Paleomagnetic work during Leg 178 consisted of long-core measurements of the natural remanent magnetization (NRM) of archive-half core sections before and after alternating field (AF) demagnetization, and magnetic remanence measurements on discrete samples collected from the working half of core sections.

Long-core remanence measurements and AF demagnetization were performed using a long-core cryogenic magnetometer (2-G Model 760-R) with an in-line AF demagnetizer capable of reaching peak fields of 80 mT. Archive halves of all core sections were measured, unless precluded by drilling-related deformation. The number of demagnetization steps applied to each core section was controlled by time constraints. A four-step demagnetization scheme taking ~15 min per section was applied initially and consisted of stepwise demagnetization at 0 (NRM), 10, 20, and 30 mT. This scheme was revised to a three-step demagnetization at 0, 20, and 30 mT to speed core flow through the laboratory. The decision to commence at the 20-mT demagnetization level was justified by the high coercivity of the drill-string overprint. The low maximum peak demagnetization fields ensured that the archive halves remain useful for shore-based high-resolution (U-channel) studies of magnetic remanence. At Site 1095, we used a 4-cm interval for long-core remanence measurements, starting 12 cm above the core section top (leader) and ending 12 cm below the core section base (trailer). At all subsequent sites, we used a 5-cm measurement interval with 15-cm-long headers and trailers. The large leader and trailer distance was used to allow future deconvolution of the long-core data. For the shipboard analysis, we did not use measurements within 8-10 cm of the ends of each section, which are compromised by an edge effect, the false apparent low intensities and inaccurate directions occurring where the wide response function of the SQUID sensors averages empty space with the core signal. Core images were examined to delete data from disturbed or missing intervals. For each site, we produced a set of CD-ROM tables that contain these cleaned data (see "Related Leg Data" contents list). The raw data are available in the JANUS database.

Discrete samples were collected from working halves in standard 8-cm³ plastic cubes with an arrow on the bottom of the sampling box pointing upcore. The sampling frequency was generally two samples per core section at one hole per site. Intervals of drilling-related core deformation were avoided. Discrete samples were collected by two different sampling methods. Very soft intervals were sampled by pushing the plastic cube directly into the sediment. Stiffer intervals were sampled using an extruder. These two sampling methods yield a 180º difference in the orientations of both the +X and +Y axes, relative to the standard up arrow (the -Z direction) drawn on the bottom of the sample box (Fig. F13). The shipboard long-core cryogenic magnetometer was used to measure the NRM of the discrete samples. Samples were measured in a tray designed for seven samples. All discrete measurements were done using the "NORMAL X & Y" option in the cryogenic magnetometer control program, and the samples were placed into the tray such that they had the normal working-half orientation. Magnetic susceptibility was measured on the discrete samples using a Bartington MS2 Susceptibility Meter with a dual frequency M.S.1.B Sensor. The discrete samples will be further analyzed on shore to study the magnetic properties of the sediments and the mineralogy of remanence carriers.

All cores were affected to some extent by a vertically downward overprint attributed to the drill string. Cores with higher clay and water content also had a radial overprint, manifested as declinations of 0º along the entire length of a section. Shallow cores from Hole 1095A required demagnetization at 30 mT to remove the radial overprint.

Where magnetic cleaning successfully isolated the characteristic component of remanence, paleomagnetic inclinations were used to define polarity zones. The revised time scale of Cande and Kent (1995), as presented in Berggren et al. (1995), was used as a reference for the ages of Cenozoic polarity chrons (Fig. F12; Table T1).