Hole 1256C, our single-bit pilot hole, was cored 88.5 m into basement, and Hole 1256D, the deep reentry hole, was cored 502 m into basement during Leg 206. Hole 1256D is located ~30 m due south of Hole 1256C (see Fig. F12). The basement/sediment interface was encountered at 251.8 and ~250 mbsf in Holes 1256C and 1256D, respectively, in a water depth of 3634.7 m. Recovery from the upper part of the basement in Hole 1256C was excellent, averaging 81.8% in Cores 206-1256C-5R through 11R (252.4-312.8 mbsf). The last two cores (206-1256C-13R and 14R) penetrated 18.3 m and recovered only 0.16 m, or 0.9% of the cored interval, because of blocked core catchers, resulting in an average recovery in basement for Hole 1256C of 61.3%. Coring of Hole 1256D began at 276.1 mbsf, or 26.1 m below the sediment/basement interface. The average recovery from 276.1 to 350.3 mbsf (92.8-m interval) in Hole 1256D was 93%. Recovery over the total cored interval of 475.9 m in Hole 1256D averaged 47.8%.
We drilled a massive ponded flow (~32 to >74 m thick) in both Holes 1256C and 1256D, which is a clear marker unit for correlation of the igneous stratigraphy between holes. Therefore, we describe the igneous rocks from both holes together in this section. A summary of the igneous stratigraphy is presented in Figure F56. We divided the basement into 22 units in Hole 1256C and 26 units in Hole 1256D (Tables T26, T27). The igneous basement is dominated by thin (tens of centimeters to ~3 m) basaltic sheet flows separated by chilled margins, which make up ~60% of the recovered interval in both holes. Massive flows (>3 m thick) are the second most common rock type in both holes, including the thick ponded flow near the top of each hole. Minor intervals of pillow lavas (~20 m) and hyaloclastite (a few meters) were recovered in Hole 1256D.
Groundmass grain size varies throughout the section from glassy in the hyaloclastite and in chilled margins of flows to cryptocrystalline (<0.1 mm) or microcrystalline (0.1-0.2 mm) in the interiors of most of the sheet flows to fine-grained (>0.2 mm) in the interior of the massive ponded flow. The basalts are mostly aphyric, with minor intervals of sparsely olivine- or plagioclase-phyric lavas.
We describe the igneous units below in more detail based on hand-sample and binocular microscope examination, thin section examination, and shipboard major and trace element data obtained by inductively coupled plasma-atomic emission spectroscopy (ICP-AES).
The upper 27 m of basement in Hole 1256C (251.8-278.7 mbsf) is composed of thin basaltic sheet flows a few tens of centimeters to ~3 m thick, separated by chilled margins and containing rare intervals of recrystallized sediment (Fig. F56). The basalts are aphyric to sparsely phyric with plagioclase, olivine, and clinopyroxene phenocrysts (in order of decreasing abundance). The groundmass is cryptocrystalline to microcrystalline in flow interiors, decreasing in average grain size toward the margins (Fig. F57) and preserving glassy margins in some cases (Fig. F56; Table T28).
Units 1256C-18 and 1256D-1 each consist of a single cooling unit of cryptocrystalline to fine-grained basalt, which we interpret as a ponded lava flow. A total of 32 m of this unit was cored in Hole 1256C, of which 29 m was recovered. In Hole 1256C the top of the unit was encountered at 280.27 mbsf (calculated based on piece lengths from the top of Core 206-1256C-8R to the top of Unit 1256C-18) and consists of ~75 cm of cryptocrystalline to glassy aphyric basalt representing the deformed surface of the flow (Fig. F58). The groundmass grain size increases toward the interior of the flow, becoming microcrystalline within 2 m of the top and gradually increasing to fine grained ~5 m from the top. The groundmass remains fine grained throughout the interior of the flow (Fig. F59) until it abruptly decreases to cryptocrystalline (over ~15 cm; interval 206-1256C-11R-6, 120-135 cm) ~1.5 m above the base of the flow. Because of the relatively coarse groundmass, no phenocrysts were apparent in hand-sample or binocular microscope examination, although they can be distinguished in thin section (dominantly olivine with trace plagioclase and clinopyroxene) (see "Petrography" below). The base of the flow has been deformed and recrystallized, probably during and shortly after emplacement (Fig. F60).
The flow is much thicker in Hole 1256D than in Hole 1256C. An exact thickness cannot be calculated because the top of the ponded flow was drilled out prior to setting of the casing and was not cored. The first core in Hole 1256D (Core 206-1256D-2R) consists of aphyric microcrystalline basalt rather than the deformed cryptocrystalline flow top recovered in Hole 1256C. Nevertheless, the minimum thickness of 74.2 m of the flow in Hole 1256D can be determined from the cored interval (276.1-350.3 mbsf), of which 68.4 m was recovered, and is more than twice that in Hole 1256C. Despite this increase in overall thickness, the groundmass grain size in the interior of Unit 1256D-1 does not exceed 1 mm and thus remains microcrystalline to fine grained. We correlate Units 1256C-18 and 1256D-1, despite this difference in thickness, based on the similarity in mineralogy (both in hand sample and in thin section), general appearance, similarity in the depth of the top of each unit, and geochemical similarities between the two units (see "Hard Rock Geochemistry"). Although the top of Unit 1256D-1 is somewhat shallower than the top of Unit 1256C-18 (above 276.1 mbsf, compared to 280.27 mbsf in Hole 1256C), we interpret the two units to have been parts of a single ponded lava flow where the interior at the locations of both holes was liquid at the same time. The dramatic increase in thickness over 30 m of lateral distance suggests steep paleotopography, with Hole 1256D deeper in the depression that was filled in by the flow.
Massive Lavas and Thin Sheet Flows. The remainder of the section in Hole 1256D (with the exception of Units 1256D-3, 4a, 4c, 8c, 16d, and 21) consists of sheet flows tens of centimeters to ~3 m thick with uncommon massive flows 3.5-16 m thick (see Table T27). These flows are aphyric to sparsely phyric cryptocrystalline to microcrystalline basalts and are distinguished by chilled margins (Fig. F61) or by increasing grain size toward the interiors of flows where the margins were not recovered. In the lowermost units (Units 1256D-19 to 26), we grouped sequences of flows that were mineralogically similar and did not break out separate subunits for each cooling unit. An example of a chilled margin is shown in Figure F61, and the locations of glass and altered glass in the hole are compiled in Table T28.
Pillow Lavas. We distinguished pillow basalts from thin sheet flows based on the presence of curved glassy margins oblique to the sides of the core and radial pipe vesicles oriented perpendicular to the chilled margins (Fig. F62). We cored one ~20-m-thick interval of pillow basalts near the top of the section (Unit 1256D-3), with ~8.5 m recovered. The pillow basalts consist of aphyric to sparsely phyric cryptocrystalline basalts with glassy chilled margins. Additional pillow basalt intervals will probably be recognized from analysis of downhole FMS and UBI data (see "Downhole Measurements"; Fig. F194).
Volcaniclastics. We recovered hyaloclastite intervals in two places in the basement: interval 206-1256D-21R-1, 80-146 cm, through 21R-2, 0-50 cm (Subunit 1256D-4c) (Fig. F63), and interval 51R-1, 85-150 cm, through 51R-2, 0-134 cm (Unit 1256D-21) (Fig. F64). These intervals consist of angular to rounded clasts of basaltic glass several centimeters to <10 cm in diameter and smaller (<1 cm) curved shards of glass within a matrix of altered glass. We also recovered an interval of volcanic breccias composed of angular fragments of cryptocrystalline basalt embedded in a matrix of altered glass (Subunit 1256D-4a) (Fig. F65). A fragmented interval along a subvertical crack that extends across several pieces (Section 206-1256D-21R-1 [Pieces 16-20]) (Fig. F66) was also recovered, and we interpret this as brecciated material filling a crack that opened on a flow top. Other types of breccia encountered in the section are noted in Figure F56 but are tectonic in origin and are described in "Structural Geology".
Dikes. We cored one intrusive contact between a dike and microcrystalline host basalt in interval 206-1256D-32R-2, 110-120 cm (Fig. F67). One other potential dike was encountered in interval 206-1256D-26R-2, 86-95 cm, although the contact relationships are not as clear.
The mineralogy, grain size, and modal analyses of thin sections are recorded in the thin section description sheets and summarized in Table T29. A complete list of digital photomicrographs available from the ODP data librarian is given in Table T30. Because most samples are glassy to aphanitic lavas with cryptocrystalline to microcrystalline grain sizes, modal analyses of these samples were done only with respect to phenocryst abundances based on counting 2000 points per thin section. For the coarser-grained lavas from Units 1256C-18 and 1256D-1, the groundmass mineral modes for selected samples were determined based on counting 1000 points.
The basalts show a large variation in grain size and texture, from holohyaline in the outermost chilled margins of lava flows and hyaloclastite clasts, through aphanitic groundmass, consisting of cryptocrystalline varioles, to the coarser intergranular textures in the lava pond. The basaltic lava is dominantly aphyric to sparsely phyric, with 72% of the examined thin sections having <5 vol% phenocrysts and 18% being aphyric (Fig. F68). The modal peak for Site 1256 is slightly shifted toward higher abundances of phenocrysts when compared to sheeted dikes from Hole 504B, which formed at an intermediate spreading rate but which are also dominantly aphyric (Dick, Erzinger, Stokking, et al., 1992). In contrast, flows and dikes from the slow-spreading Mid-Atlantic Ridge show bimodal phenocryst abundance with peaks at <10 and >18 vol% (Bryan and Moore, 1977; Hekinian, 1982; Hodges, 1978; O'Donell and Presnall, 1980; Sato et al., 1978; Shipboard Scientific Party, 1988).
Phenocrysts are dominantly olivine (average = 68% of the phenocrysts in Hole 1256C and 70% in Hole 1256D) with subordinate amounts of plagioclase (average = 31% of the phenocrysts in Hole 1256C and 25% in Hole 1256D) and <1.7 vol% clinopyroxene phenocrysts (average = 5.4% of the phenocrysts in Hole 1256C and 4.7% in Hole 1256D). Most clinopyroxene is augite, but rare pigeonite is present as a prismatic intergrowth with augite, resembling the groundmass pyroxenes in the coarser-grained lava from Units 1256C-18 and 1256D-1. Spinel was identified as tiny inclusions in completely altered olivine phenocrysts in a few samples.
The ratio of the three phenocrystic phases (clinopyroxene, plagioclase, and olivine) is shown in Figure F69 compared with the sheeted dikes from Hole 504B and lava flows and dikes from the Mid-Atlantic Ridge. Nearly 50% of the basalts from Site 1256 plot on the plagioclase/olivine join. Hole 504B dikes have a slightly higher proportion of clinopyroxene than the Site 1256 lavas. In contrast, the majority of the Mid-Atlantic Ridge lavas and dikes plot on the plagioclase/olivine join. With respect to both the phenocrystic abundance and proportions, the Site 1256 lavas are intermediate between these two extremes.
Olivine shows a seriate grain size variation and is as small as 0.02 mm in diameter in some samples. Although it forms euhedral crystals that seldom include other groundmass minerals, some olivines have outer rims that partially include plagioclase laths, suggesting some crystallization of olivine at the groundmass stage. However, complete alteration of olivine and its surrounding glassy matrix to clay minerals (saponite and celadonite) makes it hard to conclusively identify the presence of groundmass-stage olivine. Because of this, olivine is regarded only as a phenocryst phase.
Olivine is the most common phenocryst phase (0.1-11 vol%), but fresh olivine was found only in fresh glass of some chilled flow margins and hyaloclastite (Fig. F70A). Although it has been completely replaced by alteration minerals in most samples, the presence of olivine is deduced from the pseudomorph outline and textural relationships to surrounding minerals. Most olivine phenocrysts are euhedral equant crystals that rarely occur in clusters with other mineral species. Larger olivine crystals (up to 1.2 mm in diameter) from lower units of Hole 1256D (Unit 1256D-18 and lower) tend to be skeletal and coexist with more equant, smaller crystals (Fig. F70B). Rare tiny inclusions of spinel were found in some altered olivine. Units 1256D-23 and 24 contain olivine phenocrysts in crystal clots with reversely zoned plagioclase and with or without augite.
Plagioclase is the second most abundant phenocryst phase in abundances up to 6 vol%. The average size of plagioclase phenocrysts ranges from 0.1 to 0.8 mm in length, and some are up to 2 mm long. They are mostly subhedral to euhedral crystals present in clots with either other plagioclase crystals or with clinopyroxene, though euhedral platy to stubby discrete crystals are also present.
Most plagioclase phenocrysts from Hole 1256C and the upper units of Hole 1256D are unzoned or have minimal normal and/or oscillatory zoning. Zoning is more common in lower units from Hole 1256D (igneous Units 1256D-18 through 26) and may be either normally or reversely zoned or both. When zoning is apparent, the plagioclase crystals can be subdivided into a core, mantle, and rim. Normally zoned plagioclases commonly have clear cores with euhedral outlines, surrounded by thin and less calcic rims. Inclusions are uncommon except for some glass blebs and clinopyroxene aligned along twin planes of the host plagioclase (Fig. F71A). Much less common is plagioclase with resorbed cores, mottled with bleblike inclusions of clinopyroxene, magnetite, and glass and enclosed in a less calcic mantle. In contrast, reversely zoned plagioclases have dusty resorbed cores with euhedral, more calcic mantles enclosed by sodic rims. The dusty inclusions are small skeletal to dendritic magnetite, tiny acicular clinopyroxene, and thin plagioclase laths and pale brown glass mostly altered to clay minerals. These inclusions are distributed in the cores of the phenocrysts or commonly concentrated in bands between the core and mantle. Dusty cores of plagioclase tend to be more altered and replaced by clay minerals than the normally zoned clear cores. In one thin section (Sample 206-1256D-55R-2, 63-64 cm), both normally and reversely zoned plagioclases form clots with clinopyroxene and olivine; however, reversely zoned plagioclase cores are seldom in direct contact with olivine and clinopyroxene in the same clots (Fig. F71B).
Although clinopyroxene is the least abundant phenocryst phase (<0.17 vol%), it is present in more than 40% of the thin sections examined. The most common variety is black to dark green augite in hand specimen and pale green brown to colorless in thin section. The other variety is a pale yellowish green prismatic pyroxene that resembles the groundmass pigeonite in the coarse basalt lavas from Units 1256C-18 and 1256D-1.
Augite phenocrysts are typically subhedral to euhedral, stubby to short prismatic crystals 0.1-0.7 mm long. Most augite phenocrysts are small (0.15-0.25 mm in diameter) and commonly form crystal clots with platy plagioclase. In finer-grained samples, such phenocrystic augite is readily distinguished from the groundmass clinopyroxene. However, augite in coarser-grained samples shows more or less seriate grain size variation and changes in form from granular through short prismatic to more equant, stubby crystals as the groundmass grain size increases, making it difficult to distinguish phenocrysts from the groundmass. Most phenocrysts are either unzoned or only slightly zoned where simple zoning patterns are the most common, but sector zoning is also observed. Some large augite phenocrysts show patchy irregular extinction, suggesting grain growth through the coalescence and crystal lattice reorientation of smaller groundmass clinopyroxene crystals. Much larger phenocrysts (1-2 mm in length) tend to be more euhedral and equant and subophitically to ophitically include platy to stubby plagioclase (Fig. F72). These phenocrysts are almost free from inclusions except plagioclase.
Pale yellowish green pyroxene is present as rare crystal clots in hand specimen and as colorless prismatic crystals under the microscope. Although we could not observe a conoscopic figure because of the scarcity of this pyroxene in thin section, the crystal habit, colorless elongate prismatic forms, development of cleavages perpendicular to the c-axis, and intergrowths sandwiched by augite lamellae are strongly reminiscent of the groundmass pigeonite in the ponded basalt lavas that are described below.
Constituent Minerals. The groundmass is composed of plagioclase, augite, and magnetite, with or without minor amounts of glass, pigeonite, green clinopyroxene, quartz, apatite, and granophyric intergrowths of quartz and sodic plagioclase. Dark green clinopyroxene is present as the outermost rims surrounding augite and pigeonite and as a discrete prismatic crystal in the coarser-grained basalt from igneous Units 1256C-18 and 1256D-1. It is presumably of late magmatic origin and is described in "Alteration". The groundmass is composed primarily of plagioclase and clinopyroxene (augite and pigeonite) with the plagioclase slightly more abundant than the clinopyroxene in fine-grained parts of the lava pond, whereas the proportions appear to be reversed in cryptocrystalline to microcrystalline basalt flows. However, the very fine to cryptocrystalline groundmass grain size makes it very difficult to determine exact proportions of plagioclase to clinopyroxene with an optical microscope.
