Our procedures for documenting the structural geology of the Leg 190 cores largely followed those given in the recent account by Kimura, Silver, Blum, et al. (1997), which in turn were based on the approach first developed at Nankai Trough during Leg 131 (Taira, Hill, Firth, et al., 1991). As during all legs, a persistent problem in describing structural features in the cores was the distinction between natural structures and those induced by the drilling process. It was often difficult to decide if a structure was natural, drilling induced, or a natural feature that had been distorted or exaggerated by the drilling process. Core disturbance is not reported here unless we judged that it had some natural significance, and in line with previous legs, we adopted a conservative approach. Also following the practice of previous legs, we were wary of steeply inclined fractures, especially where they lacked strongly and regularly developed slickensides, and we generally considered a feature to be largely of natural origin if it occurred entirely within an intact piece of core.
Each structure observed during Leg 190 was recorded manually on a structural description sheet. Figure F4 shows the sheet layout that we found useful. One innovation is the column headed "piece interval" for recording the length of the single intact piece of core in which a structure was contained. This information was needed during paleomagnetic reorientation (see below). Features such as a horizontal bedding plane have two identical interval depth values, whereas an inclined structure, such as a fault zone, has an interval top and bottom. The thickness of a dipping structure differs from the length of the interval in which it occurs and was therefore documented in a separate column labeled "thickness." We described structures in the cores using the list of deformational structures developed during Legs 131 (Taira, Hill, Firth, et al., 1991), 156 (Shipley, Ogawa, Blum, et al., 1995), and 170 (Kimura, Silver, Blum, et al., 1997), which were in turn based on the recommendations of Lundberg and Moore (1986). For the purposes of the visual core description forms, the structural identifiers were converted to the symbols listed in Figure F2.
Orientations of structural features were measured using the well-established protractor-goniometer method pictured in Figure F5 and explained in detail in the "Explanatory Notes" chapter of the Leg 131 Initial Reports volume (Shipboard Scientific Party, 1991). For a linear feature, such as sometimes seen on breakage surfaces within the core, a direct measurement of plunge and trend was usually possible, though it was often more convenient in practice to measure the orientation of a toothpick inserted into the core and aligned with the linear structure. In most instances, determining the orientation of planar structures required the measurement of two apparent dips, one being the intersection of the plane with the core face and the other some second intersection, commonly that seen in a surface cut perpendicular to the core face. Again, it was helpful in practice to measure the orientation of a toothpick inserted in line with this second intersection. The two apparent measurements were recorded in columns 7-10 of the structural description sheet (Fig. F4). The true spatial orientation was then derived on a stereographic projection program by finding the unique great circle that includes the two apparent dips. The derived value was added to the "true orientation in core" column on the description sheet.
This explicit reference to orientation within the core is necessary because the actual geographic orientation of most cores is not known, having been broken into pieces and differentially rotated during drilling. The core reference system we adopted was as follows. The length of the core was taken to represent vertical; hence, the direction at right-angles to the core axis was the horizontal, to which dip angles and plunges were referred (Fig. F6). The conventions for measuring azimuths have varied during previous legs. In some cases, probably because observations were being made chiefly on the archive half, the 000° azimuth was assigned to the back of that half of the core. Because we found it fruitful to deal with the structures on both the archive and working halves and convenient to be in line with paleomagnetic convention, we referred all structural data to a 000° core azimuth toward the double line at the back of the working half of the core. The back of the archive half was therefore the 180° azimuth.
All the numerical data and some commentary were extracted from the description sheets and entered into separate spreadsheets that are included as tables in each site chapter. From this structural database, information was extracted as necessary for statistical manipulation and to produce the various graphical representations of the structural geology presented in the site summaries. We put particular importance on attempting to restore the orientations of structures in the cores to their natural dispositions before drilling.
