31. petrology of basalt drilled from the galapagos spreading center, deep sea drilling project leg...
TRANSCRIPT
31. PETROLOGY OF BASALT DRILLED FROM THE GALAPAGOS SPREADING CENTER,DEEP SEA DRILLING PROJECT LEG 54
R. V. Fodor, Department of Geosciences, North Carolina State University, Raleigh, North CarolinaJ. L. Berkley, K. Keil, and J. W. Husler, Department of Geology, and Institute of Meteoritics,
University of New Mexico, Albuquerque, New Mexicoand
M.-S. Ma and R. A. Schmitt, Department of Chemistry and the Radiation Center,Oregon State University, Corvallis, Oregon
ABSTRACT
Basalt recovered from Site 424, 22 km south of the Galapagosspreading center, East Pacific Ocean, is FeO*- and TiO2-rich (—13.3and 1.85 wt. %, respectively, FeO* expressing the total Fe content)oceanic tholeiite containing groundmass clinopyroxene (Fs16_35),plagioclase (An60_70), titaniferous magnetite, pyrrhotite, and in-terstitial material, much of which is altered but SiO2-rich wherefresh. Trace-element abundances in these ferrobasalts (FeOVMgO~ 2.1) show low Ni contents (30 to 90 ppm), a depletion in lightREE, and a negative Eu anomaly. Basalt recovered from Site 425, 62km north of the spreading center, is high-SiO2 (51.5 wt %) and low-Na2O (—2.0 wt °/o) oceanic tholeiite and is composed of phasessimilar to those of Site 424 ferrobasalt, except that clinopyroxene isricher in MgO (Fslo_3O). Basalts at both sites are relatively fresh andfree of notable hydrothermal alteration.
The ferrobasalts are compositionally similar to those previouslyreported from the region of the Galapagos spreading center, and arecharacteristic of a "normal" mid-oceanic ridge environment wheresubstantial clinopyroxene and plagioclase (and minor olivine) frac-tionation occurred in light REE and LIL element-depleted parentalmagma. The Site 425 basalt, although fractionated, is substantiallycloser to primitive oceanic tholeiite than ferrobasalt and is probablythe product of largely plagioclase and lesser olivine fractionationfrom primitive oceanic tholeiite magma.
INTRODUCTION
The crustal spreading axis in the region of theGalapagos Islands, Eastern Pacific Ocean, is a complexeast-west-trending ocean ridge system extending east-ward from the intersection of the Pacific, Cocos, andNazca lithospheric plates at about 2°N and 102°W(Figure 1). As defined by transform faults, three majorsegments compose this spreading system: the Cocos-Nazca rise, the Galapagos spreading center, and theCosta Rica rift. Drilling during Leg 54 was done alongthe central segment of the system, the Galapagos spread-ing center. This region is of special interest for its hy-drothermal fields defined by Fe-Mn oxide-encrustedsediment mounds (about 10 m high, 25 m across; Lons-dale, 1977); high heat flow (Sclater and Klitgord, 1973;Williams et al., 1974); and high water temperature(~15°C; Corliss and Ballard, 1977). Basaltic crust atthe spreading center was an important objective of thedrilling, for the characterization of this material isessential for a proper understanding of oceanic hydro-
thermal activity, mineralization, and the formation ofsediment mounds. Accordingly, we drilled and recov-ered basalt for study from two sites at the Galapagosspreading center.
GENERAL DESCRIPTIONSSite 424 is located about 22 km south of the spreading
center axis (Figure 1) near O°35'N and 86°07'W, at adepth of about 2700 meters of water. This site comprisesfour closely spaced drill holes (each about 300 m aparton a north-south line) designated Holes 424, 424A,424B, and 424C. On the basis of magnetic anomalies,the crustal age there is about 0.60 m.y.; attempts atK-Ar dating were futile owing to excess Ar (E. H.McKee, USGS, written communication). Shipboard ex-aminations and analyses of the rocks from the fourholes revealed them to be relatively fresh aphyric ferro-basalt. Of the four holes, samples selected for post-cruise study are from Hole 424, where drilling was car-ried out through a mound and where the deepest basaltpenetration was achieved (45 m of basalt; 8.5 m of core
737
R. V. FODORETAL.
Figure 1. Index map showing locations of Sites 424 and 425. Inset shows relativepositions of the four holes drilled at Site 424.
recovery), and from off-mound Hole 424B (14.5 m ofbasalt; 3 m of core recovery).
Site 425 is located about 62 km north of the spreadingcenter axis near 1°24'N and 86°04'W at a water depthof about 2850 meters (Figure 1). The crustal age here isabout 1.8 m.y. Basalt penetration was about 30 meters,and 5.6 meters of basalt core was recovered. Shipboardexamination and analyses showed that the rock wasrelatively fresh aphanitic basalt, and compositionallymore "normal" oceanic tholeiite.
In view of the potential genetic significance betweenthe sediment mounds and underlying basement rock inthe area of Site 424 (where hydrothermal fluids perco-lating through the basement could be responsible formound development; Lonsdale, 1977), special attentionwas directed at the basalt from this region. Also, it washoped that this study could contribute to the characteri-zation and understanding of ferrobasalts. Therefore,this report presents the petrography, mineral chemistry,and the major element, trace element, and rare-earthelement (REE) abundances for samples taken at rela-tively close intervals from Holes 424 and 424B. Similarstudies, with the exception of analyses for REE, weremade of three samples from Site 425.
ANALYTICAL PROCEDURESWhole-rock analyses were obtained by the following
techniques. Silica was determined gravimetrically. TotalFe, A12O3, MgO, CaO, Na2O, K2O, TiO2, and MnOwere determined by atomic absorption using calibrationcurves prepared from U.S. Geological Survey standardbasalt, BCR-1. For determination of P2O5, an adapta-tion of the Molybdenum Blue spectrophotometricmethod was employed. Water was determined by igni-tion loss, but the amount of Fe oxidation was taken into
account by K2Cr2O7 titration of ferrous iron in both theignited and unignited sample.
The trace elements Co, Cr, Sc, Ni, V, and REE forSite 424 were determined by instrumental neutronactivation analyses; Sr plus all trace elements for Site425 were obtained by atomic absorption. Mineral com-positions were determined with an ARL-EMX-SM elec-tron microprobe operated at 15 keV accelerating voltageand 0.015 to 0.20 µA sample current. Corrections fordifferential matrix effects were made following the pro-cedures of Bence and Albee (1968) and Albee and Ray(1970).
