Geochemistry of DSDP Holes 60-453 and 60-454A

Pliocene and Pleistocene volcanic glass fragments from Mariana Trough sediments at Sites 453 (16 samples) and 454 (4 samples), located near the western edge of the trough and just west of the spreading axis, respectively, have been analyzed for major elements with an electron microprobe. They derive from volcanic activity on the present Mariana active arc. The glasses from Site 453 are all tholeiitic with a wide range of SiO2 contents. Those less than 2 m.y. old have slightly lower TiO2 and higher K2O contents than the older ones. The glasses from Site 454 are all Pleistocene and resemble the younger glasses at Site 453. Major element compositions of the older basaltic glasses at Site 453 are similar to those of the Mariana Trough basalts drilled on Leg 60. Both older and younger suites of glasses differ from the composition of rocks exposed on the active arc, which are assumed to be younger than any of the samples studied (i.e., about 200,000 y.). A third suite is represented by the arc rocks exposed on the volcanic islands. These have a smaller range of SiO2 contents and contain more A12O3 but less K2O, TiO2, and FeO1 (total Fe as FeO) than the sediment glasses studied. Further, a plot of FeO1 against MgO for the arc rocks does not follow the island arc tholeiite trend of the trough sediment glasses. Using the major element compositions of the arc rocks and sediment glasses, we can recognize three phases of volcanic activity, as indicated. The first evidence of the oldest phase of activity occurs 5 Ma, about 4.5 m.y. after the trough started to form. The second commenced about 2 Ma, and the last, including present-day activity, began within the last 200,000 y. Initially the rocks had major element affinities with the tholeiitic Mariana Trough seafloor, but this influence declined as the trough widened.

Pleistocene and late Miocene Calcareous nannofossil biostratigraphy of ODP Hole 150-905A

Continental rise Site 905 yielded upper Miocene and Pliocene uniform hemipelagic mud (a contourite) from approximately 215 to 540 meters below seafloor. The nannofossil biostratigraphy of this interval was reexamined using closely spaced samples from core interiors. Additionally, total nannofossil abundances and dominant species and species group abundances were determined to evaluate the potential of this section for extracting sequence stratigraphic information. The data indicate that the putative hiatuses at the end of the late Pliocene (Zones NN17 and NN18) and in the early Pliocene (Zones NN13 and NN14) probably are condensed intervals, but the base of the late Miocene is almost certainly marked by an unconformity. Judging from carbonate content and sedimentation rate both, nannofossil abundance may be governed by carbonate dissolution or by siliciclastic dilution. Consequently, condensed sections cannot be identified by the abundance of pelagic component in the sediment alone, as is possible in equivalent age Gulf of Mexico sediments. Where nannofossil preservation is adequate in consecutive samples, as in the early Pliocene and latest late Miocene, total nannofossil abundance fluctuates regularly and with a periodicity of less than 105 yr, which suggests that dilution of the pelagic component occurred with a frequency probably related to astronomical forcing.

Sulfur concentrations and isotopic composition of serpentinized oceanic peridotites

The mineralogy, contents, and isotopic compositions of sulfur in oceanic serpentinites reflect variations in temperatures and fluid fluxes. Serpentinization of <1 Ma peridotites at Hess Deep occurred at high temperatures (200°-400°C) and low water/rock ratios. Oxidation of ferrous iron to magnetite maintained low fO2 and produced a reduced, low-sulfur assemblage including NiFe alloy. Small amounts of sulfate reduction by thermophilic microbes occurred as the system cooled, producing low-delta34S sulfide (1.5? to -23.7?). In contrast, serpentinization of Iberian Margin peridotites occurred at low temperatures(~20°-200°C) and high water/rock ratios. Complete serpentinization and consumption of ferrous iron allowed evolution to higher fO2. Microbial reduction of seawater sulfate resulted in addition of low-delta34S sulfide (~15 to ~43?) and formation of higher-sulfur assemblages that include valleriite and pyrite. The high SO4/total S ratio of Hess Deep serpentinites (0.89) results in an increase of total sulfur and high delta34S of total sulfur (mean ~8?). In contrast, Iberian Margin serpentinites gained large amounts of 34S-poor sulfide (mean total S = 3800 ppm), and the high sulfide/total S ratio (0.61) results in a net decrease in delta34S of total sulfur (mean ~ -5?). Thus serpentinization is a net sink for seawater sulfur, but the amount fixed and its isotopic composition vary significantly. Serpentinization may result in uptake of 0.4-14 * 10**12 g S/yr from the oceans, comparable to isotopic exchange in mafic rocks of seafloor hydrothermal systems and approaching global fluxes of riverine sulfate input and sedimentary sulfide output.

