Samples of muddy and silty sediments were selected from Units I and II of each Leg 149 hole to compare the fabric within a consistent lithology. Samples were limited to fine-grained sediments, first, because fine sediments are appropriate for obtaining a stable paleomagnetic direction, and second, because the rolling of particles on the surface causes the fabric in coarser parts of a turbidite unit to be perpendicular to the flow direction (Tarling and Hrouda, 1993). We expect that magnetic lineations of fine grains will be parallel to the paleocurrent direction. Samples were collected at Sites 897, 898, 899, and 900. The samples were taken using 12-cm3 plastic cubes. Remanent magnetization of each sample was measured and demagnetized using an automatic fluxgate spinner magnetometer to identify a stable magnetization that was then used to restore the original geographic position. Stepwise alternating-field demagnetization up to 45 mT was performed and the magnetic direction was analyzed using a vector endpoint diagram. Samples that failed to yield a stable magnetization were rejected for paleocurrent analyses, but were used for AMS analyses. Consequently, 223 samples were reoriented using magnetic remanence to dipole direction. Normally magnetized samples were oriented to the north (0°), and reversely magnetized samples were oriented to the south (180°) horizontally. After demagnetization, the AMS of samples was measured using a magnetic susceptibility meter (Kappabridge KY-2) at the University of Hawaii (Table 1). In this paper, I use the following parameters to describe the orientation of the magnetic fabric:
L = Kmax/Kint (Balsey and Buddington, 1960),
F = Kint/Kmin (Stacey et al., 1960),
q = (Kmax - Kint)/[0.5(Kmax + Kint) - Kmin] (Granar, 1958),
where Kmax is the maximum axis, Kint is the intermediate axis, and Kmin is the minimum axis of the susceptibility ellipsoid.