The structure and evolution of the lower ocean crust is a significant issue in earth science. The present study, along with Leg 118, provides the first opportunity to examine a stratigraphically coherent section of the lower oceanic crust sampled in situ. In all, 866 m of rock was recovered from 1004 m of drilled section during Leg 176 from Hole 735B on the Southwest Indian Ridge. This section reports on the igneous petrology of the cores and begins the discussion of its relevance to previously established models for the structure of the ocean crust. The details of the descriptive methods used to characterize these rocks are given in the "Explanatory Notes" chapter of this volume.
The Hole 735B cores between 504 and 1508 mbsf (Leg 176) were divided into 457 intervals numbered 496 through 952, following in numerical succession the 495 intervals identified in Hole 735B from the Leg 118 core (Dick et al., 1991a). Gabbroic rocks constitute more than 99 vol% of the igneous lithologies, and these, together with 0.5 vol% felsic veins, make up the core. Most of these rocks are coarse grained with a granular to intergranular texture. Rhythmic modal and grain size layering is present in 22 intervals (12 vol% of the core; see Fig. F2 and "Igneous Layering"). Fine-grained "microgabbro" intrudes the coarse-grained gabbros and constitutes 2 vol% of the core. Oxide-rich gabbroic rocks, which make up 7 vol% of the recovered samples, occur throughout the core, but are concentrated between 500 and 700 mbsf. The oxide-rich gabbros form lenses or networks intruding or replacing normal oxide-poor gabbros (see "Rock Types and Geochemistry"). Gabbronorites and orthopyroxene-bearing gabbros, with or without oxides, are present in the Leg 176 core only between 536 and 714 mbsf and make up ~8 vol% of the total volume.
The lithologic intervals generally have considerable variety in mode and grain size; thus, their reported characteristics are averages. Rock types deduced from thin sections and chemical analyses of small samples are not necessarily representative of the average rock type in the interval.
Rock types are defined in "Rock Classification" (in "Igneous Petrology") in the "Explanatory Notes" chapter. The main rock types are distinguished by variations in grain size, and variations in the abundance of olivine, oxides, and orthopyroxene. Two relatively rare rock types in the Leg 176 core, pyroxenite and anorthositic gabbro, are characterized by their respective high modal abundances of augite and plagioclase. We use the term augite to cover the entire range of high-Ca clinopyroxene observed in hand specimens and thin sections. We find only orthopyroxene as the low-Ca pyroxene in the Leg 176 rocks, but pigeonite is present in a few Leg 118 cores. The relative abundances, modes, and grain sizes of the main rock types, based on macroscopic study of the core, are presented in Table T3, along with the average chemical composition of each rock type. Chemical compositions were determined by X-ray fluorescence analysis of small samples from the core (see "Geochemistry"). Rock names for analyzed samples were determined by point counting of thin sections from adjacent core. The modal compositional variation of the 457 intervals is graphically illustrated in Figures F3 and F4.
In the description below of the lithologic units, no distinction is made between the various oxide-rich rock species except for oxide gabbronorite. Thus oxide gabbro comprises oxide gabbro, oxide olivine gabbro, and troctolitic oxide gabbro.
Only a few nongabbroic intervals were found: four troctolites, three diorites, and four clinopyroxenites (or clinopyroxenitic gabbros); the latter are local augite-rich cumulus layers and probably do not represent separate intrusive units. Many more troctolites occur locally in patches within intervals of varitextured gabbro. Felsic veins constitute approximately 0.5% of the core but were only defined as separate intervals if their width exceeded 5 cm (see "Felsic Veins"). The majority of veins are leucodiorite; other types are diorite, trondhjemite, tonalite, and granite.
Most of the rocks are intergranular or granular in texture. Although rocks with subophitic textures are not abundant, they are important within several of the major lithologic units (see "Lithologic Units"). The microgabbros are mainly olivine gabbros, but they have compositions ranging from troctolite through olivine gabbro and gabbro to oxide gabbro; they are presented separately in Table T3. Pegmatitic gabbro constitutes 4 vol% of the core in 29 separate intervals. Photographs of typical rock types are presented in Figures F5, F6, F7, F8, F9, F10, and F11.
