METHODS

Over 60 samples of pillow basalts representative of all nine sites were prepared and analyzed for the mineralogic and petrographic studies by optical microscopy, electron microprobe, X-ray diffraction (XRD), X-ray fluorescence (XRF), loss on ignition (LOI), and quantitative ferrous iron analyses.

Identification of mineral species was achieved by means of optical examinations under microscope (transmitted and reflected light) and XRD analyses on concentrated powder, using a Philips PW3710 diffractometer at the University of Genova, Italy. Samples were run between 2.5° and 70°2, with a generator potential of 30 kV, a generator current of 22 mA (using a CuK radiation), a Ni filter, and a scan speed of 1°/min. For XRD of the clay-sized fraction, the material was scraped from surface coating, open fractures, veins, vesicles, and cavities using dental tools and, when necessary, concentrated by settling in a water column. In addition, the clay-sized samples were saturated with ethylene glycol for 12 hr and reanalyzed to help identify expandable clay minerals. The software used for XRD data reduction was Philips PC-APD Diffraction Software and MacDiff 3.0.6c. Some nonclay minerals were determined by single crystal or separated aggregates analyses using a DIFFLEX II diffractometer equipped with a Gandolfi Camera (CuK, Ni filter, 35kV/25mA, scan time = 6 hr).

Routine electron microprobe analyses were carried out with a Philips SEM 515 equipped with an energy-dispersive spectrometer (EDAX PV9100) at the University of Genova, Italy, and an ARL-SEMQ (WDS) microprobe at the University of Modena, Italy. Characteristic X-rays were detected using wavelength-dispersive spectrometers. Microprobe operating conditions were 15-kV accelerating voltage, 2- to 15-nA beam current, and 1- to 10-µm beam diameter (up to 25 µm for the glass and the clay minerals). Counting times of 120 s were reduced up to 60 s for clay minerals to prevent damage. Calibration was accomplished with a range of synthetic and natural standards.

The modal abundance of secondary phases was calculated by point counting (based on 500 cts per section) on about 100 thin sections used for shorebased and shipboard analyses. The results were further checked against comparison charts for visual percentage estimation.

Representative samples of the various lithologies were selected and prepared as fused beads for major oxide analysis by XRF on a Philips PW1400 with a Rh anode X-ray tube. The gabbro MRG-1 and the basalt AII-92-29-1 international reference standards were used to calibrate the XRF spectrometer before running any unknown samples.

On 30 samples representative of all nine sites, quantitative ferrous iron analyses were performed at the Udine University, Italy, by KMnO₄ titration on bulk rock powders after a HF + H₂SO₄ acid attack. The values represent the average of five separate analytical steps; the precision and the accuracy of this technique were checked periodically against the gabbro MRG-1 and the basalt AII-92-29-1 international reference standards, and analytical errors were always around ±0.1 wt%.

Shipboard analyses done aboard the JOIDES Resolution included standard petrographic study with quantitative point counting, mineral analyses by XRD, whole-rock analyses by XRF, and LOI (Davis, Fisher, Firth, et al., 1997). These have been used to complete and compare the shorebased data set.