Augite in the groundmass shows a variety of textures that correspond to different cooling rates or the degree of undercooling. Type 1 is cryptocrystalline aggregates of fibrous crystals (Fig. F73A); Type 2 is comb-shaped or sheaflike plumose crystals (Fig. F73B); Type 3 is granular-acicular subhedral to anhedral crystals (Fig. F73C); and Type 4 is prismatic to stubby euhedral to subhedral crystals (Fig. F73D). Two or more of these textural types commonly coexist in a single sample. In the chilled margin of lava flows, the texture of the groundmass augite changes from completely glassy through Types 1 and 2 to Type 3 in a short distance (1-5 cm) likely because of a large thermal gradient caused by drastic chilling of the flow surface under water. Even in the center of a thick sheet flow, textural Types 3 and 4 may coexist, indicating varying degrees of undercooling during the groundmass crystallization.
Pigeonite is much less abundant than augite and is typically colorless elongate prismatic crystals with well-developed cleavages perpendicular to the prisms. It commonly forms intergrowths with augite, where prismatic pigeonite is sandwiched between pale brown-green augite lamellae. In coarser-grained samples, pigeonite can easily be identified but is difficult to identify in very fine-grained to cryptocrystalline samples.
Groundmass plagioclase is present as very tiny acicular crystals, thin planar laths, curved very thin planar crystals, skeletal fan-shaped or bowtie-shaped crystals in finer-grained samples, and platy and more stubby crystals in coarser-grained samples. Individual samples may contain one or more textural types as seen with the groundmass augite. These textural types change roughly in accordance with the augite crystallization type.
Fe-Ti oxide minerals are a ubiquitous groundmass phase. They typically form dendritic chains to equant skeletal crystals 1-30 µm across in cryptocrystalline to microcrystalline basalts and equant polyhedral crystals up to 0.2 mm in some fine-grained basalts. In coarser-grained samples, host magnetite has exsolved ilmenite lamellae during cooling.
Fine-grained and some microcrystalline basalts contain interstitial mesostasis of granophyric to vermicular quartz-albite intergrowths, quartz, granular to prismatic clinopyroxene (mostly altered to secondary clay minerals), acicular apatite, and dendritic to skeletal magnetite (Fig. F74). The intergrowths may be concentrated within and near late magmatic veins in the coarse basalt lavas from igneous Units 1256C-18 and 1256D-1, which are composed of the above mesostasis minerals, green clinopyroxene, and a trace amount of brown mica. Similar mesostasis is reported from the sheeted dikes in Hole 504B off Costa Rica (Shipboard Scientific Party, 1992) and lavas from Sites 417 and 418 in the western Atlantic (Sinton and Byerly, 1980).
Groundmass Texture. Most basalt lavas from Site 1256 have a common groundmass texture consisting of clinopyroxene and plagioclase ± disseminated magnetite. Both clinopyroxene and plagioclase are radially arranged to form spheroidal or fan-shaped crystal aggregates, finer-grained varieties of which have been traditionally called varioles (MacKenzie et al, 1982; Bryan, 1972).
Devitrification of a quenched flow surface begins within a few millimeters of the surface as patches of fine plumose crystallites, predominantly clinopyroxene, form. The number and density of the crystallite patches and the crystal density in individual patches rapidly increase toward the flow interior and occupy the entire groundmass with dark brownish varioles that are barely translucent under the microscope. Many varioles near the front of the dense variole zone have tiny microlites of plagioclase or olivine in their cores and are elongated to the orientation of the microlites, showing that they nucleated on the preexisting microlites. Typical varioles within a zone 1-2 cm beneath the surface are composed of very fine grained fibrous aggregates of clinopyroxene with minor amounts of thin plagioclase laths. Tiny magnetite grains may be disseminated in and between the pyroxene-plagioclase aggregates. Samples taken from farther into the interior of lava flows show preferential growth of clinopyroxene over plagioclase, where clinopyroxene forms larger curved sheaves of acicular crystals with the intergrown plagioclase laths much less abundant ("fine varioles") (Fig. F73A, F73B). As the groundmass crystals coarsen, both plagioclase and clinopyroxene thicken to form more stubby and prismatic crystals. Eventually the growth rates of plagioclase and clinopyroxene are reversed, and farther into the interiors of thick massive flows plagioclase becomes skeletal to platy with bowtie-like crystals larger than clinopyroxene forming subhedral to euhedral prisms broadening toward the exterior of varioles ("medium varioles") (Fig. F73). There is also a transitional type that has both plagioclase and clinopyroxene crystals similarly developed in the cores of thick sheet flows and massive lava.
Further development of variolitic texture is only observed in the very thick massive lavas from igneous Units 1256C-18 and 1256D-1. Stubby clinopyroxene coagulates near the center of a larger plagioclase crystal to form a hub, from which a few crystals of platy plagioclase radiate ("coarse varioles") (Fig. F73D). The variole hub is usually composed of several clinopyroxene crystals, but in some cases only one or two clinopyroxene crystals showing a patchy extinction pattern are intergrown with radiating plagioclase crystals. This suggests that gradual coalescence and reorientation of several clinopyroxene crystals into a single continuous crystal took place in the slowly cooling lava. An extreme variety can be seen in the coarsest part of the lava pond where large plagioclase crystals poikilitically enclose clinopyroxene crystals with different crystallographic orientations (Fig. F75).
The distribution, morphology, and size of varioles appear to change in response to the distance from the chilled margin for thin sheet and massive lava flows, which is the main factor that determines the degree of undercooling. In the core of lava flows or lava ponds, fine and medium varioles tend to form as randomly spaced isotropic spheroids, consistent with crystallization under static conditions. However, when varioles grow under differential movement beneath the deforming surface crust of flowing lava, varioles may be flattened and oriented subparallel to the shear plane or else disintegrated into bands of granular pyroxene and platy plagioclase intercalated with flattened variole-rich bands.
Unit 1256D-10 is a 10.8-m-thick (minimum = 7.25 m thick) massive lava unit composed of sparsely clinopyroxene-plagioclase-olivine-phyric cryptocrystalline basalt with a medium variolitic groundmass. Although it does not have any intervening flow boundaries, there are abrupt variations in grain size through this flow, suggesting at least three intraflow cooling units (see "Site 1256 Core Descriptions"). Two grain size minima are encountered in Sections 206-1256D-37R-2 (Piece 3) and 37R-3 (Pieces 2 and 3). In interval 206-1256D-37R-2, 14-18 cm, light gray discontinuous aphanitic bands run oblique to the core and represent chilled flow lobe margins that were incorporated into the interior of the flow. The bands are 1-2 mm thick and 5-6 cm long with undulating outlines (Fig. F76A). Microscopic examination shows that the bands are composed mainly of granular clinopyroxene and magnetite with sporadic plagioclase laths. The bands consist of finer- and coarser-grained domains; dark finer-grained domains show folding and flow structures embedded in light coarser-grained matrix. Dark patches appear to be sheared and disrupted surface crust mingled with coarser-grained lava interior. Folding and attitude of inclined plagioclase laths are consistent with the flow structure seen in the upper lavas, where the groundmass plagioclase laths show sheared bands with left lateral movement in the partially disrupted variolitic matrix (Fig. F76B).
The groundmass texture of the band is apparently not of igneous origin. Magnetite grains are mostly equant polyhedra 10-17 µm in diameter and rarely show skeletal forms. Clinopyroxene grains, ~5-6 and 9-12 µm in diameter in the fine and coarse domains, respectively, are almost equigranular in each domain. The grain boundaries of clinopyroxene show nearly 120° intersections. Plagioclase longer than 20 µm preserves igneous lath shapes, but smaller grains (<5 µm) have polygonal equant forms like clinopyroxene. The upper lavas show a fine-grained quench texture in contact with the band, and plagioclase laths grow perpendicular to the contact (Fig. F76C). In contrast, the lower lavas do not show any variation in grain size away from the contact and have well-developed spheroidal varioles, suggestive of crystallization under static conditions. The groundmass plagioclase and clinopyroxene in direct contact with the band are surrounded by very fine rounded plagioclase laths and granular clinopyroxene that appear to be remnants of original groundmass crystals being recrystallized into more equant crystals (Fig. F76D).
These textures can be interpreted as recrystallized chilled margins of a lava flow that were incorporated into the flow interior when a later flow lobe was emplaced onto this flow lobe. The overlying lava was quenched against the lower contact. Subsequent flow of the upper lava caused shear deformation of its solidifying lower margin and probably also of the underlying chilled margin atop the lower lava. After emplacement of the upper lava and incorporation into the inflating sheet flow, the duration of heating was sufficient for recrystallization of the chilled margin to take place. The lower lava does not show any gradual grain size variation near the chilled margin and crystallized under static conditions, suggesting that some of the chilled margin was eroded away prior to emplacement of the upper lava. Crystallization of the lower lava in contact with the recrystallized margin occurred after the lava supply ceased and the flow interior became static. Flow lobe margins incorporated into the flow interior because of coalescence of inflating flow lobes can be seen in subaerial pahoehoe sheet flows and tumuli in Hawaii (Umino et al., 2002) and are found in submarine sheet flows and tumuli from the Oman Ophiolite (S. Umino, unpubl. data, 2002).
Unit 1256C-18 is a >30-m-thick massive lava body that begins at 280.3 mbsf with a holohyaline-cryptocrystalline lava surface and develops downhole into intergranular to coarse variolitic fine-grained massive basalt and to cryptocrystalline recrystallized basalt at the bottom (312 mbsf). The uppermost 0.7 m is composed of aphanitic lava with folded glassy chilled margins and volcanic rubble, including glassy clasts, and is interpreted as the folded and jumbled surface crust of a lava flow. The basal 1.6 m is unusual aphanitic basalt consisting of recrystallized variolitic groundmass and late magmatic veins that shows recrystallization and deformation textures described below (see "Recrystallization of the Base of Unit 1256C-18"). A similar fine-grained massive lava was encountered in Hole 1256D (Unit 1256D-1) from the first core at 276.1 mbsf, ~4 m above the top of Unit 1256C-18 and continued downhole to 350 mbsf. This unit is lithologically correlated to the thick massive lava unit in Hole 1256C, but it is much thicker (>74 m thick) and lacks both the quenched upper surface and the basal recrystallized basalt. A thick lava flow could potentially have formed either as a lava pond, where rapidly delivered lava accumulates in a depression, or as an inflated sheet flow, with slowly delivered lava confined by its own chilled margin. We interpret these massive basaltic units to be a thick ponded lava and not an inflated sheet flow on the following grounds:
In Hole 1256C where almost the complete succession of the lava pond was recovered, the basalt is a macroscopically homogeneous, massive microcrystalline to fine-grained lava with a coarse variolitic to intergranular texture (apart from the uppermost 0.7 m and the basal 1.7 m). Neither the top nor the base of the lava pond was recovered in Hole 1256D. Although Unit 1256D-1 is more than twice as thick as Unit 1256C-18, the lava is similar to that in Unit 1256C-18 both in hand specimen and under the microscope, consisting of microcrystalline to fine-grained basalt with similar groundmass textures. A total of 58 samples from Unit 1256C-18 and 18 samples from Unit 1256D-1 were made into thin sections; these were examined to determine the average maximum grain size of the groundmass plagioclase and augite. The measurement procedures followed those of Shipboard Scientific Party (1992) and Umino (1995), except for the measurements on the recrystallized base from Hole 1256C. Twelve circular areas of ~13 mm2 were chosen from each thin section. Selection of measured areas was made so as not to overlap each other and to minimize the area occupied by phenocrysts. The length and width of the largest crystals of plagioclase and augite were measured under the microscope. The longest dimensions of plagioclase and augite are measured as the length, and the widths are measured perpendicular to elongation at the widest portion of the crystals. A plagioclase grain is usually composed of twinned crystals, which are often displaced along the twin planes and terminate at different positions. In such a case, only the longer single crystal was measured. This measurement was repeated in every circle, and the largest and smallest values of the 12 measurements were excluded; the remaining 10 values were then averaged as the maximum grain size. For recrystallized samples from the base of Unit 1256C-18, the grain size was determined as follows: the largest 10 grains of plagioclase, augite, and magnetite were chosen from a ~40-mm2 area under the microscope and their lengths and widths were measured as described above and averaged to give the average maximum grain size. The results of the measurement are given in Table T31.
In Unit 1256C-18, the average maximum plagioclase grain size rapidly increases with depth in the upper 2 m from the surface of the lava pond (Fig. F77A) and then shows stepwise increases up to 2.7 and 0.53 mm in length and width, respectively, at 289 mbsf. From this depth to 293 mbsf, the plagioclase grain size fluctuates widely from 1.4 to 2.6 mm in length (0.7-1.8 mm in diameter of a circle with an equivalent area to the crystal [EQD]). Below this depth range, the plagioclase grain size remains almost constant (length = 1.56 ± 0.21 mm and EQD = 0.83 ± 0.16 mm). The width/length ratio of plagioclase varies systematically with the size variations. It remains low (0.11-0.12) for the first 2 m of the unit but increases quickly up to 0.35, from which it varies between 0.13 and 0.41 down to 293 mbsf. Below this depth the width/length ratio remains almost constant at ~0.23 but shows an overall slight decrease downhole. As the width/length ratio indicates, the zone of coarse-grained plagioclase at 289-293 mbsf has larger, more equant plagioclase than elsewhere in Unit 1256C-18.
The average maximum augite grain size shows a similar pattern to plagioclase, but the peak grain size occurs slightly deeper downhole at 296 mbsf (Fig. F77B). The width/length ratio of augite shows rather different variations from those of plagioclase, especially in the lower half of the unit. Below the zone of coarse-grained augite, the width/length ratio further increases up to 0.64 at 301 mbsf and then remains constant at 0.72 ± 0.11 down to 310 mbsf. The coarse augite zone is a result of the abundant elongate prismatic augite crystals with pigeonite cores that have low width/length ratios. The lower half of Unit 1256C-18 also has some elongate augite prisms; however, they are much less abundant than the stubby to equant augite crystals.
Plagioclase from Unit 1256D-1 shows variations in the maximum grain size similar to those in Unit 1256C-18 (Fig. F77C). Unlike Unit 1256C-18, the largest grain size is present near the top of Unit 1256D-1 (286 mbsf) because the uppermost part of the lava pond was not recovered. However, the width/length ratio of plagioclase has the highest value of 0.30 below the depth of the peak maximum grain size, where the ratio is as low as the majority of the lava pond. The largest plagioclases at 286 mbsf in Hole 1256D form long but thin, platy crystals, suggestive of rapid growth under a large degree of undercooling.
The coarse-grained plagioclase and augite zones in Holes 1256C and 1256D are enriched in incompatible elements (see "Hard Rock Geochemistry") and coincide with intervals of high magnetic susceptibilities (see "Paleomagnetism"), which can be explained by the concentration of differentiated melt in the upper part of the massive lava body. This suggests that the lava body solidified mainly from the bottom and the residual melt became more enriched in incompatible elements and volatiles as it accumulated into the upper part of the body, promoting the growth of large grains of plagioclase and augite. Such a bottom-up solidification of a lava body differs from that of subaerial inflating sheet flows (Kauahikaua et al., 1998) but is well known from observations based on direct drilling into solidifying lava lakes in Hawaii (Helz et al., 1989).
The lowermost 1.6-m-thick core of Unit 1256C-18 is an aphyric cryptocrystalline basalt with an unusual groundmass texture of equigranular clinopyroxene and magnetite with sparse plagioclase laths and that reflects recrystallization at near-magmatic temperatures (Fig. F78). The sample from the deepest portion of Unit 1256C-18 (Sample 206-1256C-11R-7, 130-133 cm) has a variolitic texture, where varioles 0.1-0.3 mm in diameter are composed of slightly elongate, subrounded clinopyroxene crystals 10-15 µm in length aligned in curved lines radiating from the center of varioles and uncommon plagioclase laths 10-40 µm in length with disseminated granular magnetite. Unlike the cryptocrystalline basalt with similar fine variolitic textures and grain sizes, skeletal magnetite is less common than equant crystals. Plagioclase phenocrysts and larger laths have an embayed subhedral outline where original plagioclase is disintegrated into tiny (<10 µm) granular crystals of plagioclase. Recrystallization is more advanced upward, and at 1 m above the base of Unit 1256C-18 (Sample 206-1256C-11R-7, 32-35 cm), clinopyroxene is completely recrystallized and magnetite barely preserves elongate dendritic forms. At 1.2 m above the base (Sample 206-1256C-11R-7, 9-12 cm), clinopyroxene in the original groundmass is completely recrystallized into equant equigranular neoblasts and magnetite scarcely shows the remnant of skeletal crystal forms. Nevertheless, the igneous variolitic texture is still identifiable from the alignment of granular clinopyroxene and elongate blebs of plagioclase (Fig. F78). This is especially true for larger varioles, suggesting that recrystallization is strongly dependent on the original grain size and mineral species. Figure F79 shows the groundmass grain size variations through the recrystallized base of Unit 1256C-18. Both clinopyroxene and magnetite show a steady increase in grain size toward the main body of the lava pond. This can be explained if the driving force of recrystallization was heat supplied from the thick ponded lava above. In contrast, plagioclase does not show any trend of grain size variation with depth, which is consistent with the view that plagioclase retains primary igneous textures and is most resistive to recrystallization.