Of the major cumulate rock types expected to form in the course of differentiation of a mid-ocean-ridge basalt magma, dunite, troctolite, olivine gabbro, gabbro, gabbronorite, oxide gabbronorite, and oxide gabbro, all are present in the Leg 176 core with the exception of dunite. Given that gabbroic Layer 3 at the SWIR is believed to be only about 2 km thick (Muller et al., 1997), the combined rocks of Legs 118 and 176 may well constitute a fair representation of the gabbroic lower crust generated at a slow-spreading ridge (see also "Rock Types and Geochemistry"). The overall mode and chemical composition of the Leg 176 core is presented in Table T4.
Six new major lithologic units were identified in the core recovered during Leg 176, based on modal mineralogy and the relative abundance of rock types as defined by the 457 reported intervals (see Fig. F2 and below).
Correlations between depth, the distribution of rock types, and unit boundaries are presented in Figure F2 . The boundary positions and the characteristics of lithologic Units VI through XII are listed in Table T5. Unit VI continues from 382 mbsf in the Leg 118 core. The bottom of Unit XII was not reached when drilling ended during Leg 176.
Unit VI is the compound olivine gabbro recovered during Leg 118 and which continues down into the section drilled during Leg 176. It was defined during Leg 118 on the basis of the occurrence of troctolitic and olivine-rich gabbros with minor oxide-bearing gabbros, orthopyroxene-bearing gabbro, and gabbronorites. The lower part of Unit VI recovered during Leg 176, below 504 mbsf, continues with abundant olivine gabbro and troctolitic gabbro intervals and some intervals that are mixtures of the two rock types. The occurrence of Cr-diopside and the high relative abundance of troctolitic rocks with high Mg numbers are evidence of the primitive character of much of Unit VI.
The transition to lithologic Unit VII at 536 mbsf is marked by the occurrence of orthopyroxene-bearing gabbro and gabbronorite, with and without oxides, as well as the disappearance of troctolitic gabbro. Intervals of oxide-bearing rocks are more abundant, and the average grain size decreases relative to the Unit VI rocks recovered during Leg 176. In contrast to the previous unit, the first part of Unit VII has no subophitic intervals. This unit is composed of ~70% orthopyroxene-bearing rocks, which are relatively rare elsewhere in Hole 735B except for Units IX and I in the uppermost part of the Leg 118 core (Unit I extends 0-37 mbsf). The oxide-rich rocks clearly crosscut the gabbros, and some have sheared contacts. Unit VII is divided into two subunits, VIIA (intervals 512-526) and VIIB, at 560 mbsf. The bottom of Subunit VIIA at 560 mbsf coincides with a major fault, the disappearance of olivine gabbro, and the occurrence in Subunit VIIB (intervals 527-546) of intervals with subophitic texture.
The start of Unit VIII at 599 mbsf is marked by a strong increase in grain size and a return to a higher abundance (73%) of olivine gabbro intervals than is present in Unit VI. Some of the intervals in Unit VIII contain subophitic textures, which are rare in Unit VII. Some gabbronorite, oxide-rich gabbronorite, and oxide-rich gabbro intervals are also present in Unit VIII.
Below 670 mbsf, Unit IX is dominated by orthopyroxene-bearing gabbro and gabbronorites (~65%) but contains fewer orthopyroxene-bearing oxide gabbros and oxide gabbronorites than Units VII and VIII. Olivine is not abundant in this unit, there are no subophitic intervals, and the grain size is markedly finer than in Unit VIII. Thus, there are several similarities between Units VII and IX. No pegmatitic gabbro or microgabbro intervals are present in Unit IX. The transition to Unit X is marked by the lowest occurrence of orthopyroxene-bearing rock types.
