Volume-normalized magnetic susceptibility (k) on the JOIDES Resolution was measured at a spacing of 2 cm on Site 1075 whole-core sections using a Bartington Instruments MS-2 susceptibility meter. The sensing loop has an 80-mm diameter and uses an inducing field of 0.1 mT at a frequency of 565 Hz. The meter was set on SI units and the raw values were converted to SI volume-normalized susceptibilities.

We took 218 samples (wet volume = ~7 cm3) for detailed rock magnetic characterization from two depth intervals at Cores 175-1075-3H through 5H, between 14 and 32 mcd (0.21-0.09 Ma), and the older at Cores 175-1075A-15H through 16H, between 141 and 163 mcd (1.31-1.54 Ma). Sampling was conducted at one sample every 20 cm, or at ~1.3 and 2.0 k.y., respectively (Table T1). The comparison between samples from different holes is possible by using the meters composite depth scale, which was established by interhole correlation procedures during Leg 175 (Berger et al., 1998a; see also Hagelberg et al., 1992, for procedures). Samples from the Hole 1075B part of the splice are included in Table T1 but are not represented in the figures because of a mismatch in the composite depth section.

The low-field mass-normalized susceptibility (c) of the specimens was measured using a KLY-2 Kappabridge (Geofyzika Brno; inducing field = 0.5 mT) at the Institute for Rock Magnetism at the University of Minnesota. Hysteresis properties were determined on the Princeton Measurements Corporation vibrating sample magnetometer (microVSM) at the Institute for Rock Magnetism. An anhysteretic remanent magnetization (ARM) was imparted to each sample by exposing it to a constant 0.05-mT field and a slowly decaying 100-mT field. Partial anhysteretic remanence magnetizations (pARM) were imparted on 13 selected samples by switching on a direct current field between two specified values of alternating field. Grains with coercivity within that window acquire an ARM while the rest of the assemblage is demagnetized. By moving the window over a range of alternating fields, a pARM curve is obtained, representing the spectrum of coercivities in the sample (Jackson et al., 1988; Jackson et al., 1989). A 1-T isothermal remanent magnetization (IRM) was imparted with an ASC IM-10 impulse magnetizer and a backfield isothermal remanent magnetization (BIRM) imparted at -0.3 T. The acquired magnetizations were measured using the 2G three-axis superconducting rock magnetometer at the Institute for Rock Magnetism. Low-temperature demagnetization of the saturation isothermal remanent magnetization was conducted on a Quantum Design magnetic property measuring system. A 2.5-T field was imparted at 20 K and the magnetic moment was recorded in 5-K steps during the warm up from 20 to 300 K in zero field.

We used the XRF core scanner (Jansen et al., 1998; Röhl and Abrams, 2000) at the University of Bremen to determine the chemical element composition of the two depth intervals. The XRF core scanner is a nondestructive analysis system developed to scan the surface of split cores at high resolution. The central sensor unit consists of a molybdenum X-ray source (3-50 kV) and a Peltier-cooled Si (PSI) detector (KEVEX) with a 125-µm beryllium window and a multichannel analyzer with a 20-eV spectral resolution. The system is computer controlled and allows the analysis of elements from the atomic number 19 through 38 (K through Sr). The analyses were performed at predetermined positions and counting times. The X-ray source and the detector are lowered on the core surface during analysis; a slit defines the dimensions of the irradiated 1-cm2 core surface. The analyzed area is flushed with helium gas to avoid a loss of energy because of scattering in air. The XRF data of all split sections were collected at 5-cm intervals over a 1-cm2 area. A test-run calibration resulted in the use of a 15-s count time and an X-ray current of 0.15 mA to obtain statistically significant data of the investigated elements (K, Ca, Fe, Ti, and Mn). Although Sr and Cu measurements were also obtained, the counts at the selected counting time were too low to be interpreted.