We measured the natural remanent magnetization (NRM) of all archive halves from Hole 1136A with the pass-through cryogenic magnetometer using measurement intervals of 5 and 2.5 cm for sediment and basalt sections, respectively. After measuring the NRM, all sediment cores were demagnetized with a peak alternating field (AF) of 20 mT. Two discrete samples were taken per section and were stepwise AF demagnetized up to 60 mT. All basalt archive-half cores were stepwise AF demagnetized with a peak field of 60 mT. Discrete basalt samples were stepwise AF demagnetized in a peak field of 60 mT or thermally demagnetized in temperatures of up to 620°C. We determined the anisotropy of magnetic susceptibility (AMS) of discrete basement samples to obtain information about the magnetic fabric.
We obtained reliable results in undisturbed cores with high recovery, and correlated normal and reversed segments with biostratigraphic zones (see "Biostratigraphy"). However, RCB cores do not provide a continuous magnetostratigraphic record as a result of core disturbance and incomplete recovery. Secondary components of remanent magnetization could be removed by AF demagnetization at 10 mT (Fig. F35A). Stable remanent magnetizations were obtained after AF demagnetization at 20 mT, especially for sediments with a higher median destructive field (MDF) (Fig. F35B). For the magnetostratigraphic study of Hole 1136A (Fig. F36), we used the data selection criteria as described in "Paleomagnetism" in the "Explanatory Notes" chapter. The selection criteria were that (1) the intensity of remanent magnetization after AF demagnetization at 20 mT was > 2 × 10-4 A/m and hence above the noise level of the magnetometer in rough-sea conditions, (2) the inclination was > ±30°, (3) at least three consecutive values (which corresponds to a 15-cm length of split core) had the same polarity, and (4) there was no significant core disturbance. We calculated characteristic remanent magnetic directions of discrete samples using component analysis. Characteristic inclinations from discrete samples generally agree well with selected inclinations from whole-core measurements (Fig. F36).
Correlation of biostratigraphic data and polarity reversals (Fig. F36) suggests that the reversed and normal chrons in Hole 1136A are middle Eocene to Early Cretaceous in age (see "Biostratigraphy"). Both Sections 183-1136A-2R-CC (14 mbsf) and 3R-CC (23 mbsf) lie within nannofossil Subzone CP12b (see "Biostratigraphy"). This reversed polarity interval at ~15 mbsf correlates with Chron C21r. We correlated the next reversed polarity interval at ~27 mbsf to Chron C21r, based on the paleontological identifications; Section 183-1136A-CC lies within nannofossil Subzone CP12a and upper foraminifer Zone AP7 (see "Biostratigraphy"). We correlate reversed and normal polarity intervals at ~35 mbsf and between 42 and 53 mbsf to Chrons C22r and C23n, respectively, according to biostratigraphic data (see "Biostratigraphy"). Normal polarity sequences at ~80 and ~120 mbsf can be correlated to early Maastrichtian Chron C31r and Albian Chron C34n, respectively, according to nannofossil zoning (see "Biostratigraphy").
A summary of NRM intensities and susceptibilities (see "Physical Properties") for the four lithostratigraphic units (see "Lithostratigraphy") is presented in Table T9. In the upper part of Unit II (nannofossil ooze) we observed negative susceptibilities and weak NRM intensities. Remanent magnetization was less stable, and magnetization of discrete samples collected from this part was too weak to measure reliably with the shipboard magnetometer during rough seas. We observed weak and positive susceptibilities and stronger NRM intensities in the lower part of Unit II than in the upper part. Progressive AF demagnetizations and measurements of discrete samples were successful and provided a reliable paleomagnetic record. High susceptibilities (average >2 × 10-4 SI units) and strong NRM intensities (average >1 × 10-2 A/m) were observed in Units IV and V. We obtained stable remanent magnetizations during progressive AF demagnetization experiments (Fig. F35). Stable and reliable paleomagnetic data were obtained from both half-core and discrete measurements in this interval.
We obtained a continuous record of the NRM inclination and intensity from measurements of half-core basement sections. We chose two discrete samples for thermal demagnetization and one sample for AF demagnetization from each flow unit (basement Units 1 and 2; see "Igneous Petrology"). We found negative inclinations, and hence, normal polarity in the basement rocks of Site 1136.
To observe the magnetic properties of each basement unit and to determine the variation of magnetic properties within each unit, we performed whole-core measurements of remanent magnetization and susceptibility (Fig. F37). Three different types of susceptibility measurements generally show consistent results. Anomalously low MST susceptibility values and low AMST values are commonly caused by gaps in recovery (see "Paleomagnetism" in the "Explanatory Notes" chapter). In basement Unit 1, susceptibilities are constant except for the lower 2 m of the flow, where we observe slightly weaker susceptibilities. The average susceptibility of basement Unit 2 is almost the same as that of basement Unit 1 (Table T9). We found more significant variations in susceptibility within basement Unit 2 than in basement Unit 1. Weaker susceptibilities were observed in the lower half of this unit. The variation in susceptibility is probably caused by variations in the magnetic mineral type or variations in the content of magnetic minerals in the basement units with depth. No significant difference is observed in average NRM intensities between the two basement units. After AF demagnetization at 40 mT, we found a clear difference between the intensities of remanent magnetization of basement Units 1 and 2. NRMs of Unit 2 lose their intensity easily by AF demagnetization (Fig. F37). This is caused by differences in size of magnetic mineral and/or mineralogical differences.
