Plagioclase variation
Revision of known area of Dufek intrusion J . C. BEHRENDT U.S. Geological Survey Denver, Colorado 80225
D. J . DREWRY and E. JANKOWSKI Scott Polar Research Institute Cambridge University Cambridge, England
A. W. ENGLAND U.S. Geological Survey Reston, Virginia 22092
The data we collected along a 4,200-kilometer traverse during December 1978 substantially enlarged the known area of the Dufek layered mafic intrusion. An aeromagnetic and gravity survey in 1965 indicated that the 172±4 million-year-old intrusion covered 34,000 square kilometers, making it the second largest such structure in the world. Our 1978 data, however, indicate anomalies that can be traced about 200 kilometers farther north from the Forrestal Range and Dufek Massif than previously observed over the Filchner
Ice Shelf near Berkner Island at approximately 80°30'SI 45°W. Radar ice soundings enabled us to calculate Bouguer anomalies at locations where free air anomalies had been previously measured east of the Forrestal Range. The aeromagnetic and gravity data indicate that the Dufek intrusion probably extends southeastward to the vicinity of 83°45'S/43°W. Together, these results suggest a minimal areal extent of about 50,000 square kilometers. The Dufek intrusion is comparable in area to the Bushveld complex in Africa. As measured a few hundred meters directly over outcrops of the Dufek intrusion, observed aeromagnetic anomalies ranged in amplitude from about 50 to about 3,600 nT, thereby reflecting a variation in measured remanent magnetization and susceptibility extending over three orders of magnitude (Beck, 1972). Topography on the bedrock surface is as much as 4 kilometers based on depths beneath the Filchner Ice Shelf previously determined from seismic soundings compared with the highest peaks. The grounded ice is as much as 2.5 kilometers thick over the intrusion. Using surface and subsurface topography determined from the radar sounding profiles, we calculated magnetic models to fit the observed magnetic data compatible with the 5°-10° southeastward dip of the strata, which require normal and reversed magnetization ranging from 10_4 to 102 electromotive units per cubic centimeter. The models suggest initial cooling of the intrusion through the Curie isotherm through reversals in the Earth's magnetic field as proposed by Beck (1972) on the basis of paleomagnetic results. Reference Beck, M. E. 1972. Paleomagnetism and magnetic polarity zones in the Jurassic Dufek instrusion, Pensacola Mountains, Antarctica. Geophys. J . R. Astro. Soc., vol. 28, p. 49.
Petrologic studies of Dufek intrusion: Plagioclase variation KATHLEEN D. ABEL Pennsylvania Topographic and Geologic Survey Pittsburgh, Pennsylvania 15222
GLENN R. HIMMELBERG Department 01 Geology University of Missouri Columbia, Missouri 65221
ARTHUR B. FORD U.S. Geological Sun'e' Menlo Park, California 94025
This study is part of a continuing, detailed petrologic investigation of the Dufek intrusion, a stratiform body of mafic rock that makes up nearly the entire northern third of the Pensacola Mountains (Ford, 1976). The structure and rock textures of the intrusion and the relation of rock and cumulus-mineral chemistry to magmatic stratigraphy indicate an origin of the rocks by crystal accumulation on a magma-chamber floor (rocks of this nature are termed cumulates). One of our objectives, therefore, is a detailed investigation of the minerals and their chemical variation so as to develop an understanding of magma differentiation and cooling. Previous studies have investigated cumulus pyroxenes and iron-titanium oxide minerals (Himmelberg and
