21. PETROGRAPHY AND RELICT MINERALOGY OF SERPENTINITES FROM DEEP SEADRILLING PROJECT LEG 841

Sherman H. Bloomer, Scripps Institution of Oceanography2

ABSTRACT

Eleven serpentine samples from DSDP Leg 84 and four serpentinized ultramafic samples from Costa Rica and Gua-temala were described and their relict mineral compositions measured by electron microprobe to try to determine the or-igin of the Leg 84 serpentinites and their relationship to the ultramafic rocks of the onshore ophiolites. The Leg 84 sam-ples comprise more than 90% secondary minerals, principally serpentine, with hematitic and opaque oxides, and minortalc and smectites. Four distinct textural types can be identified according to the distribution of opaque phases andsmectite. Remnants of spinel, olivine, orthopyroxene, and clinopyroxene occur variously in the samples; spinal occursin all the samples. Textural evidence suggests that the serpentinites were originally clinopyroxene-bearing harzburgites.Relict mineral compositions are refractory and relatively uniform: olivine, Fo^ ^ ^ 9; orthopyroxene, E n ^ ^ ; clinopy-roxene, Wo47 En50 Fs3; spinels, Cr/Cr + Al = 0.4-0.6. 567A-29-2, 30-35 cm has slightly more magnesian olivines(Fo92) and orthopyroxene, and more aluminous spinels (Cr/Cr + Al = 0.3). These compositions are similar to thoseinferred for refractory upper-mantle materials and also fall within the range of compositions for relict minerals in abys-sal peridotites. They could be of oceanic origin. The onshore samples include serpentinites, a clinopyroxene-bearingharzburgite, and a clinopyroxenite. They too have magnesium-rich silicate assemblages, but relative to the drilled sam-ples have more iron-rich olivines (Fogo) and more aluminous and sodic pyroxenes; spinels which are clearly relicts arevery aluminum-rich (Cr/Cr + Al = 0.1-0.25). These samples are most likely mantle materials, but significantly lessdepleted. Their relationship to the drilled samples is unclear. Serpentinites were the most common basement materialsrecovered during Leg 84, and there appears to be a bimodal assemblage (basalt/diabase and serpentine) of igneousrocks sampled from the trench slope. Diapirism of serpentine throughout the trench slope and forearc is suggested as anexplanation for this distribution of samples.

INTRODUCTION

One of the objectives of DSDP Leg 84 was to samplethe basement beneath the landward trench slope of theMiddle America Trench off Guatemala. It had been sug-gested, on the basis of geophysical data, that this slopewas composed of imbricated, landward-dipping slices ofocean crust (Seely et al., 1974; Ladd et al., 1978). Gravi-ty and magnetic data, particularly a magnetic high par-allel to the trench-slope break, are consistent with thepresence of high-density crustal rocks at shallow depths,and indicate that these rocks may be an extension ofmid-Tithonian-upper Santonian ophiolitic rocks exposedin coastal Costa Rica (Woodcock, 1975; Couch 1976;Couch and Woodcock, 1981). The recovery of clasts ofserpentinite, basalt, and chert in cores and dredges alongthe trench slope supports this interpretation (Ibrahim etal., 1979).

On DSDP Leg 67, a partial transect was drilled acrossthe trench and forearc, but the primary objective of drill-ing to basement in the nearshore slope was abandonedwhen large amounts of gas hydrated were encounteredin the slope (von Huene, et al., 1980). The offshoreslope is floored by basalts petrographically and chemi-cally similar to normal ocean-ridge and spilitic ocean-ridge basalts (Maury et al., 1982); these are overlain bylower Miocene chalks (Coulbourn et al., 1982). The one

von Huene, R., Aubouin, J., et al., Init. Repts. DSDP, 84: Washington (U.S. Govt.Printing Office).

2 Present address: Department of Geology, Duke University, Box 6729 College Station,Durham, NC 27708.

onshore hole which penetrated basement (494A on thelower trench slope) yielded greenschist-facies metavol-canics which are similar to arc andesites or basaltic an-desites in both clinopyroxene and bulk-rock chemistry(low Ti, Ni, high SiO2 contents; Maury et al., 1982).These are overlain by Campanian-Maestrichtian lime-stones, indicating that this portion of trench is not cur-rently undergoing accretion (Coulbourn et al., 1982).

The occurrence of Cretaceous sediments overlying base-ment in Hole 494A is another indication that there maybe a relationship between the forearc basement and theupper Cretaceous ophiolitic rocks exposed on the SantaElena, Nicoya, Oso, and Azuro peninsulas of CentralAmerica. The Nicoya Complex is a deformed volcani-clastic unit that includes metamorphosed pillow basaltsoverlain by chert and siliceous sediments; it was proba-bly formed in mid-Tithonian to late Santonian time andemplaced in the margin in Santonian-Campanian time(Galli-Olivier, 1979). It was probably emplaced as a nappe,and was later uplifted in the Eocene-Oligocene (Kuij-pers, 1980). The Nicoya Complex has been interpretatedas oceanic material, on the basis of a very few analysesof basalts and the occurrence of pelagic sediments over-lying the pillow basalts (Weyl, 1969; Pichler and Weyl,1975). The Santa Elena Peninsula includes a complexof ultramafic rocks with some gabbros, thrust over asection of basic volcanic rocks; these two sections areseparated by a breccia (Galli-Olivier, 1979; Azema andTournon, 1982). The peridotites are serpentinized harz-burgites with olivine, orthopyroxene, brown spinel, andsmall amounts of clinopyroxene, and are cut by dikesof gabbro and dolerite (Azema and Tournon, 1982).

643

S. H. BLOOMER

The volcaniclastic unit over which these are thrust maybe an extension of the Nicoya Complex; it is overlain byUpper Cretaceous-Paleocene sediments with ultrabasicclasts (Azema and Tournon, 1982).

Leg 84 was planned in part to complete the objectiveof Leg 67: to sample the basement of the trench slopeand to examine the relationship between that basementand the onshore complexes. Indeed, at five of six sites,"ophiolitic"-type materials—most commonly serpentin-ites, with lesser amphibolitized gabbros and sheared me-tavolcanics—were recovered (site chapters, this volume).The oldest sediments recovered overlying this basementwere upper Cretaceous limestone in Hole 567A, consis-tent with the age of ophiolitic rocks in Costa Rica.

The only primary minerals remaining in the serpen-tinite samples are scattered grains of olivine, orthopy-roxene, clinopyroxene, and spinel. The compositions ofthose minerals provide the only information available toestablish the initial compositional character of these ser-pentinites and the relationship of the Leg 84 serpentin-ites to the peridotites of the costal ophiolites. To thatend, analyses of primary minerals in 11 samples fromLeg 84 and 4 samples from the onshore ultramafic com-plexes of Central America are presented here.

SAMPLES AND METHODS

Cores from four of the five holes from which serpentinites wererecovered (566, 567A, 566C, and 570) were examined, and sampleswhich appeared to have some relict minerals were selected. The sam-ples examined are listed in Table 1. Four samples from ultramafic out-crops in Guatemala and Costa Rica were also analyzed. CR-39 andCR-205 are a serpentinite breccia and a peridotite clast in a breccia, re-spectively, and are both from exposures along the Rio Potrero Grandein the Santa Elena Peninsula (samples supplied by N. Lundburg, Uni-versity of California, Santa Cruz). D-50 and DL-90 are two Guatema-lan peridotites similar in age to those in Costa Rica (approximatelyBarremian-Cenomanian, emplaced after the early Campanian; T. W.,Donnelly, pers. comm., 1983). D-50 is from Minas de Oxec, Alta Vera-paz (Rosenfeld, 1981); DL-90 is from Quebrada Patache, northeasternSanarte Quadrangle (Lawrence, 1975); both were supplied by T. W.,Donnelly, State University of New York Binghamton.

