Paleomagnetic analyses were performed on the archive sections of cores recovered at Site 1195 using the three-axis pass-through cryogenic magnetometer. The core sections were first measured for natural remanent magnetization (NRM), and then were subjected to 5-mT and 30-mT field demagnetizations in the alternating-field coils of the cryogenic magnetometer to measure their successive magnetization. The maximum destructive field used was 30 mT.
In order to reduce the magnetic overprint from drilling, the APC cores from Hole 1195B were collected with a nonmagnetic Russian-made PDC bit, whereas a standard C3RBI bit was used for Hole 1195A. Comparing the APC records from Hole 1195A and 1195B (Figs. F11, F12A), we can see the downward overprinting effect of the drill bits on the paleomagnetic record. As seen at the previous sites, magnetostratigraphic data in Hole 1195A are all shifted toward positive inclinations, an indication of the downward overprint of the drilling process (Fig. F11). This effect can result in misinterpreting a normal polarity event to be a reversed one. Results from Hole 1195B (Fig. F12A) have no downward magnetization overprint, and most of the values are only slightly offset toward the negative side of the axis. Surprisingly, some of the normal polarity intervals seem to be unresolved, possibly because of an uncleaned modern-day magnetic field overprint, which, in fact, should not affect the reversal record at Site 1195.
The NRM intensity generally varies between 10-3 and 10-2 A/m with an average intensity of 10-2.5 A/m for the top 100 mbsf (Figs. F11, F12). Average intensity decreases to 10-3.5 A/m between 100 and 400 mbsf and slightly increases to a mean value of 10-3 A/m below 400 mbsf. These intensity variations appear to be a function of the quantity of magnetic mineral input rather than core-top anomalies, as observed in the previous sites. When all of the core sections are demagnetized to 30 mT, which is the minimum considered necessary to remove the effects of viscous remanent magnetization (VRM), the intensity drops approximately one order of magnitude throughout the section. The low intensity levels suggest the dominance of diamagnetic and paramagnetic minerals, whereas high-intensity levels indicate the presence of strongly magnetic materials. We have identified ten levels of high intensity; four are found within the first 80 mbsf and occur in Holes 1195A and 1195B (Figs. F11, F12). The equivalent depths of these high intensity zones in Hole 1195B are systematically 6 m shallower than in Hole 1195A. Most high-intensity levels appear to correlate well with glauconite-rich layers (see "Lithostratigraphy and Sedimentology"). Comparison of results with sedimentological descriptions indicate that other minor intensity increases are also linked with glauconite occurrences. Glauconite is a family of sheet silicate (mica) minerals resembling muscovite with high proportions of paramagnetic ions such as Fe3+ and Fe2+ in its two structural sites. Whether the glauconite is authigenic or transported, it suggests higher concentration of magnetic minerals like magnetite and hematite. However, other sections further downcore do not show a one-to-one relation between glauconite occurrence and magnetic susceptibility (see "Core Physical Properties" and "Downhole Measurements").
Although there are indications of VRM, the magnetization measured at 30 mT represents predominantly the primary NRM. Results for the top 100 mbsf (late Miocene to Pleistocene) in Holes 1195A and 1195B (Figs. F11, F12A) record a sequence of magnetic polarity reversals. The correlation of the observed pattern of normal and reversed polarity interval with the geomagnetic polarity timescale (GPTS) was assigned based on the preliminary age estimates from calcareous nannofossils. In this correlation, the prominent polarity intervals are easily matched, but some short polarity intervals are masked and cannot be unambiguously identified. Some of the observed magnetostratigraphic intervals show a shorter duration than expected, indicating temporarily reduced or interrupted sedimentation. These intervals are sometimes found to coincide with the location of anomalous intensity increases, lending support to the presence of a hiatus. Between 100 and 200 mbsf (Fig. F12B), the prominent late Miocene chrons have been identified. Between 200 and 300 mbsf (Fig. F12C), low recovery hindered correlation with the GPTS. Despite this difficulty, the prominent Chrons C5n, C5r, C5An, and C5Ar have been assigned to the observed polarity intervals. Below 300 mbsf (Fig. F12D, F12E), in the lower Miocene section, coring gaps preclude magnetostratigraphic interpretation. Nevertheless, an attempt was made to identify some magnetic polarity zones (Fig. F12E) based on the distinctive C6n Chron (Table T6). The five short normal intervals observed between 500 and 530 mbsf are tentatively assigned to the C6Ar, C6Aan, and C6Aar (early Miocene) (Fig. F12F). The overall magnetostratigraphic interpretation is shown in Figure F13. The resulting age-depth curve is plotted in Figure F14 (see "Age Model").