983) Major elements given in weight percent, trace elements in ppm. LAVAS: 1-2, Livingston Island; 3, Robert Island; 4-7, King George Island. INTRUSIONS: 8, Half Moon Islnnd; 9-10, King George Island; 11-12, Low Island. 10 Deception Island lavas range from olivine basalts to rhyodacites, with CeN/YbN ratios of approximately 2, and universally high Na20/K20.The rhyodacites are high in incompatible elements such as Zr, Y, and the rare earth elements, and have negative Eu anomalies suggesting plagioclase fractionation.Overall, incompatible elements increase and Sr decreases with increasing silica, suggesting fractional crystallization as the dominant process (Weaver et al., 1979).The most primitive basalt reported from Deception Island has only 6% MgO, 114 ppm Cr, and 35 ppm Ni.This suggests that even the most primitive Deception Island lavas have undergone considerable fractionation of olivine and clinopyroxene (Weaver et al., 1979). Bridgeman Island is composed of subalkaline basaltic andesites similar to those on Deception Island (Figure 4).The Bridgeman lavas show little compositional variation, and have low abundances of incompatible elements such as Ti, P, Sr, Zr, Ce, and La at the same time as having low abundances of compatible elements such as Cr and Ni.Weaver et al. (1979) concluded that this requires a larger degree of partial melting of the Bridgeman Island source, possibly under more hydrous conditions. Penguin Island lavas are mildly alkaline olivine basalts with high MgO contents, suggesting that they are closer to primary mantle melts (Weaver et al., 1979).The most primitive Penguin lavas are high in Cr and Ni, but these elements decrease rapidly with increasing concentrations of incompatible elements.This characteristic, as well as their small compositional variation (Figure 4), is consistent with fractional crystallization of olivine, clinopyroxene, and chromite.Compared to the Deception Island basalts, the Penguin Island basalts are higher in K, Rb, Ba, Sr, Ti, P. Na, Ni, and Cr at a given Zr content, and have CeN/YbN ratiosof approximately 4. Weaver et al. (1979) interpreted this as evidence for garnet in the Penguin Island source, and model the Penguin Island lavas as being created by <5% melting of a garnet peridotite source. Rare outcrops of Quaternary lavas also occur on three of the older islands in the South Shetland Islands arc.These are: Melville Peak on King George Island, Gleaner Heights on LivingstonIsland, 11

No20 K20 K 8 .1.

K 6

r±i Deception Is

x Bridgemon Is Penguin Is K Si02 wt %

I I I SO 55 60 65 70

Figure 4. Silica versus total alkalies plot of previously analyzed samples from Deception, Bridgeman, andPenguin Islands (from Weaver et al.,1979). Line LB divides alkaline and subalkaline fields of Irvine and Baragar (1971).K lines delimit high alumina basalt field of Kuno (1966). 12 and Mount Plymouth on Greenwich Island.Melville Peak lies on the northern margin of the Bransfield Strait along a known fault line, and has also been attributed to the rifting in the Bransfield Strait (Keller and Fisk, 1989b; see Appendix A).The only published analysis for a Melville Peak lava is a tholeiitic basalt with 52% Si02 and 11% MgO described as having large (>7mm) phenocrysts of clinopyroxene and olivine (Godoy et al., 1987). Lavas from Gleaner Heights on Livingston Island and Mount Plymouth on Greenwich Island are mildly alkaline olivine basalts. They are broadly similar to the Penguin Island basalts, but have lower Ba, Sr, Zr, Ce, and Ga, and higher Ca and La.Also, their CeN/YbN ratios (1-4) are lower than the Penguin lavas, suggesting a shallower (garnet free) source (Smellie et al., 1984).The occurrence of these Quaternary basalts on Livingston and Greenwich Islands has been interpreted as due to 'intra-plate tensional tectonics' rather than subduction (Smellie et al.,1984). The only other known location of Quaternary lavas in the Bransfield Strait area is a line of seamounts down the middle of the deepest part of the strait.This seamount chain was identified as a possible rift axis by Ashcroft (1972) when he first suggested that the South Shetland Islands were rifting away from the Antarctic Peninsula.Dredged samples from two of these seamounts yielded late Pleistocene K-Ar dates (Fisk, in press).

G eochronology

Basement

The oldest rocks on the South Shetland Islands are from the blueschist facies metamorphic complexes of Smith, Elephant, and Clarence Islands (Smeilie, 1984).Outcrops are rare, and no radiometric ages are available for these rocks, but correlation to 13 lithologically similar metamorphic complexes on the South Orkney Islands and northern Antarctic Peninsula are appropriate (Hoskins, 1963; Thomson, 1968).Biotite/hornblende separates from the mica schists and amphibolites of the South Orkney Islands yield K-Ar ages slightly less than 200 Ma (Tanner et al., 1982).Whole rock Rb-Sr ages of a gneiss from Marguerite Bay (Antarctic Peninsula) fall between 170 and 200 Ma.These ages coincide with the accepted age of Gondwanan metamorphism (Gledhill et al.,1982).

Jurassic to Tertiary Arc Volcanism

The oldest volcanic rocks on the South Shetland Islands are found on the western end of the chain on Low, Snow, and Livingston Islands.Volcaniclastic rocks sampled there have been dated paleontologically as Upper Jurassic-Lower Cretaceous (Smellie et al., 1980). Also, a granodiorite intruding the volcaniclastic beds yielded a K-Ar age of 121 Ma (Smellie et al., 1984).Further radiometric evidence suggests relatively continuous volcanism in the South Shetland Islands since the Late Jurassic, although activity was apparently not present everywhere at all times.Pankhurst and Smellie (1983) used extensive K-Ar age data to argue that arc volcanism migrated slowly eastward along the island chain, reaching the western end of King George Island in the Paleocene, and the eastern end of King George Island in the Oligocene.Arc volcanism apparently ceased, however, in the Early Miocene, as no arc activity younger than 20 Ma has been found (Birkenmajer et al., 1988).It is not clear, however, if this is an artifact of limited outcrop.It must be kept in mind that over 90% of the South Shetland Islands are covered by permanent ice; exposures are usually limited to isolated outcrops and seacliffs. 14 Quaternary Volcanism

Quaternary volcanic activity in the South Shetland Islands is much less extensive than the arc activity that preceeded it. Locations of recent activity are: Deception Island, Gleaner Heights on Livingston Island, Mount Plymouth on Greenwich Island, Penguin Island, Melville Peak on King George Island, and Bridgeman Island. Livingston and Greenwich Islands.Basalt samples from Gleaner Heights on Livingston Island and Mount Plymouth on Greenwich Island had undetectable amounts of radiogenic argon, and are considered to be less than 200,000 years old (Pankhurst and Smellie, 1983). Deception Island.Volcanic activity has been recorded on Deception Island many times since it was first mapped in 1829 (Kendall, 1831), with the most recent eruption having been in August of 1970 (Roobol, 1973).Dates of pre-historic eruptions of the volcano are poorly constrained, but Smellie (1988) suggested that the oldest activity maybe less than 750,000 years old. Penguin Island.The small size and unglaciated terrain of Penguin Island attest to its brief and recent history.Basalts at the base of the volcano are intercalated with Holocene beach sands. Lichenometric dating of some of the volcanic features on the island suggests that at least 4 eruptions have occurred in the past 300 years (Birkenmajer,1980). Melville Peak on King George Island.Birkenmajer (1981) outlined a two phase history for the Melville Peak volcano: a late Pleistocene phase, followed by a Holocene phase correlatable to the Deacon Peak formation of Penguin Island.The first phase has since been K-Ar dated as middle to late Pleistocene (Birkenmajer and Keller, submitted, see Appendix B).This supports Birkenmajer's original conclusion, and suggests that the second phase is indeed of Holocene age. Bridgeman Island.Bridgeman Island resembles the lower units of Deception Island, but has suffered extensive glacial erosion and weathering (Weaver at al., 1979).No radiometric ages are available 15 from the island.Mid-nineteenth century sailors reported fumarole activity on Bridgeman, but González-Ferrán and Katsui (1970) visited the island and could find no evidence for such recent activity.They suggested that the location of the fumarole activity was misreported, and may in fact have been on Penguin Island. 16

PRESENT WORK

Introduction

One of the primary objectives of this project was to examine the nature of volcanism during the early stages of the formation of a marginal basin.To this end, we obtained samples from all of the Pliocene or younger volcanoes in the Bransfield Strait, with the exception of Bridgeman Island.The sampled subaerial volcanoes are Deception Island and Penguin Island, and Melville Peak on King George Island.Also, samples were dredged from two seamounts in the Bransfield Strait.Locations and brief descriptions of all of these samples are given below. Previous studies of the Bransfield Strait indicated that a linear chain of seamounts bisect the deepest part of the strait, and Ashcroft (1972) proposed these seamounts as the axis of rifting in the Bransfield Strait.A more recent geophysical survey associated these seamounts with a prominant positive magnetic anomaly characteristic of recent volcanism (Parra et al., 1984).Also, a previous cruise to the area discovered hydrocarbons within the sediments of the strait that were surprisingly mature for such a young basin, and anomalously high heat flow was suspected as a cause of this rapid maturation (Whiticar et al., 1985; Suess et al., 1987).For these reasons, submarine volcanism in the strait was stronglysuspected. To test this suspicion, five prominent features of the floor of the strait were dredged from theR/V Polarsternin November and December of 1985 (Table 2).A seamount between Low and Deception Islands yielded altered and eroded basalts that appeared to be glacial erratics, and these samples were not analyzed nor included in this study.The same is true for samples dredged from a seamount on the southern side of the strait near the Antarctic Peninsula margin, and a seamount on the northern edge of the strait near the shelf of King George Island. 17

Table 2. Bransfield Strait Dredge Stations and Results. ANT IV/2

Station Latitude Longitude Number S W Area Results 285 63°9.8' 61°45.5' hill in Low Island Basin glacial boulders 286 63°11.8' 61°13.0' hill in Low Island Basin no sample 290 62°15.2' 58°10.7' King George Island shelf glacial boulders 292 62°12.3' 57°30.3' seamount, King George Basin pillow basalts 297 62°15.4' 57°24.5' floor, King George Basin fresh basalts 300 62°14.1' 57°23.5' seamount, King George Basin fresh basalt 307 62°18.0' 57°32.8' hill, King George Basin glacial boulders 309 62°13.4' 57°28.6' floor, King George Basin fresh basalt 310 62°12.9' 57°28.8' seamount. King George Basin nillow basalts Five dredge hauls recovered samples from a pair of seamounts in the deepest part of the strait just southeast of Penguin Island. These are the samples used in this study.These two seamounts are part of a linear chain of seamounts in line with Deception and Bridgeman Islands that is believed to be the current axis of rifting of the Bransfield Strait.Dredges 292 and 310 were taken on the western of the two seamounts at a depth of approximately 1900 m (Figure 5).Dredges 297, 300, and 309 were taken from the eastern of the two seamounts at depths ranging from 1700 to 2000 m.Both dredge targets yielded glassy, highly vesicular pillow lavas and more massive flows of fresh basalt and basaltic andesite, mostly classifiable as olivine tholeiites. During a brief shore visit to Deception Island, 13 samples were collected near the north shore of Whalers Bay (Figure 6) by Martin Fisk.The samples were collected at the outwash of glacial streams, so their field relationships are not known, but they appeared representative of the local lithology.As with most of the exposed rock on Deception Island, these samples are from the 'post-caldera' phase of eruptive activity on the island (Smellie et al., 1984).Their compositions range from subalkaline basalt to dacite and rhyodacite, and comparison to published analyses of Deception Island volcanics (Weaver et al., 1979) shows that our samples represent the range in compositions found on the island (Figure 4). Five samples from Penguin Island were donated by Prof. K. Birkenmajer of the Polish Academy of Sciences.Three of the samples (A172, A260, and A263) are from the basal Marr Point Formation (Figure 7).The other two samples are from a dyke (A174) and a plug (A261) of the Deacon Peak Formation in the principal cone (Birkenmajer, 1982).Penguin Island is only about 1.5 km across, and consists of a single stratocone with little chemical variation.The compositions of our samples are similar to published analyses from Penguin Island (Weaver et al., 1979; Figure 4), that is, mildly alkaline olivine basalts (Table 3 and Figure 8). Four samples from Melville Peak on King George Island were also donated by Prof. Birkenmajer (locations are given in Birkenmajer and Keller, submitted; see Appendix B).Melville Peak is 19

57°18'W 57°40' 38' 36' 34' 32' 30' 28' 26' 24' 22' 20'

12'S

21'

Figure 5. Bathymetry and dredge locations in eastern end of Bransfield Strait.Dredge locations are numbered in bold. Bathymetry data were collected on the Antarktis IV cruise of the R/V Polarstern, and are contoured in meters. 4i

60 30

Kendot Terioce 'Dp_ Mocorne, not nomevd \ V°°'

-6255 S\permonene.ce \ Teletnn Ridge -: M0 Cone 4'- :-.

k.Oont Pond

, \ \ w.oaaey.ie

8..con I 1 ) RI Fotn.rcde - HC.d Bdy I PORT FOSTER I)

c_a

Vapour Col Buy Sooth Eant - Cuthedrul * Ce60n Point -6300S Enmance M ' P0.-it K.rkwood permanent Ce

k5 Inlund km

R.tr,i .Octflt.c1.6.0.0.. .00ron's-n..Liii]

C.?. .01.! 0*000,10 ST-00L0CRn ORPOSITS

F*go,. .rnø.dStmntnk.n Icon, and.II

£010 lu!t COO. d.00s,Is CALOtCA COLLAPSE

000 COSOKAhn....On'. 0.000(0

poE-CALOEMnceoslis

PemoLalo lock.(.000?.....I.O6

Figure 6. Geologic sketch map of Deception Island(from Smellie, 1984) showing general geology of the island andlocations of samples analyzed in this study (Dl, D2, and D7through D13). WNW ESEjSW NE m Deacon Peak m - 300 ,- 3oa PRINCIPAl. CONE CENTRAL - 200 CONE W 200

EL. CRATER SUPPOSED (lake) -100 SUBMARINE 100 CRATER g LPETR30 (

0 100 O 300rfl I I -

Figure 7. Geologic cross-section of Penguin Island (from Birkenmajer, 1982) showing locations of samples analyzed in this study (A172, A174, A260, A261, and A263).Circled numbers refer to the following structures and formations: 1-2, Marr Point Formation; 3-6, Deacon Peak Formation; 7, Petrel Crater Formation; 8, collapse structures; 9, crater fill and present beach. 10

...... - a a. W Seamount a

I - o E Seamount trachyte 8 +Penguin _-- ...... -- I '.' .-- I 0 Melville

--- I I + C Low Head --

--- I 0 .--.. I C1 6 £ Deception I z(5 I

+ ... 1. I I ..... * I I I ....?

I I .-..*.... I i basaltic . basalt andesite andesite dacite rhyolite

I I I. I I 45 50 55 60 65 70 wt% S102

Figure 8. Silica versus total alkalies plot of samples analyzed in this study.Half-tone line divides alkaline (above line) and subalkaline fields of Irvine and Baragar (1971).

t) 23 Table 3.Whole Rock Major* and Trace** Element Chemistry and Loss on Ignition***

Sample BS292-13BS292-17 BS292-18BS292-24 BS292-53B S 292-54 Meas. Acc. Si02 48.56 47.52 48.34 49.17 47.95 48.02 49.5549.90 Ti02 1.14 1.13 1.15 1.33 1.13 1.16 2.79 2.69 A1203 14.47 14.07 14.13 15.79 14.11 14.31 13.5413.85 Fe203 9.63 9.71 9.79 9.89 10.00 9.62 12.2412.23 MnO 0.16 0.16 0.16 0.16 0.22 0.16 0.17 0.17 MgO 12.30 12.59 12.78 8.54 12.34 11.88 7.18 7.31 QO 10.38 10.17 10.21 10.86 10.27 10.40 11.3211.33 Na20 3.24 2.95 2.99 3.47 2.93 3.01 2.31 2.29 K20 0.42 0.40 0.45 0.43 0.43 0.45 0.49 0.54 P205 0.17 0.16 0.17 0.17 0.22 0.16 0.27 0.28 Total 100.46 98.85 100.16 99.79 99.61 99.16 99.86100.59 LOl 0.08 0.12 0.00 0.15 0.05 0.15

Ba 194 145 168 199 198 188 cr 262 262 243 244 234 233 Cu 76 77 70 78 66 78 Li 5 5 5 5 5 5 Ni 309 332 323 133 312 303 Rb 10 11 11 10 9 12 Sr 270 268 270 281 288 275 Zn 64 64 63 66 64 63

WesternSeamount-Dredge310 StandardGL-0 StandardJB-la

SampleBS3lO-1BS31O-2 BS31O-7BS31O-13 Meas. Acc. Meas. Acc. Si02 48.77 48.70 49.08 48.15 52.85 52.22 Ti02 1.53 1.49 1.54 1.46 0.06 0.07 A1203 15.60 15.29 16.05 14.72 7.94 7.75 Fe203 10.43 10.60 10.21 10.81 19.79 20.11 MnO 0.16 0.17 0.16 0.18 0.02 0.01 MgO 7.49 8.60 6.75 9.37 4.51 4.58 CD 10.55 10.48 10.74 10.21 1.00 0.98 Na20 3.57 3.50 3.61 3.12 0.00 0.04 K20 0.46 0.46 0.55 0.42 8.34 8.16 '205 0.20 0.17 0.18 0.17 0.34 0.38 Total 98.75 99.46 98.88 98.61 94.85 94.30 LOl 0.19 0.01 0.45 0.10 Ba 193 173 187 172 480 497 Q 227 252 190 293 352 415 Cu 72 76 81 74 56 56 Li 5 5 5 5 12 12 NI 118 155 100 184 136 140 Rb 9 9 9 9 40 41 Sr 285 263 288 264 467 443 Zn 72 73 72 71 78 82 24

SampleBS297- 1 BS297-2 BS300-2 BS300-1OBS300-16BS300-19BS300-21 BS309-1 Si02 50.98 51.12 52.95 52.27 51.85 52.53 51.64 53.79 Ti02 1.05 1.07 1.74 1.41 1.40 1.73 1.35 1.75 A1203 16.71 16.78 15.46 15.90 15.80 15.30 15.79 15.54 Fe203 7.85 8.27 11.55 10.15 10.23 11.46 10.05 11.32 MnO 0.14 0.14 0.18 0.17 0.17 0.17 0.17 0.17 MgO 6.98 7.35 3.88 5.00 4.90 3.87 5.03 3.57 00 11.20 11.13 7.85 9.28 9.23 7.77 9.35 7.39 Na20 3.10 3.07 4.54 4.04 3.95 4.31 3.89 4.56 K20 0.38 0.36 0.58 0.58 0.52 0.75 0.45 0.90 P205 0.12 0.12 0.23 0.17 0.18 0.23 0.17 0.22 Total 98.49 99.42 98.96 98.96 98.23 98.13 97.87 99.20 Ba 116 116 192 175 142 177 161 206 Cr 219 203 6 14 14 6 14 4 Cu 59 59 65 71 71 64 69 63 Li 5 4 8 5 6 5 5 8 Ni 100 100 13 24 23 14 26 12 Rb 6 6 11 9 9 11 8 15 Sr 300 294 322 299 299 321 304 366 Zn 56 58 87 75 76 91 73 89

Sample A172 A174 A260 A261 A263 MeasuredAccepted Sf02 47.75 48.59 48.00 48.00 47.47 Ti02 1.30 1.31 1.11 1.26 1.17 A120316.18 17.03 15.72 15.99 15.42 Fe203 9.84 9.71 9.61 9.71 9.86 MnO 0.16 0.16 0.16 0.16 0.16 MgO 8.50 8.10 10.07 9.62 10.67 00 10.11 10.17 10.19 10.04 9.76 Na20 3.61 3.92 3.34 3.67 3.41 K20 0.63 0.64 0.48 0.60 0.56 P205 0.50 0.29 0.23 0.27 0.30 Total98.57 99.92 98.91 99.33 98.77 LOT 0.10 -0.01 -0.32 -0.11 -0.23

Ba 233 230 222 230 191 283 251 Cr 237 206 356 321 339 66 60 Cu 88 157 114 75 121 199 198 Li 6 7 6 6 6 8 7 Ni 142 122 201 181 237 41 39 Rb 6 6 6 5 6 13 13 Sr 664 705 560 643 593 427 395 Zn 73 73 71 71 74 100 106 25

