We drilled about 740 m of sediment at Site 900. However, only from 30.4 to 145.3 mbsf, which was composed of the nannofossil ooze and clay of Unit I (see "Lithostratigraphy" section, this chapter), were the recovered cores sufficiently magnetized to permit alternating field (AF) demagnetization and measurements using the cryogenic magnetometer. Below 151.3 mbsf (Cores 149-900A-18R through -79R), sediments are very weakly magnetized (generally between 1 × 10-4 and 2 × 10-4 A/m). Consequently, only natural remanent magnetization was measured in these cores. We measured sections of the basement rocks at 2- to 5-cm intervals, using the pass-through cryogenic magnetometer. Thus far, a total of nine discrete samples has been progressively AF demagnetized to verify the reversed magnetizations indicated by the cryogenic magnetometer measurements.
The quality of paleomagnetic data depends strongly on the lithology of the recovered material; thus, the discussion here of magnetic properties at Site 900 is organized on the basis of lithology. From Cores 149-900A-1R to -4R, we recovered mud-dominated turbidites. These cores were too mechanically disturbed by drilling to allow for meaningful measurements.
Cores 149-900A-5R through -17R are mainly nannofossil clay and ooze and have a strong magnetic signal (NRM intensities typically in the range 1 to 10 mA/m) that was easily measured. The whole-core pass-through measurements agree well with those from the discrete samples. Figure 17 shows typical examples of AF demagnetization diagrams for the discrete samples. The magnetic behavior of these sediments is relatively straightforward, with easily identified characteristic magnetizations in the orthogonal demagnetization diagrams.
Except for two short intervals, the sediments from 181.5 to 748.9 mbsf in Unit II (see "Lithostratigraphy" section, this chapter) were almost entirely characterized by weak magnetic remanence (<0.5 mA/m). We could not measure these sediments accurately on board the ship. We hope that land-based studies can successfully analyze the weak remanence in these sediments.
Below 748.9 mbsf, the recovered cores are fine-grained, metamorphosed mafic rocks (see "Igneous and Metamorphic Petrology and Geochemistry" section, this chapter) and have not yet provided any reliable magnetic results, mainly because of their weak magnetizations.
The best evidence for polarity reversals occurs in the interval from the upper part of Core 149-900A-5R to the top section of Core 149-900A-17R (30.4-145.3 mbsf). The combined shipboard paleomagnetic and biostratigraphic data (see "Biostratigraphy" section, this chapter) have enabled us to construct a tentative magnetostratigraphy for the sediments in this interval (Fig. 18). The Brunhes/Matuyama boundary was not observed in Hole 900A. However, because biostratigraphic samples between 12.0 and 65.5 mbsf have been tentatively assigned ages in the range from 0.8 to 2.6 Ma, the observed predominantly negative inclinations in this interval suggest that these sediments were deposited within the Matuyama Chron. Thus, the shift in polarity from reversed to normal at about 65.0 mbsf may represent the Matuyama/Gauss boundary (2.6 Ma). In addition, the preliminary biostratigraphic data suggest that sediments at 46.9 mbsf may be of late Pliocene age. This information suggests that the shift of polarity from reversed to normal at about 49.0 mbsf corresponds to the upper Olduvai boundary (1.88 Ma). Finally, the upper boundaries of both nannofossil Zone NN15 and foraminiferal Zone N19 have been tentatively placed at about 80 mbsf, which would suggest that the Gauss/ Gilbert boundary (3.5 Ma) should occur in the vicinity of this depth. As mentioned above, there are two exceptions within Unit II where the sediments are not too weakly magnetized. These occur in Cores 149-900A-51R (469.9 mbsf) and -77R (720.0 mbsf). Interestingly, pass-through measurements revealed that sediments at both these two depths are reversely magnetized. Although biostratigraphic evidence suggests that they may be of early Oligocene and early Eocene age, respectively, so far we have been unable to relate the reversed polarities to the geomagnetic reversal time scale.
The magnetic susceptibility of all cores for Hole 900A was measured at intervals of 3 cm (Fig. 19). In general, the whole-core susceptibility measurements vary in similar fashion to the NRM signals. For Unit I, the magnetic susceptibilities have average values consistently above 4 × 10-4 SI units, similar to those observed at earlier Leg 149 sites. Average susceptibilities of the weakly magnetized sediments of Unit II and the metamorphosed mafic rocks in the basement are only about 2 × 10-4 SI units. The two peaks of susceptibility at about 470 and 720 mbsf correspond to Cores 149-900A-51R and -77R, respectively, which also displayed relatively high magnetizations.