Carbonate concentration of the analyzed samples ranges between 7 and 90 wt% (Figs. F2, F3). The mean value is ~50 wt%. The postcruise work shows carbonate variations at three spatial scales: ~100 m, >1 m, and <1 m. The <1-m-scale variations are known only from the proxy reflectance data. These carbonate variations are primary; shipboard descriptions also indicate centimeter-scale through millimeter-scale glauconite laminations that formed during diagenesis (Shipboard Scientific Party, 2000).
Organic carbon concentration is generally low (<0.25 wt%) and uniform throughout Hole 1139A (Fig. F3). The mean of organic carbon values is 0.13 wt%. Values range up to 0.35 wt%.
Median grain size diameter ranges between 2.3 and 7.6 µm. Smaller sizes are present in the carbonate-rich samples and vice versa (Fig. F2) (R2 = 0.83)
Opal ranges from 1.5 to 4.1 wt%. The lowest values (~2 wt%) are present in carbonate-rich samples; high values (~4 wt%) are present in carbonate-poor samples (Fig. F2) (R2 = 0.62)
The average values of major and trace elements in the treated sediments (carbonate and opal removed) are compared with the average values of major and trace elements in the volcanic basement of Skiff Bank in Table T8. Cr and Ni are significantly enriched in the sediments. In a future study, the isocon method should be used to compare the Skiff Bank sediments and basement more rigorously.
Shipboard X-ray diffraction (XRD) analyses show chlorite in Core 183-1139A-13R (in the upper carbonate minimum) and a trace of smectite in Core 19R (middle carbonate) (Table T3). A sample from Core 183-1139A-27R (lower carbonate minimum; 52 wt% CaCO3) contains a prominent smectite peak. The five samples that were analyzed following decarbonation contain the same assemblage of igneous minerals (plagioclase and sanidine) as identified during shipboard studies (Shipboard Scientific Party, 2000). They do not, however, appear to contain significant amounts of clay minerals.
Bulk sediment 
18O measurements range between 0
and ~1 (Fig. F2C). They vary considerably in replicate analyses of the same sample (especially low-carbonate samples), and the average values of replicates chosen from a range of carbonate contents are similar (~0.4). These values are approximately 1 lighter than values from benthic foraminifers, consistent with the bulk sediment being dominated by calcareous nannofossils.
Benthic foraminiferal 18O values generally decrease downcore from ~2 in Core 183-1139A-19R to <1 in Section 40R-CC (Fig. F3). In Core 183-1139A-19R, ratios range between 1.5 and 2.0. Values <1 are present in Cores 183-1139A-24R and 25R, followed by an abrupt downcore increase to >1.6.
Benthic foraminiferal 13C values range from as high as 1.75 in Core 183-1139A-19R to as low as 0.38 in Core 36R. They generally decrease downcore, although the value in Section 183-1139A-40R-CC is ~1.
Paleobathymetry at Site 1139 cannot be determined by backtracking because the Kerguelen Plateau has a complex origin and structural evolution (Frey et al., 2000). Benthic foraminifers provide the best tool for reconstructing paleobathymetric history in this setting. Oligocene samples from Hole 1139A yielded too few benthic foraminifers to conduct quantitative analyses of faunal abundance changes. Nonetheless, presence/absence of key depth indicator species yields some paleobathymetric information (M. Katz, pers. comm., 2000).
Species typical of bathyal (200-2000 m) to upper abyssal (2000-3000 m) depths are scattered throughout the Oligocene section recovered from Hole 1139A, including Astrononion pusillum, Bulimina macilenta, Cibicidoides bradyi, Cibicidoides dickersoni, Cibicidoides eocaenus, Laticarinina pauperata, Melonis barleeanum, Nonion havanense, Osangularia mexicana, Planunlina renzi, and Sphaeroidina bulloides. Species present at Site 1139 that were most common at bathyal depths (200-2000 m) in the Oligocene include Planulina costata, Planulina renzi, Rectuvigerina multicostata, Uvigerina mexicana, and Uvigerina spinulosa. Anomalinoides pseudogrosserugosus and Anomalinoides semicribratus are typically found in Oligocene sediments deposited in water depths >600 m and are present in some samples. Bulmina impendens (>500 m; most common at 1000-2000 m) is present throughout, whereas buliminids most common (but not limited to) abyssal depths (>2000 m; B. glomarchallengeri, B. semicostata, and B. grata) are absent. Similarly, the shallowest species of Cibicidoides (C. crebbsi and C. alazanensis) and deepest species of Cibicidoides (C. havanensis and C. lamontdohertyi) are absent. Cibicidoides praemundulus and Laticarinina pauperata are most abundant at >1000 m and are found throughout our samples. Benthic foraminiferal species such as Cibicidoides alazanensis, Neoeponides campester, Rectuvigerina transversa, and Siphonina tenuicarinata that were limited to water depths shallower than 1000 m in the Oligocene are absent from this section. Based on the above depth ranges, we find that the Oligocene section at Site 1139 was deposited at 1000-2000 m depth. Because benthic foraminifers are sparse, it is not possible to speculate on smaller, short-term paleodepth fluctuations based on benthic foraminifers alone.
Benthic foraminiferal assemblages from Site 1139 are similar to those identified for the latest Eocene-Oligocene sections recovered at Kerguelen Plateau Sites 747 and 748 (Schlich, Wise, et al., 1989; Mackensen and Berggren, 1992). Based on benthic foraminifers, Oligocene sediments recovered from Site 747 were deposited at lower bathyal-abyssal depths (>1000 m), and Site 748 was at lower bathyal depths (1000-2000 m) in the Oligocene (Schlich, Wise, et al., 1989).
The chronology for Hole 1139A is based on calcareous nannofossils and magnetostratigraphy (Shipboard Scientific Party, 2000) (Fig. F4). Basement 40Ar/39Ar age is 68-69 Ma (Duncan, 2002).
The overall sedimentation rate is 23 m/m.y. The depth-age plot (Fig. F4) shows that the sedimentation rate peaked at ~29 and ~22 Ma with a local minimum at ~24 Ma.
