Because core recovery was very poor, the following description is based on only a few thin sections and X-ray fluorescence (XRF) analyses taken at widely spaced intervals (Fig. F1). However, although differing considerably in appearance and degree of alteration, we have ascribed all of this material to a single unit: a gabbro and its progressive alteration adjacent to a fault zone.
The hole began in a soft, light-colored clayey material, which had a soapy feel and did not contain any obvious gritty particles beyond a few sparse clasts (one of which was sampled for thin section examination [see paragraphs below]). This clayey material was present in the first two cores: Core 1R with a recovery of 33% and Core 2R with recovery <3%, which suggests that the incoherent material had bottomed out near the top of this core. This material was interpreted as fault gouge and was subsequently found, by use of XRF, to consist of a mixture of the following minerals: talc, chlorite, calcite, ankerite, and serpentine (Table T3). In the fault gouge all three main polytypes of serpentine are present--chrysotile, lizardite, and antigorite. From the X-ray diffraction (XRD) analyses there appears to be a systematic distribution of these mineral types with depth throughout the gouge. Quartz and amphibole are present only in the deepest sample (Section 2R-CC). Talc is almost ubiquitous and chrysotile is present throughout most of Core 1R, whereas calcite and chlorite are present in most samples and also in Section 2R-CC. Antigorite is present only in the uppermost sample and lizardite only in one of the deeper intervals. We assume that differing mineralogical compositions are reflected in the color banding observed in the core (Fig. F2), although this assumption has not been tested. Minerals of the group talc-serpentine-chlorite are common as alteration products of mafic and ultramafic rocks (Deer et al., 1982), suggesting that the protolith for this material can be expected to fall within these compositional types, which is confirmed by fresher material from deeper in the hole (see description below).
This fault gouge contains very sparse, random clasts as noted above. These consist of black, shiny material reminiscent in hand specimen of glass. However, thin-section examination shows that they consist largely of talc with chlorite, calcite, and magnetite. Although not completely homogeneous, these clasts do not seem to contain any clear information on the structure of the protolith.
Continuing downward, the next sectioned rock (Core 2R) also consists of chlorite and talc with accessory magnetite. It has a distinct structure, which may be reflecting the original texture of the protolith. This consists of patches of chlorite, 0.5-1.0 cm in diameter, surrounded by fibrous amphibole possibly of the tremolite-actinolite group (Fig. F3). Lath-shaped bodies may be pseudomorphs after original plagioclase, although this could not be substantiated.
The two thin sections from Core 3R show a similar mineral assemblage and, again, a texture that may be reflecting the original protolith, in spite of the fact that replacement has been 100%.
At 57 mbsf (Section 7R-1) true cohesive metamorphic rocks were recovered, various types of mylonites and cataclasites showing pervasive shearing. They are epidote-rich rocks with layers of quartz, epidote, and latest calcite-rich veins. Sulfides (precise nature not determined) are present in small amounts, either in veins or sporadically distributed. Similar mylonites and cataclasites are found downward to Core 11R, where gabbro makes an appearance. The metamorphic rocks fall into the general category of fault rocks. Mineralogically they are composed of epidote, quartz, chlorite, calcite, and variable amounts of finely comminuted material, which is not resolvable under the microscope. The amount of clasts varies widely, so that some rocks with many clasts are termed breccias, whereas others have only small amounts of clasts and consist largely of the finely crushed material. These finer grained rocks are either mylonites or cataclasites. The former have a strong directional fabric, for example, the sample from Section 9R-1, at ~66 mbsf, where foliation planes within the mylonite are marked by chlorite layers (Fig. F4). This rock also has recognizable angular clasts of clinopyroxene, whereas another rock has clasts of plagioclase, thought to be residual from the protolith. Other rocks (Fig. F5) are cataclasites with no directional fabric. Some are traversed by zones where the crushing is so extreme that the amount of recognizable clasts is so low as to warrant the terms "ultracataclasite" and "ultramylonite" (see "Igneous and Metamorphic Petrology" in the "Explanatory Notes" chapter). Both angular and lensoid clasts consist of epidote and/or quartz, which may result from boudinaged layers formed early in the metamorphism by hydrothermal alteration, or quartz, plagioclase, and pyroxene, which are residual from the gabbro protolith. All these rocks are traversed by late-stage veins of calcite, quartz, and epidote, sometimes with sulfides.
By a depth of 85 mbsf (Section 11R-1), relatively fresh quartz gabbro (Figs. F6, F7, F8) was recovered with a grain size up to ½ cm and a hypiodiomorphic texture. Feldspar is still largely sericitized, although remaining anisotropic, indicating that alteration is only partial. The rocks contain up to 10% quartz.
Below this in Section 12R-1 (~95 mbsf) brecciated gabbro was again recovered. This rock has a large amount of chlorite and epidote with fresh plagioclase, which allows the protolith to be identified as gabbro, although all the pyroxene has been replaced. As in previous cases, this rock is cut by late-stage epidote-quartz veins. It has no directional fabric and is termed a protocataclasite because the amount of clasts is high.