Ford, 1976, 1977). This study describes the distribution, chemistry, and fractionation trend of the cumulus plagioclase in the cumulate sequence and of noncumulus plagioclase in granophyre above the cumulates. Plagioclase is the dominant cumulus phase in most rocks of the intrusion. In the cumulate sequence, it is generally absent as a cumulus phase only in basal parts of thin, pyroxene cumulate (pyroxenite) layers. In most rocks, plagioclase also occurs as a postcumulus phase, either as small, irregular interstitial fillings or, more commonly, as rims around cumulus grains. Plagioclase from 30 samples containing previously analyzed pyroxene and iron-titanium oxide minerals were analyzed with an ARL EMX-SM electron microprobe (Abel, 1978). Natural feldspars of known composition were used as standards; matrix corrections were made by the procedures of Bence and Albee (1968) and the correction factors of Albee and Ray (1970). Anorthite content for all analyzed plagioclase shows a range of An 79 _ 49 . Most grains are compositionally zoned; cores are generally 1-2.5 percent more anorthitic than rims, but some are up to 6.5 percent greater. Some zoning may be cumulus, but most probably formed by postcumulus growth around original settled primocrysts. The anorthite content of plagioclase generally decreases with stratigraphic height (see figure), a fractionation trend that is characteristic in layered intrusions of this type (Wager and Brown, 1968). Significant exceptions to this overall trend are compositional reversals, with height, in the Neuburg Pyroxenite and Frost Pyroxenite members of the Aughenbaugh Gabbro, as well as between 20 and 100 meters and between 750 and 860 meters above the base of the Saratoga Gabbro. The fractionation trends of plagioclase and pyroxenes are significantly different (see figure). Whereas pyroxene trends show slight or no iron enrichment upward through a 1-kilometer-thick section of the Aughenbaugh Gabbro above the Neuburg Pyroxenite member, the anorthite content of plagioclase decreases steadily except for the reversals noted above. Reversals in the plagioclase trends are not paralleled by reversals in the pyroxene trends, except for the higher one in the Saratoga Gabbro. These differences may in large part be related to differences in the settling characteristics of the minerals. Plagioclase is generally less dense than the magma from which it crystallized (Bottinga and Weill, 1970; Campbell, Roeder, and Dixon, 1978), although it may be carried down along with denser minerals by magmatic currents (Irvine, 1978) or as part of composite chains that include denser minerals (Campbell, Roeder, and Dixon, 1978). Field evidence such as cut-and-fill channels shows that the pyroxenc cumulates of the Aughenbaugh Gabbro and plagioclase cumulates of the Saratoga Gabbro were deposited from currents moving along the floor of the magma chamber (Ford, 1976). The experiments of Irvine (1978) suggest that plagioclase may be carried upward in suspension currents, accumulate near the roof, and then be transported downward by density-current flow along chamber walls. Relations shown in the figure can be interpreted in terms of Irvine's experiments, assuming that pyroxenes continuously settle and are not episodically concentrated near the roof, as are some plagioclases, by local upward
CATION RATO MgjMcj F M; IN PYROXENES 03 04 05 06 07 08 Lexington X GXphy
0 e- Cq-urn poor
pyroxene
U
10
A 0 A I pX P0XPM Mlbp
U
00 0
bYg Py , oxeroe Mernbe, *
-
-------------
50 50 MOLE PEPCEC ANORTHITE N LL0CLOCLASE
70
80
Variation in anorthite content of plagioclase and in cation ratio, Mg/(Mg+Fe+ Mn), of pyroxenes with stratigraphic height in Dufek intrusion. Analyzed minerals are noncumulus In Lexington granophyre and cumulus below the Lexington.
