7. geochemistry of the lower sheeted dike complex, …
TRANSCRIPT
Erzinger, J., Becker, K., Dick, H.J.B., and Stokking, L.B. (Eds.), 1995Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 137/140
7. GEOCHEMISTRY OF THE LOWER SHEETED DIKE COMPLEX, HOLE 504B, LEG 1401
Joel W. Sparks2
ABSTRACT
Sixty-three samples representing 379 m of sheeted dikes from Deep Sea Drilling Project/Ocean Drilling Program Site 504Bhave been analyzed for major and selected trace elements by X-ray fluorescence. The samples range from microcrystalline aph-yric basalts to moderately phyric (2%-10% phenocrysts) diabase that are typically multiply saturated with plagioclase, olivine,and clinopyroxene, in order of relative abundance. All analyzed samples are classified as Group D compositions with moderateto slightly elevated compatible elements (Mg'-value = 0.65% ± 0.03%; A12O3 = 15.5% ± 0.8%; CaO = 13.0% ± 0.3%; Ni = 114± 29 ppm), and unusually depleted levels of moderate to highly incompatible elements (Nb < 1 ppm; Zr = 44 ± 7 ppm; Rb < 0.5ppm; Ba 1 ppm; P2O5 = 0.07% ± 0.02%). These compositions are consistent with a multistage melting of a normal oceanridge basaltic mantle source followed by extensive fractionation of olivine, plagioclase, and clinopyroxene.
Leg 140 aphyric to sparsely phyric (0%-2% phenocrysts) basalts and diabases are compositionally indistinguishable fromsimilarly phyric samples at higher levels in the hole. An examination of the entire crustal section, from the overlying volcanicsthrough the sheeted dikes observed in Leg 140, reveals no significant trends indicating the enrichment or depletion of CostaRica Rift Zone source magmas over time. Similarly, significant trends toward increased or decreased differentiation cannot beidentified, although compositional patterns reflecting variable amounts of phenocryst addition are apparent at various depths.Below 1700 mbsf to the bottom of the Leg 140 section, there is a broadly systematic pattern of Zn depletion with depth, theresult of high-temperature hydrothermal leaching. This zone of depletion is thought to be a significant source of Zn for thehydrothermal fluids depositing metal sulfides at ridge-crest hydrothermal vents and the sulfide-mineralization zone, located inthe transition between pillow lavas and sheeted dikes.
Localized zones of intense alteration (60%-95% recrystallization) are present on a centimeter to meter scale in many litho-logic units. Within these zones, normally immobile elements Ti, Zr, Y, and rare-earth elements are strongly depleted comparedwith "fresher" samples centimeters away. The extent of compositional variability of these elements tends to obscure primaryigneous trends if the highly altered samples are not identified or removed. At levels up to 40% (or possibly 60%) recrystalliza-tion, Ti, Zr, and Y retain their primary signatures. Although the mechanisms are unclear, it is possible that these intense alter-ation zones are a source of Y and rare-earth elements for the typically rare-earth-element-enriched hydrothermal vent fluids ofmid-ocean ridges.
INTRODUCTION AND GEOLOGIC SETTING
Leg 140 marked the sixth excursion of the Deep Sea DrillingProject/Ocean Drilling Program (DSDP/ODP) to core Hole 504B, lo-cated approximately 200 km south of the Costa Rica Rift Zone(1°13.61 l'N, 83°43.818'W). The 379 m drilled during this leg deep-ened the hole to a total depth of 2000.4 meters below seafloor (mbsf),making it the deepest penetration into oceanic crust (1725.9 m). Hole504B has already provided an unprecedented view of the oceaniccrust as it has penetrated 572 m of extrusive volcanics (Layer 2A),209 m of a mixed extrusive/sheeted dike transition zone (Layer 2B),and 945 m of sheeted dikes (Layer 2C). The main intent of Leg 140was to drill through and sample the Layer 2C-3 transition, from sheet-ed dikes to layered gabbros, the existence of which is inferred fromanalogies to ophiolite complexes (e.g., Coleman, 1977) and to plu-tonic rock sequences sampled in submarine fracture zones (e.g., VonHerzen, Robinson, et al., 1991).
The Costa Rica Rift Zone (CRRZ) is an intermediate-rate spread-ing ridge on the easternmost extremity of the Cocos-Nazca SpreadingCenter, bounded by the Ecuador Fracture Zone to the west, and thePanama Fracture Zone to the east (Hey et al., 1977; Lonsdale andKlitgord, 1978). The asymmetric spreading rates of 3.6 cm/yr to the
'Erzinger, J., Becker. K., Dick, H.J.B., and Stokking, L.B. (Eds.), 1995. Proc. OOPSci. Results, 137/140: College Station, TX (Ocean Drilling Program).
department of Geology, Stanford University, Stanford, CA 94305, U.S.A.
south and 3.0 cm/yr to the north (Hey et al., 1977) are consistent withthe characteristically uniform seafloor topography of an intermedi-ate-rate ridge. Hole 504B is situated between magnetic Anomalies3A and 4, in 5.9-m.y.-old oceanic crust.
The basalts and diabases from Hole 504B are typically multiplysaturated with respect to olivine, plagioclase, and clinopyroxene, andrange from 0% to 10% phenocrysts, of which plagioclase is the mostcommon. Considerable work from previous legs has focused on char-acterizing Hole 504B igneous compositions based on bulk-rock ma-jor and trace element analyses (Autio and Rhodes, 1983; Etoubleauet al., 1983; Hubberten et al., 1983; Marsh et al., 1983; Sharaskin etal., 1983;Emmermann, 1985; Kemptonet al., 1985;Tualetal., 1985;Autio et al., 1989), isotopic determinations (Barrett and Friedrichsen,1982; Barrett, 1983; Friedrichsen, 1985; Kusakabe et al., 1989;Shimizu et al., 1989), and electron microprobe analyses of freshglasses and selected mineral phases (Natland, et al., 1983; Kempton,1985; Kempton et al., 1985). Using the classification scheme of Au-tio and Rhodes (1983) and Kempton et al. (1985), over 98% of thesamples analyzed from this location are Group D (depleted) compo-sitions. The remaining 2% belong to either Groups M (moderatelyevolved), E (enriched), or T (transitional). Groups M and E corre-spond to normal Type I and Type II mid-ocean-ridge basalt (MORB)compositions (e.g., Bryan et al., 1976), whereas Group T is transi-tional between them. Group D bulk-rock and glass compositions varyfrom slightly olivine- to quartz-normative tholeiites, and are similarin many respects (compatible major- and trace-element abundances;Sr and Nd isotopes) to moderately evolved, normal Type I MORB.
81
J.W. SPARKS
Despite these similarities, Group D compositions exhibit marked lev-els of depletion in moderate to highly incompatible element abun-dances relative to Type I MORB. These characteristics have beenmodeled as either unusual primitive melt compositions (Natland etal., 1983), or the result of multistage melting of a normal MORBmantle source, followed by extensive crystal fractionation (Autio andRhodes, 1983; Autio, 1984; Kempton et al., 1985; Autio et al., 1989;Kusakabe et al., 1989; Shimizu et al., 1989).
Nearly all samples recovered from the sheeted dike section showmineralogical and compositional signs of subsolidus alteration rang-ing from slight (<10%) to extensive (95%) recrystallization. Thesechanges result from a complex interplay between alteration undergreenschist facies conditions (250°-380°C) near the ridge axis, pre-cipitation of anhydrite from deep penetrating seawater as the crustmoved off-axis into recharge zones, and low-temperature formationof zeolites (<150°C) as the section moved further off-axis (Alt, 1984;Alt et al., 1985, 1986a, 1989b; Shipboard Scientific Party, 1992).Bulk-rock compositional changes that take place during alterationhave been documented at Hole 504B for major elements (Honnorezet al., 1983; Hubberten et al., 1983; Noack et al., 1983; Alt and Em-mermann, 1985; Emmermann, 1985; Kawahata and Furuta, 1985; Altet al., 1989b), selected trace elements (Belyi et al., 1983; Hubberten,1983; Hubberten et al., 1983; Alt and Emmermann, 1985; Kawahataand Furuta, 1985; Alt et al., 1989a, 1989b), as well as stable and ra-diogenic isotopes (Barrett, 1983; Barrett and Friedrichsen, 1982,1983; Belyi et al., 1983; Honnorez et al., 1983; Friedrichsen, 1985;Mitchell and Terrell, 1985; Alt et al., 1986a, 1986b, 1989a, 1989b;Alt and Chaussidon, 1989; Kusakabe et al., 1989; Shimizu et al.,1989; Ishikawa and Nakamura, 1992). These compositional varia-tions are controlled by the style of alteration, which in turn, is depen-dent upon the given section of oceanic crust that is of interest. Low-temperature (0°-100°C), high water/rock ratio (10-100) seawater al-teration dominates the highly porous pillow lava section down to thebeginning of the transition into the sheeted dikes. Below the transi-tion, higher temperatures (200°-380°C; greenschist facies mineralo-gy) and significantly lower water/rock ratios (1-5), characterize thehydrothermal alteration in the sheeted dikes. Basalts in this sectionare heterogeneously altered (on a scale of centimeters to tens or hun-dreds of meters), exhibiting trends toward losses of SiO2, CaO, TiO2;variable changes in A12O3; gains in MnO, Na2O, CO2, H2O
+, S, MgO(slight), and oxidation of Fe; and some mobility of light rare-earth el-ements (Alt and Emmermann, 1985).
Ti, Zr, and Y are generally considered to be relatively immobileelements and extremely useful for evaluating igneous processes inbasalts that have been exposed to a range of subsolidus alteration pro-cesses since initial emplacement (e.g., Cann, 1970; Winchester andFloyd, 1976; Humphris and Thompson, 1978; Pearce and Norry,1979). However, recent work on the secondary alteration of basalts(e.g., Bienvenu et al., 1990; Price et al., 1991) has shown that theseelements are not always immobile, and the preliminary findings ofTi, Zr, and Y depletion in localized areas of intense alteration (Ship-board Scientific Party, 1992), appear to support this notion.
The main goals of this study are to document and discuss the vari-ations in primary igneous chemistry of the sheeted dike complex, asthe hole approaches the Layer 3 gabbros, and to document the varia-tions of Ti, Zr, Y, and selected rare-earth elements (REE) due to hy-drothermal alteration. Particular emphasis is placed on how thesevariations affect normal primary igneous trends. Finally, a discussionis presented concerning the presence and implications of a Zn deple-tion zone near the base of the Leg 140 section.
SAMPLING AND ANALYTICAL METHODS
Most samples used in this study were chosen as being representa-tive of the lithologic units described by the Shipboard Scientific Party
(1992), including three samples from the Leg 137 core (8.8 m totalrecovery). In addition, several lithologic units displaying a widerange of hydrothermal alteration were sampled in detail to documentthe compositional variability of Ti, Zr, Y, and REE resulting fromsubsolidus alteration.
Prior to analysis, samples were trimmed and ground to removeany large (>0.5 mm wide) hydrothermal veins and outside surfacesthat may have been contaminated by exposure to the coring bit. Sawmarks were removed on a silicon carbide grinding wheel, and sam-ples were cleaned with distilled water in an ultrasonic bath. Cleanedrock chips were reduced in a tungsten carbide (WC) jaw crusher, andground to a powder (<200 mesh) in a 20 mL WC-shatterbox (gra-ciously lent to the author by J.M. Rhodes, University of Massachu-setts). The amount of powder produced for each sample ranged from8 to 43 g, with most samples over 15 g. Powders to be used for majorelement analysis were ignited for 4 to 6 hr at 1000°C to determineloss on ignition and to oxidize iron to the trivalent state. Duplicateglass discs for each sample were prepared for X-ray fluorescence(XRF) major element analysis by fusing ignited sample powder witha lanthanum-bearing lithium borate flux (Norrish and Hutton, 1969).Samples for XRF trace analysis were prepared with approximately 5g of unfired sample powder pressed into a boric acid backing using ahydraulic press (Norrish and Chappell, 1977). Major and trace ele-ment (Nb, Zr, Y, Sr, Rb, Ga, Zn, Cu, Ni, Cr, V, Cl, and S) abundanceswere analyzed at Stanford University using an automated Rigaku S-MAX wavelength-dispersive spectrograph, equipped with an end-window, Rh-target X-ray tube. Major element matrix correctionswere made following the method of Norrish and Hutton (1969). Be-cause ignited powders were used, all major element analyses are re-ported on an anhydrous basis, with total iron as Fe2O3 (Fe2O3*). Losson ignition (LOT) is corrected for the oxidation of FeO to Fe2O3 (as-suming the average FeO content of a typically altered sample is 85%of the total iron content). Trace element analyses were corrected fornonlinear backgrounds and spectral interferences after Norrish andChappell (1977). Matrix absorption effects for all XRF trace ele-ments except Cl and S were measured using a modified Comptonscattering method (Reynolds, 1963; 1967; Walker, 1973; Willis,1989). Mass absorption coefficients for S and Cl were calculatedfrom major element analyses using mass absorption data tables (e.g.,Norrish and Chappell, 1977). All XRF analyses are presented in Ta-ble 1. Estimates of accuracy and precision (Table 2) were determinedby analyzing well-characterized basalt and diabase standards (USGSstandards BIR-1 and DNC-1, and ODP Leg 111 inter-laboratory stan-dard BAS- 111) together with the unknowns. These standards werechosen for their broad compositional similarities to the highly deplet-ed Leg 140 diabase to provide a realistic assessment of analytical pre-cision.
