The nature and evolution of the oceanic crust and spreading-center magma chambers are two of the most important issues in modern marine geology. Considerable research efforts have been directed toward revealing the structure and constituents of the lower oceanic crust and understanding magma chamber processes at spreading centers. Models for the nature of oceanic magma chambers have traditionally been derived from studies of layered intrusions in orogenic and cratonic environments (Wager and Brown, 1967; McBirney, 1995; Cawthorn, 1996) and from ophiolitic complexes of uncertain provenances (e.g., Moores and Vine, 1971; Pallister and Hopson, 1981). However, recent geophysical surveys and dredging conducted at mid-oceanic ridges have greatly enhanced and completely transformed the way we view the nature of oceanic magma chambers, their solidification processes, and magma migration and its crustal interactions (e.g., Fox and Stroup, 1981; Bloomer et al., 1989; Meyer et al., 1989; Sinton and Detrick, 1992). Of particular relevance in this context have been ODP's drilling efforts into the lower oceanic crust, first at Site 735 near the Atlantis II Fracture Zone of the Indian Ocean during Leg 118 (Robinson, Von Herzen, et al., 1989), then in the Hess Deep of the Atlantic Ocean during Leg 147 (Gillis, Mével, Allan, et al., 1993), and, subsequently, near the Kane Fracture Zone of the Atlantic Ocean during Leg 153 (Cannat, Karson, Miller, et al., 1995). Recently, the JOIDES Resolution revisited the SWIR during Leg 176 in an attempt to deepen Hole 735B (Dick, Natland, Miller, et al., 1999) and was successful in penetrating to about 1500 mbsf with excellent recovery. During Leg 179, a 158-m section of oceanic gabbros was cored at Site 1105, which is offset ~1.2 km from Hole 735B. Hole 1105A provides the first opportunity to assess the lateral variability of magma-chamber processes in the region.
The following overview of the petrology of the core recovered from Hole 1105A during Leg 179 is based on the initial visual core descriptions (VCDs) as well as examination of the available thin sections and chemical analyses performed on board during the leg. The large variability in modes, grain sizes, and textures within individual intervals necessitates that the reported characteristics are given as averages. For the same reason, the information obtained from individual thin sections and X-ray fluorescence analyses are not necessarily representative of the broader interval from which they were obtained.
The core was divided into 141 intervals based on variations in mineralogy, mineral mode, texture, and grain size as described in "Igneous Petrology and Geochemistry" in the "Explanatory Notes" chapter. The detailed information on these lithologic intervals is given in the VCDs. Gabbroic rocks constitute by far the majority of the recovered igneous lithologies (>99% by volume) with only a very minor component of felsic veins (<1%). The gabbroic rocks have mostly medium- to coarse-grained granular textures and display modal, textural, and grain-size variations. Regularly to irregularly developed igneous layering and rare laminations are present within intervals, and many of the interval contacts are marked by layer boundaries. The grain-size variation shows extreme ranges with pegmatitic gabbro in 2%, coarse-grained gabbro in 67%, medium-grained gabbro in 28%, and fine-grained gabbro in 3% of the core. Modally, the gabbroic rocks range from gabbro (36%), to olivine gabbro (43%), oxide gabbro (17%), oxide olivine gabbro (4%), and minor amounts of gabbronorite. The principal gabbroic rock types used in the description are defined based on the presence of olivine, orthopyroxene, and/or Fe-Ti oxide minerals in proportions as defined in "Rock Classification" in ("Igneous Petrology and Geochemistry") in the "Explanatory Notes" chapter.
The lithologic intervals have been grouped into summary units for the ease of description and to facilitate comparison with the core recovered during Leg 118 (Robinson, Von Herzen, et al., 1989; Dick et al., 1991a). The units are based on modal and grain-size characteristics and are illustrated in Figure F21. The downhole variations in modal composition and grain size are illustrated in Figures F22, F23, and F24.
Lithologic Unit I includes intervals 1 through 23 (15.00-48.14 mbsf, Sections 179-1105A-1R-1 to 7R-1) and is composed of 33.14 m of alternating gabbro, olivine gabbro, and oxide olivine gabbro (Fig. F21). The lower boundary is defined by the appearance of a higher proportion of oxide gabbro and is clearly revealed by the abrupt change in average magnetic susceptibility (Fig. F22). Medium- to coarse-grained oxide olivine gabbro (8%) is confined to the lower part of the unit and only is present in intervals 19 through 22 (37.48-38.40 mbsf, Sections 179-1105A-4R-4 to 5R-2). The rest of Unit I is composed of medium- to coarse-grained gabbro and olivine gabbro (92%).
Extensive grain-size variation is present within individual intervals and ranges from fine to medium (e.g., intervals 13 and 16) and medium to pegmatitic (e.g., intervals 1, 8, and 17). The dominant grain size is coarse grained (68%) with minor components of medium-grained (31%) and pegmatitic gabbro (1%). Unit I is characterized by anhedral to subhedral granular textures with olivine (1%-20%), plagioclase (50%-70%), and clinopyroxene (20%-30%; all in volume percent). Coarser grains of poikilitic clinopyroxene are commonly developed and include smaller grains of plagioclase. Fe-Ti oxides and sulfides are present in trace amounts in gabbro and olivine gabbro. Oxide olivine gabbro is characterized by irregularly distributed, interstitial, and disseminated Fe-Ti oxides and sulfides, which may be highly concentrated in localized patches and bands. These oxide-rich zones are commonly accompanied by ductile deformation and the development of a strong foliated fabric in places (see "Structural Geology"). Primary igneous layering is defined by irregular to gradational grain-size layering and less common weak modal layering. Igneous lamination is not commonly observed in Unit I despite the igneous textures of the rocks. Felsic veins constitute a very minor component (e.g., Sections 179-1105A-2R-3, 5R-2, and 6R-2; Fig. F25). Unit I shows only sporadically developed deformational foliation and alteration (Fig. F26).