Sample A372 A375 A405 A406 LH1 Measured Accepted Si02 50.50 52.48 51.32 50.74 47.08 Ti02 0.91 0.97 0.92 0.89 0.77 A120315.36 17.06 16.86 16.67 15.32 Fe203 8.10 7.92 8.04 7.97 9.65 MnO 0.13 0.13 0.14 0.13 0.13 MgO 10.83 7.36 7.02 6.94 8.46 9.85 11.49 11.40 11.39 13.65 Na20 2.94 3.30 3.04 3.00 2.56 K20 0.72 0.86 0.91 0.91 0.24 P205 0.11 0.13 0.13 0.13 0.12 Total99.46 101.70 99.78 98.77 97.99 LOl -0.05 0.01 0.14 -0.14 5.26 Ba 166 230 346 294 142 331 318 cr 358 239 202 203 362 72 81 Cu 86 96 88 79 44 46 45 Li 5 6 6 7 8 14 14 NI 218 61 49 50 285 37 36 Rb 12 14 15 15 3 35 36 Sr 533 607 663 669 827 288 294 Zn 53 54 55 55 56 66 68

DeceptionIsland

Sample Dl D2 D7 D8 D9 D1O D12 D13 Si0264.16 61.83 67.44 66.51 64.52 50.83 68.58 79.00 Ti02 0.99 1.22 0.66 0.70 0.56 1.70 0.59 0.12 A120315.40 15.42 14.63 14.57 13.98 16.52 14.80 11.03 Fe2036.20 7.19 5.03 5.24 4.87 9.86 4.76 0.79 MnO 0.17 0.17 0.15 0.15 0.15 0.16 0.15 0.04 MgO 1.32 1.57 0.61 0.71 0.52 5.14 0.56 0.58 00 3.11 3.76 1.95 2.11 1.67 9.54 1.76 1.11 Na20 7.45 7.00 7.49 7.34 7.22 4.57 7.73 4.51 K20 1.59 1.37 1.87 1.82 1.82 0.48 1.92 2.71 P205 0.25 0.43 0.14 0.14 0.12 0.25 0.12 0.02 Total 100.63 99.96 99.97 99.28 95.42 99.05100.97 99.91 LOl 0.65 0.17 -0.01 0.00 0.09 -0.58 -0.10 1.26

Ba 203 222 242 244 250 140 253 536 Cr 6 3 2 2 2 105 3 4 Cu 7 15 7 8 5 53 6 1 Li 15 22 29 27 26 9 27 6 Ni 10 9 9 8 7 34 9 11 Rb 22 19 29 28 31 5 29 72 Sr 186 289 140 148 132 447 131 45 Zn 87 94 92 93 89_ 74 92 16 * Major element chemistry measured by XRF and expressed in weight percent oxide. **Trace element chemistry measured by AAS and expressed in ppm. ***After one hour in 1000°C oven.Negative values mean weight gained. the only location of proven post-Miocene volcanic activity on King George Island (Birkenmajer, 1981; and Birkenmajer and Keller, submitted; see Appendix B).The samples are olivine tholeiites (Table 3 and Figure 8) that yielded middle to late Pleistocene K-Ar ages (Birkenmajer and Keller, submitted; see Appendix B). An additional sample from King George Island was donated by Dr. S. Porebski of the Polish Academy of Sciences.This is a brecciated basalt from the Low Head Member of the Polonez Cove Formation at Low Head on the southern coast of King George Island (LH-1 in Figure 9).Stratigraphic evidence suggests an Early Oligocene age for the Low Head member.This basalt is used here to represent the chemistry of the basaltic eruptions on King George Island prior to the rift activity associated with the opening of the Bransfield Strait. A total of 51 samples were prepared for major element analyses: 32 dredged samples, 5 Penguin Island samples, 5 King George Island samples (4 from Melville Peak and 1 from Low Head), and 9 Deception Island samples.The results of major element analyses suggested some duplicate sampling in the dredge hauls, so some of the samples were not analyzed for trace elements or radiogenic isotopes.Results for 36 samples analyzed for major and trace elements are in Table 3.These results, as well as the results of analyses for rare earth elements (Table 4), radiogenic isotope ratios (Table 5), and phase chemistry (Tables 6-10) are discussed in their respective sections below.Complete petrographic descriptions are in Appendix C. 27

flI*t1h!Is1!IL1 ti.I!iiIW!I1 _I[5I Sample BS292-17 BS292-24BS31O-7 BS297-1 BS300-2 BS309-1 Meas. Acc. K 3303 3542 - 3161 4706 - 1407314200 Rb 11.29 9.21 8.88 6.3 11.7 17.0 55.6246.6 0.308 0.34 0.359 0.2 0.5 0.6 0.95 0.95 Sr 269.4 263.90 273.6 280.7 320.2 357.0 329.7330 Ba 137.9 110.30 125,9 64.8 135.3 180.3 674.2675 La 7.3 6.10 - 5.4 - 15.2 - 26 Ce 18.9 16.50 - 14.9 30.3 37.4 5.8 53.9 Nd 12.9 12.10 19.4 10.7 20.3 24.8 28.5 29 Sm 3.43 3.39 4.08 3.0 5.3 4.5 6.53 6.6 Eu 1.23 1.21 1.25 1.1 1.7 1.8 1.93 1,94 Gd 3.37 - - 6.0 5.8 - 6.6 Dy 3.54 4.14 4.55 - 6.2 6.3 6.43 6.3 Er 2.07 2.40 2.7 - 3.8 3.8 3.6 3.59 Yb 2.06 2.16 2.49 - 3.7 3.8 - 3.36 Lu ------0.55

KingGeorgeIsland PenguinIsland MelvillePeak Low Head DeceptionIsland Sample A174 A261 A372 A406 LH1 D2 D9 D10 K - 4951 5891 - 1198 - 15443 4016 Rb 6.3 5.3 11.58 15.80 3.34 20.3234.42 4.99 Cs 0.1 0.1 0.35 0.31 1.75 1.21 0.74 0.17 Sr 639.5 581.9 499.90622.40 373.20 274.40128.00417.80 Ba 172.6 - 122.10248.70 - 207.40262.00 77.40 La - 9.7 9.10 15.30 - 21.90 23.70 10.00 Ce 24.3 23.6 22.10 34.70 - 54.30 60.10 25.50 Nd 14.4 13.9 13.50 19.20 - 33.60 34.00 17.10 Sm 3.2 3.1 3.32 4.01 - 8.20 8.27 4.75 Eu 1.1 1.2 1.21 1.23 - 2.31 2.04 1.82 Gd - 3.0 - - - 8.18 - 5.11 Dy 2.8 2.7 2.84 3.40 - 9.34 9.90 5.41 Er 1.5 1.4 1.81 1.95 - 5.80 6.39 3.21 Yb 1.3 1.2 1.82 1.80 - 5.62 6.65 2.95 *Trace and rare earth element chemistry measured by isotope dilution mass spectrometry and expressed in ppm. Hyphen (-) denotes "not analyzed" or "unsuccessful analysis." Western Seamount EasternSeamount Sample BS292-17 BS292-18 BS292-24 BS3IO-7BS297-1BS300-2BS300-21 BS309-1 87Sr/86Sr 0.703420 0.7034 16 0.703460 0.7035060.7030360.7033700.703256 0.703246 ± 0.000013 0.000016 0.000014 0.000011 0.000010 0.000020 0.000010 0.000029

143NW144 0.512977 0.512951 0.512954 0.512945 0.513010 0.512958 0.512958 0.512938 ± 0.0000100.0000300.0000070.0000050.0000070.0000070.0000100.000004 eNd 6.6 6.1 6.1 5.9 7.2 6.2 6.2 5.8

206Pb/204Pb18.757 - 18.742 - 18.739 18.745 - - ± 0.000 - 0.001 - 0.002 0.001 - -

207Pb/204Pb15.613 - 15.602 - 15.624 15.598 - - ± 0.000 - 0.001 - 0.003 0.001 -

208Pb/204Pb38.556 38.502 - 38.546 38.505 - - ± 0.001 - 0.002 - 0.010 0.002 -

KingGeorge Island Penguin Island MelvillePeak Low Head DeceptionIsland Sample A174 A261 A372 A406 LH1 D2 D9 D10

87Sr/86Sr 0.7038550.7038040.7031700.7033900.7035930.7034330.7035300.703530 ± 0.0000140.0000140.0000120.0000200.0000100.0000120.0000020.000020

143Nd/144NJ 0.5129030.5129050.5129750.5129250.5129550.5130060.5129940.513019 ± 0.0000050.0000060.0000040.0000150.0000080.0000040.0000050.000006 eNd 5.1 5.2 6.5 5.5 6.1 7.1 6.9 7.4

206pb/204pb18.711 18.724 18.691 - 18.754 18.742 - 18.753 ± 0.001 0.000 0.001 - 0.001 0.002 - 0.002

207pb/204pb15.594 15.608 15.600 - 15.603 15.624 - 15.614 ± 0.001 0.000 0.001 - 0.001 0.002 - 0.001

208Pb/204Pb38.475 38.522 38.44 1 - 38.539 38.539 - 38.538 0.002 0.002 0.003 - 0.003 0.009 - 0.003 Notes: Hyphen (-) denotes "not analyzed." Uncertainties are machine errors given as one standard error. Western Seamount Sample BS292-17 B5292-17 BS292-17 B5310-26 B5310-26 BS3IO-26 BS31O-26 ref point 19.3 20.3 20.4 89.17 89.18 89.np 89.np.2 Si02 49.96 49.75 49.35 50.15 50.41 50.00 50.32 Ti02 1.56 1.38 1.46 1.50 1.76 1.75 1.84 A!203 15.05 15.84 15.23 15.69 15.47 15.42 15.26 FeO 10.43 8.10 8.86 9.50 9.58 9.58 9.47 MnO 0.17 0.14 0.16 0.19 0.12 0.19 0.13 MgO 6.57 5.51 6.70 5.23 5.46 5.46 5.39 aO 10.08 12.60 11.25 10.54 10.34 10.35 10.06 Na20 0.96 2.43 1.39 3.47 3.49 3.60 3.45 K20 - 0.48 0.54 0.47 0.50 P205 - 0.18 0.29 0.16 0.22 Total 94.82 95.80 94,43 96.93 97.45 96.97 96.64

Eastern Seamount SampleBS300-21 BS300-21 BS309-1BS309-1 BS309-1 BS309-1 BS309-1 ref point72.3 72.4 74.1 74.6 74,7 74.8 74.9 Si02 53.56 53.90 51.43 55.41 55.73 56.64 56.18 Ti02 1.73 1.74 0.67 2.06 2.07 2.04 2.12 A1203 14.14 14.54 17.77 14.22 14.06 14.11 14.03 FeO 10.16 8.88 5.84 11.59 10.91 11.62 11.37 MnO 0.19 0.18 0.12 0.18 0.19 0.19 0.16 MgO 4.72 4.60 6.58 3.07 3.08 3.15 3.08 QO 8.18 8.89 14.93 6.99 6.88 7.12 6.92 Na20 0.78 2.51 2.73 2.19 2.55 1.45 2.57 K20 - 0.55 - 0.78 0.80 0.72 0.75 P205 - 0.27 - 0.29 0.29 0.35 0.32 Total 93.46 96.05 100.08 96.78 96.56 97.39 97.50 Note:Hyphen (-) denotes "not analyzed" or "unsuccessful analysis." 30

r SampleBS292-17BS292-17BS292-17B5292-17BS292-17 ref point 19.1 19.2 19.4 20.1 20.2 5i02 40.18 40.25 40.07 40.49 40.37 FeO 12.00 11.96 12.26 13.58 10.47 MnO 0.20 0.20 0.20 0.16 0.16 MgO 48.60 48.57 47.87 46.71 49.80 0.22 0.22 0.29 0.31 0.20 Cr203 0.03 0.03 0.05 0.02 0.06 NO 0.26 0.26 0.22 0.17 0.37 Total 101.51 101.56 100.98 101.51 101.49

Western Seamount SampleBS31O-26 BS31O-26 BS31O-26 BS31O-26 BS31O-26 BS3IO-26 BS31O-26 ref point 89.2 89.3 89.4 89.5 89.6 89.7 89.8 Si02 39.29 38.80 38.91 39.29 39.04 39.13 38.66 FeO 16.61 16.44 16.67 16.13 16.98 15.97 16.63 MnO 0.22 0.21 0.22 0.18 0.36 0.20 0.24 MgO 44.30 43.94 44.86 44.87 44.01 45.31 44.20 0.28 0.25 0.32 0.25 0.31 0.19 0.33 Cr203 NO ------Total 100.70 99.64 100.97 100.72 100.70 100.79 100.07

WesternSeamount Sample BS31O-26BS310-26BS310-26BS310-26BS310-26BS310-268S310-26BS31O-26 ref point90.9 90.10 90.11 90.12 90.13 90.14 90.15 90.16 Si02 39.13 38.55 39.15 39.66 39.18 39.50 39.41 38.89 FeO 15.96 17.61 16.08 16.03 16.83 16.33 15.73 17.12 MnO 0.22 0.30 0.24 0.28 0.28 0.28 0.25 0.29 MgO 44.80 43.45 44.75 44.66 43.67 44.58 44.77 43.28 0.28 0.34 0.24 0.23 0.28 0.21 0.24 0.30 Cr203 ------NO ------Total 100.38100.24 100.45100.85 100.24100.90100.40 99.88

WesternSeamount Sample BS31O-7 BS31O-7 BS3IO-7 BS31O-7 BS3IO-7 BS31O-7 BS31O-7 ref point 83.3 84.5 84.7 84.6 84.8 83.2 83.1 Si02 39.14 40.00 39.21 39.47 39.65 40.01 40.10 FeO 17.38 15.91 17.20 16.13 16.57 12.97 12.95 MnO 0.28 0.26 0.25 0.25 0.26 0.21 0.17 MgO 44.24 45.17 45.21 45.44 45.44 48,01 48.19 00 0.28 0.26 0.25 0.24 0.27 0.24 0.25 Cr203 0.00 0.03 0.01 0.05 0.00 0.02 0.00 NO 0.15 0.17 0.17 0.22 0.19 0.24 0.22 Total 101.50 101.86 102.35 101.84 102.38 101.76 101.95 31

Eastern Seamount SampleB5297-2 BS297-2BS297-2BS297-2 BS297-2BS297-2 BS297-2BS297-2 ref point47.1 47.2 48.1 48.2 48.3 48.4 48.5 48.6 5i02 40.69 40.82 40.16 40.21 39.48 39.91 38.16 40.93 FeO 8.40 8.43 10.98 11.33 15.12 14.76 22.23 8.54 MnO 0.12 0.10 0.22 0.19 0.24 0.25 0.36 0.16 MgO 51.22 51.20 49.55 49.45 46.74 46.64 40.11 51.31 QO 0.19 0.17 0.21 0.23 0.26 0.28 0.27 0.19 Cr203 0.03 0.08 0.00 0.00 0.02 0.04 0.00 0.07 NO 0.42 0.40 0.24 0.21 0.13 0.17 0.07 0.45 Total 101.10 101.22101.39101.63 102.04102.07101.27 101.65

Eastern Seamount Sample BS300-21 BS309-1 BS309-1 BS309-1 BS309-1 ref point 72.2 74.5 73.3 74.3 74.4 Si02 37.78 37.36 37.04 36.94 37.46 FeO 23.66 29.56 30.19 29.73 27.67 MnO 0.39 0.45 0.47 0.49 0.42 MgO 38.26 33.45 33.80 33.94 36.46 CD 0.31 0.35 0.27 0.25 0.21 Cr203 0.00 0.00 0.01 0.00 0.00 NO 0.03 0.00 0.00 0.00 0.00 Total 100.57 101.67 101.88 101.42 102.26

PenguinIsland Sample A174 A174 A174 A174 A174 A174 A174 ref point 93.2 94.7 93.5 93.3 94.3 93.1 94.4 Si02 37.76 39.04 38.87 39.08 38.95 39.16 39.34 FeO 30.33 19.33 18.27 18.04 18.00 18.25 16.50 MnO 0.63 0.3 0.34 0.29 0.27 0.31 0.26 MgO 32.86 42.31 43.13 43.63 43.86 43.88 45.01 C) 0.44 0.24 0.28 0.26 0.26 0.24 0.20 Cr203 0.00 0.00 0.01 0.00 0.00 0.00 0.03 NO 0.06 0.07 0.11 0.1 0.1 0.04 0.13 Total 103.33 101.37 101.07 101.45 101.45101.96 101.52

PenguinIsland Sample A263 A263 A263 A263 A263 A263 ref point 99.2.11 99.2.12 99.2.13 99.2.14 99.2.1599.2.16 Si02 40.69 41.37 40.81 36.49 40.08 39.04 FeO 11.98 18.15 9.57 29.45 13.09 15.87 MnO 0.17 0.29 0.20 0.56 0.19 0.28 MgO 48.42 30.33 50.09 32.91 47.50 44.02 QO 0.20 1.72 0.15 0.29 0.19 0.21 Cr203 NO- - - - Total 101.46 91.86 100.82 99.70 101.04 99.41 32

King GeorgeIsland SampleA372 A372 A372 A372 A406 A406 A406 A406 ref point101.6 101.5 101.3 101.4 107.6 107.4 107.3 107.5 Si02 39.74 39.81 39.99 40.66 37.77 37.60 39.26 39.80 FeO 16.30 14.29 13.04 9.68 26.60 26.02 17.94 17.23 MnO 0.24 0.22 0.20 0.11 0.47 0.47 0.26 0.24 MgO 44.35 47.20 47.30 50.41 36.50 36.63 44.48 44.89 0.23 0.19 0.20 0.17 0.18 0.20 0.20 0.19 Cr203 0.00 0.02 0.00 0.03 0.00 0.00 0.02 0.04 NO 0.16 0.17 0.22 0.34 0.05 0.04 0.09 0.15 Total 101.05101.90 101.03101.43101.57101.11 102.30 102.56

DeceptionIsland

Sample D10 D10 D10 D10 D10 D10 ref point121.4 122.2 122.1 121.2 121.1 121.3 Si02 34.67 36.03 35.87 36.48 36.53 36.59 FeO 42.08 36.52 34.85 33.26 31.53 31.77 MnO 0.79 0.62 0.62 0.53 0.49 0.54 MgO 24.48 28.90 29.69 32.37 32.64 33.18 0.29 0.34 0.35 0.33 0.28 0.36 Cr203 0.00 0.02 0.00 0.00 0.00 0.02 NO 0.00 0.04 0.02 0.03 0.05 0.02 Total 102.44 102.54 101.50103.04 101.56 102.53 Note:Hyphen (-) denotes "not analyzed" or "unsuccessful analysis." 33

Table 8. Oxide Analyses by Electron Micronrobe

WesternSeamount Sample BS292-17 BS292-17 BS292-17 BS292-17 ref point 19.7 19.8 20.6 20.7 Si02 0.04 0.10 0.09 0.15 Ti02 1.01 0.94 0.94 1.15 A1203 22.92 22.31 19.36 21.76 Fe203 - - FeO 27.28 24.28 25.41 26.57 MnO 0.27 0.23 0.25 0.24 MgO 12.74 13.98 12.51 13.23 CaO 0.07 0.04 0.02 0.03 Na20 0.00 0.05 0.00 0.07 Cr2Q3 36.16 37.40 40.39 35.99 v205 - - - - NiO 0.12 0.11 0.15 0.11 Total 100.61 99.44 99.12 99.30 FefMg Cr/Cr+Al

Western Seamount Sample BS31O-26 BS31O-26 BS31O-26 BS31O-26 BS31O-26 BS31O-26 BS31O-26 ref point 89.15 89.16 90.1 90.2 90.3 90.4 90.5 Si02 ------Ti02 2.13 1.62 1.99 2.24 2.16 1.27 2.23 A1203 18.05 18.71 15.61 17.74 18.81 16.39 18.68 Fe203 15.60 14.38 14.08 16.38 15.89 14.39 16.05 FeO 19.63 18.92 20.16 19.68 20.39 19.09 20.51 MnO 0.00 0.00 0.00 0.00 0.00 0.00 0.00 MgO 10.57 10.89 10.00 10.56 10.38 10.42 10.06 CaO ------Na20 ------Cr203 31.51 33.33 36.13 30.88 30.90 36.72 29.69 V205 0.61 0.48 0.53 0.69 0.55 0.41 0.61 NiO ------Total 98.11 98.33 98.51 98.17 99.08 98.69 97.82 Fe/Mg 0.51 0.49 0.53 0.51 0.52 0.51 0.53 Cr/Cr+Al 0.54 0.54 0.61 0.54 0.52 0.60 0.52 ii

