Geochemistry on rocks from Shackleton Range

During the GEISHA expedition (Geologische Expedition in die Shackleton Range 1987/88), the Pioneers Escarpment was visited and sampled extensively for the first time. Most of the rock types encountered represent amphibolite facies metamorphics, but evidence for granulite facies conditions was found in cores of garnet. These conditions must have been at least partly reached during the peak of metamorphism. For the Pioneers Escarpment a varicolored succession of sedimentary and bimodal volcanic origin is typical. It comprises: quartzites muscovite quartzite, sericite quartzite, fuchsite quartzite, garnet-quartz schists etc.; pelites: mica schists and plagioclase or plagioclase-microcline gneisses, aluminous schists; marls and carbonates: grey meta-limestones, carbonaceous quartzites, but also pure white, often fine-grained, saccharoidal marble, or a variety of tremolite marble, olivine (forsterite) marble, diopside-clinopyroxene-tremolite marble, etc.; basic volcanic rocks: amphibole fels, amphibolite schist, garnet amphibolite, and acidic to intermediate volcanic rocks: garnet-biotite schist, epidote-biotite-plagioclase gneiss, microcline gneiss. These rocks are considered to be a supracrustal unit, called the Pioneers Group. In the easternmost parts of the Pioneers Escarpment, e.g. at Vindberget, nonmetamorphic shales, sandstones and greywackes crop out, which are cover rocks of possibly Jurassic age. These metasediments, which represent a quartz-pelite-carbonate (QPC) association, indicate that deposition took place on a stable shelf, i.e. on the submerged rim of a craton. Marine shallow-water sedimentation including marls and aluminous clays form the protoliths. The volcanics may be part of a bimodal volcanics-arkose-conglomerate (BVAC) association. Geochemical analyses support the assumption of volcanic protoliths. This is demonstrated especially by the elevated amounts of the immobile, incompatible high-field-strength elements (HFSE) Nb, Ta, Ti, Y, and Zr encountered in some of the gneisses. Microscopic investigation suggests the existence of ortho-amphibolites. This is confirmed by the geochemistry. A bimodal volcanic association is evident. The amphibolites plot in both the tholeiite and calc-alkaline fields. The acidic volcanics are mainly rhyolitic. The sediments and volcanics were subjected to conditions of 10-11 kbar and 600°C during the peak of metamorphism, i.e. granulite facies metamorphism, which can be deduced from the Fe mole ratios of 0.71-0.73 in the garnet cores. Due to the relatively low temperatures, no anatectic melting took placc. The rims of the garnets show a Fe mole ratio of 0.84-0.86, and the coexisting mineral association garnet-biotite-staurolite-kyanite indicate amphibolite facies. The thermobarometry shows P-T conditions of 5-6 kbar and 570-580°C for this stage. The metamorphic history indicates deep burial at depths down to 35 km (subduction?) i.e. high pressure metamorphism, followed by pressure release due to uplift associated with retrograde metamorphism. This may have happened during a pre-Ross metamorphic event or orogeny. The Ross Orogeny at about 500 Ma probably just led to the weak greenschist facies overprint that is evident in the rocks of the Pioneers Group. Finally, sedimentation resumed in the area of the present Shackleton Range, or at least in the eastern part of the Pioneers Escarpment, probably when detritus from erosion of the basement (Read Group and Pioneers Group) was deposited, forming sandstones and greywackes of possibly Jurassic age. There is no indication that these sediments belong to the former Turnpike Bluff Group.