Below 714 mbsf, orthopyroxene disappears and the abundance of olivine increases. The upper part of Unit X contains the most intense swarm of microgabbro intervals (Fig. F2). Pegmatitic intervals are also common in the upper part of Unit X. Rhythmic igneous layering is present at 827-914 mbsf (see "Igneous Layering"). Oxide gabbros are extremely rare. In the lower part of Unit X, several intervals (698, 700-702, and 704) have metasomatized patches, which appear to have been generated by late percolating melts. The base of Unit X coincides with a shear zone at 960-990 mbsf (Fig. F2).
This unit, starting at 960 mbsf, is distinguished by the occurrence of abundant, thin intervals of oxide gabbro, which are most pronounced around 1025 mbsf. The overall abundance of oxide gabbro, however, is much lower than in Units VII and IX. Rhythmic layering (1138-1220 mbsf), very coarse olivine, and rare subophitic textures are present in the unit. The mylonite interval 879 was not chosen as the boundary between Units XI and XII because interval 880 is geochemically more similar to the overlying rocks than to those of Unit XII (see "Geochemistry").
In Unit XII from 1314 mbsf, olivine gabbro and minor troctolitic gabbro dominate, almost to the total exclusion of other rock types. Only rarely do thin oxide gabbros break this monotonous olivine-rich unit. From the top of the unit to 1390 mbsf, abundant thin microgabbros intrude the main coarse-grained olivine gabbro. This zone resembles the swarm of microgabbro in the upper part of Unit X. Also occurring at this depth in the core are several instances of subvertical, meandering, dikelike pipes of microgabbro. Zones with crosscutting microgabbros (e.g., 880, 882, and 884) were logged as composite intervals. Zones of leucocratic gabbro, often more or less vertically arranged, are present in intervals 930 and 937. Very coarse olivine is present in parts of this unit.
Igneous contacts provide important information about the order of intrusion and the degree of crystallinity at the time of arrival of new batches of magma. A total of 457 lithologic intervals were recognized in the Leg 176 core on the basis of abrupt or gradational changes in grain size, modal mineralogy, or texture. Contacts between them were classified as intrusive, sutured, sharp-sheared, tectonic, gradational, not preserved, and not recovered.
Two contact types are thought to be the result of intrusive relationships between one lithology and another; these are "intrusive" and "sutured." These two contact types are illustrated in Figure F12. Intrusive contacts are defined by brittle deformation of the minerals composing the country rock, resulting in a contact that cuts across grain boundaries (Fig. F12A). A much more subtle form of igneous contact is defined as a "sutured" contact. Such a contact follows the outlines of grain boundaries, and grains in the country rock are not broken to make room for the invading magma (Fig. F12B). This is an obvious textural feature of sutured contacts, and thus the "sutured" term is a purely descriptive one, rather than implying a particular genesis of these contacts. Our interpretation of sutured contacts is that a magma intruded a partly consolidated crystal mush, so that grains of the invaded country rock were easily separated from one another during intrusion.
A third type of contact, "sheared," is ambiguously tectonic or intrusive in origin, because its original character is now obscured by shearing. Such contacts are typically quite sharp, and a minor foliation is developed next to the contact in one or both bordering lithologies and is pulled into the contact; this distinguishes them from gradational sheared contacts, where the shearing is distributed over a thicker zone (several centimeters). There are 24 sheared contacts preserved in the Leg 176 section. Tectonic contacts show extensive shearing, foliation in the surrounding lithologies, and in many cases the development of a mylonite along the contact. Where the mylonite is thicker than 5 cm, it is logged as a separate lithology (as in interval 879). Such contacts may imply that a fault zone has juxtaposed differing lithologies that may not have been in contact originally.
In the case of gradational contacts, one rock type grades into another without a clear, sharp border. Such gradational contacts are interpreted to be intraformational, in that small differences in volatile contents or crystal nucleation and growth rate in the residing magma may have resulted in the formation of the different rock types, rather than the arrival of a new batch of magma. Other gradational contacts may have resulted from the percolation of melt along fractures and grain boundaries beyond the boundaries of the main intruding magma body. This may have been the case for some of the oxide-rich veins and networks present in some parts of the core. Further types include not-preserved contacts, in which case the contact has been completely obscured by metamorphism or low-temperature, brittle deformation, and not-recovered contacts, where the contact occurs between two recovered pieces of the drilled section.