Geochemistry, isotopes, trace elements and fluid inclusion geothermometry at DSDP Hole 83-504B

Stockwork-like metal sulfide mineralizations were found at 910-928 m below seafloor (BSF) in the pillow/dike transition zone of Hole 504B. This is the same interval where most physical properties of the 5.9-m.y.-old crust of the Costa Rica Rift change from those characteristic of Layer 2B to those of Layer 2C. The pillow lavas, breccias, and veins of the stockwork-like zone were studied by transmitted and reflected light microscopy, X-ray diffraction, and electron microprobe analysis. Bulk rock oxygen isotopic analyses as well as isolated mineral oxygen and sulfur isotopic analyses and fluid inclusion measurements were carried out. A complex alteration history was reconstructed that includes three generations of fissures, each followed by precipitation of characteristic hydrothermal mineral parageneses: (1) Minor and local deposition of quartz occurred on fissure walls; adjacent wall rocks were silicified, followed by formation of chlorite and minor pyrite I in the veins, whereas albite, sphene, chlorite and chlorite-expandable clay mixtures, actinolite, and pyrite replaced igneous phases in the host rocks. The hydrothermal fluids responsible for this first stage were probably partially reacted seawater, and their temperatures were at least 200-250° C. (2) Fissures filled during the first stage were reopened and new cracks formed. They were filled with quartz, minor chlorite and chlorite-expandable clay mixtures, traces of epidote, common pyrite, sphalerite, chalcopyrite, and minor galena. During the second stage, hydrothermal fluids were relatively evolved metal- and Si-rich solutions whose temperatures ranged from 230 to 340° C. The fluctuating chemical composition and temperature of the solutions produced a complex depositional sequence of sulfides in the veins: chalcopyrite I, ± Fe-rich sphalerite, chalcopyrite II ("disease"), Fe-poor sphalerite, chalcopyrite III, galena, and pyrite II. (3) During the last stage, zeolites and Mg-poor calcite filled up the remaining spaces and newly formed cracks and replaced the host rock plagioclase. Analcite and stilbite were first to form in veins, possibly at temperatures below 200°C; analcite and earlier quartz were replaced by laumontite at 250°C, whereas calcite formation temperature ranged from 135 to 220°C. The last stage hydrothermal fluids were depleted in Mg and enriched in Ca and 18O compared to seawater and contained a mantle carbon component. This complex alteration history paralleling a complex mineral paragenesis can be interpreted as the result of a relatively long-term evolution of a hydrothermal system with superimposed shorter term fluctuations in solution temperature and composition. Hydrothermal activity probably began close to the axis of the Costa Rica Rift with the overall cooling of the system and multiple fracturing stages due to movement of the crust away from the axis and/or cooling of a magmatic heat source.

Biogeochemical measurements of cold seep sediments inhabit by vesicomyid clams in the Japan Deep Sea Trench

Vesicomyidae clams harbor sulfide-oxidizing endosymbionts and are typical members of cold seep communities associated with tectonic faults where active venting of fluids and gases takes place. We investigated the central biogeochemical processes that supported a vesicomyid clam colony as part of a locally restricted seep community in the Japan Trench at 5346 m water depth, one of the deepest seep settings studied to date. An integrated approach of biogeochemical and molecular ecological techniques was used combining in situ and ex situ measurements. During the cruise YK06-05 in 2006 with the RV Yokosuka to the Japan Trench, we investigated a clam colony inhabited by Abyssogena phaseoliformis (former known as Calyptogena phaseoliformis) and Isorropodon fossajaponicum (former known as Calyptogena fossajaponica). The targeted sampling and precise positioning of the in situ instruments were achieved with the manned research submersible Shinkai 6500 (JAMSTEC, Nankoku, Kochi, Japan). Sampling was first performed close to the rim of the JTC colony and then at the center. Immediately after sample recovery onboard, the sediment core was sub-sampled for ex situ rate measurements or preserved for later analyses. In sediment of the clam colony, low sulfate reduction (SR) rates (max. 128 nmol ml**-1 d**-1) were coupled to the anaerobic oxidation of methane (AOM). They were observed over a depth range of 15 cm, caused by active transport of sulfate due to bioturbation of the vesicomyid clams. A distinct separation between the seep and the surrounding seafloor was shown by steep horizontal geochemical gradients and pronounced microbial community shifts. The sediment below the clam colony was dominated by anaerobic methanotrophic archaea (ANME-2c) and sulfate-reducing Desulfobulbaceae (SEEP-SRB-3, SEEP-SRB-4). Aerobic methanotrophic bacteria were not detected in the sediment and the oxidation of sulfide seemed to be carried out chemolithoautotrophically by Sulfurovum species. Thus, major redox processes were mediated by distinct subgroups of seep-related microorganisms that might have been selected by this specific abyssal seep environment. Fluid flow and microbial activity was low but sufficient to support the clam community over decades and to build up high biomasses. Hence, the clams and their microbial communities adapted successfully to a low-energy regime and may represent widespread chemosynthetic communities in the Japan Trench.