Lithologic Unit II includes intervals 24 through 123 (48.14-136.38 mbsf, Sections 179-1105A-7R-1 to 26R-2) and is composed of 88.24 m of alternating gabbro, oxide gabbro, and olivine gabbro (Fig. F21). The lower boundary is defined by a decrease in the proportions of oxide-rich intervals (Fig. F22). Unit II is composed of medium- to coarse-grained oxide and oxide olivine gabbro (29%), gabbro (55%), and olivine gabbro (16%); there is no systematic distribution of the three main rock types. The main difference between Unit I and Unit II is a higher proportion of oxide gabbro in the latter.
Similar to Unit I, many of the defined intervals have considerable grain-size variations from fine to medium (e.g., intervals 30, 38, 40, 42, 56, 77, 118, and 120), fine to coarse (e.g., intervals 89 and 99), medium to coarse (e.g., intervals 33, 39, 67, 70, 78, and 97), and coarse to pegmatitic (e.g., intervals 54, 65, 71, and 100). In addition, the average grain size in the lower part of the unit appears to be larger than in the upper part of the unit (Figs. F23, F24). This allows Unit II to be divided into two subunits. Subunit IIA is defined by intervals 24 through 78 (48.14-92.79 mbsf, Sections 179-1105A-7R-1 to 16R-2) and Subunit II B by intervals 79 through 123 (92.79-136.38 mbsf, Sections 179-1105A-16R-2 to 26R-2). This subdivision corresponds well with the magnetic susceptibility measurements, which show a significant drop in average values at the interval 78/79 boundary (see "Physical Properties"). A reasonable breakdown of Unit II in grain-size categories, based on the estimated dominating lithologies, is for Subunit IIA fine (7%), medium (34%), coarse (56%), and pegmatitic (3%) and for Subunit IIB fine (1%), medium (23%), coarse (72%), and pegmatitic (4%), reflecting the increased grain size in the lower part of Unit II. Uniformly fine-grained gabbro intervals are evenly distributed through Unit II (e.g., intervals 32, 53, 55, 73, 75, 91, and 108).
The two main components of Unit II are oxide-poor gabbroic rocks and oxide-rich gabbroic rocks. Petrographically they are very similar to those found in Unit I. The gabbro and olivine gabbro intervals are composed of olivine (2%-20%), plagioclase (55%-70%), and clinopyroxene (20%-35%) with anhedral to subhedral granular and coarser poikilitic textures. Fe-Ti oxides are a minor component for all gabbro and olivine gabbro intervals. Oxide gabbro is typically composed of olivine (0%-15%), plagioclase (50%-60%), clinopyroxene (25%-40%), Fe-Ti oxides (5%-15%), minor amounts of sulfides, and, rarely, orthopyroxene and apatite. The mode of occurrence of the Fe-Ti oxides is very similar to that observed in Unit I. The Fe-Ti oxides are disseminated grains or grains concentrated in irregular seams. High concentrations of Fe-Ti oxides are commonly characterized by the development of foliated fabrics.
The primary igneous layering is mainly associated with extreme grain-size variation in addition to some modal variation that is not easily distinguished in hand specimen. Unambiguous igneous lamination is sporadic (Fig. F26). Felsic intrusive veins are less commonly observed in the lower part of the Unit (Fig. F25) with the notable exception of an aplitic felsic vein in interval 48. Fine-grained gabbro intervals are throughout Unit II (e.g., intervals 32, 53, 55, 73, 75, 91, and 108). Notable development of deformation and alteration effects are mainly restricted to intervals 76, 77, and 78.
Lithologic Unit III is composed of 14.22 m of gabbro (9%), olivine gabbro (88%), and minor amounts of Fe-Ti oxide gabbro (3%). The unit is defined by intervals 124 through 128 (136.38-150.60 mbsf, Sections 179-1105A-26R-2 to 29R-2; Fig. F21). The upper boundary is located at the bottom of lithologic interval 123 (136.38 mbsf) and is defined by an abrupt decrease in the modal abundance of Fe-Ti oxide as well as in the intensity of the magnetic susceptibility (Fig. F22). The lower boundary is located at the bottom of lithologic interval 128 (150.60 mbsf) and is defined by an increase in the modal abundance of Fe-Ti oxide and in the intensity of the magnetic susceptibility (Fig. F22). Unit III includes five lithologic intervals of which interval 126 (139.12-139.56 mbsf) is composed mainly of Fe-Ti oxide gabbro, whereas the other four intervals are olivine gabbro and olivine-bearing gabbro. As defined, Unit III is a sequence of dominantly olivine gabbro that is underlain and overlain by units characterized by a ubiquitous presence of oxide-rich gabbros. Because the rock types found in Unit III are similar to those in both Subunit IIB and Unit IV, the defined unit boundaries are without genetic connotations.
The principal rock types in Unit III are olivine gabbro and olivine-bearing gabbro. The olivine content ranges from 2% to 10% in visual modal estimates; interval 128 contains an accessory amount of olivine (2%), whereas intervals 124, 125, and 127 contain 10%, 5%, and 5%, respectively. Overall, the modal abundance of olivine appears to decrease downward in the unit. Most of the rocks are massive and coarse to pegmatitic in grain size with large, subhedral clinopyroxene oikocrysts (<55 mm) ubiquitously enclosing small euhedral plagioclase and more rarely subhedral olivine grains. The subhedral granular texture predominates throughout the unit. In some parts, gradational variations in mineral grain size (i.e., fining downward) are present with no clearly defined boundaries. In other areas, weak lamination, defined principally by alignment of elongated minerals, is developed.
