PALEOMAGNETISM

Hole 1274A penetrated 155.8 m of peridotite and dunite and crossed a fault zone at ~110 mbsf. The mean recovery was 22.2%, with the highest recovery (>50%) between Cores 209-1274A-5R and 7R. Remanence measurements were made on a total of 135 archive-half pieces (35 of these were >20 cm in length) and 18 discrete samples. Both types of data reveal stable positive inclinations that are presumably of normal polarity, although the mean inclination (13°) is significantly shallower than the expected dipole inclination at the site (28°).

The relatively high recovery from Hole 1274A allows recognition of some depth trends in the magnetic properties (Fig. F61). The natural remanent magnetization (NRM) intensity increases from ~0.5 A/m in the uppermost cores to ~5 A/m in the lower half of lithologic Unit I (see "Igneous and Mantle Petrology"). NRM intensities of a few amps per meter also characterize the peridotites of lithologic Unit III, whereas the poorly consolidated fault gouge of Unit II (Cores 209-1274A-23R and 24R) have lower intensities. These general trends in NRM intensity are paralleled by variations in the magnetic susceptibility.

The archive halves from Hole 1274A were subjected to stepwise alternating-field (AF) demagnetization, typically to a peak field of 50 mT. Most intervals have a low-stability magnetization component with a steep inclination that is presumably acquired during the drilling/coring process (Fig. F62A, F62B). The relative importance of this low-stability overprint, however, is quite variable, and some intervals exhibit nearly univectorial demagnetization trajectories (Fig. F62C). Demagnetization at peak fields of 20–30 mT is generally sufficient to remove the low-stability overprint and to isolate a characteristic remanence (ChRM) with shallow positive inclinations.

The stability of the remanence, as measured by the ratio of the remanence remaining after 20-mT demagnetization to the original NRM (J_20mT/J_NRM), generally decreases downhole (Fig. F61). Particularly low stabilities and unstable remanence directions characterize the lower portion of Unit II. The downhole trend in magnetic stability is opposite that observed for NRM intensity and magnetic susceptibility, as well as for the degree of alteration (see "Metamorphic Petrology"). This inverse relationship can be explained by the increasing abundance of low-coercivity, coarse-grained magnetite with depth (as a result of the degree of serpentinization of the peridotite). As a result of its low coercivity, the coarse-grained multidomain magnetite is most susceptible to a drilling-induced magnetization. The steeper initial inclinations and lower overall magnetic stability in the lower portion of lithologic Unit II can plausibly be attributed to the greater abundance of such coarse-grained magnetite.

Remanence Data

A total of 18 discrete 9.2-cm³ cubic samples of peridotite (15) and dunite (3) were analyzed from Hole 1274A (Table T8). These samples were stepwise demagnetized up to peak fields of 40–60 mT using the D-Tech off-line degausser. At each step, samples were measured in the magnetometer in three orthogonal positions, so that tray positioning and superconducting quantum interference device (SQUID) calibration errors could be averaged at each step. In order to correct for direct-current residual fields in the demagnetizer, double AF demagnetization was applied for peak fields >20 mT (see "Paleomagnetism" in the "Explanatory Notes" chapter).

As observed from the archive-half data, the discrete samples from Hole 1274A exhibit a variable low-stability overprint that may be attributed to drilling (Fig. F62D, F62E, F62F). But, as noted for previous sites from Leg 209 (e.g., see "Paleomagnetism" in the "Site 1270" chapter), the ChRM is more easily isolated in the discrete samples than in the archive-half data. The characteristic component in discrete samples was normally isolated at low AF fields (8–15 mT), after which the remanence decays univectorially to the origin. The median destructive field (MDF, the alternating field required to reduce the vector difference sum to 50% of its original value) ranges 3–13 mT, and the ChRM represents as much as 25%–50% of the NRM. The fraction of NRM represented by the ChRM is significantly higher than the average fraction in other peridotites recovered from Sites 1268 through 1272.

