At Site 1126 archive halves of all APC, XCB, and RCB cores were measured in the 2G 760-R magnetometer, except for those core sections obviously disturbed by drilling or with inadequate recovery for meaningful measurements. The cores were routinely measured as natural remanent magnetization (NRM) and after 20-mT demagnetization. A few were measured with additional intermediate steps up to 30-mT demagnetization. One core was measured as a whole core and the results were compared with archive-half measurements performed later. Strongly lithified fragments whose orientation in the core could be determined were measured as discrete samples. Rock magnetic analysis was performed on samples from soft sediments, on lithified pelagic limestones, and on sandstones. These analyses were performed using the methods described in "Paleomagnetism" in the "Explanatory Notes" chapter.
The results of long-core measurements of the nannofossil ooze cores from Holes 1126A and 1126B were disappointing. The NRM was dominated by a vertically downward coring contamination that was largely removed by 20-mT demagnetization, whereupon the signal was almost uniformly reduced to the noise level of the instrument. In addition, there were anomalous peaks in intensity near the top of most cores that were so large that we interpret them as some form of contamination introduced either by coring or by treatment on board. Only a single core (Core 182-1126B-19H) gave a sequence of reversals that could be correlated with the geomagnetic polarity time scale (Fig. F8). A comparison between whole-core and archive-half core measurements showed that the two measurements of inclination are not significantly different, but there are systematic differences in declination that remain to be interpreted. Although not as convincing as the Core 182-1126B-19H results, possible reverse-to-normal reversals were observed in Cores 182-1126B-5H (41.0 mbsf), 182-1126C-5H (44.2 mbsf), 182-1126B-9H (81.2 mbsf), and 182-1126B-17H (131.4 mbsf). Intervals of uniform normal polarity were observed in Cores 182-1126B-3H and 5H, and intervals of reversed polarity were observed in Core 182-1126B-17H. A common feature of many of the cores was that the upper halves of cores had major spikes in intensity and unrealistically rapid fluctuations in inclination that were not present in the lower halves. Poor recovery between Cores 182-1126B-19H and 27X precluded further long-core observations, but measurements were successful on nannofossil chalk recovered in Cores 182-1126B-27X, 28X, and 29X, where reversals were again observed.
In Hole 1126D the sandstone in Cores 28R to 33R gave a strong and stable NRM that had a predominantly steep upward or normal inclination, with an average value of 67.3° and a standard deviation of 14° (175 determinations). This corresponds to a paleolatitude of ~50°S.
Discrete samples were analyzed to aid interpretation of the NRM. In most cases, Zijderveld plots failed to define characteristic magnetizations primarily in the nannofossil oozes, although in Cores 182-1126B-18H to 19H, and particularly in 19H, the remanence was stronger, permitting further cleaning before the signal became too weak for measurement. However, only the nannofossil chalks and the sandstones gave good Zijderveld plots with maximum angular deviation angles of <5°. Results for Sample 182-1126D-33R-2, 87-89 cm, are shown in Figure F9. A soft component that is oriented steeply downward was demagnetized by 5 mT. There was then a univectorial decay close to the origin, indicating the isolation of a characteristic direction of magnetization.
Rock magnetism analysis consisted of measurements of weak-field susceptibility at two frequencies, progressive isothermal remanent magnetization (IRM) acquisition, alternating field demagnetization of anhysteretic remanent magnetization (ARM) and IRM, and low-temperature observations of warming curves of IRMs induced at liquid nitrogen temperature. The rock magnetism characterizations of the samples are presented in the form of modified Cisowski plots (Cisowski, 1981). In these plots the acquisition of IRM and the demagnetization of NRM, ARM, and IRM are shown in absolute values. Examples of these are presented in Figure F10 for weakly and more strongly magnetized nannofossil ooze, nannofossil chalk, and sandstone.