Table 1. Drilled samplesused in this study, withtextural type.

Sample(interval in cm)

566-6-1, 31-33566-9-1, 9-11566C-6-1, 31-33566C-7-1, 13-14566C-7-1, 78-80567A-28-1, 121-122567A-29-2, 30-35567A-29-2, 140-142570-42-1, 40-41570-42-2, 53-55

Texturaltype a

2, 331112, 33244

1 = clear serpentine with red-dish or orange spots.2 = serpentine with abun-dant, fine-grained opaques,disseminated and in stringers.3 = serpentine with dark grayto opaque sports or stains.4 = clear serpentine with scat-tered, fine opaque grains andsmectites.

Polished, 16-mm-diameter sections were prepared for each sample.Microprobe analyses were done on a Cameca CAMEBAX automatedelectron-probe microanalyzer with working conditions of 15 kV and15 nA. Count time was 20s; spot size averaged 15 µm, but varied be-tween 2 and 20 µm for some samples. Smithsonian and BRGM stan-dards were used in the calibration. Modal analyses in Table 2 are ap-proximate because of the very small area counted.

PETROGRAPHY

All the drilled samples can be termed serpentinitessensu stricto; all comprise greater than 90% secondarymineral assemblages dominated by serpentine (Table 2).The only clue to the original composition of the samplesis in the few grains of unaltered primary minerals. Spinelsare the most common of these relict grains, occurring insmall percentages in all samples. They are red-brown todark red-brown in thin section, range in size from 0.3 to1.5 mm, and occur most commonly in elongate holly-leaf or anhedral habits and are fractured and broken (Fig.1). In some samples there is a weak lineation, marked byelongation of single spinel grains and by alignments ofgroups of grains. Olivines are relatively uncommon asrelict grains, reflecting their susceptibility to serpentini-zation. They occur as 0.1-mm clear, anhedral fragments,rarely with well-developed kink-bands suggesting an ear-lier, high-temperature deformation. Areas of simultane-ous extinction among groups of relict grains define small(0.5-mm), decidely elongate areas in some samples. Or-thopyroxene occurs as 0.4 to 1.4-mm relicts in bastites;although some have exsolution lamellae of clinopyrox-ene, well-developed lamellae are the exception rather thanthe rule. Most orthopyroxenes are completely replacedby bastite-type structures of clear serpentine, serpentinewith semiparallel concentrations of opaques, or serpen-tine with linear exsolutions of dark reddish spinels. Thecorrespondence of these bastite structures to replacedpyroxene is confirmed by the composition of the serpen-tine, which has distinctly higher contents of Cr and Al,and lower Ni, than does the adjacent massive serpen-tine. Such compositions indicate a pyroxene progenitor(Table 3). Many of the bastites are bent or broken (as in-dicated by opaque lamellae); some appear to have been

Table 2. Representative modal analyses of Leg 84 serpentinites and on-shore peridotites (in vol.%).

SerpentineOpaquesBastiteOlivineOrthopyroxeneSpinelClinopyroxeneVeinsOtherSum% Secondary

minerals

566C-7-1,13-14 cm

88.61.46.9

——0.11.11.9

—100.0

98.8

566C-6-1,31-33 cm

75.20.2

18.31.21.01.02.90.10.2 a

100.1

94.0

567A-29-2,140-142 cm

90.7b8.8

—0.3——0.2c

10.0

99.7

570-42-1,111-113 cm

75.8b

12.80.81.80.71.5

°•5Λ6.2d

100.1

95.3

DL90

12.1e

72.512.60.81.9

0.2100.1

12.3

a Talc-tremolite.Abundant opaques, included with serpentine.

c Exsolution in bastite.Smectites.

e — indicates not detected.

644

567A-28-1, 121-122 cm

' ' " : - . ' - ' !**- "V V - ' . - ' . - ', r / • / *

566C-4-1, 31-33 cm<*l $•.**,&

2.0 mm 2.0 mm

570-42-1, 111-113 cm 567A-29-2, 36-38 cm2.6 mm 1.3 mm

567A-29-2, 140-142 cm1.0 mm

566C-7-1, 13-14 cm_ -^ i.? -.'*

1.0 mm

Figure 1. Photomicrographs of textures in serpentinites from Leg 84. Upper left, basite with spotty and disseminated opaques; up-per right, clear serpentine with relict olivine, spinel, orthopyroxene, and clinopyroxene; middle left, Hole 570 sample with relictorthopyroxene and clinopyroxene and patches of greenish brown smectites; middle right, serpentine with some opaque spots, rel-ict olivine, spinel, and orthopyroxene; lower left, serpentinite with abundant, disseminated opaques, one relict spinel; lowerright, serpentinite with opaque spots, bastite, and relict clinopyroxene.

sheared into slightly elongate patches (Fig. 1). There areminor amounts of hydrogarnet developed in some sam-ples. Clinopyroxene, when it occurs, appears to be quiteresistant to serpentinization, and occurs as 0.1 to 0.8-mm anhedral, fractured grains. Well-developed exsolu-tion lamellae of orthopyroxene are not common. Theclinopyroxenes most often are associated with bastite

pseudomorphs, which may occur in clumps of 2 to 5.The clinopyroxene appears to be a part of the original,equilibrium mineral assemblage in the rock.

The serpentine is pale green, typically with a mesh-work structure, though fibrous serpentine is not uncom-mon. Veins of clear serpentine often cut through themeshwork matrix. Talc occurs in some of the bastites,

645

S. H. BLOOMER

Table 3. Analyses of secondary materials in Leg 84 serpentinites.

SiO2

TiO2

Al2$3FeOT

MnOMgOCaONa2OK2OP2O5NiOCr 2O 3

Total

570-42-2, 53-55 cm

smectite

36.220.000.46

11.90.09

34.430.100.10.060.000.140.15

83.65

Serpentine

Bastite

36.040.010.865.870.09

37.100.230.080.030.040.110.64

81.11

Massive

37.550.030.074.290.13

38.460.140.090.030.000.460.03

81.29

567A-29-2, 30-35 cm

Talc/serpentine

44.170.033.434.000.06

34.070.080.150.030.010.120.58

86.73

Serpentine

Bastite

34.680.042.415.210.16

36.880.190.010.020.000.030.61

80.25

Massive

38.330.000.554.680.15

38.020.140.020.050.040.330.01

82.32

Note: T = total iron as FeO.

particularly those which include some relict orthopyrox-ene, but tremolite was not identified in any of thesesamples.

There are four distinct textural varities of serpentin-ite, distinguished principally on the basis of secondaryopaque phases. The first type comprises a very clear, palegreen serpentine with scattered grains of a blood-red,translucent mineral, most probably hematitic in com-position (Fig. 1). Similar hematitic material occurs inveins in all textural types. The second variety has paleserpentine with discrete (0.03-0.2 mm) dark to opaquepatches, probably a result of varying concentrations ofopaque phases (Fig. 1). The third type has abundant (upto 15 modal %) fine-grained (<0.03 mm), disseminatedopaque phases. There are samples which grade in tex-ture between the latter two types. These differences intexture may reflect local differences in oxygen fugacitiesand water/rock ratios during serpentinization. The tex-tural type to which each sample corresponds is keyed inTable 1. Samples from Site 570 exhibit a fourth texture,in which low-temperature retrograde metamorphism hasproduced crosscutting patches and veins of green to palebrown smectite, in a clear serpentine with a few scat-tered opaques.