Geochemistry of ODP Leg 127 igneous rocks

Legs 127 and 128 of the Ocean Drilling Program cored basement samples from two sites in the Yamato Basin (Sites 794 and 797) and one site in the Japan Basin (Site 795) of the Japan Sea. These samples represent sills and lava flows erupted or shallowly intruded in a marine environment during backarc extension and spreading in the middle Miocene. In this paper, we describe the geochemical characteristics of these igneous units using 52 new instrumental neutron activation analyses (INAA), 8 new X-ray fluorescence (XRF) analyses, and previous shipboard XRF analyses. The sills intruded into soft sediment at Sites 794 and 797 were subject to extensive hydrothermal activity, estimated at <230° C under subgreenschist facies conditions, which heavily to totally altered the fine-grained unit margins and moderately to heavily altered the coarse-grained unit interiors. Diagenesis further altered the composition of these igneous bodies and lava flows at Sites 794, 795, and 797, most intensely at unit margins. Our study of two well-sampled units shows that Mg, Ca, Sr, and the large-ion lithophile elements (LILE) mobilized during alteration, and that the concentrations of Y, Yb, and Lu decreased and Ce increased in the most severely altered samples. Nevertheless, our study shows that the rare-earth elements (REE) were relatively immobile in the majority of the samples, even where secondary mixed-layer clays comprised the great majority of the rock. Fresher Yamato Basin samples are compositionally heterogenous tholeiitic basalts and dolerites. At Site 794 in the north-central portion of the basin, Units 1 to 5 (upper basement) comprise mildly light rare-earth element (LREE) enriched basalts and dolerites (chondrite-normalized La/Sm of 1.4-1.8), while the stratigraphically lower Units 6 to 9 are less enriched dolerites with (La/Sm)N of 0.7-1.3. All Site 794 samples lack Nb and Ta depletions and LILE enrichments, lacking a strong subduction-related incompatible element geochemical signature. At Site 797 in the western margin of the basin, two stratigraphically-definable unit groups also occur. The upper nine units are incompatible-element depleted tholeiitic sills and flows with strong depletions of Nb and Ta relative to normal mid-ocean ridge basalt (N-MORB). The lower twelve sills represent LREE-enriched tholeiites (normalized La/Sm ranges from 1.1 to 1.8), with distinctly higher LILE and high field-strength element (HFSE) contents. At Site 795 at the northern margin of the Japan Sea, three eruptive units consist of basaltic andesite to calc-alkaline basalt (normalized La/Sm of 1.1 to 1.5) containing moderate depletions of the HFSE relative to N-MORB. The LILE-depleted nature of these samples precludes their origin in a continental arc, indicating that they more likely erupted within a rifting oceanic arc system. The heterogenous nature of the Japan Sea rocks indicate that they were derived at each site from multiple parental magmas generated from a compositionally heterogenous mantle source. Their chemistry is intermediate in character between arc basalts, MORB, and intraplate basalts, and implies little involvement of continental crust at any point in their genesis. Their flat chondrite-normalized, medium-to-heavy rare earth patterns indicate that the primary magmas which produced them last equilibrated with and segregated from spinel lherzolite at shallow depths (<30 kbar). In strong contrast to their isotopic compositional arrays, subduction-related geochemical signatures are usually poorly defined. No basin-wide temporal or geographic systematics of rock chemistry may be confidently detailed; instead, the data show both intimate (site-specific) and widespread backarc mantle heterogeneity over a narrow (2 Ma or so) range in time, with mantle heterogeneity most closely resembling a "plum-pudding" model.

Mineralogy of southern Kerguelen Plateau sediments

Kaolinite, goethite, minor hematite, and gibbsite were found in fluvial upper Lower Cretaceous basal sediment from the Southern Kerguelen Plateau, Sites 748 and 750, 55°S latitude. This mineral assemblage, derived from the weathering of basalt, indicates near-tropical weathering conditions with high orographic rainfall, at least 100 cm per year. The climate deteriorated by the Turonian or Coniacian, as indicated by the decline in kaolinite content of this sediment. The Upper Cretaceous sediment at Site 748 consists of 200 m of millimeter-laminated, sparsely fossiliferous, wood-bearing glauconitic siltstone and clay stone with siderite concretions deposited on a shelf below wave base. Some graded and cross beds indicate that storms swept over the shelf and reworked the sediment. Overlying this unit is 300 m of intermittently partly silicified, bryozoan-inoceramid-echinoderm-rich glauconitic packstones, grainstones, and wackestones. The dominant clay mineral in both units is identical to the mineral composition of the glauconite pellets: randomly interstratified smectite-mica. The clay fraction has a higher percent of expandable layers than the mineral of the glauconite pellets, and the clay of the underlying subunit has a higher percentage of expandable layers than the clay of the carbonate subunit. Potassium levels mirror these mineral variations, with higher K levels in minerals that have a lower percentage of expandable layers. The decrease in expandability of the mineral in the upper subunit is attributed to diagenesis, the result of higher porosity.

