A variety of gabbroic rocks (primarily oxide gabbro/gabbronorite and troctolite; see "Igneous and Mantle Petrology") were recovered from Holes 1275B and 1275D. These two holes were drilled within ~100 m of each other, and both had relatively deep penetration (109 and 210 m, respectively) and good recovery. Thus, the site is particularly well suited for continuous measurements of archive halves. Pass-through measurements yielded stable remanence directions for >500 core pieces (~350 from Hole 1275D and ~200 from Hole 1275B). In addition, stable directions were determined from 60 discrete samples (Table T7). Both Holes 1275B and 1275D are characterized by shallow, dominantly negative inclinations (mean inclinations = –10° and –4°, respectively) that presumably represent a reversed polarity remanence. Some intervals in Hole 1275D, particularly from the top 50 m, have positive inclinations and therefore may be of normal polarity. The presence of both polarities is not entirely unexpected, given the location of the site near Anomaly 2 (Fujiwara et al., 2003). The average inclinations in both Holes 1275B and 1275D are significantly shallower than the value expected from a geocentric axial dipole (±28°), apparently requiring significant tectonic rotations since remanence acquisition.
Remanence measurements were made on all archive halves longer than ~7 cm. The archive halves were subjected to stepwise alternating-field (AF) demagnetization at 5- to 10-mT steps up to maximum peak fields of 70 mT. The characteristic remanent magnetization (ChRM) directions were calculated by principal component analysis (PCA) (Kirschvink, 1980) at about the midpoint of each piece or at 10-cm intervals for pieces >15 cm. Whole-core (or archive half for Hole 1275D) susceptibility data obtained from the MST (see "Magnetic Susceptibility" in "Physical Properties") were filtered to preserve only data corresponding to the intervals where ChRM directions were determined. Thus, the resulting remanence and susceptibility data sets (Figs. F90, F91) are significantly less affected by artifacts resulting from small core pieces than are the original unfiltered data sets.
The most significant variations in the natural remanent magnetization (NRM) intensity and susceptibility for archive halves from Hole 1275B are well correlated with the lithology (Fig. F90). The diabase and oxide gabbro(norite) samples from lithologic Unit I (see "Igneous and Mantle Petrology") are characterized by very low susceptibilities (~10–3 SI) and remanent intensities (1–100 mA/m). The lowest susceptibility values in this unit could be attributed almost entirely to a paramagnetic contribution (e.g., 10% FeO corresponds to a susceptibility of ~6 x 10–4 SI) and thus are apparently at odds with the visual estimates of Fe-Ti oxides (1% to >5%) (see "Igneous and Mantle Petrology"). Reflected-light observations indicate that the oxide minerals in the oxide gabbros are predominantly hemoilmenites with little or no magnetite (which has a much higher susceptibility of ~3 SI). The dominance of hemoilmenite can account for both the low susceptibility as well as the low NRM intensities and high magnetic stability of this unit (see below).
In contrast to lithologic Unit I, the troctolites and oxide gabbros of lithologic Units II and III are characterized by high susceptibilities (generally 0.01 to >0.1 SI) and NRM intensities (typically 1–10 A/m), indicative of much higher concentrations of magnetite. The maximum magnetic susceptibility and NRM intensity values were observed in Cores 209-1275B-19R to 20R. In this interval, magnetic susceptibility values exceeded the maximum value (0.1 SI) measurable with the MST susceptibility meter. The susceptibility of many individual archive-half pieces was subsequently measured to identify and correct these intervals of clipped data (see "Igneous and Mantle Petrology").
The differences in magnetic mineralogy between lithologic Unit I and the lower units is also reflected in the directions of the natural remanence. NRM inclinations for the more magnetite-rich lower units are scattered but include values ranging from +20° to +90°. In particular, Cores 209-1275B-19R to 20R have the most uniformly steep NRM inclinations (+50° to +80°) and this interval is also characterized by uniformly high susceptibility values. This association of steep initial inclinations and high susceptibility may be explained by the presence of coarse-grained magnetite that has low coercivity and thus is most susceptible to acquiring a drilling remanence. Much of the variability in NRM inclination in Units II and III may therefore be plausibly attributed to variations in the amount of such coarse-grained magnetite. The diabase and oxide gabbros from Unit I have shallow NRM inclinations that reflect the paucity of coarse magnetite.
Stepwise demagnetization of the archive halves reveals a variable contribution from a steep, presumably drilling-related, low-stability overprint (Fig. F92). This low-stability component is generally removed by AF demagnetization at 15–25 mT. The ChRM directions from Hole 1275B were typically calculated from the remanence remaining after demagnetization at 30–70 mT (Table T8). The ChRM directions were classified into two categories; samples showing a linear decay trend toward the origin (e.g., Fig. F92D) were labeled as Class A and are considered to be a reliable record of the stable remanence, and samples with a curving demagnetization trajectory (e.g., Fig. F92F) were identified as Class C and provide a less reliable record of the paleomagnetic inclination as a result of the significant overlap between the ChRM and the low-stability drilling overprint. Class C directions, however, can provide useful information for reorientation of structural features in the core because the vertical drilling overprint causes no significant variations in declination. Class A components represent on average 40% of the NRM, whereas Class C components represent approximately <10% of the NRM.