PETROGRAPHYAll samples from Holes 424 and 424B have identical
modal mineralogy of plagioclase, clinopyroxene, andoxides, but display a gradational textural range from:plagioclase microlites (0.1 mm) in hyalo-ophitic rocks(Figure 2A) to fine-grained intergranular rocks (0.3 to0.5 mm) with minor intersertal areas (Figure 2B) tomoderately coarse grained (1.0 mm) subophitic rockswith some intersertal areas (Figure 2C). In some rocks,the plagioclase and clinopyroxene have feathery vario-litic relationships, and in the finer grained samples pla-gioclase laths are somewhat skeletal in form. Quartzwas not observed optically, but microprobe analyses re-vealed free SiO2 in several samples as discrete grains andas micron-sized SiO2 concentrations in interstitialmaterial.
Although material that was originally glass makes upan important part of these rocks, fresh glass is now rare.It often contains crystallites of plagioclase and pyroxene(Figure 3). As indicated by weak birefringence, mostglass seems to be altered to clay(?) minerals. These inter-sertal areas of "glass" display substantial concentra-
738
PETROLOGY OF BASALT
Figure 2. (A) Hyalo-ophitic texture in basalt sample from Section 424B-5-1 showing mainly plagioclase microlites andlesser clinopyroxene grains in a matrix that was originally glassy; (B) intergranular texture in basalt from Section424-5-2; and (C) intergranular to subophitic texture in basalt from Section 424-6-2. Transmitted light, crossednicols; scale bars equal 1 mm.
Figure 3. Interstitial area in basalt from Section 424-6-2 in (A) plane transmitted light; (B) transmitted light, crossednicols; and (C) reflected light. Original interstitial glass contains skeletal crystals and is altered in part to clay-likematerial. Opaque phases are concentrated in these areas. Scale bars equal 0.2 mm.
739
R. V. FODORETAL.
tions of opaque minerals, mainly titaniferous magne-tite, with minor sulfide minerals as well (Figure 4). Infact, the sulfides in Site 424 are spherules 1 to 3 µm indiameter that are largely restricted to these intersertalareas; such occurrences seem to be characteristic ofoceanic basalt (Mathez and Yeats, 1976; Czamanskeand Moore, 1977). Nearly all the samples have someother minerals such as clays or zeolites filling vesicles.
The two Site 424 samples selected for microprobeanalyses represent different aspects of the texturalscheme: Sample 424-4-6, 112-116 cm, piece 3 is fine-grained intergranular (Figure 5A) with some intersertalareas, and Sample 424-6-2,2-5 cm, piece 1A is moder-ately coarse grained subophitic (Figure 5B) with someplagioclase large enough to be considered micropheno-crysts.
Samples from Site 425 have mineralogy and texturesvery similar to Site 424 rocks. Textures range fromhyalo-ophitic with crystallites of plagioclase, to inter-granular with small intersertal areas. Also, most of theareas that were once glassy now have weak birefringenceand appear altered. The glassy areas also contain sul-fides. Nearly all rocks have vesicles with secondary min-erals, and microprobe examination revealed the pres-ence of free SiO2 as grains and as micron-sized SiO2concentrations in interstitial areas. Unlike Site 424samples, however, some Site 425 rocks have pheno-crysts, mainly of plagioclase, and, more rarely, of clino-pyroxene as well (Figure 6B).
The Site 425 samples selected for microprobe analy-ses represent two different textures: Sample 425-7-2,52-54 cm, Piece 7A is intergranular (Figure 6A) andSample 425-8-1, 130-132 cm, Piece 15 is porphyriticwith glomerocrysts of plagioclase, minor clinopyroxenephenocrysts, and an intergranular and partly intersertalgroundmass (Figure 6B).
Figure 4. Sulfide spherule (far left) next to titaniferousmagnetite within an interstitial formerly glassy areaof basalt in Section 424-6-2. Reflected light; scale barequals 0.2 mm.
Figure 5. (A) Intergranular texture of plagioclase andclinopyroxene in basalt from Section 424-4-6, and (B)the subophitic texture of basalt from Section 424-6-2.Transmitted, plane-polarized light; scale bars equal0.5 mm.
BULK COMPOSITIONWhole-rock analyses and normative mineralogy of
five samples from Hole 424, three samples from Hole424B, and three from Hole 425 are presented in Table 1.Relative positions of the cores sampled for analyses areillustrated in Figure 7. Sampling was done on ship-board, and material was selected that would potentiallyprovide as diverse a representation as possible.
Holes 424 and 424BBecause of the high total FeO* contents (FeO* =
FeO + 0.9 Fe2O3 13.3 wt. °/o) of each sample, allmaterial recovered is designated clearly as ferrobasalt.Low K2O (-0.07 wt. %) and high SiO2 and normativequartz contents of these rocks further identify them as
740
PETROLOGY OF BASALT
Figure 6. (A) Intergranular texture of plagioclase andclinopyroxene in basalt from Section 425-7-2, and (B)the porphyritic texture, phenocrysts of plagioclaseand clinopyroxene, of basalt from Section 425-8-1.Transmitted light, crossed nicols; scale bars equal1 mm.
quartz tholeiites. Notable, too, are the relatively highTiO2 (~ 1.85 wt. °/o) and low A12O3 (~ 13.3 wt. %) con-tents. These are all features diagnostic of other ferro-basalts (sometimes called FETI basalts in reference tohigh Fe and Ti contents [Melson et al., 1976]) dredgedfrom the region of the Galapagos spreading center(Anderson et al., 1975; Schilling et al., 1976), anddrilled from the nearby Peru basin (Kempe, 1976).
The low Ni (30 to 90 ppm) and Co (53 to 69 ppm)contents are compatible with the high FeOVMgO ratios(~2.1), both values suggesting substantial fractionationof Mg-rich olivine from a precursor magma. If Cr-spinels were associated with such olivine, this could alsoaccount for the relatively low Cr (-100 ppm) contentsof the rock. High V abundances are compatible with thehigh Ti values; V is expected to accompany Ti, especial-
ly in the Fe- and Ti-oxide phases. All other trace-element abundances, such as the low Sr contents, arecompatible with the composition of most mid-oceanridge basalts (e.g., Kay and Hubbard, 1978).
