METHODS

On the basis of shipboard descriptions (visual core descriptions, barrel sheets, smear-slide descriptions; Shipboard Scientific Party, 1996a, 1996b), 83 ash layers and ash-bearing sediments were sampled at one or two depth intervals (tops and bottoms of the layers), depending on layer thickness. Thin sections were made from all samples.

Texture, mineralogical and lithological composition, and alteration of all samples were studied in thin sections. Thirty-five samples containing fresh glass particles were selected for chemical analyses. Major-element compositions and volatile contents (S, Cl, and F) of fresh glass particles were determined using a CAMECA SX50 electron microprobe. The microprobe was operated at an accelerating voltage of 15 kV and a beam current of 10 nA, with 20 s on-peak + 10 s on-background counting times. To minimize sodium loss, felsic glasses were analyzed with the beam rastered (TV mode), scanning an area of 9 µm × 13 µm (magnification 10,000× ). The results (Table 2) show that under the chosen conditions sodium loss (as well as potassium loss) is negligible. Natural glass standards JDF-D2 (mid-ocean ridge basalt [MORB]-glass; C.H. Langmuir, pers. comm., 1993), ALV-981R23-5 (submarine basaltic glass; Fine and Stolper, 1986; Metrich and Clocchiatti, 1989), LIPARI (obsidian, Lipari, Cannetto Lami lava; Hunt and Hill, 1993), and KN 18 (rhyolitic glass; Nielson and Sigurdsson, 1981) were run for major-element analyses. SCAPOLITE USNM R6600-1 and SPHALERITE (synthetic zinc sulfide) standards were used to calibrate the electron microprobe for sulfur, SCAPOLITE for chlorine, and SI-F-APATITE USNM 104021 for fluorine. Additional basaltic and rhyolitic glass standards (e.g., VG-2) and interlaboratory monitors (e.g., ISL 89/15 Icelandic tholeiitic glass) were analyzed periodically to monitor the accuracy of the microprobe. We considered the effect of sulfur speciation (S⁶⁺ or S^2-) on the S-K spectral peak position, assuming sulfur to be S⁶⁺ in SCAPOLITE and S^2- in SPHALERITE in the analyzed samples (measured spectrometer positions corresponding to sin: SCAPOLITE = 61370; SPHALERITE = 61407; selection of the analyzed samples = 61404 ± 6). For the analyzed standards, the standard deviation (2) for major elements is <0.4 wt% (Table 2). Sulfur concentrations of 1160 ± 120 ppm (2) obtained for the standard ALV-98 correspond to the values of 1170 ± 90 ppm reported by Metrich and Clocchiatti (1989) and 1050 ± 140 ppm reported by Metrich et al. (1991).

To smooth out compositional heterogeneities, as many as 10 points were analyzed on each glass particle. The data presented here are averages of these analyses. As many as 32 individual glass particles were analyzed for each ash deposit.