A thin interval of Fe-Ti oxide-rich gabbro (interval 126) is present in Section 179-1105A-27R-1 (139.12-139.56 mbsf). This Fe-Ti oxide gabbro is highly variable in grain size (0.5-35 mm), is moderately to intensely altered with the formation of secondary plagioclase and green amphibole, and shows variably developed foliation. The Fe-Ti oxides are aggregates forming seams or irregular layers as well as separate grains with interstitial and irregular outlines. Fe-Ti oxide is closely associated with abundant apatite; this association can be seen in areas where the gabbros retain igneous textures. In contrast, apatite is noticeably lacking in the oxide-free olivine gabbros. This suggests that the Fe-Ti oxide-rich zones represent crystallization products of significantly more evolved magmas than those from which the olivine gabbros crystallized.
Felsic veins and alteration veins are not common in Unit III, although minor thin and discontinuous veins with narrow alteration zones are sporadic throughout the unit (Fig. F25). Mylonitic and porphyroclastic intervals are lacking in this unit.
Lithologic Unit IV is the lowest unit of the core recovered at Hole 1105A (150.60-157.44 mbsf, Section 179-1105A-29R-2 to end of the core in Section 179-1105A-30R-3; Fig. F21). This unit represents a 21.06-m sequence in which Fe-Ti oxide-rich gabbro (37%) and gabbro (63%) intervals alternate. The Fe-Ti oxide-rich gabbro in Unit IV is generally similar to that in Unit III (e.g., interval 126), whereas the gabbro in Unit IV contains either an accessory amount of olivine (0.5%-3%) or completely lacks olivine. The contact with Unit III is marked by significant changes in mineralogy, texture, and physical properties as seen by an abrupt increase in the modal abundance of Fe-Ti oxide and in the intensity of the magnetic susceptibility (Fig. F22). The top 0.85 m of Unit IV (intervals 129, 130, and 131) are relatively massive, but the rest of the unit shows variably developed foliation. This weakly to moderately developed foliation persists throughout most of Unit IV and also distinguishes it from the previous unit.
The Fe-Ti oxide content in the oxide gabbro intervals of Unit IV ranges from 5% to 10% in visual modal estimates. The remainder of the gabbroic rocks are mainly composed of plagioclase and clinopyroxene; the former generally exceeds the latter volumetrically. The oxide gabbros are in 0.15- to 1.18-m-thick intervals (intervals 129, 133, 135, 137, and 140). These gabbros are mostly fine to medium grained, but two coarse-grained intervals (129 and 133) occur with foliated and weakly banded structures accentuated by grain size and modal variations on a centimeter scale. Under the microscope, the banding is characterized by alternating coarser and finer grained layers that are accompanied by thin irregular seams of mafic minerals and/or Fe-Ti oxide minerals. The finer grained layers typically show porphyroclastic textures. The coarser grained layers are in most cases deformed to variable extents, but they retain igneous textures that are mostly subhedral granular. The banding may be relict magmatic modal variations, but the intense overprinted deformation prevents us from elaborating on such interpretations.
The gabbroic rocks in Unit IV are in 0.23- to 1.34-m-thick intervals (intervals 130, 131, 132, 134, 136, 138, 139, and 141). These gabbros contain only minor amounts of olivine and Fe-Ti oxides, or lack olivine, and are generally subhedral granular with sporadic subophitic intergrowth. Except for intervals 130, 131, and 141, which are composed of massive gabbros, the gabbros in Unit IV display variously developed foliation primarily defined by the preferred dimensional orientations of clinopyroxene and plagioclase grains or their recrystallized mineral aggregates. Despite the presence of deformation fabric, portions of these gabbro intervals commonly retain igneous textures. Felsic veins are rare, and thin alteration veins are not common, except in interval 141 (Fig. F25).
The nature of the contacts and the petrographic variability between and within individual intervals provides important constraints on the magmatic and tectonic history of Hole 1105A gabbros. In this section we review the findings from the visual inspection of the core and emphasize the primary igneous aspects. The deformational and structural aspects are dealt with in "Structural Geology". The intervals were visually defined based on differences in grain size, modal content, texture, and structure. We discuss the most important types of contacts, but first briefly consider the forms of the contacts.
The forms of contacts between intervals were defined as intrusive, planar, irregular, gradational, and tectonic. Of the 140 contacts described, 35% were not recorded because of breaks in recovery, 19% were classified as gradational, 38% as planar, 4% as irregular, and 4% as tectonic. None of the contacts was classified as intrusive. The clearly intrusive felsic veins were too small to be defined as individual intervals or lack preserved contacts. Fine-grained gabbro exists throughout the core as small intervals or as parts of intervals.
The tectonic contacts are either very sharp or irregular and are associated with the development of foliation and occasionally mylonitic fabrics (e.g., intervals 24 and 52). If it was possible to identify the original igneous contact, these were described accordingly and not as tectonic. Consequently, the contacts described as tectonic represent those for which the nature of the original igneous contacts could not be inferred.