Coarser-grained late magmatic veins are seen in interval 206-1256C-11R-7, 20-110 cm (see Fig. F80; "Structural Geology" and "Alteration"). These veins are composed of plagioclase, quartz, magnetite, brownish clinopyroxene with pale to dark green rims, and granophyric to vermicular intergrowths of sodic plagioclase and quartz. This mineral assemblage is identical to the mesostasis in the fine-grained basalt of the massive lava pond. A thin section taken from Sample 206-1256C-11R-7, 32-35 cm, shows progressive recrystallization of earlier, more intensely deformed vein minerals that are cut by later, planar veins with chilled margins against the host basalt (Fig. F81). Earlier veins are more progressively recrystallized into equigranular neoblasts and show evidence of subsolidus intracrystalline deformation such as undulose extinction and kink bands. This, together with the undulating margins of the veins, suggests either that the deformation took place under supersolidus conditions or that the rate of replacement of deformed crystals with neoblasts always exceeded the rate of intracrystalline deformation.
We analyzed 16 basaltic samples from Hole 1256C and 55 samples from Hole 1256D for major and trace element analyses on the JY2000 ICP-AES. Effort was taken to collect the freshest material from the cores in order to obtain a downhole record of primary magmatic compositions. Altered rocks were avoided, with the exception of Samples 206-1256D-30R-1, 41-58 cm (hyaloclastite pebble of fresh + altered glass), 51R-2, 107-110 cm (hyaloclastite pebble of fresh + altered glass), and 57R-3, 1-3 cm (highly altered reddish brown basalt). Loss on ignition (LOI) values indicate, however, that alteration exists in several samples we analyzed despite the caution taken to avoid this (Tables T32, T33; Fig. F82). All ICP-AES values reported in this section have been normalized to 100% including LOI (i.e., major elements + trace elements + LOI = 100%). Plots of all elements analyzed vs. MgO are presented in Figure F83.
During the six shipboard ICP-AES runs, within-run reproducibility was assessed based on multiple analyses of BAS-148, a basaltic standard from Leg 148, and BAS-206, an interlab standard created during this leg. Precision on the JY2000 was typically <3% for major elements and <6% for trace elements, with the exception of barium, which was reproducible only to ~45% (Table T34; see also Table T15 in the "Explanatory Notes" chapter). Barium is reported in the data tables but should be used only with caution.
Downhole variations in geochemistry are presented with the igneous units and lithology in Figure F82. There is considerable scatter in the data for any given element downhole, but some general trends can be recognized. For example, Mg#, chromium, nickel, and Ca/Al ratios broadly increase with depth, whereas TiO2, Fe2O3, Zr, Y, Nb, V, and Sr broadly decrease with depth (Fig. F82); other elements do not show systematic behavior downhole. Superimposed on these broad trends are smaller-scale variations, for example, near-constant Mg# in lavas of Units 1256D-2 through 6, which is higher than those in the units immediately above and below. On a Zr-Y-Nb ternary diagram (Fig. F84) (Meschede, 1986), all lavas from Site 1256 plot in the N-MORB field.
TiO2 and Zr behave similarly, as demonstrated by a relatively coherent pattern of increasing TiO2 with increasing Zr (Fig. F85A). Two groups are apparent on this diagram, one more evolved group with high Zr and TiO2, which includes all samples from Hole 1256C, Unit 1256D-1, and a single sample from deeper in Hole 1256D (Sample 206-1256D-52R-1, 5-7 cm). The rest of the deeper samples from Hole 1256D form the low-Ti, low-Zr group. Relatively high Ti and Zr in the shallow samples is consistent with their lower Mg# and higher concentrations of incompatible elements and suggests that they are more evolved than deeper lavas. Two anomalous groups of lavas are notable compared to this general trend of decreasing evolution with depth: one group of samples within Unit 1256C-18 that have very high K2O (Fig. F83) and one group of four lavas with anomalously high Zr for a given TiO2 value (Fig. F85) (Samples 206-1256D-13R-1, 106-110 cm; 14R-2, 130-133 cm; 15R-1, 72-75 cm; and 46R-1, 19-23 cm). These two groups are discussed below.
The massive ponded flow forms the majority of the evolved group distinguishable on the Zr vs. TiO2 diagram (Fig. F85A). Samples of the massive lava have lower Mg# (see Fig. F82) and higher concentrations of incompatible elements than lavas from deeper in Hole 1256D. However, a group of samples from Hole 1256C have exceptionally high K2O: from 294 to 306 mbsf (approximately the middle to lower two-thirds of Unit 1256C-18) an order of magnitude increase in K2O occurs (from 0.08 to 0.74 wt% K2O) and coincides with an increase in NGR measurements (Fig. F86; and "Natural Gamma Radiation" in "Physical Properties"). This anomaly is not apparent in samples from Unit 1256D-1. An increase in K2O would be expected with increasing differentiation; however, the magnitude of the increase in potassium is much greater than what would be consistent with the change in Mg# between the most and least evolved samples in the unit. Furthermore, the high potassium is not likely due to pervasive alteration because the degree of alteration is not noticeably different between these and other samples from the ponded flow (see "Alteration"). Assimilation of K-rich feldspar can be ruled out because there is no corresponding change in Al (Fig. F82). Thus, some other explanation must be invoked, such as an along-rift geochemical zonation in source composition, tapping a small pod of more evolved magma, or local assimilation of an unidentified high-K sediment or altered lava. The anomalously K-rich samples in Hole 1256C but not in Hole 1256D could in these scenarios result either from a geographical restriction on the presence of the K-rich component (for example, if that component erupted from one part of a fissure that fed the lava pond at the location of Hole 1256C but not Hole 1256D) or because the K-rich component was diluted by the larger volume of more typical magma present at the location of Hole 1256D. Distinguishing between these two alternatives is not possible without additional geochemical data.
Directly below Unit 1256D-1 in Samples 206-1256D-13R-1, 106-110 cm; 14R-2, 130-133 cm; and 15R-1, 72-75 cm, there is a sharp increase in Mg# accompanied by a simultaneous increase in some trace elements with widely varying compatibility (e.g., Zr, Sr, Ni, and Cr) (Fig. F82). Despite the relatively high LOI for these three samples (1.8-2.3 wt%), significant alteration may be ruled out as a cause of the "anomalous" enrichment in Mg#, Sr, and Zr because immobile element ratios (e.g., Zr/Ti and Zr/Y) (Figs. F82, F85) are substantially different from the majority of samples from this site. These patterns are also inconsistent with downhole geochemical variations that would result from differentiation for a single parent magma and indicate instead that there is a variation in the primitive magma composition. A similar composition was also measured in Sample 206-1256D-46R-1, 19-23 cm (Unit 1256D-18), and is interpreted to be from the same magmatic type.
Seawater alteration of the upper oceanic crust exerts important controls on the composition of seawater and the crust and influences the physical and magnetic properties of the crust. The seawater chemical and isotopic signatures in altered oceanic crust are recycled with the crust at subduction zones, influencing arc volcanism and mantle heterogeneities. A major goal of Leg 206 was to investigate the type and distribution of alteration effects and processes that occur in a section of upper crust formed at a superfast spreading center and whether these differ from those documented in crust formed at slow and intermediate spreading rates.
Basalts recovered from Hole 1256C are generally dark gray and slightly to moderately altered (1%-20%), with saponite and trace pyrite replacing olivine and interstitial glass and filling vesicles and pore spaces (Fig. F57). In Samples 206-1256C-12R-3, 92-95 cm, and 60-63 cm, however, olivine phenocrysts are locally replaced by talc + magnetite. Igneous sulfide globules and titanomagnetite crystals are mostly unaltered. With the exception of some variations within the massive basalt of Unit 1256C-18 (280.27-312.80 mbsf) (Fig. F56A), there is no consistent variation in alteration type or mineralogy with depth.
Veins compose 0.7% of the recovered basalts at an average frequency of 18 veins/m, decreasing to 2.4 veins/m toward the middle of massive Unit 1256C-18 at 293.9-303.3 mbsf (Fig. F87). Throughout most of Hole 1256C, veins are 0.1-5 mm (average = 0.3 mm) wide and filled mainly by saponite and minor pyrite (and trace marcasite). Locally, these veins are bounded by 0.1- to 2-mm zones where the rock is moderately to highly altered to saponite. Veins of silica (quartz and chalcedony) and calcium carbonate minerals (aragonite and calcite) are less common than saponite veins. Iron oxyhydroxide and celadonite were identified in only four veins in Cores 206-1256C-6R (261.43 mbsf) and 8R (275-285 mbsf) (Figs. F88, F89). Although based on only a few thin sections, the general sequence of secondary mineral formation is (1) celadonite and/or iron oxyhydroxide (where present), (2) saponite ± pyrite, and (3) chalcedony or calcium carbonate ± pyrite.
In the coarser-grained central portion of massive Unit 1256C-18, trace amounts of a blue-green phyllosilicate (tentatively identified as chlorite) are present in interstitial areas along with saponite. Small amounts of biotite (C. Laverne, unpubl. data) (Fig. F90) in interstitial areas associated with magnetite and/or granophyric albite-quartz intergrowths are probably a late magmatic phase. Titanomagnetite in the coarser parts of Unit 1256C-18 contains ilmenite exsolution lamellae, reflecting slow cooling of the lava. Local 0.2- to 5-mm-wide late magmatic veins are also present in this unit (see Figs. F91, F92, F93). These contain granophyric intergrowths of quartz and albite + apatite at their margins. Primary clinopyroxene is partly replaced by green clinopyroxene (possibly aegirine augite or diopside) along these veins. The centers of the veins are filled with euhedral quartz and later saponite and pyrite.
Albite partly replaces plagioclase in 1- to 2-cm-wide light gray alteration halos in the host rock adjacent to some of these veins (Fig. F94). Secondary green clinopyroxene is also present as reaction rims on primary augite within these alteration halos, commonly associated with interstitial late magmatic albite-quartz intergrowths (Fig. F95).
Siliceous interflow sediment at the base of massive Unit 1256C-18 has been recrystallized to quartz, amphibole, hematite, pyrite, chalcopyrite, magnetite, and trace garnet(?) through heating by the overlying massive flow.
Rocks from throughout Hole 1256D exhibit pervasive background alteration. These rocks are dark gray and are slightly to moderately altered (2%-20%) (Fig. F96). Saponite replaces olivine and interstitial glass and fills pore space (vesicles and miarolitic voids), and disseminated secondary pyrite and lesser marcasite are common along with rare chalcopyrite (Fig. F97).
The dark gray background alteration is more or less uniform throughout most of Hole 1256D, but there is one significant variation with depth. Secondary albite and saponite sporadically replace the cores of plagioclase phenocrysts, glomerocrysts, and microphenocrysts below 627 mbsf (Core 206-1256D-57R); this becomes more common below 696 mbsf (Core 63R) (Figs. F98, F99).
The proportion of secondary minerals in a given rock is a function of the rock texture and primary mineralogy, with greater amounts of saponite in rocks containing more abundant olivine, interstitial glass and pores, and vesicles. Veins of saponite ± pyrite are common along with lesser amounts of silica and carbonate minerals. Locally, saponite veins are bounded by 0.1- to 2-mm zones where the rock is moderately to highly altered to saponite (Fig. F97).
Basaltic glass at chilled flow margins is variably altered. The altered glass is dark olive brown to black in hand specimen and yellow brown in thin section (see "Breccia and Interflow Sediment"). This material is probably a mixture of hydrated and altered glass plus phyllosilicates. The innermost material adjacent to fresh glass is very fine grained and perhaps partly amorphous, whereas the outer replacement rims on glass shards are birefringent, suggesting either less amorphous material or greater crystallite size of phyllosilicates. Trace amounts of phillipsite were identified associated with altered glass in one sample (206-1256D-21R-1, 126-127 cm) by XRD.
In the coarser-grained portions of thicker and more massive units, several millimeter- to centimeter-sized dark greenish gray patches are more altered than the surrounding basalt. The total secondary mineral abundance in these dark patches is ~50%. These patches are volumes of rock that contain more abundant primary intercrystalline pore space, which is now filled with secondary saponite and minor pyrite, thus imparting the dark color to the rock (Fig. F100). Clinopyroxene in these patches is also partly replaced by saponite, contributing further to the dark color of the patches. Where pores are sufficiently large (several millimeters), saponite lines the walls of the pores and later silica minerals (opal, chalcedony, and quartz) fill the centers, resulting in irregularly shaped filled pores within the dark patches. The three silica minerals can be present within the same sample; for example, in Sample 206-1256D-37R-2, 12-15 cm, one large pore is filled with opal, whereas another pore <2 mm away is filled with chalcedony and quartz. Secondary pyrite is also common in many of the dark gray patches.
The alteration of the thick ponded lava of Unit 1256D-1 is generally similar to the dark gray background alteration of the underlying thinner flows. The rocks of Unit 1256D-1 are dark gray and mainly slightly to moderately altered, with saponite and trace pyrite replacing olivine and interstitial glass and filling pore spaces and veins (Fig. F101). As in Hole 1256C, however, late magmatic/hydrothermal effects are apparent in the ponded lava of Unit 1256D-1 (Fig. F99). These effects include interstitial granophyric intergrowths of plagioclase and quartz, local secondary green clinopyroxene reaction rims on primary augite (Fig. F102), trace interstitial biotite, and trace amounts of a blue-green phyllosilicate (possibly chlorite) in interstitial areas (Figs. F103, F104). Albite partly replaces plagioclase in alteration halos along veins containing granophyric quartz-albite intergrowths.
The dark gray saponitic background alteration is pervasive in basalts throughout Hole 1256D, but alteration related to veins is also common locally as different-colored alteration halos along veins. Specific vein-related alteration types identified in Hole 1256D include black halos, brown halos, and mixed halos. Twenty percent of the veins logged in Hole 1256D have alteration halos. Black halos are the most abundant, composing 2.2% by volume of recovered cored material. They are present throughout most of the cored section, with higher concentrations at 375-425 mbsf (Fig. F105). Mixed halos are less common and make up 1.2% by volume of recovered cored material, predominantly in the intervals 425-450 and 650-750 mbsf. Brown halos are the least abundant, composing only 0.08% by volume of the recovered core, and their presence is restricted to 350-400 and 625-650 mbsf.
In hand specimen, alteration halos are generally associated with saponite veins, but they are also present along celadonite, silica, iron oxyhydroxide- and pyrite-bearing veins, and a very small percentage of carbonate veins.
What are termed "black" halos range from very dark gray to dark green (e.g., intervals 206-1256D-19R-1, 36-38 cm, and 23R-1, 83-85 cm) to black. Some black halos that are incorporated into mixed halos in Cores 206-1256D-61R through 73R are dark blue green (see "Mixed Halos" below). The black halos range in width from 1 to 30 mm but are most commonly 1 to 10 mm wide (Fig. F106). The percentage of secondary minerals in black halos is similar to that in the adjacent dark gray host rock, but the mineralogy is different (Fig. F107). Black halos are characterized by celadonite replacing olivine and interstitial material and filling pore spaces (Fig. F108). Celadonite is identified by its green color in thin section or its blue-green color and brittle texture in hand specimen. This phase may be celadonite or nontronite or, perhaps more likely, a physical mixture of the two or mixed-layer nontronite-celadonite. Saponite may also replace olivine and interstitial material and fill pore space to a lesser degree than celadonite, although it is generally a later phase than celadonite, as indicated by vein and vesicle filling sequences. Iron oxyhydroxides may also be present in small amounts, intergrown with or staining celadonite, and are generally more abundant close to the vein.