The chondrite-normalized REE abundances areplotted on Figure 8. The trend is that of light REEdepletion relative to chondrites and a negative europiumanomaly. Similar REE concentrations have been shownfor other ferrobasalts dredged and drilled from theGalapagos region (Figure 8). Relative to less-fraction-ated (i.e., lower FeO*) basalt of the East Pacific regions(e.g., Thompson et al., 1976) and to mid-ocean ridgebasalt in general (e.g., Kay et al., 1970), there is sub-stantial enrichment of all REE. As pointed out by Schil-ling et al. (1976), systematic increases of all REE abun-dances can accompany increasing FeO* contents inbasalt, where higher amounts of both REE and FeO* re-flect greater degrees of fractionation from more primi-tive basaltic magmas.
The basalt compositions are similar throughoutHoles 424 and 424B, showing virtually no differences inmajor- or trace-element abundances, except for Ni. Thesmall variation in Ni concentrations may be due to arare appearance of olivine in a few samples, althougholivine was not observed optically during this study(shipboard examinations, however, did reveal the pres-ence of rare olivine in a few samples). Similarly, there isessentially no horizontal change in composition betweenthe two holes which are about 400 meters apart. Textur-al variations indicate that probably several lava flowsmake up material cored in each area but, clearly, allflows represent the same source magma.
Site 425
Basaltic samples at Site 425 have FeO* contentssimilar to typical oceanic basalt (~ 10 wt. %), but SiO2contents are exceptionally high and Na2O and TiO2 con-tents very low compared with most oceanic tholeiite(e.g., Kay et al., 1970; Engel et al., 1965). The highnormative-quartz content shows that these rocks arewell within the quartz tholeiite basalt field.
Rocks having similar FeO* and SiO2 contents havebeen dredged (Anderson et al., 1975; Schilling et al.,1976) from the Galapagos region and drilled (Kempe,1976) from the Peru basin, but none is reported to be sodepleted in Na2O as those described here. The less-frac-tionated nature (compared with Site 424 rocks) of Site425 basalt applies also to very low P2O5, Sr, and Ba inthe three samples, and very low K2O in two of them.Trace-element analyses show less V and more Cr thanthe ferrobasalts; the higher Cr contents are compatiblewith these rocks displaying less fractionation than theferrobasalt. Basalt with such high SiO2 and low large-ion lithophile element contents are uncommon amongoceanic tholeiites, but a few examples have been discov-ered through drilling in the Indian Ocean. Melson et al.(1976) referred to this basalt type as HISI.
The core displays some vertical variation. For exam-ple, the middle sample from Section 425-8-1 (Table 1,column 10), is higher in A12O3, CaO, and K2O, andlower in FeO*, MgO, TiO2, and Na2O than the other
741
R. V. FODORETAL.
TABLE 1Bulk Compositions (in wt. %) and Molecular Norms of Basalt from Site 424 (Holes 424 and 424B)
and Hole 425, Galapagos Spreading Center, East Pacific Ocean
Core-SectionInterval (cm)
SiO2
TiO2
A12O3
Fe2O3
FeOMnOMgOCaONa2OK2OH2°~H2O+
P2° 5Total
FeO*/MgO
qorabandi
hyil
mtap
Ba (ppm)SrCoCrCuSc
NiVZnLaCeNdSmEuTbDyYbLuHf
4-6,110
50.751.87
13.263.09
10.570.216.49
10.112.470.060.320.670.15
100.02
2.06
3.860.37
22.8925.7220.0320.80
2.683.330.32
8055
102
4880
451
3.515
84.71.521.098.15.0
0.763.2
5-1,70
50.541.87
13.203.29
10.390.216.40
10.142.370.060.580.870.15
100.07
2.09
4.490.37
22.0826.1520.0320.28
2.703.560.32
80
56101
4870
440
3.51411
4.51.541.067.7
5.00.763.1
Hole 424
5-3, 12
50.501.80
13.163.06
10.270.196.55
10.152.520.050.460.800.15
100.26
1.99
3.560.31
23.4625.3520.6820.41
2.593.310.32
805399
48
30447
3.4
15154.61.500.988.04.80.733.1
6-1,60
50.421.88
13.103.12
10.340.206.43
10.142.480.060.250.930.15
99.50
2.05
3.880.37
23.1425.4020.6920.092.723.380.32
8064
103
4890
443
3.514124.51.531.117.94.90.743.2
6-2,5
50.861.86
13.243.15
10.380.196.13
10.202.440.090.180.670.15
99.54
2.16
4.650.55
22.7225.8320.5019.322.683.410.32
8064
104
4740
435
3.4
1513
4.61.581.158.3
4.90.753.3
5-1,25
51.441.85
12.802.90
10.680.216.309.962.370.080.300.770.15
99.81
2.11
5.460.49
22.0724.9520.2320.66
2.673.140.32
8059
105
4970
438
3.414
184.51.571.078.04.90.753.3
Hole 424B
5-2, 15
50.531.88
13.303.33
10.230.196.50
10.282.530.080.230.620.15
99.85
2.03
3.480.49
23.4325.4820.9019.622.703.590.32
80
69105
4940
421
3.4
15
104.61.521.088.14.90.773.3
6-1, 70
50.341.83
13.242.60
10.910.226.25
10.442.410.080.180.860.15
99.51
2.12
3.230.49
22.4726.0321.3820.60
2.642.820.32
80
56110
4950
446
3.514144.51.561.088.05.10.763.1
7-2, 50
51.691.12
14.402.658.140.177.30
11.702.170.040.080.680.08
100.22
1.44
3.900.24
19.8129.9322.8518.70
1.582.810.17
<186037
22391
57336
90
Hole 425
8-1,130
51.230.74
15.803.576.110.146.80
12.581.800.310.260.500.06
99.90
1.37
5.051.87
16.4834.8022.7614.051.053.800.12
<186036
26077
64
26863
9-2,130
51.520.96
14.152.128.310.167.78
11.952.040.040.130.780.07
100.01
1.31
3.110.24
18.6429.8624.0220.35
1.362.250.14
<186041
21889
71
35088
two Hole 425 samples. This sample is plagioclase-phyricand its difference in composition relative to the othertwo can be explained easily (except for K2O) by theaddition of An-rich plagioclase phenocrysts into a mag-ma identical to the other two samples (i.e., Ca and Alhigher in phyric rock; Fe, Mg, Na, and Ti lower). Thehigher K2O content might be due to a contribution fromsecondary mineralization in the form of celadonite or
smectite, in spite of the relatively moderate water con-tent. (Shipboard descriptions pointed out a somewhataltered aspect of the core fragment from which thesample was taken.)
