IGNEOUS PETROLOGY

We encountered basement at 1130.37 mbsf (curated depth) at Hole 1183A and drilled 80.7 m of basalt flows that have been divided into eight separate basement units (Table T10) primarily on the basis of limestone and/or hyaloclastite interbeds (see "Igneous Petrology" in the "Explanatory Notes" chapter and "Lithostratigraphy"). The interbeds were designated as "A" subunits and the basalt beneath as "B" subunits. The age of the sediment directly overlying basement Unit 1 is early Aptian (see "Biostratigraphy," Table T5). The limestone interbeds never exceeded 12 cm (in curated thickness) (see Figs. F34, F35; "Lithostratigraphy"; Table T10), but the contacts with the overlying and underlying basalt were not recovered, so these interbeds may be thicker. They are extensively recrystallized, but an approximate age of early Aptian was determined for the limestone interbed forming Unit 3A (see "Biostratigraphy"). The boundary between Units 5 and 6 was defined by the presence of small pieces of hyaloclastite. The boundary between Units 6 and 7 was defined by the presence of a thin (~2 cm) calcite-cemented hyaloclastite breccia containing clasts of basalt glass and recrystallized limestone (Fig. F35). The boundary between Unit 7 and 8 was defined by the presence of highly altered, calcite-cemented hyaloclastite. There is an indication of variations in magnetic inclination in Units 5B and 7 (see "Paleomagnetism"), but we saw no petrologic evidence to subdivide these units. The basement units range in thickness from 0.36 m (Unit 1) to 25.7 m (Unit 6; Table T10). Although the drilling rate became progressively slower as we penetrated deeper into basement, recovery was generally good, averaging 56% and ranging from 25% to 94% (see Fig. F36).

The basement units are typically aphyric to moderately phyric, aphanitic to fine-grained basalts that are slightly to moderately altered (see "Alteration"). Olivine (completely pseudomorphed by smectites with subordinate calcite and/or pyrite) is the dominant phenocryst phase, with smaller amounts of unaltered plagioclase (Fig. F36). Glassy rinds are present throughout the recovered basement section (Fig. F36), and some of the glass is unaltered (e.g., Figs. F37, F38). Immediately adjacent to the glassy rinds are zones of aphanitic basalt 1-2 cm wide; slightly farther from the rind is a 1- to 2-cm sparsely vesicular zone (Fig. F37). Elongate vesicles in this zone are generally aligned perpendicular to the glassy rind and rarely exceed 0.5 cm in length. Grain size increases to fine grained farther from the narrow vesicular zone (Fig. F39). Based on these observations, we concluded that we had cored through a sequence of pillow basalts (see Fig. F40, in which a curved pillow margin is illustrated). Where the glassy rind was not recovered, the same morphology was evident, allowing us to identify aphanitic pillow rims and fine-grained pillow interiors. Core 192-1183A-68R, Unit 8 (Fig. F41), shows evidence of pillow inflation. Apart from the narrow (1-2 cm), sparsely to moderately vesicular zones that are adjacent to pillow margins, the basement units are generally nonvesicular.

Round to subround plagioclase-rich xenoliths ranging from 0.5 to 3 cm in diameter are present throughout the basement sequence (Figs. F36, F42, F43). They are most common in Units 5B and 7 (Fig. F36) but are found in all units except Unit 1, where only 0.36 m was recovered, and Unit 8, the base of which was not reached.

One of the objectives for Site 1183 was to determine the eruptive environment (subaerial, shallow submarine, or deep submarine) of the final volcanic activity in the shallowest part of the Ontong Java Plateau. Based on the recovery of an essentially nonvesicular series of pillow basalts, volcanism was entirely submarine and either eruption took place at depths that prevented abundant vesicle formation (probably >800 m; Moore and Shilling, 1973) and/or the magma was volatile poor.

We described 22 thin sections from both pillow rims and interiors, including some samples selected for xenolith and alteration features. The degree of alteration is highly dependent on the proximity to secondary veins but is generally slight to moderate (see "Alteration"). For the complete thin section descriptions see "Site 1183 Thin Sections".