A narrow band (50-300 µm wide) of disseminated secondary pyrite ± marcasite commonly separates the black halo from the adjacent dark gray host rock (Fig. F109). The amount of secondary pyrite in this "pyrite front" varies, but pyrite typically fills pores and replaces igneous minerals.
Brown alteration halos, ranging from 1 to 5 mm in width, are associated with a few veins composed of saponite and/or celadonite (Fig. F110). The orange-brown coloration of these halos results from the staining of primary minerals and filling of vesicles and microfractures with iron oxyhydroxides (Fig. F111), which point counts show amount to 10%-30% of the rock.
The mixed alteration halos result from superposition of brown halos on black halos and commonly exhibit an impressive color zonation. There are typically two well-defined zones, an inner 1- to 10-mm-wide reddish brown or yellowish brown zone adjacent to the vein or fracture, flanked by a grayish black to dark green zone adjacent to the dark gray host rock (Fig. F112).
The dark outer portions of the halos are typically mineralogically indistinct from the normal black halos described above, with celadonite replacing olivine and filling vesicles and interstitial spaces and with lesser saponite in some cases. Generally, the inner brown zone appears similar in thin section to the brown halos described above, but also contains iron oxyhydroxides staining primary minerals and intergrown with the celadonite that fills vesicles and interstitial spaces and replaces olivine. This is consistent with the formation of an early black celadonite-rich alteration halo, which is subsequently partially overprinted by an iron oxyhydroxide-rich brown halo.
There is some variation in the modal compositions of the mixed halos. In a few cases there is very little celadonite in either the dark or the brown zone of the mixed halo. For example, interval 206-1256D-13R-1, 20-22 cm, contains an 0.8-mm-wide iron oxyhydroxide and saponite vein bounded by a 6- to 8-mm-wide mixed halo. In this particular case the distinction between the host rock and the halo is the percentage of secondary minerals. Point counting indicates ~20% saponite in the host rock and ~30% saponite and iron oxyhydroxide in the mixed halo, with a more iron oxyhydroxide-rich inner brown zone nearest the vein (~20% compared with more typical ~5%). In other cases (e.g., in Cores 206-1256D-61R through 73R), the dark portions of black halos are blue green, both in hand specimen and thin section (Fig. F112C), due to a greater proportion of celadonite.
The dark zone of the mixed halo is commonly separated from the neighboring dark gray basalt by a pyrite front, typically 50-300 µm in width. Interval 206-1256D-59R-1, 0-40 cm, contains a vein net composed of saponite and silica with a few percent iron oxyhydroxide, pyrite, and calcium carbonate, which is flanked by a spectacular 25-mm-wide mixed halo with a coarse (up to 1 mm wide) pyrite front (Fig. F112D). Pyrite fronts associated with mixed halos were logged less frequently than those with black halos (12% and 24%, respectively) (see Table T5 in the "Explanatory Notes" chapter). However, this may be an artifact of the vein logging, since many alteration halos have recognizable pyrite fronts in thin section even if one was not visible in hand specimen.
Breccias are divided into two types: hyaloclastite and basalt breccias. The total amount of breccia logged in Hole 1256D makes up 2% by volume of the recovered core, and approximately 90% of this breccia material is hyaloclastite. Secondary minerals in breccia (altered glass + breccia cement) make up 0.5% by volume of the recovered core. The locations of the breccias are shown in Figure F113.
Hyaloclastites were recovered in Cores 206-1256D-20R, 21R, 30R, and 51R (see "Igneous Petrology and Geochemistry"). These hyaloclastites consist of fresh and altered glass clasts up to several centimeters in width in a dark phyllosilicate matrix (Fig. F114). The clasts originally had angular outlines, but the formation of alteration rims on the clasts has led to the presence of rounded kernels of glass in a matrix of altered glass and phyllosilicate cement.
The thin sections of hyaloclastite Samples 206-1256D-20R-1, 27-29 cm, and 30R-1, 41-58 cm, contain 60%-70% fresh glass, 20%-30% altered glass, and 10% matrix. Glass fragments smaller than ~500 µm are totally altered (Fig. F115A). The cores of the clasts are replaced by subisotropic yellowish brown material, which may be hydrated, and altered glass and/or fine-grained and randomly oriented phyllosilicate. The rims of these altered clasts are replaced by a yellowish brown low-birefringence phyllosilicate (saponite?). The larger glass fragments (up to 10 mm) exhibit 50- to 300-µm alteration rims, where glass is replaced by yellow-brown phyllosilicate and altered glass (Fig. F115). Trace amounts of pyrite are also present in the altered glass as small (<1-10 µm) grains. A similar yellowish brown phyllosilicate forms the breccia matrix along with trace amounts of quartz. Sparse plagioclase phenocrysts in the glass are unaltered.
The thin section of hyaloclastite Sample 206-1256D-51R-2, 14-16 cm, contains mainly fresh glass (95%) as a single large (>2 cm) fragment cut by myriad small (2-50 µm) cracks. The glass is altered in 10- to 50-µm-wide zones along the cracks to yellowish brown subisotropic material and phyllosilicate and trace pyrite.
Chemical analysis of a hyaloclastite sample, when normalized to constant Ti, shows gains in LOI, Si, Al, Fe, Mg, K, Cr, and Ni and loss in Ba (Fig. F116). This sample has much higher Cr and Ni contents than the rest of the unit (Unit 1256D-8) (see "Igneous Petrology and Geochemistry"), and it is unlikely that these differences are the result of alteration. Elevated K, Fe, LOI, and Mg contents are likely related to glass alteration and the presence of phyllosilicate breccia cement.
Intervals 206-1256D-57R-1, 0-12 cm, and 62-78 cm, consist of several centimeter-sized angular to rounded basalt fragments cemented by colorless silica (chalcedony), red jasper, quartz, saponite, large (several millimeter) pyrite crystals, and a tabular mineral tentatively identified as anhydrite (Fig. F117).
Elsewhere, local basalt breccias are present over small (1-3 cm) intervals (see "Breccia" in "Structural Geology"). Submillimeter to several-millimeter-sized angular basalt fragments are cemented primarily by saponite, but also locally by celadonite.
Intervals 206-1256D-26R-2, 15-32 cm, and 48-66 cm, consist of centimeter-sized fingers of basalt glass intruding a breccia composed of several centimeter-sized clasts of crystalline basalt, all cemented by mottled dark brown to black + white material (Fig. F118). This cement consists of abundant green and brownish green amphibole, bluish green phyllosilicate (chlorite?), and later quartz (Fig. F119).
Material identified as sediment is present at several depths in Holes 1256C and 1256D (Fig. F113). In some cases sediment is near the base of flows, and elsewhere it is isolated fragments or within fractures in the basalt (Fig. F120). Some of the latter cases are not clearly associated with flow margins and may not be sediment but instead could be of hydrothermal origin.
A 41-cm-long interval of basalt in Core 206-1256D-57R (648.06 mbsf) is highly altered to a blue-green phyllosilicate (tentatively identified as celadonite), a colorless phyllosilicate, and iron oxyhydroxide, imparting blue-green and brick-red colors to the rock (Fig. F121). A sharp contact cuts diagonally across dark gray basalt in interval 206-1256D-57R-2, 116-120 cm, with underlying blue-green basalt grading downward into patchy brick-red and blue-green basalt in interval 57R-2, 123-139 cm. Similar material continues in interval 206-1256D-57R-3, 0-18 cm. The contact with the underlying dark gray basalt was not recovered.
A thin section of Sample 206-1256D-57R-3, 0-1 cm, reveals the rock to be 80%-90% altered. Celadonite ± iron oxyhydroxide replace olivine and fill pore spaces, and clinopyroxene is intensely replaced by a colorless phyllosilicate ± iron oxyhydroxide (Fig. F122). Plagioclase is generally only slightly altered to a colorless phyllosilicate and secondary feldspar but may be locally highly altered to secondary feldspar (Fig. F122). Local 0.5- to 3-mm irregular vugs are present, and the smaller ones are filled with the colorless phyllosilicate or celadonite ± iron oxyhydroxide. Larger vugs are lined with the colorless phyllosilicate ± iron oxyhydroxide and the centers filled with celadonite ± iron oxyhydroxide, or else they are lined with celadonite and the centers filled with quartz or carbonate (Fig. F123).
Chemical analysis of a sample of brick-red altered basalt, normalized to constant Ti, indicates gains in Si, Al, Fe, Mg, K, P, Cr, and Ni and losses in Ca and Mn (Fig. F116). The K and Fe increases result from formation of celadonite and iron oxyhydroxide. The Mg gain most likely relates to the colorless smectite, and quartz contributes to the Si increase. The increases in Cr and Ni are outside analytical error, but it is not clear whether these changes reflect primary chemical variation within the unit or if they really result from hydrothermal alteration.
A total of 5036 veins were identified in core recovered from Hole 1256D, with an average frequency of 23 veins/m of recovered core (Fig. F124). Secondary minerals in veins make up 1% by volume of the recovered core (Fig. F125).
Saponite-bearing veins are the most common (Figs. F96, F101), with 4865 of these veins accounting for 72% of the veins in the core. Saponite is also the most abundant secondary mineral in veins, making up 0.7% of the total core. Saponite veins range from 0.1 to 5 mm in width but average 0.3 mm (e.g., Fig. F97). Pyrite is commonly intergrown with saponite in veins, and saponite typically lines veins filled with later carbonate and silica minerals. XRD of selected saponite veins indicates a trioctahedral smectite structure (Table T35).
Pyrite was identified in 813 veins. Pyrite-bearing veins make up 16% of the total, and pyrite accounts for 0.03% by volume of the recovered core. Vein pyrite is most common within saponite veins (Fig. F126), of which ~5% are pyrite bearing.
Celadonite was identified in 367 veins, or 7% of the total, making up 0.06% by volume of the total core (Figs. F107, F127). Identification of celadonite in veins is based on its green color in thin section or its blue-green color and brittle texture in hand specimen, but this phase may be celadonite or nontronite or, most likely, some combination of the two (mixture or mixed layering). Celadonite is commonly intergrown with iron oxyhydroxides and is typically followed by later saponite filling veins. XRD of selected veins suggests intergrowths of celadonite with saponite (Table T35).
Iron oxyhydroxide was documented in 629 veins, or 12% of the veins in the core, and makes up 0.03% by volume of the recovered core (e.g., Fig. F127). Iron oxyhydroxide is present as a mixture with other minerals, mainly saponite and celadonite (Fig. F107).
Silica minerals (opal, chalcedony, and quartz) were grouped together, and these phases are present in 180 veins, or 3.6% of the total, accounting for 0.14% by volume of the recovered core. Silica minerals typically fill fractures and vugs lined with saponite. Silica-bearing veins range from 0.1 to 15 mm wide (average = 1.6 mm) (Figs. F128, F129).
Carbonate was identified in 90 veins, or 0.02% of the total, making up 0.02% by volume of the recovered core. Calcite and aragonite were tentatively identified based on crystal morphology. Carbonate minerals typically make up a small proportion of veins filled with other minerals (e.g., saponite and silica minerals), but in rare cases carbonate is the dominant component (up to 90%) in 1-mm-wide veins.
Rocks from throughout Holes 1256C and 1256D exhibit a dark gray background alteration, where the rocks are slightly to moderately altered and olivine is replaced by and pore spaces are filled with saponite and minor pyrite. This background alteration is reflected in the distribution of dark gray rocks (Fig. F105) and of pyrite and saponite (e.g., Fig. F99) and is the result of low-temperature seawater interaction at low cumulative seawater/rock ratios.
The local effects of late magmatic/hydrothermal fluids are restricted to within the massive ponded lava near the top of the section (Units 1256C-18 and 1256D-1; i.e., mainly at 276-330 mbsf). These effects include granophyric intergrowths of plagioclase and quartz in veins and interstitial areas, secondary green clinopyroxene reaction rims on primary augite, trace interstitial biotite and blue-green phyllosilicate (chlorite?), and partial replacement of primary calcic plagioclase by albite.
Vein-related alteration is manifest as colored alteration halos along veins. Black halos contain celadonite and have been interpreted to result from the upwelling distal low-temperature hydrothermal fluids enriched in iron, silica, and alkalis (Edmond et al., 1979; see summary in Alt, in press). Brown halos are later features that formed by circulation of oxidizing seawater along fractures that were not bordered by previously formed black halos. Mixed halos result from the superposition of brown halos on previously formed black halos.
This vein-related alteration occurs irregularly throughout Hole 1256D below the massive Unit 1256D-1 but is concentrated in two zones: 350-450 and 635-750 mbsf (Fig. F105). These zones correspond to peaks in frequency and proportion of celadonite and iron oxyhydroxide veins and minima in the abundance of pyrite veins (Figs. F124, F125). These were likely zones of greater permeability and, consequently, increased fluid flow.
Vein carbonate is most common above ~530 mbsf, but despite lower frequency, higher overall abundance of carbonate occurs at greater depths (Figs. F125, F130).
There are three peaks in glass abundance at 400, 460, and 600 mbsf (Fig. F105), corresponding to hyaloclastites in the core. These are important because of the substantial degree of glass alteration and because saponite cementing the breccia results in corresponding peaks in the abundance of secondary minerals (saponite) at these depths (Fig. F130).
The appearance of albite and saponite partly replacing plagioclase below 625 mbsf indicates a change in alteration conditions (Fig. F99). This change may result in part because of slightly higher temperatures at depth or more evolved fluid compositions (e.g., decreased K/Na and elevated silica).
Overall, the basalts recovered from Site 1256 do not exhibit a general decrease in seawater interaction with depth and there is no simple decrease in the amount of alteration halos or iron oxyhydroxide with depth. In contrast, alteration appears to have been concentrated into different zones that may be related to the architecture of the basement, such as lava morphology, distribution of breccias and fracturing, basement topography, and the influence of these on porosity, permeability, and fluid circulation.
Alteration of the basement section of Hole 1256D is compared to other sites in Figure F131. Compared to most of these sites, Hole 1256D contains much fewer brown, mixed, and black alteration halos. Site 1256 is, however, quite similar to another section of crust generated at a fast spreading ridge, Site 801. The latter site contains two hydrothermal deposits and associated intense hydrothermal alteration, however. One important feature is the lack of any oxidation gradient with depth in Hole 1256D, in contrast to the stepwise disappearance of iron oxyhydroxide and celadonite in Hole 504B and the general downward decrease in seawater effects at Sites 417 and 418.
Primary igneous and postmagmatic structures were described in Holes 1256C and 1256D. The former include syn- to late-magmatic structures, partially linked to the flow of lava. Postmagmatic structures include veins, microfaults, joints, and breccias. The distribution of these features in Holes 1256C and 1256D are summarized in Figures F132 and F133, respectively. Our structural observations are logged on the structural description forms and entered in the structural logs (see Table T7 in the "Explanatory Notes" chapter). The Structural Log includes the record of morphology, width, mineralogy, and orientation of veins, joints, and faults. We measured more than 2000 veins and microfaults in Holes 1256C and 1256D. In addition, we compiled a record of the textural features and composition of breccia in the Breccia Log (see Table T8 in the "Explanatory Notes" chapter). Techniques and methods utilized are discussed in "Structural Geology" in the "Explanatory Notes" chapter.
Primary igneous features recognized in Holes 1256C and 1256D include the magmatic fabric, laminations, flattened vesicles, folds, shear-related structures, and late magmatic veining and fracturing.
Magmatic fabric in the recovered basalts is mainly defined by the preferred orientation of mineral shapes that do not show dynamic recrystallization, by minor distortion of crystal lattices (interval 206-1256C-8R-4 [Piece 5, 114-117 cm]; thin section 25), and by brittle and brittle-ductile intracrystalline microstructures. The first two features are related to magma flow; the last one occurs during the solidification and subsolidus cooling of lava.
Weak to pronounced shape-preferred orientation of plagioclase was observed near chilled margins of the lava sheet flows as well as at the top and at the bottom of the massive ponded lava (Sections 206-1256C-8R-4 and 11R-7). Oriented plagioclase laths (<0.1 mm) are either scattered within a glassy or cryptocrystalline groundmass or are clustered in thin bands (Fig. F134). In some cases, lattice-preferred orientation is also present (interval 206-1256C-11R-7 [Piece 1A, 9-12 cm]).