As with the remanent intensity and susceptibility, there is a distinct difference in magnetic stability (as measured by the class of the ChRM and the percentage of the NRM represented by the ChRM) between lithologic Unit I and Units II and III. Samples from Unit I have dominantly (90%) Class A remanence directions (Fig. F92D) and the ChRM represents a large fraction (average = 80%) of the NRM (Fig. F90). In contrast, the ChRM constitutes a smaller fraction (average = 25%) of the NRM for samples from Units II and III (Fig. F92E, F92F) and these units have a greater proportion of Class C directions.
The NRM intensity and susceptibility values of gabbroic rocks from Hole 1275D (Fig. F91) are comparable to those measured on samples from Hole 1275B. NRM intensities are typically 1.0–10 A/m, with maximum intensities as high as 35 A/m. This maximum value essentially corresponds to the highest number of flux counts that can be measured by the 2G magnetometer (105 counts). The high magnetic moment of the archive halves frequently resulted in residual counts even at the slowest track speeds (1 cm/s) available with the current version of the LongCore program. Susceptibility values were also high enough to cause clipping of the whole-core data measured by the MST. In order to provide an accurate measure of the susceptibility variations, the archive halves were scanned with the MST (see "Igneous and Mantle Petrology"). The smaller volume resulted in very few intervals that still exceeded the maximum measurable susceptibility (0.1 SI).
As with Hole 1275B, the downhole variations in NRM intensity and susceptibility for Hole 1275D generally reflect lithologic variations. The most notable lithologic control on magnetic properties is for intervals of diabase (Fig. F91). These finer-grained units are generally associated with lower susceptibilities (10–3 to 10–2 SI) and NRM intensities that are generally <1 A/m (e.g., near 108 and 167 mbsf). The troctolites of lithologic Unit I (0–52.99 mbsf) have somewhat lower NRM intensities than do units from lower in the section. However, the troctolites in Unit III (81.00–90.94 mbsf) are not obviously distinct from adjacent units in either their susceptibility or NRM intensity. Overall, the variations in the magnetic properties are less closely related to the lithologic units defined for Hole 1275D than was the case for Hole 1275B.
Most NRM inclinations for Hole 1275D are 30°–90° (mean = ~60°) (Fig. F91). The only exceptions are the shallow initial inclinations associated with some diabase units (e.g., the lower portion of Core 209-1275D-39R and the brecciated diabase in Core 1R). Stepwise AF demagnetization of the archive halves from Hole 1275D reveals the presence of a significant low-stability component with steep inclination that is presumably related to the drilling/coring process (Fig. F93). This low-stability drilling remanence is effectively removed by AF treatments of 15–25 mT. The NRM inclinations from Hole 1275D are, in general, steeper than those from Hole 1275B, reflecting the relatively larger contribution of a drilling remanence in cores from Hole 1275D.
ChRM directions were calculated from the archive-half demagnetization data following the same procedure as for Hole 1275B. Class A components could be calculated from nearly all intervals, even in samples with substantial drilling overprints and where the ChRM represents as little as 1% of the initial NRM (Table T8). The intensity of the ChRM varies significantly as a function of lithology. Gabbros yielded ChRM intensities that ranged ~0.1–1 A/m, or even higher (e.g., Cores 209-1275D-31R, 32R, and 34R). In contrast, diabases and microgabbros (e.g., Cores 209-1275D-23R, 29R, 33R, and 36R) have ChRM intensities as much as an order of magnitude lower than the coarser gabbroic rocks.
The stable inclinations from Hole 1275D may be subdivided into three groups (Fig. F91). The troctolite samples from lithologic Unit I have dominantly positive inclinations, although some isolated shallow negative inclinations (e.g., Core 209-1275D-1R) are also present. The ChRM in this interval generally represents >10% of the initial NRM. Cores 209-1275D-11R to 30R (~50–140 mbsf) constitute a second inclination group. The inclinations in this interval are shallow and negative, with a mean of about –10° that is similar to the mean inclination obtained from Hole 1275B. An abrupt shift in inclination to values near –20° occurs between Cores 209-1275D-30R and 31R. The steeper inclinations below ~140 mbsf constitute a third inclination group. This inclination group broadly corresponds to lithologic Unit V (olivine gabbro and oxide gabbronorite; see "Igneous and Mantle Petrology"). Within this lower inclination group, intervals of diabase and microgabbro are characterized by inclinations that are distinctly more positive than the surrounding gabbro.
Stepwise demagnetization of the NRM and magnetic susceptibility measurements were carried out on 27 discrete samples (9.2-cm3 cubes) from Hole 1275B and 36 samples from Hole 1275D. When possible, directional data obtained from AF demagnetization were used to approximately restore the magnetic anisotropy data of the samples into a geographic reference frame.