Geochemistry of Baltic Sea sediments

The Baltic Sea has experienced three major intervals of bottom water hypoxia following the intrusion of seawater ca. 8 kyrs ago. These intervals occurred during the Holocene Thermal Maximum (HTM), Medieval Climate Anomaly (MCA) and during recent decades. Here, we show that sequestration of both Fe and Mn in Baltic Sea sediments generally increases with water depth, and we attribute this to shelf-to-basin transfer ("shuttling") of Fe and Mn. Burial of Mn in slope and basin sediments was enhanced following the lake-brackish/marine transition at the beginning of the hypoxic interval during the HTM. During hypoxic intervals, shelf-to-basin transfer of Fe was generally enhanced but that of Mn was reduced. However, intensification of hypoxia within hypoxic intervals led to decreased burial of both Mn and Fe in deep basin sediments. This implies a non-linearity in shelf Fe release upon expanding hypoxia with initial enhanced Fe release relative to oxic conditions followed by increased retention in shelf sediments, likely in the form of iron sulfide minerals. For Mn, extended hypoxia leads to more limited sequestration as Mn carbonate in deep basin sediments, presumably because of more rapid reduction of Mn oxides formed after inflows and subsequent escape of dissolved Mn to the overlying water. Our Fe records suggest that modern Baltic Sea hypoxia is more widespread than in the past. Furthermore, hypoxia-driven variations in shelf-to-basin transfer of Fe may have impacted the dynamics of P and sulfide in the Baltic Sea thus providing potential feedbacks on the further development of hypoxia.

Chronostratigraphy of sediments from the Bahamas

Continuous cores drilled during the Bahamas Drilling Project (BDP) and the Ocean Drilling Program (ODP) Leg 166 along a transect from the top of Great Bahama Bank to the basin in the Straits of Florida provide a unique data set to test the assumption in seismic stratigraphy that seismic reflections are time lines and, thus, have a chronostratigraphic significance. Seismic reflections that are identified as seismic sequence boundaries (SSBs) were dated by means of biostratigraphy in the five ODP sites and by a combination of biostratigraphy, magnetostratigraphy and Sr isotope stratigraphy in the two BDP sites. The seismic reflection horizons are carried across a variety of facies belts from shallow-water carbonates over slope carbonates to drift deposits in the Straits of Florida. Within this system 17 SSBs were identified and dated. Despite the fact that the seismic reflections cross several facies belts, their ages remain remarkably constant. The average offset in all sites is 0.38 Myr. In no cases do the seismic reflections cut across time lines. The age differences are the combined result of the biostratigraphic sampling frequency, the spacing of marker species that required extrapolation of ages, and the resolution of the seismic data. In addition, uncertainties of age determination in the proximal sites where age-diagnostic fauna are rare add to the age differences between sites. Therefore, it can be concluded that the seismic reflections, which mark the SSBs along the Bahamas Transect, are time lines and can be used as stratigraphic markers. This finding implies that depositional surfaces are preferentially imaged by reflected seismic waves and that an impedance contrast exists across these surfaces. Facies successions across the sequence boundaries indicate that the sequence boundaries coincide with the change of deposition from times of high to low sea level. In the carbonate setting of Great Bahama Bank, sea-level changes produce changes in sediment composition, sedimentation rate and diagenesis from the platform top to the basin. The combination of these factors generates differences in sonic velocity and, thus, in impedance that cause the seismic reflection. The impedance contrasts decrease from the proximal to the distal sites, which is reflected in the seismic data by a decrease of the seismic amplitude in the basinal area.