return flow from density currents on the floor. Plagioclases of more anorthitic composition may thus reside for some time near the roof while later crystallized, less anorthitic plagioclase accumulates on the chamber floor. Accordingly, the plagioclase-trend reversals, where not paralleled by pyroxene-trend reversals, may be interpreted in terms of episodic downward transport of more anorthitic plagioclase by density currents. Once on the floor, most plagioclase grains are probably prevented from floating free again by the physical characteristics of the magma (Irvine, 1978). The plagioclase-trend reversal in the middle part of the Saratoga (;abbro appears to have a different origin. It is the only major reversal paralleled by pyroxenetrend reversals (see figure) and it occurs near a stratigraphic level at which field evidence suggests the emplacement of an additional magma batch at a late stage of consolidation of the intrusion (Ford et al., 1979). This finding of possible evidence for multiple emplacement of magma adds a new complexity to the interpretation of the record of fractional crystallization of the Dufek intrusion. This research represents an outgrowth of 1965-66 fieldwork on the Dufek intrusion that was supported by National Science Foundation grant GA 222 to the U.S. Geological Survey. The electron microprobe used in the study was bought with aid of National Science Foundation grant GA 18445 to the University of Missouri. 'I
References Abel, K. D. 1978. Plagioclases of the Dufek intrusion, Antarctica. M.A thesis, University of Missouri. Albee, A. L., and L. Ray. 1970. Correction factors for electron probe microanalysis of silicates, oxides, carbonates, phosphates, and sulfates. Analytical Chemistry, 42: 1408-14. Bence, A. E., and A. L. Albee. 1968. Empirical correction factors for the electron microanalysis of silicates and oxides. Journal of Geology, 76: 382-403. Bottinga, Y., and D. F. Weill. 1970. Densities of liquid silicate systems calculated from partial molar volumes of oxide components. American Journal of Science, 269: 169-82. Campbell, I. H., P. L. Roeder, and J . M. Dixon. 1978. Plagioclase buoyancy in basaltic liquids as determined with a centrifuge furnace. Contributions to Mineralogy and Petrology, 67: 369-77.
Potassium-argon ages of Dufek intrusion and other Mesozoic mafic bodies in the Pensacola Mountains R. W. KISTLER
and A.
B. FORD
U.S. Geological Survey Menlo Park, Cal!fornia 94025
Conventional potassium-argon (K-Ar) and argon-40/ argon-39 (40Ar/39 Ar) age determinations show that the stratiform gabbroic Dufek intrusion (Ford, 1976) is about the same age as tholeiitic diabase sills of the Pecora Escarpment (85°37'S168 040'W) at the south end of the Pensacola Mountains. The igneous activity corresponds with the main Early Jurassic activity of the Ferrar Group of Grindley (1963) elsewhere in the Transantarctic Mountains. Samples from the Dufek intrusion dated in our study include plagioclase of three plagioclase cumulates (anorthosites) from widely spaced stratigraphic intervals in the Dufek Massif and Forrestal Range; pyroxene from a pyroxene-cumulate (pyroxenite) layer in the Dufek Massif; and fine-. grained noncumulus whole-rock gabbro from an inferred chilled contact zone near Mount Lechner in the Forrestal Range. In addition, we also determined ages on plagioclase, pyroxene, and whole-rock samples from diabase sills of the Pecora Escarpment and basalt and diabase dikes of the Cordiner Peaks. The sills intrude the Permian Pecora Formation, and the dikes cut the Devonian Dover sandstone. and the overlying Gale mudstone (Ford et al., 1978). Conventional K-Ar ages on samples from the Dufek intrustion range from 174.1 ± 4.4, 171.2 ± 4.3, and
Ford, A. B. 1976. Stratigraphy of the layered gabbroic Dufek intrusion, Antarctica. U.S. Geological Survey Bulletin, no. l405-D. Ford, A. B., R. L. Reynolds, Carl Huie, and S. J . Boyer. 1979. Geological investigation of the Dufek intrusion, Pensacola Mountains, 1978-79. Antarctic Journal of the United States (this issue). 1-Iimmelberg, G. R., and A. B. Ford. 1976. Pyroxenes of the Dufek intrustion, Antarctica. Journal of Petrology, 17: 21943. Himmelberg, G. R., and A. B. Ford. 1977. Iron-titanium oxides of the Dufek intrusion, Antarctica. American Mineralogist, 62: 623-33. Irvine, T. N. 1978. Density current structure and magmatic sedimentation. In Carnegie Institution of Washington Yearbook, 77: 717-25. Wager, L. R., and G. M. Brown. 1968. Layered Igneous Rocks. Edinburgh, Scotland: Oliver & Boyd.