In fine-grained to microcrystalline basalts, pyroxene and plagioclase crystals underwent intracrystalline deformation at different times. In plagioclase, microcracks are planar and regularly distributed and crosscutting relationships can be easily recognized. Healed arrays of fluid and/or solid inclusions, interpreted as syn-crystallization features, are reactivated or cut by microcracks, mainly controlled by crystallographic cleavages and interpreted as cooling and shrinkage features. Healed intragranular microcracks coated by alteration phases (e.g., saponite replacing glass) are interpreted to be late to postmagmatic, as they overprint or reactivate syn-crystallization inclusions and cooling-related microcracks (e.g., intervals 206-1256C-11R-7 [Piece 1B, 32-35 cm] and 206-1256D-2R-1 [Piece 1C, 53-56 cm] and 22R-2 [Piece 5, 111-113 cm]) (Fig. F135).
Primary igneous layering or flow banding in the recovered basalts is mainly manifested by trails of elongate flattened and stretched vesicles and amygdules in the massive lavas and by millimeter-scale laminations (between 0.5 and 2 mm thick) (intervals 206-1256C-12R-1 [Piece 4, 54-67 cm] and 206-1256D-52R-1 [Piece 5, 112-123 cm]), as well as by changes in basalt grain size (see "Igneous Petrology and Geochemistry") in sheet flows. Flow banding in glass near chilled margins is commonly defined by aligned coalesced vesicles forming "necklacelike" domains. At the base of the lava pond we observed that the preferred orientation of plagioclase laths is consistent with the flattening of vesicles (e.g., Section 206-1256C-11R-7).
The dip of primary layering varies from horizontal to nearly vertical (Fig. F136), and there is no apparent relationship between dip variation and depth in the hole. Trails of stretched vesicles are visible only in some sections (206-1256C-11R-7 and 12R-1 through 12R-4 and 206-1256D-22R-1 through 22R-4, 36R-1, 37R-1 through 37R-2, 40R-1, and 41R-2), so measurements of vesicle trails may not be representative of the upper crust at Site 1256 but can be related to the intersection of the hole with different parts of lobate lava flows with differing primary dips.
The only exception is the lower part of the massive ponded lava flow, Subunit 1256C-18i, where measurements and observations on the orientation of flow banding are available throughout 130 cm of core (Section 206-1256C-11R-7).
Some core pieces contain evidence for folding and multiple refolding, which affect millimeter-scale flow banding, vesicle alignment, "necklacelike" vesicle domains, or late magmatic veinlets.
At the top of lava flows and in glassy breccia clasts (Sections 206-1256C-8R-4 through 8R-5 and 206-1256D-51R-2) near chilled glassy margins, millimeter-scale layers of flattened and coalesced varioles are folded within a glassy matrix (Fig. F137). Folds mainly have subhorizontal or gently dipping axial planes parallel to the chilled contact and show similar or isoclinal geometries. Quarter folds around clasts of altered glass are also present (Section 206-1256D-51R-2).
In the top of the massive lava flow of Unit 1256C-18 (intervals 206-1256C-8R-4 [Piece 5, 114-126 cm] and 8R-5 [Piece 3A, 18-21 cm]) (Fig. F138), multiple refolding affects veinlets filled with granoblastic fine-grained pyroxene + plagioclase + magnetite in a cryptocrystalline sparsely phyric groundmass (Fig. F139). At least two steps of progressive shear folding superposed by large-wavelength folds are observed. The first two shear fold generations mainly have subhorizontal or gently dipping axial planes, similar or isoclinal geometries, and high amplitude/wavelength ratios. The last fold generation is characterized by gentle folds with a subvertical axial plane and very low amplitude/wavelength ratios.
In Holes 1256C and 1256D (e.g., Section 206-1256C-11R-7 and interval 206-1256D-37R-2, 12-15 cm) (Fig. F140) at the bottom of ponded lava or at the contact between two sheet lobes, layers of recrystallized fine-grained basalts contain folds lined by late magmatic veinlets.
At the bottom of the lava pond in Hole 1256C (Unit 1256C-18), folds affect both fine-grained and coarse-grained veins. Folds of fine-grained veins are mainly sheath fold generations with subisoclinal profiles and with subvertical axial planes. Folds in coarse-grained veins are open folds, and crosscutting relationships show different generations of late magmatic vein infilling.
Boudinlike disruption of the veinlets is present in interval 206-1256C-11R-7 [Piece 1, 10-110 cm] (Fig. F141). The folding and disruption of layering probably occurred during the flow of the magma while it was relatively viscous.
Microstructures caused by ductile simple shear, such as rotated filled vesicles and glomerocrysts resembling mantled porphyroclasts with sigma- and delta-type shape (interval 206-1256C-7R-1, 62-66 cm) and partially rotated tear drops of mingling lava (interval 206-1256C-8R-5, 18-21 cm), were observed in aphyric basalts.
Late magmatic veins are present in both Holes 1256C and 1256D mainly in the interval between 275 and 315 mbsf (see Table T7 in the "Explanatory Notes" chapter and occurrence of measured structures in Figs. F132 and F133; Sections 206-1256C-8R-4, 9R-3 through 9R-7, 10R-1, 11R-6, and 11R-7 and 206-1256D-2R-1, 3R-3, 4R-1, 4R-2, and 9R-2). They are mostly intragranular nonfibrous veins or intragranular face-controlled veins; arrays of tension gashes and sigmoidal pull-aparts are present as well (intervals 206-1256C-11R-7, 64-68 cm, and 11R-7, 121-124 cm, respectively) (Figs. F142, F143; see also Fig. F134). In coarser-grained basalts, late magmatic veins develop along trails of aligned vesicles or along the intragranular microcracks that affect both plagioclase and pyroxene. Their morphology ranges between irregular, riedel, or stepped types (see Fig. F10 in the "Explanatory Notes" chapter).
Late magmatic veins can be divided according to their infilling and width, with felsic veins usually >0.5 mm and glassy veins usually <0.5 mm wide. Felsic veins are composed of quartz + alkali feldspar symplectites showing a granophyric texture ± clinopyroxene ± magnetite ± plagioclase ± apatite (see "Igneous Petrology and Geochemistry" and "Alteration"). Glassy veins are nearly completely replaced by alteration minerals, principally saponite, that grow with face-controlled geometries (e.g., Sample 206-1256D-11R-3 [Piece 1A, 39-42 cm]). These veins are commonly linked to vesicles and amygdules and appear to be the feeders for late magmatic fluids (e.g., Sample 206-1256D-37R-1 [Piece 8, 106-113 cm]) (Fig. F144).
In interval 206-1256C-11R-7, 9-12 cm, a 15-mm-wide vein is present in a recrystallized microcrystalline basalt. The increase in grain size from the margin to the middle of the vein is striking (Fig. F145). Near the vein wall plagioclase laths grow perpendicular to the wall, but in the middle clinopyroxene and plagioclase have a preferred orientation parallel to the vein wall. In the wallrock, plagioclase laths of the host basalt show both shape- and lattice-preferred orientations parallel to the vein border. An array of en echelon tension gashes are also present near the vein. These features, together with extension fissures parallel to the fold arc (interval 206-1256C-8R-5, 18-21 cm) (Fig. F146) or associated with wrinklelike structures (e.g., interval 206-1256D-46R-1, 55-73 cm) (Fig. F147) and filled stepped structures (intervals 206-1256C-11R-7, 9-12 cm, and 76-79 cm), show that localized shear deformation and cracking occurred in partially crystallized magma. Interstitial melt, either a residual segregation or an immiscible liquid fraction with different viscosity, has been drawn into the fractures during progressive deformation as the cooling flow passed through the ductile-brittle transition.
The progressive transition from dominant ductile to brittle-ductile deformation is also displayed by shear bands and gashes filled mainly by plagioclase + clinopyroxene ± magnetite ± quartz that cut magmatic-related shear folds with the same sense of shear (Sections 206-1256C-8R-4 and 11R-7) (Figs. F139, F140).
Veins are the most prominent structural features observed in rocks recovered from Holes 1256C and 1256D. Veins are extensional open fractures filled by a variety of minerals. We also identified shear veins, which are veins with minor shear displacement, coated by slickenfibers or overlapping fibers (see "Structural Geology" in the "Explanatory Notes" chapter). Approximately 600 veins were logged in Hole 1256C and 1700 veins in Hole 1256D during core description. These veins represent roughly one-third of the total veins in the cores of the two holes (see "Alteration") because we logged and measured only structures present in oriented pieces. In addition, most veins and joints showing typical morphologies of rock fracturing during cooling were not measured. Among these are veins with Y-shaped intersections and sinuous, mostly steeply dipping veins intersected by radiating veins (e.g., interval 206-1256D-22R-1, 85-123 cm) (Fig. F148). These veins are more common in Hole 1256D (Core 206-1256D-13R through Section 24R-3) than in Hole 1256C. The geometry of these structures is reminiscent of thermal contraction cooling and thus cannot provide reliable data for establishing the regional stress field. Therefore, there is a pronounced data gap in Hole 1256D from Cores 206-1256D-13R through 26R (Units 1256D-2 through 6).
Veins mostly have a planar or slightly curved morphology. Many veins have irregular margins but have a general planar trend. Individual veins commonly branch into a number of diverging "splays" (see "Structural Geology" in the "Explanatory Notes" chapter) at their ends. Multiple veins are common in anastomosing geometries and, where veining is pervasive, develop into vein networks (e.g., interval 206-1256D-8R-2, 76-82 cm) (Fig. F149) or the initial stages of brecciation (e.g., Section 206-1256D-59R-2) (see "Breccia" below). In many cases veins are oriented in en echelon Riedel shear arrays. Stepped veins are common in both basement holes (Cores 206-1256C-9R and 11R and throughout Hole 1526D) and commonly are characterized by millimeter-scale pull-aparts filled with secondary minerals (Fig. F150).
Some veins (mostly planar) form conjugate sets with a conjugate angle of ~60° or more (e.g., Section 206-1256D-39R-1) (Fig. F151). In Hole 1256C they are present in Cores 206-1256C-5R (massive Unit 1256C-1) and 13R. In Hole 1256D, conjugate sets of veins are more common and are mostly concentrated in the coarser-grained massive flows (e.g., Cores 206-1256D-6R and 11R [massive Unit 1256D-1], 13R-1 [Unit 1256D-2], 26R and 27R [massive Unit 1256D-8], 39R [Unit 1256D-13], and 65R [Unit 1256D-26]).
Shear veins are mostly planar, 1-6 mm wide, and are commonly filled with overlapping fibers of dark green saponite (e.g., interval 206-1256D-8R-7, 38-55 cm) (Fig. F152). In Hole 1256C, they are most abundant in the massive flow (Unit 1256C-18). In Hole 1256D, shear veins are concentrated from Cores 206-1256D-3R through 12R (massive Unit 1256D-1) and in sheet flows from Cores 27R through 43R, where they mainly show, based on the geometry of the infilling fibers, a reversed sense of shear. From Core 206-1256D-57R to the bottom of the hole (Core 73R), shear veins have strike-slip or normal sense of shear.
Veins range in thickness from ~0.1 to ~15 mm (see also "Veins" in "Alteration"). The vein infilling minerals are mostly saponite, celadonite, iron oxides, or pyrite, silica (quartz and chalcedony), and carbonate. Veins are commonly composed of a combination of these minerals, but several end-member types were defined based on the most abundant mineralogy (see "Alteration"). The crystal morphology of the infilling minerals may be fibrous or nonfibrous, blocky, and composite. Fibrous veins are filled with clay minerals (mostly saponite). Stretched carbonate fibers are rare (interval 206-1256C-8R-3 [Piece 13, 136-140 cm]) (Fig. F153), but we observed radiating aragonite fibers in composite veins. In most fibrous clay veins, the fibers extending from each wall meet along an irregular suture line near the center of the vein (syntaxial vein). Clay fibers are commonly straight and their elongated direction generally orthogonal to the vein wall, but in curved or irregular veins they can be oblique or radiating. Nonfibrous veins consist of iron oxyhydroxide, celadonite, and chalcedony. Composite veins are characterized by the coexistence of both nonfibrous and fibrous minerals. These veins appear to have formed by reopening of veins (mostly ataxial veins; e.g., interval 206-1256C-8R-1 [Piece 5A, 42-47 cm]) (Fig. F154). In some samples, composite veins show evidence of shear deformation, as indicated by tension gashes with iron oxyhydroxide-stained fillings (interval 206-1256C-8R-3 [Piece 13, 136-140 cm]) and sheared saponite fibers (interval 206-1256C-8R-1 [Piece 5A, 42-47 cm]) (Fig. F155).
Joints are open fractures with no differential displacement. We measured very few joints in oriented cores. Those measured commonly show a planar morphology. Drilling-induced fractures (mostly saddle shaped and disc fractures) were recognized but not included in the structural logs.
Microfaults are fractures with kinematic evidence for shear displacement across them in which the scale of offset is millimeters to centimeters. Microfaults detected in Holes 1256C and 1256D are mostly associated with cataclastic infilling (microcataclasite) consisting of millimeter-scale clasts of the host basaltic rock embedded in saponite matrix. Microfaults occur only in the massive units of the two holes (Unit 1256C-18: Cores 206-1256C-9R and 11R, and Unit 1256D-1: Cores 206-1256D-3R, 5R, 6R, 7R, and 11R). In thin section, the cataclastic infilling of the faults are composed of fibers and microbreccia (interval 206-1256C-11R-3 [Piece 1D, 67-71 cm]) (Fig. F156).
Five main types of breccia were intersected in Holes 1256C and 1256D (see Table T8 in the "Explanatory Notes" chapter): hyaloclastite, talus breccia, breccia with interflow sediment, incipient jigsaw puzzle, and hydrothermal. Hyaloclastite breccia forms on the seafloor by the fragmentation of glassy pillow rims (see also "Alteration"). In Hole 1256C, hyaloclastite breccia is observed at the top of Unit 1256C-18 (30 cm long, at the bottom of Section 206-1256C-8R-4 and top of Section 8R-5). The clasts consist of fresh glass and aphyric basalt with subangular to subrounded shapes and range in size from 2 mm to 4 cm. Clasts are embedded in a fine-grained clay matrix. In Hole 1256D hyaloclastite breccia forms a 115-cm-long interval in Unit 4 (Subunit 4c; Sections 206-1256D-21R-1 and 21R-2). It consists of clasts of volcanic glass, altered glassy shards, and lithic basalt, surrounded by a matrix composed of fine-grained clay and a white mineral (e.g., interval 206-1256D-21R-1 [Piece 20, 15-24 cm]) (Fig. F157). Clasts have angular to subangular and subrounded shape and range in size between 1 and 20 mm. They are fractured and cut by saponite veinlets. Glassy shards are millimeter sized and are replaced by colliformic light brown and whitish minerals that grew from the rim of the shard into open space (Fig. F157). In interval 206-1256D-21R-1, 100-145 cm, the contact between glassy clasts and intact basalt wallrock is mostly steeply dipping (Fig. F157). Hyaloclastite breccia is also present in Unit 1256D-8 (Subunit 1256D-8c; Sections 206-1256D-29R-2 and 30R-1) and Unit 1256D-21 (Section 206-1256D-51R-1 and Core 51R), where it covers an interval of 110 cm.
A second type of breccia is present in Unit 1256D-4 (interval 206-1256D-20R-1, 32-50 cm). It is a polymictic, matrix-supported breccia consisting of clasts of both aphyric basalt and glass, ranging in size from 2 mm to 3 cm, embedded in a fine-grained matrix made up of glass and clay (saponite?). Lithic clasts have angular to subangular shapes and straight margins, generally outlined by a millimeter-sized dark alteration halo. Clasts may include vesicles and are often cut by veins. Smaller lithic clasts have angular shapes and fill the space between larger clasts (interval 206-1256D-20R-1 [Piece 7, 43-50 cm]) (Fig. F158). Glassy clasts are subrounded. The matrix is locally cut by fractures and brittle shear zones, which also run along the clast margins. This type of breccia may derive from talus material.
In interval 206-1256D-26R-2 (Piece 3, 15-20 cm), we observed a third type of breccia characterized by sediment infilling. Piece 3 preserves the contact between cryptocrystalline intact basalt and breccia. The breccia consists of lithic clasts with elongated angular shape, ranging in size between 1 mm and 3 cm. From their shapes, these clasts can be fitted back together, similar to jigsaw-puzzle breccia. Other clasts consist of glass and chilled basalt with subangular shape. Clasts are embedded in a clay matrix (5 vol%). The shape of clasts and the low percentage of matrix suggest that brecciation was probably in situ and ranges from incipient to advanced. In interval 206-1256D-26R-2 (Piece 7, 48-88 cm) aphyric basalt is more completely brecciated than in interval 20R-1 (Piece 3). Basaltic clasts are angular and range in size from 1 mm to 4 cm and are bordered by a thin dark green alteration rim. These clasts include or are in contact with subangular or subrounded fragments of aphanitic to altered glass that are apparently chilled against the aphanitic basalt (interval 206-1256D-26R-2 [Piece 4, 20-32 cm]) (Fig. F159). The matrix (<10%) consists of siliceous sediment and clay (saponite), plus green and brownish green amphibole, bluish green phyllosilicate (chlorite?), and later quartz and/or chalcedony.