Table T6 lists contacts observed in the Leg 176 section by rock type. Because each contact divides two rock types, each is counted twice in the table, and the total is twice the number of igneous lithologic intervals observed. Some systematics are observable in the data when sorted by rock type. Intrusive contacts are rare in the Leg 176 core (<3% of the total). Sutured contacts were more common (~30% of the total), accounting for 138 contacts (276 in Table T6). Gradational contacts account for 40% to 50% of the contacts for most rock types, with the exception of metagabbros and microgabbros that have mostly sutured and intrusive contacts (~60%), and less abundant gradational contacts (~20%).
Approximately 13% of the contacts (59 of 457) were not recovered. Many of these may be gradational contacts where the recorded contact position was placed between two recovered pieces for convenience in logging. It is unlikely that there is a mechanical reason that gradational contacts may have broken more readily during coring than other contact types. The only rock type with a low percentage of gradational contacts and a high percentage of contacts not recovered is microgabbro, which may be the result of a mechanical weakness associated with intrusive contacts.
The maximum and minimum grain sizes of each of the major constituent minerals (olivine, plagioclase, augite, and orthopyroxene) in each lithographic interval were measured in hand specimen, and in thin section for selected samples during microscopic descriptions. In both cases measurements were made along the long axis of each mineral. The results for the microscopic measurements and the visual core descriptions are given in Figures F13 and F14, respectively.
The most general feature observed in both hand specimen and thin-section measurements is a positive correlation between the grain sizes of olivine, plagioclase, and augite (Figs. F13, F14). These correlations are independent of rock type. The average grain size of samples recovered during Leg 176 varies from fine grained (<1 mm) to pegmatitic (>30 mm), with average grain sizes generally in the range of coarse (5-15 mm) to very coarse (15-30 mm). In general, the relative grain sizes measured in hand specimen for augite, plagioclase, and olivine are in the order augite > plagioclase > olivine. In thin section, measured grain sizes for plagioclase and augite are approximately the same, with plagioclase generally slightly larger than augite, and both larger than olivine. This difference between hand specimen and thin-section measurements probably reflects the ease with which large poikilitic augite crystals can be seen in hand specimen and the misidentification of augite grain boundaries in thin sections where the augite crystals are slightly distorted, so that isolated parts of the same poikilitic crystal have slightly different orientations and, therefore, different birefringence and angle of extinction.
Peaks in average grain size occur at 510, 635, 825, 940, 1100, 1215, 1300, 1425, and 1480 mbsf (Fig. F14). Although the variation in average grain size is dominated by variations in the plagioclase grain size, the grain size data for augite, plagioclase, and olivine all follow similar trends (Fig. F14), with only a few discrepancies. Plagioclase has only very weak peaks at 510 and 1480 mbsf, where augite and olivine have relatively strong peaks. Augite, on the other hand, has a weak peak at 1425 mbsf, where plagioclase and olivine have relatively strong peaks. Olivine has extra peaks at 590 and 650 mbsf that are only relatively minor or do not appear in the plots of plagioclase and pyroxene. Peaks at 1215 and 1300 mbsf are very pronounced in the olivine plot, whereas in both the pyroxene and plagioclase plot, the whole zone from 1215 to 1300 mbsf is coarse, and the two individual peaks are not as well developed. No overall systematic downcore variation in either plagioclase or pyroxene grain size was observed over the depth range from 500 to 1500 mbsf. Olivine, however, has a slight overall decrease in grain size from 500 to 1200 mbsf, and then an abrupt increase to higher grain sizes over the range of 1200 to 1500 mbsf.
There are a number of thin pegmatitic intervals and thin fine-grained intervals in the core. These intervals generally have abrupt contacts with the surrounding coarse-grained gabbro, and both types tend to occur in clusters. With the exception of the intervals around 710 and 800 mbsf, where both pegmatitic and fine-grained intervals are common, these two rock types appear to be distributed antithetically (see Fig. F2).
