radiation generated at 40 kV and 35 mA was used. Peaks were scanned from a 2
of 2° to 32°, with a step size of 0.02°, and a counting time of 2 sec/step.
Before splitting igneous and metamorphic rock cores, the whole cores were examined for structural features. Sediments in contact with the hard rocks were examined for evidence of chilling, baking, and alteration. Contiguous pieces of core were numbered sequentially from the top of each core section and labeled according to standard ODP procedures. Cores were split in such a way as to allow important features and structures to be represented in both the working and archive samples. The archive half was described on the visual core description (VCD) form and was photographed before storage. Only the working half was sampled.
The standard visual core description forms were used to document the location of samples taken from the igneous rock cores (see "Cores" section, this volume) using the following notation: XRD = X-ray diffraction analysis; XRF = X-ray fluorescence analysis; TSB = petrographic thin section. When describing sequences of rocks, the core was subdivided into lithologic units on the basis of changes in texture, grain size, mineral occurrence and abundance, rock composition, and rock clast type. Rocks for which the protolith is completely obscured by metamorphism were given separate lithological names, whereas the prefix "meta" or the term "altered" is used as a modifier with the name of an identifiable protolith. We reserved the termed "altered" for igneous rocks that contain only those secondary minerals consistent with low-temperature seafloor alteration (e.g., calcite, phillipsite, celadonite, nontronite, amorphous iron oxide). When describing the relative proportion of primary and secondary minerals, the general term "alteration" was used to include all secondary minerals. Descriptive information recorded in the database (HARVI) for coarse- and fine-grained rocks:
Basalt is called aphyric (<1%), sparsely phyric (1%-2%), moderately phyric (2%-10%), or highly phyric (>10%), depending upon the proportion of phenocrysts visible with the hand lens and binocular microscope. Basalts are further described by phenocryst type (e.g., a moderately plagioclase-olivine phyric basalt contains 2%-10% phenocrysts, mostly plagioclase, with subordinate olivine). The
abundance of vesicles, their shape, and type of mineral fillings were also noted. More specific rock names were given where chemical analyses or thin sections were available.
A breccia is defined as any rock composed of angular broken rock fragments held together by finer fragments or glassy material. They form in many ways including volcanic, hydraulic, tectonic, and impact. Volcanic breccias form as accumulated volcanic gases expand suddenly producing a chaotic array of differently sized angular volcanic fragments which may be welded and altered at high temperature. Hydraulic breccias form as accumulated water vapor expands suddenly to produce a chaotic array of different sized angular fragments, perhaps welded and altered at low temperature. Tectonic breccias form along fault or shear zones during displacement, producing angular fragments that moved along the fault zone. The explosive volcanic and hydraulic breccias would typically have random chaotic collections of angular fragments, whereas tectonic breccias would typically have irregular fragments concentrated along two-dimensional planar fault surfaces. A large range of fragment sizes would be expected in explosive breccias relative to fault breccia.
The texture of plutonic igneous rocks was described in terms ofcrystallinity, granularity (the absolute and the relative sizes of crystals), crystal shapes, and crystal arrangement. Rock names were assigned according to essential primary minerals using a triangular diagram and based on the primary minerals being either plagioclase, olivine, orthopyroxene, or clinopyroxene (Fig. 9). Genetically significant accessary minerals were used to modify the rock name further (e.g., spinel or plagioclase lherzolite). The ranges of grain size in the plutonic rocks were recorded using the terms: very coarse-grained (crystal diameters >30 mm), coarse-grained (crystal diameters of 5-30 mm), medium-grained (crystal diameters of 1-5 mm), and fine-grained (crystal diameters of <1 mm).
Oxide and sulfide minerals were classified on the basis of type and texture and, where possible, identified as of primary or secondary origin. The percentage of these minerals was visually estimated, and individual minerals were typically undifferentiated in hand specimen. Iron oxide staining was recorded on partially altered minerals and along fractures. The texture of the oxide and sulfide minerals was described in terms of the size, shape, and orientation of the mineral and its relationship with adjacent minerals. Their form was generally described as euhedral, subhedral, or anhedral.
Textural and mineralogical variations were recorded, noting the interval over which a particular feature was observed, and its relation to other features. Modally graded zones were defined as those where the primary mineralogy showed gradational changes in grain size. Texturally graded zones were defined as those where a gradational change in texture was observed (e.g., from poikilitic to granular). The direction of grading was related to stratigraphic position. Crystal alignment of probable igneous origin was described as planar, curved, or irregular. The orientation of any planar structure was recorded by measuring apparent dips relative to the core axis.
Alteration, veining, and metamorphism were recorded to provide three types of data: (1) the extent of igneous mineral replacement by metamorphic or secondary minerals; (2) the extent to which metamorphic minerals contribute to subsolidus fabric; (3) the abundance, character, and orientation of veins, along with their associated haloes; and (4) the thickness and orientation of shear zones. Overprinting relationships between secondary minerals, particularly cross-cutting relationships of veins and shear zones, were reported to document the relative timing of these events. Sheared rocks were described as mylonites, cataclasites, or porphyroclasites, depending on the extent to which crushing, recrystallization, and fabric development had modified the primary igneous rock. Identification of vein-filling material was frequently checked using XRD.
