Interstitial water samples for Leg 178 were obtained from 5- to 10-cm, whole-round core samples cut immediately after the core arrived on deck. Specific sampling strategies varied among sites, as described in each site chapter, with allowance for the recovery of suitable core material. In general, we took at least one sample per core within the upper 50 mbsf and one every third core thereafter, typically from the bottom of the third section, adjacent to the headspace sample taken from the top of the fourth section. At Site 1098, however, we took one sample per section for a high-resolution, shore-based study of carbon and oxygen isotopes.

Interstitial water was collected using titanium squeezers modified after the standard ODP stainless-steel squeezer of Manheim and Sayles (1974). Each whole-round sample was carefully scraped using a stainless-steel spatula to remove the outer rind, squeezed through two Whatman No. 1 filters pre-rinsed in high-purity water, and then passed through a 0.45-µm Gelman polysulfone disposable filter into a 50-mL plastic syringe. Interstitial water was extruded by applying pressures up to 40,000 lb using a Carver Laboratory Press (Model 2702). After collection of 40-50 mL of interstitial water, the syringe was removed, a fresh 0.45-µm Gelman filter was attached, and aliquots were dispensed into plastic vials for shipboard analyses and into acid-washed plastic vials and 5-mL glass ampules (heat sealed) for future shore-based work.

Interstitial water was routinely analyzed for salinity as total dissolved solids with a Goldberg optical handheld refractometer (Reichart). Alkalinity and pH were measured immediately after squeezing by Gran titration using a Metrohm autotitrator and Brinkmann pH electrode, respectively. Chloride was measured by manual titration with AgNO₃. Calcium, Mg, K, and SO₄ were measured to within 3% and 5% on 1/200 diluted aliquots in nanopure water using a Dionex DX-100 ion chromatograph. In general, this instrument yields less accurate results than other methods; however, the relative trends are usually similar and can serve as a second check of results generated by the other methods. Silica, ammonium, phosphate, and fluoride were determined by colorimetric methods using a Milton Roy Spectronic spectrophotometer with a 1-cm cell and sample introduction by Mister Sipper. The chemical methods employed follow those of Gieskes et al. (1991) and Grasshoff et al. (1983). For most of these analyses, the International Association of Physical Sciences Organizations seawater was used for standardization.

Strontium, iron, and manganese were determined using a Varian Spectra AA-20 atomic absorption spectrophotometer. Samples for Fe and Mn were acidified immediately after collection with concentrated HCl (50 µL/5 mL) and analyzed using an air-acetylene flame, whereas Sr was analyzed using a N₂O-acetylene flame. Standards were matched in matrix composition to the samples, and lanthanum chloride was used as an ionization suppressant for Sr and Mn analyses. Manganese was determined on 1/5 diluted aliquots, Sr was determined on 1/10 diluted aliquots, and Fe was determined without dilution. The precision of these techniques is ~<4%, except for Fe and Mn concentrations <10 µM, for which the precision can be significantly lower.

The Fe data should be interpreted with caution for two reasons: (1) it was not possible to squeeze sediment samples under oxygen-free conditions and maintain timely processing of samples through the chemistry laboratory; thus, some dissolved Fe⁺² may have been oxidized, trapped on filters, or adsorbed on the walls of the syringe, and (2) some colloidal iron oxides can pass through 0.45-µm filters. The first effect would cause Fe⁺² concentrations to be underestimated, whereas the second could result in an overestimate. Despite these potential problems, the Fe data reported here probably represent at least qualitatively the true in situ Fe concentrations.

X-ray fluorescence (XRF) analyses of trace elements (Nb, Zr, Y, Sr, Rb, Zn, Cu, Ni, Cr, V, and Ba) were performed on sediments from Sites 1095, 1096, and 1097 in an attempt to detect a change in provenance of detrital sediments related to uplift and unroofing of the Antarctic Peninsula or to glacial-interglacial climate variability. Samples of ~10 g were freeze-dried, ground gently in an agate mortar, washed with 150 mL of nanopure water to remove excess salts, sonicated for 30-40 min, and centrifuged. This washing probably also removed weakly adsorbed ions on sediment grain surfaces. Some discoloration of the supernatant solution was observed in most samples, probably reflecting loss of some fraction of colloidal phases, and possibly traces of the finest clays; however, this should not significantly affect the bulk composition of the primary terrigenous phases. Centrifuged sediments were dried for at least 12 hr at 120ºC and powdered in a tungsten carbide mixer mill. A ~1-g aliquot was ignited for 10 hr at 950ºC to determine loss on ignition and to oxidize all iron to the trivalent state. This sample was retained for postcruise major-element analysis. Pressed pellets for trace-element analyses were made from ~6.5 g of dried powder mixed with ~50 drops of Chemplex liquid binder and pressed to ~6 t pressure.

Trace-element abundances were determined using an automated ARL 8420 wavelength-dispersive spectrograph, equipped with an end-window, Rh-target X-ray tube and operated under the conditions outlined in Table T5. The actual analyses were performed on board ship during the early part of Leg 179. Trace-element analyses were corrected for nonlinear backgrounds, spectral interferences, and matrix absorption effects using the ARL software package and a range of geologic standards. Precision and accuracy were assessed by multiple analysis of Standard Reference Material MAG-1 (Table T6), a marine mud prepared by the U.S. Geological Survey (Govindaraju, 1989).

X-ray diffraction (XRD) analyses were performed using a Philips APD 3720 diffractometer, operated by Philips software PC-APC, version 3.6. The diffractometer was operated using Cu- radiation at 35 mA and 40 keV, a focusing graphite monochromator, and a 12.5-mm theta compensating slit. Bulk mineralogy by XRD was determined on a split of the final powder from each XRF sample and on other freeze-dried samples that were powdered in an agate mortar, without washing. These samples were scanned from 2º to 70º 2. Clay mineralogy was examined by XRD on separate 3-g samples that were freeze-dried, placed in a 50-mL centrifuge tube with distilled water, and sonicated for 15 min. After flocculation, samples were centrifuged, and washing was repeated until the clays dispersed. Some samples required treatment with a buffered solution (pH 5) of sodium acetate and acetic acid to remove carbonates. Sodium pyrophosphate was added to reach a 1.5% solution. These samples were sonicated for 15 min and centrifuged at 1000 for 5 min; a sample was taken off the top 1 cm of the suspension. This clay-sized fraction was collected by vacuum filtration on a 0.45-µm Millipore filter. The clay was transferred to a glass slide by placing the filter on the slide and gently rolling the back of the filter with a small roller. Two slides were prepared for each sample: one slide was air dried and scanned, heated to 550ºC to collapse kaolinite and smectite, and scanned again. The second slide was solvated with ethylene glycol for at least 8 hr at 60ºC to determine the presence of expandable clays. Clay samples were scanned from 2º to 35º 2. Mineral identification followed standard procedures as outlined in Brown (1980), Brown and Brindley (1980), and Moore and Reynolds (1989).