Though these samples are serpentinized, the distribu-tion of bastite pseudomorphs and relict orthopyroxeneand clinopyroxene indicates that most were probably cli-nopyroxene-bearing harzburgites before serpentinization.Spinel probably made up 0.1-1 modal °7o, orthopyrox-ene 10-20%, clinopyroxene 1-3%, and olivine 76-89%of the original rocks. Despite more extensive serpentini-zation, the drilled samples are similar in texture to mate-rials from Santa Elena described as "serpentinized harz-burgitic peridotite showing relicts of olivine, orthopyrox-ene, brown spinel and, less frequently, clinopyroxene"(Azema and Tournon, 1982).

The onshore samples are petrographically more di-verse than the drilled samples. CR-39 is a clast-support-ed breccia with lithic serpentine fragments ranging insize from pebbles to fine sand. Opaque spinels are theonly relict minerals identifiable, and the interpretationof some of these as relict is suspect. CR-205 is a clast

from a similar breccia, and is a pale green-brown ser-pentinite with scattered opaques. There are abundant 0.1to 1-mm olivine remnants that commonly occur as sin-gle grains or in groups of grains defining 0.7-mm-wideextinction areas; kink-bands are also common. Ortho-pyroxene occurs as large grains (up to 5 mm) with exso-lution lamellae and spinel and clinopyroxene inclusions.It more commonly occurs as 0.5- to 1-mm grains whichare distinctly undulose. Clinopyroxene occurs as smaller,anhedral grains with some exsolution lamellae of ortho-pyroxene; spinels are small and greenish brown with blackrims.

DL-90 is a less serpentinized clinopyroxene-bearingharzburgite whose modal analysis is quite similar to theestimates for the original mineralogy of the drilled sam-ples (Table 2). It comprises large (0.1-6 mm) interlock-ing olivine grains and lesser orthopyoxene grains cut bya meshwork of serpentine. The olivines are undulose,commonly have well-developed kink-bands, and showmuch grain-boundary recrystallization. The orthopyrox-enes have few exsolution lamellae, and are often kinkedand bent. Anhedral 1- to 3-mm clinopyroxenes with fineexsolution lamellae occur interstitially between the oliv-ine and orthopyroxene. Spinels are greenish brown andhighly irregular in shape.

D-50 is a rare petrographic type among the Guatema-lan peridotites (T. W. Donnelly, pers. comm., 1982). It isa pyroxenite, dominantly of orthopyroxene (0.5-6 mm),with lesser clinopyroxene (0.5-2 mm) and some intersti-tial olivine. The thin exsolution lamellae in the orthopy-roxene are deformed, bent, and sheared off; lamellae inthe clinopyroxene show evidence of the same deforma-tion. There is minor green-brown spinel, rarely inter-grown with orthopyroxene, and interstitial, anhedral ol-ivines, which are extensively altered. The sample is veinedand partially altered to serpentine and smectite.

MINERAL COMPOSITIONS

Mineral analyses were done on 26-mm-diameter pol-ished sections with the CAMEBAX automated electron-probe microanalyzer at the Scripps Institution of Ocean-ography. Working conditions were 15 kV acceleratingvoltage and 15 nA sample current, with count times of20 s. Spot size averaged 15 µm, but varied between 2and 20 µm for some analyses.

Olivines in all samples have uniformly high concen-trations of MgO and NiO, and low amounts of CaO(Table 4). The olivine in drilled samples cluster aroundF090.8 and Fo92 (Fig. 2). The latter, very high numbermay reflect more extensive subsolidus reequilibration in567A-29-2, 30-35 cm. Olivine compositions in the on-shore samples, though very similar in Ni and Ca con-tents, are distinctly lower in Mg, ranging in compositionfrom Fo89.7 to F090.5.

The pyroxenes in the drilled samples are also Mg-rich,plotting primarily in the diopside and enstatite fields(Fig. 2; Tables 5 and 6). The major variation in compo-sition is in the Ca contents, as reflected in the spread ofpoints parallel to the Wo-En join. The minerals in 567 A-29-2, 30-35 cm, particularly the orthopyroxene, areslightly more Mg-rich than those in the other samples

646

PETROGRAPHY AND RELICT MINERALOGY OF SERPENTINITES

Table 4. Selected analyses of olivine (oxides in wt.%, elements as atoms per 4 oxygens).

SiO 2

TiO 2

A12O3

FeOMnOMgOCaONiOC r 2 O 3

Total

SiTiAlFe + 2MnMgCaNiCr

FO

n ^

567A-29-2,30-35 cm

41.050.010.007.850.07

51.140.030.430.02

100.60

0.9930.0000.0000.1590.0011.8440.0000.0080.000

92.0

40.560.010.007.640.06

50.700.010.400.00

99.39

0.9920.0000.0000.1560.0011.8490.0000.0070.000

92.2

566C-6-1,31-33 cm

40.440.030.008.930.08

49.830.030.340.02

99.69

0.9910.0000.0000.1820.0021.8210.0000.0070.000

90.8

40.520.010.008.970.06

50.130.030.350.03

100.10

0.9900.0000.0000.1830.0011.8250.0000.0060.000

90.8

CR 205

40.540.000.009.580.04

49.510.030.330.03

100.06

0.9930.0000.0000.1960.0001.8080.0000.0060.000

90.2

+/-

40.250.030.009.160.04

49.030.010.340.03

98.89

0.9960.0000.0000.1890.0001.8080.0000.0060.000

90.5

fcT~w m a r

D-50

40.260.010.009.550.08

49.060.040.420.00

99.41

0.9930.0000.0000.1970.0011.8050.0000.0080.000

90.1

itit> mat

40.200.010.009.540.08

49.220.040.430.00

99.52

0.9910.0000.0000.1960.0011.8080.0010.0080.000

90.2

•p»ria1

DL

40.230.010.009.820.05

48.870.050.370.01

99.39

0.9930.0000.0000.2020.0001.7990.0010.0070.000

89.8

rhf> n v r

-90

40.140.000.009.960.05

49.090.040.370.01

99.69

0.9890.0000.0000.2050.0011.8040.0010.0070.000

89.7

nvpnf+7\ A 7\

CaSiO,

MgSiO, FeSiO,

En

0.05

* . ! + +

ΛB•0.95

)© CC

0.85

90 91 92

Fo

Figure 2. Olivine, orthopyroxene, and clinopyroxene compositions foronshore and drilled samples. Symbols as in Figure 3.

(Fig. 2). Orthopyroxenes have low TiO2 contents (lessthan 0.04%), moderate A12O3 (2.15-4.1%) and moder-ate Cr 2O 3 (0.57-1%) (Table 5). The clinopyroxenes like-wise generally have A12O3 in the range 1.8-2.9% andCr 2O 3 of 0.5-1.0%, as well as low contents of Na2O (Ta-ble 6). These mineral compositions are consistent with aninterpretation of the serpentinite progenitor as a refrac-

pyroxene fragments were toosmall to evaluate effects of core-rim variation.

The pyroxenes in the onshore samples, though alsoMg-rich, are distinct from the pyroxenes in the drilledsamples in their consistently higher Fe contents (Fig. 2;Tables 5 and 6). In onshore sample D-50, the clinopy-roxenes tend to be lower in Ca content and the orthopy-roxenes significantly higher in Ca content than those intheir drilled counterparts. Though their Cr contents aresimilar, the orthopyroxenes are consistently higher inA12O3 content (4-6%). The clinopyroxenes have muchhigher abundances of A12O3 (5.6-7.4%). Sample DL^90also has clinopyroxenes with very high A12O3 (7%), highTiO2 (0.35-0.48%), and high Na2O (0.8%) contents sug-gesting that these are much less refractory materials thanthose in the drilled samples. These compositions are dis-tinct from most pyroxene analyses from alpine or oce-anic-type peridotites, which have Na2O contents in therange of 0.12-0.25% and A12O3 contents of 1.5-5.5%(Sinton, 1979; Arai and Fujii, 1979; Clarke and Loubat,1977). However, they are not as high in Na as jadeiticclinopyroxenes in eclogite xenoliths (Shee and Gurney,1979), and are similar in some ways to pyroxene compo-sitions from Cr-diopsied megacrysts in alkalic basalts(Jagoutz et al., 1979; Nixon and Boyd, 1979).

