= 21), indicating that only part of the overprint was successfully removed. The magnetic declinations of the oriented cores are also quite scattered (Fig. F7) and mostly directed toward the south.
All archive-half core sections from Holes 1215A and 1215B, except those where the sediment was clearly disturbed, were measured on the shipboard pass-through cryogenic magnetometer. A total of 83 core sections were measured from 15 of the 21 cores recovered in the two holes. The natural remanent magnetization (NRM) was measured at 5-cm intervals in each core section, followed by four-five steps of alternating-field (AF) demagnetization up to a maximum peak field of 20 mT. The maximum peak field was set at this level to avoid compromising the archive sections for possible shore-based (U-channel or discrete sample) studies. In addition, 34 discrete samples were taken to conduct more detailed progressive demagnetization. Several of the measured cores from the middle and bottom parts of the holes are in poor condition, primarily because of drilling disturbance. This deformation has the potential to render the pass-through magnetometer data useless as a result of the averaging of divergent magnetic vectors when drilling slurry is in the instrument's sensing region. Most of the cores, however, displayed consistent remanent magnetization directions. NRM magnetization intensities were in the order of 10-1 to 10-2 A/m and decreased to ~10-3 to 10-2 A/m after partial AF demagnetization (Fig. F6). A large group of NRM inclinations showed steep downward directions (~70°), indicative of a drilling-induced overprint. This overprint was effectively removed with AF demagnetization, typically disappearing by the 10- to 15-mT demagnetization step. Most magnetic directions did not reach a stable point between 5 and 20 mT, suggesting that the characteristic remanent magnetization (ChRM) has been only partially isolated. As a result, more detailed shore-based measurements will be necessary to obtain reliable magnetization directions.
The Tensor tool was used to orient cores starting from Cores 199-1215A-4H and 199-1215B-3H. The orientation was successful in aligning the declination between most cores. Figure F7 illustrates the distribution of declination prior to orientation and the improved grouping after applying the Tensor correction. The Tensor tool reorientation of declinations was very useful in assessing polarity changes in the Eocene unit because the magnetic inclination is very shallow and cannot be used by itself to establish polarity changes.
Oriented discrete samples (8-cm3 cubic plastic boxes) were collected from every undisturbed section from Hole 1215A. These samples were subjected to stepwise AF demagnetization up to 100 mT to assess the magnetic stability of the sediments and estimate the demagnetization step suitable for removing the secondary magnetization (Fig. F8). Many samples show a strong and steep magnetic overprint, though an AF demagnetization of 20 mT is typically sufficient to determine the polarity.
In the measured archive halves of Hole 1215A, the 20-mT AF demagnetization directions have inclination values clustering around 29° with a large scatter ( = 21), indicating that only part of the overprint was successfully removed. The magnetic declinations of the oriented cores are also quite scattered (Fig. F7) and mostly directed toward the south.
Paleomagnetic data acquired from several sections of Cores 199-1215A-8H and 199-1215B-4H were discarded because of excessive drilling disturbance, in part related to the presence of chert nodules (see "Unit II" in "Lithostratigraphy"). In many places, it was noted that identifiable bedding features in these sediments were highly deformed, having been pushed downward at the edges of the core. If these cores are to produce a reliable magnetic stratigraphy, it must come from postcruise analyses of discrete samples taken in the middle part of the core.
In those parts of the record where the remanence directions are not noisy, it is possible to interpret the inclination and declination, after 20-mT AF cleaning, in terms of polarity zones (Fig. F9). The inclination in the uppermost 8 m of Hole 1215A can be interpreted as a record of C1n (Brunhes Chron), C1n.r1 (Jaramillo Subchron), C2n (Olduvai Subchron), and the top of C2An (Gauss Chron). This is the only part of lithologic Unit I where a chronology of the sedimentary record can be obtained. As cores were oriented below 20 mbsf in Hole 1215A, the magnetic polarity was determined using the virtual geomagnetic pole latitude, which combines the information from both inclination and declination. From the base of Hole 1215A, we identify Chron C25n followed by Chron C24r, which contains the P/E boundary. At 35.5 mbsf a normal chron is tentatively interpreted as C24n. The normal chron between 29 and 25 mbsf is then interpreted as C23n. Alternatively, the interval between 35.5 and 25 mbsf could span through Chron C24n because it comprises two short intervals of reversed magnetization (Cande and Kent, 1995).
Hole 1215B has a more scattered record than Hole 1215A because of the presence of disturbed sediment, which hampers the unambiguous correlation with the geomagnetic polarity timescale (GPTS). At the base of Hole 1215B ~53 mbsf, we identify a positive chron that is also observed in Hole 1215A. Because there appears to be no correlation of this chron to the GPTS, it might represent one of several cryptochrons in Chron C24r (Cande and Kent, 1995). We interpret the reversed polarity interval between 49 and 39 mbsf as Chron C24r (also in good agreement with Hole 1215A). Core 199-1215B-4H is completely disturbed (see "Unit II" in "Lithostratigraphy") and hence not usable for paleomagnetic analysis. Five geomagnetic field reversals were found between 19 and 31 mbsf and are tentatively interpreted as Chrons C22n, C22r, C23n, C23r, and part of C24n. Similarly to Hole 1215A, we identify C1n (Brunhes Chron), C2n (Olduvai Subchron), and C2An (Gauss Chron) based only on the inclination record. From 8 mbsf to the top of lithologic Unit II (nannofossil ooze), the record cannot be interpreted in terms of polarity chrons.
ChRM directions for the AF-demagnetized discrete samples show a mean inclination of 21.8° and -45° for normal and reversed directions, respectively. The corresponding paleolatitudes are 11.3° and 26.5°, respectively. The reliability of these results, however, is suspect owing to the large scatter of paleomagnetic directions. ChRM inclinations calculated from the continuous measurements on oriented cores produce a mean inclination of 25.6°, which is similar to the mean inclination computed from discrete samples with a normal magnetization (21.8°). These preliminary results need to be followed up by shore-based thermal demagnetization to fully isolate the primary magnetization direction. However, our initial results suggest a mean paleolatitude of 12.3°N, very close to the value implied by the 57-Ma paleopole for the Pacific plate (~15°N; Petronotis et al., 1994).