Sample 182-1126B-1H-7, 6-8 cm (Fig. F10A), is representative of the weakly magnetized nannofossil ooze. The IRMs and ARM are well above the 10-4 A/m (or 10-1 mA/m) noise level of the instrument even after demagnetization, whereas the NRM is within the noise level and shows no systematic behavior. The magnetic material in this sample has an extremely hard component and cannot be simply magnetite. Indeed, from the IRM acquisition curve it is evident that our use of a 100-mT field for ARM would not excite the hardest particles. Sample 182-1126B-19H-3, 59-61 cm (Fig. F10B), which provided the good record of reversals, is quite different, having an NRM that demagnetizes systematically and is above the noise level even after 40-mT demagnetization. The acquisition of IRM is consistent with magnetite as the dominant carrier. The high ARM:IRM ratio suggests that it is in a single-domain state, consistent with a biological source. Sample 182-1126D-11R-1, 14-16 cm (Fig. F10C), comes from the nannofossil chalk and is magnetically similar to the more strongly magnetized nannofossil oozes with an NRM that systematically demagnetizes above the noise level of the instrument and a high ARM:IRM ratio. Sample 182-1126D-33R-1, 58-60 cm (Fig. F10D), from the sandstone has a saturation IRM two orders of magnitude greater than the carbonates and the NRM is two orders above the noise level of the instrument. The ratio of NRM to IRM is appropriate for a depositional remanence.
Susceptibility was measured at the two available frequencies on the Bartington MS2 (0.465 kHz and 4.65 KHz) to estimate superparamagnetic content (Mullins and Tite, 1973). The lower negative susceptibility at the higher frequency demonstrates that there is indeed a superparamagnetic component. However, the negative susceptibility of the nannofossil oozes and chalk, reflecting the diamagnetism of the dominant carbonates, precludes determination of the amount of superparamagnetic material present. In contrast, in the sandstone, which has positive susceptibility, the difference is ~10%. This is comparable with values for dusts or unweathered surface soils but is higher than most values for sandstones. Thus, it appears that this sandstone has an unusual amount of very fine magnetite that could reflect eolian input.
The low-temperature observations consisted of warming curves of IRMs induced at liquid nitrogen temperature to identify the changes in remanence at the Verwey transition of magnetite and the Morin transition of hematite (e.g., Nagata et al., 1964). There were indications of the presence of magnetite in the form of knees in the curves at the Verwey transition of magnetite in samples from the nannofossil ooze that gave the good paleomagnetic signal and in the sandstone, whereas in the nannofossil chalk, a distinct dip was seen at the transition, implying low-temperature memory of magnetization across the transition (Nagata et al., 1964). In the sandstone sample, there was no sign of the hematite transition.
To summarize, the magnetic properties of the various samples differ in a manner that is consistent with the variation in their NRM. In the nannofossil oozes the NRM is for the most part below the sensitivity of the new 2-G instrument and the IRM and ARM are smaller than in the samples for which we could measure NRM.
With only a single core displaying indisputable reversals in the nannofossil ooze cores, potential for magnetostratigraphy is limited. However, the nannofossil record demonstrates that Core 182-1126B-19H lies between 13 and 18 Ma and probably between 16 and 18 Ma. In addition, a very rough estimate of the overall sedimentation rate is 10 mm/103 yr. Hence, we seek a sequence of reversals following the pattern we have observed with an approximate time interval corresponding to 7 m of section, equivalent to 700,000 yr. C5Cr-C5Cn-C5Br provides a possible candidate pattern; accordingly, we provisionally interpret this sequence of reversals as representing the sequence from the end of C5Cr to the beginning of C5Br.
In the nannofossil chalk the stronger magnetization permitted the detection of reversals, but the discontinuous nature of the record made interpretations tentative. However, it appears that C7n is recorded in the interval with the top at 217.8 mbsf.
Only normal magnetizations were observed in the sandstone, which would be consistent with magnetization in the Cretaceous normal superchron but could also reflect rapid sedimentation in a normal chron. The observed inclination gives a paleolatitude of ~50°S for the site at the time of magnetization. This is higher than the present latitude of the site and is consistent with the northerly motion of this part of the ocean floor as it moved away from the Antarctic-Australian spreading center to its present latitude.