wet,
is determined using the bulk mass, Mwet, and volume, Vwet.
The bulk density can be expressed as (assuming Vpw 
Vwater)Evaluation of physical properties at Site 1195 included nondestructive measurements of bulk density, bulk magnetic susceptibility, natural gamma radiation, and P-wave velocity on whole cores using the multisensor track (MST). P-wave velocity (x-, y- and z-direction) and moisture and density (MAD) were measured on split cores and plug samples. Color reflectance was measured on the archive halves of split cores. Thermal conductivity was measured on whole cores and semilithified core samples.
Bulk density at Site 1195 was computed from gamma ray attenuation (GRA) using unsplit cores, and from MAD measurements on plug samples. Bulk density decreases from 1.70 to 1.45 g/cm3 in the upper 56 m of Hole 1195A and Hole 1195B (Fig. F19). Bulk density abruptly increases to 1.7 g/cm3 at 90 mbsf, below which it generally increases down to ~345 mbsf. Between 345 mbsf and the base of the hole (521 mbsf), bulk density is rather uniform, with an average of 2.07 g/cm3. Small-scale variations in density occur down the entire hole. Composite profiles of independently derived MAD bulk density estimates have similar trends and amplitude ranges (Fig. F19). With the exception of a few outliers, the GRA bulk density agrees well with the MAD values between 0 and 210 mbsf. Cores from this interval were recovered using the APC. After 210 mbsf, cores were cut using the XCB. The discrepancy in bulk density values in this section could be a function of core diameter variations produced by XCB coring, problems with the GRA calibration, excessive drying of the core prior to sampling for MAD measurements, and/or mass loss during the sample drying and pycnometer measurements. Nevertheless, the repeatability of the MAD measurements suggests that the consistent difference between GRA and MAD densities most likely arises because of the variability in core diameter.
Grain density averages 2.89 g/cm3 but shows large scatter, ranging from 2.4 to 3.6 g/cm3 between 0 and 380 mbsf (Fig. F20A, F20B). Below this interval, the grain density is approximately constant at 2.7 g/cm3. Given that the density of calcite and dolomite is 2.710 and 2.866 g/cm3, respectively, grain density values approaching 3.6 g/cm3 are problematic. The abrupt change and tight grouping of grain density at depths greater than 380 mbsf (Fig. F20B) relate to the replacing of a dysfunctional pycnometer. This pycnometer apparently compromised the quality of dry volume estimates and thus the grain density between 0 and 380 mbsf. The replacement pycnometer was used to determine the volume estimates of cube samples between 380 and 450 mbsf. The suspicion concerning a dysfunctional pycnometer has been confirmed by remeasuring grain densities between 300 and 400 mbsf (Fig. F20B) using the replacement pycnometer. Good agreement exists between the repeat measurements of beaker samples between 300-400 mbsf and the tightly grouped cube sample density values determined for the 380- to 450-mbsf interval.
A quick analysis explains
how an erroneous dry volume measurement leads to an incorrect grain density even
though the MAD bulk density agrees with the GRA bulk density (Fig. F19).
Bulk density, wet,
is determined using the bulk mass, Mwet, and volume, Vwet.
The bulk density can be expressed as (assuming Vpw
Vwater)
wet
= Mwet/Vwet = (pwVpw
+ solidVsolid)/Vwet,
where Vsolid is the dry (solid) volume and Vpw is the fluid volume (including salt).
The differential change in bulk density relative to the grain density is
wet/solid
= Vsolid/Vwet,
and given that (1 - 
)Vwet
= Vsolid, where
is porosity, then the above equation can be written as
wet
= (1 - )solid.
From this relationship, it is clear that for large porosity (i.e., shallow within the section), the changes in bulk density are relatively insensitive to the changes in grain density. Such a relationship explains the observed agreement between GRA bulk density and MAD bulk density, even though the MAD-derived grain densities are clearly incorrect. As porosity decreases, however, the bulk density will increasingly be affected by volume measurement errors just like the grain density, which helps to explain the increasing scatter in the MAD bulk density and porosity with depth (<380 mbsf; Fig. F20A).