This further rotation, to true vertical and true north in the geographic reference frame, was remarkably successful during Leg 190. We made some use of the Tensor orientation tool but most of the reorientations were completed using natural remanent magnetization (NRM). The Tensor tool is routinely run with APC cores and provides a compass measurement relating declination from true north to the orientation of the double line on the core liner. Therefore, for APC cores, measured attitudes were reoriented to true geographic coordinates by adding the angle between core reference frame 000° (the double line ruled on the core liner) and true north to the azimuth in the structural measurement.
Because XCB and RCB drilling involve individual core segments differentially rotating, knowing the core liner orientation is not enough to restore the orientations. Instead, NRM declination data were used to reorient the individual pieces of core. We based our procedure on that detailed in Shipboard Scientific Party (1991, p. 44, table 6). However, rather than having to pass individual pieces of core through the cryogenic magnetometer specifically for this purpose, we made use of the routine data that are now readily accessible in the Janus database (see "Paleomagnetism").
We had already recorded the interval of an intact segment of rotated core (or several adjacent and clearly similarly oriented pieces) in which a structure of interest occurs as a "piece interval" on the structural description sheet (see above and Fig. F4). We then consulted the Janus paleomagnetic database to find the relevant points at which NRM declination measurements had been made as part of routine paleomagnetic procedures. Two or more determinations per piece are needed to give reliable results, and because the routine measurements are made 5 cm apart, this means that viable segments must be at least ~7 cm in length. Moreover, we considered as suitable only those pieces in which the paleomagnetic declinations were reasonably consistent: normally that was a difference <10°. A check on the integrity of this procedure was performed on several APC cores by making both Tensor and NRM corrections on the same structural data; the results were found to be very consistent, always agreeing within 10° and in most cases differing by less than 2°. The relevant NRM value extracted from Janus was entered into our structural spreadsheet. We added a simple algorithm based on the steps given in Shipboard Scientific Party (1991, p. 44, table 6) into the spreadsheet so that once the NRM declination had been entered, the true geographic orientation of the structure in question was automatically calculated and stored.
In line with the leg's focus on deformation and fluid flow, a permeameter was brought and operated by the shipboard structural group. The digital gas-probe permeameter, loaned by the Rock Deformation Research group of the University of Leeds, introduces compressed nitrogen at a constant mass rate through an injection tip pressed against the surface of a sample (Fig. F7). The measured mass flow rate is recorded by the computer and, at the ambient temperature and pressure, converted to a gas permeability. Although the instrument is designed primarily for the rapid evaluation of dry, coarse clastic sediments, pilot studies showed that systematic patterns were obtainable on water-saturated medium- to fine-grained sediments. While aboard ship, we found that although the geometry and physics of the nitrogen flow were unknown, reproducible but uncalibrated permeabilities were readily obtainable, including values down to 10-17 m2 and less. Therefore, with the ability of the instrument to rapidly obtain large numbers of closely spaced measurements along the cores, the permeameter was able to give a valuable reconnaissance view of the likely fluid flow behavior of the core materials at unprecedented resolution for ODP. However, it is unknown how the reported absolute values relate to true permeability.
A gas pressure of about 100 kPa sealed a soft neoprene disc against the split core face. Sediments below about 100 mbsf were stiff enough to withstand this sealing pressure without undue disturbance, and some measurements were possible on material from as shallow as 7 mbsf. Nitrogen was introduced through a 5-mm central hole in the disc, typically at pressures of a few tens of kilopascals or less in order to avoid internal disturbance to the sediment fabric. Equilibrium flow was usually established in a matter of seconds for coarser, more permeable sediments and after some minutes for materials at the lower part of the instrument's range. The software that accompanies the permeameter is designed for oilfield work and required some modification for ODP use, and the nature of the structure or lithology being probed had to be recorded manually. The computer-stored variables and calculated gas permeabilities were exported to an Excel spreadsheet and combined with the drilling depth information obtained from the ODP database in order to derive plots of the variation of apparent gas permeabilities with downhole depths. Although the permeability values, being based on nitrogen injection into a partially saturated medium, are at best semiquantitative, eventual laboratory-based "back-calibration" may lead to quantitative values and a detailed portrait of the influence of deformation structures and fine lithologic variations on the drainage of the prism sediments.