Consolidation and strength assessment of ODP Leg 123 sediments

Twenty-one samples, ranging in depth from 0 to 150 meters below seafloor (mbsf), were obtained from Leg 123 Sites 765 and 766. All samples were tested for Atterberg limits: 14 for laboratory vane shear strength and seven for uniaxial consolidation. Based on the determined Atterberg limits, along with shipboard measurements of water content, the sediment appears to be underconsolidated from 0 to 40 mbsf at Site 765 and from 0 to 80 mbsf at Site 766. Normal consolidation trends were observed for the sediments below these depths. Vane shear strengths, when compared with calculated values for a normally consolidated clay, indicate underconsolidated sediment at both sites. However, the use of Atterberg limit and vane shear strength data to assess consolidation state is complicated by the presence of silt-sized calcium carbonate in the form of nannofossil ooze. Thus, uniaxial-consolidation test data were analyzed to determine the overconsolidation ratios (OCR) and sediment compressibilities. OCR values were found to be less than one (underconsolidated) at both sites, using two separate methods of analysis.

Description, geochemistry and mineral composition of ODP Site 193-1189 jasperoids, PACMANUS field

Bright red "jasperoids" were recovered at three positions during Leg 193 drilling below Roman Ruins (Site 1189) in the PACMANUS hydrothermal field. These do not represent fossil exhalative oxide deposits equivalent to those associated with sulfide chimneys at the Roman Ruins seafloor. Rather, they constitute an integral, relatively early stage involving oxidized fluids in the development of veins and breccias that characterize the mostly sulfidic stockwork zone intersected below Roman Ruins in Hole 1189B. They formed by growth of quartz in open spaces created by hydrofracturing, the characteristic feature being mostly euhedral cores dusted by tiny hematite flakes. In one occurrence there are also frondlike aggregates and possible earlier cavity linings of hematite, overgrown by quartz, that potentially formed by maturation of ferruginous gels first deposited in the openings. The trace element geochemistry of the jasperoids, apart from minor enrichment in uranium, provides no indication that they represent subsurface conduits for fluids that deposit Fe-Mn-Si at the seafloor, though this remains a possibility for some such deposits.

Sedimentation and sediment redistribution on the Cocos Ridge from two seismic acquisition cruises, NEMO-3 and MV1014

We use digital seismic reflection profiles within a 1° * 1° survey area on the Cocos Ridge (COCOS6N) to study the extent and timing of sedimentation and sediment redistribution on the Cocos Ridge. The survey was performed to understand how sediment focusing might affect paleoceanographic flux measurements in a region known for significant downslope transport. COCOS6N contains ODP Site 1241 to ground truth the seismic stratigraphy, and there is a seamount ridge along the base of the ridge that forms a basin (North Flank Basin) to trap sediments transported downslope. Using the Site 1241 seismic stratigraphy and densities extrapolated from wireline logging, we document mass accumulation rates (MARs) since 11.2 Ma. The average sediment thickness at COCOS6N is 196 m, ranging from outcropping basalt at the ridge crest to ~ 400 m at North Flank Basin depocenters. Despite significant sediment transport, the average sedimentation over the entire area is well correlated to sediment fluxes at Site 1241. A low mass accumulation rate (MAR) interval is associated with the 'Miocene carbonate crash' interval even though COCOS6N was at the equator at that time and relatively shallow. Highest MAR occurs within the late Miocene-early Pliocene biogenic bloom interval. Lowest average MAR is in the Pleistocene, as plate tectonic motions caused COCOS6N to leave the equatorial productivity zone. The Pliocene and Pleistocene also exhibit higher loss of sediment from the ridge crest and transport to North Flank Basin. Higher tidal energy on the ridge caused by tectonic movement toward the margin increased sediment focusing in the younger section.