Modal abundances of the primary magmatic phases were estimated in thin section using a Swift point counter set up to move in increments of 0.5 mm. Approximately 1500 points were counted for each standard (2.5 cm × 4.5 cm) thin section and 2000 to 2500 points for each oversized (5 cm × 7.5 cm) thin section. In the case of secondary minerals, the original precursor minerals were counted. Common pseudomorphs include smectite, talc, and magnesian amphiboles after olivine, and hornblende after clinopyroxene. Where amphibole occurs together with iron-titanium oxides in intergranular areas, the amphibole was counted as primary. Point counting was done in transmitted light; therefore, opaque minerals (oxides and sulfides) were not differentiated.
The proportions of major silicate phases in different rock types from Leg 176 are summarized in Figure F15 and Table T7. On the basis of Mg numbers (see "Rock Types and Geochemistry") and previous work on Hole 735B gabbros (Ozawa et al., 1991; Bloomer et al., 1991), troctolite is the most primitive rock type, followed in turn by olivine gabbro, gabbro, orthopyroxene-bearing gabbro, gabbronorite, oxide gabbronorite, oxide gabbro, and last by oxide olivine gabbro. From troctolite to gabbronorite, the proportion of olivine decreases, whereas that of orthopyroxene increases. From gabbronorite to oxide olivine gabbro, the trend is reversed with increasing olivine and decreasing orthopyroxene abundance. No pigeonite was recognized in the samples recovered during Leg 176, although some was found in thin sections from Leg 118.
The abundance of opaque minerals in thin section is typically less than 0.5% in troctolite, olivine gabbro, and gabbro; close to 1% in gabbronorite; 8.4% on average in the oxide gabbronorite, with a maximum of 18%; and 2% to 3.5% in the oxide gabbro and oxide olivine gabbro. Ilmenite is the dominant opaque phase in the Fe-Ti oxide-bearing gabbros with lesser amounts of magnetite and sulfide. Amphibole is a common accessory phase in many of the gabbros, averaging between 0% and 0.4% in the primitive gabbros and 0.5% to 1.7% in the evolved Fe-Ti oxide-bearing gabbros. Apatite is common in the Fe-Ti oxide-bearing gabbros, reaching a maximum of 2.5% and averaging 0.4% in the oxide gabbro and oxide gabbronorite.
The average phase proportions in Hole 735B troctolites, troctolite gabbros, and olivine gabbros closely resemble cotectic proportions observed in low-pressure experiments on MORB. The average olivine:plagioclase ratio of 30:67 in the troctolites and troctolitic gabbros is close to the 30:70 cotectic ratio determined experimentally (e.g., Grove and Baker, 1984; Grove et al., 1992; Toplis and Carroll, 1995). The average proportions of olivine, plagioclase, and clinopyroxene in the olivine gabbros (9.8:59.5:29.9; Table T7) are nearly identical to the 11:59:30 proportions observed by Grove et al. (1992) in 2-kbar experiments on MORB. These similarities suggest that the main body of olivine gabbro at Hole 735B crystallized at relatively shallow depths (<6 km) and solidified after efficient expulsion of residual melts. If residual melts had not been expelled, the cotectic proportions would not have been preserved, and the olivine gabbros would contain a higher proportion of accessory phases such as ilmenite and apatite.
The more evolved Fe-Ti oxide-bearing gabbros with an average of 1%-2% olivine, 51% plagioclase, 34%-40% clinopyroxene, 0.2%-3.5% orthopyroxene, and 6%-9% Fe-Ti oxides (Table T7) do not have such experimental cotectic analogs. The closest cotectic assemblages are those of Juster et al. (1989), who found proportions of 46%-51% plagioclase, 18%-35% augite, 13%-30% pigeonite, and 1.5%-4.5% ilmenite + magnetite in crystallization experiments on Fe-Ti basalt from the Galapagos spreading center and those of Toplis and Carroll (1995), who reported proportions of 0%-14% olivine, 37%-48% plagioclase, 38%-48% clinopyroxene, and 4%-22% ilmenite + magnetite in crystallization experiments on a synthetic ferrobasalt. The wide disparity in crystallizing assemblages reflects the reaction relationship between low calcium pyroxene and olivine as well as variations in oxygen fugacity, which control the stability of ilmenite and magnetite. From the modal proportions in Hole 735B gabbros, it appears that some of the oxide gabbros formed by in situ crystallization of pooled residual melt. In contrast, other oxide gabbros with excessive (>20%) oxide abundances must have formed by infiltration of Fe-Ti-rich melts or by concentrating oxides from a much larger volume of magma, which was later expelled.