As at Site 424, the textures of the samples here showenough variation to indicate several discrete flows, butall are essentially identical in composition and wereprobably derived from the same source. Only vertical
742
PETROLOGY OF BASALT
SITE424
SITE425
424 424Bdepth 2730 2740 m
Core 5 P-'
Figure 7. Relative core positions (arrows) of samplesselected for whole-rock analyses. Shaded areas repre-sent recovered rock; voids are not necessarily propor-tional to the amount of material missing (not recov-ered) from sections.
movement of early formed plagioclase and some minoralteration seem to have affected bulk composition.
MINERALOGY
ClinopyroxeneRepresentative point analyses of pyroxene grains are
listed in Table 2, and intragrain and intergrain composi-tional variations are illustrated on Figure 9. Crystalliza-tion trends favor increasing FeO* and decreasing CaOcontents, characteristic of tholeiitic basalts (Fodor etal., 1975). However, there are strong tendencies forgroundmass pyroxene in all samples to approach pig-eonite compositions, filling in the miscibility gap be-tween high- and low-Ca pyroxene. This is possiblybecause of rapid cooling and crystallization, althoughthe subophitic sample curiously demonstrates this crys-tallization trend, too. Alternatively, the pyroxenes mayhave exsolved submicron-sized, low-Ca lamellae notresolvable by the electron probe.
The Fe-rich composition of the Site 424 samplesshows up well in the pyroxene analyses plotted onFigure 9. The lowest Fs content is about 16 mole percent, compared with some groundmass pyroxenes inSite 425 basalt having Fs,0 compositions. The pyroxenephenocrysts in the phyric Site 425 sample are the mostMg- and Cr-rich pyroxene observed, although thereis compositional overlap with coexisting groundmassgrains. The minor-element abundances in groundmass
50-
40-
30-
FeO* = 18.4%
FeO* = 11.7%
Tb Dy
RARE EARTH ELEMENTS( ionic radii )
Figure 8. Chondrite-normalized rare-earth patterns forSite 424 FETI basalts (average of eight samples; solidline). Dashed lines represent patterns for other ferro-basalts from the Galapagos spreading center (Schil-ling et al., 1976). FeO* values are given for each pat-tern to illustrate the increasing FeO* with increasingabsolute REE abundances. For comparison withplume-derived ferrobasalt (i.e., formed in a con-trasting tectonic environment), the shaded areashows the range for Site 214 and 216 ferrobasaltsfrom the Ninety east Ridge, Indian Ocean (Berkley,1977). Normalized abundances were plotted versusREE ionic radii (VI coordinated), because ionic radiirepresent a more meaningful chemical parameter infractionation processes compared to REE atomicnumber.
pyroxenes vary in accordance with decreasing FeO* andincreasing CaO. Most notable is the decrease in Alwhich may reflect the competition for this element dur-ing simultaneous crystallization of pyroxene and plagio-clase. The lower A12O3 content of the pyroxene pheno-crysts in the sample from Section 425-8-1 comparedwith coexisting groundmass pyroxene may indicatedecreasing silica activity in the magma during the inter-im from phenocryst formation to groundmass crystalli-zation.
PlagioclaseRepresentative point analyses of plagioclase are listed
in Table 3, and intragrain and intergrain compositionalvariations are illustrated on Figure 10. Plagioclase in theSite 424 ferrobasalts is labradorite in composition, withthe exception of some microphenocrysts in the subo-phitic sample that are zoned into the bytownite field.Groundmass plagioclase in the Site 425 basalt is compo-sitionally zoned from calcic labradorite to sodic bytow-nite, and the phenocrysts of Section 425-8-1 are exceed-ingly rich in CaO
743
R. V. FODORETAL.
TABLE 2Representative Compositions of Clinopyroxene Phenocrysts (p) andGroundmass Grains (g) in Basalt from Sites 424 and 425, Galapagos
Spreading Center, East Pacific Ocean (in wt. %)
Core-Section
SiO2
TiO2
A12O3
Cr2O3
FeOMnOMgOCaONa2O
Total
g
49.01.33.60.09
13.50.27
13.518.30.24
99.80
Hole 424
4-6
g
49.41.32.20.01
20.40.42
13.213.20.19
100.32
6-2
g
51.40.902.90.099.90.19
16.018.70.24
100.32
Number of ions on the basis of 6 oxygens
SiAliv
Alvi
TiCrFeMnMgCaNaZXY
Total
Molecular end r
FsEnWo
1.8620.1380.0230.0370.0030.4290.0090.7650.7450.0182.0002.029
4.029
nember
22.139.538.4
1.8990.100
-0.038
-0.6560.0140.7560.5440.0141.9992.022
4.021
s
33.538.727.8
1.9040.0960.0310.0250.0030.3070.0060.8840.7420.0172.0002.015
4.015
15.945.738.4
g
52.80.471.10.02
14.50.33
18.911.80.12
100.04
1.9600.0400.0080.0130.0010.4500.0101.0460.4690.0092.0002.006
4.006
22.953.223.9
7-2
g
51.80.532.60.029.20.33
17.417.10.15
99.13
1.9260.0740.0400.0150.0010.2860.0100.9640.6810.0112.0002.008
4.008
14.849.935.3
g
52.40.421.80.02
11.00.25
19.813.20.13
99.02
1.9430.0570.0220.0120.0010.3410.0081.0940.5240.0092.0002.011
4.011
17.455.826.8
Hole 425
P
52.90.222.10.634.70.15
18.519.90.17
99.27
1.9390.0610.0300.0060.0180.1440.0051.0110.7810.0122.0002.007
4.007
7.452.240.4
8-1
g
52.10.572.70.068.30.18
17.918.20.13
100.14
1.9140.0860.0310.0160.0020.2550.0060.9800.7160.0092.0002.015
4.015
13.150.236.7
g
50.50.561.70.05
17.20.42
16.312.30.15
99.18
1.9270.0730.0030.0160.0020.5490.0140.9270.5030.0112.0002.025
4.025
27.746.925.4
As is characteristic of intermediate and calcic plagio-clase of tholeiitic basalt, K is virtually absent (Keil et al.,1972). Also, as expected, higher FeO* is present in thegroundmass plagioclase of the ferrobasalts than in thatof Site 425 basalt.