The spinels show the most distinct variations in chem-ical compositions of any of the relict minerals, as mightbe expected, since their complex chemistry makes themsensitive indicators of petrogenetic processes and condi-tions. The spinels in the drilled samples fall in the rangeofMg/Mg + F e + 2 = 0.55-0.67 and Cr/Cr + Al = 0.4-0.58, again with the exception of 567A-29-2, 30-35 cm,in which the spinels have significantly lower Cr and Mgcontents (Figs. 3 and 4; Table 7). All are low in Ti andFe + 3, and there is a tendency to slightly higher F e + 3

with higher Cr contents (Fig. 4). Compositions withinsamples are uniform, and zoning is slight, tending to

647

S. H. BLOOMER

Table 5. Selected analyses of orthopyroxene (oxides in wt.%, elements as atoms per 6 oxygens).

SiO2

TiO2

A12O3

FeOMnOMgOCaONa2OK2OP2O5NiOCr 2 O 3

Total

SiTiAl IVAl VIFe + 2MnMgCaNaKPNiCr

WoEnFs

566C-6-1,31-33 cm

55.420.032.156.010.08

34.121.15———

0.060.57

99.59

1.9260.0010.0730.0130.1740.0021.7680.043_——

0.0020.016

2.189.08.7

55.440.022.225.870.07

33.851.58———

0.140.62

99.81

1.9240.0010.0750.0140.1700.0021.7510.059—__

0.0040.017

2.988.4

8.5

567A-29-2, 30-35 cm

54.060.024.105.090.12

33.560.770.020.020.000.151.05

98.97

1.8860.0010.1140.0540.1480.0041.7450.0280.0020.0010.0000.0040.028

1.590.87.7

54.800.043.754.940.06

33.501.97———0.111.00

100.15

1.8920.0000.1070.0440.1420.0011.7230.072—_—0.0020.026

3.788.97.3

54.830.043.985.100.05

34.420.40———

0.120.90

99.85

1.8910.0010.1080.0530.1470.0011.7690.015———

0.0030.024

0.791.67.6

570-42-1,111-113 cm

55.020.032.405.770.09

34.390.56———

0.060.59

98.90

1.9200.0010.0790.0190.1680.0031.7890.020___

0.0020.016

1.090.4

8.4

54.920.041.865.810.05

34.520.63———

0.120.43

98.38

1.9290.0010.0700.0050.1710.0011.8070.023———

0.0030.012

1.190.3

8.5

CR 205

Small

54.130.104.686.320.06

33.480.29———

0.130.49

99.68

1.8780.0030.1210.0690.1830.0021.7310.011___

0.0040.013

0.00589.99.5

Core

52.870.115.665.960.08

32.191.19———

0.090.66

98.81

1.8540.0030.1450.0880.1740.0021.6820.044—_—

0.0020.018

2.388.59.1

D50

Core

53.380.054.635.880.10

31.293.10———

0.091.01

99.53

1.8700.0010.1290.0610.1720.0031.6340.115———

0.0020.028

5.985.08.9

Rim

53.490.044.756.510.08

31.741.66———

0.090.99

99.35

1.8730.0010.1260.0690.1900.0021.6570.062_—_

0.0030.027

3.286.79.9

DL90

Core

53.210.115.506.190.08

31.761.35———

0.130.69

94.01

1.8630.0030.1360.0900.1810.0021.6580.050———

0.0040.019

2.687.79.5

Rim

53.380.175.266.230.09

31.941.34———

0.070.58

99.07

1.8680.0040.1310.0850.1820.0031.6670.049———

0.0020.016

2.587.89.5

Note: Small denotes analyses of grains on same order of size as spot (generally less than 30 µm); core and rim denote, respectively, analyses ofthe center and edge of larger grains; notation in subsequent tables is the same. — indicates not determined.

Table 6. Selected analyses of clinopyroxene (oxides in wt.%, elements as atoms per 6 oxygens).

SiO2

TiO2

A12O3

FeOMnOMgOCaONa2OK2OP2O5NiOCr 2 O 3

Total

SiTiAl IVAl VIFe + 2MnMgCaNaKPNiCr

WoEnFs

566-9-1,9-11 cm

53.190.121.831.690.10

18.1022.990.140.000.000.070.89

99.11

1.9440.0030.0560.0230.0520.0030.9860.9000.0100.0000.0000.0020.025

46.450.92.7

52.300.092.022.190.07

18.2423.98

__—

0.090.08

99.85

1.9110.0020.0870.0000.0670.0020.9920.938——_

0.0030.025

46.949.6

3.3

566C-6-131-33 cm

52.720.032.552.210.04

17.7023.280.040.020.030.080.84

99.54

1.9240.0010.0750.0340.0670.0010.9620.9100.0030.0010.0010.0020.024

46.949.6

3.4

52.710.072.912.190.09

17.8023.080.020.010.000.041.14

100.05

1.9140.0020.0850.0380.0660.0030.9630.8980.0010.0000.0000.0010.033

46.649.9

3.4

570-42-1,111-113 cm

52.780.052.511.950.11

17.7623.390.030.000.010.070.90

99.55

1.9250.0010.0740.0330.0590.0030.9650.9140.0020.0000.0000.0020.026

47.149.7

3.0

52.480.041.912.000.09

17.8823.090.040.020.020.090.55

98.21

1.9380.0010.0610.0210.0610.0030.9840.9140.0030.0010.0010.0030.016

46.650.2

3.1

D 50

Rim

51.410.135.733.640.10

20.1118.840.040.010.030.151.24

101.42

1.8350.0020.1640.0760.1080.0021.0700.7200.0020.0000.0000.0030.034

37.956.35.6

Core

51.600.105.623.790.12

19.8418.270.080.020.010.081.27

100.80

1.8490.0020.1500.0870.1130.0031.0600.7010.0050.0000.0000.0010.035

37.456.56.0

DL90

Core

51.610.357.083.110.15

17.0619.220.880.000.000.071.09

100.62

1.8520.0080.1470.1510.0930.0040.9120.7390.0610.0000.0000.0010.030

42.352.2

5.3

Rim

50.270.487.412.890.15

15.6220.770.830.010.000.031.08

99.54

1.8320.0130.1670.1510.0870.0050.8490.8110.0580.0000.0000.0010.031

46.448.5

4.9

Note: See note to Table 5.

648

PETROGRAPHY AND RELICT MINERALOGY OF SERPENTINITES

1.0

0.8

0.6

0.4

0.2

Δ 566-6-1, 31—33 cm

Δ 566-9-1, 9-11 cm

H 566C-6-1, 31-33 cm

B 566C-7-1, 13-14 cm

D 566C-7-1, 78-80 cm

+ 567A-28-1, 121-122 cm

• 567A-29-2, 30-35 cm

- 567A-29-2, 140-142 cm

* 570-42-1,40-41 , vcm

• 570-42-1, 1 11-1 13 cm

A 570-42-2, 53-55 cm

+ D50

O 90

O CR 39

© CR 205

1.0 0.5 0.1

Mg/Mg + Fe+2

Figure 3. Cr/Cr + Al vs. Mg/Mg + Fe + 2 for relict spinels in onshoreand drilled samples. Dotted field is that of spinels from abyssalperidotites, after Dick and Bullen (in press).