Using compositions determined by XRF as the basis for selection,rare-earth elements (La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm,Yb, and Lu) and Ba were determined on a subset of samples by in-ductively coupled plasma mass spectrometry (ICPMS) at Union Col-lege, Schenectady, New York, by J. Tepper. Unfired sample powderswere digested using an HNO3-HF digestion technique, followed bydilution with 0.1N HNO3. Internal standards (Cs and Ta) were addedto the solutions for monitoring instrumental drift. All samples wereanalyzed in duplicate, and the resulting concentrations were calibrat-ed against basalt standard NBS-688. Estimates of precision weremade using the formula
s = A1Σ(X~X)2/(Π-\)
where x and x are the observation and mean for replicate analyses ofthe same element in one sample, and n is the number of samples ana-lyzed (e.g., Autio and Rhodes, 1983). This method produces error
GEOCHEMISTRY, HOLE 504B
estimates that reflect the true errors associated with the concentra-tion levels observed in the Leg 140 diabases. Basalt and diabasestandards BIR-1 and DNC-1 were analyzed together with theunknowns as a measure of accuracy. The ICPMS analyses and ana-lytical statistics are presented in Table 3.
Various authors have reported significant sample contaminationfor W, Co, Ta, and to a lesser extent, Nb in samples that are groundin a WC shatterbox (Thompson and Bankston, 1970; Joron et al.,1980; Hickson and Juras, 1986). Joron et al. (1980) observed an av-erage 2 ppm increase in Nb content of basalts ground in WC vs. thoseground in agate, whereas Hickson and Juras (1986) suggest a smallerincrease ( 0.5 ppm) under similar conditions. In the present study,Nb contamination does not appear to be significant ( « 1 ppm), be-cause the maximum measured concentration is only 1.0 ppm, andmany of the analyzed samples are close to the lower limit of detection(0.3 ppm; 2 σ counting error). This reduced level of contaminationmay be due, in part, to decreased sample hardness resulting from sub-solidus alteration, relatively short grinding times ( 90 s), or the useof a "cleaner" tungsten carbide in the fabrication of the shatterboxused in this study (Spex Industries, Inc., pers. comm., 1992).
Two terms referred to throughout this study are the "degree of al-teration" and Mg'-value. The degree of alteration is the total percent-age of secondary mineralization as determined by either thin-sectionpoint counts or visual inspection of sliced core samples (ShipboardScientific Party, 1992). A rough estimate of accuracy and precision isapproximately ± 10% (absolute). Mg'-value is the molar ratio MgO/(MgO + FeO), where Fe2+/(Fe2+ + Fe3+) is equal to 0.90 (Basaltic Vol-canism Study Project, 1981).
IGNEOUS CHEMISTRY
The mineralogy, water content, and the abundance of normallymobile elements (Na, K, Rb, Sr, Cu, S, and Cl) indicate that all sam-ples selected for this study have undergone varying degrees of hydro-thermal alteration. The effects of this alteration must be quantified toassess primary igneous trends, many of which could be partially orfully obscured by compositional variations. To this end, samples thathave undergone more than 60% recrystallization of their groundmassphases are arbitrarily classified as "highly altered," and will not beused in discussions of igneous compositions unless specifically men-tioned. The selection of this alteration indicator is discussed later inthis study.
The following descriptions and discussion of igneous composi-tions include 74 samples analyzed by the Shipboard Scientific Party(1992), using analytical methods and calibration standards similar tothis study. For comparison, 10 of the analyzed shipboard sampleswere measured at Stanford, and the mean differences between thetwo data sets are listed in Table 2. Only Zn and Cr exhibit significantdeviations (11 and 12 ppm lower than this study, respectively), andeach shipboard analysis has been adjusted accordingly for more pre-cise comparisons. Data selected from previous legs to Hole 504Bwere produced at the University of Massachusetts (Autio and Rhodes1983; Kempton, et al., 1985; Autio et al., 1989). The Leg 111 inter-laboratory standard, analyzed at the University of Massachusetts(Autio et al., 1989) and in this study, is listed in Table 2 for compar-ison.
All Leg 137/140 dike samples can be classified as Group D com-positions (Autio and Rhodes, 1983; Kempton et al., 1985), with mod-erate to slightly elevated levels of compatible elements (MgO =7.7%-10.0%; Fe2O3* = 8.1%-11.7%; CaO = 11.7%-13.8%; A12O3 =14.2%-17.9%; Ni = 74-189 ppm; Cr = 127-456 ppm; V = 174-374ppm), unusually high CaO/Na2O ratios (up to 9.6), and highly deplet-ed concentrations of moderately to highly incompatible elements(TiO2 = 0.67%-1.14%; K2O 0.01%; P2O5= 0.04%-0.13%; Nb < 1ppm; Zr = 31-62 ppm; Sr = 43-77 ppm; Rb < 0.5 ppm).
Analyzed samples typically fall into three petrographic classes:microcrystalline to fine-grained aphyric basalt (<1% phenocrysts),fine- to medium-grained sparsely phyric diabase (l%-2% phenoc-rysts), and fine- to medium-grained, moderately phyric diabase (2%-10% phenocrysts). Although there is significant compositional over-lap, from aphyric to porphyritic samples (Table 4), small increases inA12O3 (15.0%-15.9%), Ni (98-119 ppm), Cr (271-337 ppm), and de-creases in TiO2 (0.99%-0.86%), Fe2O3* (10.5%-9.7%), Zr (51-42ppm), Y (26-22 ppm), and V (314-274 ppm) reflect the addition ofolivine, plagioclase, (Cr-rich) clinopyroxene, and Cr-spinel compo-nents.
The average composition of aphyric to sparsely phyric samplesfrom Leg 137/140 is nearly identical to that of the Group D basaltsfrom Layer 2B-2C transition zone-sheeted dikes (Legs 83 and 111),and the Layer 2A volcanic section of Legs 69 and 70, provided nor-mally mobile elements, K and Rb, are discounted (Table 5). The aph-yric samples are olivine-normative tholeiites with Mg'-values (0.60-0.70) and Ni abundances (74-143 ppm) similar to most moderatelyevolved mid-ocean-ridge basalts (Figs. 1 and 2). In contrast to mostMORB, however, these compositions are markedly lower in TiO2
and Zr at a given level of evolution such as Mg'-value or Ni abun-dance. Included in Figures 1 and 2 is Sample 111-504B-159R-1, 52-56 cm (Piece 9), the most evolved Group D composition (Mg'-value= 0.56; Ni = 55 ppm) observed at Hole 504B (Autio et al., 1989). A1-atm fractionation path calculated using the computer programCHAOS (e.g., Nielsen, 1985, 1990) is shown for comparison. Thestarting composition was primitive aphyric Sample 69-504B-35R-1,100 cm (Mg' value = 0.70; Autio and Rhodes, 1983), and fraction-ation increments of 2% were used (see also, Naslund et al., this vol-ume). Based on this path, the range of observed compositionsrepresents roughly 35% fractionation of olivine + plagioclase, fol-lowed by olivine + plagioclase + clinopyroxene. These calculationsagree closely with fractionation paths based on experimental (1 atm)melts of Group D Costa Rica Rift Zone compositions of Autio(1984).
Group D compositions, characterized by Zr and Nb concentra-tions that are extremely depleted relative to Type I MORB (i.e.,Group M) at similar stages of differentiation, have Zr/Nb ratios thatare similar to Type I basalts (Autio and Rhodes, 1983; Kempton etal., 1985; Autio et al., 1989). Figure 3 shows that all Leg 140/137dike samples fall into the same category—depleted incompatible el-ement signature and Zr/Nb ratios greater than 30. Figure 4 depictsvariations in Zr and Zr/Y for the basalts and diabases of Hole 504Bcompared with other mid-ocean-ridge basalts. Most Hole 504B sam-ples are highly depleted in Zr and Zr/Y with respect to typical mid-ocean-ridge basalts, and this level of depletion is not significantly af-fected by simple multiphase fractionation paths of Autio et al. (1989).
COMPOSITIONAL VARIATIONS WITH DEPTH
The incompatible element signatures of the Leg 137/140 dikesamples are relatively uniform between 1620 mbsf and 2000 mbsf(Fig. 5). Levels of Nb (< 1 ppm), Zr (44 ± 7 ppm), and Zr/Y (1.9 ± 0.2)are well within the typical ranges defined by previous legs to this site.Moreover, when samples from the entire hole are considered, thereare no significant systematic trends toward enrichment or depletionof incompatible elements with depth. Similarly, there are no system-atic differentiation trends with depth, as indicated by compatible ele-ments (Figs. 6 and 7). Instead, there are regions of increasedcompositional variability (at approximately 1300 and 1700 mbsf)separated by a section of minimal variation, near 1550 mbsf. Thispattern is consistent with variable addition of olivine, plagioclase,and clinopyroxene components through influxes of slightly moreprimitive magmas, or simple crystal settling within individual dikes.These patterns may be exaggerated by the unusually low rate of re-
J.W. SPARKS
Table 1. X-ray fluorescence analyses of Leg 137/140 basalts and diabases.
Core, section:Interval (cm):Piece:Depth (mbsf):Lithologic unit:Rock name:Alteration (%):
SiO, (wt%)TiO,Al,0,Fe,O,*MnOMgOCaONa,0
p2b,Total (anhyd.)
LOI* (wt%)Cl (ppm)S
Nb
ZrYSrRbGaZn
CuNiCrV
Mg -valueCaO/‰OZr/YTi/Zr
174R-1108-110
171577.4
195OPCn.d.
50.60.90
15.2910.500.188.16
12.901.67n.d.0.06
100.2
1.6297
776
<0.342.123.546.6<0.517.1708296
249298
0.6317.721.79
128
I74R-254-58
101578.3
195OPCn.d.
50.50.91
15.5710.420.187.95
12.861.72n.d.0.06
100.2
1.5682
789
0.443.624.746.7<0.515.5717995
262313
0.6277.481.77
125
177R-129-32
91604.8
202OPCn.d.
50.71.03
14.5311.330.208.03
12.431.81n.d.0.07
100.1
1.5273
1023
0.746.627.045.8<0.516.28381
142335
0.6096.871.73
132
186R-1116-119
I6B1627.5213A-D50
47.91.00
16.0810.210.189.28
13.461.62n.d.0.08
99.8
2.27416100
1.059.526.264.9<0.515.55727
131382290
0.6678.312.27
101
186R-218-21
61628.0
213A-D50
47.91.04
16.599.920.179.34
13.281.83n.d.
100.2
2.54408143
0.761.525.667.5<0.514.95831
143374285
0.6757.262.40
101
186R-250-53
101628.3
213A-D50
47.11.00
16.3610.690.189.09
13.561.42n.d.0.08
99.5
2.47436
28
0.659.225.S63.6<0.516.05613
137378299
0.6529.552.29
101
187R-15-9
21632.1
214A-D
10
50.21.09
14.5911.160.208.25
12.492.13n.d.0.07
100.1
1.51110
1115
0.955.727.560.6<0.516.3879085
186339
0.6195.862.03
117
187R-138^0
91632.4
215A-D50
51.00.92
14.4610.260.178.38
12.772.36n.d.0.06
100.3
2.21583100
<0.342.324.456.7<0.514.8591284
210320
0.6435.411.73
131
189R-170-73
161651.7218
M-POC10
50.40.94
15.1110.420.188.32
12.721.88n.d.0.07
100.1
1.58115
1146
0.744.424.24X.7<0.516.071
10695
273314
0.6376.771.83
127
189R-2110-113
161653.6
218M-POC
70
51.10.68
15.299.630.178.71
12.341.90n.d.0.05
99.9
3.44356153
0.530.117.749.7<0.512.96910
109367262
0.6666.481.70
135
190R-15-9
21655.2
218M-POC
70
53.70.60
14.0910.070.178.21
12.191.13n.d.0.04
100.2
5.09283364
<0.327.116.827.6<0.512.471
998
313239
0.64210.79
1.61134
I93R-117-20
61674.7
220M-POC
50
50.20.90
15.0210.480.198.50
12.502.01n.d.0.06
99.9
1.85262438
0.540.424.049.1<0.515.373
10391
268320
0.6416.221.68
134
I93R-128-31
91674.8
220M-POC
80
51.00.46
15.928.950.159.22
12.542.04n.d.0.03
100.3
3.51270108
<0.318.212.75 1.5<0.512.071
9113
406232
0.6946.151.43
151
193R-144-46
13A1675.0
220M-POC
56
50.50.97
15.499.870.188.33
13.041.94n.d.0.07
100.4
2.15236584
0.646.225.748.4<0.515.072
10399
291314
0.6506.721.80
125
193R-160-63
141675.1220
M-POC95
54.30.34
13.588.660.147.40
13.450.80n.d.0.02
98.7
7.46337
14132
<0.314.712.33 1.3<0.511.1537
83320156
0.65316.811.20
139
Table 1 (continued).