Sample BS310-26 BS310-26 BS31O-26 BS3IO-7 BS31O.7 BS31O-7 ref point 90.6 90.7 90.8 84.11 83.np 84.1 Si02 - - - 0.03 0.10 0.00 Ti02 2.26 2.23 2.28 2.04 1.92 2.05 A1203 18.55 18.07 18.33 17.94 19.55 19.58 Fe203 16.61 15.55 16.11 - - - FeO 19.27 19.90 19.24 32.70 33.96 33.08 MnO 0.00 0.00 0.00 0.26 0.28 0.29 MgO 10.82 10.67 10.81 10.95 11.15 11.21 - - - (IAfl fl14 Na20 - - - 0.07 0.02 0.04 Cr20.3 29.53 32.13 30.15 33.19 32.93 32.36 V205 0.59 0.57 0.62 - - NiO - - - 0,17 0.15 0.11 Total 97.64 99.12 97,54 97.35 100.06 98.86 Fe/Mg 0.50 0.51 0.50 - Cr/Cr+Al 0.52 0.54 0.53 -

EasternSeamount Sample BS297-2 BS297-2 BS297-2 BS300-21 BS309-1 BS309-1 BS309-1 ref point 48.9 48.1 48.11 72.6 73.6 73.7 74.11 Si02 0.00 0.00 0.03 6.84 0.21 0.45 0.06 Ti02 0.73 0.35 0.45 0.23 13.72 14.05 14.07 A1203 13.22 13.30 15.74 1.42 3.50 3.41 3.40 Fe203 ------FeO 15.61 20.64 20.47 63,14 73.53 74.05 74.73 MnO 0.18 0.27 0.30 0.06 0.33 0.39 0.36 MgO 16.52 12.38 12.29 0.25 3.82 3.91 3.91 CaO 0.00 0.12 0.08 0.54 0.24 0.18 0.07 Na20 0.00 0.04 0.00 1.09 0.00 0.00 0.12 Cr203 55.99 54.85 50.25 0.00 0.02 0.06 0.06 v205 ------NiO 0.21 0.12 0.12 1.53 0.00 0.07 0.03 Total 102.46 102.07 99.73 75.10 95.37 96.57 96.81 Fe/Mg Cr/Cr+Al 35

Table 8.(Continued

Penguinisland Sample A263 A263 A263 A263 A263 A263 A263 ref point 99.1.1 99.1.2 99.1.3 99.1.4 99.1.5 99.1.6 99.1.7 5i02 - - - - - Ti02 0.47 0.46 0.45 0.48 0.54 0.74 0.75 AJ203 13.51 13.39 13.07 12.89 14.42 22.07 19.83 Fe203 6.22 6.31 6.14 6.08 7.60 9.95 9.47 FeO 17.19 17.34 17.08 17.26 17.09 16.15 16.94 MnO 0.00 0.00 0.00 0.00 0.00 0.00 0.00 MgO 11.16 11.06 11.06 10.97 11.25 12.65 12.12 CaO ------Na20 ------Cr203 50.60 50.66 50.80 51.07 47.77 36.66 40.09 \'205 0.14 0.13 0.14 0.11 0.18 0.27 0.27 NiO ------Total 99.28 99.35 98.74 98.84 98.84 98.50 99.47 Fe/Mg 0.46 0.47 0.46 0.47 0.46 0.42 0.44 Cr/Cr+Al 0.72 0.72 0.72 0.73 0.69 0.53 0.58

PenguinIsland Sample A263 A263 A263 A263 A263 A263 A263 refpoint 99.1.8 99.1.9 99.1.10 99.1.11 99.1.12 99.1.13 99.1.14 Si02 ------Ti02 0.82 1.00 1.15 1.61 1.20 7.08 5.53 Al203 20.35 21.23 21.18 21.56 19.29 5.00 7.47 Fe203 10.94 12.82 19.97 17.19 11.97 33.67 27.12 FeO 17.10 17.84 11.88 18.72 26.35 33.19 31.23 MnO 0.00 0.00 0.00 0.00 0.02 0.66 0.57 MgO 11.84 11.60 15.22 11.22 6.20 2.75 3.41 CaO ------Na20 ------Cr203 36.96 33.83 27.59 27.29 34.45 15.54 22.18 VO5 0.23 0.28 0.30 0.42 0.28 1.57 1.16 NiO ------Total 98.25 98.59 97.29 98.00 99.76 99.45 98.68 Fe/Mg 0.45 0.46 0.30 0.48 0.70 0.87 0.84 Cr/Cr+Al 0.55 0.52 0.47 0.46 0.55 0.68 0.67 36 Table 8.(Continued

Penguin Island Sample A263 A263 A263 A263 A263 A263 A263 ref point 99.1.15 99.w/n 99.wfl 99.2.1 99.2.2 99.2.3 99.2.4 Si02 ------Ti02 1.55 16.76 45.10 0.27 0.44 0.76 37.72 A1203 23.02 2.14 0.30 1.66 15.72 16.05 0.17 Fe2Q 16.42 31.71 0.00 61.31 6.15 9.17 0.00 FeO 18.85 43.83 45.65 30.74 12.56 20.36 50.32 MnO 0.00 0.53 1.02 0.02 0.00 0.00 0.38 MgO 11.42 0.89 2.00 0.19 14.09 9.50 2.10 CaO ------Na20 ------Cr203 27.11 0.04 0.00 3.95 48.72 43.35 0.07 2O5 0.43 2.33 3.19 0.09 0.13 0.20 3.39 NiO ------Total 98.80 98.22 97.25 98.23 97.82 99.38 94.15 Fe/Mg 0.48 0.97 0.93 0.99 0.33 0.55 0.93 Cr/Cr+A1 0.44 0.01 - 0.62 0.68 0.64 0.23

P Sample A263 A263 A263 A263 A263 A263 A263 ref point 99.2.5 99.2.6 99.2.7 99.2.8 99.2.9 99.2.10 99.np Si02 ------Ti02 2.20 0.79 3.13 0.79 4.15 8.10 29.34 A1203 12.13 17.76 5.37 16.15 9.23 6.95 1.61 Fe2O 23.47 9.66 16.57 9.48 17.81 18.92 5.75 FeO 15.50 16.57 26.03 21.05 26.49 31.16 53.57 MaO 0.25 0.00 0.21 0.00 0.02 0.03 0.76 MgO 11.77 11.92 5.00 8.86 6.34 4.90 1.23 CaO ------Na20 ------Cr203 28.92 41.35 38.69 41.85 33.52 25.27 0.10 VO5 0.50 0.24 1.51 0.19 0.91 1.36 2.56 NiO ------Total 94.73 98.28 96.50 98.37 98.47 96.68 94.91 Fe/Mg 0.43 0.44 0.75 0.57 0.70 0.78 0.96 Cr/CrAl 0.62 0.61 0.83 0.64 0.71 0.71 Note: Hyphen(-)denotes "not analyzed" or "analysis failed." 37

WesternSeamount Sample BS292-17 BS292-17 BS292-17 ref point 19.5 19.6 20.5 Si02 50.41 51.23 50.27 Ti02 0.06 0.09 0.06 A1203 30.23 29.86 30.15 FeO 0.69 0.81 0.64 MnO 0.01 0.03 0.03 MgO 0.26 0.31 0.24 CaO 14.93 14.35 14.61 Na 3.23 3.35 3.20 K20 0.07 0.10 0.09 P205 0.05 0.03 0.08 Total 99.93 100.17 99.35 Fe/Mg Ab Or An

WesternSeamount Sample BS31O-26 BS31O-26 BS31O-26 BS31O-26 BS31O-26 BS31O-7 BS31O-7 ref point 89.1 89.13 89.14 90.17 90.18 84.9 83.np Si02 53.29 45.64 47.37 47.61 49.86 48.03 50.27 Ti02 - - - - - 0.06 0.07 A1203 27.96 32.61 33.05 32.62 31.53 31.06 30.50 FeO 1.03 0.62 0.53 0.59 0.54 0.63 0.65 MnO - - - - - 0.04 0.00 MgO 0.35 0.27 0.10 0.12 0.12 0.18 0.19 CaO 11.93 17.17 16.54 16.15 14.94 16.16 15.31 Na20 4.71 1.60 2.11 2.22 3.07 2.40 3.08 K20 0.09 0.02 0.01 0.04 0.06 0.04 0.07 - - - - - 0.07 0.09 Total 99.35 97.93 99.71 99.36 100.12 98.65 100.23 Fe/Mg 0.62 0.56 0.76 0.74 0.71 Ab 41.42 14.41 18.76 19.88 27.03 - - Or 0.54 0.09 0.06 0.23 0.32 - - An 58.04 85.50 81.18 79.89 72.65 - - Table 9.(Continued)

Eastern Seamount Sample BS297-2 BS297-2 BS300-21 BS300-21 BS300-21 BS300-21 BS300-21BS309-1BS309-1 refpoint47.3 48.8 71.1 71.4 71.7 71.8 72.5 74.1 73.4 Si02 50.38 51.69 52.70 50.88 51.41 52.91 52.52 50.04 53.50 Ti02 0.07 0.08 0.08 0.06 0.04 0.09 0.05 0.07 0.08 A1203 30.04 29.74 27.32 28.73 30.32 28.66 28.88 28.49 28.22 FeO 0.59 0.54 0.88 0.57 0.63 0.87 0.76 0.69 0.74 MnO 0.03 0.02 0.00 0.00 0.00 0.00 0.02 0.01 0.02 MgO 0.25 0.26 0.17 0.14 0.13 0.21 0.20 0.12 0.13 CaO 15.19 14.03 12.22 13.49 14.06 12.80 13.41 12.56 12.33 Na20 3.25 3.59 4.37 3.80 3.49 4.31 4.03 4.56 4.62 K20 0.05 0.08 0.00 0.00 0.06 0.07 0.05 0.05 0.06 P205 0.02 0.00 0.00 0.00 0.02 0.00 0.02 0.00 0.00 Total 99.89 100.04 97.74 97.67 100.14 99.93 99.95 99.58 99.69 Fe/Mg ------Ab------Or------An------

PenguinIsland Sample A263 A263 A263 ref point 99.2.17 99.2.18 99.rim Si02 49.16 49.06 48.62 Ti02 - - - A1203 32.05 31.73 32.15 FeO 0.72 0.80 0.93 MnO - - - MgO 0.12 0.04 0.01 CaO 15.13 14.60 15.17 Na20 2.83 2.92 2.72 K20 0.09 0.15 0.09 P205 - - - Total 100.10 99.30 99.70 Fe/Mg 0.77 0.91 0.99 Ab 25.13 26.33 24.36 Or 0.53 0.89 0.55 An 74.34 72.78 75A)9 Note: (Hyphen (-) denotes "not analyzed" or "analysis unsuccessfuL" 39

WesternSeamount Sample BS31O-26 BS3IO-26 BS31O-26 BS31O-26 BS31O-7 BS31O-7 BS31O-7 ref point 89.9 rim 89.10 89.11 89.12 84.1 84.2 84.3 Si02 50.29 52.00 50.19 49.25 49.58 49.68 50.91 Ti02 0.82 0.53 0.79 0.90 0.93 0.82 0.67 A1203 4.31 2.81 4.14 5.19 4.76 4.33 3.59 Fe203 1.44 0.76 1.97 2.90 - - FeO 3.91 4.59 3.69 2.69 5.61 5.39 5.54 MnO 0.09 0.14 0.05 0.11 0.14 0.15 0.13 MgO 15.48 16.91 15.36 15.67 15.69 15.74 1639 CaO 21.74 20.69 21.92 21.38 22.00 22.08 21.70 Na20 0.31 0.27 0.33 0.34 0.34 0.33 0.31 Cr203 0.89 0.53 0.94 0.88 0.89 0.91 0.55 NiO - - - - 0.04 0.04 0.03 Total 99.29 99.24 99.38 99.30 99.98 99.47 100.02 Fe/Mg 0.13 0.14 0.12 0.09 - - - Wa 46.86 43.19 47.43 47.13 - - En 46.42 49.11 46.25 48.05 - - - Fs 6.73 7.70 6.32 4.82 - -

Eastern Seamount Sample BS297-2 BS300-21 BS300-21 BS300-21 BS300-21 BS300-21 BS309-1 BS309-1 BS309-1 refpoint48.7 71.2 71.3 71.5 71.6 72.1 74.2 73.1 73.2 Si02 48.62 52.07 51.32 51.02 51.05 52.11 48.20 48.98 49.12 Ti02 1.67 0.44 0.60 0.66 0.61 0.46 1.68 1.48 1.16 A1203 4.36 1.70 2.78 2.87 2.57 2.00 4.38 3.50 3.40 Fe203 ------FeO 9.77 6.69 6.59 7.02 6.40 5.60 10.66 11.28 10.45 MnO 0.24 0.16 0.19 0.16 0.16 0.15 0.24 0.30 0.23 MgO 14.99 17.66 17.00 17.15 16.76 17.41 13.61 14.45 14.48 CaO 18.59 20.35 20.87 20.39 20.68 21.36 19.99 19.27 19.98 Na20 0.32 0.22 0.28 0.29 0.28 0.25 0.47 0.39 0.37 Cr203 0.03 0.02 0.04 0.01 0.04 0.04 0.00 0.00 0.02 NiO 0.00 0.03 0.00 0.02 0.00 0.04 0.02 0.00 0.00 Total 98.59 99.34 99.67 99.59 98.55 99.42 99.25 99.65 99.21 Fe/Mg ------Wo------En------Fs------PenguinIsland Sample A174 A174 A174 A174 A174 A263 ref point 94.6 93.4 94.2 94.1 94.5 99.w41ag Si02 47.38 47.89 47.72 48.30 50.69 49.44 Ti02 1.48 1.25 1.21 1.04 0.54 1.03 A1203 5.47 6.12 6.05 5.73 3.67 5.56 Fe203 - - - 1.87 FeO 7.28 6.68 6.11 6.06 5.14 3.94 MnO 0.12 0.11 0.10 0.11 0.13 0.06 MgO 14.04 14.06 14.39 14.44 16.34 15.05 CaO 22.45 22.57 22.80 22.96 22.32 21.70 Na20 0.39 0.41 0.35 0.34 0.35 0.31 Cr203 0.28 0.21 0.40 0.42 0.46 0.39 NiO 0.01 0.00 0.00 0.00 0.04 - Total 98.90 99.30 99.13 99.40 99.68 99.34 Fe/Mg - - - - - 0.13 Wo - - - - - 47.43 En - - - - - 45.75 Fs - - - - - 6.82

King GeorgeIsland Sample A372 A372 A372 A372 A372 A372 A406 A406 ref point 101.2 102.4 102.2 102.3 101.1 102.1 107.7 107.2 Si02 50.82 49.91 50.70 50.62 50.61 51.52 51.02 52.33 Ti02 0.59 0.68 0.47 0.53 0.58 0.44 0.54 0.31 A1203 4.48 4.50 3.59 4.04 3.94 3.17 3.73 2.16 Fe203 - - - - - FeO 4.51 4.85 4.17 4.56 4.66 4.25 4.47 3.99 MnO 0.10 0.12 0.08 0.08 0.06 0.10 0.14 0.11 MgO 16.00 16.20 16.65 16.65 16.73 17.33 16.54 17.66 CaO 22.72 22.52 22.49 21.86 22.70 22.22 22.82 22.68 Na20 0.43 0.27 0.25 0.31 0.25 0.28 0.22 0.19 Cr203 0.35 0.79 0.75 0.35 0.38 0.58 0.38 0.29 NiO 0.01 0.03 0.04 0.00 0.04 0.00 0.04 0.00 Total 100.01 99.87 99.19 99.00 99.95 99.89 99.90 99.72 Fe/Mg Wo En Fs 41

Table 10.(Continued)

DeceptionIsland Sample D2 D2 refpoint 111.2 111.1 Si02 52.59 52.54 Ti02 0.55 0.38 A1203 0.72 0.53 Fe203 - FeO 21.79 21.68 MaO 0.83 0.80 MgO 22.38 23.36 CaO 1.97 1.74 Na20 0.03 0.05 Cr203 0.02 0.00 NiO 0.03 0.00 Total 100.91 101.08 Fe/Mg Wo En Fs Methods

Sample Selection

A total of 127 dredged samples from 7 dredge hauls within the Bransfield Strait were examined in hand sample.Samples that appeared to be glacial erratics or duplicates of a single lithology were not selected for analysis.Initially, 32 samples from 5 dredge hauls were chosen for further study.All 19 Penguin Island, King George Island, and Deception Island samples were also chosen for analysis. All of these samples were analyzed for major elements (Table 3). After examining the major element data it was obvious that some of the samples were redundant.For this reason 14 of the dredged samples and one Deception Island sample were not analyzed further. Fourteen diversely representative samples were analyzed for rare earth elements and certain trace elements by isotope dilution. Sixteen samples were analyzed for radiogenic isotopes. Special prepartion was necessary before analyzing sample LH-1 from King George Island.Slight secondary contamination in the form of thin calcite veins were removed from the sample by leaching for several hours in warm acetic acid.This technique was used only on LH-1, and then only before the isotope dilution and radiogenic isotope analyses.Thus, for LH-1, the data in Table 3 are not directly comparable to the data in Tables 4 and 5.

Major Elements

Major element concentrations were determined by x-ray fluorescence spectroscopy with the assistance of Greg Campi at Oregon State University (Table 3).One gram of powdered sample -'S in --- V V V 4--- 20 V V 4 si v icetongue 4--- SI .-.--- columnar 0 , i ° 0 200m V' O

-. V

:::: :::: miiiChopin, LMAZUREK POINT FM. IIjiid'. 5J

3400 0° 65° 120° 250 270°

Figure 9. Geologic cross-section of Low Head on King George Island (from Porebski and Gradzinski, 1987) showing location of sample analyzed in this study (LH-1).Circled numbers 1-8 are locations of measured field stations.Circled letters are abbreviations as follows: KG- Krakowiak Glacier Member, LH-Low Head Member, S-Siklawa Member, and OC-Oberek Member. was fused with five grams of lithium borate flux and cooled to form a glass bead.This bead was then powdered in a ball mill and 2.5 g of this powder were pressed into a pellet using 5 drops of polyvinyl alcohol as a binding agent.Pellets were analyzed in a fully automated multi-spectrometer Philips PW1600 energy dispersive x- ray fluorescence machine simultaneously for Si, Ti, Al, Fe, Mn, Mg, Ca, Na, K, and P.Standards BHVO-1 (U. S. G. S. basalt) and GL-0 (A. N. R. T. glauconite) were analyzed as unknowns (measured and accepted values included in Table 3) to test analytical accuracy.Counting time on each sample was 120 s, with background and monitor corrections made automatically between samples.X-ray counts to elemental concentrations conversion factors were established by measuring basalt standards JB-1, JB-2, and JB-3 from the Geological Survey of Japan, and BIR-1 from the U. S. Geological Survey, and diorite standard DR-N from Association Nationale de la Recherche Technique, and converting their respective x-ray counts to accepted concentrations found in Govindaraju (1984).Several Tahitian basalts used as internal standards were also included in each sample run. Weight percent loss on ignition (Table 3) was determined at Oregon State University.Approximately one gram of powdered sample was left in an 88°C oven overnight to evaporate surficial water.The sample was then placed in an unglazed porcelain crucible, weighed, and placed in a 1000°C oven for one hour.The sample was then cooled for 30 minutes in a dessicator, reweighed, and the percent weight loss calculated and recorded.