For homogeneous sediments
that are not overpressured, porosity may be approximated by an exponential
function of depth (e.g., Athy, 1930). For Site 1195, the porosity is relatively
low at the seafloor (60%-65%) and initially increases to ~80% at ~52 mbsf (Fig. F20A).
The low surface porosity may reflect surficial reworking, sorting and efficient
grain packing by oceanographic currents. This porosity trend may reflect
increased current activity starting at ~3.8 Ma (52 mbsf; see "Age
Model"). Below 52 mbsf, the porosity generally decreases
downhole. The general behavior of the porosity as a function of depth is broadly
consistent with Athy's relationship (see "Core Physical
Properties" in the "Explanatory
Notes" chapter). A least-squares estimation of these parameters
gives a surface porosity (o)
of 75.8% and a compaction decay constant (k) of 0.0025 m-1 (Fig. F20A;
correlation coefficient of 0.88). The inverse of the decay constant (400 m) can
be physically interpreted as the depth over which porosity is halved with
respect to the surface or initial value.
Porosity profiles generally reflect a combination of stress history and sedimentologic and diagenetic effects, such as variability in compressibility, permeability, sorting, grain fabric and cementation. Porosity is calculated from the pore water content of soft sediment samples and cubes of lithified sediment, assuming complete saturation of the wet sediment sample (Blum, 1997) (see "Core Physical Properties" in the "Explanatory Notes" chapter). The porosity curve mirrors that of the bulk density, with minor differences caused by variations in grain size, sedimentary facies and grain density (Fig. F20A). Porosity at Site 1195 shows a general decrease with depth. Superimposed on this trend are shorter wavelength variations over a length scale of 80-100 m. Because of the pycnometer problems, it is difficult to be confident about the geological significance of these trends. Nevertheless, it is interesting to note that abrupt changes in the downhole trend correlate approximately with lithologic unit boundaries and grain size distribution. Comparison of the porosity residuals, obtained by removing the least-squares porosity trend from the observed porosity, with the grain size distribution (see "Lithostratigraphy and Sedimentology") indicates, to first-order, a dependence of porosity on carbonate texture and/or grain size. For example, the change in grain size from wackestone to grainstone and back to wackestone between 0 and 100 mbsf correlates with the porosity residual variations shown in Figure F20C. Similarly, abrupt offsets in the porosity residuals, for example at 100, 125, 205, 265, 300 and 450 mbsf, are approximately coincident with both lithologic unit and subunit boundaries (e.g., 105 mbsf is coincident with Subunit IIA/IIB boundary; 125 mbsf is coincident with Subunit IIB/IIC boundary; and 265 mbsf is coincident with Subunit IIC/IIIA boundary).
The compaction properties
of carbonate sediments contrast with those of siliciclastic systems. Compaction
of sand siliciclastic systems is dominated by the deformation of the matrix. In
contrast, shale compaction occurs initially by the dewatering of the clays
followed by the deformation of the matrix, after which sand and shale systems
eventually compact in a similar way at depths 
1000
m. The carbonate equivalent of siliciclastic sands and shales are mudstones/wackestones
and grainstones/rudstones. It is suggested that the micritic mud supporting the
matrix of mudstones and wackestones, because it does not have an
intracrystalline porosity, responds to increasing overburden pressure by
directly deforming the matrix. This behavior is analogous to siliciclastic sand
systems. In contrast, grainstones and rudstones are grain or clast supported
sediments such that during compaction, the generally high porosity clasts (e.g.,
rhodoliths) need to deform first in order for the matrix to deform. This
compaction behavior is thus analogous to shale compaction. Alternatively, the
relatively higher porosity of grainstones may reflect a high initial porosity
that is maintained during the early compaction process. This analogy for
carbonate compaction helps to explain the correlation of residual porosity with
carbonate facies (Fig. F20C).