Highly dealkylated copper and nickel etioporpyfrins in marine sediments

Copper porphyrins have been recognized as natural constituents of marine sediments only within the past 5 years (Palmer and Baker, 1978, Science201, 49-51). In that report it was suggested that these pigments may derive from and be markers for oxidized terrestrial organic matter redeposited in the marine environment. In the present study we describe the distribution of copper porphyrins in sediments from several north Pacific and Gulf of California DSDP/IPQD sites (Legs 56,63,64). These allochthonous pigments have now been found to be accompanied by identical arrays of highly dealkylated nickel etioporphyrins. Evaluation of data from this and past studies clearly reveals that there is a strong carbon-number distribution similarity betweeen coincident Cu and Ni etioporphyrins. This homology match is taken as reflecting a common source for the tetrapyrrole ligands of this population of Cu and Ni chelates. Predepositional generation of these highly dealkylated etioporphyrins is concluded from the occurrence of these pigments in sediments continuing essentially all stages of in situ chlorophyll diagenesis (cf. Baker and Louda, 1983). That is, their presence is not regulated by the in situ diagenetic continuum. Thus, the highly dealkylated Cu and Ni etioporphyrins represent an 'allochthonous' background over which 'autochthonous' (viz. marine produced) chlorophyll derivatives are deposited and are undergoing in situ diagenesis.

Physical properties, compressional-wave velocity and consolidation characteristics of ODP Holes 133-815B and 133-817B

Sites 815 and 817 were drilled near the Townsville Trough during Leg 133 of the Ocean Drilling Program. The physical properties, compressional-wave velocity, and consolidation characteristics indicate that the periplatform carbonate sediments maintain more water content and lower compressional velocity near the Queensland Plateau than the clayey hemipelagic sediments, which have a clay content of up to 60%. Bulk density, void ratio or porosity, water content, and compressional-wave velocity are shown to have a linear relationship with burial depth. Between 3.5 and 5 Ma (about 100-500 mbsf), these physical properties maintained a constant rate vs. the depth in core because of the fast sedimentation-rate effect at Site 815. However, compressionalwave velocity still increases downward in this section. The clay content in this section causes an increase of bulk modulus and compaction effect. At Site 817, scarce terrigenous mud content and abundant carbonate content (88%-97%) cause a straight line relationship between physical properties and burial depth. During the consolidation test, we show that dominant micritic particles may cause faster acoustic velocity than sediments composed mainly of coccoliths. The bulk modulus ratio increasing rate in the clay-rich carbonate sediments is almost 4.5 times higher than in the clay-free periplatform carbonate sediments.

Hydrocarbon gases in ODP Leg 104 samples

Geochemical studies at three ODP Leg 104 sites on the Wring Plateau help define the distribution of hydrocarbon gases in sediment of this prominent feature of the Norwegian continental margin. Low levels of hydrocarbon gas were encountered in sediment of the outer part of the plateau, but sediment of the inner part of the plateau is very gassy. The molecular composition of inner plateau gases (>99.9% methane) and the carbon isotopic composition of the methane (avg. = -76 per mil relative to the PDB standard) clearly show that the gas is biogenic. Heavier hydrocarbon gases accompany this methane, and their presence is probably a result of both chemical and microbial low-temperature diagenesis. Although these heavier hydrocarbons were not detected in sediment of the outer part of the plateau during shipboard analyses, subsequent shore-based analyses showed that these compounds are present at very low concentrations. Methane in the gassy sediment of the inner part of the plateau may be present as gas hydrates, judging from sedimentological and inorganic geochemical considerations, but no discernible gas hydrates were recovered during drilling.