Core, section:Interval (cm):Piece:Depth (mbsf):Lithologic unit:Rock name:Alteration (%):
SiO, (wt%)TiO,A1,O,FeX>,*MnOMgOCaONa,O
p,b5Total (anhyd.)
LOI* (wt%)Cl (ppm)S
NbZrYSrRbGaZnCuNiCrV
Mg -valueCaO/Na,OZr/YTi/Zr
208R-139-41
101778.4
238S-PCO
10
49.90.76
15.679.310.179.40
13.251.71n.d.0.05
100.2
1.3173
661
<0.335.020.248.8<0.514.76399
141432261
0.6907.751.73
130
2O8R-1114-118
231779.1
239M-OPC
10
50.00.85
15.929.620.178.68
13.131.80n.d.0.06
100.2
1.3180
836
0.342.421.758.8<0.514.56085
113364260
0.6657.291.95
120
208R-269-72
121780.1
239M-OPC
15
49.70.82
16.359.330.178.6?
13.301.83n.d.0.06
100.2
1.3483
777
0.540.420.859.4<0.514.858
120375257
0.6717.271.94
122
209R-113-17
31787.6
240M-OPC
15
49.90.81
15.809.340.178.94
13.251.78n.d.0.06
100.0
1.2769
741
0.339.720.657.9<0.514.76385
114426256
0.6787.441.93
122
209R-1135-139
161788.9
240M-OPC
50
49.10.74
17.758.360.158.85
13.481.74n.d.0.05
100.2
1.81127571
<0.336.018.057.5<0.514.454
158134389229
0.7007.752.00
123
209R-2102-105
17A1790.0
240M-OPC
80
52.30.37
13.5711.250.179.58
11.281.56n.d.0.03
100.1
4.05510896
<0.318.012.448.4<0.510.469
8130458187
0.6527.231.45
124
211R-183-87
181799.3241
S-POC15
49.30.80
16.529.550.159.08
13.041.79n.d.0.06
100.3
1.73146282
0.338.621.653.9<0.514.053
115144358247
0.6777.281.79
124
213R-110-11
41812.6
242A-B60
50.81.04
14.4210.790.178.19
12.752.07n.d.0.07
100.3
1.82169112
0.748.627.354.5<0.516.7495674
169337
0.6266.161.78
129
213R-122-25
81812.7
242A-B60
50.51.09
14.5910.810.178.15
12.762.06n.d.0.07
100.2
1.79178210
0.348.426.254.4<0.515.150
10575
160347
0.6246.191.85
135
213R-179-83
211813.3
243S-POC
30
49.80.99
14.9210.470.188.43
12.812.04n.d.0.07
99.7
1.2489
838
0.549.425.654.7<0.516.1799095
264317
0.6396.281.93
120
213R-1121-123
28A1813.7
243S-POC
90
48.90.50
17.318.800.129.57
12.682.08n.d.0.03
100.0
2.66506
62
<0.319.814.254.9<0.513.85111
142409214
0.7056.101.39
150
214R-158-61
61819.2
244M-POC
90
49.10.84
16.249.250.169.14
13.391.80n.d.0.06
99.9
2.05143591
0.540.422.753.3<0.516.058
194147370254
0.6857.441.78
124
214R-195-101
121819.6
244M-POC
90
49.00.56
16.469.720.149.79
12.291.79n.d.0.04
99.8
2.46630
18
<0.328.015.751.7<0.513.253
9164446211
0.6896.871.78
119
214R-258-64
141820.6
244M-POC
90
49.20.56
16.888.670.139.57
12.971.86n.d.0.04
99.9
2.05602
66
<0.325.515.353.7<0.513.24811
160477222
0.7086.971.67
132
218R-117-19
51847.1
247A-D70
49.41.06
15.0610.710.208.61
12.662.33n.d.0.07
100.1
2.09411100
0.650.427.055.5<0.516.1514488
252346
0.6395.431.87
126
Notes: Rock name: A = aphyric (%); S = sparsely phyric (%); M = moderately phyric (%); D = diabase; B = aphanitic groundmass. Phenocrysts are listed in order of abundance foreach unit. P = plagioclase; O = olivine; C = clinopyroxene (Shipboard Scientific Party, 1992). OPC = phenocrysts as just described, but with no relation to relative abundance (firstthree samples only, from Leg 137). Alteration % = percentage of secondary minerals in the groundmass. n.d. = not determined.
GEOCHEMISTRY, HOLE 504B
Table 1 (continued).
194R-15-9
21680.5
220M-POC
15
49.80.84
16.129.560.178.58
13.081.85n.d.0.05
100.0
1.77158603
<0.339.422.246.7<0.515.36891
110348291
0.6647.071.77
128
196R-10-3
11696.5
221A10
50.51.11
14.4711.200.198.27
12.292.10n.d.0.08
100.2
1.71113933
0.654.329.349.1<0.516.1
1097988
172353
0.6195.851.85
122
197R-192-95
201703.7222
M-POC15
50.81.02
14.6311.030.198.14
12.611.86n.d.0.07
100.3
1.3382
843
0.347.226.846.1<0.516.0778286
228322
0.6196.781.76
130
197R-I105-107
231703.9
223S-POC
15
50.71.06
14.5311.270.208.31
12.371.89n.d.0.07
100.4
1.44119622
<0.350.128.643.7<0.516.4798291
220358
0.6196.541.75
127
198R-14-62
1712.2223
S-POC15
49.20.79
15.829.770.179.56
12.921.68n.d.0.05
100.0
1.5792
729
<0.336.921.754.4<0.514.26888
153371247
0.6837.691.70
128
198R-175-78
191713.0224AD60
50.81.01
14.6510.180.188.48
12.872.07n.d.0.07
100.3
1.68543
43
0.548.926.059.1<0.515.9591390
270336
0.6476.221.88
124
199R-171-76
171720.1226
M-PCO15
49.40.78
16.419.100.169.69
13.051.76n.d.0.05
100.4
1.5876
838
0.339.819.964.2<0.514.67088
168432242
0.7017.412.00
117
199R-181-84
191720.2226
M-PCO15
49.40.77
16.239.160.169.59
13.011.69n.d.0.05
100.1
1.5482
776
0.538.919.363.8<0.513.67084
171430232
0.6977.702.02
119
200R-153-57
121729.1227
M-POC15
49.30.75
16.539.090.159.43
13.101.75n.d.0.05
100.2
1.5053
780
<0.337.918.767.0<0.514.86082
170415218
0.6957.492.03
119
200R-I86-88
17A1729.5227
M-POC15
49.50.75
16.518.930.169.35
13.201.74n.d.0.05
100.2
1.3958
805
0.439.119.065.4<0.513.76487
161437232
0.6977.592.06
115
201R-15-9
21737.9228A-D60
51.20.73
14.8310.530.198.69
12.052.00n.d.0.05
100.2
1.80482
21
0.433.419.760.2<0.514.8681093
250287
0.6456.031.70
131
203R-14-7
21749.1
231A-D15
50.20.92
15.0110.500.188.78
13.031.89n.d.0.06
100.6
1.35n.d.n.d.
n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.
0.6486.89n.d.n.d.
203R-119-21
61749.2231A-D15
50.10.93
14.8210.350.198.61
13.271.96n.d.0.06
100.3
1.2093
687
0.343.024.852.6<0.514.97790
104377310
0.6476.771.73
130
203 R-l23-25
71749.2231A-D15
49.80.90
15.1510.450.188.86
12.871.81n.d.0.06
100.0
1.1884
951
0.444.125.355.0<0.514.78196
107367290
0.6517.111.74
123
205R-178-79
181757.8
234A-B20
49.91.10
14.4411.190.198.15
12.632.17n.d.0.07
99.8
1.25122950
0.755.627.959.9<0.515.7848881
181336
0.6165.821.99
118
205R-191-92
221757.9235
M-COP10
50.11.08
14.3911.000.198.32
12.822.22n.d.0.07
100.2
1.31139426
0.855.127.760.2<0.515.4759981
181341
0.6255.771.99
118
207R-10-4
11768.4236
A60
50.61.01
14.4210.700.208.58
12.771.96n.d.0.07
100.3
1.36235259
0.549.426.455.2<0.515.2596795
276341
0.6386.521.87
122
Table 1 (continued).
220R-10-2
11865.5
251A-D30
50.91.07
14.9410.160.187.92
13.062.31n.d.0.07
100.6
1.34449
39
0.450.826.861.2<0.516.0422679
231339
0.6325.671.90
126
222R-1100-104
191885.6
255S-OCP
60
n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.
n.d.184344
0.348.325.950.6<0.515.7569481
187324
n.d.n.d.1.86n.d.
222R-1121-125
231885.8
256A-D70
49.20.96
15.6910.210.159.02
12.821.67n.d.0.08
99.8
2.67348
27
0.657.325.165.6<0.515.54510
137462290
0.6607.682.28
101
224R-178-82
171904.5259A-D30
50.81.14
14.1711.740.198.01
11.692.25n.d.0.08
100.1
1.53538504
0.553.930.353.6<0.516.6578976
127374
0.6005.201.78
127
225R-113-16
31912.3259A-D50
49.80.98
15.299.680.188.81
13.201.92n.d.0.06
99.9
1.34359187
<0.356.724.277.3<0.515.340
107112416278
0.6676.892.34
103
225 R-l103-106
261913.2259A-D50
49.70.92
15.469.650.178.98
13.131.85n.d.0.06
99.9
1.50406
95
0.552.123.277.0<0.514.84058
115414174
0.6727.102.25
106
226R-318-24
21923.1260
M-COP60
49.20.77
16.919.340.169.10
13.321.76n.d.0.05
100.6
1.74153333
<0.335.920.054.7<0.515.248
148151363244
0.6827.571.80
129
227R-226-30
51926.1261
M-COP30
50.50.88
15.7210.000.178.28
13.091.77n.d.0.06
100.5
1.4049
735
<0.343.123.254.0<0.515.37680
101278279
0.6467.401.86
122
228R-10-5
11934.0262
M-OCP20
49.60.82
16.039.880.188.85
13.061.79n.d.0.06
100.2
1.8094
1063
0.639.021.952.7<0.515.26294
133394261
0.6637.291.78
125
230R-15-8IB
1953.1265
S-POC50
50.91.01
14.7910.760.188.29
12.182.15n.d.0.07
100.3
1.48446
90
<0.346.126.151.6<0.515.0444584
178327
0.6295.671.77
131
232R-14-6
11957.3265
S-POC5
50.50.98
14.9110.830.198.33
12.162.28n.d.0.07
100.2
0.6988
159
<0.347.726.052.5<0.516.4908688
168347
0.6295.331.83
123
233R-14-7
21967.5266A60
50.30.89
14.8510.130.158.90
13.241.80n.d.0.06
100.3
1.57647
23
<0.341.224.547.8<0.515.1392092
207323
0.6597.361.68
129
235R-10-3
11976.1267
M-PCO40
50.50.92
15.609.970.188.48
13.151.96n.d.0.07
100.8
1.1682
826
<0.344.723.460.2<0.515.4588697
358276
0.6526.711.91
123
235R-118-20
61976.3268
S-PC40
50.80.94
14.7510.350.158.46
12.871.88n.d.0.07
100.3
1.68183167
0.444.423.848.9<0.516.0429889
266324
0.6436.851.87
127
236R-10-4
11980.7269
M-POC30
49.50.78
17.028.820.158.55
13.591.85n.d.0.06
100.3
1.52117599
0.638.920.061.2<0.515.449
114125369243
0.6817.371.95
120
238R-I22-25
81992.3269
M-POC30
50.10.86
15.659.600.178.53
13.001.91n.d.0.06
99.9
1.14108795
<0.341.622.159.3<0.515.96384
105377273
0.6626.811.88
124
J.W. SPARKS
Table 2. X-ray fluorescence analytical statistics and comparisons.
