The results described above have revealed important information about the origin of remanence and magnetic minerals present in the Leg 183 cores. The Leg 183 cores recovered from the six drilling sites displayed variable rock magnetic properties. Three general types of behavior were found in the rock magnetic measurements. One group has a single phase of Ti-poor titanomagnetite. The majority of subaerial basaltic samples from Sites 1136, 1137, 1138, 1141, and 1142 belong to this group. Results for various massive subaerial basalt samples (Table T2) show Curie temperatures between 480º and 605ºC, compatible with those of Ti-poor titanomagnetites. The thermomagnetic curves of these samples exhibit very little difference between heating and cooling of the samples. Samples with titanomagnetite also exhibit a strong Verwey transition in the vicinity of 110 K and frequency-dependent susceptibility curves that resemble those of synthetic Ti-poor titanomagnetites. These results are in good agreement with the hysteresis ratios suggesting that the bulk magnetic grain size is in the PSD boundary (e.g., with lower Hcr/Hc values). Subaerial basalts have been thoroughly studied (e.g., Lindsley, 1962; Grommé et al., 1969; Ade-Hall et al., 1971; Tucker and O'Reilly, 1980). They have high Curie temperatures (500°-580ºC) and rarely have two Curie points (Dunlop and Özdemir, 1997). They undergo deuteric high-temperature oxidation, which is responsible for the high Curie temperature. It is this high-temperature oxidation, especially when carried to completion with the production of fine-grained hematite, that makes subaerial basalts excellent recorders of the geomagnetic field (Dunlop and Özdemir, 1997). In this regard, a subaerial sample (Sample 183-1136A-16R-2, 140-142 cm) from Site 1136 may have experienced the high-temperature oxidation, as the Mössbauer spectrum of this sample shows sign of presence of hematite. Therefore, the subaerial basalts from this group are most likely good paleomagnetic recorders that can preserve original and stable magnetic remanences.
The second group is characterized by a Curie temperature of 200°-400ºC and is mainly represented by pillow basalt samples from Site 1140. Samples in this group apparently went through low-temperature oxidation, which is another form of alteration for titanomagnetite (Dunlop and Özdemir, 1997). Thermomagnetic analysis using the low-field susceptibility method shows that the remanent magnetization in this group is carried by a thermally unstable mineral that breaks down at ~400ºC. Thermomagnetic analysis using the high-temperature vibrating sample magnetometer shows evidence for unoxidized titanium-rich titanomagnetite (TM60) (Fig. 2B). The low-temperature curves for this sample and the rest of the samples of the group do not show Verwey transition. The frequency-dependent relationships are distinctly different from those in the first group and show little or no sign of pure titanomagnetite characteristics. The magnetic grain sizes of the pillow basalt samples fall toward the boundary between SD and PSD with several samples fall in the SD size (Fig. F3). The rock magnetically inferred fine grain size indicates a rapid cooling environment for the pillow lavas of Site 1140. Altogether, these rock magnetic data seem to be sensitive indicators of low-temperature oxidation and support the contention that titanium-rich titanomagnetite is responsible for the magnetic signatures displayed in the pillow basalts.
The third group has more than one Curie temperature, which suggests the presence of multiple magnetic phases. Four of the six samples from Site 1139 (Table T2) belong to this group. Although the hysteresis ratios for rocks in this group still fall in the PSD region, the cluster is centered toward the MD region (with higher Hcr/Hc ratios). Low-temperature curves do not clearly show the Verwey transition (Fig. F6C). The thermomagnetic signature indicates the inversion of titanomaghemite to a strongly magnetized magnetite, as showing by the irreversible cooling curve (Fig. F2C). Although chemical remanent magnetization resulting from oxidization of titanomagnetite and inversion of titanomaghemite has been shown to parallel the original thermoremance (Johnson and Merrill, 1974; Hall, 1977; Özdemir and Dunlop, 1985; Dunlop and Özdemir, 1997), it is difficult to assess whether these rocks retain stable remanent magnetization from the data collected in this study.
Variations of Curie temperatures also suggest that a stratigraphically distinct change in magnetic minerals exists in Leg 183 drill sites. It is significant to note that the variations come from relatively deeper sections of the basement that coincide with changes in lithology and alteration. For example, rocks in the lower part of Site 1136 basement exhibit lower Curie temperatures (Sections 183-1136A-18R-2 through 19R-2) (see Table T2). Preliminary shipboard observations also show that vesicle distribution in these cores increases significantly, from 1% in the overlying sections to 10%-15% (Shipboard Scientific Party, 2000a). The vesicles and veins are filled with dark olive-green clays, which not only suggests that alteration is most likely the product of both weathering and low-temperature fluid-rock reaction but also emphasizes the importance of distinctions between primary and secondary causes of magnetic property variation.
Although not every problem in paleomagnetism has a rock magnetic solution, there is clearly a strong interest in constraining remanent magnetization by detailed rock magnetic studies (e.g., Banerjee, 1992). Measurement of rock magnetic parameters has been demonstrated to be useful for studying various rock-forming and rock-altering geological processes. The results described in this study have revealed important information about the origin of remanence and the magnetic minerals present in the Leg 183 cores. The rock magnetic data from Leg 183 samples clearly indicate that titanomagntite is the dominant mineral and the primary remanence carrier. The generally good magnetic stability and other properties exhibited by these rocks support the inference that the ChRM isolated from the Cretaceous sites were acquired during the Cretaceous Normal Superchron. The stable inclinations identified from these samples are therefore useful for future tectonic studies.
The rapidly cooled pillow basalt samples from Site 1140 (34 Ma) are fresh to slightly altered and have a more uniform magnetic grain size distribution. From the high-field magnetic moment curves and Curie points, it may be inferred that Ti-rich titanomagnetites are present in these submarine basalts and they are expected to give accurate results. Basalts recovered at Site 1142 were questioned as pillowed basalts in the shipboard petrological description (see Table T2). Results of our rock magnetic investigation on a limited set of these rocks, however, would suggest they were most likely erupted in a subaerial environment, which is similar to their counterpart at Site 1141.
Most of samples from Site 1139 showed the third group behavior, which raises the question of whether the reversed signal observed in six Site 1139 cores is due to field or self reversals. We do not believe these samples have a self-reversed thermal remanent magnetization because the thermal magnetic curves for the normal and reversed magnetized samples are indistinguishable. However, we noticed that Site 1139 lies on the Skiff Bank, which has been proposed as the current site of the Kerguelen hotspot. Although hundreds of meters of sediment on parts of the elevated feature argue against the Skiff Bank originating entirely as a result of recent volcanism, the shipboard petrologic observation for these altered volcanic rocks and trachyte basalt favor the hypothesis that the remanence may be a recent overprint during alteration. Magnetic grain-size measurements in this study, as determined from hysteresis ratios, show that the magnetic minerals of Site 1139 cores are of PSD-MD size and have undergone low-temperature oxidation. This information, together with other observations and a lack of evidence to the contrary, would suggest that substantial alteration has subsequently taken place and may have reset the original magnetization. Thus, we favor the speculation that the reversed magnetization in the six samples from Site 1139 involves resetting of the direction of remanence by alteration. Additional work is still needed to constrain the magnetic interpretation.