Intersertal and Interstitial MaterialRepresentative point analyses of material intersertal
and interstitial to plagioclase groundmass grains arepresented in Table 4. These compositions presumablyrepresent the composition of the residual trapped li-quids of original magma that are still fresh enough forelectron probe analysis; most of the original glass ap-pears to be partially altered and readily volatilizes underthe electron beam. The intersertal and interstitial mate-rial in rocks of both sites is SiO2-rich and has varyingamounts of A12O3 and alkalis. One point analysis is thatof essentially pure SiO2 (Table 4, column 6). In keepingwith the tholeiitic nature of these Galapagos samples,no K-enrichment occurred during final crystallizationand cooling.
Opaque OxidesIn spite of the greater amount of FeO* in the Site 424
samples, the titaniferous magnetite does not containsignificantly different FeO* and TiO2 from those in Site425 samples (Table 5). In fact, the overall compositionof the oxides, which have 60 to 65 mole per cent ulvo-spinel content, is like those found in other oceanictholeiites (e.g., Mazullo et al., 1976; Fodor et al., 1977).
Similarly, the V2O3 content in the oxide of the ferro-basalt does not reflect the higher V concentration in thewhole rock. Apparently, the higher FeO*, TiO2, and Vcontents in the ferrobasalts are expressed by a greatermodal percentage of oxides relative to the Site 425samples. A modal analysis (1000 points) of two samplesproved this: Section 424-4-6 has 5.8 volume per centoxides, whereas Section 425-7-2 has 3.4 volume percent.
SulfidesSulfide grains are present in only trace amounts in all
of the basalts examined and are usually less than 5 µm insize. In the ferrobasalts, sulfides range in compositionfrom nearly pure pyrrhotite, to pyrrhotite with minorCu to substantial Cu contents (-24 wt. %) (Table 6).Pyrite and Cu-rich pyrrhotite were observed in the Site425 basalts. In addition to the one analysis of a sulfidegrain in a Site 425 basalt presented in Table 6, othergrains were found to have 16 and 24 weight per cent Cuand <l.O weight per cent Ni; small grain sizes, how-ever, preclude summations of their analyses to near 100weight per cent.
DISCUSSION AND CONCLUSIONSThe ferrobasalts of the Galapagos Spreading Center,
Site 424, are remarkably uniform in major- and trace-element compositions, both vertically (over 40 m) andhorizontally (-400 m). Since compositionally similarferrobasalt is present at both mound and nonmoundholes (Holes 424 and 424B, respectively), little informa-tion can be extracted regarding the role of basementrock composition in the formation of the sedimentmounds in the Galapagos hydrothermal field. TheseFe-Mn-oxide encrusted mounds are believed to haveoriginated over fault zones (structurally controlled),where sea water is transformed into hydrothermal fluidduring its descent and passage in the fractured basalticbasement (Lonsdale, 1977). The petrologic results showonly that a particular rock composition is no assuranceof mound development, and that based on the freshnessof these rocks, the upper-most basement, where drilled,had little to do with mound development.
Defining hydrothermal mineralization is a key ob-jective of drilling in the Galapagos region. There isgood evidence that hydrothermal solutions at spreadingridges are the cause of substantial Fe-Cu-sulfide miner-alization (Bonatti et al., 1976; Rona, 1978). In thesecases, veins of sulfides and evidence of hydrothermalalteration (e.g., greenschist facies metabasalt; Bonatti etal., 1976) are present. No such mineralization was notedin the Site 424 basalts and only traces were found in Site425 basalts (Schrader et al., "Ore Petrology," this vol-ume). The Fe-Cu-sulfides in these basalts are much likethose in typical magmatic systems in which primary sul-fides are associated with immiscible silicate liquids.These sulfides commonly form as globules in glass (Des-borough et al., 1968; Mathez and Yeats, 1976; Czaman-ske and Moore, 1977).
FETI basalts like those from Site 424 are found else-where in the Pacific Ocean at the East Pacific Rise and
744
PETROLOGY OF BASALT
424 4-6
phenocrysts O
groundmass
Figure 9. Pyroxene quadrilateral showing individual point analyses for clinopyrox-ene phenocrysts and groundmass in basalt from Sites 424 and 425, plotted interms of molecular end members enstatite (En, MgSiO3), ferrosilite (Fs, FeSiO3),diopside [Di, (Ca, Mg)Si2O6], and hedenbergite [Hd, (Fe, Ca)Si2O6J.
TABLE 3Representative Compositions of Plagioclase Phenocrysts (p), Microphenocrysts (m), and Groundmass Crystals (g)
in Basalt from Sites 424 and 425, Galapagos Spreading Center, East Pacific Ocean (in wt. %)
Core-Section
SiO2
A12O3
FeOMgOCaO
Na2OK2O
Total
Number of ions
SiAlFe
MgCaNaKZXY
Total
4-6
g
52.229.30.990.22
12.93.70.03
99.34
g
54.428.2
1.10.32
12.34.2
0.03
100.55
Hole 424
m
51.330.50.740.16
14.13.3
0.03
100.13
on the basis of 32 oxygens
9.5576.3220.1520.0602.5301.3130.007
15.8794.062
19.941
Molecular end members
An
AbOr
65.734.1
0.2
9.8185.9980.1660.0862.3781.4690.007
15.8164.106
19.922
61.738.1
0.2
9.3436.5470.1130.0432.7511.1650.007
15.8904.079
19.969
70.129.7
0.2
6-2
g
52.229.5
1.2
0.2813.1
3.50.03
99.81
9.5236.3420.1830.0762.5601.2380.007
15.8654.064
19.929
67.332.50.2
g
53.429.10.810.13
12.24.10.04
99.78
9.6996.2290.1230.0352.3741.4440.009
15.9283.985
19.913
62.037.8
0.2
g
51.830.70.590.33
14.42.90.02
100.74
9.3596.5370.0890.0892.7871.0160.005
15.8963.986
19.882
73.226.7
0.1
7-2
g
53.129.50.780.23
13.13.80.04
100.55
9.5956.2820.1180.0622.5361.3310.009
15.8774.056
19.933
65.434.4
0.2
Hole 425
P
45.934.30.410.19
17.41.10.02
99.32
8.5017.4870.0630.0523.4520.3950.005
15.9883.967
19.955
89.610.30.1
8-1
g
49.631.10.500.27
15.22.50.02
99.19
9.1376.7510.0770.0743.0000.8930.005
15.8884.049
19.937
77.022.9
0.1
g
51.230.10.670.22
13.63.30.02
99.11
9.4016.5130.1030.0602.6751.1750.005
15.9174.015
19.932
69.430.50.1
745
R. V. FODORETAL.
Figure 10. Feldspar ternary diagram showing individual point analyses for plagioclase phenocrysts,microphenocrysts, and groundmass grains in basalt from Sites 424 and 425, plotted in terms ofmolecular end members albite (Ab, NaAlSi3O8), orthoclase (Or, KAlSi3O8), and anorthite (An,CaAl2Si2O8).