The majority of the contacts are described as planar and are seen as a relatively abrupt changes in either grain size or modal composition. The gradational contacts show one rock type grading into another without a clearly defined interface on a scale of typically <1 cm. Such contacts may reflect differences in nucleation and growth rates. This simplistic view may require modifications for two reasons. First, it is possible that some contacts described as gradational result from late magmatic or subsolidus recrystallization. Second, melt migration within a semisolid crystal-liquid mush may significantly reshape primary igneous contacts. Such migrations may be driven either by liquid density relations, simple compaction, or deformation controlled compaction. It was argued by the Leg 118 participants that such melt migration could have played a major petrogenetic role at Site 735 (Robinson, Von Herzen, et al., 1989). The similarities in the contact types of the oxide-free and oxide-rich intervals from Hole 1105A do not suggest petrogenetic differences. Both groups have approximately constant planar/gradational abundance contact ratios (0.67/0.33).
Igneous lamination is defined by a magmatic planar alignment of the dominating primocrysts resulting in a relatively well-defined planar fabric to the rock. Well-developed and unambiguous igneous lamination has only been described from two oxide gabbro intervals in Unit II (intervals 43 and 51; Figs. F26, F27). At first sight, this result may be somewhat surprising in view of the plutonic nature of the gabbros. However, thin-section examinations suggest that lamination actually may be more common than indicated by the core descriptions. An example of a relatively well-developed lamination is illustrated in Figure F27 from a fine-grained olivine gabbro (interval 55; Section 179-1105A-12R-2) and is defined by the parallel alignment of plagioclase and mafic phases, including a distinct regular crystallographic alignment of plagioclase twin planes. Abundant examples of igneous lamination, however, are lacking based on thin-section descriptions.
The contacts of the intervals were visually characterized as mainly caused by modal, grain size, and textural differences, or a combination thereof. The summary of statistics for all intervals are (1) modal and grain-size change (58%), (2) grain-size, modal, and textural changes (23%), (3) grain-size change only (5%), (4) grain-size and textural changes (4%), (5) modal changes only (3%), (6) textural changes only (1%), and (7) layering contacts affected by ductile deformation (6%).
It is interesting that in excess of 80% of the contacts involve modal and grain-size changes typical of cumulate layering in continental basic intrusions and in ophiolites. The frequency of textural changes tends to increase with depth in the core. This can be correlated with a downhole increase in the proportion of pegmatitic to coarse grain sizes with depth (Figs. F23, F24). Figure F26 shows the recorded distribution of layering. It can be observed that well-documented layering is mainly concentrated in the upper part of Subunit IIA and in Subunit IIB and Unit IV.
A relatively nondistinct and commonly poorly defined modal layering characterizes 84% of the intervals with contacts described as either planar (67%) or gradational (33%). This type of layering is caused by gradational or abrupt changes in the mineral proportions, typically in the ratio between mafic and felsic minerals or by the appearance or disappearance of a phase (e.g., olivine and Fe-Ti oxide minerals). Such modal variations may give rise to alternating mafic- and felsic-rich layers or, in extreme cases, melanocratic and leucocratic layers. An example of this is seen in Figure F28A from a coarse-grained oxide gabbro (interval 89, Section 179-1105A-18R-2) as mafic layers composed of pyroxenes and oxide minerals that are separated by nearly monomineralic anorthositic layers a few centimeters thick. Another example is illustrated in Figure F28B where the mafic layer shows a finer scaled alternation between mafic and felsic parts, perhaps accentuated by deformation (interval 57, Section 179-1105A-13R-1). This type of layering is sporadic throughout the core and is not only confined to the interval contacts, but also down to a centimeter scale within individual intervals. This finer scaled layering appears to be generally similar to the layering used to define the individual intervals.
A significant part of the layering seen in the core is modally defined by the appearance or disappearance of a mineral phase not present in the layer above or below. Alternating gabbro and olivine gabbro intervals or layers represent examples of modal layering defined by the appearance or disappearance of olivine. The modal layering defined by the appearance or disappearance of Fe-Ti oxide minerals gives rise to alternating layers of olivine gabbro and oxide olivine gabbro or oxide gabbro down to a scale of decimeters. In addition, layering caused by modal variations in the amount of Fe-Ti oxide minerals appears on a small scale (centimeter-millimeter) as irregular layers or seams with high concentrations of Fe-Ti oxide minerals together with olivine and pyroxenes. These layers are mineralogically similar to gabbro in adjacent layers that have lower Fe-Ti oxide concentrations. This is illustrated in Figure F29, which shows a coarse-grained, oxide-bearing gabbro with an ~2-cm-thick central seam of highly concentrated Fe-Ti oxides (interval 71, Section 179-1105A-15R-1). Figure F28A illustrates an example with two seams of oxide gabbro separated by an oxide-bearing anorthositic band. Although the concentrations of Fe-Ti oxides may locally be high (75%), on the scale of an interval the concentrations may average only 2%, classifying some intervals as an oxide-bearing gabbro despite the local presence of Fe-Ti oxide gabbro bands (see "Igneous Lithology, Interval Definitions, and Summary," (in "Igneous Petrology and Geochemistry") in the "Explanatory Notes" chapter).
It is not possible, based on the present observations, to evaluate the origin(s) of the modal layering. Any model must explain both the alternating meter- to decimeter-scale layering as well as the finer scale layering mostly defined by the concentrations of the Fe-Ti oxides. Some of the available information indicates that redistribution of interstitial melt may have played a role either because of compaction or deformation. It is not clear whether a similar process can account for the meter-scale alternation of the two main components in the core (olivine gabbro and oxide olivine gabbro). A solution may be related to the magnitude and distance of migration of interstitial melt, either laterally or vertically, but in situ crystallization or gravitational and flow-controlled crystal accumulation, sorting, as well as nucleation and growth-rate differentiating processes may also have played a role during solidification.
