We squeezed four whole-round core samples for interstitial water from depths between 280 and 380 mbsf at Site 1097 (Table T5). Poor core recovery and the presence of large rocks in most depth intervals limited the availability of suitable sample material. Furthermore, one of the four samples (Sample 178-1097A-42R-2, 114-124 cm) yielded only 10 mL of interstitial water; another (Sample 178-1097A-45R-1, 140-150 cm) yielded only 5 mL because of the coarse-grained, semilithified nature of the recovered sediment. We consequently could not perform a complete suite of analyses for all samples, and we provide here a necessarily limited interpretation of the results.

In general, the interstitial water chemistry (Fig. F21) fits the pattern expected for terrigenous sediments buried to a few hundred meters depth with relatively small amounts of biogenic components. Low concentrations of sulfate, manganese, and iron indicate that these sediments lie below the sulfate reduction zone. Maximum alkalinity and ammonium concentrations (7.4 and 1.9 mM, respectively) approach those observed at comparable depths in the rise sediment (Sites 1095 and 1096) and confirm that decaying organic matter influences the shelf sediment. Phosphate concentrations remain quite low (<3 µm), however, in the sampled depth range, and fluoride also reaches low concentrations (<20 µm) relative to seawater (70 µm). These factors suggest that fluoride uptake, perhaps through apatite precipitation, occurs somewhere in the upper sediment column. The range of dissolved silica concentrations (0.64-0.84 mM) exceeds that of normal seawater but remains below the solubility limit of opal-A (~1.1 µM; Kastner et al., 1977). Thus, although some silica dissolution has occurred, the shelf sediment apparently does not contain an excess of opal. Also, calcium concentrations deviate only slightly from seawater, perhaps not a surprise given the nearly carbonate-free composition of the shelf sediment (see "Organic Geochemistry"). Magnesium, potassium, and strontium concentrations lie significantly below seawater values and indicate that the shelf sediment acts as a sink for these elements, probably through clay mineral reactions.

Five samples were analyzed for their bulk and clay mineralogy by X-ray diffraction. The two shallowest samples (from 73 and 132 mbsf) represent glacial till, and the three deeper samples (from 218, 295, and 362 mbsf) represent glacial marine sediment that contains varying amounts of ice-rafted debris (see "Lithostratigraphy"). These shelf deposits contain the same mineral assemblage as that identified on the nearby continental rise (Sites 1095 and 1096): primarily quartz, feldspars, and a blend of clays consisting of chlorite, illite, and a mixed-layer clay as well as traces of amphibole. However, the shelf sediment contains significantly more feldspar and less clay relative to quartz. The average intensity of the 3.19-Å plagioclase peak equals 66% of the 3.34-Å quartz peak (Table T6), compared to only 40% for Sites 1095, 1096, and 1101 (see Table T32, in the "Site 1095" chapter; Table T26,  in the "Site 1096" chapter; and Table T2, in the "Site 1101" chapter), whereas the 7-Å chlorite peak averages only 9% of the quartz peak at Site 1097, compared to 12% at the rise sites. Also, average chlorite/illite ratios are highest in the shelf sediments (Table T7; Fig. F22 ). The one exception to these observations is the glaciomarine mud from 218 mbsf, which has feldspar and clay contents similar to those observed in the rise sediments. The richer feldspar and poorer clay content of the shelf sediment may reflect factors such as larger average grain size and closer proximity to sediment sources. Alternatively, some of these differences may reflect a different provenance for a part of the rise sediment.

Trace-element concentrations were measured by X-ray fluorescence on splits of all five samples that were analyzed by X-ray diffraction (Table T8). These samples have similar trace-element compositions, although the sample from 218 mbsf generally has higher concentrations of most elements, perhaps because of its higher clay content. The concentrations of some elements in the shelf sediment differ from those observed in the rise sediment (Sites 1095 and 1096). Barium concentrations reach as low as any measured at the rise sites and presumably reflect a purely detrital signal, with little or no Ba deposited as a result of high biological productivity on the shelf. Most other trace elements, particularly Rb and Cr, have lower concentrations than were measured in most samples from Sites 1095 and 1096. Like the XRD data (see "X-Ray Diffraction Mineralogy"), these results suggest a different provenance for part of the rise sediment.