TABLE 4Representative Compositions of SiC>2-Rich Material within Interstitial
and Intersertal Areas in Basalt from Sites 424 and 425, GalapagosSpreading Center, East Pacific Ocean (in wt%)
Core-Section
SiO2
A12O3
FeO*MgO
CaO
Na2θ
K2O
Total
76.3
14.2
0.89
0.25
2.62.4
0.04
96.68
4-6
70.7
13.9
1.80.77
4.84.1
0.07
96.14
Hole 424
78.7
11.4
1.9
0.48
3.1
3.2
0.06
98.84
6-2
75.9
16.61.2
0.10
2.8
0.62
0.01
97.23
72.613.6
0.86
0.08
5.5
6.4
0.06
99.10
Hole 425
7-2
98.60.09
0.17
<O.Ol
<O.Ol
<O.Ol
<O.Ol
98.86
8-1
82.2
4.2
6.2
0.78
1.2
1.7
0.19
96.47
the Juan de Fuca Ridge. Other FETI basalts occur in theIndian Ocean at the Ninetyeast Ridge and MozambiqueRidge (Berkley, 1977; Thompson et al., 1978). Althoughcompositionally all of the ferrobasalts are more or lessalike, details of their compositions vary according totheir tectonic environments. Three locations are atspreading centers (Galapagos, East Pacific Rise, andJuan de Fuca Ridge), whereas the Indian Ocean FETIbasalts occur along aseismic ridges. The Site 424samples are members of the Fe-rich and light REE-depleted basalts (Figure 8) of the Galapagos SpreadingCenter that Schilling et al. (1976) described as character-istic of "normal ridge segments" and associated withmaterial derived from a REE-depleted mantle source.The same holds true for the ferrobasalts from the Juande Fuca Ridge (Kay et al., 1970) and the East PacificRise (Schilling and Bonatti, 1975). This is in contrast tothe Fe-rich tholeiitic rocks of the nearby Galapagos Is-lands which are relatively enriched in light REE (Schill-
TABLE 5Representative Compositions of TitaniferousMagnetite in Basalt from Sites 424 and 425,
Galapagos Spreading Center,East Pacific Ocean (in wt. %)
Core-Section
SiO2
TiO2
A12O3
Cr 2 O 3
v 2 o 3FeO*MnO
MgO
Total
Recalculated
FeO
F e 2 O 3
Total
Usp
Hole 424
4-6
0.1622.3
1.6
0.02
0.5070.4
0.510.43
95.92
49.822.9
98.2
65.3
6-2
0.2022.4
1.9
0.030.85
71.50.45
0.75
98.13
49.824.0
100.38
63.8
Hole 425
7-2
0.2620.3
1.4
0.11
0.4072.1
0.52
0.23
95.32
48.2
26.6
98.02
59.8
8-1
0.3021.3
1.7
<O.Ol0.65
69.90.450.35
94.66
48.623.7
97.05
63.6
ing et al., 1976). According to the hypothesis of Schill-ing (1975), this archipelago probably represents materialderived from a mantle plume rising beneath the islands.The effects of the mantle plume are observed as REEenrichment in rocks located up to 450 km from thecenter of the Galapagos archipelago; Site 424 is about500 km away.
746
PETROLOGY OF BASALT
TABLE 6Representative Compositions3 of Sulfide Grains in Basalt from
Sites 424 and 425, Galapagos Spreading Center,East Pacific Ocean
CuNiFeS
Total
59.238.1
97.3
Section 424-4-6
18.2
43.235.2
96.6
0.550.21
58.237.8
97.06
22.00.62
42.034.5
99.12
24.10.12
38.834.5
97.52
Section 424-6-2
0.151.1
59.036.9
97.15
18.10.32
44.834.9
98.12
59.438.9
98.3
Section425-7-2
0.100.20
58.538.6
97.4
aSome analyses total low, owing to small grain sizes and pitted surfaces.
FETI basalts relatively enriched in light REE are notunique to the Galapagos Islands. Similar REE patterns(Figure 8) and LIL element concentrations are displayedby ferrobasalts of the aseismic Ninetyeast Ridge (Thomp-son et al., 1978; Ludden et al., 1977) and the Mozam-bique Ridge (Berkley, 1977) of the Indian Ocean, indi-cating a possible mantle-plume source for these basaltsas well.
Computer modeling (Clague and Bunch, 1976; Schill-ing et al., 1976) has shown that the origin of FETIbasalts can be explained by shallow-level crystal frac-tionation. Given "normal" mid-ocean ridge basalt as aparent, ferrobasalts can be produced by the fractiona-tion of largely plagioclase (which accounts for thenegative Eu anomaly) and smaller amounts of clino-pyroxene and lesser olivine. Clague and Bunch (1976)compute respective proportions for these fractionat-ing phases of 9.3:7.7:1. Our least-square calculations(Wright and Doherty, 1970) for generating the Site 424FETI basalts from a Site 425 parent (Section 425-9-2)suggest that the weight proportions of the fractionatingphases are 12.8:11.3:1 for clinopyroxene, plagioclase,and olivine, respectively (residuals for each oxide areless than 1 per cent of the amounts present when com-paring the calculated parent rock with an actual parentrock). This fractionation scheme requires about 50 percent crystallization. Since significant clinopyroxenefractionation can cause a light-REE enrichment in resid-ual liquid, the light REE-depleted patterns displayed bySite 424 basalts (Figure 8) suggest derivation from ahypothetical parent which may have been even more de-pleted in light REE than the Site 424 basalt.