The two most common changes of layer contacts are modal and grain-size changes. Grain-size layering is used to help define 90% of the intervals. The grain-size layering increases significantly in intensity downward in the core (Fig. F26) and may be related to an increase in the grain size and overall coarsening of the gabbros (Figs. F23, F24). This type of layering is defined by variation in grain size from typically medium- to coarse-grained and pegmatitic gabbro. An example of grain-size layering is shown in Figure F30 . Typically, grain-size layering is seen as well defined, gradational to planar contacts with the coarser rocks occupying layers, patches, and pockets in finer grained rocks with perhaps limited lateral extent.
Fine-grained gabbro (or microgabbro) is present in most of the core as a minor component (4%), but it is also in layers and irregular pockets in coarser grained intervals. The fine-grained gabbros range in thickness from a few meters (interval 73, Section 179-1105A-15R-1) to <5 cm (intervals 32, 53, 55, 75, 91, 108, 134, and 138) and possess typically sharp contacts to the coarser grained gabbro (Fig. F31). They are most commonly olivine gabbros with equigranular textures showing rather sharp to irregular contacts and marginal coarsening of the fine-grained gabbro (Fig. F32).
Commonly, the grain-size layering is associated with textural changes from granular to inequigranular to poikilitic textures seen by the development of large clinopyroxene oikocrysts enclosing, or partially enclosing, plagioclase. Grain-size gradation within layers can be "normal" or "reverse." In general, the estimated maximum grain sizes are independent of rock type and increase for both plagioclase and pyroxene downward in the section (Figs. F23, F24). The fact that the dominating variability in the recovered core is related to variations in both modal abundances and grain size suggests that fluctuation in nucleation and growth rates were important during phase equilibrium changes and may have been controlled by fluctuation in the extent of magma mixing, compositional variability of a volatile component, and cooling rate.
Late magmatic veins of felsic material are present throughout most of the core and may represent a large compositional range including trondhjemite and diorite but are too thin to be defined as separate intervals. Despite their volumetrically minor importance, they are of considerable interest since they may represent the final product of magmatic differentiation. Felsic veins are particularly concentrated in the olivine gabbro and gabbro of Unit I and III (Figs. F25, F26). There are a few millimeter- to centimeter-thick veins with sharp and planar subparallel boundaries, which indicate intrusive relations, but irregular veins and patches are also common. A typical example is shown in Figure F33. Assimilation is sporadically suggested by included gabbroic material in some veins.
Rock types are defined in "Rock Classification," (in "Geochemistry") in the "Explanatory Notes" chapter. The main rock types are distinguished by variations in grain size and in the abundances of plagioclase, clinopyroxene, orthopyroxene, olivine, and Fe-Ti oxide. The term clinopyroxene is used for the entire compositional range of Ca-rich pyroxene; most of these are likely augitic pyroxenes, but their compositions are unknown. A minor amount of orthopyroxene is present in the Leg 179 gabbroic rocks, but pigeonite was not found despite being reported in a few cores from Leg 118 (Robinson, Von Herzen, et al., 1989). The relative abundance, modes, and grain-size variations of the main constituent minerals, based on VCDs, are presented in Figures F22, F23, F24, and F34.
A variety of gabbroic and metagabbroic rocks are recovered from Hole 1105A. Thin-section observations indicate that these include olivine gabbro (52%), gabbro (16%), oxide olivine gabbro (14%), oxide gabbro (16%), and oxide gabbronorite (2%). The proportions shown in parenthesis were obtained from the samples for which thin-section analyses were available and, therefore, are not identical to the visual estimates presented above. Obviously, the thin-section populations by no means accurately reflect the actual proportions of each rock type present in the core, but only give rough estimates of their proportions. In addition to these gabbroic rocks, felsic rocks are present as crosscutting veins in the core. We did not recover other types of gabbroic rocks such as troctolite, norite, and hornblende gabbro or any type of ultramafic rocks. Brown hornblende is ubiquitous throughout the core; it is found in >70% of the samples examined under the microscope, but its modal proportion is generally <1%-2% and never >10%. We found two highly olivine-rich samples (troctolitic gabbros) and one mafic mineral-poor sample (anorthositic gabbro); however, this is probably an artifact caused by either uneven distributions of minerals in coarse-grained rocks or small-scale layering. Olivine gabbro and olivine-bearing gabbro are by far the dominant rock types, constituting >70% of the recovered core.
Many of the gabbroic rocks display deformation textures ranging from weakly foliated to porphyroclastic or mylonitic (see "Structural Geology"). The gabbroic rocks with deformation and recrystallization textures represent >40% of the sample population examined under the microscope. This suggests that many of the gabbroic rocks in Hole 1105A are affected by deformation and recrystallization to various extents.
Gabbro (<5% modal olivine) and olivine gabbro (>5% modal olivine) are the most abundant rock types distributed throughout the core. They are the main rock types in all of the lithologic units (Units I-IV). The primary mineral assemblages and modal compositions of gabbro and olivine gabbro are generally olivine (0%-30%), plagioclase (50%-70%), clinopyroxene (10%-40%), orthopyroxene (3%, found in one sample), trace amounts of primary brown hornblende, and opaque minerals (Fe-Ti oxide and sulfides). The modal proportions of olivine are generally <10%. These modal proportions were visually estimated under the microscope, coupled with point-counted modes from 25% of the thin sections. These modal estimates vary significantly because of both the coarse-grained nature of the gabbros and the presence of modal layering (see "Igneous Contacts and Layers"). The gabbro and olivine gabbro are mostly medium to coarse grained but, in many cases, they are pegmatitic.