SiO, (wt%)TiO,A1,0,FejO,*MnOMgOCaONa,0K20p2ö,LÖITotal
Cl (ppm)SNbZrYSrRbGaZnCuNiCrV
BIR-1
Mean
47.80.96
15.6511.380.1749.65
13.311.830.0280.022
-0.75100.1
n.d.n.d.0.5
15.614.9
108<0.516.373.0
129159380311
Std. dev.(iσ>
0.20.010.090.050.0030.090.050.030.0020.002
n.d.n.d.0.10.40.310.40.20.41124
DNC-
Mean
46.90.49
18.5810.040.151
10.1211.33
1.900.2330.066
-0.1099.7
60648
1.336.916.6
1433.3
14.570.689
246276149
!
Std. dev.(1CJ)
0.20.000.090.070.0030.070.050.040.0020.002
420.10.40.200.10.40.41113
BAS-111
Mean
49.50.79
16.439.040.1568.56
13.221.820.0110.0520.42
100.0
79729
0.539.620.159.9<0.514.668.387
127370245
Std. dev.
(iσ)0.20.010.070.040.0020.070.030.030.0010.002
350.10.20.20.40.20.4
0.51122
BAS-140
50.60.99
14.5811.080.198.14
12.411.85n.d.0.065n.d.
99.9
87
<0.344.826.645.5<0.515.6858485
192342
Detectionlimits (2<3)
n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.
420.30.50.50.50.50.8
11113
Mean difference
(Stanford-Shipboard)
-0.2-0.02-0.17
0.10-0.002-0.21-0.03-0.01
n.d.0.014
n.d.n.d.
-0.10.3
-0.1-0.4
n.d.n.d.
-110
-1-12
-2
Std. dev.
(iσ)0.250.010.130.150.010.120.080.06n.d.0.03
n.d.n.d.0.20.50.70.8
n.d.n.d.32366
Note: BAS-140 was ignited before major element analysis; therefore, its total is reported on an anhydrous basis.
Table 3. Barium and rare-earth element abundances, lithologic Unit 220.
Core, section:Interval (cm):Piece number:Depth (mbsf):
Ba (ppm)LaCePiNdSinEuGdTbDyHoErTinYbLu
I93R-117-20
61674.7
1.00.863.480.734.101.750.712.500.533.510.812.440.352.410.33
193R-128-31
91674.8
1.50.761.670.332.090.880.441.440.302.200.471.520.201.440.21
193R-14 4 ^ 6
13A1675.0
1.40.923.920.814.821.870.742.940.604.040.902.680.402.630.37
193R-160-63
141675.1
0.70.301.500.332.250.920.671.450.291.960.421.310.181.180.16
194R-15-9
21680.5
0.70.773.200.654.161.670.652.490.483.390.742.260.332.060.31
Precision(±lGppm)
0.250.030.100.010.090.060.040.050.020.070.020 . 1 2
0.010.080.01
BIR-1
210.772.590.462.781.310.582.110.473.030.742.080.302.240.33
Standards
DNC-1
102
3.608.651.125.181.570.672.180.432.830.702.100.332.060.30
NBS-688
1804.98
12.351.788.282.400.992.900.503.250.712.030.312.010.31
N
-
0.330.880.1120.600.1810.0690.2490.0470.3220.0700.2000.0300.2000.034
Notes: Precision = calculated precision (G) for Unit 220 (see text). Chondrite-normalizing value (N) for Dy is from Boynton (1989); all others from Haskin et al. (1968).
covery around 1550 mbsf, due to numerous problems with coringduring Legs 111, 137, and 140 (e.g., Shipboard Scientific Party,1992).
A significant trend of Zn depletion with depth (Fig. 8A) occurs inthe bottom 350 m of core, and is observed in all samples ranging from5% to 95% alteration (Shipboard Scientific Party, 1992; this study).With the exception of the sulfide mineralization zone (910-928 mbsf)and other transition zonebasalts, typical Zn concentrations for 504Bvolcanics and sheeted dikes vary between 60 and 80 ppm. However,beginning at about 1650 mbsf, the mean Zn concentration ( 70 ppm)decreases to about 42 ppm at 2000 mbsf. Similarly, the minimum Znabundance decreases (55-30 ppm) through the same depth interval.The relative Zn depletion as depicted in Figure 8A, is not as pro-nounced as the one reported by the Leg 140 Scientific Party. The dif-ference is that the Zn data produced by the shipboard XRF lab,although precise, was found to be consistently 12 ppm lower thanthose determined by this study and from previous legs, and has beencorrected for a more precise comparison. The resulting depletion pat-tern is different in one significant aspect only—the depth at which the
trend begins is very likely between 1700 and 1800 mbsf, and nothigher up in the sheeted dike section ( 1300 mbsf), as it may have ap-peared before (i.e., Shipboard Scientific Party, 1992; Dick et al.,1992). A comparable depletion trend for Mn or Cu has not been iden-tified (Fig. 8B, -C). Below 1600 mbsf, the pattern of MnO concentra-tions is similar to that of Fe2O3*, and Cu is highly scattered aroundtypical values of 75 to 110 ppm.
VARIABILITY OF "IMMOBILE" ELEMENTSTI, ZR, AND Y
As previously noted, all samples recovered from Leg 140 haveundergone minor to extensive levels of subsolidus alteration (Ship-board Scientific Party, 1992), as indicated by mineralogy (5%-95%secondary minerals), water content (LOF = 0.7%-7.5%), and con-centrations of relatively mobile elements (Cu = 7-194 ppm; S = 18-14132 ppm; Sr = 28-77 ppm; and Na2O = 0.8%-2.4%). The extent of
GEOCHEMISTRY, HOLE 504B
Table 4. Comparison of samples with variable phenocryst content.
SiCX (wt%)TiO,A1203
Fe.O,*MnOMgOCaO‰ OK,0p,ö,LOI* (wt%)Cl (ppm)S
NbZrYSrRbGaZnCuNiCrV
Mg -valueCaO/‰OZr/YTi/Zr
Mean
50.00.99
14.9810.520.188.59
12.861.98n.d.0.07
1.67335342
0.550.525.059.5<0.515.5625798
271314
0.646.611.95
120
Aphyric( 1% phenocrysts)
1 std. dev.
1.10.100.670.540.010.420.470.23n.d.0.01
0.41188383
0.27.32.38.0n.d.0.7
1936219843
0.021.030.24
12
Min.
47.10.73
14.179.650.157.92
11.691.42n.d.0.05
1.188421
0.333.419.747.8
n.d.14.7391074
127174
0.605.201.68
101
Max.
51.21.14
16.5911.740.209.34
13.562.36n.d.0.08
2.54647
1115
1.061.530.377.3n.d.
16.7109107143416374
0.679.552.40
135
Mean
50.20.93
15.1710.370.178.67
12.691.92n.d.0.06
1.39150474
0.344.424.650.5<0.515.56588
105267308
0.656.691.80
127
Sparsely phyric(l%-2% phenocrysts)
1 std. dev.
0.70.120.680.690.020.530.380.20n.d.0.01
0.31112295
0.15.52.73.7n.d.0.9
1618299141
0.030.840.074
Min.
49.20.76
14.539.310.158.14
12.161.68n.d.0.05
0.697390
0.335.020.243.7
n.d.14.0424581
168247
0.625.331.70
120
Max.
50.91.06
16.5211.270.209.56
13.252.28n.d.0.07
1.73446843
0.550.128.654.7n.d.
16.490
115153432358
0.697.751.93
131
Mean
49.90.86
15.919.730 . 1 7
8.6913.051.83n.d.0.06
1.54109739
0.441.622.255.8<0.515.26595
119337274
0.667.171.88
123
Moderately phyric(2%-10% phenocrysts)
1 std. dev.
0.50.090.800.740.010.490.280.12n.d.0.01
0.2555
200
0.24.22.76.8n.d.0.89
21298337
0.030.510.115
Min.
49.10.74
14.398.360.157.95
12.431.67n.d.0.05
1.1449
333
0.335.918.045.8
n.d.13.64879
142218
0.615.771.68
115
Max.
50.71.08
17.7511.330.209.69
13.592.22n.d.0.07
2.15262
11460.8
55.127.767.0n.d.
17.183
158171437341
0.707.752.06
134
this alteration is extremely heterogeneous on a scale of centimeters tometers, commonly ranging from slight, to nearly complete recrystal-lization within 10 cm of core. Localized areas of extensive alteration,dubbed alteration "patches" or "halos," typically represent 70% to90% replacement of primary minerals by actinolite, chlorite, albite,titanite, and trace epidote, prehnite, and anhydrite, whereas, sulfideminerals are virtually absent (Alt et al., 1989b; Shipboard ScientificParty, 1992). For detailed descriptions and compositions of alterationminerals in the sheeted dike complex, see Alt et al. (1985, 1986a,1989b), Ishizuka (1989), Kempton et al. (1985), Kusakabe et al.(1989), and various authors, this volume. At high degrees of alter-ation (>70%), elements that are normally considered to be relativelyimmobile during hydrothermal alteration (i.e., TiO2, Zr, Y) becomehighly variable, and exhibit even greater levels of depletion than haspreviously been observed at this site (as noted earlier, the degree ofalteration is simply the total percentage of secondary mineraliza-tion—predominantly chlorite and actinolite—as determined by eitherthin-section point counts or visual inspection of sliced core samples).Comparing TiO2, Zr, and the ratio Zr/Y with highly mobile Cu, illus-trates this point (Figs. 9-11). For samples with less than 50% alter-ation of the groundmass, Cu is typically between 75 and 115 ppm,with minimal variation in TiO2 (0.9% ± 0.1%), Zr (43 ± 5 ppm), andZr/Y (1.9 ± 0.1). With increased alteration Cu becomes highly vari-able, with the addition or removal of Cu-Fe sulfides, ranging from 10to 158 ppm at 50% to 60% alteration, and 7 to 194 ppm at greater than60% alteration. In contrast, the values of TiO2, Zr, and Zr/Y in sam-ples that contain up to 60% secondary minerals, are more variable,but they are well within typical ranges for Hole 504B. At greater lev-els of alteration, however, even these elements exhibit marked varia-tion, particularly toward depletion (TiO2 = l.l%-0.3%; Zr = 66-15ppm; Zr/Y = 2.6-1.2). For this reason, an arbitrary limit of 60%groundmass alteration is used by this study to filter out "altered" sam-ples when describing variations in igneous chemistry. Alternatively,a minimum limit of 35 to 60 ppm Cu could be used to achieve thesame purpose. Under this scheme, however, some of the units withZr/Y ratios greater than 2.2 may be unduly omitted.
The largest TiO2, Zr, and Y anomalies were recorded in lithologicUnit 220, a unit in which the average groundmass alteration is ap-
proximately 10%, but in localized patches reaches 60% to 95%(Shipboard Scientific Party, 1992). The composition of these patchesrelative to less altered samples in the unit is shown in Figure 12.Three samples (140-504B-193R-1, 17-20 cm; -193R-1, 44-46 cm;and -194R-1, 5-9 cm) represent slightly to moderately altered (LOI*= 1.9% ± 0.2%; Cu = 99 ± 7 ppm) sections of Unit 220. These sampleswere averaged and used as a "fresh" composition for normalizationagainst two highly altered samples (140-504B-193R-1, 28-31 cm;and -193R-1, 60-63 cm). Mobile elements in the highly altered sam-ples are exceedingly variable and/or depleted with respect to theslight to moderately altered samples (Na2O = 0.4-1.1 times that of theaverage "fresh" composition; S = 0.2-26'; Cu = 0.1'; Sr = 0.7-1.1').Concentrations of TiO2, Zr, Y, P, and most rare-earth elements, onthe other hand, are all depleted the same amount—roughly one-halfthose observed in the fresher samples—resulting in Zr/Y and Ti/Zrvariations that are more limited than the overall changes in abun-dance. Zr is more depleted than Ti, Y, and the middle to heavy rare-earth elements in the altered samples, producing a small degree of de-coupling. La and Eu exhibit behavior slightly different from the restof the rare-earth elements (Figs. 12 and 13). Both altered sampleshave small, positive Eu anomalies (Eu/Eu* < 2), and one sample(140-504B-193R-1, 28-31 cm) displays an unusual enrichment in Larelative to the trend of other light REEs. Positive Eu anomalies couldbe attributable to the presence of secondary albite (Humphris et al.,1978), or the interaction with seawater (e.g., Michard et al., 1983).However, the relative level of Na abundance in the two altered sam-ples is the opposite of what would be expected if Eu was indicativeof albite concentration—the sample with the strongest Eu anomalycontains the least Na. It is not known if the REE pattern for Sample140-504B-193R-1, 28-31 cm, represents a La enrichment, or a nega-tive Ce anomaly associated with the interaction of seawater. If it isgenuine La enrichment, the light rare-earth pattern differs from thepatterns associated with palagonitization of glassy volcanic pillows(e.g., Ludden and Thompson, 1979; Alt and Emmermann, 1985). Inthese cases, the light rare-earth elements from La to Sm are all en-riched relative to the fresh, glassy interiors, in contrast to the singularincrease in La as seen here.