AF demagnetization of discrete samples from Hole 1275B (Fig. F92) yielded variable results depending on the lithology, but stable characteristic components after treatment to 20–30 mT were observed in most of the samples. Diabase samples from the upper Unit I have the lowest NRM intensities but very stable magnetizations (e.g., Fig. F92A). The drilling overprint in this sample is insignificant, and the ChRM constitutes nearly 100% of the initial NRM. Oxide gabbro(norite) samples show a variable drilling overprint that is removed by AF treatments of 15–20 mT (Fig. F92B, F92C). However, a linear characteristic component trending toward the origin can generally be calculated from the 30- to 100-mT demagnetization intervals.
As observed at previous sites, the characteristic component is more clearly isolated in the discrete sample data than in the long-core measurements. For example, Sample 209-1275B-19R-1, 76–78 cm, shows a linear decay trend from 30 to 100 mT (Fig. F92C), from which a PCA direction of Class A can be calculated. The same interval in the long-core data does not show a clear isolation of the higher-coercivity component but yields a smooth curve from the vertical drilling overprint component to the shallow directions of the last demagnetization steps (Class C direction).
As observed for samples from Hole 1275B, the demagnetization of discrete samples of gabbros from Hole 1275D reveals a stable characteristic component after demagnetization to 20- to 30-mT peak fields. This stable component typically represents 10%–100% of the initial NRM (Fig. F93), shows a linear decay toward the origin, and has maximum coercivities between 70 and 100 mT (Class A direction). Overall, the discrete sample data agree well with the long-core data, particularly with respect to the remanent declination.
Most gabbros, both from Holes 1275B and 1275D, have remanence inclinations that are either subhorizontal or negative. However, a significant number of troctolite, diabase, and microgabbro samples recovered from Hole 1275D yielded less uniform inclinations and include clearly positive inclinations (Fig. F94). The characteristic component for some of these samples has lower coercivities (typically <50 mT) than the average of the coarser-grained gabbros. For example, a diabase sample from Core 209-1275D-33R is nearly completely demagnetized by 15 mT (Fig. F94C) but has an apparently stable final component with positive inclination that represents <1% of the NRM (Class C direction). However, some positive inclinations are associated with Class A demagnetization behavior. For example, a troctolite sample from Core 209-1275D-9R (Fig. F94A) yielded a well-isolated characteristic component with an inclination of +12°, representing 23% of the NRM. More rarely, samples with two shallow, nearly antipodal components were recognized (Fig. F94B), with the higher-stability component having a positive inclination. A similar pattern, though less clearly defined, is evident in the demagnetization data from the corresponding archive-half interval (Fig. F94E).
The anisotropy of magnetic susceptibility (AMS) was determined for all discrete samples from Site 1275 (Table T9). Gabbroic samples from Site 1275 exhibit low to moderate degrees of susceptibility anisotropy, and the AMS fabrics are triaxial to oblate. The degree of anisotropy (P = maximum/minimum eigenvalue) ranges ~1.00–1.23, though typically the degree of anisotropy is <1.10. Although only a small number of diabase samples were measured, these fine-grained lithologies have nearly isotropic susceptibility tensors (Table T9).
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 reversed 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 reversed polarity direction (180°). After this reorientation, the minimum eigenvectors of the AMS tensors for gabbros from Hole 1275B form a distinct cluster at steep inclinations (Fig. F95). The corresponding maximum eigenvectors delineate a nearly horizontal girdle, possibly dipping slightly to the northwest. The restored AMS data from Hole 1275D are much more scattered, although the mean orientation of the minimum eigenvectors is also steep.
Inclinations of Class A components from 186 archive-half pieces agree well with the inclination distribution from the discrete sample data from Hole 1275B (Fig. F96). The discrete sample data yield a mean inclination of –9.5° (+3.9°/–3.7°; 
= 50.3; N = 26; using the inclination-only method of McFadden and Reid, 1982), which is significantly lower than that expected for the geocentric axial dipole at the site (28°). Despite their different magnetic properties, both the gabbros and the diabases from the upper part of Hole 1275B show similar directional results.
Overall, the inclinations from Hole 1275D are in agreement with the data from Hole 1275B. However, Hole 1275D samples show less uniform results, both with respect to the lithology and as a function of depth. The mean remanence inclination for discrete samples from Hole 1275D is –3.5° (+7.0°/–6.7°; = 11.9; N = 34), a value that is statistically indistinguishable from the mean inclination for Hole 1275B.
Although both Holes 1275B and 1275D apparently have dominantly reversed polarity magnetizations, some normal polarity magnetizations are apparently also present in Hole 1275D. In particular, the troctolites and diabases of Unit I have scattered but generally positive inclinations that may be interpreted as normal polarity magnetizations. Some of the intervals in Unit I (e.g., from Cores 209-1275D-9R and 10R) (Fig. F91) have similar magnetic stability to the presumed reversed polarity samples elsewhere in the hole. In addition, the presence of one sample with two nearly antipodal magnetization components (Fig. F94B) suggests that both normal and reversed polarity magnetizations are recorded at this site. At present, the origin of this normal polarity (positive inclination) signal is uncertain. It may reflect either an ancient thermoremanence or a more recent chemical or viscous remanence.