Some of the samples recovered during Leg 176 were deformed and, as a result, have porphyroclastic to mylonitic textures (see "Structural Geology"). In these deformed samples, most of the olivine and much of the plagioclase have been recrystallized to the point that the original rock textures cannot be determined. Other samples have undergone various percentages of metamorphic alteration such that original textures have been at least partly obscured (see "Metamorphic Petrology"). Many samples, however, have not been extensively deformed or altered and appear to preserve original igneous textures. These samples are described in detail in thin-section descriptions, and their petrographic features are summarized in this section.
Most of the fine-grained gabbros, and some of the medium-grained and coarse-grained gabbros have granular textures in which subhedral to anhedral crystals form an interlocking matrix. Most of the coarse samples, however, have poikilitic to ophitic textures in which large, anhedral augite, and to a lesser extent olivine, oikocrysts enclose or partly enclose euhedral to subhedral plagioclase crystals, and intergrown subhedral to anhedral plagioclase crystals fill the volume between the large oikocrysts. In a few samples, euhedral to subhedral olivine crystals are enclosed in later subhedral to anhedral plagioclase and anhedral augite oikocrysts. In at least one sample, olivine is enclosed in plagioclase, and both olivine and plagioclase are enclosed in augite. Some of the more plagioclase-rich samples have intergranular textures in which subhedral to anhedral plagioclase laths form an interlocking framework, filled with smaller anhedral augite and olivine crystals in the interstices. A few samples have cumulate textures, in which coarse to pegmatitic euhedral to subhedral cumulus crystals of plagioclase, augite, and olivine form an interlocking framework filled with smaller anhedral crystals in the interstices. A few samples have preferred mineral orientations. In most of these samples, however, the original igneous fabric is overprinted by a later deformation fabric (see "Structural Geology").
Undeformed plagioclase crystals vary from euhedral to anhedral in form. Euhedral and subhedral crystals typically occur as laths with aspect ratios (width:length) of 1:3 to 1:5. Undeformed olivine occurs as rounded subhedral to anhedral crystals, anhedral amoeboidal crystals, and anhedral ophitic to poikilitic oikocrysts. Subhedral olivine crystals are elongate parallel to extinction directions and typically have aspect ratios of 1:1.5 to 1:2. Augite occurs as tabular subhedral crystals (aspect ratio 1:2 to 1:3), anhedral crystals, and anhedral ophitic to poikilitic oikocrysts. Orthopyroxene is not an abundant phase in the Leg 176 samples. Where present, orthopyroxene is often subhedral with aspect ratios of 1:2 to 1:3. Amphibole is a common, anhedral, interstitial phase in most samples. In some samples amphibole clearly derives from late-stage alteration of original olivine or augite, but in some samples it appears to represent an original igneous phase. In a very few samples, brown amphibole is present as coarse, subhedral to anhedral crystals that enclose smaller, earlier subhedral silicate minerals. Apatite and zircon occur as rare euhedral interstitial crystals or euhedral chadacrysts in opaque oxides and amphibole. Titanite, which ranges from euhedral to anhedral, is an abundant minor constituent in some samples. In most samples, titanite is a replacement of ilmenite or spheroidal euhedra in veins; however, some may have crystallized from late magmatic or deuteric liquids. Quartz and feldspar are present in some of the felsic veins as anhedral crystals or as complex granophyric intergrowths.