Assuming the parent basalt for Site 424 crystallizedclinopyroxene, plagioclase, and olivine in the propor-tions 12.8:11.3:1 (50 per cent crystallization), we usedappropriate partition (mineral/liquid) coefficient dataand trace-element abundances in average Site 424 basaltto calculate expected trace-element concentrations in itsparent. The calculated values (in ppm) are: La 1.84, Ce7.7, Nd 6.1, Sm 2.68, Eu 0.95, Tb 0.65, Dy 4.8, Yb 3.0,Lu 0.45, Sc 78, Co 50, and Ni 108. The expected REEpattern (Figure 8) has no Eu-anomaly and is similar tothe lowest REE-abundance patterns observed in theabundance range for "normal" ocean ridge basalts. TheSc value of 78 ppm is high compared with Sc abun-dances commonly observed for ocean ridge basalts thathave similar REE patterns (e.g., 42 to 51 ppm; Berkley,1977). However, since the calculated Sc abundance is
strongly dependent on the D(s/0 value for clinopyrox-ene, lowering of the DSc (cpx/liq) from 3.3 to 2.0 pro-duces agreement between expected and observed Scabundances in ocean ridge basalts with low REE-abun-dances. Reductions in either the quantities of fractional-ly crystallized olivine and clinopyroxene, or in thevalues of the partition coefficients for Ni in theseminerals, will lower the expected Ni content of 108 ppmto values more in line with the observed 71 ppm of Ni inSection 425-9-2 basalt (Table 1).
The contribution of clinopyroxene as a fractionatingphase in the generation of FETI basalts is significant,because this phase is believed to have a relatively unim-portant role in fractionation of "normal" oceanic tho-leiite (e.g., Frey et al., 1974; Shibata, 1976). The relativeeffects of more or less clinopyroxene fractionation onbulk chemistry are shown in Figure 11. ' 'Normal" ridge-type tholeiites generally lie along a plagioclase subtrac-tion line, whereas FETI basalts lie along a vector moreinfluenced by clinopyroxene subtraction (Figure 11).Assuming shallow-level fractionation for FETI basalts,significant clinopyroxene fractionation is expected. Inthe basalt system, Fo-An-Qz (e.g., Walker et al., 1972)cotectic crystallization of olivine and plagioclase is fol-lowed by clinopyroxene plus plagioclase crystallizationat advanced stages of fractionation. Thus, the mosthighly fractionated oceanic basalts, such as the Site 424basalts, are the most likely basalts to show evidence forsignificant clinopyroxene fractionation.
Based largely on FeOVMgO and low LIL contents,Site 425 basalts are much closer to primitive basalt thanthe FETI basalts from Site 424. Yet their moderate MgOand low Ni contents indicate that they have undergoneat least some fractional crystallization. The most primi-tive basalts from mid-ocean ridge environments areconsidered to have MgO >8.0 weight per cent (i.e., aslightly lower FeOVMgO ratio than that of 1.4 for theSite 425 basalts; Frey et al., 1974; Thompson et al.,1976). Our computer modeling suggests that Site 425 ba-salts may be the product of about 28 per cent fractionalcrystallization of plagioclase (An70) and olivine (FOÇQ) inthe proportion 3:1, respectively, from a primitive tholei-ite composition similar to that given by Frey et al. (1974;table 3, analysis 7). No clinopyroxene was needed toproduce the resultant least-squares fit. Thus, in terms offractionation history, the Site 425 basalts representmoderately fractionated oceanic tholeiites derived froma more primitive tholeiite parent by mainly plagioclaseplus olivine fractionation. On the other hand, the Fe-rich Site 424 basalts represent highly fractionated mag-mas derived from a light REE and LIL element-depletedparent, possibly similar in composition to the Site 425basalts.
ACKNOWLEDGMENTSThis work was supported in parts by the National Aero-
nautics and Space Administration, Grants NGL 32-004-064(K. Keil, principal investigator), Grant NGL 39-002-039 (R. A.Schmitt, principal investigator), and by the Engineering Foun-dation of North Carolina State University.
747
R. V. FODOR ET AL.
20
18
16
Ninetyeast Ridge
/"214~~ 25?\I • • j
\ 216 /\jjfy
31
425 7-2,-,J F R • U
15
424CJIJFR~EPR
INDIANPrimitive
^Φ "/
^ A T L A N T I Cj Primitive
η"^ I Galapagos
^/ / ^ w Primitivef U425 8-1
O 425 9-2
10 12
CaO
Figure 11. Plot of bulk A12O3 versus CaO in weight percent for Site 424 FETI basalts and Site 425 tholeiites,with comparative analyses and pertinent phenocrystsubtraction lines. Circles are for "normal" oceanictholeiites; squares are for ferrobasalts; open symbolsrepresent analyses from this report. Comparativeanalyses: Atlantic primitive oceanic basalt and highlyfractionated Site 15 glass from Frey et al. (1974); In-dian Ocean primitive and other basalts from Ninety-east Ridge (averages from Sites 214, 216, 254, 257)from Berkley (1977); Galapagos primitive basalt(Sample D-7) from Anderson et al. (1975); Site 319represents the high MgO/FeO* basalt from the Naz-ca plate from Thompson et al. (1976); ferrobasaltsfrom the East Pacific Rise (EPR), and Juan de FucaRidge (JFR)from Clague and Bunch (1976).
REFERENCES
Albee, A. L. and Ray, L., 1970. Correction factors for elec-tron probe microanalysis of silicates, oxides, carbonates,phosphates, and sulfates, Anal. Chem., v. 42, p. 1408-1414.
Anderson, R. N., Clague, D. A., Klitgord, K. D., Marshall,M., and Nishimori, R. K., 1975. Magnetic and petrologicvariations along the Galapagos spreading center and theirrelation to the Galapagos melting anomaly, Geol. Soc. Am.Bull., v. 86, p. 683-694.
Bence, A. E. and Albee, A. L., 1968. Empirical correction fac-tors for the electron microanalysis of silicates and oxides, /.Geol., v. 76, p. 382-403.
Berkley, J. L., 1977. A petrochemical characterization of cer-tain subaqueous DSDP basalts and andesites from the In-dian Ocean, Unpublished Ph.D. dissertation, University ofNew Mexico.
Bonatti, E., Guerstein-Honnorez, B-M., and Hornorez, J.,1976. Copper-iron sulfide mineralizations from the equa-torial mid-Atlantic ridge, Econ. Geol., v. 71, p. 1515-1525.
Clague, D. A. and Bunch, T. E., 1976. Formation of ferro-basalt at East Pacific midocean spreading centers, J. Geo-phys. Res., v. 81, p. 4247-4256.
Corliss, T. B. and Ballard, R. D., 1977. Oasis of life in the coldabyss, National Geogr., v. 152, p. 441-453.
