SUMMARY OF RESULTS (continued)

Paleomagnetism
Paleomagnetism tends to be a difficult profession during carbonate legs. However, during Leg 189 we were fortunate enough to get some magnetostratigraphy at all sites, and at Sites 1170, 1171, and 1172 the paleomagnetic record was sufficiently good to permit construction of useful age-depth plots. Our work is described in terms of its three main aspects: (1) magnetostratigraphy, (2) rock magnetism, and (3) investigation of magnetic overprints and measurement difficulties.

Magnetostratigraphy
The interpretation of the paleomagnetic record at Sites 1168, 1169, and 1170 in terms of magnetostratigraphy was restricted to the Pliocene-Pleistocene and late Miocene. At Sites 1171 and 1172, the paleomagnetic records were adequate to generate an age-depth plot for the Neogene independent from biostratigraphic datums. At Site 1172, magnetostratigraphy was established down to the middle Miocene. Although the quality of the magnetic record deteriorated in older intervals, a relatively complete magnetostratigraphy was achieved, especially across the Eocene/Oligocene boundary. Unfortunately, at both sites the interpretation of the Eocene record from the inclination was problematic, which led to the development of an approach based upon the sign of the z component. This generated distinctive magnetostratigraphic boundaries, where they were almost totally obscured in the inclination record. The magnetostratigraphy for Leg 189 is summarized in Table 3, which gives the depths of the observed boundary chrons at the various sites. Where there was more than one core, the depths are from the best record.

Rock Magnetism
Throughout Leg 189, rock magnetism studies have revealed a remarkable homogeneity of magnetic material. Most discretes samples had anhysteretic remanent magnetization (ARM) intensities of order 10^-3 mA/m and isothermal remanent magnetization (IRM) intensities one order greater. We interpret the ARM/IRM ratio as typical for a detrital source. However, at Site 1168 this ratio was higher, suggesting that the magnetic record was carried in fine, single-domain grains probably of biogenic origin. At Sites 1170 and 1171, attempts to correlate magnetic properties (ARM/IRM, IRM 20/IRM, and IRM 500-IRM 200/IRM 500 ratios) with lithologic units were unsuccessful. Down to the Eocene, magnetic properties were very similar and seemed to characterize a similar source, as if magnetic inputs were not influenced by environmental changes. However, there were some changes weakly correlated with lithology, and the origin of the hard carriers observed in some sections remains to be determined from shore-based studies.

Measurement Techniques
The techniques used during this leg varied a little from the standard methods. For the most part, we measured unsplit sections instead of archive-half sections. This was carried out ahead of the multisensor track (MST) measurement. Because the paleomagnetic measurement is faster than the MST track, the paleomagnetic analysis of the unsplit section was conducted in dead time, and thus, core flow was improved. A long-standing problem in the paleomagnetism of ODP recovered sediments, and in particular in results from weak sediments such as carbonates encountered on this leg, has been the bias toward the "0" declination reading. The analysis of the drift of the 2G cryogenic magnetometer indicates that one possible explanation of this phenomenon is improper application of drift corrections, although this needs to be investigated in more detail onshore. The use of nonmagnetic core-barrel assemblies permitted comparisons between magnetic results from standard and nonmagnetic assemblies, which did not appear to make a major difference to the magnetization of recovered cores, although further analysis onshore is needed again.

In summary, during Leg 189 we encountered the usual difficulties of carbonate legs, but considerable magnetostratigraphy was achieved and aspects of measurement and analysis techniques have been improved. Of particular note are the measurement of unsplit sections to increase intensities and reduce potential of overprint from splitting and the utilization of the intensity of the z component to interpret the magnetostratigraphy.