The fourth type of breccia is incipient jigsaw puzzle and is mainly present in Unit 24 (Section 260-1256D-59R-1), but several similar brecciated pieces are scattered in Sections 206-1256D-29R-1, 43R-1, 45R-1, 54R-3, 55R-1, 56R-1, and 58R-1. This breccia is characterized by jigsaw-puzzle fabric in which the various clasts have angular shape, straight boundaries, and can be fitted back together (e.g., interval 206-1256D-56R-1 [Piece 2, 9-18 cm]) (Fig. F160). The matrix in this type of clast-supported breccia mainly consists of clay (saponite) ± white minerals (carbonate or anhydrite?) and chalcedony. Where brecciation is incipient, the contact with the basalt host rock is commonly preserved and breccia passes into a complex vein network (interval 206-1256D-45R-1 [Piece 2, 4-14 cm]) (Fig. F161). Incipient brecciation is evident where a few small basalt clasts "float" in the vein and are completely supported by matrix. The characteristics of this type of breccia suggest in situ formation, with fragmentation occurring along a vein (clast supported) within a vein network cutting intact basalt (tectonic breccia?). Disaggregation of the host basalt would be aided by contractional cooling joints. Even though the displacement of clasts away from the intact rock is small (millimeter scale), the nature of the matrix and the shape of clasts suggest that brecciation could have been assisted by relatively high pore fluid pressure in an extensional regime.
The fifth type of breccia is present in intervals 206-1256D-57R-1, 0-12 cm, and 57R-2, 62-78 cm. It consists of several centimeter-sized angular to rounded basalt fragments cemented by colorless silica (chalcedony), red jasper, quartz, saponite, large (several millimeters) pyrite crystals, and a tabular mineral tentatively identified as anhydrite (see "Alteration"; Fig. F162). The small interval of breccia sampled and the relatively large size of the clasts precludes any clear-cut interpretation of the origin of this breccia. We tentatively interpret it as being of hydrothermal origin, but a tectonic nature cannot be ruled out.
Only true dip data obtained by the measurement of structure orientations are considered because the cores have yet to be reoriented with respect to geographic north. The absence of structural marker planes (e.g., sedimentary bedding and basaltic layering) precludes reorientation of the measured veins and the calculation of dip azimuths. Paleomagnetic and logging results may enable the reorientation of some of the veins on the larger pieces during shore-based analysis of the data (see "Paleomagnetism" below), which would allow calculation of true dip azimuths.
Rose diagrams of true dip values of all oriented structures for Holes 1256C and 1256D (Fig. F163) show that, in general, structures of Hole 1256D are gently dipping, having most common dip angles of ~15° ± 5°; dip values of ~70° are well represented mainly in the lowermost 100 m of cores. Other dip angles are equally distributed throughout the hole. In Hole 1256C, true dip angles show a maximum in frequency between 10° and 20°; however, dip values of ~50°-55° and 90° are common as well.
In Hole 1256C, the distribution of true dips in each lithologic unit (Fig. F162) shows that in the upper units between 258 and 271.82 mbsf (Units 1256C-5 through 15) the dip values are bimodally distributed in sets at an angle of 50°-60°. This suggests conjugate systems in the upper part of the hole, whereas in the middle and lower parts of the hole, true dips are clustered around one orientation. In the lower three units (Units 1256C-18 through 22), structures mainly show gentle dips.
In Hole 1256D, the distribution of true dip angles with depth (Fig. F164) does not show any systematic variation. In the massive flow Unit 1256D-1 gently dipping veins predominate, especially in the coarser-grained sections of Subunit 1256D-1c.
These observations suggest that the geometry and the intensity of rock fracturing may be controlled also by the physical properties and morphology of the lithologic units, such as the grain size and the rock type.
True dip values for late magmatic veins and shear veins and microfaults from Holes 1256C and 1256D are plotted in the rose diagrams in Figure F165. Late magmatic veins are mostly gently dipping in Holes 1256C and 1256D, showing the highest frequency at 15° and 5°, respectively. In contrast, shear veins are moderately to steeply dipping in the two holes (maximum frequency ranges from 45° to 75°).
To characterize the paleomagnetic signal and resolve the magnetization components recorded in the igneous rocks of Holes 1256C and 1256D, we measured and analyzed the magnetic remanence of selected archive-half sections, discrete rock pieces (<15 cm long) from the archive half, and discrete cubes (~8 cm3) from the working half. The remanence data were collected before and after progressive AF or thermal demagnetization. A primary goal of our study was to assess the role that the extrusive rocks of the upper crust play in generating marine magnetic anomalies. To properly do this, a variety of rock magnetic data will be required along with more remanence data that build upon our initial remanence observations. Even so, the data presented here provide some intriguing information about the stability of the magnetization of the extrusive upper crust.
All split cores, core pieces, and discrete samples have a sizable drilling overprint, which is characterized by a steep downward direction and a radial-horizontal component that points toward the center of the core. In the ODP core orientation system, the latter results in a strong bias in the declinations toward 0° for archive-half samples and 180° for working-half samples (given the orientation we used during Leg 206; see "Core Orientation" in "Paleomagnetism" in the "Explanatory Notes" chapter). Initial NRM measurements are thus characterized by inclinations of ~60° or more and declinations of ~0° or 180°. For many intervals, we were unable to remove this overprint with AF or thermal demagnetization without the magnetization direction becoming erratic.
As discussed below, the drilling overprint appears to swamp any pre-drilling remanence in many of the units, more so than has been observed for other cruises that have penetrated into basalts. In general, 25-mT AF demagnetization removes much of the drilling overprint but also reduces the magnetization by more than 95%, often leaving no or only a poorly resolved hint of the ChRM direction. In some units where we may have removed the drilling overprint, the ChRM appears to be steep, which is unexpected for a low-paleolatitude site unless a geomagnetic transitional or excursional field was recorded.
Our primary data set was based on cutting 2-cm x 2-cm x 2-cm cubes from the working half of the core and measuring their magnetization during progressive AF demagnetization. The demagnetization increment was a maximum of 5 mT, sometimes 2-4 mT below 20 mT, and all demagnetizations were continued to the maximum capability of the demagnetizer at 80 mT. In general, little information about ChRM was added after 50 mT, with direction and intensity either roughly stable or with direction showing incoherent variation at very low intensities (Table T36).
For Hole 1256C, all samples had steep NRM inclinations (>58°) and little tendency for inclinations to change during demagnetization. Most samples have low coercivities, with generally 3%-20% of NRM intensity remaining after only 10-mT demagnetization. The range of behavior on demagnetization is shown in Figure F166. Directional stability at higher demagnetization levels clearly correlates with slower loss of intensity at low demagnetization levels. Most intensities for NRM are high for Miocene MORB at 10-30 A/m, but not after moderate demagnetization. Uniquely stable Sample 206-1256C-11R-3, 62-64 cm, shows a clear overprint with steep inclination and southerly declination being removed from a ChRM with a northwesterly declination (Fig. F166D). For all other samples from Hole 1256C, any magnetization predating a steep overprint can only be recognized after careful consideration, if at all.
For the deeper Hole 1256D, we observed more variety of behavior on demagnetization. For all of the very thick igneous Subunits 1256D-1b and 1c, samples resembled the lowest-coercivity samples from Hole 1256C, with near-vertical NRM giving way to scattered directions and very weak intensities by 30- to 50-mT demagnetization (Fig. F167A). Below Unit 1256D-1, it becomes common for inclinations to trend toward nearly horizontal with demagnetization. Some samples are readily interpreted as a nearly horizontal ChRM with a partial drilling overprint (Figs. F167B), but others show more complex behavior. A moderate number of the samples show trends away from the expected drilling overprint up to 30-50 mT but follow a different trend during higher-field demagnetization (Fig. F167C, F167D, F167E). Because the higher-field trend is often similar for different samples within a single demagnetization run but often varies between runs, we interpret them as artifacts of failure to null the magnetic field in the demagnetizer, with variations between runs resulting from changing orientation of the ship or electric currents.
Progressive thermal demagnetization experiments most commonly indicate that >80% of the magnetization is removed by ~300°C, with a steep gradient in intensity loss between ~240° and 300°C (Fig. F168A). Above 500°C, the directions and intensities sometimes become erratic, indicating conversion of the magnetic minerals. Both observations are consistent with the presence of titanomaghemite. In most samples some magnetization remains beyond 300°C. This remaining magnetization, although <20% of the NRM in most samples, unblocks mainly above 500°C, as might be expected for low-Ti titanomagnetite. In a few samples, there is no steep gradient in intensity loss at ~300°C. Instead, the magnetization mostly unblocks within a relatively narrow temperature range (<50°C) above 400°C (Fig. F168B), which again is consistent with the presence of titanomagnetite. As with AF demagnetization, thermal demagnetization indicates the occurrence of both steep and shallow directions.
Because it can be difficult to determine whether the steeper inclinations reflect a faithful record of an unusual magnetic field or simply a nearly complete overprint during drilling, we compared results from the working-half minicubes with measurements of the archive-half pieces at corresponding depths. Because the drilling overprint has a strong radial component, there is potential to distinguish original field from overprint using the extent of agreement of directions between the halves of the core.
We measured entire pieces of the archive half of the core using progressive AF demagnetization but stopping at 25-40 mT to preserve some magnetization for potential future studies. In the massive basalt unit near the top of basement, igneous Units 1256C-18 and 1256D-1 (see "Massive Ponded Flow" in "Macroscopic Description" in "Igneous Petrology and Geochemistry"), we measured large split-core sections in pass-through mode at 1- to 2.5-cm intervals. These sections were progressively demagnetized in 2- to 4-mT steps up to 12 mT, then 2-mT steps up to 24 mT, followed by a final demagnetization step at 25 mT. All sections measured were characterized as being nearly complete unfractured pieces, which allows for high-quality continuous measurements along the section. Any data collected near the few fractured zones or across the few gaps that were present in the selected sections were removed to avoid interpreting erroneous data. This clean data set is given in Tables T37 and T38. Deeper in Hole 1256D, where pieces are more irregular and usually smaller, we selected pieces or subpieces of length ~6 to 15 cm to measure in discrete sample mode (Table T39). Comparisons between pieces and discrete samples are discussed below (see "Discussion and Conclusions").
The clean pass-through split-core data set is characterized by directions that have declinations near 0° and an average inclinations >50° before and after AF demagnetization up to 25 mT (Figs. F169, F170). Interestingly, the mean inclination before demagnetization is 58° and actually gets steeper, averaging 66° after 25-mT demagnetization. It could be that this steep inclination after demagnetization at 25 mT is still the drilling overprint or, alternatively, that the ChRM really is steep. The former interpretation is supported by the unremoved radial overprint, which is why the declinations never deviate far from 0°. We, however, favor the latter interpretation for several reasons. First, the radial overprint is often more persistent than the steep drilling overprint in cores with steep ChRM directions because the relative contribution of the horizontal component of the ChRM is small. Second, even if the radial overprint persists in the split-core sections, it appears to be removed in discrete samples, as illustrated by Sample 206-1256C-11R-3, 62-64 cm (Fig. F166D). In those samples, the steep direction remains. Likewise, we think most of the steep drilling overprint has been removed from the split-core sections by 25-mT demagnetization and that a steep ChRM direction remains.
In addition, demagnetization at 25 mT results in removal of more than 95% of the initial NRM. This could indicate that the massive unit has a magnetic mineralogy that has only relatively soft coercivity. Alternatively, the unit may have higher-coercivity minerals that are not contributing to the remanent signal, which could happen if the unit cooled when the geomagnetic field was weak, like that during geomagnetic transitions and, hence, the unit was never strongly magnetized. Such a phenomenon would also explain why the ChRM direction is steep rather than shallow, as would be expected for the stable geomagnetic field at low paleolatitudes.
To better assess the role of the drilling overprint, a 1-cm-thick whole-round sample was taken from massive basalt at interval 206-1256D-26R-4, 74-75 cm, in igneous Subunit 1256D-8a. The sample was then labeled and cut into 25 small cubes, each of volume <1.50 cm3, and an interval was assigned to allow data to be uploaded into the database. Each sample was weighed, and five samples had a volume analysis conducted in the pychnometer to determine the average density. This allowed the volume of all pieces to be calculated (Table T40). The susceptibility of each piece was then measured before AF demagnetization was conducted. All pieces were progressively demagnetized in steps of 2-5 mT up to a peak field of 80 mT.
Prior to demagnetization, the inclination of the samples has a tendency to point steeply downward; inclinations are positive and typically >60° across the cross section, being steepest in the central samples (Fig. F171). The strong radial symmetry indicates that the NRM is dominated by drilling overprint, but the slight breakdown in the symmetry in the center suggests that the more central samples are less affected by the overprint and thus are a combination of the overprint and the ChRM. The median intensity is 35.3 A/m prior to demagnetization but is reduced to 0.96 A/m after 20-mT demagnetization. Thus >95% of the magnetization of the samples is removed prior to this step; however, the overprint is generally still apparent.
These samples illustrate that the ChRM can be resolved by detailed demagnetization of samples taken from near the center of the core (Fig. F172), where typically 20- to 30-mT demagnetization removes the drilling overprint. Samples near the periphery, however, are seriously affected by the drilling overprint, often requiring more than 50 mT to remove the overprint; in some cases it is not obvious that any other component is present. By 60-mT demagnetization, only a few percent of the initial NRM remains, making it difficult to accurately estimate the direction of the remaining ChRM. Significant biases probably thus persist in most of the split-core and piece measurements for which AF demagnetization did not exceed 40 mT. Small to moderate biases probably persist in most of the 2-cm x 2-cm x 2-cm minicubes, which include portions from near the edge of the core.
The extent of agreement between archive- and working-half measurements varied substantially. Representative examples are presented in Figure F173. Samples with better direction behavior on demagnetization, showing resolvable different directions for overprint and ChRM, also tended to show better agreement between working and archive halves. We found it useful to assign a subjective quality index based on the extent that the pre-overprint direction could be recognized. An index value of 5 (poor) indicates little change from the overprint direction, and an index of 1 (excellent) indicates several consecutive demagnetization steps showing the same direction, distinct from the overprint direction. An index of 3 (good) indicates a steady progression away from the overprint direction but with failure to stabilize on a new direction. We presume that shore-based analysis of adjacent samples using better-shielded demagnetizers will yield stable directions in these cases. There was a strong correlation between the index value and the rate of loss of intensity with demagnetization. Samples rated better than good retained at least 5% of their NRM intensity at 20-mT demagnetization, and often more than 10%, whereas samples rated poor retained <3% at 20 mT, and often <1%. There was a moderate correlation between NRM intensity and index value, with intensities above 15 A/m common in the poor samples and rare for the samples judged better than good. These generalities on intensities can help evaluate whether a sample retaining a steep inclination during demagnetization has a steep primary inclination or is simply dominated by overprint.
The downhole distribution of quality index values judged from direction behavior shows some correlation between index value and lithology (Fig. F174). For Figure F174 and Table T41, index values were lumped by igneous unit or subunit for simplicity. Samples rated poor are restricted to massive basalts (Subunits 1256D-1b, 1c, 8d, and 24a), judged to be from flows thicker than 3 m (see "Macroscopic Description" in "Igneous Petrology and Geochemistry"). Other massive units, however, were rated good to very good (Subunits 1256D-4d and 8a and Unit 1256D-15). Within units, more stable directional behavior was often observed in samples from chilled margins than from flow interiors. Clearly, cooling rate is an important factor in determining magnetic stability for these lavas.
Inclination values from the units judged better than good quality scatter about nearly horizontal values for Hole 1256D (Table T41), as expected for the nearly equatorial original latitude. These inclination values should be considered tentative and approximate, subject to possible revision if shore-based studies remove more of the drilling overprint. If they are reasonably accurate, the changes in inclination between most of the units suggest time intervals of at least many decades between eruption of the successive units.