Two samples from the basalt flows of Site 1136 were demagnetized with an AF of up to 60 mT. Both samples possess single-component magnetizations, as is evident from the straight lines on the Zijderveld plots (Fig. F38). We expect single-domain magnetite or titanomagnetite particles to be the magnetic carrier in these samples because of their high NRM intensity (7.5 and 9.6 A/m) and high MDF (43 and 37 mT). Four samples were stepwise thermally demagnetized up to 620°C (Fig. F39). We measured the susceptibility of the samples after every heating step to detect changes of their magnetic minerals (Fig. F39). We determined an unblocking temperature of ~580°C in Samples 183-1136A-18R, 132-134 cm, and 15R-2, 116-118 cm (Fig. F39A). This temperature corresponds to the Curie temperature of magnetite or titanium-poor titanomagnetite, which we consider to be responsible for the remanent magnetization of these samples. The constant decrease in susceptibility with temperature is probably caused by a mineral change from (titano)magnetite to (titano)hematite. The decay curve of Sample 183-1136A-18R, 132-134 cm (Fig. F39B) during thermal demagnetization indicates that two magnetic phases are responsible for the remanent magnetization. Most of the magnetization was removed by ~300°C, and the small remaining magnetization disappeared by ~580°C. The low-temperature phase (300°C) is maghemite or titanomaghemite. The increase in susceptibility at 400°-500°C and the following decrease at 500°-600°C correspond to changes from (titano)maghemite to (titano)magnetite and from (titano)magnetite to (titano)hematite, respectively. Maghemite and titanomaghemite are low-temperature oxidation products of magnetite and titanomagnetite, respectively. Sample 183-1136A-17R-1, 22-24 cm, seems to contain components of both (titano)magnetite and (titano)maghemite (Fig. F39C). Titanomaghemite and titanomagnetite have been previously reported from basement basalt of the southern Kerguelen Plateau (Heider et al., 1992).
We obtained characteristic remanent magnetic directions of discrete samples using component analysis. The declinations of all samples are arbitrary because the cores were unoriented. The six samples have inclinations ranging from -69° to -76° (Table T10). For basement Units 1 and 2, we obtained mean inclinations of -74° and -70°, respectively. The small difference between these two values could be insignificant, given the limited shipboard measurements. A certain amount of inter-unit difference is expected from secular variation of the geomagnetic field, depending on the elapsed time between lava-flow eruptions. The average inclination of the two units is -72° and agrees with the mean inclination obtained from whole-core measurements (Fig. F37). The mean inclination of whole-core measurements is -71° and was obtained using inclinations (after AF demagnetization at 40 mT) from long igneous cores that are longer than the effective sensitivity of the pass-through magnetometer (~15 cm). The paleolatitude calculated from the mean inclination (-72°) assuming a geocentric dipole field is 56°S. Thus, the difference between the paleolatitude and the present latitude of Site 1136 (60°S) is only 4°. Previous paleomagnetic results show a difference of 13° to 16° between the present latitude and paleolatitude of basement rocks from the southern Kerguelen Plateau (Inokuchi and Heider, 1992). The difference between our results and the previous study can be explained by (1) inaccurate paleolatitude estimates caused by limited measurements, (2) secular variation of the geomagnetic field, (3) possible recent overprints of the remanent magnetization, (4) tilting of the basement units, (5) differences in age of the units measured, or (6) divergent motion of separate tectonic blocks. As there is no evidence for the last three explanations, we conclude that the inconsistent results may be caused by inadequate measurements, secular variation, and/or overprints.
From 20 discrete samples of basement Units 1 and 2, we determined the AMS using susceptibility measurements in 15 different orientations (Fig. F40). Generally, we found only a low degree of anisotropy, with an average magnitude of 1.016. We used the three principal susceptibilities for each sample to determine the shape of the fabric. Most of the samples have a positive shape parameter, which corresponds to oblate (disk) shapes, while negative values correspond to prolate (rod) shapes. The minimum principal axes are grouped at an inclination of 90° and the maximum and intermediate directions are within a girdle 90° from the minima (Fig. F40), hence there is a predominant magnetic foliation. The foliation plane in basement rocks generally lies close to the flow plane, and the minimum susceptibility axis is perpendicular to the flow plane. The distribution of the minimum axes (Fig. F40) indicates horizontal outflow for Units 1 and 2.