In general, the medium- and coarse-grained olivine gabbros have subhedral granular textures (Fig. F35). In some cases, these rocks are equigranular, but in others they are inequigranular with subhedral to anhedral, interlocking plagioclase and clinopyroxene crystals. Most of the coarse-grained and pegmatitic olivine gabbros, however, have poikilitic or subophitic textures in which large subhedral clinopyroxene oikocrysts enclose or partly enclose small euhedral to subhedral plagioclase and, to a much lesser extent, olivine (Figs. F36, F37). Although these rocks show pervasively igneous textures, deformation bands and undulatory extinction in olivine crystals (Fig. F38), and curved, tapering twin lamellae and undulatory extinction in plagioclase laths, are common (Fig. F39).
Plagioclase is found as euhedral to subhedral laths. These plagioclase crystals show weak compositional zoning, although it becomes difficult to recognize zoning in the deformed plagioclase crystals in which wavy extinction and deformation twins are commonplace. Symmetrical extinction angles obtained on the polysynthetic albite twins suggest that the compositional zoning of plagioclase is rather narrow, ranging from An55 to An65.
Olivine is present as rounded subhedral crystals or rounded anhedral crystals. The optic angles (2V) of olivine, estimated from interference figures of 70º to 80º, indicate relatively high magnesian compositions. Compositional zoning is not generally optically noted in olivine crystals; however, in one sample, a rounded, zoned olivine crystal enclosed in plagioclase has higher interference colors in the core than the rim with a fairly sharp interface. In several samples, we found a cluster of olivine crystals, which, as a whole, is similar to surrounding plagioclase and clinopyroxene crystals in grain size and appears to retain a single crystal outline (Fig. F39). In rare cases where olivine might have recrystallized, clusters of olivine crystals show a polygonal mosaic texture in which grain boundaries form 120º triple junctions. Olivine crystals generally display deformation bands and undulatory extinction even in gabbros that show typical igneous texture.
Clinopyroxene is present as prismatic to blocky, euhedral to anhedral crystals. Obvious compositional zoning of clinopyroxene is not detected under the microscope. Fine exsolution lamellae of orthopyroxene parallel to the c-axis are ubiquitously present in many clinopyroxene crystals. In several samples, clinopyroxene shows consertal intergrowth with adjacent clinopyroxene (Fig. F40). Some clinopyroxene crystals are clouded by acicular opaque minerals, which are aligned and regularly spaced. In one medium-grained gabbro (Sample 179-1105A-8R-1, 68-70 cm), we found orthopyroxene along with plagioclase, clinopyroxene, and olivine. This sample contains 2.8% modal Fe-Ti oxide (obtained by point counting). The orthopyroxene crystals are subhedral to anhedral with abundant, thin exsolution lamellae and blebs of clinopyroxene.
As noted previously, brown hornblende is an accessory mineral ubiquitous throughout the core (Fig. F41). It is mainly present as fringes to clinopyroxene crystals as well as patchy blebs inside clinopyroxene crystals. Less commonly, it also forms rims around Fe-Ti oxide and olivine. Brown hornblende crystals, in particular those that rim clinopyroxene, tend to be gradationally zoned toward green hornblende (i.e., those close to clinopyroxene are brown and gradually become greenish outward). Also, brown hornblende is generally replaced by fibrous, colorless amphibole (actinolite), which also replaces clinopyroxene. It is not certain whether this brown hornblende is a primary mineral. However, we believe that much of the brown hornblende is a primary igneous miner al for the following reasons:
Previous studies, however, have shown that extensive amounts of brown hornblende in porphyroclastic gneissic gabbro from Hole 735B could be related to high-temperature alteration (e.g., Stakes et al., 1991). As discussed in "Alteration and Metamorphism," some of the brown hornblende in the deformed gabbros from Hole 1105A also could be a product of high-temperature metamorphism. In a few medium-grained gabbro samples, vermicular, symplectic intergrowths of clinopyroxene and plagioclase along the grain boundaries can be found (Fig. F42).
Most of the fine-grained gabbros (microgabbro) are not truly igneous and apparently formed from coarser-grained gabbro by ductile to brittle deformation. There are, however, a few microgabbro samples that display truly igneous textures with a little or no sign of deformation. These microgabbros are anhedral granular with a relatively even distribution in grain sizes. Some of these display distinct igneous lamination defined by subparallel alignment of plagioclase and mafic minerals (Fig. F27). Obvious compositional zoning in these minerals is generally not optically detected, but, in rare cases, plagioclase displays a complex zoning pattern (Fig. F43).
Oxide gabbro (>5% modal Fe-Ti oxide; <5% modal olivine) and oxide olivine gabbro (>5% modal Fe-Ti oxide; >5% modal olivine) are the main rock types in lithologic Units II and IV; they are also present in subordinate amounts in lithologic Units I and III. The primary mineral assemblages and modal compositions of oxide gabbro and oxide olivine gabbro are generally olivine (0%-12%), plagioclase (50%-80%), clinopyroxene (5%-44%), orthopyroxene (0%-4%), opaque minerals (Fe-Ti oxides, sulfides; 5%-12%), and trace amounts of primary brown hornblende, apatite, and titanite. The modal compositions are visual estimates coupled with point counting of 25% of the thin sections. One oxide-rich gabbro sample contains orthopyroxene visually estimated as high as 5% (Fig. F44; Sample 179-1105A-22R-2, 74-78 cm).