Stable carbon and oxygen isotope ratios of benthic foraminifera from Paleocene to Oligocene sediments

Benthic oxygen and carbon isotopic results from a depth transect on Maud Rise, Antarctica, provide the first evidence for Warm Saline Deep Water (WSDW) in the Paleogene oceans. Distinct reversals occur in the oxygen isotopic gradient between the shallower Hole 689B (Eocene depth ~1400 m; present-day depth 2080 m) and the deeper Hole 690B (Eocene depth ~2250 m; present-day depth 2914 m). The isotopic reversals, well developed by at least 46 Ma (middle middle Eocene), existed for much of the remaining Paleogene. We do not consider these reversals to be artifacts of differential diagenesis between the two sites or to have resulted from other potentially complicating factors. This being so, the results show that deep waters at Hole 690B were significantly warmer than deep waters at the shallower Hole 689B. A progressive decrease and eventual reversal in benthic to planktonic delta18O gradients in Hole 690B, demonstrate that the deeper waters became warmer relative to Antarctic surface waters during the Eocene. The warmer deep waters of the Paleogene are inferred to have been produced at middle to low latitudes, probably in the Tethyan region which contained extensive shallow-water platforms, ideal sites for the formation of high salinity water through evaporative processes. The ocean during the Eocene, and perhaps the Paleocene, is inferred to have been two-layered, consisting of warm, saline deep waters formed at low latitudes and overlain by cooler waters formed at high latitudes. This thermospheric ocean, dominated by halothermal circulation we name Proteus. The Neogene and modern psychrospheric ocean Oceanus is dominated by thermohaline circulation of deep waters largely formed at high latitudes. An intermediate condition existed during the Oligocene, with a three-layered ocean that consisted of cold, dense deep waters formed in the Antarctic (Proto-AABW), overlain by warm, saline deep waters from low latitudes, and in turn overlain by cool waters formed in the polar regions. This we name Proto-oceanus which combined both halothermal and thermohaline processes. The sequence of high latitude, major, climatic change inferred from the oxygen isotopic records is as follows: generally cooler earlier Paleocene; warming during the late Paleocene; climax of Cenozoic warmth during the early Eocene and continuing into the early middle Eocene; cooling mainly in a series of steps during the remainder of the Paleogene. Superimposed upon this Paleogene pattern, the Paleocene/Eocene boundary is marked by a brief but distinct warming that involved deep to surface waters and a reduction in surface to deep carbon and oxygen isotopic gradients. This event coincided with major extinctions among the deep-sea benthic foraminifers as shown by Thomas (1990 doi:10.2973/odp.proc.sr.113.123.1990). Salinity has played a major role in deep ocean circulation, and thus paleotemperatures cannot be inferred directly from the oxygen isotopic composition of Paleogene benthic foraminifers without first accounting for the salinity effect.

Nannofossil and stable isotope record across the Paleocene/Eocene Thermal Maximum in ODP Hole 183-1135A

The Paleocene/Eocene Thermal Maximum (PETM, ca. 55 Ma) is an abrupt, profound perturbation of climate and the carbon cycle associated with a massive injection of isotopically light carbon into the ocean-atmosphere system. As such, it provides an analogue for understanding the interplay between phytoplankton and climate under modern anthropogenic global-warming conditions. However, the accompanying enhanced dissolution poses uncertainty on the reconstruction of the affected ecology and productivity. We present a high-resolution record of bulk isotopes and nannofossil absolute abundance from Ocean Drilling Program (ODP) Site 1135 on the Kerguelen Plateau, Southern Indian Ocean to quantitatively constrain for the first time the influence of dissolution on paleoecological reconstruction. Our bulk-carbonate isotope record closely resembles that of the classic PETM site at ODP Site 690 on the opposite side of the Antarctic continent, and its correlation with those from ODP Sites 690, 1262 and 1263 records allows recognition of 14 precessional cycles upsection from the onset of the carbon isotopic excursion (CIE). This, together with a full range of common Discoasteraraneus and an abundance crossover between Fasciculithus and Zygrhablithusbijugatus, indicates the presence of the PETM at Site 1135, a poorly known record with calcareous fossils throughout the interval. The strong correlation between the absolute abundances of Chiasmolithus and coccolith assemblages reveals a dominant paleoecological signal in the poorly preserved fossil assemblages, while the influence of dissolution is only strong during the CIE. This suggests that r-selected taxa can preserve faithful ecological information even in the severely-altered assemblages studied here, and therefore provide a strong case for the application of nannofossils to paleoecological studies in better-preserved PETM sections. The inferred nannoplankton productivity drops abruptly at the CIE onset, but rapidly increases after the CIE peak, both of which may be driven by nutrient availability related to ocean stratification and vertical mixing due to changed sea-surface temperatures.