Trace Elements

Trace elements (Ba, Cr, Cu, Li, Ni, Rb, Sr, and Zn) were determined by atomic absorption spectrophotometry at Oregon State University. Powdered samples weighing approximately 500 mg were digested in anHF:HNO3solution and a small amount of CsCI was added to suppress thermal ionization in the flame source. Absorptions were determined on a Perkin-Elmer 5000 flame source spectrophotometer which automatically converted absorption to concentration-in-solution by comparison to prepared standard solutions of known concentrations.All samples were counted over two consecutive integrations of five seconds each.The concentrations in ppm reported herein are simple averages of two duplicate weighings, dissolutions, and countings.External standardization and accuracy was maintained by including Japanese Geological Survey rock standards JA-3, JB-la, and JB-3 amongst the analyzed samples and normalizing to their accepted values (included in Table 3).

Rare Earth Elements

Rare earth element concentrations (Table 4) were determined by isotope dilution mass spectrometry at Cornell University under the guidance of Dr. William White and following the procedures outlined in Hooker et al. (1975), Thiriwall (1982), and White and Patchett (1984).Forty to fifty milligrams of powdered sample and 600 mg of a spike solution of known isotopic composition were carefully weighed.This mixture was digested in anHF:HNO3solution and pipetted into a ion exchange resin column that separated the solution into alkalies (K, Rb, and Cs), Sr, Ba, and rare earth element (REE) fractions.The REE fraction was pipetted into a teflon powder ion exchange column that separated the solution into La, light REE (Ce, Nd, Sm, and Eu), and heavy REE (Gd, Dy, Er, Yb, and Lu) fractions. Each of these six fractions were run on a VG Sector thermal ionization mass spectrometer.This instrument has six computer-controlled Faraday cup collectors, so all relevant isotopes of an element were collected simultaneously.Count times on samples and backgrounds, and data correction and reduction procedures are given in White and Patchett (1984).A blank consisting of 5 drops of the spike solution was processed through the chemistry lab, but no readable signal could beobtained on themass spectrometer; so theblankis considered to bezero. The U. S.G. S. standard BCR-1 wasrun asan unknown,and the resultscompare well to acceptedvalues(Table 4).

Isotopes

Radiogenic isotope ratios (87 Sr/86Sr, 143Nd/' 44Nd, 2O6pb/2O4p b, 207Pb/204Pb, and 208Pb/204Pb) (Table 5) were also measured on the mass spectrometer at Cornell University with the assistance of Dr. White.For Nd and Sr, powdered samples weighing approximately 60 mg were leached with HC1, digested in HF:HNO3, and pipetted into an ion exchange resin column that separated Sr and Nd fractions from the other elements.Further details of the chemical and analytical techniques used for Sr and Nd are given in White and Patchett (1984).Pb analyses were performed by Dr. White using procedures described in White and Dupré (1986).

Electron Microprobe

Electron microprobe data were collected at Oregon State University and at the University of Washington.Analyses were done on glass (Table 6), olivine (Table 7), oxides (Table 8), plagioclase (Table 9), and pyroxene (Table 10).X-ray counts were calibrated to elemental concentrations using U. S. G. S. and N. B. S. standards: Ten polished thin sections (BS292-17, BS297-2, BS300-21, BS309-1, BS31O-7, A174, A372, A406, D2, and D10) were carbon coated and analyzed in the JEOL Superprobe 733 at the University of Washington.Count times of 10 s on peak and background, a 15 nA beam, and standard Bence-Albee corrections were used (Bence and Albee, 1968; Albee and Ray, 1970). 47 Analyses of 2 thin sections (BS31O-26, A263) were performed on the Cameca SX 50 electron microprobe at Oregon State University using a 20 nA electron beam.Count times were 10 s on peaks and 5 s on backgrounds on each side of a peak. ZAF corrections of five iterations were performed automatically.

Whole Rock K-Ar Age Dating

Whole rock potassium-argon ages were determined at Oregon State University by conventional methods (Dairymple and Lanphere, 1969). Samples A372 and A375 originally contained large olivine phenocrysts, but these olivines were removed by hand in case they contained excess 40Ar.After olivine removal the samples could be classified as fine-grained basalts and basaltic andesites with no glass nor alteration, and less than 3% phenocrysts.Potassium concentrations were determined using the same atomic absorption spectrophotometry technique as that described for trace elements. Argon isotope data were collected with the assistance of Dr. Robert Duncan on an AEI-MS1O mass spectrometer with an on-line vacuum extraction system.Total fusion of whole rock chips was obtained using an external radio-frequency induction heating coil.The ratios 36Ar/38Ar and 40Ar/38Ar were measured, and the absolute abundance of 40Ar was determined by adding a spike with a known volume of 38Ar.Precise monitoring of atmospheric 40Ar/36Ar allowed very small amounts of radiogenic 40Ar to be detected in the sample (e. g., Cassignol and Gillot, 1982). A correction factor for instrumental mass fractionation was obtained by measuring the 40Ar/36Ar of an air spike (accepted 4OAr/36Ar of 295.5). Results

The results of the chemical analyses of whole rock samples are in Tables 3-5.The Melville Peak and the dredged samples are tholeiitic basalts and basaltic andesites (Figure 8), while the Penguin Island lavas are mildly alkaline basalts, and the Deception Island lavas range from mildly alkaline basalt to trachyte.Samples from the western seamount and Penguin Island, as well as the one basalt from Deception Island, contain up to 5% normative nepheline (Appendix D).The rest of the samples are silica saturated, although only the Deception Island dacites contain significant amounts of normative quartz.Most of the variation in major element chemistry in the Bransfield Strait samples can be accounted for by fractional crystallization of olivine and oxides (Figure 10).The more evolved (lower MgO in Figure 10) Deception Island lavas also appear to have undergone considerable fractionation of pyroxene and plagioclase. Radiogenic isotope ratios (87Sr/86Sr, l43Nd/'44Nd, 206Pb/204Pb, 2O7pb/204pb, and 208Pb/204Pb; Table 5) show the small distinction between the two seamounts, and their collective similarities to Melville Peak and Deception Island, and differences from Penguin Island (Figures 11 and 12). Complete petrographic descriptions of 55 seamount and island samples are given in Appendix C.Abbreviated descriptions are given here, with mineral chemistry from Tables 6-10 given where appropriate. Western seamount. (47.5-49.2% Si02) Dredge 292.Olivine: 1-7%, euhedral to subhedral, mostly <2mm but some up to 4.5mm;Piagioclase: 0-5%, euhedral to subhedral, <1mm, occasionally in glomercrysts with olivine; Oxides: rare, euhedral to subhedral, <0.3mm, in olivine and groundmass; Groundmass: olivine+plagioclase; Vesicles: 20-50%, 4mm, some with glassy rims. Dredge 310.Olivine: 1-3%, euhedral to subhedral, l .5mm, Fo82..84, some slightly altered along cracks;Plagioclase: 2-3%, euhedral to subhedral, 1.5mm, An81..85;Pyroxene: rare, euhedral to 12

10 08 CM

CM Z4

0 0 2 4 6 8 10 12 14

1.8

1.6

1.4 CM 21.2 I- 1.0

0.8

0.6

0.4 0 2 4 6 8 10 12 14

12 1 0

.. 10 +e +++ .:.. 0 CM W Seamount u84) E Seamount ..! U & + I-'enguin isiano £ MeMile Peak 6 a Low Head A Deception Island

4 . 0 2 4 6 8 10 12 14 wt% MgO Figure 10. Major element variation versus MgO for samples analyzed in this study.Notice that most of the major element variation can be explained by fractionation of olivine and oxides. Low MgO Deception Island samples have also been affected by plagioclase fractionation. 0.5 133

0.5131

0.5129 z

'-4 Z 0.5127

0.5 125

0.5123 0.7020 0.7030 0.7040 0.7050 0.7060 87SrI86Sr

Figure 11. Plot of 87Sr/86Sr versus '43Nd/144Nd of samples analyzed in this study.Fields (from Hickey et aL, 1986) are included for mid-ocean ridge basalt (MORB), ocean island basalt (OIB), southern volcanic zone of the Andes (SVZ), and island arcs.Symbols representing Bransfield Strait samples analyzed in this study are as follows:solid diamond-western seamount; hollow diamond-eastern seamount; plus sign-Penguin Island;solid square-Melville Peak; hollow square-Low Head; and triangle-Deception Island. 51

39.5

39.0

38.5 Arc

37.5 15.4 15.5 15.6 15.7 2O7Pb/2O4Pb

Figure 12. Plot of 207Pb/204Pb versus 208Pb/204Pb of samples analyzed in this study.Fields (from Hickey et al., 1986) include mid- ocean ridge basalt (MORB), ocean island basalt (OIB), South Sandwich Island Arc (S. Sandwich Arc), and southern volcanic zone of the Andes (SVZ).Symbols represent Bransfield Strait samples as given with Figure 11. 52 subhedral, <0.6mm, in glomerocrysts with olivine and plagioclase, En46..49 Fs5..8; Oxides: rare, euhedral to subhedral, <0.06mm, in larger olivines;Groundmass: 0-90% glass,+plagioclase (An58)+olivine; Vesicles: 15-25%, many containing vesicular glass. Eastern seamount.(51.0-53.8% SiO2) Dredge 297.Olivine:<1%, euhedral to subhedral, mostly <0.6mm but some up to 3mm;Plagioclase:1-3%, euhedral to subhedral, 0.6mm, occasionally in glomercrysts with olivine;Oxides: rare, euhedral to subhedral, 0.l5mm, in olivine and groundmass; Groundmass: 5-70% glass,+olivine+plagioclase;Vesicles: 15-30%, 1.Smm, some with glassy rims. Dredge 300.Olivine:very rare, subhedral to anhedral, <1mm; Plagioclase: 2%, euhedral to subhedral, mostly <1mm but some up to 3.5mm;Oxides: rare, euhedral to subhedral, <0.05mm;Groundmass: 10-80% glass,+olivine+plagioclase; Vesicles: 5-35%, lOmm, some with glassy rims. Dredge 309.Olivine:<1%, euhedral to anhedral, <0.4mm; Plagioclase: 2%, euhedral to subhedral, <0.9mm;Oxide: 1%, euhedral to subhedral, <0.2mm, mostly in groundmass, but some in olivine; Groundmass 25-85% glass,+olivine+plagioclase;Vesicles rare and <1.5mm. Penguin Island: (47.5-48.6% SiO2) Olivine: 0.5-15%, euhedral to subhedral, <4mm, Fo83..90;Plagioclase: 7-15%, euhedral to subhedral, <2mm, An7375, larger phenocrysts show zoning;Clinopyroxene: 1- 5%, euhedral to subhedral, <3mm, En46 Fs7;Oxides: 0-2%, euhedral to subhedral, <0.3mm, mostly in olivine;Groundmass: plagioclase+ clinopyroxene+olivine ± oxides;Vesicles: 0-10%, usually 1-2mm but up to 5mm. King George Island: Melville Peak. (50.5-52.5% SiO2) Olivine: 1-5%, euhedral to subhedral, <2.5mm, some large glomerocrysts with clinopyroxene, some with minor alteration;Plagioclase: 0.5-2%, euhedral to subhedral, mostly lmm but some up to 2.3mm;Clinopyroxene: rare-2%, euhedral to subhedral, <3mm, many are zoned;Oxides: none to rare, euhedral to subhedral, <0.03mm, in olivine but only those not 53 in glomerocrysts with cpx;Groundmass: plagioclase+olivine+oxides ± pyroxene;Vesicles: 0-25%, <2.5mm. Low Head. (47.1% Si02) Olivine: unaltered ones rare, subhedral to anhedral, <1mm;Plagioclase: 10%, subhedral to euhedral, <1mm, corroded edges;Oxides: 0.5%, subhedral to anhedral, 0.lmm, in altered mineral and groundmass; Altered mineral (serpentine?): 20%, anhedral to subhedral, lmm;Groundmass: plagioclase+ alteration; Several 0.1-1.5mm wide veins of secondary mineral that may be CaCO3. Deception Island: (50.8-79.0% Si02) Ranges from plagioclase and oxide phyric subalkaline basalt to vesicular, oxide phyric dacite and rhyodacite to high silica alteration product. DISCUSSION

Three possible sources may contribute to the chemical composition of arc magmatism: dehydration and/or partial melting of subducted oceanic crust (basalt + alteration + sediment + seawater), partial melting of subarc mantle, and assimilation of overlying arc crust as the magmas pass to the surface.Inversions of the geochemistry of arc magmas into relative contributions of each of these components, combined with data from experimental petrology, have led to a general consensus that the primary source of arc magma is hydration-related partial melting of the subarc mantle wedge. But the chemical contributions of subducted sediment and assimilated arc crust to the composition of arc volcanism remain controversial.Some authors (e. g., Armstrong, 1971; Kay, 1980; Hickey et al., 1986; and White and Dupré, 1986) argue in favor of subducted sediments and older arc crust contributing to arc magma chemistry, while other authors (e.g., Sinha and Hart, 1971; Arculus, 1981; and Arculus and Powell, 1986) argue against these contami- nants having significant affects on arc chemistry. The chemical characteristics of arc magmas that these authors endeavor to explain are their enrichment in alkali and alkali earth elements relative to rare earth and high field strength elements. Most of this enrichment can be accommodated by processes unique to the subduction zone environment (Hickey et al., 1986).Many authors have suggested that alkalies and alkali earths are preferentially mobilized in the hydrous environment just above the dehydrating subducted slab, and are therefore concentrated in the partial melts that ascend to the arc.Some of these authors believe that the excess alkalies and alkali earths are scavenged from the mantle itself (e.g., Arculus and Powell, 1986) while others think the only possible source of such high concentrations of these elements must be subducted sediment (e. g., Kay, 1980).In either case the subarc mantle becomes permeated with regions highly enriched in those incompatible elements that are mobile under hydrous conditions.These enriched regions are melted and incorporated in 55 the partial melt of the surrounding peridotite, and the eruptions of this partial melt are consequently enriched in certain incompatible elements.The chemical expression of this enrichment process can be seen by comparing the trace element patterns of ocean island basalt (OIB in Figure 13), representing normal (undepleted) mantle, to arc basalts from the Andes (Figure 14).Notice that the arc basalts are enriched in alkalies, and in some cases the alkali earths, relative to the rare earth elements, while the ocean island basalt is slightly depleted in alkalies relative to rare earth elements. Saunders et al. (1980) proposed that minerals capable of retaining the rare earth and high field strength elements (e.g., ilmenite, sphene, zircon, and apatite) may be stabilized by the high PH2O and f02 present in the mantle wedge just above the dehydrating subducted slab.This phenomenon would serve to further increase the ratio of alkalies and alkali earths to rare earths and high field strengthelements. Stern (1980) and Lin et al. (1989) presented geochemical evidence that arc chemistry is the starting point for the geochemical development of marginal basins.The chemistry of marginal basin volcanism develops with age from arc-like to MORB-like as successive eruptions repeatedly deplete the mantle source. The short history of the Bransfield Strait does not allow testing of this model using Bransfield Strait samples alone, so we compare the chemistry of the Bransfield Strait lavas to volcanism in three other marginal basins that are in different stages of development.But first we must account for the chemical variation between the different lavas in the Bransfield Strait. 56

N-MORB 1000

100

110

1 Cs Rb Ba K La Ce Sr Nd Sm TI Yb

Figure 13. Bulk earth normalized trace element plot of typical caic-alkaline basalt (CAB), enriched mid-ocean ridge basalt (E-MORB), island arc tholeiite (IAT), normal mid-ocean ridge basalt (N-MORB), and ocean island basalt (OIB).Data are from Sun (1980).Data were normalized to bulk earth using the following values compiled in Hickey et al. (1986): Cs-0.012, Rb-0.35, Ba-3.5, K-120, La-0.31, Ce- 0.81, Sr-li, Nd-0.602, Sm-0.196, Ti-620, Yb-0.21. 57

1000

100

10 o Andes 26S U Andes 38S Patagonia 46S

1 Cs Rb Ba K La Ce Sr Nd Sm TI Yb

Figure 14. Bulk earth normalized trace element plot for Andean and Patagonian rocks.Andes 26°S data from Kay et al. (1988), Andes 38°S data from Mufloz B. and Stern (1989), and Patagonia 46°S data from Hawkesworth et al. (1979).Normalization values given with Figure 13. Chemical variationwithintheBransfieldStraitsamples

The variation in chemistry of the lavas within the Bransfield Strait can be accounted for by different degrees of partial melting of the same source, followed by low-pressure crystal fractionation.The rare earth element patterns for Melville Peak (Figure 15), Penguin Island (Figure 16), and two representative dredged sample (Figure 17) exhibit variation that can be accounted for by different degrees of partial melting of a single source, most likely a garnet peridotite with residual garnet retaining some of the heavy rare earth elements (Weaver et al., 1979; Lin et al., 1989). Modeling of rare earth element patterns resulting from different degrees of partial melting of a garnet peridotite versus a spinel peridotite source (Figure 18, from Lin et al., 1989) produced crossing patterns and slopes for the garnet peridotite partial melts that match the Bransfield Strait patterns (plotted together in Figure 19). The Penguin sample has the steepest pattern, suggesting it was formed by the least amount of partial melting.This also agrees with the slightly higher alkali content of the Penguin samples (Table 5 and Figure 8).The rare earth element pattern of the dredged sample has the shallowest slope, suggesting it was created by a slightly higher degree of partial melting, but of the same source as the Penguin and Melville lavas.The differences in degrees of partial melting were apparently quite small however, as comparison with Figure 18 suggests a range of 12% to 17% partial melting can explain the variations in rare earth element pattern slopes. Deception Island contains considerable geochemical variation, but its rare earth element pattern (Figure 20) shows that this variation can be accounted for by fractional crystallization from a parent quite similar to the dredged sample included in Figure 20. Sample D10 is the most primitive of the Deception samples, and it appears to be related to the dredge sample by fractionation of a phase that is incompatible with rare earth elements (e.g., olivine). The other two Deception samples are trachytes that are much more evolved than sample Dl0, and appear to be related to D10 by 59

100

I- 0 C 10 C)

a. .1 E 1 Melville Peak i A372

La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure15. Chondrite normalized rare earth element patterns for two Melville Peak samples. 100

Penguin Island

xA174 + A261

1 La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure16. Chondrite normalized rare earth element patterns for two Penguin Island samples. 100

0 1

10 C) 0

1 La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure 17. Chondrite normalized rare earth element patterns for two dredged seamount samples. 62

GARNET L1-IERZOLITE SPINEL LHERZOLITE [x = K : 101 .59.138 101 .509 18 En1.2111.41 IEn .11 .086 101 .67 I 101 .19 .421 I .2301 .11 .629 LJ

I 0a: z 0 UI - .... UI UI a: 10

SOURCE (3 X CHONORITE)

La Ce Nd Sm Eu Gd Dy Er Yb

Figure 18. Results of batch melting models of garnet peridotite (solid lines) and spinel peridotite (dashed lines) from Lin et al. (1989).X values are mineralogy and P values are melt fractions.Lines are calculated rare earth element patterns with the attached numbers being the percent melting that created that line. 63

100

-S I-

0 . 10 C)

0. E

1 La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure 19. Chondrite normalized rare earth element plot of three representative samples from Melville Peak, Penguin Island, and a dredged seamount.Crossing patterns and change in slope are characteristic of different extents of partial melting of a garnet bearing source. 100

I- C 0 .c 10 Deception Island C) £D2

E A 09 a010 (I) Seamount BS292-17

I I La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure 20. Chondrite normalized rare earth element plot for three Deception Island samples and a dredge sample (same as the one in Figure 19) analyzed in this study.Similarities suggest that Deception Island samples are related to dredge sample by fractional crystallization. fractionation of a phase compatible with Eu (e. g., plagioclase). Lowered Sr and Ti in the trachytes also suggests fractional crystallization of plagioclase, clinopyroxene, ilmenite and magnetite. Radiogenic isotope data support the conclusion that all of the Bransfield Strait samples have the same source.The Penguin Island samples have higher 87Sr/86Sr and lower '43Nd/144Nd, but similar Pb isotope ratios to the rest of the Bransfield Strait samples (Table 5 and Figures 11 and 12).The fact that the Pb isotopes of the Penguin Island lavas are indistinguishable from the other Bransfield Strait lavas argues against significantly greater sediment contamination being responsible for the differences in 87Sr/86Sr and 143Nd/144Nd of the Penguin Island samples. The differences in Sr and Nd isotope data must therefore be accounted for by assimilation of arc crust during ascent of the Penguin Island lavas.Published 87Sr/86Sr data for the South Shetland Island arc volcanics (Smellie, 1984) range from 0.7034 (and 620 ppm Sr) for Oligocene volcanics to 0.7241 (and 35 pm Sr) for Cretaceous volcanics (these are measured, not initial, ratios).Published 87Sr/86Sr data for pre-Jurassic basement rocks of the South Shetland Islands and Antarctic Peninsula (Pankhurst, 1983) range from 0.7090 (and 1010 ppm Sr) for a granodiorite to 0.7293 (and 59 ppm Sr) for a gneiss.Mixing calculations using these values show that the higher 87Sr/86Sr of the Penguin Island samples can be explained by less than 5% assimilation of basement rocks.Nd isotopic data for the basement rocks are not available. Alternatively, there may be some small scale isotopic heterogeneities in the Bransfield Strait source.This would be expected in a mantle that has been receiving subducted oceanic crust for at least 200 Ma (Weaver et al., 1979). Thus, the intervolcano major and trace element variation, as well as the Pb isotope variation, can be accounted for by different degrees of partial melting of a garnet peridotite source, followed by low pressure fractional crystallization.The Sr and Nd isotopic data are also compatible with this model, although the Penguin Island lavas show the effects of assimilation of small amounts of the more radiogenic arc crust. Comparisons with otherlocations

It is not clear whether the volcanism in the Bransfield Strait should have geochemical affinities with continental arc volcanism or with marginal basin volcanism, or both.Since the Bransfield Strait has no contemporary analogues with which to make a direct comparison, the Bransfield Strait data reported here will be compared to published data from two examples of marginal basin rifting in tectonic settings somewhat similar to the present Bransfield Strait (the arc/back-arc systems of the Scotia Sea and the Andes). We will also compare the Bransfield Strait lavas to published data from the Sarmiento ophiolite in southern Chile.The Sarmiento ophiolite has previously been interpreted as a Cretaceous marginal basin that was very similar to the present day Bransfield Strait (e. g., Dalziel, 1989).