Compressional wave velocity was measured using the PWS1 (z-direction) and PWS2 (y-direction) insertion probe system on split cores (within the core liner), and the PWS3 contact probe system on both split cores (within the core liner) and ~9.5-cm3 cube samples of semilithified and lithified sediments. The cubes were used to measure velocity in the transverse (x and y) and longitudinal (z) directions. Continuing problems with the transducer-liner coupling of the MST P-wave logger forced its discontinuation during Site 1195 drilling.
Velocity obtained using the PWS3 contact probe, irrespective of the direction measured, shows abrupt shifts, for example between 0 and 35 mbsf, 85 and 113.5 mbsf, and 208 and 245 mbsf (Fig. F21). Similar problems were experienced at other sites (e.g., "Core Physical Properties" in the "Site 1194" chapter). These abrupt velocity steps appear to be caused by a pressure gauge linked to the PWS3 sensor. When activated, the acquisition software does not correct for movement of the gauge, thereby overestimating the calculated velocity. To correct the various shifts, all PWS3 x-direction data between 0-35 mbsf and 85-113.5 mbsf were decreased by 100 m/s, whereas the x-, y-, and z-direction data were decreased by 310 m/s between 208 and 245 mbsf (Fig. F21).
Corrected velocity values increase gradually from ~1558 to a maximum of 2250 m/s at a depth of 360 mbsf, after which the velocity is scattered around a mean of 2100 m/s (Fig. F21). This general behavior correlates with carbonate content (see "Geochemistry"), which is ~80% over a depth range of 0-300 mbsf. From 300 mbsf to the base of the hole, the carbonate content is characterized by large fluctuations superimposed on an overall decreasing trend toward basement. No simple relationship between the lithologic units and the velocity trends is observed (see "Lithostratigraphy and Sedimentology").
Velocity anisotropy is significant and ranges from -10% to 5% (Fig. F21) (see "Core Physical Properties" in the "Explanatory Notes" chapter). In general, positive anisotropy is more common in layered sediments, where sound transmission is relatively more efficient parallel to bedding rather than across. Dolomitization, since it represents chemical modification of the sediment structure, should be conducive to an isotropic velocity distribution. At Site 1195, most of the anisotropy is negative, indicating that the fast direction is vertical. Seismic anisotropy shows no obvious relationship with either dolomitization or lithologic units (Fig. F21; see "Lithostratigraphy and Sedimentology").
A crossplot of velocity vs. porosity for Site 1195 shows a general inverse relationship (Fig. F22). For the time-average empirical equation of Wyllie et al. (1956), the travel time of an acoustic signal through rock is the sum of the traveltime through the solid matrix and the fluid phase. However, the porosity and velocity data from Site 1195 do not match the time-average equation but can be described with a power law relation,
)
= a-b,
where VP is the compressional wave velocity, and a (63,119 m/s) and b (1.55) are empirical constants determined from a least-squares regression (correlation coefficient = 0.87; Fig. F22). Deviations from the velocity-porosity model do not show any simple relationship with either dolomitization or lithologic units.
Thermal conductivity values at Site 1195 show an overall increase with depth, ranging from ~0.8 W/(m·K) near the seafloor to ~1.4 W/(m·K) at the base of Hole 1195B (Fig. F23). Scatter tends to increase downsection, which is broadly consistent with the porosity data (Fig. F20A). A direct inverse relationship should exist between porosity and thermal conductivity because of the power law dependence of bulk thermal conductivity on the solid matrix grain thermal conductivity and the thermal conductivity of the interstitial fluid (see "Core Physical Properties" in the "Explanatory Notes" chapter) (Keen and Beaumont, 1990).