+ 3Al Fe

Figure 4. Trivalent constituents of spinels from onshore and drilledsamples. Symbols as in Figure 3; for clarity, only selected analysesare plotted.

rims less Cr-rich than cores. In some cases alteration ofthe spinels has produced black rims of magnetite andcompositions intermediate between the Cr-spinel, fer-richromite, and magnetite (Fig. 4).

The three least-serpentinized onshore samples have Al-rich spinels, as suggested by their greenish black color inthin section. Those of D-50 are closest to 567A-29-2,30-35 cm in composition; these have low Ti contents as

well. The spinels in CR 39 are the most Cr-rich of anysampled; they occur, however, in a totally serpentinizedand slightly deformed breccia, and it is not clear thatthese are indeed primary compositions. Alteration ofspinels will tend to drive them to higher Cr and Fe con-tents, generally like the ferrichromite compositions in570-42-2, 53-55 cm (Fig. 4). The extensive serpentiniza-tion and deformation in this sample make its interpreta-tion as a primary composition suspect.

Although diverse in composition, most of the spinelsanalyzed have Cr/Cr + Al ratios less than 0.60. Thecompositions of spinels in the Leg 84 serpentinites, thoughnot definitely of oceanic character, fall within the rangeof compositions found for spinels in abyssal peridotites(Aumento and Loubat, 1971; Sigurdsson, 1977; Arai andFujii, 1979; Dick and Bullen, in press). The onshore sam-ples, though distinct from the drilled samples, fall in therange of abyssal peridotite compositions as well. The pe-ridotites at Santa Elena are associated with basalts thathave been interpreted as ocean-ridge tholeiites (Weyl,1969; Pichler and Weyl, 1975), and with gabbros whichhave common cumulate olivine and poikilitic orthopy-roxene (Galli-Olivier, 1979; Azema and Tournon, 1982),both common characteristics of ocean-ridge gabbros(Hodges and Papike, 1976; Tiezzi and Scott, 1980).

Geothermometry using the analyzed mineral compo-sitions indicates that there has been subsolidus reequili-bration in these samples. The olivine-spinel geothermom-eter calibration of Roeder et al. (1979) gives temperaturesof 730-850° C for the drilled samples and 820-990°Cfor the onshore samples. Despite the differences in min-eral composition between Sample 567A-29-2, 30-35 cmand the other drilled samples, calculated temperaturesfor all are similar, suggesting that the differences reflectinitially distinct compositions rather than differing de-grees of reequilibration. Mori's (1977) calibration of thetwo-pyroxene geothermometer indicates temperaturesof 1110-1170°C for the drilled samples, much closer tothose expected for mantle materials, but still indicativeof some reequilibration. The onshore samples yield tem-peratures of 1220-1440° C; like the olivine-spinel calcu-lations, these temperatures are higher than those calcu-lated for the drilled samples.

DISCUSSION

The petrography of the serpentinites drilled on Leg84 indicates that they were clinopyroxene-bearing harz-burgites before serpentinization. Their mineral chemis-try is refractory: high-Mg olivines (Fo91_92) and orthopy-roxene (En8 9_9 1); low-Ti and -Fe + 3 spinels and low-Aland -Na pyroxenes. Such compositions are consideredcharacteristic of rocks interpreted to be upper-mantle ma-terials residual from an episode of partial melting (Men-zies, 1977; England and Davies, 1973; Loney et al.,1971). Sample 567A-29-2, 30-35 cm is distinct from theremainder of the samples in its more magnesian olivinesand orthopyroxene and more aluminous spinels. Beyondsuggesting a mantle affinity for these materials, it is dif-ficult to be more specific about their origin. Their min-eral compositions, particularly those of the spinels, aresimilar to those reported for peridotites from the ocean

649

O

Table 7. Selected analyses of spinels (oxides in wt.%, elements as atoms per 32 oxygens).

SiO2

TiO2

A12O3

Fe2O3

FeOMnOMgOCaONiOCr2O3

Total

SiTiAl VIF e + 3F e + 2MnMgCaNiCrCr/Cr + Al

566-9-1, 9-11 cm

Core

0.000.11

23.122.92

16.210.20

12.870.010.08

45.74

101.27

0.0000.0216.6270.5353.2960.0434.6650.0000.0168.7930.570

Rim

9.000.16

23.462.74

16.800.15

12.760.020.04

45.21

101.85

0.0000.0276.8190.4983.3920.0324.5910.0050.0058.6280.558

Small

0.010.15

25.753.08

17.170.23

12.640.040.08

42.63

101.69

0.0050.0277.2820.5573.4350.0434.5260.0100.0168.0940.526

567A-28-1,121-122 cm

Core

0.000.12

31.161.39

17.670.14

12.650.010.08

37.15

100.48

0.0000.0228.7370.2483.5030.0274.4740.0000.0166.9690.443

Rim

0.000.04

32.121.56

17.330.13

12.970.030.10

36.67

100.96

0.0000.0058.8990.2753.4040.0274.5460.0050.0226.8130.433

567A-29-2,30-35 cm

Small

0.000.05

40.180.50

12.870.20

16.S10.020.06

30.63

101.33

0.0000.010

10.5150.0842.3910.0385.5650.0050.0105.3780.338

Core

0.010.08

39.270.78

12.580.18

17.050.030.07

31.88

101.91

0.0000.010

10.2580.1302.3290.0325.6320.0050.0105.5890.352

566C-6-1, 31-

Core

0.000.06

32.251.85

15.390.21

14.480.020.15

37.88

102.29

0.0000.0108.7580.3202.9640.0434.9700.0050.0276.8990.440

Rim

0.000.09

30.482.28

15.130.17

14.470.010.12

39.17

101.93

O.(XX)0.0168.3600.3982.9420.0325.0190.0000.0217.2080.463

33 cm

Core

0.000.06

30.662.23

15.080.18

14.530.020.19

39.46

102.39

o.ooo0.0108.366

2.9210.0375.0130.0050.0327.2230.463

570-42-1, 111-113

Core

0.000.05

31.911.93

13.710.13

15.450.030.09

38.20

101.50

0.0000.0058.6800.3352.6470.0215.3140.0050.0166.9720.445

Rim

0.000.08

33.681.03

14.340.14

15.220.010.09

36.82

101.41

0.0000.0109.1150.1782.7510.0275.2090.0050.0166.6830.423

cm

Small

0.000.08

34.361.67

15.020.17

14.870.010.10

35.44

101.72

0.0000.0109.2730.28T2.8750.0325.0750.0050.0216.4170.408

570-42-2, 53-55

Core

0.000.08

30.192.06

15.210.12

14.550.020.07

40.33

102.61

0.0000.016S.2360.3592.9440.0215.0230.0050.016-.3820.472

Core

0.000.020.00

67.7231.56

0.030.090.030.132.99

102.57

0.0000.0050.000

15.2777.9150.0050.0420.0100.0310 . 7 l l1 .(XX)

cm

Core

0.000.05

32.601.31

14.870.17

14.740.060.13

37.80

101.74

0.0000.0108.8590.2272.8670.0325.0660.0160.0276.8920.437

CR 39

Core

0.170.13

19.614.31

19.100.16

10.650.030.07

46.45

100.66

0.0440.0275.8140.8154.0180.0333.9920.0050.0169.2360.613

Rim

0.000.14

19.624.88

19.070.20

10.400.000 10

45.75

100.16

0.0000.0275.8530.9304.0400.0443.9260.0000.0169.1590.610

CR 205

Core

0.000.05

54.110.61

12.080.08

18.890.000.26

15.96

102.04

0.0000.016

13.2590.0962.1020.0105.8540.0000.0432.6220.165

Core

0.000.06

56.090.70

11.570.07

19.340.000.27

13.52

101.61

0.0000.010

13.6610.1091.9980.0105.9570.0000.0432.2060.139

D 50

("ore

0.000.09

43.891.10

13.240.13

16.820.010.18

25.24

100.71

0.0000.106

11.3900.1822.4370.0225.5180.0050.0334.3930.278

Core

0.000.03

50.120.76

12.380.08

17.960.080.36

19.44

101.21

0.0000.006

12.5910.1222.2040.0165.7060.0160.0603.2740.206

DL90

Core

0.000.09

54.290.83

11.840.10

18.860.010.31

14.34

100.95

0.0000.016

13.4640.1302.0690.0165.8810.0000.0492.3710.149

Core

0.030.08

53.461.38

11.010.07

18.940.020.31

14.20

99.52

0.0050.011

13.3640.2211.9540.0115.9880.0050.0562.3790.151

Note: See note to Table 5.