Comparison with the Scotia arc/back-arc system

The Scotia Rise is an active back-arc spreading center in the eastern Scotia Sea, behind the South Sandwich Islands volcanic arc (Figure 2).Saunders et al. (1982) dredged fresh, highly vesicular tholeiites from the Scotia Rise ranging from olivine normative to quartz normative, with major element chemistry broadly similar to MORB.The trace element pattern for a representative Scotia Rise sample (Figure 21) illustrates this similarity to MORB (compare 'Scotia Spreading Center' sample in Figure 21 with 'E-MORB' and 'N- MORB' in Figure 13), except the Scotia Rise sample is slightly more enriched in K and Rb, even compared to E-MORB.Saunders et al. (1982) consider this typical chemistry for back-arc basin volcanism, and attribute it to partial melting of a nondepleted mantle source that may have been enriched in alkalies by the adjacent subducting slab. 1000-

U Scotia spreading center ISo Sandwich isis -c 100- a So Sandwich isis IAT J w

10-

1- I Cs Rb Ba K La Ce Sr Nd Sm TI Yb

Figure 21. Bulk earth normalized trace element plot for Scotia Rise spreading center and South Sandwich Islands volcanic arc. Scotia Rise data from Saunders et al. (1982) and South Sandwich Islands data from Hawkesworth etal. (1977).Normalization values given with Figure 13. Trace element patterns of two samples from the South Sandwich Islands (Figure 21) are among the most primitive of the samples analyzed by Hawkesworth et al. (1977).Their depleted rare earth element chemistry (CeN/YbN < 1) contrasts with their enrichment in alkalies and alkali earths.This is typical of island arc tholeiite (compare with IAT in Figure 13), and they are in fact classified that way by Hawkesworth et al. (1977). The Bransfield Strait dredge samples are chemically similar to the Scotia Rise sample in their slight enrichment in light rare earths relative to heavy rare earths (CeN/YbN = 2-2.5), but the Scotia Rise sample is slightly depleted in alkalies and alkali earths compared to the flat trace element patterns of the Bransfield Strait dredge samples (Figure 22).Since high alkalies and alkali earths relative to rare earths are diagnostic of a mantle source influenced by subduction zone processes (Hickey et al., 1986), this suggests that the Bransfield Strait source has been enriched to a larger degree by subduction zone processes.This is not surprising considering the distance from the South Shetland Trench to the Bransfield Strait is only 100-120 km, compared to 440 km from the Scotia Arc to the Scotia Rise.It is also possible that the longer history and faster spreading rate of the Scotia Rise (8 million years at 70-90 mm/yr vs. 1-2 million years at <50 mm/yr for the Bransfield Strait) have simply given the Scotia Rise the opportunity to deplete more of the alkalies and alkali earths from its source. The Bransfield Strait samples have no apparent similarities to any of the South Sandwich Island samples, reiterating that the Bransfield Strait samples are clearly not calc-alkaline in chemistry. There are some similarities in radiogenic isotopes, however.Figure 12 shows Pb isotope data for the Bransfield Strait samples along with fields representing MORB, OIB, detrital oceanic sediment, and the South Sandwich Islands.Data for the Bransfield Strait and South Sandwich Islands plot on a trend away from the MORB and OIB fields toward the oceanic sediment field.Clearly the sources of the Bransfield Strait and South Sandwich Islands volcanics contain some component of subducted sediment. 0 >-

I-

E U)

I- Co

C-)

-I

0 0 0 o0 1 qe )Hn8/aIdwS

Figure 22. Bulk earth normalized trace element plot for dredged samples analyzed in this study.Normalization values given with Figure 13. 70 Comparison with the Andean arc/back-arc system

The Andean cordillera is a classic and well-studied continental magmatic arc containing significant chemical variability within its volcanic rocks.Most of the cordjllera is caic-alkaline volcanics, but alkaline types are also present (e. g., Baker et al., 1981). Figure 14 graphically shows the chemical variability, as well as the typically caic-alkaline enrichment in alkali and alkali earth elements to concentrations almost 700 times bulk earth, even though these samples are among the most primitive volcanics analyzed in their respective studies.The sample with the lowest abundances of trace elements ('Andes 38S') is chemically similar to island arc tholeiites (high Cs/Rb and Ba/La; compare to trace element pattern'IAT' in Figure 13).Hickey et al. (1986) explain the chemistry of '38S' by partial melting of undepleted (OIB) mantle that has been enriched by alkalies and alkali earths derived from the subducted crust of the Nazca Plate. The back-arc activity of the Andean cordillera consists of alkaline plateau basalts associated with diffuse back-arc extension of the continental crust of South America (Baker et al., 1981).The representative trace element pattern in Figure 14 (labeled 'Patagonia 46S') shows general similarities to OIB, and lacks evidence of contamination by the subducted Nazca Plate.However, no Ba or Cs data are available for the Patagonian basalts, and these are two of the elements most diagnostic of the chemical influence of a subduction zone. None of the Bransfield Strait samples resemble the Andean samples to any significant extent.The Bransfield Strait samples are not as enriched in incompatible elements as the calc-a!ka!ine Andes samples, nor do they have the smooth, concave downward trace element patterns of the alkaline Patagonia basalt (compare Figures 22, 23, 24, and 25 to Figure 14).This again supports the statement that the Bransfield Strait samples are not from caic-alkaline arc, or even alkaline back-arc, volcanism. 71

I-

E U)

0 z

1 U)

-J

0 00 !<

0, C)

0 0 0 o c q1Jc2 lInaI9IdweS

Figure 23. Bulk earth normalized trace element plot for Deception Island samples analyzed in this study.Normalization values given with Figure 13. 72

0 >-

I-

E U)

Cl)

C.)

U) -J

- (0 Qc0

C.)

o 0 r 0 0 r0 1 1flJ3 ng/edwes

Figure 24. Bulkearth normalized trace element plot for Melville Peak samplesanalyzed in this study.Normalization values given with Figure 13. 73

-o >-

E U)

-Dz

I-

Cl)

-j

-a

c -,--

m cO)X+ a)

-D

o 0 0 o 1 qjie inioIdweS

Figure 25. Bulk earth normalized trace element plot for Penguin Island samples analyzed in this study.Normalization values given with Figure13. 74 Figures 11 and 12 show that there are some similarities between the Andes and the Bransfield Strait in radiogenic isotopes, however.Pb isotopes show that the Andes and Bransfield Strait volcanics have been contaminated to similar extents by subducted sediment.Like the South Sandwich Island and Bransfield Strait volcanics, the Andes volcanics have Pb isotopes trending away from the MORB and OIB fields toward the oceanic sediment field. The Bransfield Strait samples have 87Sr/86Sr and '43Nd/144N d ratios similar to many island arcs (Figure 11).This could be used to suggest that the source of the Bransfield Strait lavas is similar to an arc source, but this is not a unique solution since, unlike in the Pb isotope plots, the island arc and OIB mantle fields completely overlap.The Penguin Island samples are the most similar to the Andes data shown in Figure 11.This is probably due to assimilation by the Penguin Island lavas of older South Shetland Island arc crust that is similar to Andean crust.

Comparison with the Sarmiento ophiolite

The Sarmiento ophiolite has been interpreted by Saunders et al. (1979) and Stern (1980) as an obducted Cretaceous marginal basin, and has previously been suggested as an ancient analogue of the present day Bransfield Strait based upon geographical (Saunders and Tarney, 1982), geophysical/structural (Daiziel, 1989), and geochemical (Keller and Fisk, l989a) data.All of the known samples from the Sarmiento except those from the lowest level cumulate gabbros appear to have been depleted in K and Rb by hydrothermal alteration at the original seafloor spreading center (Saunders and Tarney, 1982), so care must be taken in examination of their geochemistry.The rare earth element chemistry of the Sarmiento appears intact though (Stern, 1980), so these elements are the focus of this geochemical comparison. The rare earth element chemistry of the Bransfield Strait samples (Figures 19 and 20) is strikingly similar to data from the Sarmiento ophiolite (Figure 26).They have similar mild enrichment in light rare earths (CeN/YbN 2), and similar absolute abundances of rare earths.Stern (1980) suggested that the limited chemical variation in the crystalline rocks of the Sarmiento could not be due to different extents of partial melting of a homogeneous source, but rather to similar extents of partial melting of a source progressively depleted by previous partial melting events.If this is true then the rare earth element data suggest that the Sarmiento was a marginal basin at a stage of maturity similar to the present Bransfield Strait. Unfortunately, no isotopic data are available from the Sarmiento to further test its chemical similarity to the Bransfield Strait. Structural data suggest that upon obduction in the mid- Cretaceous the Sarmiento was a marginal basin approximately 50 km wide (deWit, 1977), quite similar to the present 65 km width of the Bransfield Strait.This is further evidence in support of the interpretation that the Sarmiento was in a similar stage of maturity to the present day Bransfield Strait. 76

100

a- C 0 .c 10 C)

0.. E

1 La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure 26. Plot of rare earth element data (from Saunders et al., 1979) for a dike in the Sarmiento ophiolite.Notice similarity to rare earth element patterns of Bransfield Strait samples (Figure 19). 77

CONCLUSIONS

The unusual geochemical features of the Quaternary volcanism in the Bransfield Strait are a result of the transitional nature of its tectonic setting and its source region of melt production.While the volcanism is causally related to rifting of the Bransfield Strait marginal basin, the chemistry of the lavas exhibit some of the characteristics of arc volcanism.Geochemical evidence for enrichment in alkali and alkali earth elements and radiogenic isotopes in the Bransfield Strait source by Jurassic to Quaternary(?) subduction beneath the South Shetland Islands and Antarctic Peninsula supports the conclusion that subduction zones are sites of geochemical flux from the crust to the mantle. Chemical variation within the Quaternary lavas of the Bransfield Strait can be accounted for by slightly different extents of partial melting of a garnet peridotite source, followed by fractionation of olivine, oxides, clinopyroxene, and plagioclase, and, for Penguin Island, small amounts of assimilation of older South Shetland Islands arc crust. Geochemical comparisons between the Bransfield Strait lavas and volcanism in the arc/back-arc systems of the Andes and the South Sandwich Islands show that Bransfield Strait lavas are unique. No published data suggest the existence of any contemporary analogues of the Bransfield Strait marginal basin.A geochemical comparison with the Sarmiento ophiolite in southern Chile shows striking similarities, however.Structural and geochemical data suggest that the Sarmiento was a marginal basin approximately 50 km wide when it was obducted onto the Andean margin of South America in the Cretaceous (deWit, 1977; Saunders, 1979).The similarities in size and geochemistry between the Sarmiento and the Bransfield Strait support the interpretation that the Sarmiento was a marginal basin very similar to the present day Bransfield Strait. This conclusion lends additional support to the interpretation that many ophiolites are obducted marginal basin crust. BIBLIOGRAPHY

Albee, A. L., and L. Ray, 1970, Correction factors for electron probe microanalysis of silicates, oxides, carbonates, phosphates, and sulfides. Anal. Chem., v. 42,p. 1408-1414. Arculus, R. J., 1981, Island arc magmatism in relation to the evolution of the crust and mantle. Tectonophysics,v. 75, p. 113-133. Arculus, R. J.,and R. Powell, 1986, Source component mixing in the regions of arc magma generation. J. Geophys. Res., v. 91,p.5913- 5926. Armstrong, R. L., 1971, Isotopic and chemical constraints on models of magma genesis in volcanic arcs. Earth Planet. Sci. Letters, v. 12, p.137-142. Ashcroft, W. A., 1972, Crustal structure of the South Shetland Islands and Bransfield Strait. British Antarct. Surv. Sci. Reports, No. 66, 43 pp. Baker, P. E., W. J. Rea, J. Skarmeta, R. Caminos, and D. C. Rex, 1981, Igneous history of the Andean Cordillera and Patagonian plateau around latitude 46°5. Phil. Trans. R. Soc. London, v. 303,p. 105- 149. Barker, P. F., 1976, The tectonic framework of Cenozoic volcanism in the Scotia Sea region, a review. In 0. González-Ferrán (Ed.), Proceedingsofthe Symposium on Andean and Antarctic Volcanology Problems, IAVCEI Spec. Srs.,p. 330-346. Barker, P. F., 1982, The Cenozoic subduction history of the Pacific margin of the Antarctic Peninsula: ridge crest-trench interactions. J. geol. Soc. London, v. 139,p.787-801. Barker, P. F., and J. Burrell, 1977, The opening of the Drake Passage. Marine Geology, v. 25,p. 15-34. Barker, P. F., and I. W. D. Dalziel,1983, Progress in geodynamics in the Scotia Arc region, In Geodynamics in the Eastern Pacfic Region, Caribbean and Scotia Arcs. S. J. R. Cabré (Ed.), Am Geophys. Union Geodynamics Srs. 9,p. 137-170. 79