Pressure, tilt, temperature and conductivity monitoring within the Immerged Santorini Caldera during Jul 2012 - Sep 2013

Bottom pressure, tilt and seawater physical-properties were monitored for a year using two instruments within the immerged Santorini caldera (Greece). Piggy-backed on the CALDERA2012 cruise, this geodetic experiment was designed to monitor evolution of the 2011-2012 Santorini unrest. Conducted during a quiescent period, it allowed us to study oceanographic and atmospheric signal in our data series. We observe periodic oceanographic signals associated with tides, and seiches that are likely linked to both the caldera and Cretan basin geometries. In winter, the caldera witnesses sudden cooling events that tilt an instrument towards the Southeast, indicating cold-water influx likely originating from the north-western passage between Thirasia and Oia. We do not obtain evidence of long-term vertical seafloor deformation from the pressure signal, although it may be masked by instrumental drift. However, tilt data suggests a local seafloor tilt event ~1 year after the end of the unrest period which could be consistent with inflation under or near Nea Kameni. Seafloor geodetic data recorded at the bottom of the Santorini caldera illustrates that the oceanographic signature is an important part of the signal, which needs to be considered for monitoring volcanic or geological seafloor deformation in shallow-water and/or nearshore areas.

Chemical composition of bulk sediments from ODP Hole 199-1221C

A record of inorganic geochemical variability was produced from a contiguous sequence of 35 samples, with 1 cm spacing, recovered from Hole 1221C. This record covers from 153.91 to 154.27 meters below seafloor and spans the Carbon Isotope Excursion (CIE) associated with the Paleocene/Eocene boundary interval. Elemental concentrations were determined for Al, As, Ba, Ca, Fe, K, Mg, Mn, P, Si, Sr, Ti, Cd, Co, Cr, Cu, Hf, Mo, Nb, Ni, Pb, Pt, Re, Sc, V, Y, Zn, La, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu. Most concentration profiles exhibit a marked peak coincident with or just prior to the CIE. In addition, the rare earth element pattern exhibits a significant flattening of the typical, prominent negative Ce anomaly across the same interval.

Chemical composition of altered volcanic rocks from ODP Sites 193-1188 and 193-1189

The conventional model of leaching volcanic rocks as a source of metals in a seafloor hydrothermal systems has been tested by examining the behavior of Pb and other trace elements during hydrothermal alteration. ODP Leg 193 drill sites 1188 (Snowcap) and 1189 (Roman Ruins) on Pual Ridge in the eastern Manus Basin offshore eastern Papua New Guinea provide a unique three-dimensional window into an active back-arc hydrothermal system. We investigate by means of a LA-ICP-MS microbeam technique the capacity of Pb to be leached from a host volcanic rock exposed to various types and intensities of alteration. Our results are in general agreement with previous studies that utilized bulk analytical techniques but provide a more detailed explanation of the processes. Fresh representative dacitic lavas from the Pual Ridge have an average whole rock Pb content of 5.2 ppm, an average interstitial glass Pb content of 5.6 ppm and an average plagioclase Pb content of 1.0 ppm. Altered matrix samples have highly variable Pb values ranging from 0 to 52.4 ppm. High Pb values in altered samples are associated with a low temperature chlorite and clay mineral assemblage, in some cases overprinted by a high temperature (up to 350°C) silica-rich "bleaching" alteration. Only the most highly altered matrix samples have REE patterns that differ from the fresh Pual Ridge dacite. This may represent either different lava histories or alteration characteristics that have affected normally immobile REEs. Altered samples with the highest Pb values have similar REE patterns to those of the local unaltered lavas. They are compositionally similar to typical Pual Ridge dacites indicating a genetic relationship between the main regional volcanic suite and the subseafloor hydrothermally altered, Pb-enriched material. Relative loss/gain for Pb between the analyzed altered samples and a calculated precursor show a maximum relative gain of 901%. Samples with relative Pb gain from both drill sites are associated with lower temperature alteration mineral assemblages characterized by pervasive chloritization. The related lower temperature (220-250°C) neutral to slightly acidic fluids have been ascribed by others to return circulation of hydrothermal fluids that did not interact with seawater. Because altered samples have a higher Pb content than the fresh precursor, leaching of fresh volcanic rocks cannot be the source of Pb in the hydrothermal systems.