The anisotropy of magnetic susceptibility (AMS) and the anisotropy of anhysteretic remanence (AARM) (McCabe et al., 1985) were determined for all discrete samples from Hole 1274A (Tables T9, T10). The AMS tensor was determined using a Kappabridge KLY-2 (Geofyzika Brno) and the standard 15-position measuring scheme. To determine the AARM tensor, the sample was given an ARM (60-mT AF; 0.1-mT bias field) along the six sample axial directions (i.e., +x, +y, +z and –x, –y, –z). After subtraction of a baseline demagnetization step for each ARM acquisition direction, the remanence anisotropy tensor was then calculated in a manner analogous to that used for the susceptibility tensor. These two measures of anisotropy provide complementary information on the preferred orientation/distribution of ferromagnetic minerals in the peridotites. AMS is primarily sensitive to the coarsest-grained magnetite particles, whereas smaller remanence-carrying grains dominate the AARM fabric. In serpentinized peridotites, AMS fabrics have most commonly been found to reflect the distribution of magnetite in serpentine veins (e.g., Lawrence et al., 2002; MacDonald and Ellwood, 1988).

Peridotite samples from Hole 1274A exhibit moderate degrees of susceptibility anisotropy and dominantly oblate AMS fabrics (Fig. F63). The degree of anisotropy (P = maximum/minimum eigenvalue) ranges 1.06–1.18. The corresponding remanence anisotropy is more pronounced (P = 1.09–1.46). The higher values in this range are sufficient that some significant deflection of the remanence is possible. The magnitude of this effect will be examined during shore-based studies.

In order to compare the orientations of magnetic fabrics from different samples, some common reference frame is required. Under the assumption that the stable remanent magnetization approximates the time-averaged normal polarity direction (see below), the magnetic fabric data have been restored to a common reference frame by a simple vertical axis rotation that restores the remanent declination to the presumed normal polarity direction (360°). After this reorientation, the maximum eigenvectors of the AMS tensors define a northeast-southwest–trending girdle while the corresponding minimum eigenvectors have shallow northwesterly or southeasterly directions. Thus, the AMS data define a magnetic foliation that is steeply dipping and strikes approximately northeast–southwest. The eigenvectors from the AARM tensors define a similar pattern, suggesting that a broad range of magnetite grain sizes have a common preferred orientation.

The direction of the ChRM for both long-core and discrete sample data was estimated by principal component analysis (Kirschvink, 1980), providing orientation information for 135 core pieces (Tables T8, T11). A single discrete sample has a steep negative inclination (–52°). This inclination most likely reflects an orientation error during sampling/curation (in addition, the piece is rounded on all sides and thus the vertical orientation is suspect).

Both the archive-half and discrete samples from Hole 1274A have inclinations that are significantly shallower than the expected dipole inclination at the site (28°) (Fig. F64). The mean inclination calculated from 17 discrete samples is 13.4° (+7.4°/–8.1°, = 20.3 using the inclination-only technique of McFadden and Reid, 1982). The characteristic inclinations determined from the archive-half data are more dispersed but most lie between –5° and +35° (Fig. F64), with an average value of 18° (+2°/–3°, N = 178). Note that for the longest core pieces (up to 1 m in length), multiple ChRM directions were calculated at ~10-cm intervals. This procedure gives more weight to the data from longer core pieces, an approach validated by the higher scatter evident in directions from pieces <10 cm in length (see below). Given the proximity of the site to the ridge axis and the normal polarity magnetization inferred from inversion of sea-surface magnetic anomaly data (Fujiwara et al., 2003), the shallow positive inclinations most likely represent a normal polarity remanence acquired during the Brunhes Chron (<0.78 Ma) (Cande and Kent, 1995).

The declination of the remanence shows a cluster at ~310° in core coordinates (Fig. F64). This clustering reflects the systematic selection of a splitting plane for the cores, such that the dominant fabric dips toward 090° in the core reference frame. Restoring the mean remanent declination to 360° (the time-averaged normal polarity declination) by a simple vertical axis rotation will result in a southeasterly dip direction for the dominant foliation (see "Structural Geology" for a more complete discussion of the orientation of structural features within the core).

Although there is general agreement between the ChRM directions determined from archive halves and discrete samples (Fig. F64), the former exhibit significantly greater scatter. This additional scatter may, at least in part, be attributed to artifacts associated with the measurement of half cores, particularly for small core pieces (see "Appendix"). If only archive-half data from pieces 10 cm are considered, the scatter is significantly reduced (Fig. F65). The combined archive-half and discrete sample data from Hole 1274A reveal no trend with depth in lithologic Unit I. The inclinations for Unit III are much closer to the expected value for the site. Because these two units are separated by fault gouge material, it is possible that some relative motion between the two blocks has occurred. However, given the small number of samples available from below the fault zone, no definitive conclusion can be reached.