Czamanske, G. K and Moore, J. G., 1977. Composition andphase chemistry of sulfide globules in basalt from the Mid-Atlantic Ridge rift valley near 37 °N lat, Geol. Soc. Am.Bull., v. 88, p. 587-599.
Desborough, G. A., Anderson, A. T., and Wright, T. L.,1968. Mineralogy of sulfides from certain Hawaiian ba-salts, Econ. Geol., v. 63, p. 636-644.
Engel, A. E., Engel, C. G.,.and Havens, R. G., 1965. Chem-ical characteristics of oceanic basalts and the upper mantle,Geol. Soc. Am. Bull., v. 76, p. 719-734.
Fodor, R.V., Husler, J. W., and Keil, K., 1977. Petrology ofbasalt recovered during DSDP Leg 39B. In Supko, P. R.,Perch-Nielsen, K., et al., Initial Reports of the Deep SeaDrilling Project, v. 39: Washington (U.S. GovernmentPrinting Office), p. 513-523.
Fodor, R. V., Keil, K., and Bunch, T. E., 1975. Contributionsto the mineral chemistry of Hawaiian rocks IV. Pyroxenesin rocks from Haleakala and West Maui volcanoes, Maui,Hawaii, Contrib. Mineral. Petrol., v. 50, p. 173-195.
Frey, F. A., Bryan, W. B., and Thompson, G., 1974. AtlanticOcean floor: geochemistry and petrology of basalts fromLegs 2 and 3 of the Deep Sea Drilling Project, /. Geophys.Res., v. 79, p. 5507-5527.
Kay, R. W. and Hubbard, N. J., 1978. Trace elements inocean ridge basalts, Earth. Planet. Sci. Lett., v. 38, p.95-116.
Kay, R. W., Hubbard, N. J., and Gast, P. W., 1970. Chemicalcharacteristics and origin of oceanic ridge volcanic rocks, /.Geophys. Res., v. 75, p. 1585-1613.
Keil, K., Fodor, R. V., and Bunch, T. E., 1972. Contributionsto the mineral chemistry of Hawaiian rocks II. Feldspars inrocks from Haleakala and West Maui volcanoes, Maui,Hawaii, Contrib. Mineral. Petrol., v. 37, p. 253-276.
Kempe, D. R. C , 1976. Petrological studies on DSDP Leg 34basalts: Nazca plate, eastern Pacific Ocean. In Yeats, R. S.,Hart, S. R., et al., Initial Reports of the Deep Sea DrillingProject, v. 34: Washington (U.S. Government Printing Of-fice), p. 189-213.
Lonsdale, P., 1977. Deep-tow observations at the moundsabyssal hydrothermal field, Galapagos rift, Earth Planet.Sci. Lett., v. 36, p. 92.
Ludden, J., Thompson, G., and Bryan, W. B., 1977. An eval-uation of fractional crystallization and the origin of lavasfrom the 90 °E ridge, eastern Indian Ocean, Geol. Soc. Am.Abstracts with Programs, v. 9, p. 1077-1078.
Mathez, E. A. and Yeats, R. S., 1976. Magnetic sulfides inbasalt glass from DSDP Hole 319A and Site 320, Nazcaplate. In Yeats, R. S., Hart, S. R., et al., Initial Reports ofthe Deep Sea Drilling Project, v. 34: Washington (U.S.Government Printing Office), p. 363-374.
Mazzullo, L. J., Bence, A. F., and Papike, J. J., 1976. Petrog-raphy and phase chemistry of basalts from DSDP Leg 34,Nazca plate. In Yeats, R. S., Hart. S. R., et al., InitialReports of the Deep Sea Drilling Project, v. 34: Washing-ton (U.S. Government Printing Office), p. 245-262.
Melson, W. G., Valuer, T. L., Wright, T. L., Byerly, G., andNelen, J., 1976. Chemical diversity of abyssal volcanic glass
748
PETROLOGY OF BASALT
erupted along Pacific, Atlantic, and Indian Ocean sea-floorspreading centers, Am. Geophys. Union Mono. 19, p.351-367.
Rona, P. A., 1978. Criteria for recognition of hydrothermalmineral deposits in oceanic crust, Econ. Geol., v. 73, p.135-160.
Schilling, J-G., 1975. Azores mantle blob: rare-earth evidence,Earth Planet. Sci. Lett., v. 25, p. 103-115.
Schilling, J-G., Anderson, R. N., and Vogt., P., 1976. Rareearth, Fe and Ti variations along the Galapagos spreadingcentre, and their relationship to the Galapagos mantleplume, Nature, v. 261, p. 108-113.
Schilling, J-G. and Bonatti, E., 1975. East Pacific ridge (2°S-19°S) versus Nazca intraplate volcanism: rare-earth evi-dence, Earth Planet. Sci. Lett., v. 25, p. 93-102.
Sclater, J. G. and Klitgord, K. D., 1973. A detailed heat flow,topographic and magnetic survey across the Galapagosspreading centre at 86 °W, /. Geophys. Res., v. 78, p. 6951-6975.
Shibata, T., 1976. Phenocryst-bulk rock composition relationsof abyssal tholeiites and their petrologic significance,Geochim. Cosmochim. Ada, v. 40, p. 1407-1417.
Thompson, F., Bryan, W. B., Frey, F. A., and Dickey, T. S.,1978. Basalts and related rocks from Deep Sea Drilling sitesin the central and eastern Indian Ocean, Marine Geol., v.26, p. 119-138.
Thompson, F., Bryan, W. B., Frey, F. A., Dickey, T. S., andSuen, C. J., 1976. Petrology and geochemistry of basaltfrom DSDP Leg 34, Nazca Plate. In Yeats, R. S., Hart, S.R., et al., Initial Reports of the Deep Sea Drilling Project,v. 34: Washington (U.S. Government Printing Office), p.215-226.
Walker, D., Langhi, T., and Hays, T. F., 1972. Experimentalpetrology and origin of Fra Mau rocks and soil, Proc.Third Lunar Sci. Conf., v. 1, p. 797-817.
Williams, D. L., von Herzen, R., Sclater, R. P., and Ander-son, R. N., 1974. The Galapagos spreading centre: litho-spheric cooling and hydrothermal circulation, Geophys. J.Roy. Astron. Soc, v. 38, p. 587-608.
Wright, T. and Doherty, P., 1970. A linear programming andleast squares computer method for solving petrologic mix-ing problems, Geol. Soc. Am. Bull., v. 81, p. 1995-2008.
749