Barton, C. M., 1965, The geology of the South Shetland Islands. III. The stratigraphy of King George Island. British Antarct. Surv. Sci. Reports, No. 44, 33pp. Bence, A. E., and A. L. Albee, 1968, Empirical correction factors for electron microanalysis of silicates and oxides. J. Geol., v. 76,p. 382- 403. Birkenmajer, K., 1980, Age of the Penguin Island volcano, South Shetland Islands (West Antarctica), by the lichenometric method. Bull. Polish Acad. Sci., Earth Sci., v. 27,p. 69-76. Birkenmajer, K., 1981, Structural evolution of the Melville Peak volcano, King George Island (South Shetland Islands, West Antarctica). Bull. Polish Acad. Sci., Earth Sci., v. 29,p. 341-351. Birkenmajer, K., 1982, The Penguin Island volcano, South Shetland Islands (Antarctica): Its structure and succession. Studia Geologica Polonica, v. 74, p.155-173. Birkenmajer, K., E. Soliani, Jr., and K. Kawashita, 1988, Early Miocene K-Ar age of volcanic basement of the Melville Glaciation deposits, King George Island, West Antarctica. Bull. Polish Acad. Sci., Earth Sd., v. 36,p. 25-34. Birkenmajer, K., and R. A. Keller, submitted, Pleistocene age of the Melville Peak volcano, King George Island, West Antarctica, by K- Ar Dating. Antarctic Science. [see Appendix II]. Cassignol, C., and P-Y. Gillot, 1982, Range and effectiveness of unspiked potassium-argon dating: Experimental groundwork and applications. In G. S. Odin (Ed.) Numerical Dating in Stratigraphy. John Wiley and Sons, New York,p. 159-179. Dairymple, G. G., and M. A. Lanphere, 1969, Potassium-Argon Dating. W. H. Freeman Co., San Francisco, 258pp. Daiziel, I. W. D. (Ed.), 1989, Tectonics of the Scotia Arc. 28th International Geological Congress Field Trip Guidebook T180, Am. Geophys. Union, Wash., D. C., 206pp. deWit, M. J., 1977, The evolution of the Scotia Arc as a key to the reconstruction of southwestern Gondwanaland. Tectonophysics, v. 37, p. 53-81. Fisk, M. R., in press, Volcanism in the Bransfield Strait, Antarctica. J. South Am. Earth Sd. Forsyth, D. W., 1975, Fault plane solutions and tectonics of the South Atlantic and Scotia Sea. J. Geophys. Res., v. 80,p.1429-1443. Gledhill, A., D. C. Rex, and P. W. G. Tanner, 1982, Rb-Sr and K-Ar geochronology of rocks from the Antarctic Peninsula between Anvers Island and Marguerite Bay, In C. Craddock (Ed.), Antarctic Geoscience, Univ. Wisconsin Press, Madison, p. 315-323. Godoy, E., R. Harrington, and E. Tidy, 1987, Sobre el carácter "anómalo" del volcanismo reciente en las islas Shetland del Sur. Inst. Antart. Chileno, Ser. Cient., v. 36,p. 21-32. González-Ferrán, 0., and Y. Katsui, 1970, Estudio entegral del volcanismo Cenozoico Superior de las Islas Shetland del Sur, Antartica. Inst. Antart. Chileno, Ser. Cient., v. 1,p. 123-174. Govindaraju, K., 1984, 1984 compilation of working values and sample description for 170 international reference samples of mainly silicate rocks and minerals. Geostandards Newsletter, v. 8, 88 pp. Hawkesworth, C. J., R. K. O'Nions, R. J. Pankhurst, P. J. Hamilton, and N. M. Evensen, 1977, A geochemical study of island arc and back-arc tholeiites from the Scotia Sea. Earth Planet. Sci. Letters, v. 36,p. 253 -262. Hawkesworth, C. J., M. J. Norry, J. C. Roddick, P. E. Baker, P. W. Francis, and R. S. Thorpe, 1979, 143Nd/'44Nd, 87Sr/86Sr, and incompatible element variations in caic-alkaline andesites and plateau lavas from South America. Earth Planet. Sci. Letters, v. 42,p.45-57. Hickey, R. L., F. A. Frey, D. C. Gerlach, and L. Lopez-Escobar, 1986, Multiple sources for basaltic arc rocks from the Southern Volcanic Zone of the Andes (34°-41°S): Trace element and isotopic evidence for contributions from subducted oceanic crust, mantle, and continental crust. J. Geophys. Res., v. 91,p. 5963-5983. Hooker, P. J., R. K. O'Nions, and R. J. Pankhurst, 1975, Determination of rare-earth elements in USGS standard rocks by mixed-solvent ion exchange and mass -spectrometric isotope dilution. Chemical Geology, v. 16, p. 189-196. Hoskins, A. K., 1963, The basement complex of Neny Fjord, Graham Land.British Antarct. Surv. Sd. Reports, No. 43, 49pp. Irvine, T. N., and W. R. A. Baragar, 1971, A guide to the chemical classification of the common volcanic rocks. Can. J. Earth Sci., v. 8,p. 523-5 48. Kay, R. W., 1980, Volcanic arc magmas: Implications of a melting- mixing model for element recycling in the crust-upper mantle system. J. Geol., v. 88, p. 497-522. Kay, R. W., N. J. Hubbard, and P. W. Gast, 1970, Chemical characteristics and origin of ocean ridge volcanics. J. Geophys. Res., v. 75, p. 1585-1613. Kay, R. W., and R. G. Senechal, 1976, The rare earth geochemistry of the Troodos Ophiolite complex. J. Geophys. Res., v. 81,p. 964-970. Kay, S. M., V. Maksaev, R. Moscoso, C. Mpodozis, C. Nasi, and C. B. Gordillo, 1988, Tertiary Andean magmatism in Chile and Argentina between 28°S and 33°S: correlation of magmatic chemistry with a changing Benioff Zone. J. South Am. Earth ScL, v. 1,p. 21-38. Keller, R. A., and M. R. Fisk, 1989a, Bransfield Strait, Antarctica: A modern analogue of the Sarmiento Ophiolite of southern Chile [abs.]. Terra Cognita, in press. Keller, R. A., and M. R. Fisk, 1989b, Rifting and volcanism in the Bransfield Strait and South Shetland Islands. Antarctic J. U. S., in press, [see Appendix I]. Kendall, E. N., 1831, Account of the Island of Deception, one of the New Shetland Isles. J. Royal geogr. Soc., v. 1,p. 62-66. Kuno, H., 1966, Lateral variation of basalt magma across continental margins and island arcs. Geol. Surv. Canada Paper 66-15, p. 317- 336. Lin, P.-N., R. J. Stern, and S. H. Bloomer, 1989, Shoshonitic volcanism in the northern Mariana arc 2. Large-ion lithophile ands rare earth element abundances: Evidence for the source of incompatible element enrichments in intraoceanic arcs. J. Geophys. Res., v. 94,p. 4497-4514. Muñoz B., J., and C. R. Stern, 1989, Alkaline magmatism within the segment 38°-39°S of the Plio-Quaternary volcanic belt of the southern South American continental margin. J. Geophys. Res., v. 94, p. 4545-4560. Pankhurst, R. J.,1982, Sr-isotope and trace element geochemistry of Cenozoic volcanic rocks from the Scotia Arc and the northern Antarctic Peninsula. In, C. Craddock (Ed.), Antarctic Geoscience, Univ. Wisconsin Press, Madison,p. 229-234. Pankhurst, R. J., 1983, Rb-Sr constraints on the ages of basement rocks of the Antarctic Peninsula. In R. L. Oliver, P. L. James, and J. B. Jago (Eds.), Antarctic Earth Science, Austral. Acad. Sci., Canberra, and Cambridge Univ. Press, Cambridge,p. 367-371. Pankhurst, R. J., and J. L. Smellie, 1983, K-Ar geochronology of the South Shetland Islands, Lesser Antarctica: apparent lateral migration of Jurassic to Quaternary island arc volcanism. Earth Planet. Sci. Letters, v. 66,p. 214-222. Parra, J. C., J. Bannister, and 0. Gonzalez-Ferran, 1984, Aeromagnetic survey over the South Shetland Islands, Bransfield Strait and part of the Antarctic Peninsula. Revista Geol. Chile, No. 23,p. 3-20. Pelayo, A. M., and D. A. Wiens, 1989, Seismotectonics and relative plate motions in the Scotia Sea region. J. Geophys. Res., v. 94,p. 7293-7 320. Porebski, S. J., and R. Gradzinski, 1987, Depositional history of the Polonez Cove Formation (Oligocene), King George Island, West Antarctica: A record of continental glaciation, shalow-marine sedimentation and contemporaneous volcanism. Studia Geol. Polonica, v. 93, Pt. 7, p. 7-62. Roach, P. J., 1978, The nature of back-arc extension in Bransfield Strait Ilabs.]. Geophys. J. Royal Astron. Soc., v. 53,p. 165. Roobol, M. J., 1973, Historic volcanic activity at Deception Island. British Antarct. Surv. Bull., No. 32, p. 23-30. Saunders, A. D., and J. Tarney, 1979, The geochemistry of basalts from a back-arc spreading centre in the East Scotia Sea. Geochim. Cosmochim. Acta, v. 43, p. 555-572. Saunders, A. D., J. Tarney, C. R. Stern, and I. W. D. Dalziel, 1979, Geochemistry of Mesozoic marginal basin floor igneous rocks from southern Chile. Geol. Soc. Am. Bull., v. 90,P. 237-258. Saunders, A. D., J. Tarney, and S. D. Weaver, 1980, Transverse geochemical variations across the Antarctic Peninsula: Implications for the genesis of caic-alkaline magmes. Earth Plan. Sci. Letters, v. 46, p. 344-360. Saunders, A. D., J. Tarney, S. D. Weaver, and P. F. Barker, 1982, Scotia Sea floor: Geochemistry of basalts from the Drake Passage and South Snadwich spreading centers. In, C. Craddock (Ed.), Antarctic Geoscience, Univ. Wisconsin Press, Madison,p.213-222. Sinha, A. K., and S. R. Hart, 1972, A geochemical test of the subduction hypothesis for generation of island arc magmas. Carnegie Institution of Washington Yearbook 71,p. 309-312. Smellie, J. L., 1983, A geochemical overview of subduction-related igneous activity in the South Shetland Islands, Lesser Antarctica. In R. L. Oliver, P. L. James, and J. B. Jago (Eds.), Antarctic Earth Science, Austral. Acad. Sci., Canberra, and Cambridge Univ. Press, Cambridge,p. 352-356. Smellie, J. L., 1988, Recent observations on the volcanic history of Deception Island, South Shetland Islands. British Antarct. Surv. Bull., No. 81, P. 83-85. Smellie, J. L., R. E. S. Davies, and M. R. A. Thomson, 1980, Geology of a Mesozoic intra-arc sequence on Byers Peninsula, Livingston Island, South Shetland Islands. British Antarct. Surv. Bull., No. 50,p. 55- 76. Smellie, J. L., R. J. Pankhurst, M. R. A. Thomson, and R. E. S. Davies, 1984, The Geology of the South Shetland Islands: VI. Stratigraphy, Geochemistry and Evolution. British Antarct. Surv. Sci. Reports, No. 87, 85 pp. Stern, C. R., 1980, Geochemstry of Chilean ophiolites: Evidence for the compositional evolution of the mantle source of back-arc basin basalts. J. Geophys. Res., v. 85,p.955-966. Stern, R. J., and E. Ito, 1981, Oxygen, strontium and neodymium isotopic and incompatible trace element constraints on the origin of melts in the Mariana and Volcano Island arcs. I. A. V. C. E. I. Syposium on Arc Volcanism,p. 348-349. Suess, E., M. Fisk, and D. Kadko, 1987, Thermal interaction between back-arc volcanism and basin sediments in the Bransfield Strait, Antarctica. Antarctic J. U. S., v. 22, no. 5, p. 46-49. Sun, S.-S., 1980, Lead isotopic study of young volcanic rocks from mid-ocean ridges, ocean islands and island arcs. Phil. Trans. R. Soc. London, v. 297,p. 409-445. Tanner, P. W. G., R. J. Pankhurst, and G. Hyden, 1982, Radiometric evidence for the age of the subduction complex in the South Orkney and South Shetland Islands, West Antarctica. J. Geol. Soc. London, v. 139,p. 683-690. Thiriwall, M. F., 1982, A triple-filament method for rapid and precise analysis of rare-earth elements by isotope dilution. Chemical Geology, v. 35,p. 155-166. Thomson, J. W., 1968, The geology of the South Orkney Islands: II. The petrology of Signy Island. British Antarct. Surv. Sci. Reports, No. 62, 30pp. Thomson, M. R. A., and R. J. Pankhurst, 1983, Age of post- Gondwanian caic-alkaline volcanism in the Antarctic Peninsula region. In R. L. Oliver, P. L. James, and J. B. Jago (Eds.), Antarctic Earth Science, Austral. Acad. Sci., Canberra, and Cambridge Univ. Press, Cambridge, p. 328-333. Weaver, S. D., A. D. Saunders, R. J. Pankhurst, and J. Tarney, 1979, A geochemical study of magmatism associated with the initial stages of back-arc spreading. The Quaternary volcanics of Bransfield Strait, from South Shetland Islands. Contrib. Mineral. Petrol., v. 68, p.151-169. White, W. M., and Dupré, 1986, Sediment subduction and magma genesis in the Lesser Antilles: Isotope and trace element constraints. J. Geophys. Res., v. 91,p.5927-5941. White, W. M., and J. Patchett, 1984, Hf-Nd-Sr isotopes and incompatible element abundances in island arcs: implications for magma origins and crust-mantle evolution. Earth. Planet. Sci. Letters, v. 67,p. 167-185. Whitford, D. J., W. M. White, A. W. Hofman, and D. E. James, 1978, Separation and isotopic analysis of neodymium. Carnegie Institution of WashingtonYearbook 77, p. 620-623. Whiticar, M. J., E. Suess, and H. Wehner, 1985, Thermogenic hydrocarbons in surface sediments of the Bransfield Strait, Antarctic Peninsula.Nature, v. 314, p.87-90. Appendices Appendix I.Keller and Fisk. 1989b Rifting and Volcanism in the BransfieldStrait and South ShetlandIslands

Randall A. Keller and Martin R. Fisk College of Oceanography Oregon State University Corvallis, Oregon 9733 1-5503

Southeastward subduction of oceanic crust of the Drake Plate beneath the Antarctic Plate has produced volcanism on the South Shetland Islands and Antarctic Peninsula since the Mesozoic.In the late Cenozoic, subduction and its related volcanism slowed or ceased. Possiblyin response tothis,the BransfieldStrait,a young (2-3 million-year-old)marginalbasin,openedbetweentheSouth Shetland Islands and the Antarctic Peninsula (Figure 27). In November 1985, Martin R. Fisk participated in a cruise to theBransfieldStraitaboardthe German polarresearchvessel Polarsternto study the thermal interaction between sediments and volcanism in a back-arc basin (Suess, Fisk, and Kadko, 1987).During the cruise, approximately 150 kilograms of rocks were dredged from several seamounts at what bathymetrically appeared to be the axis of spreading of the Bransfield Strait (Table 11).Thirteeen samples were collected from Deception Island during a brief stop on that same cruise.Our collection was later supplemented by 10 samples of very young volcanic rocks from King George Island andPenguin Island donated to us by Krzysztof Birkenmajer of the Polish Academy of Sciences. Our present collection represents most of the volcanic activity in the South Shetland Islands and the Bransfield Strait in the past several million years.We selected a representative suite of volcanic samples and areinthe process of analyzingtheir petrology and geochemistry.To date, we have done some preliminary petrographic 87 work, two potassium-argon ages, and 51 major element analyses by X-ray fluorescence. The dredge samples are fresh pillows and some massive flows of very vesicular, olivine-rich basalts and basaltic andesites. Most samples were dredged from depths in excess of 1800 meters, so their extreme vesicularity suggests unusually high volatile contents. As expected, their major element chemistry is most similar to that of back-arc basin basalts.Two of the samples yielded potassium-argon ages of 50,000 and 100,000 years.We conclude that the submarine volcanic activity that we sampled is due to active back-arc rifting in the Bransfield Strait. DeceptionIslandisanhistoricallyactivecomposite stratovolcano near the southwest end of the South Shetland Islands volcanic arc.The samples we collected from the island span a range of compositions from subalkaline basalt to rhyodacite.While this range of rocktypesiscompatible with Deception Island beinga typical island arc volcano, its location is in alignment with the axis of riftingof the BransfieldStrait. This suggeststousthat,while Deception Island lavas may have a subduction-contaminated source, their thermal impetus for eruptionis probably rift-related. Trace element and isotope data will help us test this hypothesis in the near future. Most of King George Island was created by late Mesozoic to Cenozoic islandarc volcanism that,priortothe opening of the Bransfield Strait, was part of the Antarctic Peninsula magmatic arc. However, our samples are from thesiteof the most recent (late PleistocenetoHolocene) phase of volcanism,the Melville Peak volcano on the eastern end of the island (Birkenmajer, 1982a).These samples are chemically very similar to some of our dredge samples. Penguin Island is a young cinder cone off the southeast coast of King George Island that has been recently active (Birkenmajer,1982b). The Penguin Island samples are tholeiiticto mildly alkalicolivine basalts. Similarities in major element chemistry between our seamount samples and our King George Island and Penguin Island samples (Figure 28) suggest that the recent volcanism on these two islands is not true island arc volcanism, but instead is related to the rifting in the Bransfield Strait (Keller and Fisk, 1987).Again, we must await trace element data to test this interpretation. Thus,ourpresentsamplecollectionrepresentsvolcanism associated with the initial stages of rifting of an island arc and the formation of a back-arc basin.We are examining this transition from arc to back-arc volcanism with special attention to changes in source chemistry and thermal regime, and the possibility of a decrease in input by the subducted slab with time and with increasing distance from the trench.As part of thisproject, William White (Cornell University) is measuring strontiun, neodymium, and lead isotopes to help answer some of these questions (White, Cheatham, and Fisk, 1987). This commingling of arcand back-arc lavasintheSouth Shetland Islandsalsoholdsinterestingimplicationsforophiolite genesis, since ophiolites often contain both of these types of lavas in a small area.We, therefore, plan to compare our geochemical data to some well-studied ophiolites. The cruise in which M. R. Fisk participated in was Antarktis IV, from Rio de Janeiro(11November 1985)toPunta Arenas(1 December 1985) aboard the R/V Polarstern operated by the Alfred Wegener Institute for Polar Research, Bremerhaven, Federal Republic of Germany. We thank chief scientist Gerold Wefer for inviting us to participate in this cruise.We are grateful to Krzysztof Birkenmajer of the Polish Academy of Sciences for generously donating samples from Penguin and King George islands. This work is supported by National Science Foundation grants DPP 85-12395 and DPP 86-14022 to M. R. Fisk and a Texaco Fellowship to R. A. Keller.

References

Birkenmajer, K.,1982a,Structural evolution of the Melville Peak volcano, King George Island (South Shetland Islands, West Antarctica). Bulletin de L'Academje Polonaise des Sciences, Série des sciences de la terre, 29(4), 341-351. Birkenmajer, K.,1982b, The Penguin Island volcano, South Shetland Islands(Antarctica):Itsstructure and succession. Studia Geologica Polonica, 74, 155-173. Irvine, T.N., and W.R.A. Baragar, 1971, A guide to the chemical classification of the common volcanic rocks.Canadian Journalof Earth Sciences, 8, 523-548. Keller, R. A., and M. R. Fisk, 1987, Magmatism associated with the initial stages of back-arc rifting, Bransfield Strait, Antarctica.Eos, 68(44),1533. Suess, E., M. Fisk, and D. Kadko, 1987, Thermal interaction between back-arc volcanism and basin sediments inthe Bransfield Strait, Antarctica. Antarctic Journalofthe United States, 22(5), 46- 49. White, W. M., M. Cheatham, and M. Fisk, 1987, Geochemistry of back-arc basin volcanics from Bransfield Strait, Antarctic Peninsula. Eos,68(44), 1520.

TABLE11.BransfieldStrait DredgeStations and Results,ANT IV/2

Station Latitude Longitude Area Results Number S W

285 63°9.8' 61°45.5' hill in Low Island Basin glacial boulders 286 63°11.8' 61°13.0' hill in Low Island Basin no sample 290 62°15.2' 58°10.7' King George Island shelf glacial boulders 292 62°12.3' 57°30.3' seamount, King George Basin pillow basalts 297 62°15.4' 57°24.5' floor, King George Basin fresh basalts 300 62°14.1' 57°23.5' seamount, King George Basin fresh basalt 307 62°18.0' 57°32.8' hill, King George Basin glacial boulders 309 62°13.4' 57°28.6' floor, King George Basin fresh basalt 310 62°12.9' 57°28.8' seamount, King George Basin pillow basalts r I ..

: 600W8 Figure 27.Sketch map of the Bransfield Strait area.Contours are depths in kilometers. 91

0 c.,1 Z50

5102 wt%

Figure 28.Silica (Si02) versus alkali (Na20 + K20) plot of Bransfield Strait and South Shetland Islands samples.Squares and triangles are dredge samples.Diamonds are King George Island.Octagons are Penguin Island.Asterisks are Deception Island.lB line is the alkali/tholeiite line from Irvine and Baragar (1971). Appendix II 92 Appendix II.Birkenmajer and Keller. submitted

Pleistocene Age of the Melville Peak Volcano, King George Island, West Antarctica, by K-Ar Dating

Krzysztof BIRKENMAJER1 and Randall A. KELLER2

1 of Geological Sciences, Polish Academy of Sciences, Senacka 3, 31-002 Kracow, Poland

2College of Oceanography, Oregon State University, Corvallis, Oregon 97331-5503 USA 93 Summary

Potassium-argon dating of basaltic lavas from the extinct Melville Peak volcano (King George Island, South Shetland Islands, West Antarctica) yielded mid to late Pleistocene ages for the first phase of activity (296,000 ± 27,000; 231,000 ± 19,000; 72,000 ± 15,000 years).The second phase of activity (subordinate in importance) may be either late Pleistocene or early Holocene in age. This is the first radiometric evidence for mid-late Pleistocene volcanic activity associated with rifting of the Bransfield Strait.

Introduction

Melville Peak (549 m a.s.1.) is an extinct Quaternary stratocone situated near the eastern corner of King George Island, South Shetland Islands (Figure 29).It's volcanic nature was recognized by Barton [1,2]; Hawkes [3] described its lavas as olivine basalt with diopsidic augite clots.Birkenmajer [4] carried out detailed geological mapping of the area, and established the structure and succession of volcanic events at Melville Peak.He concluded that the older part of the volcano (first phase) is probably late Pleistocene in age while the younger part (second phase) could correlate with the Holocene Deacon Peak Formation of Penguin Island [5,6]. The present paper reports on potassium-argon dating of the first phase basaltic lavas of the Melville Peak stratocone. The samples were collected by the first author in 1981 during the 5th Polish Antarctic Expedition to H. Arctowski Station; their location is shown in Figures 32 and 33. The K-Ar dating was performed by the second author at the College of Oceanography, Oregon State University, Corvallis, Oregon.

Geologic Setting.The Melville Peak volcano is a basaltic stratocone about 3 km in diameter at the base and nearly 550 m high, situated at the head of Melville Peninsula (Figures 30 and 31). Its internal structure is relatively well exposed due to glacial and marine erosion, the best sections being available at Sherratt Bay (Figure 32), and at a ridge leading from Danowski Glacier to the peak of the mountain (Figure 33). The basement of the volcano is the Moby Dick Group sediments and volcanics: fossiliferous glacio-marine deposits of the Cape Melville Formation (Lower Miocene), underlain by fossiliferous marine tuffaceous sediments of the Destruction Bay formation (Lower Miocene), and those by terrestrial basaltic lavas of the Sherratt Bay Formation (probably pre-Miocene).The whole Moby Dick Group is densely intruded by a swarm of andesite dykes which yielded Lower Miocene K-Ar ages of about 20 Ma [9].The dykes are displaced stepwise by transverse faults.An erosional unconformity separates the Melville Peak volcanics from various lithostratigraphic units of the Moby Dick Group [7-10] (see Figure 31).