Radiolarians at DSDP Leg 58 Holes

Radiolarians were observed at all five sites drilled during DSDP Leg 58. Three sites (442, 443, 444) are south of Japan in the Shikoku Basin. The remaining two sites (445, 446) are east of Okinawa, in the Daito Ridge and Basin areas. The observations made on radiolarians during Leg 58 are understood best by considering these two areas separately. The basement ages, preservation, diagenesis, and paleoecology are similar within each area, but different between the two areas. The radiolarian zones of Riedel and Sanfilippo (1978) were used to determine the sediment age. Because of the mixed nature of the fauna, there was an opportunity to test the tropical zonation in middlelatitude sediments. A middle- to high-latitude biostratigraphy for the Pliocene and Pleistocene has been formulated (Hays, 1970; Kling, 1973; Foreman, 1975), but there is no Miocene radiolarian zonation for these latitudes. The tropical elements of the present fauna are sufficient to use the low-latitude zonation, although there is a loss of resolution in the Pleistocene. Because of poor preservation, zone boundaries are indistinct in much of the cored sediment. Determination of abundance in any sample is always subjective and varies among investigators. This work was in its final stages at the publication of Westberg and Riedel (1978), and the guidelines outlined therein are not closely followed. The abundances recorded in Tables 1 through 5 are based on strewn slides which were searched entirely if an individual of a species was found, or for 8 to 10 minutes if the species was not found.

Biostratigraphy and Sr isotopic composition of sediments from DSDP Legs 12, 14, 15, 17, 20-22, and 30

We present the data used to construct the Cenozoic and Cretaceous portion of the Phanerozoic curve of seawater 87Sr/86Sr that had been given in summary form by W.H. Burke co-workers. All Cenozoic samples (128) and 22 Cretaceous samples are foram-nannofossil oozes and limestones from DSDP cores distributed among 13 sites in the Atlantic, Pacific and Indian Oceans, and the Caribbean Sea. Non-DSDP Cretaceous samples (126) include limestone, anhydrite and phosphate samples from North America, Europe and Asia. Determination of the 87Sr/86Sr value of seawater at particular times in the past is based on comparison of ratios derived from coeval marine samples from widely separated geographic areas. These samples are characterized by a wide variety of diagenetic and burial histories. The large size and cosmopolitan nature of the data set decreases the likelihood that, among coeval data, systematic error has been introduced by a similar pattern of diagenetic alteration of the ratios. There is good clustering of data points throughout the Cenozoic and Cretaceous curve. The consistency of data is illustrated by Cenozoic and Cretaceous data plots that include a separate symbol for each DSDP site and non-DSDP sample location. More than 98% of the data points are enclosed by upper and lower lines that define a narrow band. For any given time, the correct seawater ratio probably lies within this band. A line drawn within the band represents our estimate of the actual seawater ratio as a function of time. The general configuration of the Cenozoic and Cretaceous curve appears to be strongly influenced by the history of plate interactions and sea-floor spreading. Specific rises and falls in the 87Sr/86Sr of seawater, however, may be caused by a variety of factors such as variation in lithologic composition of the crust exposed to weathering, configuration and topographic relief of continents, volcanic activity, rate of sea-floor spreading, extent of continental inundation by epeiric seas, and variations in both climate and paleooceanographic conditions. Many or all of these factors are probably related to global tectonic processes, yet their combined effect on the temporal variation of seawater 87Sr/86Sr can complicate a direct platetectonic interpretation for portions of the seawater curve.