Individual vein types were identified by color and mineralogy. Data on abundance, width (mm), orientation, and texture and abundance of vein-filling minerals were recorded for each piece containing one or more veins. In pieces having numerous veins, sequences, and patterns produced by intersecting veins and sequences of changing mineralization in the veins were recorded and described.
Thin sections of igneous rocks were examined to complement and refine hand-specimen observations. The same terminology was used for thin-section and visual-core descriptions. The percentages and textural descriptions of individual phases were recorded using a computerized database (HRTHIN). Thin-section descriptions are included in the "Cores" section (this volume) as separate tables for each core.
A Philips ADP 3520 X-ray diffractometer was used for the X-ray diffraction (XRD) analysis of mineral phases. Ni-filtered CuK radiation generated at 40 kV and 35 mA was used. Peaks were scanned from a 2 of 2° to 32°, with a step size of 0.02°, and a counting time of 2 sec/step.
Samples were ground to <200 m mesh in a Spex 8000 Mixer Mill using tungsten carbide and steel. The powder then was pressed into aluminum sample holders or smeared onto glass plates for analysis. Diffractograms were interpreted with the help of a computerized search and match routine using the Joint Committee on Powder Diffraction Standards powder files.
Before analysis, samples were crushed in a Spex 8510 shatterbox using a tungsten carbide barrel. Where recovery permitted, at least 20 cm3 of material was ground to ensure a representative sample. The tungsten carbide barrel was used despite the considerable W contamination and minor Ta, Co, and Nb contamination, which makes the powder unsuitable for later instrumental neutron activation analysis (INAA).
A fully automated wavelength-dispersive ARL8420 XRF (3 kW) system, equipped with an Rh-target X-ray tube, was used to determine the abundances of major and trace elements in whole-rock samples (Table 4). Analyses of the major oxides were performed on lithium borate glass disks doped with lanthanum as a "heavy absorber" (Norrish and Hutton, 1969). The disks were prepared from 500mg of rock powder that had been ignited for 2 hr at about 1030°C and mixed with 6.000 g dry flux (pre-weighed onshore) consisting of 80% lithium tetraborate and 20% La2O3. This mixture then was melted in air at 1150°C in a Pt-Au crucible for about 10 min and poured into a Pt-Au mold using a Claisse fluxer. The 12:1 flux-to-sample ratio and the use of the lanthanum absorber made matrix effects insignificant over the normal range of igneous rock compositions. Hence, the relationship between X-ray intensity and concentration becomes linear and can be described by:
where Ci = concentration of element i (wt%); li = peak X-ray intensity of element i; mi = slope of calibration curve for element i (wt%/counts/second); and bi = apparent background concentration for element i (wt%).
The slope mi was calculated from a calibration curve derived from the measurement of well-analyzed reference rocks, such as BEN, BR, and DRN from Geostandards, France; BHVO-1, RGM-1, and AGV-1 from the U.S. Geological Survey; JGB-1 and JP-1 from the Geological Survey of Japan; AII-92-29-1 from Woods Hole Oceanographic Institution/Massachusetts Institute of Technology; and K1919 from Lamont-Doherty Earth Observatory. Two other standard reference rocks were run with samples as unknowns. A complete list of analyses for the standards used to derive calibration curves is given in Table 5 (major elements) and Table 6 (trace elements). The background bi was determined by regression analysis from the calibration curves.
Systematic errors resulting from short- or long-term fluctuations in X-ray tube intensity and instrument temperature were addressed by counting an internal standard between no more than six unknowns in any given run. The intensities of this standard were normalized to its known values, thereby providing correction factors for the measured intensities of the unknowns. To reduce shipboard weighing errors, two glass disks were prepared for each sample. Accurate weighing was difficult on board the moving platform of the JOIDES Resolution, and was performed with particular care as weighing errors could be a major source of imprecision in the final analysis. Five weight-measurements that agreed to within 0.5 mg (each sample, and the average was used as the true weight. Loss on ignition was determined by drying the sample at 110°C for 8 hr and then by weighing before and after ignition at 1030°C in air.
Trace elements were determined on pressed-powder pellets prepared by pressing (with 7 MPa of pressure) a mixture of 5.0 g of dry rock powder (dried at 110°C for >2 hr) and 30 drops of polyvinyl alcohol binder into an aluminum cap. A modified Compton scattering technique, based on the intensity of the Rh Compton peak, was used for matrix absorption corrections (Reynolds, 1967).
Replicate analyses of rock standards showed that the major-element data are precise to within 0.5% to 2.5% and are considered accurate to approximately 1 % for Si, Ti, Fe, Ca, and K, and between 3% and 5% for Al, Mn, Na, and P. The trace-element data are considered accurate to between 2% and 3% or 1 ppm (whichever is greater) for Rb, Sr, Y, and Zr, and between 5% and 10% or 1 ppm (whichever is greater) for the other elements. The accuracy of Ba and Ce is considerably less, and they are reported primarily for purposes of internal comparison. Precision is within 3% for Ni, Cr, and V at concentrations >100 ppm, but 10% to 25% at concentrations <100 ppm.