87
J.W. SPARKS
Table 5. Hole 504B interleg comparisons-aphyric and sparsely phyric (0%-2%) samples only.
Legs 69/70 Leg 83 Leg 111
SiO, (%)TiO,A 1 2 O 3
Fe^O,*MnOMgOCaONa,OK,Op 2 bj
Nb (ppm)ZrYSrRbGaZnCuNiCrV
Mg -valueCaO/Na,OZr/Y
Mean
50.30.94
15.619.970.188.28
12.681.990.140.08
0.547.523.366.9
2.216.078
n.d.106362279
0.6476.392.04
1 std. dev.
0.30.040.560.560.020.180.240.130.090.01
0.13.01.46.02.00.67n.d.
194528
0.0120.430.09
Min.
49.80.88
14.908.830.167.99
12.171.800.040.06
0.443.02 1 . 5
57.00.4
15.266
n.d.76
282231
0.6285.66
Max.
50.71.00
16.7010.690.238.55
12.982.200.360.10
0.753.025.678.0
7.316.890
n.d.144463347
0.6717.092.18
Mean
50.11.00
14.8310.680.218.49
12.861.990.020.09
0.750.624.753.4
0.515.485
n.d.93
274292
0.6366.54
1 std. dev.
0.50.080.380.600.020.370.250.19
<O.Ol0.01
0.25.52.26.10.10.5
11n.d.7
5924
0.0210.61
Min.
48.80.84
14.319.670.177.72
12.251.68
<O.Ol
0.08
0.438.020.543.0
0.414.567
n.d.76
168245
0.5945.021.60
Max.
50.81.14
16.0811.620.269.09
13.182.440.030.11
1.159.028.066.0
0.916.4
104n.d.
102355326
0.6667.732.84
Mean
50.30.96
15.2310.460.188.33
12.721.900.010.06
0.848.324.557.9
0.115.874
n.d.104252271
0.6376.751.97
1 std. dev.
0.60.120.850.840.010.410.400.16
<O.Ol0.01
0.35.92.66.40.10.9
10n.d.
2810636
0.0290.740.10
Min.
48.90.75
13.809.260.157.16
11.701.59
<O.Ol0.05
0.338.420.143.7
0.014.155
n.d.5526
213
0.5585.271.78
Max.
51.11.24
16.9012.500.219.18
13.172.220.020.08
1.564.430.466.4
0.418.499
n.d.164388345
0.6868.212 . 1 2
+ Legs 69, 70, 83, and 111• Legs 137 and 140Δ Calculated fractionation path
0.65 0.70 0.750.50 0.55 0.60Mg'-value
Figure 1. Mg'-value vs. TiO2. Compared with over 1300 aphyric to sparsely phyric MORB compositions, the Group D basalts and diabases of Hole 504B (>270analyses) are unusually low in TiO2 for a given Mg'-value. The calculated liquid line of descent (e.g., Nielsen, 1985, 1990) represents 38% fractionation fromthe starting composition, and does not result in more "normal" MORB compositions after extensive differentiation. MORB field constructed after Autio et al.(1989). Dashed field boundary represents 95% of the MORB samples. Data from Legs 137 and 140 exclude highly (>60%) altered samples. Analytical error isapproximately the width of each plot symbol.
DISCUSSION
Primary Igneous Compositions
Discounting the mobile elements K, Rb, Sr, Cu, Mn, and Zn, the127 basalts and diabases examined in this study and by the Leg 140Shipboard Scientific Party (1992) are compositionally indistinguish-able from 98% of the lithologic units observed in the upper sheeteddikes, transition zone, and volcanics of Hole 504B. Based on Sr andNd isotopic composition (Barrett, 1983; Kusakabe et al., 1989;Shimizu et al., 1989), petrography (e.g., Natland et al., 1983; Kemp-ton et al., 1985; Shipboard Scientific Party, 1992), and compatible el-ement concentrations, Hole 504B Group D basalts and diabases aresimilar to moderately evolved, Type I MORB. On further inspection,however, Group D samples exhibit unusual levels of depletion in in-
compatible elements (Figs. 1-4, and 13) and are not related to normalType I compositions through simple fractionation. Both calculatedand experimentally determined 1-atm fractionation paths are incon-sistent with the notion that normal Type I compositions can be pro-duced through extensive fractionation of the depleted CRRZ magmas(Autio and Rhodes, 1983; Kempton et al., 1985; Autio et al., 1989,this study). The range of aphyric to sparsely (0%-2%) phyric compo-sitions observed at this location, from the most primitive to the mostevolved (Mg'-values = 0.70-0.56), requires approximately 35% frac-tionation to expiain, based on calculated levels of highly incompati-ble elements (Autio et al., 1989) and calculated liquid lines of descentfrom major element phase equilibria (e.g., Nielsen, 1985, 1990). Fur-ther differentiation will produce incompatible element levels that aretypical of Type I basalts, but only at the expense of reducing the com-patible elements to unacceptably low levels.
GEOCHEMISTRY, HOLE 504B
Table 5 (continued).
Mean
50.00.9415.0110.490.188.5212.791.930.0060.065
0.446.725.053.4<0.515.67193102280305
0.6436.831.80
Legs 137/140
1 std. dev.
0.60.120.630.680.010.470.340.160.0050.017
0.26.12.57.80.10.81517238636
0.0270.670.05
Min.
48.50.7114.178.990.157.7911.691.67
<0.0020.036
0.335.020.243.0<0.514.0425674127231
0.6005.201.64
Max.
50.81.14
16.5211.740.209.5713.272.280.0200.128
1.557.530.377.30.916.7109136155432374
0.7018.151.88
The CaO/Na2O ratios (5.2-9.6) for Group D basalts are larger thanwhat is commonly observed in normal MORB. These ratios are cal-culated to be in equilibrium with unusually anorthitic plagioclase (upto An92; Autio and Rhodes, 1983), and are in agreement with naturalplagioclase compositions from the volcanic section (up to An90; Nat-land et al., 1983) and the upper sheeted dikes (Kempton et al, 1985).Natland et al. (1983) have suggested that the highly depleted, Ca-richCRRZ basalts represent unusual primitive melt compositions, but thisnotion is inconsistent with observed concentrations of compatible el-ements such as Ni (102 ± 23 ppm in aphyric-sparsely phyric sam-ples), and the multisaturated nature of the CRRZ basalts. Otherworkers have argued that these compositions are the result of a mul-tistage melting of a normal MORB mantle source, followed by exten-sive crystallization (Autio and Rhodes, 1983; Autio, 1984; Kemptonetal., 1985; Autio etal., 1989; Kusakabeetal., 1989; Shimizu, 1989).Multistage melting produces melts that are extremely depleted in in-
compatible elements (Nb and Ba < 1 ppm), whereas compatible ele-ments, isotopic ratios (e.g., Sr and Nd), and incompatible elementratios with the more incompatible element in the denominator (e.g.,Zr/Nb) remain at normal, or slightly elevated levels (e.g., Duncan andGreen, 1980)—all features that are characteristic of Group D compo-sitions. Furthermore, multistage melting would raise CaO/Na2O ra-tios relative to normal MORB (Green et al., 1979; Duncan and Green,1980; Autio and Rhodes, 1983), and this increase would be reflectedin higher than normal anorthite content in equilibrium plagioclase.
It is possible that multistage melting is more common than previ-ously thought. Many mid-ocean-ridge basalts contain plagioclasephenocrysts that are too calcic to be in equilibrium with their hostglasses (Bryan et al., 1976; Rhodes et al., 1979; Fisk et al., 1982). Amultistage melting scenario could explain their presence, if smallamounts of depleted second-stage melts were mixed with large vol-ume, early melts during the normal production of mid-ocean-ridgebasalts. The resulting mixture would erase the depleted signature ofthe second-stage melts, but highly calcic phenocrysts would remain(Autio etal., 1989).
The mechanism(s) that produced and erupted CRRZ melts with solittle Type I MORB component is still unclear. Autio (1984) has sug-gested that early-stage melts could have been produced and removedfrom a normal MORB source at the East Pacific Rise (EPR) or theGalapagos thermal anomaly, but the difficulties of explaining thetransport of the residual mantle source and the fate of the enriched,early melts are unresolved. Furthermore, little is known about thegeographic extent of these Group D basalts. Among the variousDSDP sites along the Cocos-Nazca Spreading Center, only basaltsfrom Sites 501 and 505 (200 and 120 km south of the CRRZ, respec-tively; e.g., Etoubleau et al., 1983), and Site 425 (62 km north of theGalapagos Rift; e.g., Joron et al., 1980; Mattey and Muir, 1980;Srivastava et al., 1980) exhibit Group D characteristics. Basalts fromDSDP Site 424, approximately 20 km south of the Galapagos Rift,are clearly normal Type I MORB and represent a significant compo-sitional shift within 82 km of Site 425.
200
150
3 100
N
50
Legs 69, 70, 83, and 111
Legs 137 and 140
CRRZ fractionation path
50 100 200 250 300150Ni (ppm)
Figure 2. Ni vs. Zr. Group D compositions from Hole 504B exhibit low to moderate levels of Ni, but Zr values that are atypical of most MORB. The liquid lineof descent calculated from 1 -atm melting experiments of CRRZ basalts (Autio, 1984) demonstrates that common multiphase fractionation of CRRZ liquids willnot produce normal levels of Zr without severely depleting Ni. MORB field constructed as in Figure 1. Data from Legs 137 and 140 exclude highly (>60%)altered samples. Analytical error is within the width of each plot symbol.
J.W. SPARKS
200
Legs 69, 70, 83, and 111
Legs 137 and 140
10
Nb (ppm)
Figure 3. Nb vs. Zr. All samples from this study plot in the D Group classifi-cation of Autio and Rhodes (1983), bringing the total number of samples inthis class to over 270. Although the Zr/Nb ratios are similar to Group M (30), or normal Type I MORB, the Group D compositions are extremelydepleted in Nb and Zr. Group E (Type II: Enriched) and Type T (Transitional)are the other basalt classes observed at Hole 504B. Data from Legs 137 and140 exclude highly (>60%) altered samples. Analytical error is approxi-mately the width of each plot symbol.
Variations in Ti, Zr, Y, and REEs in HydrothermallyAltered Samples
The localized zones of Ti, Zr, Y, REE, Cu, and S depletion occurin areas of enhanced fluid flow and porosity in the halos around veinsand vugs (Alt et al., 1992; this study). The nature of these vugs is un-clear. It has been suggested that they represent primary pore space,subsequently filled with secondary minerals (Scientific Shipboard
N
+ Legs 69, 70, 83, and 111
• Legs 137 and 140
10 20 50
Zr (ppm)
100 200
Figure 4. Zr vs. Zr/Y. The depleted samples of Hole 504B have Zr/Y ratiosand Zr concentrations that are lower than typical MORB values. MORB fieldplotted as in Figure 1. Data from Legs 137 and 140 exclude highly (>60%)altered samples. Analytical error is within the width of each plot symbol.
Party, 1992), but the amount of primary porosity is likely to be van-ishingly small, given the depth and pressure of magma emplacement(e.g., Moore, 1979). An alternative explanation that the vugs are theresult of complete recrystallization of localized regions in glassy orfine-grained groundmass (Scientific Shipboard Party, 1992) is morelikely, as glass reacts more readily than the crystalline material be-tween 70° and 300°C (Seyfried and Bischoff, 1979, 1981). The deple-tion of these elements is not restricted to the lower sheeted dikecomplex. Alt and Emmermann (1985) reported TiO2 loss in hydro-thermally altered samples from Leg 83, which penetrated the lowervolcanic section, transition zone, and upper sheeted dike complex.Furthermore, data from lithologic Unit 60 (Emmermann, 1985) sug-
200
400
600
800
1000
S.1200
ΦQ
1400
1600
1800
2000
:+‰ <
i • i • i •
Volcanics
" jt Transition zone
•
P i
Sheeted dikecomplex
Nb error
i i i i i i
2 3
Nb (ppm)50 100
Zr (ppm)150
Figure 5. Depth vs. Nb (A), Zr (B), Zr/Y (C). All samples, aphyric to moderately phyric (<10%), with less than 60% alteration are plotted. Crosses = data fromLegs 69, 70, 83, and 111; solid circles = data from Legs 137 and 140. Except for Nb, analytical error is within the width of each plot symbol. The error bar forNb is given in Figure 5A.
