Every other core from Site 1266 was recovered with a nonmagnetic core barrel until the first core barrel had to be drilled over (see Table T1; "Operations"). As at other sites, no obvious difference was noted in the magnetic data between sediments recovered with the nonmagnetic barrel and those with a standard core barrel. All APC cores in Holes 1266A and 1266C were successfully oriented with the Tensor tool with the exception of Cores 208-1266A-1H, 2H, and 9H and 208-1266C-6H.
The archive halves of 67 cores from Holes 1266A, 1266B, and 1266C were measured in the pass-through magnetometer. Natural remanent magnetization (NRM) was measured on all cores. Most cores were demagnetized at 10 and 15 mT. As at other sites, a strong vertical overprint was largely removed by demagnetization to 10 mT.
After reviewing data from previous sites, questions were raised regarding the quality of the data obtained from the archive halves. It was thought that the relatively soft and weakly magnetized carbonate sediments may have suffered from additional deformation and overprinting during splitting, resulting in a biased or poorly resolved polarity record. In an attempt to determine if core splitting significantly affected the pass-through data, a comparison test between measurements made on archive-half split cores and measurements made on whole-round sections was performed at Site 1266. All measurements from Hole 1266A were performed on the archive-half split cores, whereas those from Hole 1266C were made on whole rounds prior to splitting. This resulted in the demagnetization (up to 15 mT) of the working-half section (see "Paleomagnetism" in the "Explanatory Notes" chapter) (Table T10). Following splitting, a sampling of sections from Hole 1266C was also measured as archive halves, following demagnetization to 15 mT.
The results are somewhat ambiguous, but they suggest that at Site 1266 the effects of splitting were generally not significant enough to justify the demagnetization of the working half. The soft component of magnetization (i.e., measured before demagnetization and largely corresponding to the overprint) is generally stronger and more stable in the whole-round cores, suggesting that this soft component may provide a proxy for postcoring deformation. The characteristic remanence, however, does not appear to be significantly better resolved for Hole 1266C whole rounds. We should point out, though, that the inclination record from the Hole 1266A split cores is relatively clean and stable compared to those of Sites 1263 and 1264; a similar comparison experiment for one of these sites has shown different results.
A more direct comparison of the effects of splitting can be made by comparing sections that were measured both as whole rounds and as archive halves. Data from these sections suggest that splitting can affect both directional and intensity data. Of the twelve sections measured as both whole round and archive halves, only two showed significant differences in the inclination data (Fig. F22A, F22B). Section 208-1266C-18H-3 in particular showed significant inclination shallowing following splitting (Fig F22B). These sections did not appear to be any less firm or lithified than the remainder of the sections, which showed no significant change in inclination following splitting (e.g., Fig. F22C). The intensity data from the split vs. whole-round sections generally parallel each other, although often with an offset in absolute value. Even in cases where the inclination records are identical, this offset between intensity records is sometimes observed. This may be at least partially attributed to uneven core splitting. Cores are often not split evenly in half, meaning that the volume of the archive half is not exactly half that of the whole round. This could lead to an incorrect calculation of magnetization (a volume-normalized quantity) in the archive half, where a constant volume is assumed.
Intensities of initial NRM, including the vertical overprint component, are mostly on the order of 10–3 to 10–2 A/m (Fig. F23). After demagnetization to 15 mT, intensities are about an order of magnitude lower (Fig. F23) and exhibit similar downhole trends to initial MS values (Fig. F24).
The uppermost 75 mcd (lithostratigraphic Unit I) is characterized by low depositional remanent magentization (DRM) intensities (less than ~5 x 10–3 A/m at 15 mT), lower susceptibility, and higher DRM normalized by susceptibility (nDRM15 mT) (Figs. F23, F24, F25) when compared to the interval below. These higher values of nDRM15 mT suggest a change in ferromagnetic mineral properties, if we assume no change in geomagnetic field intensity. As discussed in "Paleomagnetism" in the "Site 1264" chapter, long-term trends or offsets in average nDRM15 mT may be useful as an indicator of diagenetic changes. These elevated nDRM15 mT values are also observed at all other Leg 208 sites (with an onset in the upper Miocene), although they are largely obscured by a hiatus at Site 1263. The increase in the derived nDRM parameter is reflected in relatively lower susceptibility and magnetic intensity and in some cases is additionally characterized by a decrease in the low-coercivity (0–15 mT) component. These observations may be explained by diagenetic dissolution of magnetite, but further measurements will be necessary to confirm this interpretation.
Magnetostratigraphic age-depth tie points are given in Table T10. The Pliocene was only cored in Hole 1266A, but we believe that all the major polarity boundaries are identifiable, although some appear to fall in core breaks (base of Chron C2An and the base of Chron C3n) or are not well resolved (top of Chron C2An).
Much of the upper Miocene is either condensed or missing. The placement of Chron C5n (Fig. F26B) was chosen to agree with the bottoms of biostratigraphic datums Discoaster bellus gr. (10.48 Ma) and Catinaster coalitus (10.79 Ma). The remainder of the Miocene is poorly resolved, and the placement of Chron C6Cn was chosen to correspond to the biostratigraphic placement of the O/M boundary.
The entire Oligocene polarity sequence appears to be reasonably well resolved in both Holes 1266A and 1266C (Fig. F26C), but some ambiguities remain in the preliminary interpretation. The lack of overlap between cores from the two holes leaves some boundaries unclear, and the biostratigraphic datums from this interval suffer from significant reworking. Chron C7n through the top of C8n is placed with some confidence, and the placement of Chron C13n is constrained with the tops of nannofossil datums E. formosa at 207.7 mcd and D. saipanensis at 221.4 mcd, as at all other sites. Between these two intervals, however, several interpretations are possible. If the age and placement of the top of nannofossil datum R. umbilicus (31.7 Ma) at ~200 mcd is correct, then Chron C12n should correspond to the short normal which is observed in both holes at ~194 mcd. This interpretation results in a normal polarity interval between ~205 and 208 mcd in both Holes 1266A and 1266C that does not correspond to the geomagnetic polarity timescale. Alternative interpretations would assign this normal polarity interval to Chron C12n or to part of Chron C13n but would result in dramatic changes in sedimentation rates throughout the lower Oligocene. Detailed shore-based sampling may help resolve some of these ambiguities.
Tentative identification of a few chrons in the upper Eocene was possible, but most of the Eocene is either condensed, missing, or poorly resolved. The lower part of Chron C24n appears to be resolved for Holes 1266B and 1266C. The transition from Chron C24n to C24r was placed between 319.6 and 320.4 mcd. If this placement is correct, it implies that the portion of Chron C24r above the P/E boundary is considerably condensed compared to the portion below the boundary. Below ~325 mcd, the inclination record from all holes is based solely on data from XCB cores and is not well resolved. However, a reversed to normal transition is observed in sediments from Holes 1266B and 1266C at ~364.5 mcd. Based on the biostratigraphy, this reversal was assigned to the base of Chron C26n.