Oxide gabbro and oxide olivine gabbro are in many respects similar to the gabbro and olivine gabbro described above, especially in terms of rock textures, mineral morphologies, and mineral intergrowth. Nonetheless, there are some distinct differences between these groups of gabbroic rocks from Hole 1105A. Most notable is the high content of Fe-Ti oxides in these gabbroic rocks when compared to the gabbros and olivine gabbros. Moreover, Fe-Ti oxide minerals are generally <0.5 mm in grain size in gabbro and olivine gabbro, whereas they are generally larger than this size and up to 10 mm in oxide gabbro and oxide olivine gabbro. In particular, they are present not only as single, interstitial fillings between larger silicates but as elongated aggregates forming seams and bands (Fig. F45). In several samples, we found apatite crystals in close association with Fe-Ti oxide seams (Figs. F46, F47). Where found, apatite generally ranges from 1% to 5% modal and contains saline fluid inclusions (Fig. F48). On board, XRF analyses indicated that some of these oxide-rich gabbros have P2O5 contents as high as 4 wt% (see "Geochemistry"). Apatite is characteristically absent from the gabbro and olivine gabbro samples examined under the microscope. The presence of considerable amounts of apatite is one of the significant features that characterizes the oxide-rich gabbros from Hole 1105A.
Leg 118 Shipboard Scientific Party (Robinson, Von Herzen, et al., 1989) noted the following petrographic characteristics that distinguish Fe-Ti oxide gabbro from other types of Hole 735B gabbro:
The characteristics related to pyroxenes listed above are not generall y applicable to the Leg 179 oxide-rich gabbros. Pigeonite is totally lacking in the thin sections of Leg 179 gabbros we examined, and orthopyroxene is present in only four out of 20 oxide-rich samples. This finding differs from the Leg 118 observation, in which the oxide-rich gabbros contained ubiquitous Ca-poor pyroxene. However, since we found orthopyroxene only in one oxide-poor gabbro sample, it appears that orthopyroxene tends to be associated with oxide-rich gabbros. In rarer cases, we found a clinopyroxene simply twinned on {100} with orthopyroxene lamellae leading to herringbone texture (Fig. F49), but it is not so common as in Hole 735B oxide gabbros. Olivine does appear to be generally low in abundance in the Leg 179 oxide-rich gabbros, but gabbros having ~5% modal olivine are not uncommon.
There is little doubt that the gabbroic rocks recovered from Hole 1105A are igneous cumulates and composed of a framework of tou ching cumulus minerals concentrated through fractional crystallization (Irvine, 1982). Although the strongest support for this is provided by the whole-rock concentrations of major oxides and excluded trace elements (see "Geochemistry"), this also is in accordance with textural observations. The relatively unzoned nature of most cumulus phases and their granular interlocking textures suggest that the amount of postcumulus trapped melt is small and classify most of the gabbroic rocks as mesocumulates, adcumulates, and poikilitic adcumulates (or heteradcumulates; Wager et al., 1960; Irvine, 1982). The origin of the interstitial clinopyroxene and Fe-Ti oxide minerals found in many of the intervals in Hole 1105A is less certain, but they appear to have grown dominantly in equilibrium with the coexisting largely unzoned plagioclase and pyroxenes. Despite this, there is still some evidence for postcumulus trapped liquid crystallization. For example, in an anorthositic layer (interval 127 in 179-1105A-27R-3, 93-94 cm), clinopyroxene and olivine fill small interstices between large, euhedral to subhedral plagioclase crystals and may represent postcumulus crystallization. Well-developed evidence for interstitial crystallization (zoning and the appearance of low-temperature mineral phases) is observed in some thin sections but is not commonly found to characterize the gabbroic rocks of Hole 1105A. Additional support for the cumulate nature of the gabbros is provided by the modal layering (including monomineralic anorthositic layers) and the occasional preferred mineral orientations or lamination with plagioclase laths showing aspect ratios of ~1:4 roughly aligned subparallel to each other.
The gabbroic rocks recovered from Hole 1105A were subjected to variable degrees of static hydrothermal alteration and, more remarkably, ductile to brittle deformation accompanied by recrystallization followed by later nonpervasive fracturing and felsic veining. These postmagmatic processes modify incipiently to almost completely or obliterate the original igneous texture (see "Structural Geology"). As a result, the textures that display replacement of primary igneous minerals by secondary and metamorphic minerals or dynamic recrystallization of primary minerals are widespread throughout the core. Thin-section observation indicates that the gabbroic samples with no or only traces of secondary and/or metamorphic minerals constitute 18% of the total number of samples examined under the microscope. The remainder of the samples are altered or metamorphosed to various extents, ranging from 1% modal to as much as 30% modal secondary minerals (Fig. F25). It also appears from cursory inspection that the extent to which these gabbroic rocks are altered or metamorphosed is not closely related to the major igneous rock types such as gabbro, olivine gabbro, oxide gabbro, and oxide olivine gabbro, or to any specific lithologic unit described in "Lithostratigraphy." Instead, it appears to be largely related to the degree in which these rocks are deformed. Detailed petrologic studies of oceanic metagabbros have shown that metamorphic recrystallization is related to either deformation and/or interaction with seawater during cooling of the gabbroic sequence (e.g., Bonatti et al., 1975; Honnorez et al., 1984; Stakes et al., 1991). The alteration and metamorphism observed in Hole 1105A gabbroic rocks are generally in accord with this notion. On a mesoscopic scale, the gabbroic rocks that are altered or metamorphosed to lesser extents are typically massive, whereas those that are altered or metamorphosed to greater extents are foliated or banded. The textures of the massive gabbros grade from medium grained to pegmatitic and are generally subhedral granular and poikilitic; these massive rocks retain the igneous textures. The textures of the deformed gabbros grade from medium- to fine-grained porphyroclastic to fine-grained mylonitic. These porphyroclastic and mylonitic rocks generally no longer retain their original igneous textures (see "Structural Geology").
