All cores from Holes 1128B, 1128C, and 1128D with sufficient recovery to make long-core measurements feasible were measured as half cores using the 2-G 760-R magnetometer. Measurements were made of natural remanent magnetization (NRM) and after 20-mT demagnetization. An experimental nonmagnetic APC assembly was used to obtain alternate cores in Hole 1128C from 3H to 13H. The nonmagnetic corer produced a significant reduction in the acquisition of a "radial" coring-induced contamination (see "Appendix: Magnetics Experiment"). Discrete samples for NRM and rock magnetic analysis were taken from both soft sediments cored by APC and from biscuits in the XCB and RCB cores. The analyses were performed using the methods discussed in "Paleomagnetism" in the "Explanatory Notes" chapter.
Long-core measurement provided two almost continuous records from Holes 1128B and 1128C for the uppermost 250 mbsf. Neogene sediments have relatively high remanence intensities, averaging ~6 mA/m. Notably lower intensities, ranging from 8 × 10-4 to 2 × 10-5 A/m, are present in the Oligocene and Eocene pelagic clays that predominate below 140 mbsf. Although the quality of the record varied from excellent in the vicinity of ~100 mbsf to fair in deeper cores, a reliable record of inclination was obtained. Variable recovery from Hole 1128D made a continuous record impossible, although clearly reversed and normal inclinations were evident.
Analysis of NRM by standard Kirschvink-Zijderveld techniques yielded satisfactory results from both soft sediments and lithified material, as illustrated in Figure F13. In the uppermost Figure F13A, results from a soft sediment sample show that a soft upward moment demagnetizes to yield a progressively steeper downward (reverse polarity) magnetization, with a maximum angular deviation angle of 1.6. The second sample is from an RCB biscuit that has acquired a stronger soft upward magnetization and again gives way on demagnetization to a reversed magnetization.
Rock magnetic analysis revealed subtle differences in magnetic properties of APC cores. This is illustrated in Figure F14, in which results from Cores 182-1128B-3H-2 and 13H-2 are compared. In the nannofossil oozes that predominate in the upper cores, isothermal remanent magnetization (IRM) acquisition is consistent with fine-grained magnetite, as are the demagnetization characteristics. The ratio of anhysteretic remanent magnetization (ARM) to IRM suggests single-domain grain size. Deeper in the section, in the calcareous clays of Section 13H-2, the alternating field decay of the natural and laboratory-induced magnetizations suggest that the bulk of the remanence is carried by a phase of relatively low coercivity such as magnetite, although inductions of 300 mT are not sufficient to reach saturation. The ratios of ARM:IRM for the calcareous clays are lower than in the shallow samples. These ratios and the lower coercivity of the NRM would imply small multidomain grain sizes (in excess of several micrometers) rather than the submicrometer size of bacterial magnetite. The higher coercivities in the IRM in the calcareous clay (Section 1H-2) may indicate that small amounts of hematite are present.
The magnetic susceptibility data obtained from the multisensor track (MST) for Hole 1128B are shown in Figure F15. Above the debrites in lithostratigraphic Unit I, there is a trend of decreasing susceptibility with depth, which is interrupted by short wavelength peaks and a strong peak centered at 40 mbsf. This peak is associated with a high in the natural gamma log and a color change from lighter to darker green, although the source of the high susceptibility values has not been established. The shorter wavelength troughs reflect the contrast between the low-susceptibility packstones and turbidites, compared with the nannofossil ooze, whereas wackestones are associated with higher susceptibility peaks. The remainder of lithostratigraphic Unit I, including the debrites, have intermediate susceptibilities between 20 and 40 × 10-5 SI. Within the calcareous clays of lithostratigraphic Unit II, the susceptibility falls from values of 40 × 10-5 SI to ~5 × 10-5 SI at 240 mbsf. Frequency dependence of susceptibility is low to moderate (2%-10%), except for samples in the deepest cores of Hole 1128D, in which values reach 17%, indicating a higher concentration of superparamagnetic magnetite. Within lithostratigraphic Unit II, higher susceptibility values are observed in the interval from 70 to 130 mbsf than in the interval from 140 to 240 mbsf. Susceptibility is higher by a factor of ~5, coinciding with lower sedimentation rates. At 250 mbsf a minor peak appears to correlate with the occurrence of siltstones.
The two records for Holes 1128B and 1128C are similar and permit tentative correlation with the geomagnetic polarity time scale in the Pliocene-Pleistocene to a depth of 54 mbsf, at which point the section is dominated by gravity-flow deposits, and no meaningful record can be obtained. This Pliocene-Pleistocene section is illustrated in Figure F16 from Holes 1128B and 1128C. In Hole 1128C the long wavelength pattern of normal from 0 to ~15 mbsf, reversed from 15 to ~40 mbsf, followed by a return to dominantly normal polarity, appears to be a manifestation of the Brunhes-Matuyama-Gauss sequence. The Jaramillo and Olduvai normal zones in the Matuyama are tentatively interpreted to be at 16 and 26-27 mbsf. However, the occurrence of other brief intervals of normal polarity preclude indisputable identification. In the normal interval interpreted as the end of the Gauss, the Kaena appears at 46 mbsf. The record in Hole 1128B is less clear, but the Brunhes/Matuyama boundary is evident at 14 mbsf. The Jaramillo Subchron is possibly between 20 and 22 mbsf, and the Olduvai Subchron is possibly between 28 and 32 mbsf.
Below 70 mbsf a sequence of reversals is clearly evident in the pelagic clays (Fig. F17). A long period of reversed polarity occurs from ~138 to 213 mbsf and has been correlated with Chron C12r, the longest reversed chron in the time frame indicated by biostratigraphic data. The well-defined short reverse interval centered at 82 mbsf, bounded by two intermediate-length normal chrons, is similar to the reversal sequence of Chron C10n. The correlation suggested in Figure F17 implies that a moderate change in sedimentation rate occurs toward the end of Chron C12r, with a faster sedimentation rate in the lowermost Oligocene. It also suggests that the Eocene/Oligocene boundary lies at a depth of ~250 mbsf, although it cannot be better defined because of poor recovery in this critical interval.
Recovery was limited in Hole 1128D, and thus paleomagnetic results are difficult to interpret. However, the results are strongly biased toward normal directions from ~280 mbsf to the bottom of the core at 450 mbsf, which may correspond to the dominantly normal polarity sequence of the upper and upper middle Eocene.
In summary, the magnetostratigraphy above the debrites is interpreted as recording C1n to C2An in the Pliocene-Pleistocene and below the debrites we interpret C9r to C13r from mid-Oligocene to the Eocene/Oligocene boundary. In the RCB cores, the long predominantly normal polarity is tentatively interpreted as a record of the predominantly normal upper and middle Eocene section. However, it must be recognized that this is an equivocal magnetostratigraphic interpretation.