169.5 ± 4.2 million years for the plagioclase, through 111.9± 2.8 and 106.3 ± 2.7 million years for the wholerock gabbros, to 97.5 ± 2.4 million years for the pyroxene samples. In thin section, the plagioclase appears little altered, but the pyroxenes (augite and inverted pigeonite) show variable and commonly large effects of inversion and exsolution (Himmelberg and Ford, 1976). Kistler and Dodge (1966) suggested that the young K-Ar ages for pyroxene-.-compared to those for coexisting biotite, hornblende, plagioclase, and orthoclase from Sierra Nevada (California) quartz diorite plutons-may be related to ubiquitously observed exsolution lamellae, which decrease the effective diffusion dimensions and may permit low-temperature argon loss. Similarly, the considerably younger ages of Dufek intrusion samples containing pyroxene probably reflect low-temperature argon loss during or after subsolidus phase changes. To ascertain that these young pyroxene ages are not the result of inaccuracies in the measurement of lowK2 0 concentrations, pyroxene from two localities in the Dufek intrusion was also dated by the total-fusion 40Ar/ 39 Ar technique. Results showed that the young conventional K-Ar ages of samples containing pyroxene are probably not due to such analytical difficulties. The best present estimate for the age of the intrusion is therefore considered to be the average of the three plagioclase determinations: 171.6 ± 4.3 million years. This age is in close agreement with, and within the range of analytical uncertainties of, the value of 175 ± 5 million years reported by Elliot, Fleck, and Sutter (in press) for basalt flows of the Ferrar Group in the central Transantarctic Mountains. Samples of whole-rock holocrystalline chilled-contact basalt and of three plagioclase-pyroxene pairs from two diabase sills in the Pecora Escarpment yielded conventional K-Ar ages ranging from 223.1 ± 5.6 million years for the basalt to 177.7 ± 4.5 million years for a plagioclase. In contrast with the Dufek intrusion, the apparent ages of pyroxene in all three pairs are slightly greater than those of coexisting plagioclase and are probably affected by small but variable amounts of excess argon.
D. J . DREWRY and E. JANKOWSKI Scott Polar Research Institute Cambridge University Cambridge, England
A. W. ENGLAND U.S. Geological Survey Reston, Virginia 22092
The data we collected along a 4,200-kilometer traverse during December 1978 substantially enlarged the known area of the Dufek layered mafic intrusion. An aeromagnetic and gravity survey in 1965 indicated that the 172±4 million-year-old intrusion covered 34,000 square kilometers, making it the second largest such structure in the world. Our 1978 data, however, indicate anomalies that can be traced about 200 kilometers farther north from the Forrestal Range and Dufek Massif than previously observed over the Filchner
Ice Shelf near Berkner Island at approximately 80°30'SI 45°W. Radar ice soundings enabled us to calculate Bouguer anomalies at locations where free air anomalies had been previously measured east of the Forrestal Range. The aeromagnetic and gravity data indicate that the Dufek intrusion probably extends southeastward to the vicinity of 83°45'S/43°W. Together, these results suggest a minimal areal extent of about 50,000 square kilometers. The Dufek intrusion is comparable in area to the Bushveld complex in Africa. As measured a few hundred meters directly over outcrops of the Dufek intrusion, observed aeromagnetic anomalies ranged in amplitude from about 50 to about 3,600 nT, thereby reflecting a variation in measured remanent magnetization and susceptibility extending over three orders of magnitude (Beck, 1972). Topography on the bedrock surface is as much as 4 kilometers based on depths beneath the Filchner Ice Shelf previously determined from seismic soundings compared with the highest peaks. The grounded ice is as much as 2.5 kilometers thick over the intrusion. Using surface and subsurface topography determined from the radar sounding profiles, we calculated magnetic models to fit the observed magnetic data compatible with the 5°-10° southeastward dip of the strata, which require normal and reversed magnetization ranging from 10_4 to 102 electromotive units per cubic centimeter. The models suggest initial cooling of the intrusion through the Curie isotherm through reversals in the Earth's magnetic field as proposed by Beck (1972) on the basis of paleomagnetic results. Reference Beck, M. E. 1972. Paleomagnetism and magnetic polarity zones in the Jurassic Dufek instrusion, Pensacola Mountains, Antarctica. Geophys. J . R. Astro. Soc., vol. 28, p. 49.