GEOCHEMISTRY, HOLE 504B
200
400
600
800
1000
S.1200Φ
Q1400
1600
1800
2000
+
Volcanics
Transitionzone
+•Hftf»• + Sheeted dike _i4" - h ^ . complex
-
•
• i i i i i
13 14 15 16 17 18 19 20ALO, (wt%)2 3
8 9 10MgO (wt%)
50 100 150 200 250
Ni (ppm)
Figure 6. Depth vs. A12O3 (A), MgO (B), Ni (C). All samples, aphyric to moderately phyric ( 10%), with less than 60% alteration are plotted. Symbols and ana-lytical error are the same as in Figure 5.
200
400
600
800
1000 -
3.1200ΦQ
14001600
1800
2000
1 I ' I
+ 4
-
• i • i
TT
• I • I • ^
Volcanics
Transition •
L * +
+ Sheeted dike» * . complex
-
t
47 48 49 50 51 52 53 54SiO. (Wt%)
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8TiO. (Wt%)
8 9 10 11 12 13 14Fe,0 * (wt%)2 3
Figure 7. Depth vs. SiO2 (A), TiO2 (B), Fe2O3* (C). All samples, aphyric to moderately phyric (<10%), with less than 60% alteration are plotted. Symbols andanalytical error are the same as in Figure 5.
gests the loss of Zr (61-38 ppm), Y (27-17 ppm), Cu (118-15 ppm),and rare-earth elements as well.
Documenting the variability of Ti, Y, Zr, and REE is made easierby the fact that igneous trends involving these elements are reason-ably well defined for this location (Figs. 1 and 2). Comparing rela-tively fresh samples with the highly altered patches or halos, the mostnoticeable variation is a trend toward moderately strong depletion(30%-50%) that is echoed by many of the rare-earth elements andP2O5 (Figs. 12 and 13). The abundances of these elements in the al-teration patches are significantly lower than all other values reportedat this site and, as such, they are easily distinguishable from relatively
fresh material. Examples of alteration-induced enrichment, on theother hand, are more difficult to substantiate given that all samples liewithin the normal range of primary Ti, Y, and Zr variation. Most ofthe enriched samples, although moderately recrystallized, do not con-tain intense alteration patches, nor do they display significant differ-ences in secondary mineralogy (i.e., increased titanite or Fe-Tioxides). As a result, it is not currently known if the enrichments are aprimary signature, or an artifact of alteration.
The overall effect of the alteration zones is to significantly in-crease the compositional scatter of moderate to highly incompatibleelements (Fig. 14). Employing an alteration "filter" such as H2O
+
J.W. SPARKS
1600
1800
200020r÷ . àEVV• ,
Volcanics
Transition zone
Sheeted dikecomplex
40 60 80Zn (ppm)
100 120 0.08 0.12 0.16 0.2 0.24 0.28MnO (wt%)
50 100 150Cu (ppm)
200
Figure 8. Depth vs. Zn (A), MnO (B), Cu (C). All samples, regardless of the degree of alteration, are plotted. Open boxes = data from Legs 69, 70, 83, and 111;solid circles = data from Legs 137 and 140 with less than 50% alteration; crosses = data from Legs 137 and 140 with 50%-60% alteration; open triangles = datafrom Legs 137 and 140 with greater than 60% alteration. Analytical error is within the width of each plot symbol.
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
o, -
3%-49% alteration
50%-60% alteration
61%-95% alteration
50 100Cu (ppm)
150 200
Figure 9. Cu vs. TiO2. Most Leg 137/140 samples with up to 60% alterationfall within the normal range of Hole 504B variation, as defined by analysesof slightly altered samples in the upper levels of the hole. Samples with over60% alteration display abnormally low levels of TiO2. In the absence of pet-rographic data, highly mobile Cu could possibly be used as an alteration indi-cator. The amount of error in estimating alteration is roughly ±10%.
(e.g., Shipboard Scientific Party, 1992) or Cu content to separate theaffected samples is useful, but inexact. The arbitrary definition of <60% secondary minerals as being "unaltered" in terms of Ti, Zr, andY has been used in this study, but this method is inexact as well, andmay be too permissive if future work reveals the enriched samples tobe the result of alteration.
Rare-earth elements in hydrothermal fluids from mid-ocean-ridgevents are enriched 10 to 10,000 times relative to seawater (Michardet al., 1983; Campbell et al., 1988), and it is intriguing to think thatthe intense alteration zones could be one of the main sources of Y and
80
70
60
50E& 40
N30 -
20 -
10 -
_ ΛΔ •& + ^
+ T *• ù + < β k ^
+, ^^^^
f
* •• -ft-•
Φ
? -(0
COTj
omToE *-δ
50 100
Cu (ppm)
150 200
Figure 10. Cu vs. Zr. As in Figure 9, the Zr concentrations in most Leg 137/140 samples up to 60% alteration are within the typical Hole 504B range. Atmore intense levels of alteration, Zr is commonly lower.
REE enrichment. This notion is consistent with the typical depletionof Ti, Zr, Y, and the REE in alteration halos, and the fact that thereare no obvious regions of enrichment in the 504B crustal section.However, the REE pattern of a typical hydrothermal vent fluid has apronounced light-rare-earth enrichment, and a large (10-20), positiveEu/Eu* anomaly (Michard et al., 1983; Campbell et al., 1988). Thispattern could not be produced from the one example given here.
Similar suggestions for Ti and Zr in hydrothermal fluids are diffi-cult to make at this time because it is common practice to use titaniumcollection bottles or syringes when sampling high-temperature hy-drothermal vents (e.g., Von Damm et al., 1985; Michard and Al-
;
GEOCHEMISTRY, HOLE 504B
>-
N
50 100Cu (ppm)
150 200
Figure 11. Cu vs. Zr/Y. The Zr/Y ratio decreases slightly at high degrees ofalteration, but it is less affected than either Zr or Y, individually. There is nostrong decoupling of these two elements with increased alteration. The sam-ples with 50%-95% alteration and Zr/Y values between 2.2 and 2.6 arewithin the range described by relatively fresh samples from higher levels inthe hole. There is no significant evidence of Ti, Zr, or Y enrichment resultingfrom alteration.
barède, 1986). The level of Ti and Zr contamination precludesaccurate trace element analysis of these elements.
Zinc Depletion With Depth
The leaching of Zn and Cu, coupled with the presence of massivesulfide deposits, is a common feature in ophiolite complexes (e.g.,Troodos [Richardson et al., 1987] and Semail [Koski et al., 1987]).At the Troodos ophiolite, Zn and Cu depletion is strongly linked to
the presence of epidosite zones (quartz + epidote rocks) in the lowersections of the sheeted dike complex. The epidosites represent fossilconduits for hydrothermal vents and their associated massive sulfidedeposits. Within these zones, Zn and Cu are thought to be liberatedfrom the epidosite and surrounding diabase dikes by reactions withhydrothermal fluids at temperatures of 350°-^00°C (Richardson etal., 1987; Schiffman and Smith, 1988). Much of this hydrothermalactivity probably occurred as the crustal section moved off-axis andwas reheated by the emplacement of hypabyssal stocks (Schiffmanand Smith, 1988).
In contrast, at Hole 504B a stockwork-like sulfide-mineralizationzone (910-928 mbsf) occurs in the transition from sheeted dikes tothe overlying pillow lavas, but large-scale epidosite veins have yet tobe identified. Sulfide precipitation is thought to have resulted fromhydrothermal fluids (up to 38O°C) upwelling through low porositysheeted dikes and reacting with cooler, down-flowing seawater fromthe highly porous volcanic section (Alt, 1984; Altet al., 1985, 1986a;Honnorez et al., 1985). A likely source of hydrothermal Zn (and Cu)for sulfide mineralization zones and ridge-crest black smokers is theslightly to highly altered diabase of the (lower?) sheeted dikes. Typ-ical primary Zn concentrations at Hole 504B are approximately 70 ±10 ppm (Fig. 8A), but between 1700 and 2000 mbsf, Zn is progres-sively depleted in both "fresh" and highly altered samples. A compa-rable pattern of Zn depletion with depth has been observed in thelower section of sheeted diabase dikes of the Josephine ophiolite(Harper et al., 1988) where, as at Troodos, epidosites are thought tobe remnants of a circulation system that allowed transport of hydro-thermal fluids to ridge-crest black smokers and to sulfide mineraliza-tion zones. The lack of extensive epidosite veins at Hole 504Bsuggests that the hydrothermal circulation system in this section ofoceanic crust is less well developed than either the Troodos or Jose-phine ophiolite. Upward transport of metal-enriched hydrothermalfluids must be accomplished through open, partially mineralizedfractures that have yet to develop into large-scale epidosite zones.
High-temperature leaching experiments (Seyfried and Janecky,1985) help place constraints on the reaction temperatures that arenecessary for Zn and Cu depletion. These experiments indicate thatthe concentration of Zn and Cu in leachate solutions reaches a maxi-mum between 375° and 400°C and then decreases considerably at
10
oü
0)
03
Q.
C/D
193R-1, 28-31 cm (80% alteration)
193R-1, 60-63 cm (95% alteration)
0.1
Figure 12. Comparison of "fresh diabase" vs. alteration patches, Unit 220. XRF and ICP-MS analyses of alteration patches are normalized against the averageof three samples (140-504B-193R-1, 17-20 cm; -193R-1, 4 4 ^ 6 cm; and -194R-1, 5-9 cm) representing slight to moderate alteration. Four of the five Unit 220samples were taken within a 45-cm interval in one core section (193R-1); the remaining sample (140-504B-194R-1, 5-9 cm) was located within 10 cm of thetop of the next core.
J.W. SPARKS
10
CD
Ioü
Q.E 1CO
0.1
193R-1, 17-20 cm (50% alteration)
193R-1, 44-46 cm (56% alteration)
194R-1, 5-9 cm (15% alteration)
193R-1, 28-31 cm (80% alteration)
193R-1, 60-63 cm (95% alteration)
La CΘ Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
Figure 13. Chondπte-normalized rare-earth element patterns for Unit 220. Chondrite values from Haskin et al. (1968) and Boynton (1989).
N
• 3%-49% alteration
+ 50%-60% alteration
Δ 61 %-95% alteration
10 20 50
Zr (ppm)
100 200
Figure 14. Zr vs. Zr/Y. Highly altered samples add a considerable amount ofscatter to a relatively well defined Leg 137/140 field. The higher Zr/Y sam-ples are still within normal Hole 504B range, and it is likely that they repre-sent primary igneous signatures. MORB field constructed as in Figure 1.
425 °C. This temperature range is in broad agreement with estimatesfrom the dike section of Hole 504B. Alt et al. (1989b) estimated atemperature range of 250°-350°C in the Leg 111 section of the dikesbased on secondary mineralogy and oxygen isotope data. At greaterdepths, general increases in the proportions of actinolite replacingclinopyroxene, and the increasing stability of calcic plagioclase overalbite reflect rising temperatures with depth that approach upper-greenschist facies conditions in the Leg 140 section (Shipboard Sci-entific Party, 1992; Laverne et al., this volume).
The behavior of Cu (Fig. 8C) in the altered diabase is more com-plex than Zn, and may be related to Cu being a major component inonly one phase (chalcopyrite), whereas Zn is present in minoramounts in clinopyroxene, chlorite, actinolite, Ti-magnetite as wellas sulfides. According to Seyfried and Janecky (1985), Mn and Fe arestrongly leached at 425°C, depending on the pH of the hydrothermal
solution. With this in mind, similar Mn and Fe depletion zones maybecome apparent with further penetration of Hole 504B.
CONCLUSIONS
1. Disregarding highly mobile elements that are affected by sub-solidus alteration, the compositions of basalt and diabase samples re-covered on Leg 140 are virtually identical to those of the volcanicsection, transition zone, and the upper sheeted dike complex previ-ously sampled by Legs 69, 70, 83, and 111. All samples analyzed inthis study are classified as Group D compositions (e.g., Autio andRhodes, 1983; Kempton et al., 1985), with compatible element sig-natures that are similar to moderately evolved MORB, but with un-usual depletion in incompatible elements, and elevated levels ofCaO/Na2O ratios. These features are consistent with the multistagemelting of a normal MORB mantle source (e.g., Duncan and Green,1980), followed by extensive crystal fractionation (Autio andRhodes, 1983; Autio, 1984; Kempton et al., 1985).
2. Considering all of Hole 504B, no significant trends indicatingincompatible element enrichment or depletion of source magmasover time are apparent. Patterns of phenocryst addition and/or influxof slightly more primitive magma are identifiable at various levels inthe sheeted dikes, but broad trends signifying increased or decreaseddifferentiation are absent.