For Hole 1256C, sampled at a depth range for which all samples in Hole 1256D were rated fair to poor, the lack of change in inclination makes it more difficult to evaluate stability using direction behavior alone. Sample 206-1256C-11R-3, 62-64 cm (Fig. F166D), shows excellent directionally stability and high coercivity and by itself provides good evidence that igneous Unit 1256C-18 cooled in a steep magnetic field. Samples 206-1256C-6R-1, 40-42 cm (Fig. F166C), 6R-5, 80-82 cm, and 13R-1, 9-11 cm (Table T36), have fairly stable declinations distinct from 180° and retain 5%-7% of their NRM intensity at 20-mT demagnetization, properties shared by samples from Hole 1256D that were rated very good based on their trend to stable shallow inclinations. Although subject to confirmation by shore-based studies, these samples suggest that most or all of the lavas recovered from Hole 1256C erupted during an interval when the magnetic field was steep, perhaps during the Chron C5Br-C5Bn magnetic reversal (see "Location and Regional Geology of Site 1256" in "Background and Objectives"). The maximum time interval that such a direction might be maintained at low latitude is not well known, but is generally presumed to be a few centuries to perhaps a thousand years.
Because the vast majority of studied basalt samples have NRMs that are dominated by drilling overprint, it is difficult to assess the contribution of the in situ section to the magnetic anomalies measured at the sea surface. Most NRM intensities are stronger than the ~5 A/m that is typical of low-latitude Miocene DSDP/ODP samples (e.g., Bleil and Petersen, 1983), and a 500-m thickness of about this intensity was used to successfully model the observed anomaly amplitude (Wilson, 1996). The sample data are consistent either with an interpretation that the drilling overprint has slightly increased the NRM intensity and we have already penetrated the majority of the anomaly source or with an interpretation that the drilling overprint has strongly increased the NRM intensity and we have only penetrated a small fraction of the source. Although it is conceivable that experiments imparting an artificial thermoremanent magnetization may offer some constraints on the original magnetization, in general, heating and cooling experiments produce irreversible changes in magnetic mineralogy. Probably the best opportunity for judging the in situ magnetization is interpretation of the downhole magnetic field measured during logging (see "Magnetic Logging Results" in "Downhole Measurements") along with rock magnetic characterization of the igneous units and additional remanence measurements in a less magnetically noisy environment.
Igneous samples for microbiological studies were collected immediately after core retrieval. The core liners were split on the catwalk or in the saw room prior to curation, and core samples were chosen for microbiological analyses. Samples from Hole 1256C were collected for shipboard contamination tests, and samples from Hole 1256D were collected for shore-based microbiological studies (petrological observation, scanning electron microscope and microprobe analysis, deoxyribonucleic acid [DNA] extraction, in situ hybridization, and cultivation). Samples from Hole 1256C used for contamination tests were all massive basalt. Igneous samples from Hole 1256D selected for microbiological studies represent a range of rock types within the volcanic basement, including glassy hyaloclastite and glassy chilled margins, different types of altered basalts and vein fillings, and massive basalts. Table T42 gives a core-by-core list of samples collected for each type of analysis. Drilling fluid and surface seawater were also collected for shore-based characterization of background levels of microbes potentially introduced during the drilling process.
To estimate the amount of fluid intrusion into the recovered cores, chemical and particulate tracers were deployed as previously described in ODP Technical Note, 28 (Smith et al., 2000). Perfluoro(methylcyclohexane) was used as the chemical tracer, and 0.5-µm latex fluorescent microspheres were used as the particulate tracer. The tracer tests were conducted while coring Cores 206-125C-7R, 9R, and 11R with the RCB.
PFT was continuously fed to the drill water at a rate of 0.8 mL/min, which resulted in a concentration greater than or equal to the limit of solubility (1 µg/g) once the drill fluid reached the bit and well above the detection limit for gas chromatographic analysis (1 pg/g). Calibrations of the gas chromatograph (HP 5890) with standard solutions yielded a slope of 1.26 x 1014 area units/g of PFT (Fig. F175). The tracer was detected in the drill water and on the outer edge of the core in all cases, indicating a successful delivery. The drilling fluid intrusion into the center of the core was below the detection limit.
Latex fluorescent microspheres were deployed during collection of the same cores as the PFT. Fluorescent microspheres were rare in the crushed interior of the rock, even though they were detected on the outside of the cores and in the drill water (see Table T43). Although detected in the interior of the cores, the low abundance of microspheres indicates very low levels of drill water intrusion, which is consistent with the PFT tests.
Hole 1256D provides a rare opportunity for determining whether microbial life is present in crust formed at a fast spreading rate. We focused our search for extant and fossil microbial activity to volcanic glass from Hole 1256D. The location of fresh glass from flow margins, pillow margins, and hyaloclastites are reported in Table T28. Thin sections of volcanic glass made on board ship were examined for "microbial" and chemical alteration patterns. A total of four thin sections were examined; two from hyaloclastites and two from glassy flow margins. None of the four thin sections contained unequivocal textures characteristic of the style of alteration previously attributed to microbial alteration. Instead of the irregular alteration textures attributed to alteration by microbes, smooth alteration fronts with clay minerals replacing the isotropic glass are most common, suggesting chemical alteration as the dominant process. Additional thin sections of glass samples will be prepared and examined for shore-based studies.
One interesting discovery is filamentous textures preserved in a 6.2-mm vein cutting cryptocrystalline basalt in Sample 206-1256C-8R-3, 136-148 cm (Fig. F176). The vein is filled with chalcedony, iron oxyhydroxide, saponite, celadonite, and minor aragonite. Within the iron oxyhydroxide in the vein are curved filaments 5-10 µm wide x 50-100 µm long with curved and irregular morphologies. These filaments are similar in size and morphology to iron-oxidizing bacteria and could represent their fossilized remains. These textures will be the focus of shore-based study.
Surface and drill water samples were collected for shore-based analysis to evaluate the background cell levels and microbial population composition introduced into the formation during drilling. Surface seawater was collected several hundred meters away from the ship (in the upwind and upcurrent direction relative to the ship's position) in sterile Falcon tubes during a trip in the Zodiac. Drill water was collected when the pipe was opened to retrieve the core barrel for most cores (see Table T42). Each drill water and surface seawater sample were divided into two splits, one of which was frozen and maintained at -80°C, and the other kept at 4°C. These samples will be used for shore-based DNA extraction and community analysis for comparison with any microorganisms detected in culture.
The physical properties of the basalts cored in Holes 1256C and 1256D are characterized through a series of measurements on whole-core and split-core sections and discrete samples as described in "Physical Properties" in the "Explanatory Notes" chapter. In addition to the standard suite of physical properties measured during ODP cruises, which includes magnetic susceptibility, density, porosity, P-wave velocity, NGR activity, and thermal conductivity, we also use light reflectance properties at visible and near-infrared wavelengths to estimate qualitative degree of hydration as a factor of alteration.
Whole-core sections were measured on the MST. The measurements consisted of GRA bulk density, NGR activity, and magnetic susceptibility (Figs. F177, F178, F179, F180). Measurement spacing was 2.5 cm for GRA and magnetic susceptibility and 5 cm (count time = 20 s) for NGR measurements. Discrete samples were taken at intervals where paleomagnetic cubes were taken. Thermal conductivity was measured once per core or once every other core (see "Thermal Conductivity" in "Physical Properties" in the "Explanatory Notes" chapter).
The continuous GRA bulk density measurements are compared with discrete samples of wet bulk density (Table T44) from Holes 1256C (Fig. F177) and 1256D (Fig. F178). All density values <2.0 g/cm3 and other clearly erroneous data have been removed. The large scatter in GRA density values results from the discontinuous nature of the hard rock core (empty space and/or rubble zones). The GRA bulk densities are lower than discrete sample bulk densities because the core diameter is less than that of the core liner.
In Hole 1256C, grain density and bulk density values are fairly close, indicating low porosity and, therefore, water content (Fig. F177). From the top of the hole to ~281 mbsf there is a high amount of scatter in both parameters. At 281 mbsf, the top of the massive ponded basalt (Unit 1256C-18), the scatter decreases and values remain consistent to the base of Unit 1256C-18 at ~311 mbsf, below which the data are again highly variable.
In Hole 1256D, the highest densities are present in the more massive Unit 1256D-1 and Subunit 8a and from the lower part of Subunit 1256D-8d to the base of Unit 1256D-10. A step decrease in discrete sample bulk density from an average value of 2.9 to 2.7 g/cm3 marks the base of the massive ponded basalt (Unit 1256D-1) at ~350 mbsf. The density varies systematically downhole with relatively higher values for massive basalts and relatively lower values for sheet flows. The pillow basalts, recovered below the massive ponded unit, have the lowest densities.
Density-velocity systematics display the expected general trend of increasing bulk density correlating with increasing velocity (Fig. F181). The massive basalts generally have the highest bulk densities and velocities, whereas the sheet flow basalts have a high scatter and tend to have higher than expected velocities. The pillow basalts plot with low bulk densities and low velocities. The hyaloclastite samples fall within the range of other values, but this may be due to biased sampling of the more competent material, such that the altered matrix material would not be well represented.
The basalts from Holes 1256C and 1256D have generally low porosities, ranging from ~2% to 6% (Figs. F177, F178). Hole 1256D shows the greatest variability, with the highest porosity values exhibited by the pillow basalts.
Porosity-density systematics are as expected, with an inverse relationship of porosity decreasing with increasing bulk density (Fig. F182). The massive basalts tend to be grouped with low porosities and high velocities, although some scatter from the general trend is observed. The sheet flows have the highest range in values, which generally plot along the expected trend. The pillow basalts tend to have higher porosities and lower bulk densities.
The relationship between porosity and velocity (Fig. F183) displays a trend of increasing porosity with decreasing velocity. The massive basalts and the sheet flows, having generally low porosities, have high velocities, although the variability is also high. The pillow basalts plot at higher porosities with lower velocities.
In Hole 1256C the x-direction P-wave velocities start low at ~4500 m/s then increase sharply to values averaging 5430 m/s from ~256 to 289 mbsf (Fig. F177). From 289 to 294 mbsf, there are low values that correlate with low densities and high porosities as well as with high values in VNIS-based hydration. Below 294 mbsf to the bottom of the hole, there are generally constant velocity measurements of ~5500 m/s. Hole 1256D velocity variations correspond to the lithologic units, with high velocities in the massive basalts and lower velocities in the sheet flows. Velocity was measured in three directions (x-, y-, and z-) on the paleomagnetic cube samples to investigate possible anisotropy. The paleomagnetic cubes were cut at the same depth as the physical property samples. Because of the lack of core orientation relative to geographic north, the values should lie along a one-to-one line (45°) or at least be scattered randomly above and below it (Fig. F184). The data suggest higher velocities in the x-direction, however, which is most likely due to the difference in cutting technique used on the sides of the cubes. Y- and z-directions are cut using the same dual-blade saw, which produces a smoother edge than the single-blade saws used to cut in the x-direction. To avoid this small bias, we compare the y-direction (horizontal) velocity to the z-direction (vertical) velocity (Fig. F184). The comparison indicates that the basalts from Holes 1256C and 1256D do not have significant anisotropy.
In Hole 1256C the NGR activity is relatively low, close to background levels from ~250 to 294 mbsf, at which point there is a step increase of ~4 cps to an average of 16.2 cps for the interval from 294 to 308 mbsf (Fig. F179). This interval has high potassium content (0.53-0.74 wt% K2O) as measured by shipboard ICP-AES (see "Igneous Petrology and Geochemistry"). In Hole 1256D, the NGR activity is low from the top of the hole throughout the massive ponded flow (Unit 1256D-1). At the base of this unit the NGR activity increases slightly in Unit 1256D-2 and the pillow basalts. The values remain fairly constant downhole with slightly lower values exhibited by the less altered massive units (Fig. F180).
Susceptibility is relatively low (<4000 raw meter units) from the top of Hole 1256C down to the top of Unit 1256C-18 (the massive ponded flow) at ~281 mbsf (Fig. F179). The susceptibility is high through the thick flow with a peak at ~291 mbsf (reaching 8000 raw meter units), probably due to a high iron oxide content (more than ~15 wt% Fe2O3; see "Igneous Petrology and Geochemistry"). Beneath the massive flow at ~313 mbsf, the scatter of the susceptibility values increases. Hole 1256D has high susceptibility from the top of the hole to the base of Unit 1256D-1, where there is a step decrease in values from ~4000 to ~1500 raw meter units (Fig. F180). The values then remain relatively low until the base of Unit 1256D-6. Below Unit 1256D-6, the susceptibility measurements appear to increase slightly downhole, although there is increasing scatter with increasing depth, which makes it difficult to determine a trend.
Thermal conductivity measurements were taken once per core or once every other core for Holes 1256C (Fig. F179) and 1256D (Fig. F180). The few measurements made in Hole 1256C range from 1.55 to ~2 W/(m·K). Thermal conductivity values in Hole 1256D are slightly higher in Unit 1256D-1, below which there is a step decrease from ~1.9 to ~1.5 W/(m·K). Below Unit 1256D-1 the values remain fairly uniform at ~1.5 W/(m·K) to the base of the hole.
The extent of alteration of the Site 1256 basalts was estimated with VNIS using measurements taken on the split cores. VNIS has just begun to be used as a technique for basalt analysis, yet it has several additional advantages over traditional DSDP/ODP analytical techniques: measurements take only seconds and are nondestructive, and VNIS is particularly sensitive to hydration and smectite concentration, both of which have presented challenges to interlaboratory measurement consistency. The volumetrically dominant minerals in basalts are pyroxene and plagioclase. Plagioclase is spectrally featureless and therefore undetectable by VNIS. The dominant spectral signature in fresh basalt comes from pyroxene.
Montmorillonites, with an OH absorption band at 1400 nm and a strong water absorption band at 1930 nm, can be detected at concentrations of only a few percent (Vanden Berg and Jarrard, 2002a). Jarrard et al. (in press) found that the amplitude of the low trough depths at these two wavelengths, each normalized to adjacent wavelengths outside the absorption band, were highly correlated for the basalt rocks of Hole 801C, so they combined them into a single measure of relative abundance of montmorillonite. We employ the same technique for the basalts recovered at Site 1256.
Pore water has the same OH and water absorption bands as hydration minerals, so it is important to use VNIS only on dried samples. Archive-half cores are not actively dried, but their split-core surface is exposed to air for ~12-24 hr between the time of core splitting and VNIS measurement, with the result that the surface is visibly dry. Air drying is unlikely to drive off much loosely bound water, so the hydration level measured by VNIS is probably closer to total water than to structural water.
Comparison of VNIS-based alteration intensity to measurements of porosity, grain density, and matrix velocity allows a test of the previously noted pattern of increasing alteration in higher-porosity (and presumably higher permeability) samples. Figure F177 shows that VNIS-based hydration is positively correlated with porosity and inversely correlated with both grain density and matrix velocity in Hole 1256C, confirming this hypothesis (analysis will continue postcruise including all basalts from Site 1256). This pattern probably results from two factors: (1) alteration increases porosity, and (2) high porosity promotes alteration because of high permeability and therefore enhanced fluid flow.
All whole-round core pieces that could be rotated smoothly through 360° were imaged on the Deutsche Montan Technologie (DMT) Digital Color CoreScan system (DMT, 1996, 2000a). The pieces were scanned consistently from right to left, with the top of each piece to the left of the scanner. Contiguous pieces were imaged together where possible. In a number of cases, particularly within the massive ponded flow (Unit 1256C-18, Cores 206-1256C-8R through 11R, and Unit 1256D-1, Cores 206-1256D-2R through 12R), where recovery was high, core pieces with lengths >1 m were broken to fit the core scanner. Often pieces were broken or fractured during drilling and recovery. These were allowed to dry, fitted together, and then shrink-wrapped to hold them in place during scanning.
For each piece scanned, the length of the piece and the depth to the top of the piece was calculated. This information is required at the acquisition stage, as it is entered in the DMT software Digicore (DMT, 2000a). The length of the gap (where pieces were too small or uneven to be scanned effectively) between each scanned piece was also calculated to allow for it in the total core barrel length. All measurements were recorded in the whole-round digital image scanner piece log (Table T45).
Prior to scanning, a red line was marked on each piece to indicate the position of the cut surface. An arrow pointing toward the top was added with a W marked to the right of the line looking upcore to identify the working half. This procedure was adopted so that each piece could initially be rotated according to the ODP reference frame prior to making structural measurements.
In total, 889 whole-round core pieces (~216 m of core) were scanned using this method. This accounts for ~85% of the material recovered from Hole 1256C and ~70% of the material recovered from Hole 1256D (see Table T45). The size and quality of the images vary greatly between cores, depending on the nature of the lithology and corresponding closely to core recovery.
