A number of massive gas hydrate samples were recovered from Sites 994 and 997 on the Blake Ridge and from Site 996, an active vent site on the Blake Ridge Diapir (Fig. 1). Methane gas hydrates recovered from the Blake Ridge were ice-like, translucent white solids a few centimeters in diameter and more than 35 cm in length, whereas those from the Blake Ridge Diapir were friable vein-fillings, measuring a few millimeters thick, with near-vertical orientation (Paull, Matsumoto, Wallace, et al., 1996). After 5 to 10 min of observation and photographing on the catwalk, solid gas hydrates were hand-picked or cut from the sediment cores, transferred to pressure vessels, and stored in a freezer at -20ºC. The pressure vessels were shipped to Tokyo by air cargo from Miami and arrived in Tokyo within 36 hours after the end of the cruise.
One gas hydrate sample
from Site 994 and three samples from Site 997 were analyzed for 
18O.
These samples were well preserved, whereas vein-filling gas hydrate from Site
996 were apparently contaminated with pore waters and so were of no use. We
interpret that gas hydrates from Site 996 dissociated within the pressure
vessels during transport and storage.
Sample 2A, ~30 cm3 in volume, was selected from one of five large pieces of white, irregularly shaped nodules with soapy luster that were embedded in nannofossil-rich clay (Sample 164-994C-31X-7, 10-50 cm; 259.90-260.30 mbsf). The broken appearance of the hydrate and the disturbance of the host sediments suggest that the recovered samples were originally present as layers that were broken during coring and recovery. Occurrence of massive gas hydrate from this horizon seems to correspond to the zone of anomalously high electrical resistivity of the interval between 220 and 260 mbsf (Paull, Matsumoto, Wallace, et al., 1996).
Several large pieces of massive gas hydrate were recovered from Hole 997A (Sample 164-997A-42X-3, 0-55 cm (330.03-330.58 mbsf). The top 20 cm of the sample contained 10 to 15 angular fragments of white, fizzing hydrate, 3-10 cm in diameter, in nannofossil-rich clay that was disturbed by drilling. The next 27-cm section consists of a single piece of solid gas hydrate, which was broken into three pieces, 5, 7, and 15 cm long. Sample 14 was taken from the top 20 cm of the section (Sample 164-997A-42X-3, 0-20 cm; 330.03-330.23 mbsf), and samples 16A and 16B (Sample 164-997A-42X-3, 20-25 cm; 330.23-330.48 mbsf) and 16C (Sample 164-997A-42X-3, 25-32 cm; 330.48-330.55 mbsf) were taken from the top two pieces of the continuous gas hydrate section. The surface of these continuous gas hydrate samples was coated by thin (<1 mm), transparent shiny layer of water ice. The sediments immediately below the continuous hydrate sample contained small rectangular fragments of gas hydrate in slightly soupy, disturbed clay. Well logging conducted in Hole 997B, 25 m northwest of Hole 997A, observed anomalously high resistivity at 360-365 mbsf, 30-35 m deeper than thick massive gas hydrate horizon in Hole 997A (Paull, Matsumoto, Wallace, et al., 1996). If the high-resistivity zone corresponds to massive gas hydrate, this may suggest that the thick occurrence is not bedded, but perhaps fills an inclined fault.
Interstitial waters were
extracted from whole-round sections of sediments, 10 to 25 cm long, using a
Manheim hydraulic squeezer. These water samples were stored in flame-sealed
glass ampoules immediately after extraction. Thirteen samples from Site 994
(57.85 to 686.53 mbsf) and 21 samples from Site 997 (5.80 to 746.85 mbsf) were
selected for isotopic analysis of pore-water oxygen (18OH2O).
Each gas-hydrate sample was placed in a teflon-coated dissociation chamber to decompose at room temperature. Gas pressure within the chamber steadily increased and reached stability in about 10 min, then the gas was transferred to a small collection chamber for further analysis. The measurements include gas composition and carbon and hydrogen isotopic composition (the results are given in Matsumoto et al., Chap. 2, this volume). The volume of the residual water was measured and stored in flame-sealed glass tubes for isotopic analysis. Chloride concentration was also measured to estimate the amount of pore water that mixed with water derived from gas hydrate.
The standard CO2-equilibrium
method of Epstein and Mayeda (1953), as modified by Matsuhisa and Matsumoto
(1986), was used to measure the oxygen-isotopic composition of gas hydrate and
interstitial water. Water aliquots (1.0-1.5 mL; 28 to 42 mmol O2)
were sealed in a small flask with 0.10 mmol of CO2 of known isotopic
composition. The flasks were placed in a water bath (25.0ºC) and stirred for
15-20 hr so that oxygen-isotopic equilibrium occurred between the water and CO2
phases. After complete isotopic exchange, the CO2 gas was removed and
refined within a cryogenic vacuum line. The oxygen isotopic ratio (18O/16O)
of the equilibrated CO2 gas was determined using Finnigan Delta E and
Delta S mass spectrometers. The results are represented in per-mil delta
notation (18O,
)
relative to Standard Mean Ocean Water (SMOW). The standard deviation (2
)
of independent analyses was 0.01-0.05,
whereas the reproducibility of the measurements was ~0.10
as estimated from the duplicate analysis of CO2 gas samples that were
prepared from the identical water sample.
Chloride concentration of gas hydrate-derived waters were measured using an ion chromatograph (ICA-5000). Standard deviation of the measurements is ~2%.