PETROGRAPHY AND RELICT MINERALOGY OF SERPENTINITES

basins. There is no evidence (i.e., high-Cr spinels or adominantly clinopyroxene-free harzburgite assemblage)that might indicate a very refractory, or island-arc typeassociation (Dick and Bullen, in press).

With the exception of CR 39, the onshore set of sam-ples from Guatemala and Costa Rica are a distinctly dif-ferent assemblage from drilled samples. They are alsoakin to materials of upper-mantle origin, but have morealuminous spinels, more Fe-rich olivines and pyroxenes,and more sodic and aluminous clinopyroxenes than theLeg 84 samples. Though CR 205 and DL-90 are cpx-bearing harzburgites, and are texturally similar to theLeg 84 samples, they are chemically less depleted rocks.D-50 has an apparently cumulate texture with large or-thopyroxene and interstitial olivine, but has very magne-sian mineral compositions and low-Ti, low-Fe+3 spinelswhich may indicate a mantle origin. This sample could bea fragment of a pyroxenite dike or pod of upper-mantleorigin. The Al and Na contents of the clinopyroxenes inall the onshore samples are higher than is characteristicof pyroxenes described from oceanic fracture zones orophiolites thought to be fragments of ocean crust (Araiand Fujii, 1979; Pallister and Hopson, 1981; Aumentoand Loubat, 1971; Dick and Bullen, in press). The sam-ple is not large, however, and the onshore samples aresimilar enough in composition to the oceanic samplesthat they might be less-depleted oceanic mantle mate-rials.

Although it is clear that the Leg 84 samples are dis-tinct from the onshore samples, nothing in the data pro-hibits a general relationship between the two. Both on-shore suites could be materials derived from the oceanicmantle. Peridotite sections in ophiolites can vary greatlyon both large and small scales (Boudier and Coleman,1981; Evans, 1983). The Leg 84 sample set is fairly uni-form over the area sampled; the onshore samples arelimited in number, come from a substantial distance fromthe drill sites, and may be biased toward more atypicalvarieties because they were generally the least serpen-tinized samples available. What is intriguing is that thesample from Costa Rica (CR 205) and the samples fromGuatemala (DL-90 and D-50) have quite similar mineralcompositions, suggesting that both were formed fromsimilar sources and by similar processes, and that thosechemical compositions may be neither unusual nor un-representative. The compositional differences betweenthe two samples sets (onshore and offshore) may repre-sent two distinct and possibly unrelated types of mantlematerial.

If the drilled samples are indeed mantle materials andof ocean-crustal origin, they must have come from depthsof 9-12 km, judging by the occurrence of similar harz-burgitic rocks in ophiolites and the depth to the Mohoin seismic sections of ocean crust (Clague and Straley,1977). This raises the question of why these serpentin-ized peridotites are now exposed at relatively shallowdepths throughout the trench slope. There are at leasttwo likely ways such exposures might occur: (1) imbri-cate faulting of slices of ocean crust, or (2) diapirism ofserpentinized ultramafics along zones of weakness in thetrench slope.

Seismic profiles of the forearc have been interpretedas showing two or more large imbricate slices of oceancrust in the forearc (Ladd et al., 1978). Such imbricatestacking might well expose some upper mantle materialsat shallow depths. There are, however, some aspects ofthe drilled assemblages which contradict a model basedonly on imbricate stacking. As pictured by Ladd et al.(1978) or Azema and Tournon (1982, p. 740), the mostcommon materials sampled ought to be basalts, diabases,and gabbros, with lesser amounts of peridotitic rocks.In contrast, coring recovered serpentinite as the sole base-ment rock at Holes 566, 566A, 566C, and 570; serpentin-ite in blocks and between units of basalt and gabbro inHole 567A; and gabbro and basalt alone in Hole 569A(site chapters, this volume). Furthermore, the commonassociation reported in dredge and piston core samplesis serpentinite, basalt, and chert (Ibrahim et al., 1979);gabbro is notably lacking from these sets of samples.The coarser-grained samples from Leg 84 are predomi-nantly altered, massive gabbro or diabase, rather thancumulate-textured gabbros. The existing sample set seemsstrongly biased toward serpentinite, with lesser amountsof basalt and diabase, and minor volumes of gabbro.This bimodal distribution would be surprising in a sim-ple set of stacked crustal sheets.

The other fact that should be noted is that the "ser-pentinized peridotites" are really serpentinites—greaterthan 90% serpentine and secondary minerals. Serpen-tine is a relatively mobile material; it is found at veryshallow depths in fracture zones (e.g., Bonatti, 1976),where its emplacement is usually inferred to be diapiric;it also occurs as diapiric bodies in many deformed ter-ranes (e.g., the Franciscan Formation of California; Oake-shott, 1968). In northern Central America there are bod-ies of serpentinite along many of the high- and low-anglefaults in the region (King, 1968; Anderson et al., 1973;Weyl, 1980). Two requirements for such diapiric emplace-ment of serpentine may include (1) a source of water tocirculate into mantle to alter the peridotite, and (2) azone of weakness in the crust, such as a fault, alongwhich the serpentine can be emplaced, driven presum-ably by its low density and aided by its relative mobility.The forearc along the Leg 67 and Leg 84 transects wouldappear to satisfy both conditions. There should be faultsdeveloped as a result of the presumed initial imbricationof crustal materials in the forearc. There is also a com-plex network of high-angle faults throughout northernCentral America, which probably extends into the fore-arc. If this is the case, these faults provide both zones ofweakness in the crust and pathways along which watercan circulate into the lower crust and mantle to producethe serpentinites. Dehydration of sediments and alteredigneous rocks as they are subducted beneath the forearcmay also contribute water to the system.

Serpentine diapirism in the forearc may explain thedistribution of basement rocks at the drill sites. Wherediapirs pierce the upper crustal layers, the basement wouldconsist solely of serpentine; where they had simply pushedinto and fractured the upper crust, a complex mixture ofbasalts, diabases, and serpentine might be found. Suchdiapirism could explain the bimodal distribution of rock

651

S. H. BLOOMER

samples in the trench slope. There is some morphologicand seismic evidence to support the suggestion that ser-pentine diapirism is, or was, important in the forearc.Seely (1979) noted shallow magnetic bodies and cross-cutting structures in seismic reflection lines across theGuatemalan margin, which he interpreted as serpentinediapirs of shallow slices of ocean crust. An indicationthat sediments were dragged upward at the edges of thestructures favored the former (Seely, 1979). There is agreat deal of relief in the basement topography (site chap-ters), some of which could be due to diapirism. Serpen-tine diapirs are commonly "elliptical and slightly elon-gate in plan" (Oakeshott, 1968), and there are a numberof conical or slightly elongate, elliptical features on thetrench slope. The Seabeam map of Aubouin et al. (1982,p. 736) shows several conical and elliptical bathymetrichighs in the forearc which are often elongate subparallelto the trench axis (for example, near 12°43' N, 90°47' Wand 12°50'N, 90°54' W). Near Site 570 there is a coni-cal high closely corresponding to a magnetic high (Volpe,this volume). This feature may well be a diapiric struc-ture cored by serpentine. The magnetic characteristicsof serpentine are highly variable, and are probably notreliable in mapping such bodies—the highly variable de-velopment of secondary oxides in the samples describedhere is probably the primary cause of such variation.The fact that serpentine was recovered at such a highproportion of sites may in part reflect a tendency to drillbasement highs where the sediment cover is thinner; Site570, for example, clearly sits atop such a basement high(site chapter, this volume). Diapiric structures would beexpected to be areas of positive relief in the basement.