3. A broad zone of Zn depletion from about 1700 to 2000 mbsf re-sulting from high-temperature (375°-400°C) leaching is a likelysource of hydrothermal Zn for sulfide mineralization zones near thesheeted dike/transition zone interface, and ridge-crest black smokers.Manganese and iron may begin to exhibit similar depletion trends atgreater depths as temperatures increase (> 425°C). In contrast, Cu canbe virtually absent locally (down to 7 ppm), but there is no systematicvariation with depth.
4. Areas of intense alteration exhibit significant depletions in Ti,Zr, Y, and REE. These localized (centimeter to meter scale) alterationpatches can create significant compositional scatter in igneous trendsfeaturing these elements, if not accounted for. The patches are gener-ally associated with regions near fluid flow, such as veins or voidspace, and could be a significant source of Y and REE to the typicallyenriched hydrothermal fluids at ridge-crest vents.
94
GEOCHEMISTRY, HOLE 504B
ACKNOWLEDGMENTS
I thank L.K Autio and J.M. Rhodes for numerous discussions anddebates concerning the petrogenesis of CRRZ basalts. Their input hasbeen invaluable. I also thank J.A. Pearce and J.L. Karsten for promptreviews that added clarity and depth to the manuscript. This studywas supported by a grant from JOI/USSAC to J.W. Sparks and G.A.Mahood.
REFERENCES*
Alt, J.C., 1984. The structure, chemistry, and evolution of a submarinehydrothermal system, DSDP Site 504 [Ph.D. dissert.]. Univ. of Miami,Miami, FL.
Alt, J.C., Anderson, T.F., and Bonnell, L., 1989a. Geochemistry of sulfur in a1.3 km section of hydrothermally altered oceanic crust, DSDP Hole504B. Geochim. Cosmochim. Acta, 53:1011-1023.
Alt, J.C., Anderson, T.F., Bonnell, L., and Muehlenbachs, K., 1989b. Miner-alogy, chemistry, and stable isotopic compositions of hydrothermallyaltered sheeted dikes: ODP Hole 504B, Leg 111. In Becker, K., Sakai, H.,et al., Proc. ODP, Sci. Results, 111: College Station, TX (Ocean DrillingProgram), 27-40.
Alt, J.C., and Chaussidon, M., 1989. Ion microprobe analyses of the sulfurisotopic composition of sulfides in hydrothermally altered rocks, DSDP/ODP Hole 504B. In Becker, K., Sakai, H., et al., Proc. ODP, Sci. Results,111: College Station, TX (Ocean Drilling Program), 41-45.
Alt, J.C., Dick, H.J.B., Erzinger, J., Stokking, L.B., Agrinier, P., Allerton, S.,Boldreel, L.O., Fisk, M.R., Harvey, P.K., Iturrino, G.J., Johnson, K.T.M.,Kepezhinskas, P., Kelley, D.S., Laverne, C, Marton, F.C., McNeill,A.W., Naslund, H.R., Pariso, J.E., Pertsev, N., Pezard, P., Schandl, E.S.,Sparks, J.W., Tartarotti, P., Umino, S., Vanko, D.A., and Zuleger, E.,1992. Hydrothermal alteration of a 2 km "reference section" of the upperocean crust, DSDP/ODP Hole 504B. In The Transition From Basalt ToMetabasalt. IGCP Project 294, Davis, CA. (Abstract)
Alt, J.C., and Emmermann, R., 1985. Geochemistry of hydrothermallyaltered basalts: Deep Sea Drilling Project Hole 504B. In Anderson, R.N.,Honnorez, J., Becker, K., et al., Init. Repts. DSDP, 83: Washington (U.S.Govt. Printing Office), 249-262.
Alt, J.C., Honnorez, J., Laverne, C , and Emmermann, R., 1986a. Hydrother-mal alteration of a 1 km section through the upper oceanic crust, DeepSea Drilling Project Hole 504B: mineralogy, chemistry, and evolution ofseawater-basalt interactions. J. Geophys. Res., 91:10309-10335.
Alt, J.C., Laverne, C , and Muehlenbachs, K., 1985. Alteration of the upperoceanic crust: mineralogy and processes in Deep Sea Drilling ProjectHole 504B, Leg 83. In Anderson, R.N., Honnorez, J., Becker, K., et al.,Init. Repts. DSDP, 83: Washington (U.S. Govt. Printing Office), 217-247.
Alt, J.C., Muehlenbachs, K., and Honnorez, J., 1986b. An oxygen isotopicprofile through the upper kilometer of the oceanic crust, DSDP Hole504B. Earth Planet. Sci. Lett., 80:217-229.
Autio, L.K., 1984. Compositional diversity of mid-ocean ridge basalts: anexperimental and geochemical study with emphasis on the depletedCosta Rica Rift Zone basalts [Ph.D. dissert.]. Univ. of Massachusetts,Amherst, MA.
Autio, L.K., and Rhodes, J.M., 1983. Costa Rica Rift Zone basalts: geochem-ical and experimental data from a possible example of multistage melt-ing. In Cann, J.R., Langseth, M.G., Honnorez, J., Von Herzen, R.P.,White, S.M., et al., Init. Repts. DSDP, 69: Washington (U.S. Govt. Print-ing Office), 729-745.
Autio, L.K., Sparks, J.W., and Rhodes, J.M., 1989. Geochemistry of Leg 111basalts: intrusive feeders for highly depleted pillows and flows. InBecker, K., Sakai, H., et al., Proc. ODP, Sci. Results, 111: College Sta-tion, TX (Ocean Drilling Program), 3-16.
Barrett, T.J., 1983. Strontium- and lead-isotope composition of some basaltsfrom Deep Sea Drilling Project Hole 504B, Costa Rica Rift, Legs 69 and70. In Cann, J.R., Langseth, M.G., Honnorez, J., Von Herzen, R.P.,
* Abbreviations for names of organizations and publications in ODP reference lists fol-low the style given in Chemical Abstracts Service Source Index (published byAmerican Chemical Society).
White, S.M., et al., Init. Repts. DSDP, 69: Washington (U.S. Govt. Print-ing Office), 643-650.
Barrett, T.J., and Friedrichsen, H., 1982. Strontium and oxygen isotopic com-position of some basalts from Deep Sea Drilling Project Hole 504B,Costa Rica Rift, Legs 69, 70. Earth Planet. Sci. Lett., 60:27-38.
, 1983. Oxygen- and hydrogen-isotope composition of somebasalts from Deep Sea Drilling Project Hole 504B, Costa Rica Rift, Legs69 and 70. In Cann, J.R., Langseth, M.G., Honnorez, J., Von Herzen,R.P., White, S.M., et al., Init. Repts. DSDP, 69: Washington (U.S. Govt.Printing Office), 637-642.
Basaltic Volcanism Study Project, 1981. Basaltic Volcanism on the Terres-trial Planets: New York (Pergamon Press).
Belyi, V.M., Migdisov, A.A., Barskaya, N.V., and Grinenko, V.A., 1983.Sulfur-isotope composition of hydrothermal sulfides from Hole 504B,Deep Sea Drilling Project Leg 70, Costa Rica Rift. In Cann, J.R., Lang-seth, M.G., Honnorez, J., Von Herzen, R.P., White, S.M., et al., Init.Repts. DSDP, 69: Washington (U.S. Govt. Printing Office), 619-627.
Bienvenu, P., Bougault, H., Joron, J.L., Treuil, M., and Dmitriev, L., 1990.MORB alteration: rare-earth element/non-rare-earth hygromagmaphileelement fractionation. Chem. Geol., 82:1-14.
Boynton, W.V., 1989. Cosmochemistry of the rare earth elements: condensa-tion and evaporation. In Lipin, B.R., and McKay, G.A. (Eds.), Geochem-istry and Mineralogy of the Rare Earth Elements. Rev. Mineral, 21:1—24.
Bryan, W.B., Thompson, G., Frey, F.A., and Dickey, J.S., 1976. Inferred set-tings and differentiation in basalts from the Deep Sea Drilling Project. / .Geophys. Res., 81:4285-4304.
Campbell, A.C., Palmer, M.R., Klinkhammer, G.P., Bowers, T.S., Edmond,J.M., Lawrence, J.R., Casey, J.F., Thompson, G., Humphris, S., Rona, P.,and Karson, J.A., 1988. Chemistry of hot springs on the Mid-AtlanticRidge. Nature, 335:514-519.
Cann, J.R., 1970. Rb, Sr, V, Zr, and Nb in some ocean floor basaltic rocks.Earth Planet. Sci. Lett., 10:7-11.
Coleman, R.G., 1977. Ophiolites: Ancient Oceanic Lithosphere: New York(Springer-Verlag).
Dick, H.J.B., Erzinger, J.A., Stokking, L.B., Agrinier, P., Allerton, S., Alt,J.C., Boldreel, L.O., Fisk, M.R., Harvey, P.K.H., Iturrino, G.J., Kelley,D.S., Kepezhinskas, P.K., Laverne, C , Marton, F., McNeill, A.W.,Naslund, H.R., Pariso, J., Pertsev, N.N., Pezard, P., Schandl, E.S., Sparks,J.W, Tartarotti, P., Umino, S., Vanko, D.A., and Zuleger, E., 1992. ODPdrills deepest hole in ocean crust. Eos, 73:537.
Duncan, R.A., and Green, D.H., 1980. Role of multistage melting in the for-mation of oceanic crust. Geology, 8:22-26.
Emmermann, R., 1985. Basement geochemistry, Hole 504B. In Anderson,R.N., Honnorez, J., Becker, K., et al., Init. Repts. DSDP, 83: Washington(U.S. Govt. Printing Office), 183-199.
Etoubleau, J., Corre, O., Joron, J.L., Bougault, H., and Treuil, M., 1983.Costa Rica Rift: variably depleted basalts in the same hole. In Cann, J.R.,Langseth, M.G., Honnorez, J., Von Herzen, R.P., White, S.M., et al., Init.Repts. DSDP, 69: Washington (U.S. Govt. Printing Office), 765-773.
Fisk, M.R., Bence, A.E., and Schilling, J.-G., 1982. Major element chemistryof Galapagos Rift Zone magmas and their phenocrysts. Earth Planet. Sci.Lett., 61:171-189.
Friedrichsen, H., 1985. Strontium, oxygen, and hydrogen isotope studies onprimary and secondary minerals in basalts from the Costa Rica Rift,Deep Sea Drilling Project Hole 504B, Leg 83. In Anderson, R.N., Hon-norez, J., Becker, K., et al., Init. Repts. DSDP, 83: Washington (U.S.Govt. Printing Office), 289-295.
Green, D.H., Hibberson, W.O., and Jaques, A.L., 1979. Petrogenesis ofmid-ocean ridge basalts. In McElhinny, M.W. (Ed.), The Earth: Its Ori-gin, Structure and Evolution: London (Academic Press), 265-290.
Harper, G.D., Bowman, J.R., and Kuhns, R., 1988. A field, chemical and sta-ble isotopic study of subseafloor metamorphism of the Josephine Ophio-lite, California-Oregon. / . Geophys. Res., 93:4625^1656.
Haskin, L.A., Frey, F.A., Schmitt, R.A., and Smith, R.H., 1968. Meteoric,solar, and terrestrial rare-earth distributions. In Physics and Chemistry ofthe Earth 7: New York (Pergamon), 167-321.
Hey, R., Johnson, L., and Lowrie, A., 1977. Recent plate motion in the Gal-apagos Area. Geol. Soc. Am. Bull., 88:1385-1403.
Hickson, C.J., and Juras, S.J., 1986. Sample contamination by grinding. Can.Mineral., 24:585-589.
Honnorez, J., Alt, J.C., Honnorez, B.M., Laverne, C , Muehlenbachs, K.,Saltzman, E., and Ruiz, J., 1985. Stockwork-like sulfide mineralizationin young oceanic crust: Deep Sea Drilling Project Hole 504B. In Ander-
95
J.W. SPARKS
son, R.N., Honnorez, J., Becker, K., et al., Init. Repts. DSDP, 83: Wash-ington (U.S. Govt. Printing Office), 263-282.
Honnorez, J., Laverne, C, Hubberten, H.-W., Emmermann, R., and Mue-hlenbachs, K., 1983. Alteration processes in Layer 2 basalts from DeepSea Drilling Project Hole 504B, Costa Rica Rift. In Cann, J.R., Langseth,M.G., Honnorez, J., Von Herzen, R.P., White, S.M., et al., Init. Repts.DSDP, 69: Washington (U.S. Govt. Printing Office), 509-546.