Two volcanic phases have been recognized at Melville Peak [411. The first phase (Hektor Icefall Formation) is represented by black basaltic lavas and flow-breccias alternating with gray to yellow well- cemented large-scale cross-stratified pyroclastics. The latter grade from coarse explosion breccias and agglomerates through lapilli and tuff-sandstones to pelitic tuffs, and consist of fragmented basalt and clasts of baked pelitic sediment from the substratum of the stratocone.The lavas are massive, aphanitic, or vesicular; they show well-developed columnar jointing often arranged fan-wise. The central part of the volcano is occupied by a vertical plug of gray porphyritic augite-olivine basalt.A basaltic dyke exposed on the eastern slope of the volcano (Figure 30) has also been included in the first phase volcanics. The second phase volcanics (tentatively correlated with the Deacon Peak formation of Penguin Island[4,5,6])are represented by red basaltic agglomerates and scoriaceous lavas which fill a vent exposed on the eastern ridge of the mountain (Figures 30 and 33). The agglomerate consists mainly of volcanic bombs and lava cakes. The interpreted stratigraphy is given in Table 12. Radiometric Dating.Three basalt lavas from the first phase of volcanism have been K-Ar dated; two from the base of the stratocone, above Sherratt Bay (Samples A-372 and A-375; Figure 32), and a third from the eastern ridge of Melville Peak (Sample A- 406; Figure 33). The samples were dated by conventional K-Ar techniques [19]. Sample A-372 and A-375 originally contained large olivine phenocrysts, but these were removed out of concern for the possibility that they contained excess 40Ar.After olivine removal the samples could be classified as fine-grained basalts and basaltic andesites with no glass or alteration, and less than 3% phenocrysts. Potassium concentrations were determined by atomic absorption spectrophotometry.Argon isotope data were collected on an AEI- MS1O mass spectrometer with an on-line vacuum extraction system. Total fusion of rock chips was obtained using an external induction radio frequency heating coil.The ratios of 36Ar/38Ar and 40Ar/38Ar were measured, and the absolute abundance of 40Ar was determined by adding a spike with a known volume of 38Ar.Precise monitoring of atmospheric 4OAr/36Ar allowed very small amounts of radiogenic 40Ar to be detected in the sample.A correction factor for instrumental mass fractionation was obtained by measuringan air spike with an accepted 40Ar/36Ar of 295.5. The results of the dating are shown in Table 13.The dates fall in the middle and late Pleistocene (296x lOto 72 x 10years). The second phase of volcanism was not dated because it overlies the first phase, and if it is contemporaneous with the volcanism on Penguin Island, it would be too young to obtaina reliable date by the K-Ar method. Significance of Dating.Caic-alkaline island arc volcanism in the South Shetland Islands related to subduction of oceanic crust under the Antarctic plate ceased in the Lower Miocene, about 20 million years ago [16].The rifting in the western part of the Antarctic Peninsula that produced the Shetland microplate started nearly simultaneously, as evidenced by a system of antithetic faults along the outer margin of the Bransfield Strait, on King George Island, which displace Upper Oligocene and older rocks, and are intruded by basaltic and andesitic dykes K-Ar dated at between 21.8 and 14.4 Ma, early to middle Miocene [16, 17]. Quaternary volcanism in the Bransfield Strait occurs along two subparallel structural lines: the Deception-Bridgeman line, and the Penguin (Penguin-Melville) line.The Deception-Bridgeman line includes active (Deception Island) and inactive (Bridgeman Island) subaerial volcanoes, with a chain of seamounts between them (Figure 12). The Penguin line is represented by two subaerial volcanoes, Penguin Island (dormant) and Melville Peak (extinct), situated close to, or at the coast of, King George Island. The Quaternary volcanoes of the Bransfield Strait area produce three distinct volcanic suites: alkaline and caic-alkaline suites occur on volcanic islands of Penguin, Deception and Bridgeman, and on Melville Peak, while the seamounts are composed of tholeiites [13, 14].This difference may reflect not only the youth of the basin, but also the crustal structure below the volcanoes; normal continental crust some 30 km thick occurs below the Melville and Penguin volcanoes, while highly anomalous, modified crust has been recognized below the Bransfield basin [15]. Radiometric evidence shows there is a long time gap in volcanism in the Bransfield Strait area between the early to middle Miocene volcanic events and the recent volcanic activity at Penguin Island (late Pleistocene) [6].The middle to late Pleistocene K-Ar dates reported herein from the Melville Peak volcano shorten this gap to the late Miocene through early Pleistocene time span.

Acknowledgements

The K-Ar dates were determinedwith the help of Dr. Robert A. Duncan at Oregon State University, Corvallis, Oregon, USA.Randall Keller was supported by National Science Foundation grant DPP86- 14022 to Dr. Martin R. Fisk. INSTITUTE OF GEOLOGICAL SCIENCES, POLISH ACADEMY OF SCIENCES, SENACKA 3,31-002 CRACOW, POLAND (K. B.) (INSTYTUT NAUK GEOLOGICZNYCH PAN)

COLLEGE OF OCEANOGRAPHY, OREGON STATE UNWERSITY, CORVALLIS, OREGON 97331-5503, U.S.A. (R. A. K.)

R e fe r en c e s

[1] C.M. Barton, The geology of King George Island, South Shetland Islands.Pre!. Rept. Falkd. Is!. Dep. Surv., 12 (1961), 1-18. [2] C.M. Barton, The geology of South Shetland Islands, III.The stratigraphy of King George Island.Sci. Repts. Brit. Antarct. Surv., 44 (1965), 1-33. [31 D.D. Hawkes, The geology of the South Shetland Islands, I.The petrology of King George Island.Sci. Rept. Falkd. Is!. Dep. Surv., 26 (1961),1-28. [4] K. Birkenmajer, Structural evolution of the Melville Peak volcano, King George Island (South Shetland Islands, West Antarctica).Bull. Acad. Pol. Sci., Terre, 29 (1981), 341-351. [5] K. Birkenmajer, Age of the Penguin Island volcano, South Shetland Islands, West Antarctica.Ibid., 27 (1979), 69-76. [6] K. Birkenmajer, The Penguin Island volcano, South Shetland Islands (Antarctica): Its structure and succession.Stud. Geol. Polon., 74 (1982), 155-173. [7] K. Birkenmajer, Geology of the Cape Melville area, King George Island (South Shetland Islands, Antarctica): Pre-Pliocene glaciomarine deposits and their substratum.Ibid., 79 (1984), 7-3 6. [8] K. Birkenmajer, E. Luczkowska, Foraminiferal evidence for a Lower Miocene age of glaciomarine and related strata,Moby Dick Group, King George Island (South Shetland Islands, Antarctica).Ibid., 90 (1987), 81-123. [9] K. Birkenmajer, A. Gazdzicki, H. Kreuzer, P. Muller, K-Ar dating of the Melville Glaciation (Early Miocene) in West Antarctica. Bull. Pol. Acad. Sci., Earth-Sci., 33 (1985), 15-23. [10]K. Birkenmajer, B. Soliani, Jr., K. Kawashita, Early Miocene K-Ar age of volcanic basement of the Melville glaciation deposits, King George Island, West Antarctica.Ibid., 36 (1988), 25-34. [11]0. González-Ferrán, Y. Katsui, Estudio integral del volcanismo cenozoico superior de las islas Shetland del Sur, Antarctica.Ser. Cient. Inst. Antárt. Chil.,1 (1970), 123-174. [12]S.D. Weaver, A.D. Saunders, RJ. Pankhurst, J. Tarney, A geochemical study of the magmatism associated with the initial stage of back-arc spreading: The quaternary volcanics of Bransfield Strait, from South Shetland Islands.Contrib. Mm. Petrol., 68 (1979), 151-170. [131R.A. Keller, M.R. Fisk, W.M. White, K. Birkenmajer, Late tertiary- quaternary transition from arc to back-arc volcanism, South Shetland Islands, and Bransfield Strait, Antarctica.EOS Trans. Amer. Geophys. Un., 69 (1989),p. 1471. [14]M.R. Fisk, Back-arc volcanism in the Bransfield Strait, Antarctica, Jour, of South Amer. Earth Sci. [submitted]. [15]A. Guterch, M. Grad, J. Janik, E. Perchué, J. Pajohel, Seismic studies of the crustal structure in West Antarctica, 1979-1980: Preliminary results.Tectonophysics, 114 (1985), 411-429. [16]K. Birkenmajer, King George Island.In: I.W.D. Dalsiel et al. (eds.), Field Trip Guidebook T 180, Tectonics of the Scotia Arc. 28th Inst. Geol. Congr. (Washington, D.C., 1989), 114-121. [17]K. Birkenmajer, E. Soliani, Jr., K. Kawashita, Geochronology of Tertiary glaciations on King George Island, West Antarctica. Bull. Pol. Acad. Sci., Earth-Sci. [in press]. [18]W.A. Ashcroft, Crusta! structure of the South Shetland Islands and Bransfield Strait.Brit. Antarct. Surv., Sci. Repts, 66 (1972), 1-43. [19]G.G. Dairymple and M.A. Lanphere, Potassium-Argon Dating. W.H. Freeman Co., San Francisco, (1969), 258pp. K. Birkenmajer & R.A. Keller Objasnienia figur itabeli 610 :--- volcanoes

E /7 p 6 C / S (atlot 62°3O'S) I /O 600 Z [4: ; '-o ---o C MELVILLE PEAKI 627 1yr

d I,,- \\Y çtç JJ/F

Loa /) LjI \ 1Oo (fAH ELF ,T4200+ OOIcT(

Figure 29.Subaerial and submarine volcanoes (seamounts) in the Bransfield Strait area. Submarine relief interpreted from Ashcroft [18J: submarine escarpments barbed; land stippled; depths in metres. 0 - JWrona strike and dip jtocactics 1 ., "5.8uttress In and lava / o00 (I) horizontal }C I, 4 ±strata A. \&dyke // fault I I / Quoternary / Liisediments plug I If Lava 4 2 I' Crag -; younger dykes !1!IITJ la flo I V1st older I / vent dykes PYroclastksj MELViLLEPEA( 2nd phase Moby Dick vent Group

1' lOi I - * "4.. aiIw S )A.I5VLS Ice f.0' a '49 4-. if4 :i: 39v

SHE R RAiiTiIII!IItI1

Figure 30.Geological map of the Melville Peak volcano, after Birkenmajer [4]. w E IWNW ESE

-x MelvillePeak MelvillePenin5ula

o OO OOOn,

plug younger Cape Metville Fm- . c d2dykes 0 lava flows I:: d older dykes Sherratt Bay Fm- rcicscs L:-:j

Figure 31.Geological cross-section of the area between Melville Peak and CapeMelville, after Birkenmajer [7]. 103

Lii:iI

Figure32. Exposure in the Melville Peak stratocone above Sherratt Bay (after Birkenmajer [41).Numbers in boxes refer to K- Ar dated samples (see Table 12). 104

345

0. /,/._;/

/ 2nd phasevent /

Figure33. Exposure in the Melville Peak stratocone, eastern ridge (after Birkenmajer [4]).1, 2 - first phase volcanics and volcaniclastics (1 stratified yellow agglomerates and tuffs; 2 basalt lava); 3, 4 second phase volcanics (3 red scoriaceous lava; 4 red lava-agglomerate).Number in box refers to K-Ar dated sample (see Table 12). Table 12.Summary of structural evolution of the Melville Peak volcano (after Birkenmajer [4]), and K-Ar ages of the first phase volcanics.

E V£NT S IPHASESAMPLE:K-ArAGE(YRS= 0. Fornation of paraitic stratocone: P red explosion agglomerate Sand scotia EDEACON 2nd 14 PEAK (?)C. Farasitic vent explosion, partial G destructionof older stratocone 0

TlTftIUhR F°1T1TflTh1rniTTftflhliTIflhl B. Emplacement of plug cind

dyke . 1st z HEKTOR L A. Formation of m&n stratocone: A-375= 72,000± 15,000 AICEFALL black basaltic lava flows, yellow A-372=230,000±,000 N explosion agglomerates, grey j 0 tuffetc. AL06296,000±27,0OO - -_J_ _ - - -- -'------(base: Moby Dick Group, Lower Miocene 23-20 Ma)

C 106 Table 13.Potassium-argon ages of Melville Peak samples.

Sample wt%K Radiogenic 40Ar Percent Radio-Age±la x 10cc/g genic 40Ar xiOyrs

A372 0.694 6.229 6.7 231±19 A375 0.767 2.137 2.2 72±15 A406 0.802 9.240 5.3 296±27 Appendix III 107 Appendix III.Complete Petrographic Descriptions

A172 olivine basalt PHENOCRYSTS PERCENT SIZE. mm Morphology COMMENT OlIvine 0.5 <0.9 subeuhedral Plagioclase 15 <1.2 euh.subeuh some large ones zoned Pyroxene 5 <2.3 euh-sub Oxide v. rare <0.05 euh.sub In olivine GROUNOMASS OlIvine <1 <0.2 euh-anh Plagioclase 10 <0.2 euh-sub Pyroxene <1 <0.2 euh-anh Oxide v.v. rare <0.2 euh-sub Glass none Vesicles 10 <5 subrnd-irreg Comment remainder Is microcrystallin A174 olivinebasalt PHENOCRYSTS PERCENT SIZE, mm Morphology COMMENT 011vine 5 2 Euh-subhedralpossibly cumulate Plagloclase 10 <2 Euh- subhedrallarge grains zoned, with olivino Pyroxene 3 <2 Euh.Subhedralzoned, possibly cumulate Oxide <1 <0.3 Euh.SubhedralspInel In divine

GROUNOMASS

Olivine 1 <0.2 euh.subhedral Plagioclase 20 <0.2 euh-subhedral

Pyroxene 1 <0.2 euh-aubhedral Oxide 3 <0.2 ouh-subhedral Glass none VesIcles 10 <2 round-irreg. Comment remainder is microcrystallim undmass A260 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 5 2 Euh-subhedralpossibly cumulate Plagioclase 10 <2 Euh- subhedrsl 'large grains zoned, with olivine Pyroxene 3 <2 Euh-Subhedralzoned, possibly cumulate Oxide <1 <0.2 Euh-Subhedralspinel In olivine

GROUNOMASS

Olivine 1 <0.2 euh.subhedral Plagioclase 20 <0.2 euh-subhadral

Pyroxene 1 <0.2 euh-subhedral Oxide S 0.2 euh-subhedrat Glass none Vesicles 0% Comment iemainder Is microcrystalline undmass A261 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 7 <2.5 euh-sub Plagioclase 10 <1.5 euh-subhedral Pyroxene 5 <3 euh-sub Oxide <1 <0.2 euh.sub

GROUNDMASS Olivine 2 <0,1 euh.anh Plagioclase 15 <0,1 euh-sub Pytoxene 2 <0,1 euh-anh Oxide 3 <0.03 euh-sub Glass Vesicles

Comment remainder Ismicrocrystalline S A263 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 15 <4 euh-sub Plagioclase 7 1.1 euh-sub Pyroxene v. rare 0.2 ouh-sub Oxide 2 <02 euh-sub

GROUNOMASS Olivine 2 <0.05 euh-anh Plagioclase 7 <0.1 euh.sub Pyroxene Oxide <1 <0.02 euh-sub Glass Vesicles Comment remainder is microcrystallin A372 olivine basalt PHENOCRYSTSPERCENT SiZE, mm Morphology COMMENT Olivine 5 euh-sub some gloms w/px Plag4oclase 2 euh.subhedral

Pyroxene 1 <1 euh-sub many are zoned Oxide rare <0.05 euti-sub In olivines

GROUNDUASS Olivine 3 <0.02 sub-anh Plagioclase 10 <0.1 euh-sub Pyroxene

Oxide 1 <0.02 euh-sub Glass Vesicles 7 <2.5 round-irreg Comment remainder is microcrystallin A375 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine 2 <2.5 euh-sub in gloms w/cpx

Plagioclase 1 <0.8 euh-sub Pyroxene 2 <3.0 euh-aub some are zoned Oxide rare <0.03 euh-sub in oily

GROUNOMASS

Ohvine 1 <0.03 euh-anh Plagioclase 7 <0.1 euh-sub

Pyroxene 1 <0.05 sub-anh

Oxide 1 <0.01 euh-sub Glass Vesicles 5 <2.2 md-linear Comment remainder is microcrystaUin A405 red-brown vesicular basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT

Otryine 1 <2.5 euh-sub some w/iddingslte alteration Plagioclase 0.5 2.3 euh-sub Pyroxene rare <0.5 euh-sub In gloms w/oliv Oxide

GROUNOMASS

Olivine 1 <0.2 euh-sub Plagioclase 1 5 <0.2 euh-sub Pyroxene Oxide Glass Vesicles 25 <1.5 irreg-submnd Comment remainder is microcrystalline 109

A406 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 2 <2 euh-sub some gloms wlpx Plagloclase 0.5 <1 euh-subhedral

Pyroxene 1 <2 euh-sub In gloms w/oliv Oxide

GROUNDMASS Olivine 1 5 <02 sub-anh Plagioclase 30 <02 euh-anh Pyroxene Oxide 15 <01 euh-sub Glass Vesicles Comment remainder is microcrystallinegroundmass BS292-13 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 5 1 euh-sub Plagioclase Pyroxene Oxide rare .1 euh-sub

GROUNDMASS Olivine 2 0.1 sub-euh Plagioclase 2 <0.15 sub-euh Pyroxene Oxide Glass 5 Vesicles 30 <4 rnd-irreg Comment some vesicles contain glass; remainder is microryst. groundmass BS292-14 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 5 <2 euh-sub Plagioclase Pyroxene Oxide rare .1 euh-sub

GROUNOMASS Olivine 2 0.1 sub-euh Plagioclase 2 <0.15 sub-euh Pyroxene Oxide Glass 5 Vesicles 30 <4 rnd-irreg Comment some vesicles contain glass; remainder is rnicroryst.groundmass l3S292-17 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 2 1.5 Euh-subhedqal Plagioclase Pyroxene Oxide rare <0.1 euh in oliv and groundmass

GROUNDMASS Olivine Plagioclase 15 <0.3 Pyroxene Oxide Glass Vesicles 40 cS Comment remainder is microcrystalline groundmass 110

BS292-18 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 5 4.S euh-sub some altered along cracks Plagioclase Pyroxene

Oxide 1 <0.05 euh-sub

GROUNDMASS Olivine Plagioclase 2 .cO.2 sub-euh Pyroxene Oxide Glass 10 Vesicles 40 <2 rnd-irreg Comment some vesides have glassy rims; remainder microcryst.pmass. BS292-21 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 7 4 euh-sub Plagioclase rate <07 euhedral in gloms w!olivine Pyroxene Oxide

GROUNDMASS Olivina Plagioclase 2 <0.3 sub-euh Pyroxene Oxide Glass 10 Vasicles 40 <2 rnd-irreg Comment remainder is microcrystallinegroundmass BS29 2.24 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine I 1 Euh-subhedral Plagloclase 2 <1 euh-subhedral Pyroxene Oxide v. tare 0.3 euh in oliv

GROUNDMASS Olivine Plagioclase 15 <0.3 Pyroxeno Oxide Glass 15 Vesictes 30 <2 round to irreg. Comment remainder is microcrystallinegroundmass BS292-25 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 3 2.5 euh-sub some are skeletal Plagioclase Pyroxene Oxide rare 0.1 euh-sub

GROUNDUASS Olivine Plagioclase 2 <0.15 sub-euh Pyroxene Oxide Glass 10 Vesicles 40 <2 rnd-irreg Comment remainder is microcrystalline groundmass 111

BS292-3 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 5 Euh-subhedral Plagioclase Pyroxene Oxide rare 0.05 euh in oliv and groundmass

GROUNOMASS Olivine Plagioclase 10 <0.3 Pyroxene Oxide Glass Vesicles 20-50 <4 Comment remainder is microcrystallinegroundniass BS292-33 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 3 2.5 euh-sub some are skeletal Plagioclasa Pyroxene Oxide rare 0.I euh-sub

GROUNDMASS Olivine Plagioclase 2 <0.15 sub-euh Pyroxene Oxide Glass 10 Vesicles 40 rnd-irreg Comment remainder Is microcrystallinegroundmass BS292-37 basalt PHENOCRYSTSPERCENT SIZE, aim Morphology COMMENT Olivine I 1.5 euh-sub Plagloclase Pyroxene Oxide rare 0.05 euh-sub

GROUI4OMASS Otivine Plagioclase 2 <0.3 aub-eub Pyroxene Oxide Glass 10 Vesicles 40 <10 rnd-irreg Comment remainder Is microcrystallinegroundmass 8S292-43 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 3 2 euh-aub Plagioclase Pyroxene Oxide rare <0.1 euh-sub

GROUNOMASS Olivine Plagioclase 2 0.1 sub-euh Pyroxene Oxide Glass 10 Vesicles 30 c3 rnd-irreg Comment remainder is microcrystalline groundmass 112

BS292-53 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 3 <2.5 euh-sub Plagioclase rate 05 subhedral Pyroxane Oxide rare S0.05 euh-sub

GROUNOMASS Olivine Plagioclase 2 <0.3 sub-euh Pyroxene Oxide Glass 10 Vesicles 20 <4 rnd-irreg Comment remainder Is microcrystalline groundmass BS292-54 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 2 <2 euh-sub some are altered along cracks Plagioclase Pyroxene Oxide rare <0.1 euhadral

GROIJNDMASS Olivine Plagioclase 2 <0.2 sub-euh Pyroxene Oxide Glass 10 Vesicles 25 8 rnd-irreg Comment remainder is microcrystalline groundmass BS297-1 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT

Olivine 1 2.5 Euh-subhedral Plagioclasa 5 <1 euh-subhedralaome gloms wfolivine Pyroxene Oxide v. rare <0.05 euh in oliv and gmasa

GROUNDMASS Olivine 5 0.1 sub-anhedral Plagioclase 10 <0.3 sub-anh Pyroxene Oxide Glass Vesicles 20 8 round to irreg. Comment remainder is microcrystalline groundniasa BS297-2 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine I 2.5 euh-sub Plagioclase 5 <1 euh-sub aome in gloms w/oliv Pyroxene Oxide v. rare <0.05 euhedral