The variation in bulk thermal conductivity for the observed porosity range obtained at Site 1195 is shown in Figure F24. Except for a small number of outliers, the majority of the measured thermal conductivities lay between the shale and sandstone power law relationship, giving confidence in the viability of the observations. Given the predominance of carbonate through most of the section at Site 1195 (see "Geochemistry"), the observed thermal conductivity is consistent with the carbonate sediment facies mixed with clays and siliciclastics.
Magnetic susceptibility (MS) and natural gamma radiation (NGR) contain independent information concerning terrigenous sediment source and magnetic mineral derivation. In general, for Site 1195, NGR correlates with the MS and inversely correlates with the lightness (L*) parameter of color (Fig. F25). The MS and NGR can be divided into four main zones with distinct patterns (Fig. F25):
Carbonate content slowly decreases below ~220 mbsf (see "Geochemistry"), more or less as the MS and NGR values increase. The general decrease in lightness as both the MS and NGR increase between 0 and 80 mbsf and below 230 mbsf indicates increased terrigenous clay content (lithologic Units IIIA and IIIB; see "Lithostratigraphy and Sedimentology").
MS and NGR can be used to estimate the depth offset between Holes 1195A and 1195B by matching characteristic features in the data sets (Fig. F26). For example, the NGR shows a prominent peak at 36 mbsf in Hole 1195B (Fig. F25). The interpreted counterpart in Hole 1195A is at 36.3 mbsf, suggesting a minor 30-cm offset between the holes at this depth. This offset will not apply to other cores or necessarily to other sections in these cores. However, it represents the typical margin of uncertainty in depth measurements with the drill string. In contrast, the high-amplitude MS peak at 26-30 mbsf in Hole 1195B is not reflected in Hole 1195A although both holes contain the high-amplitude peak at 48-50 mbsf. Caution should be exercised if these MS data are to be used for cyclicity studies.
An important component of lithologic Subunits IIA, IIC, and IIIB, and Unit IV is the presence of glauconite, in places approaching 40% (see "Lithostratigraphy and Sedimentology"). Glauconite, rich in ferro- and ferric-ions, is potentially susceptible to induced magnetization. Glauconite is essentially a member of the mica and illite clay families and is thus a phyllosilicate. Consequently, glauconite may be able to scavenge radioactive minerals much in the same way as other clays and fine-grained micas. In order to examine the NGR and MS correlation with glauconite, the high-concentration glauconite zones straddling the lithologic Subunits IIC and IIIA (236-257 and 293-304 mbsf; Fig. F27A) were plotted with the MST-NGR and the Schlumberger hostile environment natural gamma ray sonde (HNGS) logging data (see "Downhole Measurements"). Superimposed onto the lithologic units is the a* color reflectance parameter (green-red value; Fig. F27A). The HNGS data are characterized by a number of peaks occurring between 234 and 256 mbsf. A local minimum in the a* occurs at 250 mbsf, superimposed on a broad low in the a* values between 230 and 290 mbsf. In detail, the a* local minimum does not correlate well with any of the NGR-HNGS peaks, although the general MST-NGR high is coincident with the general green emphasis of a* (Fig. F27A). In contrast, the high-glauconite zone recognized from the lithostratigraphy at 294-304 mbsf is not represented in either the a* or NGR-HNGS data. Furthermore, the maximum HNGS peak at 285 mbsf is associated with neither a local minimum in a* nor with a conspicuous glauconite layer in the core. A similar correlation problem exists with the MS (Fig. F27B). For the entire 235-320 mbsf section, the MS ranges in amplitude from 0 to 7x10-6SI units and shows no clear correlation with either the a* data or the location of the glauconite zones. It is concluded that glauconite does not show an appreciable MS effect, despite its potential as a strong paramagnetic mineral. Further, the association of the broad NGR-HNGS peak between 240 and 253 mbsf and the a* local minimum may reflect the existence of glauconite, but clearly, not every natural gamma radiation peak is necessarily associated with glauconite (see "Paleomagnetism").