Petrologic studies of Dufek intrusion: Plagioclase variation KATHLEEN D. ABEL Pennsylvania Topographic and Geologic Survey Pittsburgh, Pennsylvania 15222
GLENN R. HIMMELBERG Department 01 Geology University of Missouri Columbia, Missouri 65221
ARTHUR B. FORD U.S. Geological Sun'e' Menlo Park, California 94025
This study is part of a continuing, detailed petrologic investigation of the Dufek intrusion, a stratiform body of mafic rock that makes up nearly the entire northern third of the Pensacola Mountains (Ford, 1976). The structure and rock textures of the intrusion and the relation of rock and cumulus-mineral chemistry to magmatic stratigraphy indicate an origin of the rocks by crystal accumulation on a magma-chamber floor (rocks of this nature are termed cumulates). One of our objectives, therefore, is a detailed investigation of the minerals and their chemical variation so as to develop an understanding of magma differentiation and cooling. Previous studies have investigated cumulus pyroxenes and iron-titanium oxide minerals (Himmelberg and
Ford, 1976, 1977). This study describes the distribution, chemistry, and fractionation trend of the cumulus plagioclase in the cumulate sequence and of noncumulus plagioclase in granophyre above the cumulates. Plagioclase is the dominant cumulus phase in most rocks of the intrusion. In the cumulate sequence, it is generally absent as a cumulus phase only in basal parts of thin, pyroxene cumulate (pyroxenite) layers. In most rocks, plagioclase also occurs as a postcumulus phase, either as small, irregular interstitial fillings or, more commonly, as rims around cumulus grains. Plagioclase from 30 samples containing previously analyzed pyroxene and iron-titanium oxide minerals were analyzed with an ARL EMX-SM electron microprobe (Abel, 1978). Natural feldspars of known composition were used as standards; matrix corrections were made by the procedures of Bence and Albee (1968) and the correction factors of Albee and Ray (1970). Anorthite content for all analyzed plagioclase shows a range of An 79 _ 49 . Most grains are compositionally zoned; cores are generally 1-2.5 percent more anorthitic than rims, but some are up to 6.5 percent greater. Some zoning may be cumulus, but most probably formed by postcumulus growth around original settled primocrysts. The anorthite content of plagioclase generally decreases with stratigraphic height (see figure), a fractionation trend that is characteristic in layered intrusions of this type (Wager and Brown, 1968). Significant exceptions to this overall trend are compositional reversals, with height, in the Neuburg Pyroxenite and Frost Pyroxenite members of the Aughenbaugh Gabbro, as well as between 20 and 100 meters and between 750 and 860 meters above the base of the Saratoga Gabbro. The fractionation trends of plagioclase and pyroxenes are significantly different (see figure). Whereas pyroxene trends show slight or no iron enrichment upward through a 1-kilometer-thick section of the Aughenbaugh Gabbro above the Neuburg Pyroxenite member, the anorthite content of plagioclase decreases steadily except for the reversals noted above. Reversals in the plagioclase trends are not paralleled by reversals in the pyroxene trends, except for the higher one in the Saratoga Gabbro. These differences may in large part be related to differences in the settling characteristics of the minerals. Plagioclase is generally less dense than the magma from which it crystallized (Bottinga and Weill, 1970; Campbell, Roeder, and Dixon, 1978), although it may be carried down along with denser minerals by magmatic currents (Irvine, 1978) or as part of composite chains that include denser minerals (Campbell, Roeder, and Dixon, 1978). Field evidence such as cut-and-fill channels shows that the pyroxenc cumulates of the Aughenbaugh Gabbro and plagioclase cumulates of the Saratoga Gabbro were deposited from currents moving along the floor of the magma chamber (Ford, 1976). The experiments of Irvine (1978) suggest that plagioclase may be carried upward in suspension currents, accumulate near the roof, and then be transported downward by density-current flow along chamber walls. Relations shown in the figure can be interpreted in terms of Irvine's experiments, assuming that pyroxenes continuously settle and are not episodically concentrated near the roof, as are some plagioclases, by local upward
CATION RATO MgjMcj F M; IN PYROXENES 03 04 05 06 07 08 Lexington X GXphy
0 e- Cq-urn poor
pyroxene
U
10
A 0 A I pX P0XPM Mlbp
U
00 0
bYg Py , oxeroe Mernbe, *
-
-------------
50 50 MOLE PEPCEC ANORTHITE N LL0CLOCLASE
70
80
Variation in anorthite content of plagioclase and in cation ratio, Mg/(Mg+Fe+ Mn), of pyroxenes with stratigraphic height in Dufek intrusion. Analyzed minerals are noncumulus In Lexington granophyre and cumulus below the Lexington.