Composition and distribution of silica in sediments of the Broken Ridge

Within a dipping sequence of middle Cretaceous to Eocene sediments on Broken Ridge, opal-A, opal-CT, and quartz occur as minor constituents in carbonate and ash-rich sediments. Biogenic opal-A is mainly derived from diatoms and radiolarians. Opal-A and almost all siliceous microfossils disappear within a narrow (<20-m-thick) transition zone below which authigenic opal-CT and quartz are present. These latter silica polymorphs occur together within a 750-m-thick interval, but the ratio of quartz/opal-CT increases with increasing age and depth within the pre-rift sediment sequence. The boundary between opal-A- and opal-CT-bearing sediments is also a physical boundary at which density, P-wave velocity, and acoustic impedance change. This physical transition is probably caused by infilling of pore space by opal-CT lepispheres.

Chert petrology and geochemistry at DSDP Leg 62 Holes

Sixty-five chert, porcellanite, and siliceous-chalk samples from Deep Sea Drilling Project Leg 62 were analyzed by petrography, scanning electron microscopy, analysis by energy-dispersive X-rays, X-ray diffraction, X-ray spectroscopy, and semiquantitative emission spectroscopy. Siliceous rocks occur mainly in chalks, but also in pelagic clay and marlstone at Site 464. Overall, chert probably constitutes less than 5% of the sections and occurs in deposits of Eocene to Barremian ages at sub-bottom depths of 10 to 820 meters. Chert nodules and beds are commonly rimmed by quartz porcellanite; opal-CT-rich rocks are minor in Leg 62 sediments 65 to 108 m.y. old and at sub-bottom depths of 65 to 520 meters. Chert ranges from white to black, shades of gray and brown being most common; yellow-brown and red-brown jaspers occur at Site 464. Seventy-eight percent of the studied cherts contain easily recognizable burrow structures. The youngest chert at Site 463 is a quartz cast of a burrow. Burrow silica maturation is always one step ahead of host-rock silicification. Burrows are commonly loci for initial silicification of the host carbonate. Silicification takes place by volume-f or-volume replacement of carbonate sediment, and more-clay-rich sediment at Site 464. Nannofossils are commonly pseudomorphically replaced by quartz near the edges of chert beds and nodules. Other microfossils, mostly radiolarians and foraminifers, whether in chalk or chert, can be either filled with or replaced by calcite, opal-CT, and (or) quartz. Chemical micro-environments ultimately control the removal, transport, and precipitation of calcite and silica. Two cherts from Site 465 contain sulfate minerals replaced by quartz. Site 465 was never subaerially exposed after sedimentation began, and the formation of the sulfate minerals and their subsequent replacement probably occurred in the marine environment. Several other cherts with odd textures are described in this paper, including (1) a chert breccia cemented by colloform opal-CT and chalcedony, (2) a transition zone between white porcellanite containing opal-CT and quartz and a burrowed brown chert, consisting of radial aggregates of opal-CT with hollow centers, and (3) a chert that consists of silica-replaced calcite pseudospherules interspersed with streaks and circular masses of dense quartz. X-ray-diffraction analyses show that when data from all sites are considered there are poorly defined trends indicating that older cherts have better quartz crystallinity than younger ones, and that opal-CT crystallite size increases and opal-CT cf-spacings decrease with depth of occurrence in the sections. In a general way, depth of burial and the presence of calcite promote the ordering in the opal-CT crystal structure which allows its eventual conversion to quartz. Opal-CT in porcellanites converts to quartz after reaching a minimum d-spacing of 4.07 Å. Quartz/opal-CT ratios and quartz crystallinity vary randomly on a fine scale across four chert beds, but quartz crystallinity increases from the edge to the center of a fifth chert bed; this may indicate maturation of the silica. Twenty-four rocks were analyzed for their major- and minor-element compositions. Many elements in cherts are closely related to major mineral components. The carbonate component is distinguished by high values of CaO, MgO, Mn, Ba, Sr, and (for unknown reasons) Zr. Tuffaceous cherts have high values of K and Al, and commonly Zn, Mo, and Cr. Pure cherts are characterized by high SiO2 and B. High B may be a good indicator of formation of chert in an open marine environment, isolated from volcanic and terrigenous materials.