The Digicore software (DMT, 2000a), generates two files: (1) a bitmap image file (.BMP extension) and (2) a text file (.DMT extension), containing specific information for each piece, including the length of the piece, the depth to the top of the piece in mbsf, and the complete file path to the core image. After scanning, each bitmap file was edited in Adobe PhotoShop to adjust the brightness and contrast levels and remove any "empty space." The text files were edited in Microsoft Word Pad to adjust the piece length and the depth interval if the core depths exceeded the drilled interval, as in cases of >100% recovery. Any other information that was entered incorrectly at the acquisition stage was also edited at this point.
All the 360° unwrapped core images for Hole 1256D were rotated using the Digicore software immediately after scanning so that the red lines ran up the center (180° position) of the images. This was done in order to speed up subsequent rotation of images with respect to the ODP reference frame (cut surface oriented east-west) using the CoreLog Integra software during the interpretation stage (DMT, 2000c).
The scanned images were then integrated into core barrel lengths using the Core Recovery Quality Control program (DMT, 2000b). This program requires that all the information is entered in the DMT file in the correct format; otherwise, it will not plot the images. For each core barrel length, the images are plotted on a depth scale according to their ODP curated depths, leaving the appropriate gaps where material was not scanned or not recovered. The depth profiles are output as .EMF files (one for each core barrel length), which can then be opened, edited, and saved in Adobe Illustrator format.
After whole-round core scanning, the core pieces were split and labeled according to ODP convention. The slabbed archive half of every section was allowed to dry and then was scanned, prior to description by the petrologists, using the Geotek camera system. The slabbed images were added to the Adobe Illustrator files and aligned in depth alongside the whole-round images for comparison between the external and internal features of the core (Figs. F185, F186). In Figure F185, note that planar horizontal features, such as veins and fractures, in the slabbed core appear horizontal on the unrolled whole-round image, whereas planar dipping features appear as sinusoids on the whole-round image. Unit boundaries were plotted and sample locations are also included in the Adobe Illustrator files (see example Fig. F186). These files are available in the volume "Supplementary Material" (see the "Supplementary Material" contents list) and are intended for use as a template for detailed postcruise structural and core-log integration studies, in particular for whole-round core and downhole log image correlation. They will aid determination of core (and sample) depth with respect to the downhole logs and core reorientation with respect to true geographic north obtained from the General Purpose Inclinometer Tool (GPIT) on the FMS-sonic tool string.
Analysis of the unrolled whole-round core images includes the picking and measurement of planar features, such as lithologic contacts, veins, faults, and fractures. Dipping planar structures produce sinusoidal-shaped curves on the images such as those observed in Figure F185. The sinusoids can be plotted interactively, and the dip of the feature can be calculated using the DMT CoreLog Integra software (DMT, 2000c). The calculated dips can then be compared with those measured directly from the core pieces by the structural geologists and to those measured on the geographically oriented downhole image logs. Unfortunately, this software was not functional during ODP Leg 206 and detailed structural analyses will thus be performed postcruise.
Correlations between the whole-round core images and electrical or acoustic representations of the borehole wall, derived from downhole log measurements, allow determination of the true core depth (as opposed to ODP curated depth) in intervals with greater or less than 100% recovery. Distinctive features such as lithologic boundaries, fractures, veins, and so on, can be depth-matched to allow repositioning of core pieces. The certainty of the correlation will vary with recovery, the continuity of the pieces, and the number of distinctive features within a particular interval. Individual pieces or sequences of pieces that can be correlated can then be shifted to the appropriate depths with respect to the log data. This enables direct comparison between structural, physical, and chemical properties measured on the core and those recorded downhole.
Individual core pieces (and associated structural data) that can be confidently depth-matched can ultimately be reoriented or rotated so they are oriented with respect to true geographic north using data from the GPIT, which is included in the tool string with both the FMS and the UBI (see "Downhole Measurements" in the "Explanatory Notes" chapter).
Downhole electrical images were not obtained in Hole 1256C because an obstruction was encountered at ~257 mbsf, preventing acquisition of data in the basement section during the FMS-sonic run. Correlation of core images and oriented electrical image logs is thus not possible in this hole. Excellent quality downhole images were, however, recorded in Hole 1256D (see "Downhole Measurements" below).
Preliminary attempts at correlation and reorientation of core pieces, using the methods described in Haggas et al. (2001), show potentially good matches between the unrolled core images and the FMS and UBI data from Hole 1256D. Figure F187 shows several features that can be identified and reasonably well matched on the whole-round core image and in the oriented FMS and UBI image logs. To be accurate and effective, core-log integration routines must be performed with care and rigorously checked. The procedures involved are very time intensive, and as such, further correlation and core reorientation work will be completed postcruise.
Logging operations were conducted in Hole 1256D after it had been cored to a depth of 752 mbsf with a 9.75-in drill bit. Logging operations began at 0915 hr on 28 December 2002. Coring started below the 16-in casing at 276 mbsf. At the conclusion of coring, the hole was prepared for logging and a logging BHA was made up. The pipe was set to 60 m into the casing. Five tool strings were deployed (Fig. F188). The borehole proved to be in very good condition, and no constrictions impeded the passage of the various wireline strings throughout the logging program.
The triple combo tool string was deployed first. It included, from top to bottom, the HNGS, APS, HLDT, DLL tool, and the TAP tool (see "Tool String Configuration and Geophysical Measurements" in "Downhole Measurements" in the "Explanatory Notes" chapter). The tool string was successfully lowered to 750 mbsf, within 2 m of the total hole depth, and was raised, collecting data, without any difficulties. After the main pass, a short repeat pass was made to resample an interval where the WHC failed. The WHC was used on all runs to counter ship heave resulting from the mild sea conditions; however, during the first run, the WHC failed at 529 mbsf and could not be restarted during the run. The WHC was fixed after the triple combo run, and for the remaining runs, the heave compensator operated without further difficulty.
The second tool string deployed was the FMS-sonic (DSI) tool string. When the EMEX power voltage was set to the FMS tool prior to data recording, downhole communication with the tool was lost. Consequently, only the DSI data were recorded during this first run. The tool was then pulled up and the connections were tested. It appeared that the FMS would function if run by itself, so a short FMS string was deployed and three successful passes were made.
The third tool string deployed was the BGR magnetometer. Three attempts were made to deploy this tool, including the last logging run at the end of the logging program. In all attempts the tool failed before it entered the open hole. In the first two runs operations ceased at 1000 and 2500 mbrf, respectively. An initial analysis of the BGR data shows that the data flow ceased because of a malfunction of the gyro unit in the tool. All data sets indicate that a fast rotation rate coupled with rapid vertical acceleration of the tool (both >40°/s) preceded tool failure. To counteract these strong movements, the gyro attempts to compensate but consequently consumes about double the usual current, causing the supply voltage to drop below the minimum operation voltage of the instrument. This in turn terminates data acquisition in the tool. Since the startup of the gyro requires high and variable currents, a restart can not be performed downhole.
The fourth tool string was the UBI. This tool was used in hard rocks for the first time in the history of ODP. One pass was made in a high-resolution mode (100 m/hr), and a short repeat was made in a low-resolution mode between 642 and 736 mbsf (525 m/hr).
The next tool to be deployed was the WST. The air gun was used, and 12 stations located along the whole section were recorded (Table T46). For each station, the air gun fired five to seven times. WST operations were delayed for 45 min because of the proximity of pilot whales.
Finally, as time was available, the BGR magnetometer was redeployed and the speed to run the tool to pipe depth was slowed (600 m/hr, compared to the 3000 m/hr usually used). Unfortunately, at 3680 mbrf operation of the tool ceased again and logging was abandoned. Logging operations ended at 2030 hr on 30 December 2002.
The principal results are shown in Figures F189, F190, and F191. Borehole conditions were excellent during the five runs, and no ledges or obstructions were encountered. Caliper readings from both the triple combo and FMS tool strings show good borehole conditions, with a diameter typically between 11 and 14 in (Fig. F189). Slightly wider sections with thin washouts occur at 348-403, 418-435, 450-473, 530-605, and 678-694 mbsf. The excellent hole conditions over the rest of the interval resulted in particularly good measurements in the contact tools such as density, porosity, and FMS. No TAP data could be recovered from the tool, and the reasons for tool failure are unclear. Sonic velocities measured by the DSI appear to be of variable quality. S-wave velocities are well constrained, but P-wave velocities are of poor quality.
The UBI provided reliable results. This tool is designed for boreholes of smaller size than those typically drilled during ODP legs, and the optimal conditions of deployment place the transducer as close as possible to the formation (see "Ultrasonic Borehole Images" in "Tool String Configurations and Geophysical Measurements" in "Downhole Measurements" in the "Explanatory Notes" chapter). Despite the limitations of a large borehole, the transit time and amplitude of the ultrasonic waves were strong enough to provide good data. However, this tool has to be deployed very slowly to record data (100 m/hr), and, consequently, heave and sticking of the tool can influence the data quality. Numerous zones of sticking were experienced, but as with the FMS, information from the GPIT three-axis accelerometer allows these sticking zones to be corrected for using the Geoframe software. Stick and slip on the FMS and UBI logs were generally a few centimeters where they occurred, although they can be as large as 1 m in rare instances. A comparison of FMS and UBI images is presented in Figure F192. As the impedance contrast between the different rock types drilled in Hole 1256D is high, very good results are obtained with the UBI in terms of lithologic changes and excellent correlations can be made with the FMS. Overall, logging data are of good to excellent quality for most of the measured parameters.
GPIT magnetometer data were also acquired during four logging runs. Taking heave-induced depth shifts into account, comparison of the data from the four GPIT runs shows good reproducibility in magnetic field H- (horizontal) and Z- (vertical) components (Fig. F191). The horizontal magnetometer components Bx and By show sinusoidal variations resulting from tool rotation during logging. The first run (black line) was apparently subject to less rotation. The deviation values for the last three runs coincide well, whereas these values from the first run seem to be reduced by ~1°.
Downhole measurements and images recorded in Hole 1256D show a high degree of variation, reflecting the massive units, thin flows, pillow lavas, and hyaloclastite encountered in Hole 1256D (see "Igneous Petrology and Geochemistry"). Combined results of FMS and UBI images and standard geophysical measurements allow us to distinguish between the different rock types (Figs. F189, F190, F192). In sections with low recovery, the combination of standard downhole measurements and FMS images allows us to determine true thicknesses of lithologic units and flows and to orient structural measurements of Hole 1256D cores. Basement unit numbers used in the next paragraphs are those assigned from visual core descriptions (see "Macroscopic Description" in "Igneous Petrology and Geochemistry"). Three logging intervals were distinguished based on strong changes in the geophysical parameters.
The sediment/igneous basement boundary at 251 mbsf was logged in Hole 1256C, and the data are reported in Figure F55. This boundary is marked by a significant increase in density and resistivity and a corresponding decrease in porosity (Fig. F189).
The top of the igneous basement logged in Hole 1256D consists of a massive lava, producing distinctive FMS images (Figs. F189, F193). Numerous veins are well imaged by the FMS and the UBI. This first unit is characterized by high resistivity (up to 100 
m) and little variation in neutron porosity (5%-6%), density (2.87-2.9 g/cm3), capture cross-section (22-23 cu), and PEF (5.8 b/e-). At 296 mbsf in Hole 1256C and 309 mbsf in Hole 1256D, an isolated spike is observed with a strong decrease in density (1.5 g/cm3), electrical resistivity (5 m), and PEF (1.4 b/e-), suggesting a small faulted or highly altered interval.
In the massive unit, the UBI provided very good data. Veins and fractures are observed which correlate well with those recorded by the FMS. Furthermore, as the UBI records 360° images, numerous subvertical veins can be identified. Correlations with the FMS in terms of the fractures are good, especially in this upper massive unit, and veins and fractures can be matched to structures imaged on cores (see "Digital Imaging").
Just below the massive unit, lithology changes sharply. The NGR log values strongly increase, especially the potassium and, to a lesser degree, the uranium content. In contrast, thorium content significantly decreases. In this interval, massive units, pillows, and hyaloclastites can be distinguished from the electrical and acoustic images.
Massive lavas appear to be more abundant in the lower part of interval II. They are characterized by high electrical resistivity (up to 100 m), low porosity (<6%), high density (2.6-2.8 g/cm3), and low NGR (<6 gAPI) values. Pillow lobes are easily recognized on the FMS and UBI images (for example, Fig. F194 from logging interval III) in the uppermost part of interval II (igneous Unit 1256D-3). Pillow basalts appear on the image logs as circular to elliptical bright patches of varying sizes (20-80 cm diameter). The pillow rims are identified as darker regions of high conductivity, as they are more altered compared with the central part of the pillows. Hyaloclastite zones appear to be abundant in this interval (Fig. F195). They consist of highly heterogeneous material, with resistive material (basalts and basaltic glassy clasts) cemented in a conductive matrix (altered glass). Hyaloclastites do not exhibit any fractures but show a random distribution of generally lower and variable resistivity material. In these intervals, the NGR activity increases (up to 10 gAPI) and the neutron porosity increases (up to 60%). The electrical resistivity is low (<10 m), as is the density (<2.5 g/cm3). From the preliminary analysis of the downhole physical properties and the FMS and UBI data, the true thickness of several units can be determined. For example, four massive units from 472 to 532 mbsf are clearly identified (see, for example, Figs. F188, F193). The first one extends from 472 to 489 mbsf (17 m thick; igneous Subunit 1256D-8d), the second from 494 to 504 mbsf (10 m thick; igneous Unit 1256D-10), the third from 509 to 517 mbsf (8 m thick; possibly igneous Unit 1256D-13), and the last, from 526 to 532 mbsf (6 m thick; igneous Unit 1256D-15).
In this interval, variations in physical properties appear to be less pronounced than in the previous interval. The NGR values are low (~6 gAPI) from 532 to 675 mbsf. In this interval, one peak in the NGR data is recorded at 647 mbsf, corresponding to a strong increase in potassium content (Fig. F190). This peak can be correlated to Core 206-1256B-57R, where a highly altered and celadonite-rich interval was described (see "Hole 1256D" in "Alteration"). Below 673 mbsf, the NGR activity increases to 10 gAPI from 645 to 647 mbsf before decreasing gradually from ~8 to 4 gAPI.
Below 532 mbsf, some massive units can be recognized on the FMS and UBI images (resistive zones with abundant veins), but unlike the massive units in interval II, they do not exhibit high resistivity, PEF, and density values. This may be linked to a difference in alteration or vein abundance between the upper massive units and the lower ones, the units below 532 mbsf being more altered and/or more vein rich than those above (Fig. F196). In interval III, pillows can be identified on the FMS and UBI images. These pillows are mainly present between 697 mbsf and the bottom of the logged section (Fig. F194). Hyaloclastite appears to be more abundant in the upper part of interval III.
In several intervals (at 518, 614, 686, and 462 mbsf) breakouts can be identified on the UBI images. The amplitude log is indicative of both acoustic impedance and borehole wall roughness, and the transit time log records borehole geometry. Breakouts are marked by borehole enlargement in the direction of the minimum horizontal stress; it is then possible to derive the actual in situ stress orientation. In the example presented in Figure F197 (between 685.8 and 687.2 mbsf), the breakouts correspond to the two thick black lines in north and south orientations; thus, the present-day stress derived from this breakout appears to be directed east-west.
In Figure F198, magnetic field components Z and H are shown alongside inclination (I) for GPIT run 1. Magnetic subdivisions were compiled using a qualitative approach by breaking the data set into segments of short lengths of a few meters so that time variations of the magnetic field and subregional anomalies are of subordinate influence only. On the basis of the U-test statistics, magnetic subdivisions are set (Fig. F198, Magn. Subdiv. column).
Major changes in I are marked by arrows, blue for decreasing I, red for increasing I. Changes can be observed at 350 mbsf at the boundary between igneous Units 1256D-1 and 2. The remaining subdivisions are not well correlated lithologic boundaries, although they are quite distinct, particularly the subdivisions at 455, 605, and 690 mbsf, the anomalies in Z amount to several thousand nanotesla. The massive basalt flow of igneous Units 1256D-1 and 2 as well as the pillows of igneous Unit 1256D-3 show a smaller variability in the magnetic field values than the rest of the data, indicating a uniform magnetization. A distinct negative Z-field anomaly stretches across Unit 1256D-6 to Subunit 1256D-8b. A more detailed analysis of the magnetic data must take the regional Earth field, Earth field variations, any possible lateral subregional anomalies, and the local rock magnetization into account and will be performed postcruise.