Diapirism of serpentinites in other trench slope andforearc regions has been described. The inner slope ofthe Mariana Trench between 16° and 18°N is character-ized by very large amounts of serpentinite and serpen-tinized peridotites (Bloomer and Hawkins, 1983). Thereare elongate bathymetric highs along the trench-slopebreak, sometimes referred to as forearc seamounts (Hus-song and Fryer, 1982). Dredge hauls along these yieldedprimarily serpentine and volcaniclastic siltstones. Thesehave been interpreted as diapirs piercing the volcanicbasement surrounding them (Bloomer, 1983). The geo-logic setting in the Marianas forearc is certainly differ-ent from that off Guatemala, in that the forearc base-ment in the Marianas is clearly of island-arc origin. Re-gardless of the difference in setting, if there is extensive,deep faulting in the forearc and ultramafic rocks at depth,serpentine diapirism is likely to become an importantprocess.

The suggestion of diapirism of serpentine in the fore-arc along the Leg 84 transect does not negate the possi-bility of imbricate faulting. Such diapirism might haveaccompanied the disruption and faulting of the crustalmaterials, or could have occurred after that faulting.

CONCLUSIONS

1. The serpentinites sampled on Leg 84 are derivedfrom clinopyroxene-bearing harzburgites with mineralcompositions characteristic of upper-mantle rocks de-pleted to some extent by an epsiode of partial melting

(olivine, Fo90.6-92; orthopyroxene En89_91; clinopyroxeneWo47En5oFs3; spinel, Cr/Cr + Al = 0.3-0.6). The sam-ples from all sites are relatively homogeneous, except for567A-29-2, 30-35 cm, which has more magnesian oliv-ines and orthopyroxenes and more aluminous spinelsthan the other drilled samples.

2. The least-serpentinized samples from ultramaficexposures in Costa Rica and Guatemala, two of whichwere clinopyroxene-bearing harzburgites, have mineralchemistries distinct from those of the drilled samples.Their more iron-rich olivines and pyroxenes, aluminousspinels, and sodic and aluminous clinopyroxenes indicatethat these are less-depleted materials than the Leg 84samples. CR 39, a serpentinite, has spinel compositionsmore Cr-rich than those of the Leg 84 samples, but it isnot clear that these represent unaltered relict composi-tions.

3. Both the onshore and the offshore sets of samplescould be derived from oceanic-type mantle—that be-neath an ocean-crustal section formed at a spreadingridge. It is unclear, however, whether the offshore andonshore ultramafics are parts of a related terrane.

4. Diapirism of serpentine may have been an impor-tant process in the development of the forearc. Thismight account for the abundance of serpentine in therecovered materials, the bimodal distribution of rocktypes in dredged and drilled samples (serpentine andbasalt), and some of the morphologic characteristics ofthe trench slope and forearc.

ACKNOWLEDGMENTS

M. Baltuck assisted greatly in obtaining the Leg 84 samples, and T.W. Donnelly and N. Lundburg kindly provided samples from Guate-mala and Costa Rica, respectively. A. Volpe and M. Baltuck providedseveral helpful discussions. R. Fujita assisted with the microprobe anal-yses. Reviews by T. W. Donnelly and C. Evans greatly improved themanuscript. This work was supported by a grant from the Scripps In-dustrial Associates.

REFERENCES

Anderson, T. H., Burkart, B., demons, R. E., Bohnenger, O. H.,and Blount, D. N., 1973. Geology of the western Altos Cuchuma-tanes, northwestern Guatemala. Geol. Soc. Am. Bull., 84:805-826.

Arai, S., and Fujii, T., 1979. Petrology of ultrabasic rocks from Site395. In Melson, W. G., Rabinowitz, P. D., et al., Init. Repts.DSDP, 45: Washington (U.S. Govt. Printing Office), 587-594.

Aubouin, J., Stephan, J., Renard, V., Roump, J. and Lonsdale, P.,1982. A Seabeam survey of the Leg 67 area (Middle America Trenchoff Guatemala). In Aubouin, J., von Huene, R., et al., Init. Repts.DSDP, 67: Washington (U.S. Govt. Printing Office), 733-745.

Aumento, F., and Loubat, H., 1971. The Mid-Atlantic Ridge near45°N XVI. Serpentinized ultramafic intrusions. Can. J. Earth. Sci.,8:631-663.

Azema, J., and Tournon, J., 1982. The Guatemala margin, the NicoyaComplex, and the origin of the Caribbean Plate. In Aubouin, J.,von Huene, R., et al., Init. Repts. DSDP, 67: Washington (U.S.Govt. Printing Office), 739-745.

Bloomer, S. H., 1983. Distribution and origin of igneous rocks fromthe landward slopes of the Mariana Trench: implications for itsstructure and evolution. J. Geophys. Res., 88:7411-7428.

Bloomer, S. H., and Hawkins, J. W., 1983. Gabbroic and ultramaficrocks from the Mariana Trench: an island arc ophiolite. In Hayes,D. E. (Ed.), The Tectonic and Geologic Evolution of SoutheastAsian Seas and Islands, II. Am. Geophys. Union (Washington),Geophys. Mono., 27:294-317.

Bonatti, E., 1976. Serpentinite protrusions in the oceanic crust. EarthPlanet. Sci. Lett., 32:107-113.

652

PETROGRAPHY AND RELICT MINERALOGY OF SERPENTINITES

Boudier, E, and Coleman, R. G., 1981. Cross-section through the peri-dotite in the Semail ophilite, southeastern Oman mountains. J. Geo-phys. Res., 86:2573-2592.

Clague, D., and Straley, P., 1977. Petrologic nature of the oceanicMoho. Geology, 5:133-136.

Clarke, D. B., and Loubat, H., 1977. Mineral analyses from the peri-dotite-basalt-gabbro complex at Site 334, DSDP Leg 37. In Aumen-to, E , Melson, W. G., et al., Init. Repts. DSDP, 37: Washington(U.S. Govt. Printing Office), 847-857.

Couch, R., 1976. Estructuras del margén continental de Centra Amer-ica y el desarollo de una hipótesis. In Perez-Rodrequez, R., andSuarez-Zozaya, M. R. (Eds.), First Reunion Latino-Americano so-bre Ciencia Tecnologia de los Oceanos: Mexico (Secretaria de Ma-rina Mexicana), pp. 120-139.

Couch, R., and Woodcock, S., 1981. Gravity and structure of the con-tinental margins of southwestern Mexico and northwestern Guate-mala. J. Geophys. Res., 86:1829-1840.

Coulbourn, W. T., Hesse, R., Azema, J., and Shiki, T., 1982. A sum-mary of the sedimentology of Deep Sea Drilling Project Leg 67sites: the Middle America Trench and slope off Guatemala—anactive margin transect. In Aubouin, J., von Huene, R., et al.,Init. Repts. DSDP, 67: Washington (U.S. Govt. Printing Office),759-774.