Calcareous nannofossil range-distribution and zonation of sediments from three ODP Leg 205 holes

Ocean Drilling Program Leg 205 of the research vessel JOIDES Resolution was a return expedition to the Leg 170 sites located on the Costa Rica subduction zone. Here the entire sediment cover on the incoming Cocos plate, including significantly large sections of calcareous nannofossil ooze and chalk, is underthrust beneath the overriding Caribbean plate. The large amount of subducted carbonate produces characteristic styles of volcanic and seismic activity that differ from those found farther along strike in Nicaragua and elsewhere. An understanding of the fate of subducted carbonate sediment sections is an essential component to our understanding of the global biogeochemical cycling of carbon dioxide. Because Leg 205 drilling operations were performed within meters of the Leg 170 drill sites occupied during October-December 1996, minimal coring was done during Leg 205. Although the biostratigraphy of the Leg 170 sites has since been documented in detail, questions remained regarding the age and nature of a gabbro sill that was only partially penetrated by coring during Leg 170. Coring operations during Leg 205 fully penetrated the gabbro sill, followed by an additional 12 m of sediments below the sill, and then ~160 m of gabbro. Coring halted at 600 meters below seafloor (mbsf). Calcareous nannofossil age dating of the sediments immediately above the igneous sill, as well as the sediment between the sill and the lower igneous unit, indicates a minimum age of 15.6 Ma and a maximum age of 18.2 Ma for the sediments. This implies that the sill was emplaced more recently than 18.2 Ma. The calcareous nannofossil assemblage in baked sediments in contact with the top of the lower igneous unit also suggests that the maximum age for emplacement is 18.2 Ma. At Site 1254, coring was accomplished between 150 and 230 mbsf (prism section), and from 300 to 367.5 mbsf (prism and through the décollement into the underthrust section). In the interval from 150 to 322 mbsf, the biostratigraphic analysis of calcareous nannofossils suggests that the sediments are early Pleistocene age between 150 and 161 mbsf, late Pliocene age from 161 to 219 mbsf, and early Pliocene age from 219 to 222 mbsf (no younger than 3.75 Ma). The lack of marker fossils in the interval of sediments cored from 300 to 350.6 mbsf does not allow for any age determinations; however, sediments from 351.6 to 359.81 mbsf could be age dated and are also early Pliocene age, but no younger than 3.75 Ma.

Mineral and chemical composititon of sediments from ODP Hole 119-740A

ODP Hole 740A is located on the inner part of the East Antarctic continental shelf in Prydz Bay, at the seaward end of a major onshore rift structure known as the Lambert Graben. Drilling at this site led to the recovery of some 65 m of continental sediments (Prydz Bay red beds) that form part of a much thicker (2-3 km) pre-continental breakup sequence, the development of which may be related to the initiation and rifting of the Lambert Graben. Palynological and paleomagnetic studies have not been able to determine the age of the sediments; they may be equivalent to the onshore late Permian Amery Group or younger. The succession consists predominantly of sandstone, siltstone, and claystone arranged in erosively based, pedogenically influenced fining-upward sequences up to 5 m thick. These were deposited by shallow, braided streams draining an extensively vegetated alluvial plain, with sufficient topographic relief to trap fine-grained sediment and inhibit rapid channel shifting. Pedogenic processes were initiated on the alluvial plain, but climatic conditions were generally unsuitable for extensive pedogenic carbonate formation and the development of mature soil profiles. The sediments were probably derived from a rapidly uplifted fault block terrain composed of upper Proterozoic and Archaean gneisses lying to the southeast of the depositional site. Uplift may have taken place along the tectonically active seaward extension of the eastern faulted margin of the Lambert Graben, which passes immediately southeast of Hole 740A. Differences in mineralogical composition between the Amery Group and the Prydz Bay red beds probably reflect differences in rock composition in the source area. The age of the Prydz Bay red beds has still to be resolved.