The minerals positively identified as secondary or metamorphic under the microscope include talc, serpentine, smectite, magnetite, calcite, epidote, chlorite, brown hornblende, green hornblende, actinolite, and plagioclase. In addition, orange-brown to pale greenish brown phlogopitic mica is found in a trace amount in one olivine gabbro sample. It is not certain, however, whether this mica is primary or secondary. The secondary and metamorphic mineralogy of the core is limited to <12 mineral species. This apparent simplicity in mineral assemblages is partly because the gabbroic samples from which thin sections were made were not taken from the late magmatic veins or their vicinities, and we only conducted optical determination of the minerals. It seems likely that more secondary and metamorphic minerals will be identified if further, more rigorous analyses are undertaken. There are three main styles of alteration and metamorphic recrystallization that can be distinguished mainly from textural relations and occurrences observed in the thin sections:
The metamorphic minerals that belong to the first category are those found in porphyroclastic and mylonitic gabbros. The fabrics of these rocks are described in detail in "Structural Geology." In the porphyroclastic and mylonitic gabbros, relict plagioclase, clinopyroxene, olivine, and Fe-Ti oxides generally are porphyroclasts set in a finer grained, recrystallized mosaic dominated by plagioclase neoblasts. The plagioclase neoblasts are relatively strain free, commonly show sutured grain boundaries, and are easy to identify. Clinopyroxene appears to be more resistant to deformation, and olivine tends to be replaced with secondary minerals; therefore, these two minerals are less commonly found as neoblasts. There are, however, a few gneissic samples of tectonic origin, which display a fine-grained, subhedral equigranular texture with weak to strong foliation. The foliation is largely defined by the preferred dimensional orientation of plagioclase, clinopyroxene, orthopyroxene, and lenses of opaque minerals. The plagioclase crystals are relatively strain free, although some display undulatory extinction and deformation twins. In these samples, a minor amount of brown hornblende also is in close association with clinopyroxene and orthopyroxene. The two- pyroxene-bearing mineral assemblage indicates that they could have recrystallized at high temperatures (granulite-grade temperatures). Based on thin-section observation alone, however, we could not completely eliminate a possibility that some of these samples retain igneous remnants without recrystallization. If this is so, then these rocks simply represent weakly deformed noritic mineral assemblages. Further rigorous analysis is needed to resolve this ambiguity; tectonic and igneous textures apparently converge and become difficult to distinguish as the former is formed at increasing higher temperatures.
The secondary and metamorphic minerals that belong to the second category noted above include all the minerals previously mentioned. Neither zeolite group minerals nor prehnite were positively identified in our examination of thin sections. Epidote is present, but it is only found in a trace amount in one sample. In most cases, this group of minerals fringe or completely replace the primary mafic minerals such as olivine and clinopyroxene and form pseudomorphs after these minerals. Fibrous actinolitic amphibole commonly fringes clinopyroxene; less commonly, greenish amphibole, which is either ferrous actinolite or common hornblende, does the same. Some of the greenish amphibole crystals are fibrous, but others are stocky, larger in grain size and show slight brownish tints. This variety of amphibole morphologies suggests the presence of various calcic amphiboles. Chlorite, smectite, and opaque minerals are present along with these amphiboles, replacing clinopyroxene. The pseudomorphs, after olivine, are generally composed of either (1) brown smectite + talc + magnetite, (2) brown smectite + talc + serpentine + magnetite, (3) talc + serpentine + chlorite + magnetite ± smectite or (4) brown smectite + talc + serpentine + magnetite ± actinolite. Opaque minerals and smectite commonly fill fractures and cleavages in olivine. It is noteworthy that clinopyroxene and olivine appear to be more susceptible to alteration or metamorphism under static conditions than plagioclase. Plagioclase alteration is restricted only to a slightly cloudy appearance with the formation of tiny clay minerals or to formation of chlorite, smectite, and/or colorless to greenish amphibole along microfractures and cracks. These fractures and cracks in plagioclase are commonly connected with the pseudomorphs after mafic primary minerals, and no crosscutting relation is observed between them. As for the minerals that belong to the third category, only VCD data are available at the moment. These data indicate that smectite-rich, amphibole + chlorite-rich, and calcite-rich veins are sparse throughout the core.
The alteration and metamorphic reactions in gabbroic intrusions at mid-oceanic ridge environments take place essentially during successive cooling stages (i.e., under a falling-temperature regime); thus, these processes are inherently retrogressive in nature. In continental gabbroic intrusions, reactions between primary magmatic minerals and the water-rich solutions that separate from the same magma at later stages during the process of crystallization are very common and are variously referred to as deuteric alteration or autometamorphism (e.g., Skaergaard intrusion). In fact, many of the petrographic features described above for Hole 1105A gabbroic rocks are virtually indistinguishable from those of deuteric alteration. However, oxygen isotope studies on Hole 735B metagabbros indicate that infiltrating seawater has played an important role in the alteration and metamorphic processes (Kempton et al., 1991). It is very likely that similar infiltration can be found at Site 1105.
The secondary and metamorphic mineral assemblages observed in Hole 1105A gabbroic rocks suggest that, overall, the pervasive alteration or dominant metamorphism probably took place under greenschist facies conditions. Basic rocks of the greenschist facies are characterized by the assemblage albite + epidote + chlorite + actinolite + titanite. In general, albite + epidote are ubiquitously found in the greenschist facies basic rocks in place of primary calcic plagioclase. As noted previously, epidote is conspicuously lacking in the Hole 1105A metagabbros, and plagioclase appears to be more resistant to alteration or metamorphism. Stakes et al. (1991) also reported that epidote is very rare at depths shallower than 150 mbsf in Hole 735B. This is perhaps one of the significant petrographic features that characterizes alteration and metamorphism at Sites 1105 and 735.