Dick, H. J. B., and Bullen, T., in press. Chromium spinel as a petroge-netic indicator in oceanic environments. Contrib. Mineral. Petrol.

England, R. L., and Davies, H. L., 1973. Mineralogy of ultramaficcumulates and tectonites from eastern Papua. Earth Planet. Sci.Lett.,11-.416-425.

Evans, C. A., 1983. Petrology and geochemistry of the transition frommantle to crust beneath an island arc-backarc pair: implicationsfrom the Zambales Range ophiolite, Luzon, Philippines [unpubl.Ph.D. dissertation]. Univ. Calif., San Diego.

Galli-Olivier, C , 1979. Ophiolite and island-arc volcanism in CostaRica. Geol. Soc. Am. Bull., 90:444-452.

Hodges, F. N., Papike, J. J., 1976. DSDP Site 334: magmatic cumu-lates from oceanic Layer 3. J. Geophys. Res., 81:4135-4151.

Hussong, D. M., and Fryer, P., 1982. Structure and tectonics of theMariana arc and forearc. Drillsite selection surveys. In Hussong,D. M., Uyeda, S., et al., Init. Repts. DSDP, 60: Washington (U.S.Govt. Printing Office), 33-44.

Ibrahim, A. K., Latham, G. V., and Ladd, J., 1979. Seismic refractionand reflection measurements in the Middle America Trench off-shore Guatemala. /. Geophys. Res., 84:5643-5649.

Jagoutz, E., Lorenz, V., and Wanke, H., 1979. Major-trace elementsof Al-augites and Cr-diopsides from ultramafic nodules in Europe-an alkali basalts. In Boyd, F. R., and Meyer, H. O. A. (Eds.), Themantle Sample: Inclusions in Kimberlites and Other Volcanics.Proc. Second Int. Kimberlite Conf. (Vol. 2): Washington, (Am.Geophys. Union), 382-390.

King, P. B., 1968 (compiler). Tectonic map of North America, U.S.Geol. Survey Map.

Kuijpers, E., 1980. The geological history of the Nicoya ophiolite com-plex, Costa Rica, and its geotectonic significance. Tectonophysics,68:233-255.

Ladd, J. W., Ibrahim, A. K., McMillen, K. J., Latham, G. V., vonHuene, R. E., Watkins, J. E., Moore, J. C , and Worzel, J. L.,1978. Tectonics of the Middle America Trench offshore Guatema-la. Int. Symp. Guatemala 4 February Earthquake and Reconstruc-tion Process (Guatemala City, May 1978), Vol. 1.

Lawrence, D. P., 1975. Petrology and structural geology of the Sanarte-El Progreso area, Guatemala [unpubl. Ph.D. dissertation]. SUNY,Binghamton.

Loney, R. A., Himmelburg, G. R., and Coleman, R. G., 1971. Struc-ture and petrology of the alpine-type peridotite at Burro Moun-tain, California, U.S.A. J. Petrol., 12:245-309.

Maury, R. C , Bougault, H., Joron, J. L., Girard, D., Treuil, M.,Azema, J., and Aubouin, J., 1982. Volcanic rocks from the Leg 67sites: mineralogy and geochemistry. In Aubouin, J., von Huene,R., et al., Init. Repts. DSDP, 67: Washington (U.S. Govt. PrintingOffice), 557-576.

Menzies, M., 1977. Residual alpine lherzolites and harzburgites—geo-chemical and isotopic constraints on their origin. In Dick, H. J. B.(Ed.), Chapman Conference on Partial Melting in the Earth's Up-per Mantle. Oregon Dept. Geol. Min. Indus. Bull., 96:129-147.

Mori, T , 1977. Geothermometry of spinel lherzolites. Contrib. Min-eral. Petrol., 59:261-279.

Nixon, P. H., and Boyd, F. R., 1979. Garnet bearing lherzolites anddiscrete nodule suites from the Malaita alnoite, Solomon Islands,southwest Pacific and their bearing on oceanic mantle composi-tion and geotherm. In Boyd, F. R., and Meyer, H. O. A. (Eds.),The Mantle Sample: Inclusions in Kimberlites and Other Volcanics.Proc. Second Int. Kimberlite Conf. (Vol. 2): Washington (Am.Geophys. Union), 400-423.

Oakeshott, G. B., 1968. Diapiric structures in the Diablo Range, Cali-fornia. In Braunstein, J., and O'Brien, G. D. (Eds.), Diapirs andDiapirism. Mem. Am. Assoc. Pet. Geol., 8:228-243.

Pallister, J., and Hopson, C. A., 1981. Samail ophiolite plutonic suite:field relations, phase variations, cryptic variation and layering, and amodel of a spreading ridge magma chamber. J. Geophys. Res., 86:2593-2644.

Pichler, H., and Weyl, R., 1975. Magmatism and crustal evolution inCosta Rica. Geol. Rundschau, 64:457-475.

Roeder, P., Campbell, I., and Jamieson, H., 1979. A re-evaluation ofthe olivine-spinel geothermometer. Contrib. Mineral. Petrol., 68:325-334.

Rosenfeld, J. H., 1981. Geology of the western Sierra de Santa Cruz,Guatemala, Central America: an ophiolite sequence [unpubl. Ph.D.dissertation]. SUNY, Binghamton.

Seely, D. R., 1979. Geophysical investigations of continental slopesand rises. In Watkins, J. S., and Montadert, L. (Eds.), Geologicaland Geophysical Investigations of Continental Margins. Mem. Am.Assoc. Pet. Geol., 29:245-260.

Seely, D. R., Vail, P. R., and Walton, G. C , 1974. Trench slope model.In Burk, C. A., and Drake, C. L. (Eds.), The Geology of Conti-nental Margins: New York (Springer-Varlag), pp. 249-260.

Shee, S. R., and Gurney, J. J., 1979. The mineralogy of xenoliths fromOrapa, Botswana, In Boyd, F. R., and Meyer, H. O. A. (Eds.), TheMantle Sample: Inclusions in Kimberlites and Other Volcanics.Proc. Second Int. Kimberlite Conf., (Vol. 2): Washington, (Am.Geophy. Union), 37-49.

Sigurdsson, H., 1977. Spinels in Leg 37 basalts and peridotites: phasechemistry and zoning. In Aumento, F., Melson, W. G., et al., In-it. Repts. DSDP, 37: Washington (U.S. Govt. Printing Office),883-892.

Sinton, J. M., 1979. Petrology of (alpine-type) peridotites from Site395, DSDP Leg 45. In Melson, W. G., Rabinowitz, P. D., et al.,Init. Repts. DSDP, 45: Washington (U.S. Govt. Printing Office),595-602.

Tiezzi, L. J., and Scott, R. B., 1980. Crystal fractionation in cumulategabbro, Mid-Atlantic Ridge, 26°N. J. Geophys. Res., 85:5438-5454.

von Huene, R., Aubouin, J., Azema, J., Blackinton, G., Carter, J. H.,et al., 1980. DSDP Mid-America Trench transect off Guatemala.Geol. Soc. Am. Bull., 91, Pt. 1:421-432.

Weyl, R., 1969. Magmatische forderphasen und gesteinchemismus inCosta Rica. Neues. Jahrb. Geol. Palaeontol. Monatsh., 7:423-446.

, 1980. Geology of Central America: Berlin (Gebruder Born-traeger).

Woodcock, S. F , 1975. Crustal structure of the Tehuantepec Ridgeand adjacent continental margin of southwestern Mexico and west-ern Guatemala [unpubl. Master's thesis]. Oregon State University.

653