Primary composition, alteration, and origin of Cretaceous volcaniclastic rocks at DSDP Site 89-585

An upper Aptian to middle Albian series of volcaniclastic rocks more than 300 m thick was drilled at Site 585 in the East Mariana Basin. On the basis of textural and compositional (bulk-rock chemistry, primary and secondary mineral phases) evidence, the volcaniclastic unit is subdivided into a lower (below 830 m sub-bottom) and an upper (about 670-760 m) sequence; the boundary in the interval between is uncertain owing to lack of samples. The rocks are dominantly former vitric basaltic tuffs and minor lapillistones with lesser amounts of crystals and basaltic lithic clasts. They are mixed with shallow-water carbonate debris (ooids, skeletal debris), and were transported by mass flows to their site of deposition. The lower sequence is mostly plagioclase- and olivine-phyric with lesser amounts of Ti-poor clinopyroxene. Mineralogical and bulk-rock chemical data indicate a tholeiitic composition slightly more enriched than N-MORB (normal mid-ocean ridge basalt). Transport was by debris flows from shallow-water sites, as indicated by admixed ooids. Volcanogenic particles are chiefly moderately vesicular to nonvesicular blocky shards (former sideromelane) and less angular tachylite with quench plagioclase and pyroxene, indicating generation of volcanic clasts predominantly by spalling and breakage of submarine pillow and/or sheet-flow lavas. The upper sequence is mainly clinopyroxene- and olivine-phyric with minor plagioclase. The more Ti-rich clinopyroxene and the bulk-rock analyses show that the moderately alkali basaltic composition throughout is more mafic than the basal tholeiitic sequence. Transport was by turbidity currents. Rounded epiclasts of crystalline basalts are more common than in the lower sequence, and, together with the occurrence of oxidized olivine pseudomorphs and vesicular tachylite, are taken as evidence of derivation from eroded subaerially exposed volcanics. Former sideromelane shards are more vesicular than in the lower sequence; vesicularity exceeds 60 vol.% in some clasts. The dominant clastic process is interpreted to be by shallow-water explosive eruptions. All rocks have undergone low-temperature alteration; the dominant secondary phases are "palagonite," chlorite/smectite mixed minerals, analcite, and chabazite. Smectite, chlorite, and natrolite occur in minor amounts. Phillipsite is recognized as an early alteration product, now replaced by other zeolites. During alteration, the rocks have lost up to 50% of their Ca, compared with a fresh shard and fresh glass inclusions in primary minerals, but have gained much less K, Rb, and Ba than expected, indicating rapid deposition prior to significant seafloor weathering.

Mineral chemistry of primary and secondary phases in basaltic rocks

Primary magmatic phases (spinel, olivine, plagioclase, clinopyroxene, amphibole, and biotite) and secondary phyllosilicates (smectite, chlorite-smectite, and celadonite) were analyzed by electron microprobe in alkalic and tholeiitic dolerites and basalts from Ocean Drilling Program Sites 800, 801, and 802. Aphyric alkalic dolerite sills (Hole 800A) and basalt flows (Holes 801B and 801C) share common mineralogical features: matrix feldspars are strongly zoned from labradorite cores to discrete sodic rims of alkali feldspar with a high Or component, which overlaps that of quench microlites in glassy mesostasis; little fractionated clinopyroxenes are Ti-rich diopsides and augites (with marked aegirine-augite rims at Site 801); rare, brown, Fe**3+-rich amphibole is winchite; and late biotites exhibit variable Ti contents. Alkalic rims to feldspars probably developed at the same time as quenched mesostasis feldspars and late-stage magmatic biotite, and represent the buildup of K-rich hydrous fluids during crystallization. Phenocryst phases in primitive mid-ocean ridge tholeiites from Hole 801C (Mg numbers about 70) have extreme compositions with chrome spinel (Cr/Cr + Al ratios about 0.2-0.4), Ni-rich olivine (Fo90), and highly calcic plagioclase (An90). Later glomerophyric clumps of plagioclase (An75-80) and clinopyroxene (diopside-augite) are strongly zoned and probably reflect rapidly changing melt conditions during upward transport, prior to seafloor quenching. In contrast, phenocryst phases (olivine, plagioclase, and clinopyroxene) in the Hole 802A tholeiites show limited variation and do not have such primitive compositions, reflecting the uniform and different chemical composition of all the bulk rocks. Replacive phyllosilicates in both alkalic and tholeiitic basalts include various colored smectites (Fe-, Mg-, and Al-saponites), chlorite-smectite and celadonite. Smectite compositions typically reflect the replaced host composition; glass is replaced by brown Fe-saponites (variable Fe/Mg ratios) and olivine by greenish Mg-saponites (or Al-rich chlorite-smectite).