Hubberten, H.-W., 1983. Sulfur content and sulfur isotopes of basalts fromthe Costa Rica Rift (Hole 504B, Deep Sea Drilling Project Legs 69 and70). In Cann, J.R., Langseth, M.G., Honnorez, J., Von Herzen, R.P.,White, S.M., et al., Init. Repts. DSDP, 69: Washington (U.S. Govt. Print-ing Office), 629-635.
Hubberten, H.-W., Emmermann, R., and Puchelt, H., 1983. Geochemistry ofbasalts from Costa Rica Rift Sites 504 and 505 (Deep Sea DrillingProject Legs 69 and 70). In Cann, J.R., Langseth, M.G., Honnorez, J.,Von Herzen, R.P., White, S.M., et al., Init. Repts. DSDP, 69: Washington(U.S. Govt. Printing Office), 791-803.
Humphris, S.E., Morrison, M.A., and Thompson, R.N., 1978. Influence ofrock crystallization history upon subsequent lanthanide mobility duringhydrothermal alteration of basalts. Chem. GeoL, 23:125-137.
Humphris, S.E., and Thompson, G., 1978. Hydrothermal alteration of oce-anic basalts by seawater. Geochim. Cosmochim. Ada, 42:107-125.
Ishikawa, T., and Nakamura, E., 1992. Boron isotope geochemistry of theoceanic crust from DSDP Hole 504B. Geochim. Cosmochim. Ada,56:1633-1639.
Ishizuka, H., 1989. Mineral paragenesis of altered basalts from Hole 504B,ODP Leg 111. In Becker, K., Sakai, H., et al., Proc. ODP, Sci. Results,111: College Station, TX (Ocean Drilling Program), 61-76.
Joron, J.L., Briqueu, L., Bougault, H., and Treuil, M., 1980. East PacificRise, Galapagos spreading center and Siqueiros Fracture Zone hygro-magmaphile elements—a comparison with the North Atlantic. InRosendahl, B.R., Hekinian, R., et al., Init. Repts. DSDP, 54: Washington(U.S. Govt. Printing Office), 725-735.
Kawahata, H., and Furuta, T, 1985. Sub-sea-floor hydrothermal alteration inthe Galapagos spreading center. Chem. GeoL, 49:259-21 A.
Kempton, P.D., 1985. An interpretation of contrasting nucleation and growthhistories from the petrographiç analysis of pillow and dike chilled mar-gins, Hole 504B, Deep Sea Drilling Project, Leg 83. In Anderson, R.N.,Honnorez, J., Becker, K., et al., Init. Repts. DSDP, 83: Washington (U.S.Govt. Printing Office), 165-181.
Kempton, P.D., Autio, L.K., Rhodes, J.M., Holdaway, M.J., Dungan, M.A.,and Johnson, P., 1985. Petrology of basalts from Hole 504B, Deep SeaDrilling Project, Leg 83. In Anderson, R.N., Honnorez, J., Becker, K., etal., Init. Repts. DSDP, 83: Washington (U.S. Govt. Printing Office), 129-164.
Koski, R.A., Haymon, R.M., and Abrams, M., 1987. Ridge-crest andoffridge hydrothermal-discharge zones mapped beneath massive sulfidedeposits in the Samail Ophiolite, Northern Oman. Eos, 68:1545.(Abstract)
Kusakabe, M., Shibata, T, Yamamoto, M., Mayeda, S., Kagami, H., Honma,H., Masuda, H., and Sakai, H., 1989. Petrology and isotope characteris-tics (H, O, S, Sr, and Nd) of basalts from Ocean Drilling Program Hole504B, Leg 111, Costa Rica Rift. In Becker, K., Sakai, H., et al., Proc.ODP, Sci. Results, 111: College Station, TX (Ocean Drilling Program),47-60.
Lonsdale, P., and Klitgord, K.D., 1978. Structure and tectonic history of theeastern Panama Basin. Geol. Soc. Am. Bull, 89:981-999.
Ludden, J., and Thompson, G., 1979. An evaluation of the behavior of rare-earth elements during the weathering of sea-floor basalts. Earth Planet.Sci. Lett., 43:85-92.
Marsh, N.G., Tarney, J., and Hendry, G.L., 1983. Trace element geochemis-try of basalts from Hole 504B, Panama Basin, Deep Sea Drilling ProjectLegs 69 and 70. In Cann, J.R., Langseth, M.G., Honnorez, J., VonHerzen, R.P., White, S.M., et al., Init. Repts. DSDP, 69: Washington (U.S.Govt. Printing Office), 747-763.
Mattey, D.P., and Muir, I.D., 1980. Geochemistry and mineralogy of basaltsfrom the Galapagos Spreading Center, Deep Sea Drilling Project Leg 54.In Rosendahl, B.R., Hekinian, R., et al., Init. Repts. DSDP, 54: Washing-ton (U.S. Govt. Printing Office), 755-771.
Michard, A., and Albarède, F, 1986. The REE content of some hydrothermalfluids. Chem. GeoL, 55:51-60.
Michard, A., Albarède, F, Michard, G., Minster, J.F, and Charlou, J.L.,1983. Rare-earth and uranium in high-temperature solutions from EastPacific Rise hydrothermal vent field (13°N). Nature, 303:795-797.
Mitchell, J.G., and Terrell, D.J., 1985. Downhole variation of potassium,inert gas abundance, and apparent K-Ar age in basalts from the CostaRica Rift, Hole 504B, Leg 83, Deep Sea Drilling Project. In Anderson,R.N., Honnorez, J., Becker, K., et al., Init. Repts. DSDP, 83: Washington(U.S. Govt. Printing Office), 305-312.
Moore, J.G., 1979. Vesicularity and CO2 in mid-ocean ridge basalt. Nature,282:250-253.
Natland, J.H., Adamson, A.C., Laverne, C, Melson, W.G., and O'Hearn, T.,1983. A compositionally nearly steady-state magma chamber at theCosta Rica Rift: evidence from basalt glass and mineral data, Deep SeaDrilling Project Sites 501, 504, and 505. In Cann, J.R., Langseth, M.G.,Honnorez, J., Von Herzen, R.P., White, S.M., et al., Init. Repts. DSDP, 69:Washington (U.S. Govt. Printing Office), 811-858.
Nielsen, R.L., 1985. EQUIL.FOR: a program for the modeling of low-pres-sure differentiation processes in natural mafic magma bodies. Comp.GeoscL, 11:531-546.
, 1990. Simulation of igneous differentiation processes. In Nicholl,J., and Russell, J.K. (Eds.), Modern Methods of Igneous Petrology:Understanding Magmatic Processes. Mineral. Soc. Am., Rev. Mineral.,24:65-105.
Noack, Y., Emmermann, R., and Hubberten, H.-W, 1983. Alteration in Site501 (Leg 68) and Site 504 (Leg 69) basalts: preliminary results. In Cann,J.R., Langseth, M.G., Honnorez, J., Von Herzen, R.P., White, S.M., et al.,Init. Repts. DSDP 69: Washington (U.S. Govt. Printing Office), 497-508.
Norrish, K., and Chappell, B., 1977. X-ray fluorescence spectrometry. InZussman, J. (Ed.), Physical Methods in Determinative Mineralogy: NewYork (Academic Press), 201-272.
Norrish, K., and Hutton, J.T., 1969. An accurate X-ray spectrographicmethod for the analysis of a wide range of geological samples. Geochim.Cosmochim. Ada, 33:431^4-53.
Pearce, J.A., and Norry, M.J., 1979. Petrogenetic implications of Ti, Zr, Yand Nb variations in volcanic rocks. Contrib. Mineral. Petrol., 69:33-47.
Price, R.C., Gray, CM., Wilson, R.E., Frey, FA., and Taylor, S.R., 1991. Theeffects of weathering on rare earth element, yttrium and barium abun-dances in Tertiary basalts from southeastern Australia. Chem. GeoL,93:245-265.
Reynolds, R.C., Jr., 1963. Matrix corrections in trace element analysis by X-ray fluorescence: estimation of the mass absorption coefficient by Comp-ton scattering. Am. Mineral., 48:1133—1143.
, 1967. Estimation of mass absorption coefficients by Comptonscattering: improvements and extensions of the method. Am. Mineral.,52:1493-1502.
Rhodes, J.M., Dungan, M.A., Blanchard, D.P., and Long, P.E., 1979. Magmamixing at mid-ocean ridges: evidence from basalts drilled near 22°N onthe Mid-Atlantic Ridge. Tectonophysics, 55:35-61.
Richardson, C.J., Cann, J.R., Richards, H.G., and Cowan, J.G., 1987. Metal-depleted root zones of the Troodos ore-forming hydrothermal systems,Cyprus. Earth Planet. Sci. Lett., 84:243-253.
Schiffman, P., and Smith, B.M., 1988. Petrology and oxygen isotopegeochemistry of a fossil seawater hydrothermal system within the SoleaGraben, northern Troodos ophiolite, Cyprus. J. Geophys. Res., 93:4612-4624.
Seyfried, W.E., Jr., and Bischoff, J.L., 1979. Low temperature basalt alter-ation by seawater: an experimental study at 70°C and 150°C. Geochim.Cosmochim. Ada, 43:1937-1947.
, 1981. Experimental seawater-basalt interaction at 300°C and 500bars: chemical exchange, secondary mineral formation, and implicationsfor the transport of heavy metals. Geochim. Cosmochim. Ada, 45:135-147.
Seyfried, W.E., Jr., and Janecky, D.R., 1985. Heavy metal and sulfur trans-port during subcritical and super critical hydrothermal alteration ofbasalt: influence of fluid pressure and basalt composition and crystallin-ity. Geochim. Cosmochim. Ada, 49:2545-2560.
Sharaskin, A.Y., Migdisov, A.A., Rostschina, I.A., and Miklishansky, A.Z.,1983. Major- and trace-element chemistry of Hole 504B basalts and theiralteration products (Costa Rica Rift, Deep Sea Drilling Project Leg 70).In Cann, J.R., Langseth, M.G., Honnorez, J., Von Herzen, R.P., White,S.M., et al., Init. Repts. DSDP, 69: Washington (U.S. Govt. PrintingOffice), 775-789.
Shimizu, H., Mori, K., and Masuda, A., 1989. REE, Ba, and Sr abundancesand Sr, Nd, and Ce isotopic ratios in Hole 504B basalts, ODP Leg 111,Costa Rica Rift. In Becker, K., Sakai, H., et al., Proc. ODP, Sci. Results,111: College Station, TX (Ocean Drilling Program), 77-83.
96
GEOCHEMISTRY, HOLE 504B
Shipboard Scientific Party, 1992. Site 504. In Dick, H.J.B., Erzinger, J.,Stokking, L.B., et al., Proc. ODP, Init. Repts., 140: College Station, TX(Ocean Drilling Program), 37-200.
Srivastava, R.K., Emmermann, R., and Puchelt, H., 1980. Petrology andgeochemistry of basalts from Deep Sea Drilling Project Leg 54. InRosendahl, B.R., Hekinian, R., et al., Init. Repts. DSDP, 54: Washington(U.S. Govt. Printing Office), 671-693.
Thompson, G., and Bankston, D.C., 1970. Sample contamination from grind-ing and sieving determined by emission spectroscopy. Appl. Spectros.,24:210-219.
Tual, E., Jahn, B.M., Bougault, H., and Joron, J.L., 1985. Geochemistry ofbasalts from Hole 504B, Leg 83, Costa Rica Rift. In Anderson, R.N.,Honnorez, J., Becker, K., et al., Init. Repts. DSDP, 83: Washington (U.S.Govt. Printing Office), 201-214.
Von Damm, K.L., Edmond, J.M., Grant, B., Measures, C.I., Walden, B., andWeiss, R.F., 1985. Chemistry of submarine hydrothermal solutions at21 °N, East Pacific Rise. Geochim. Cosmochim. Ada, 49:2197-2220.
Von Herzen, R.P., Robinson, P.T., et al., 1991. Proc. ODP, Sci. Results, 118:College Station, TX (Ocean Drilling Program).
Walker, D., 1973. Behavior of X-ray mass absorption coefficients nearabsorption edges: Reynolds method revisited. Am. Mineral., 58:1069—1072.
Willis, J.P., 1989. Compton scatter and matrix correction for trace elementanalysis of geological materials. In Ahmedali, S.T. (Ed.), X-ray Fluores-cence Analysis in the Geological Sciences; Advances in Methodology.Geol. Assoc. Can. Short Course, 7:91-140.
Winchester, J.A., and Floyd, P.A., 1976. Geochemical magma type discrimi-nation: application to altered and metamorphosed basic igneous rocks.Earth Planet. Sci. Lett., 28:459-469.
Date of initial receipt: 20 April 1993Date of acceptance: 13 January 1994Ms 137/140SR-021