GROUNOMASS Olivine 5 0.1 sub-anh Plagioclaso 10 <0.3 sub-anh Pyroxene Oxide Glaaa Vesicles 20 4 rnd-irreg Comment remainder is microcrystalline groundmass 113

8S297-3 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine rare 0.6 euh-sub one 3mm skeletal olivina Plagloclase 3 0.6 euh-sub some in gloms w/oliv Pyroxene Oxide v. rare <0.05 sub-eub

GROUNOMASS Olivine I 0.1 euh-sub

Plagioclase 1 0.2 euh-sub Pymxene Oxide Glass 70 Vesicles 20 1.5 rnd-subrnd Comment remainder Is microcrystalline groundmass BS297-4 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine rare O.5 euh-anh large ones mostly resorbed Plagioclase I .6 euh-sub Pyroxene Oxide

GROUHDMASS Olivine 2 <0.1 euh-sub

Plagioclase 1 <0.1 euh-sub Pyroxene Oxide Glass 30 Vesicles 30 <4 rnd-irreg Comment remainder Is microcrystalline groundmass BS297-5 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine

Plagioclase 1 .9 euh-sub common orientation Pyroxene Oxide 0.5 <0.05 euh in groundmass

GROUNDMASS

Olivine 1 0.1 euh-anh Plagioclase 2 <0.2 euh-sub Pyroxene Oxide Glass Vesicles 20 1 rnd-subrnd Comment remainder is microcrystallinegroundmass BS297-6 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine rare <2 euh-sub

Plagioclase 1 c04 euh-sub Pyroxene Oxide single one 0.15 euhedral in groundmass

GROUNDMASS Olivine I 0.1 euh-sub plagioclase 2 0.2 euh-sub Pyroxene Oxide Glass Vesicles 15 2 rnd-irreg Comment remainder ismicrocrystalline groundmaas 114

BS300-1O basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine v. rare <1 sub-anh Plagioclase rare <0.9 euh-sub Pyroxene Oxide

GROUNOMASS Olivine 5 .1 euti-anh Plagloclase 20 0.2 euh-sub Pyroxene Oxide Glass Vesicles 20 <0 rnd-irreg Comment remainder Ismicrocrystallinegroundmass BS300-13 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine v. rare 0.5 euh-sub Plagloclasa I .8 euh-sub Pyroxene Oxide

GROUNOMASS Olivine I <0.1 euh-anh Plagioclase 15 <0.2 euh-sub Pyroxene Oxide Glass Vesicles 15 <2.5 rnd-irreg Comment remainder ismicrocrystallinegroundmass BS300-15 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine

Plagioclase 1 <1 euh-sub Pyroxene Oxide

GROUNOMASS Olivine 3 <0.1 euh-anh Plagioclase 15 0.2 euh-sub Pyroxene Oxide Glass Vesicles 10 5 rnd-irreg Comment remainder Ismicrocrystallinegroundmass 8S300-16 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine

Plagioclase 1 <1 euh-sub Pyroxene Oxide

GROUNDMASS Olivine 3 <0.1 euh-anh Plagioclase 15 0.2 euh-sub Pyroxeno Oxide Glass Vesicles 15 5 rnd-irreg Comment remainder is microcrystalline grouodmass 115

6S300-18 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine

Plagioclase 1 <1 euh-sub Pyroxene Oxide

GROUNDMASS Olivine 3 <0.1 euh-anh Plagioclase 15 0.2 euh-sub Pyroxene Oxide Glass Vesicles 15 5 rod-irreg Comment remainder Ismicrocrystallinegroundmass 8S300-19 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine

Plagioclase v. rare 1 euh.sub Pyroxene Oxide rare <0.05 euh-sub in vesicles only

GROUNDMASS Olivine <1 0.1 euh-anh Plagloclase 20 0.2 euh.sub Pyroxene Oxide Glass Vesicles 25 10 rnd-irreg Comment remainder Ismicrocrystallinegroundmass BS300-2 basalt PIIENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine Plagloclase v. rare 0.3 euh-subhedral Pyroxene Oxide

GROUNDMASS Olivine <1 <0.1 euh-anh Plagioclase 7 0.2 euh-sub Pyroxene Oxide Glass 20.80 Vesicles 25 c3 round to irreg. Comment remainder ismicrocrystallinegroundmass BS300-21 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine v. rare <1 euh-sub Plagioclase 2 <1.5 euh-sub Pyroxene Oxide v. rare <0.05 euh-sub in gmass

GROUNOMASS

Olivine 1 <0.1 euh-anh Plagioclase 10 0.2 euh-sub Pyroxene Oxide Glass 10 Vesicles 5 3 rnd-irreg Comment remainder ismicrocrystallinegroundmass 116

BS300-7 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine Plagioclase v. rare <0.5 euh-sub Pyroxene Oxide

GROUNDMASS

Olivine 1 <0.1 euh-anh Plagioclase 15 0.2 euh-sub Pyroxene Oxide Glass Vesicles 35 <10 rnd-irreg Comment remainder Is microcrystallinegroundmass BS309-1 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine <1 <0.4 euh-anh many are skeletal Plagioclaso 2 <0.9 euh-subhedralmost are corroded Pyroxene Oxide I 0.2 euh-sub some in olivine

GROUNOMASS Olivine <1 <0.1 euh-anh Plagioclase 7 <0.3 euh-sub Pyroxene Oxide Glass 25-85 Vesicles rare <1.5 round to subrnd Comment remainder is microcrystallinegroundmass BS3IO-1 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine I 1 euh-sub Plagioclase 2 1 euh-sub some in gloms w/olivine Pyroxene Oxide rare 0.05 eub-sub in olivine

GROUNDMASS Olivine <1 0.15 euh-anh Plagioclase 2 <0.3 euh-sub Pyroxene Oxide Glass Vesicles 15 4 rnd-irreg Comment remainder is microcrystallinegroundmass BS31O-13 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology GOMMENT Olivine 3 <1.5 euh-sub Plagioclase 2 <0.7 euh-sub Pyroxene Oxide v. rare <0.04 euh-sub in large olivines

GROUNOMASS Olivine <1 <0.1 euh-anh Plagioclase I <0.2 eub-aub Pyroxene Oxide Glass 5-90 Vesicles 5-20 .2 rnd-irreg Comment remainder is microcrystalline groundmass 117

8S310-2 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT

Olivine I 1 euh-sub Plagloclase 3 <1.5 euh-sub Pyroxene Oxide rare 0.05 euh-sub in olivine

GROUNDMASS Olivine <1 <0.1 euh-anh

Plagioclase 1 <0.2 euh-sub Pyroxene Oxide Glass 1-95 Vesicles 0-20 7 rnd-irreg Comment remainder is microcrystallinegroundmass BS3IO-21 olivine basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 3 <2.5 euh-sub Plagioclase 2 <0.7 euh-sub Pyroxene Oxide v. rare <0.04 euh-sub in large olivines

GROUNOMASS Olivine <1 <0.1 euh-anh

Plagioclase 1 <0.2 euh-sub Pyroxene Oxide Glass 5-90 Vesiclea 5-20 ..3 rnd-Irreg Comment remainder is microcrystallinegroundmass BS31O-26 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine <1 <1.5 euh-sub Plagioclase 2 <1 euh-eub Pyroxene Oxide rare <0.06 ouh-sub in some large olivines

GROUNDMASS Olivine 1 <0.1 euh-anh Plagioclase I 0.2 euh-anh Pyroxene Oxide Glass <1-90 Vesicles 0-30 <2 rnd-irreg Comment remainder is microcrystallinegroundmass BS31O-4 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT

Olivine 2 1 euh-sub Plagioclase 3 <1.5 euh-sub Pyroxene Oxide rare O.05 euh-sub in olivine

GROUNOMASS Olivine 7 0.1 sub-anh Plagioclase 10 0.3 sub Pyroxene Oxide Glass Vesicles 15 3 rnd-irreg Comment remainder is microcrystallinegroundmass 118

BS31O-7 basalt PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine <1 <1.5 Euh-subhedralmostly in glomerocrysts plagioclase 2 <1 euh.subhedral Pyroxene Oxide rare <0.05 euh-sub in large olivines

GROUNOMASS 011vine <1 <0.06 sub-anhedral Plagioclase 2 <0.2 subhedral Pyroxene Oxide Glass 0-70 Vesicles 25 <1.5 round to Irreg. Comment remainder ismicrocrystallinegroundmass 01 dacite PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine Plagioclase I 0.3 sub-anh

Pyroxene 1 0.3 sub-anh Oxide rare 0.1 euh-aub mag wrifm

GROUNDMASS Olivine Plagioclase 2 <0.1 sub-anh

Pyroxene 1 <0.1 sub-anh Oxide 7 <0.05 ouh-anh Glass Vesicles 30 <1 irreg-subrnd Comment remainder Ismicrocrystallinegroundmass 010 basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine v. rare <0.3 sub-anh some gloms w/plag Plagioclase 30 1 euh-subhedral Pyroxene v. rare <0.3 sub-anh Oxide 5 <0.2 sub-euh

GROUNOMASS Olivine 30 <0.1 sub-anh Plagioclase 20 <0.1 sub-anh Pyroxene Oxide 5 <0.05 euh-sub Glass Vesicles Comment remainder ismicrocrystallinegoundmass 012 dacite PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine rare <1.5 sub-euh in gloms w/plag Plagioclase 7 .2 sub-anh Pyroxene rare <1 sub-anh

Oxide 1 <0.3 sub-anh

GROUNOMASS Olivine Plagioclase 20 0.2 sub-snh Pyroxene rare 0.1 sub-anh Oxide 3 <0.05 sub-euh Glass Vesicles Comment remainder ismicrocrystallinegroundmass 119

013 hI-silicaalteration product PHENOCRYSTS PERCENT SIZE. mm Morphology COMMENT Olivine Plagioclase 5 <1 sub-euh Pyroxene Oxide

GROUNOMASS Olivine Plagioclase Pyroxene Oxide Glass Vesicles Comment 20% atz: 3% sulfide: remainder is highly altered 02 daclte PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine Plagioclase 3 euti-subtiedral

Pyroxene 1 <0.5 euh-anh in gloms w/plag

Oxide 1 <0.3 euh-sub

GROUNOMASS Olivine Plagioclase 10 <0.2 sub-anti Pyroxene Oxide 7 <0.08 euh-sub Glass Vesicles 10 subrnd-irreg Comment remainder is microcrystalline goundmass 05 dacite PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine 0.5 <1.5 sub-anti in gloms w/plag; embayed Plagioclase 3 <2.5 sub-euti Pyroxene rare 0.5 sub-anti

Oxide 1 0.3 sub-anl,

GROUNOMASS Olivine rare <0.1 anh-sub Plagioclase 20 0.2 sub-anti Pyroxene rare <0.2 sub-anti Oxide 2 <0.05 sub-euti Glass

Vesicles 5 1 Irreg-subrnd Comment remainder is microcrystalline groundmass 07 daclte PHENOCRYSTSPERCENT SIZE. mm Morphology COMMENT Olivine rare <2 sub-anti in gloms w/plag and oxide Plagioclase 5 2 sub-anti Pyroxene rare <0.5 sub-anti

Oxide 1 0.5 euti-anti

GROUNOMASS Olivine Plagioclase 2 .2 euti-sub Pyroxene rare 0.2 sub-anti Oxide 2 <0.05 sub-euti Glass

Vesicles 1 <1 irreg Comment remainder is microcrystalline groundmass 120

D8 dacite PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine 0.5 <1.5 sub-anh in glorna w/plag Plagioclase 3 <1.5 sub-euh Pyroxena 0.5 0.5 aub-anh Oxide 1 0.5 sub-anh

GROUNDMASS Olivine Plagioclase 30 <0.2 sub-anh Pyroxeno 3 <0.2 sub-anh Oxide 3 <0.05 sub-auh Glass Vesicles 2 <0.5 irreg Comment remainder Is microcrystallinegroundmass 09 dacita PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine v. rare sub-anh gloms w/plag Plagioclase S 2 euh-subhedral Pyroxene rare <1 sub.anh in gloms w/plsg Oxide I 0.5 sub-anh

GROUNOMASS Olivine Plagioclase 40 <0.2 sub-anh Pyroxene rare <0.2 sub-anh Oxide 7 <0.05 euh-sub Glass Vesicles Comment remainder is microcrystalline goundmass LHI brecciated basalt PHENOCRYSTSPERCENT SIZE, mm Morphology COMMENT Olivine rare <1 sub-anh altered Plagioclase 10 <1 euh-subhedralaltered rims Pyroxene Oxide 0.5 0.1 sub-anh

GROUNOMASS Olivine Plagioclase 10 <0.2 sub-anh Pyroxene Oxide Glass Vesicle s Comment remainder is alteration products Appendix IV 121

Appendix IV. Results of Normative Calculations

Western seamount Sample BS292-3BS292-13BS292-14 BS292-17 BS292-18 BS292-21 Quartz 0.00 0.00 0.00 0.00 0.00 0.00 Orthoclase 2.36 2.50 2.62 2.37 2.66 2.46 Albite 21.77 21.72 23.38 20.97 21.74 20.62 Anorthite 24.25 23.70 24.79 23.99 23.81 23.44 Nepheline 1.70 3.08 1.96 2.14 1.93 2.40 I{ypersthene0.00 0.00 0.00 0.00 0.00 0.00 Diopside 20.54 21.48 21.07 20.43 20.70 20.72 Olivine 23.04 23.04 23.12 24.03 24.32 24.20 Magnetite 1.54 1.55 1.55 1.56 1.57 1.58 lilmenite 2.15 2.17 2.22 2.14 2.19 2.15 Apatite 0.38 0.38 0.39 0.38 0.38 0.37

Western seamoun Sample BS292-24 BS292-25 BS292-33 BS292-37 BS292-43 BS292-53 BS292-54 Quartz 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Orthoclase 2.51 2.59 2.49 2.48 2.45 2.53 2.65 Albite 25.84 21.37 20.89 21.76 22.02 21.78 21.81 Anorthite 26.26 23.71 23.42 24.23 24.62 24.07 24.23 Nepheline 1.90 2.52 2.76 1.88 1.33 1.65 1.97 Hypersthene0.00 0.00 0.00 0.00 0.00 0.00 0.00 Diopside 21.55 20.99 21.04 20.70 20.40 20.49 21.16 Olivine 16.37 23.92 24.04 24.03 23.51 23.94 22.37 Magnetite 1.59 1.57 1.57 1.57 1.56 1.61 1.55 Illmenite 2.52 2.19 2.15 2.16 2.17 2.15 2.21 Apatite 0.38 0.38 0.38 0.38 0.38 0.51 0.38

Western seamoun Sample BS31O-1BS31O-2BS31O-4BS31O-7BS31O-13 BS31O-21 BS31O-26 Quartz 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Orthoclase 2.70 2.70 2.76 3.25 2.49 2.58 2.91 Albite 27.27 25.70 26.16 27.45 25.85 25.76 28.60 Anorthite 25.21 24.67 25.65 25.98 24.92 25.45 27.21 Nepheline 1.57 2.11 1.08 1.66 0.29 0.45 0.27 Hypersthene0.00 0.00 0.00 0.00 0.00 0.00 0.00 Diopside 21.09 21.31 20.60 21.35 19.99 20.08 20.72 Olivine 14.95 17.09 16.83 13.28 19.21 18.65 13.43 Magnetite 1.68 1.70 1.70 1.64 1.74 1.73 1.68 lilmenite 2.90 2.84 2.85 2.93 2.77 2.81 3.07 Apatite 0.45 0.39 0.40 0.43 0.39 0.40 0.47 122 Anpendix IV. (Continued)

Easternseamount Sample BS297-1BS297-2BS300-2BS300-7B5300-10 BS300-13 Quartz 0.00 0.00 0.01 0.00 0.00 0.00 Orthoclase 2.22 2.16 3.42 4.65 3.42 3.12 Albite 26.20 26.01 38.45 37.99 34.15 34.00 Anorthite 30.59 30.91 20.08 19.71 23.56 24.06 Nepheline 0.00 0.00 0.00 0.00 0.00 0.00 Hypersthene 11.51 11.42 15.77 14.63 11.52 11.07 Diopside 19.64 19.11 14.52 14.68 17.64 17.86 Olivine 4.12 5.45 0.00 0.79 3.07 3.48 Magnetite 1.26 1.33 1.86 1.85 1.63 1.60 Ilimenite 1.99 2.03 3.31 3.27 2.67 2.56 Apatite 0.28 0.27 0.53 0.54 0.40 0.38

Easternseamount SampleB5300-15BS300-16 BS300-18 BS300-19 BS300-21BS309-1 Quartz 0.00 0.00 0.00 0.41 0.00 0.81 Orthoclase 3.00 3.06 2.87 4.43 2.68 5.32 Albite 33.22 33.45 33.48 36.50 32.89 38.56 Anorthite 24.56 23.84 24.96 20.18 24.30 19.29 Nepheline 0.00 0.00 0.00 0.00 0.00 0.00 Hypersthene 13.59 13.02 12.90 15.86 13.39 15.20 Diopside 17.50 17.21 17.26 14.08 17.34 13.37 Olivine 1.74 2.03 2.46 0.00 1.83 0.00 Magnetite 1.61 1.65 1.61 1.84 1.62 1.82 lilmenite 2.56 2.66 2.60 3.28 2.56 3.32 Apaiite 0.38 0.41 0.39 0.54 0.38 0.51 123 Appendix IV. (Continued)

Penguin Island Sample A172 A174 A260 A261 A263 Quartz 0.00 0.00 0.00 0.00 0.00 Orthoclase 3.71 3.78 2.85 3.57 3.33 Albite 24.76 24.52 23.47 22.83 22.41 Anorthite 26.12 26.97 26.48 25.36 25.11 Nepheline 3.11 4.70 2.59 4.48 3.49 Hypersthene0.00 0.00 0.00 0.00 0.00 Diopside 16.87 17.55 18.25 18.34 17.25 Olivine 17.93 16.83 20.24 19.32 21.83 Magnetite 1.58 1.56 1.55 1.56 1.59 lllmenite 2.46 2.49 2.10 2.39 2.22 Apatite 1.16 0.67 0.53 0.63 0.69

King GeorgeIsland Melville Peak Low Head Sample A372 A375 A405 A406 LH1 Quartz 0.00 0.00 0.00 0.00 0.00 Orthoclase 4.23 5.09 5.38 5.35 1.44 Albite 24.90 27.88 25.70 25.40 15.70 Anorthite 26.59 29.22 29.68 29.34 29.58 Nepheline 0.00 0.00 0.00 0.00 3.24 Hypersthene6.60 3.72 5.36 4.44 0.00 Diopside 17.32 21.82 21.12 21.41 30.40 Olivine 15.81 9.85 8.47 8.86 13.49 Magnetite 1.30 1.27 1.29 1.28 1.55 lilmenite 1.73 1.85 1.75 1.70 1.46 Apatite 0.26 0.31 0.31 0.29 0.27

Deception Island Sample Dl D2 D5 D7 D8 D9 D10 D12 D13 Quartz 5.70 4.93 10.39 10.96 10.56 10.33 0.00 11.06 39.56 Orthoclase9.37 8.10 11.05 11.08 10.76 10.77 2.84 11.32 16.02 Albite 63.02 59.25 64.10 63.35 62.09 61.09 35.96 65.43 38.13 Anorthite 3.91 6.60 0.39 0.78 1.45 0.36 23.14 0.02 1.87 Nepheline 0.00 0.00 0.00 0.00 0.00 0.00 1.47 0.00 0.00 Hypersthene 6.28 8.14 3.99 4.06 4.53 4.16 0.00 3.73 1.01 Diopside 8.36 7.85 7.13 6.93 6.95 6.16 18.57 6.82 2.85 Olivine 0.00 0.00 0.00 0.00 0.00 0.00 10.80 0.00 0.00 Magnetite 1.00 1.16 0.80 0.81 0.84 0.78 1.59 0.77 0.13 Illmenite 1.87 2.32 1.21 1.26 1.32 1.06 3.23 1.12 0.23 Apatite 0.59 0.99 0.31 0.31 0.33 0.28 0.58 0.29 0.05