The deepest samples, from Section 13R-1 and 14R-1 (105 mbsf), are relatively fresh gabbros, similar to those mentioned above. A description of the gabbros follows with indications as to their affinities based on their mineralogy, texture, and setting.
The gabbros are coarse grained (maximum crystal sizes observed were ~8 mm), although there are probably considerable variations in grain size and a suspicion of this variability was gained from the sparse material available (Figs. F6, F7, F8). No directional fabrics or structures (such as layering or lamination) have been observed and the fabric is not that of a cumulate gabbro. Clinopyroxene is largely fresh, plagioclase is altered (clouded by fine micaceous material, although sometimes the outer rims are fresh), and relatively large amounts of quartz are present, sometimes as clear anhedral grains and sometimes in patches of granophyre (Fig. F7B). Spectacular dendritic magnetite crystals were observed in one sample, although all have skeletal magnetite. In one sample, the clinopyroxene is undergoing alteration to tremolite-actinolite. Other altered material in the gabbros is now represented by chlorite, although its original nature could not be determined.
In summary, these are relatively evolved magnetite-quartz gabbros, which have presumably crystallized in place because they have a distinctly non-cumulate texture.
Section 180-1117B-1R-1 contained three pebbles of sheared mylonitic rock, which have a glassy appearance on uncut surfaces, likely caused by the previously described alteration to talc and other finely divided layer silicates. None of these rocks were sectioned or analyzed by XRF. Two thin sections were made from the sparse recovery from Hole 1117C. One is a sample of mylonite, very similar to those already described except that the clasts make up only a relatively small part of the rock, and therefore, it would be specified as an ultramylonite. The other sample consists of a very fine grained material, which could not be identified optically, but does not appear to be the same talc-chlorite mix found above. This rock is traversed by two vein systems: an earlier one filled with quartz, and a later one containing a mixture of quartz, chlorite, and sericite.
Chemical analyses by XRF for major and trace elements of two analyzed gabbros are shown in Tables T4 and T5. The two quartz gabbros are significantly different in composition, particularly in terms of Fe2O3, TiO2 and Fe2O3/MgO ratio, an observation very common in such rocks that are not liquid compositions. An analysis of a ferrogabbro from the Samail Ophiolite (Lippard et al., 1986) has been included for comparative purposes and it can be observed that many of the elements fall in the same range.
Features common to the two gabbro samples are the relatively low TiO2 and K2O and similar SiO2 and MgO contents. There is little doubt that we are dealing with similar rock types, but the Semail Ophiolite samples appear to be generally higher in Zr, Cr, and V and lower in Cu. All samples recovered from Site 1117 are altered, the quartz gabbros only marginally, the exact significance of these differences is not clear. The brecciated quartz gabbro (Sample 180-1117A-12R-1, 11-13 cm) appears to be similar to the quartz gabbros.
A distinctly different composition is shown by the chlorite-epidote schist (Sample 180-1117A-9R-1, 85-86 cm), which is high in MgO, Ce, Ni, and Cr (spectacularly so) and low in Na2O. If this rock were derived from the quartz gabbro, it seems that either its protolith was distinctly different from the two analyzed quartz gabbros or that the process was not isochemical.
The fault gouge (Sample 180-1117C-1R-1, 9-10 cm) is very different from any of the others; it is siliceous and poor in both Fe2O3 and MgO and high in Ba, Ce, and Rb. The reason for these extreme chemical variations remains obscure.
In summary, the chemical compositions of the quartz gabbros and their mylonitized equivalent are consistent with their origin as a high-level gabbro in an ophiolite complex, although this is in no way specific and other environments are by no means ruled out. The other two rocks analyzed are distinctly different, but whether this is because of different protoliths or to metasomatic alteration is unknown.
These data are consistent with the drill entering a fault zone and continuing through zones of metamorphosed and brecciated material into relatively unaltered gabbro, which was the protolith for the rocks. The metamorphic rocks are of the type and fabric to be expected from intense shearing and fluid flow along a major fault zone.
The gabbro belongs to the noncumulate, quartz-magnetite variety found in a number of settings, including the upper levels of ophiolites, such as at Oman (Pallister and Hopson, 1981) and Troodos (Robertson and Xenophontos, 1994). Ophiolitic, high-level gabbros have the same features described here: variable grain size, evolved nature (containing quartz and magnetite), presence of zoned minerals (as in the clinopyroxene shown in Fig. F7A), lack directional fabrics or layering, and often show alteration (e.g., of pyroxene to hydrothermal amphibole as shown in Fig. F8). In conjunction with the observations from Site 1114 and to a lesser extent Sites 1110-1113, which show that the top of Moresby Seamount consists largely of dolerites reminiscent of an ophiolitic sheeted dolerite complex, the results are consistent with the hypothesis that this topographic high represents part of an ophiolite complex exhumed by extension along a major fault plane. This interpretation suggests that the gabbros sampled at Site 1117 are the high-level gabbros recognized in other ophiolites.