Deformation and the development of undulose extinction in plagioclase makes the unambiguous recognition of compositional zoning difficult. In some samples, however, multiple, rectangular zones of changing extinction angle are present, and these roughly parallel the margins of the crystals. Compositional zoning is well developed in plagioclases in some of the felsic veins, and is weakly developed in plagioclase in some of the gabbros. Weakly zoned augite crystals are present in a few samples, and zoned titanite crystals are present in at least one sample.
Exsolution textures are a common feature in most of the thin sections examined. Augite in most samples has very fine-scale, planar, parallel, exsolution lamellae. In many thin sections, two exsolution textures occur in the augite: fine-scale, planar, exsolution lamellae in very light brown host augite and coarser, bleby exsolution in clear host augite (Fig. F16). In both cases, low-Ca pyroxene is exsolved from augite, indicating that the initial augite crystallized from a magma that was saturated or nearly saturated in low-Ca pyroxene. In a few thin sections, herringbone exsolution patterns with fine-scale, planar exsolution lamellae parallel to 001 and 010 are present. In these sections, 001 exsolution was dominant over 010 exsolution, suggesting that exsolved pigeonite is more common than exsolved orthopyroxene and, hence, that the exsolution occurred at high temperatures. In a few sections, two sets of fine-scale, planar, exsolution lamellae were observed at slight angles to each other (~10º), suggesting two periods of pigeonite exsolution at different temperatures. Orthopyroxene crystals typically contain fine-scale, planar exsolution lamellae of augite. Magnetite in some samples contains lamellae of ilmenite indicating oxidation and exsolution during subsolidus cooling. No exsolution was observed in the associated ilmenite.
Many of the silicate minerals in the Hole 735B gabbros contain abundant, small crystals oriented parallel to crystallographic planes or crystallographic directions within the host crystal structure. The origin of these crystals is not clear; they may be inclusions, alteration minerals, or exsolution products. Small brown plates of amphibole are present in many augite crystals, and abundant tiny opaque rods are present in some. Olivine in many thin sections contains abundant transparent brown rods 0.01 to 0.05 mm in diameter and 0.10 to 0.30 mm in length and flattened, opaque dendritic patches up to 0.5 mm across (Fig. F17). In most specimens, the rods are transparent and the dendritic patches are opaque, but in some samples the relationship is reversed, and in a few samples it appears that the two textures may represent different crystal habits of the same mineral phase. In some samples the olivine contains fine needles of sulfide. Plagioclase in many thin sections contains abundant fine opaque rods or needles (as large as 0.001 mm t-transform: none; vertical-align: baseline">× 0.30 mm; Fig. F18). A few plagioclase crystals contain highly birefringent needles (as large as 0.001 mm t-transform: none; vertical-align: baseline">× 0.03 mm), or tiny (<0.001 mm) blebs. Orthopyroxene crystals typically contain abundant fine-scale opaque needles.
Mineral and melt inclusions are not common within the major rock-forming minerals in the Leg 176 samples. As described above, euhedral to anhedral plagioclase chadacrysts are common within augite oikocrysts. In some oikocrysts, apparently isolated anhedral plagioclase crystals have identical albite-twin patterns, suggesting that they were once part of a contiguous crystal that was later partly replaced by the augite host. Rounded to subhedral olivine crystals occur as inclusions in augite and plagioclase in some samples, and small, euhedral spinel crystals are present as inclusions in plagioclase in a few samples. Spherical masses of intergrown sulfides are common inclusions in opaque oxides and amphibole and are rare inclusions in augite.
Olivine in many samples is rimmed by augite, orthopyroxene, or amphibole. Amphibole is also common as rims around augite. Some samples contain wormy plagioclase-augite symplectites, particularly along grain boundaries between large anhedral plagioclase crystals. In a few coarse-grained samples, wormy intergrowths of medium-grained to fine-grained plagioclase and augite are common in interstitial areas and appear to represent initial simultaneous growth from residual magma as is common during the late stages of crystallization in some diabases. Ilmenite-amphibole and ilmenite-augite symplectites are present in a few samples. Granophyric quartz-plagioclase and quartz-microcline intergrowths are common in the felsic veins.