return flow from density currents on the floor. Plagioclases of more anorthitic composition may thus reside for some time near the roof while later crystallized, less anorthitic plagioclase accumulates on the chamber floor. Accordingly, the plagioclase-trend reversals, where not paralleled by pyroxene-trend reversals, may be interpreted in terms of episodic downward transport of more anorthitic plagioclase by density currents. Once on the floor, most plagioclase grains are probably prevented from floating free again by the physical characteristics of the magma (Irvine, 1978). The plagioclase-trend reversal in the middle part of the Saratoga (;abbro appears to have a different origin. It is the only major reversal paralleled by pyroxenetrend reversals (see figure) and it occurs near a stratigraphic level at which field evidence suggests the emplacement of an additional magma batch at a late stage of consolidation of the intrusion (Ford et al., 1979). This finding of possible evidence for multiple emplacement of magma adds a new complexity to the interpretation of the record of fractional crystallization of the Dufek intrusion. This research represents an outgrowth of 1965-66 fieldwork on the Dufek intrusion that was supported by National Science Foundation grant GA 222 to the U.S. Geological Survey. The electron microprobe used in the study was bought with aid of National Science Foundation grant GA 18445 to the University of Missouri. 'I
References Abel, K. D. 1978. Plagioclases of the Dufek intrusion, Antarctica. M.A thesis, University of Missouri. Albee, A. L., and L. Ray. 1970. Correction factors for electron probe microanalysis of silicates, oxides, carbonates, phosphates, and sulfates. Analytical Chemistry, 42: 1408-14. Bence, A. E., and A. L. Albee. 1968. Empirical correction factors for the electron microanalysis of silicates and oxides. Journal of Geology, 76: 382-403. Bottinga, Y., and D. F. Weill. 1970. Densities of liquid silicate systems calculated from partial molar volumes of oxide components. American Journal of Science, 269: 169-82. Campbell, I. H., P. L. Roeder, and J . M. Dixon. 1978. Plagioclase buoyancy in basaltic liquids as determined with a centrifuge furnace. Contributions to Mineralogy and Petrology, 67: 369-77.
Potassium-argon ages of Dufek intrusion and other Mesozoic mafic bodies in the Pensacola Mountains R. W. KISTLER
and A.
B. FORD
U.S. Geological Survey Menlo Park, Cal!fornia 94025
Conventional potassium-argon (K-Ar) and argon-40/ argon-39 (40Ar/39 Ar) age determinations show that the stratiform gabbroic Dufek intrusion (Ford, 1976) is about the same age as tholeiitic diabase sills of the Pecora Escarpment (85°37'S168 040'W) at the south end of the Pensacola Mountains. The igneous activity corresponds with the main Early Jurassic activity of the Ferrar Group of Grindley (1963) elsewhere in the Transantarctic Mountains. Samples from the Dufek intrusion dated in our study include plagioclase of three plagioclase cumulates (anorthosites) from widely spaced stratigraphic intervals in the Dufek Massif and Forrestal Range; pyroxene from a pyroxene-cumulate (pyroxenite) layer in the Dufek Massif; and fine-. grained noncumulus whole-rock gabbro from an inferred chilled contact zone near Mount Lechner in the Forrestal Range. In addition, we also determined ages on plagioclase, pyroxene, and whole-rock samples from diabase sills of the Pecora Escarpment and basalt and diabase dikes of the Cordiner Peaks. The sills intrude the Permian Pecora Formation, and the dikes cut the Devonian Dover sandstone. and the overlying Gale mudstone (Ford et al., 1978). Conventional K-Ar ages on samples from the Dufek intrustion range from 174.1 ± 4.4, 171.2 ± 4.3, and
Ford, A. B. 1976. Stratigraphy of the layered gabbroic Dufek intrusion, Antarctica. U.S. Geological Survey Bulletin, no. l405-D. Ford, A. B., R. L. Reynolds, Carl Huie, and S. J . Boyer. 1979. Geological investigation of the Dufek intrusion, Pensacola Mountains, 1978-79. Antarctic Journal of the United States (this issue). 1-Iimmelberg, G. R., and A. B. Ford. 1976. Pyroxenes of the Dufek intrustion, Antarctica. Journal of Petrology, 17: 21943. Himmelberg, G. R., and A. B. Ford. 1977. Iron-titanium oxides of the Dufek intrusion, Antarctica. American Mineralogist, 62: 623-33. Irvine, T. N. 1978. Density current structure and magmatic sedimentation. In Carnegie Institution of Washington Yearbook, 77: 717-25. Wager, L. R., and G. M. Brown. 1968. Layered Igneous Rocks. Edinburgh, Scotland: Oliver & Boyd.
169.5 ± 4.2 million years for the plagioclase, through 111.9± 2.8 and 106.3 ± 2.7 million years for the wholerock gabbros, to 97.5 ± 2.4 million years for the pyroxene samples. In thin section, the plagioclase appears little altered, but the pyroxenes (augite and inverted pigeonite) show variable and commonly large effects of inversion and exsolution (Himmelberg and Ford, 1976). Kistler and Dodge (1966) suggested that the young K-Ar ages for pyroxene-.-compared to those for coexisting biotite, hornblende, plagioclase, and orthoclase from Sierra Nevada (California) quartz diorite plutons-may be related to ubiquitously observed exsolution lamellae, which decrease the effective diffusion dimensions and may permit low-temperature argon loss. Similarly, the considerably younger ages of Dufek intrusion samples containing pyroxene probably reflect low-temperature argon loss during or after subsolidus phase changes. To ascertain that these young pyroxene ages are not the result of inaccuracies in the measurement of lowK2 0 concentrations, pyroxene from two localities in the Dufek intrusion was also dated by the total-fusion 40Ar/ 39 Ar technique. Results showed that the young conventional K-Ar ages of samples containing pyroxene are probably not due to such analytical difficulties. The best present estimate for the age of the intrusion is therefore considered to be the average of the three plagioclase determinations: 171.6 ± 4.3 million years. This age is in close agreement with, and within the range of analytical uncertainties of, the value of 175 ± 5 million years reported by Elliot, Fleck, and Sutter (in press) for basalt flows of the Ferrar Group in the central Transantarctic Mountains. Samples of whole-rock holocrystalline chilled-contact basalt and of three plagioclase-pyroxene pairs from two diabase sills in the Pecora Escarpment yielded conventional K-Ar ages ranging from 223.1 ± 5.6 million years for the basalt to 177.7 ± 4.5 million years for a plagioclase. In contrast with the Dufek intrusion, the apparent ages of pyroxene in all three pairs are slightly greater than those of coexisting plagioclase and are probably affected by small but variable amounts of excess argon.