The textures exhibited in thin section demonstrate the different cooling rates between pillow rims and interiors. In order to save the glassy rinds for shore-based studies, we did not sample this material on the ship. Early-crystallizing olivine and plagioclase phenocrysts are best observed in the aphanitic pillow rims (Fig. F44) and are sometimes difficult to identify in the pillow interiors where the groundmass is more coarsely crystalline (e.g., Fig. F45). Thin sections demonstrate that the majority of the basalt flows are sparsely olivine (±plagioclase) phyric and all are essentially nonvesicular. Small (0.1-1.2 mm) euhedral to subhedral pseudomorphs after olivine were observed in all thin sections. The dominant replacement minerals include smectite (nontronite group), celadonite, saponite, and calcite (e.g., Figs. F44, F46, F47; see "Alteration"). Calcite replacement becomes more prevalent at greater depths. Rare localized accumulations of olivine phenocrysts result in some samples being classified as moderately olivine (±plagioclase) phyric. Plagioclase phenocrysts (0.1-0.4 mm wide) are less abundant than olivine; they form euhedral to subhedral tabular crystals that usually exhibit oscillatory zoning. Some plagioclase phenocrysts contain minute, partly devitrified glass inclusions (Fig. F48). Rarely, olivine and plagioclase are found together as glomerocrysts (Fig. F44), and a clinopyroxene-plagioclase glomerocryst is present in one thin section (Sample 192-1183A-64R-2 [Piece 2, 15-17 cm]; Fig. F49). Strained and partially resorbed clinopyroxene crystals observed in three thin sections (e.g., Sample 192-1183A-65R-3 [Piece 2, 18-19 cm]; Fig. F50) are probably xenocrysts. Titanomagnetite is a minor yet ubiquitous groundmass phase (e.g., Fig. F51); it is unaltered and exhibits skeletal to dendritic or trellis morphology. Primary sulfide is rare in all but three of the thin sections; one from Unit 5B (Sample 192-1183A-59R-1 [Piece 7C, 107-109 cm]) and two from Unit 7 (Samples 192-1183A-64R-2 [Piece 10B, 136-138 cm] and 65R-3 [Piece 2, 18-19 cm]). In these sections, sulfide is found as 0.01-mm subround to elongate blebs in the groundmass and as inclusions in groundmass clinopyroxene and plagioclase; it is tentatively identified as pentlandite and/or possibly troilite. Otherwise, primary sulfide is found only as <0.005-mm inclusions (too small for accurate petrographic identification) in groundmass clinopyroxene and plagioclase.

The groundmass in thin sections of pillow rims close to glassy rinds ranges from partly to totally glassy and exhibits subvariolitic, variolitic or spherulitic (cryptocrystalline) textures with skeletal plagioclase and mesostasis. The mesostasis consists of elongate plagioclase, fibrous clinopyroxene, and glass (e.g., Figs. F51, F52), which is altered to brown clay (smectite group?). Rare, irregularly shaped cavities (1.5 mm) filled with green clay (nontronite and/or celadonite) are also present (see "Alteration").

Fine-grained samples from pillow interiors typically have subophitic, intergranular, and/or intersertal textures (e.g., Figs. F53, F54). Quenched domains showing a subvariolitic texture are present locally, as is subtrachytic alignment of plagioclase. The groundmass consists of plagioclase and clinopyroxene, with minor titanomagnetite and meso-stasis composed of devitrified and altered glass. Plagioclase is generally slightly more abundant than clinopyroxene and ranges in morphology from acicular-skeletal to elongate laths. Clinopyroxene is anhedral to subhedral and mainly equant. Some clinopyroxene exhibits classic bow-tie intergrowths with plagioclase.

One interesting feature is the grain-size banding exhibited in Sample 192-1183A-68R-1 (Piece 3A, 32-35 cm) from Unit 8 (Fig. F41). This core section is from a fine-grained pillow interior and shows alternating hypohyaline and hypocrystalline layers with subvariolitic to intersertal textures and localized subtrachytic patches. This banding is consistent with differential cooling rates during repeated pillow inflation through magma injection. In Sample 192-1183A-67R-1 (Piece 2C, 46-48 cm), the glass-rich layers are subparallel with the elongate subtrachytic patches. Olivine phenocrysts are concentrated in the more crystalline bands and plagioclase laths are aligned around these crystals (Fig. F55).

Thin sections of xenoliths demonstrate an interlocking network of medium to coarse grains of plagioclase (1-10 mm) with interstitial patches consisting of devitrified glassy mesostasis containing crystallites of clinopyroxene and plagioclase with brown clay and titanomagnetite. The plagioclase crystals exhibit fine-scale, complex, oscillatory zoning with truncations indicative of resorption (Fig. F56). One interstitial patch in Sample 192-1183A-57R-3 (Piece 2, 15-17 cm) contains a partially resorbed clinopyroxene crystal (Fig. F57), but, in most cases, the interstitial material appears to consist of quenched (devitrified) basaltic glass. Rarely, these interstitial patches have shapes suggesting that mafic phases may originally have been present (Fig. F58). Elsewhere, the shape of the interstitial space is controlled by the large, interlocking plagioclase crystals (Fig. F59), producing subhedral to anhedral outlines similar to that of the partially resorbed clinopyroxene in Figure F57. At the perimeter of the xenolith in Sample 192-1183A-57R-3 (Piece 2, 15-17 cm), we again observed partially resorbed clinopyroxene crystals (Fig. F60). Disaggregation of these xenoliths may be responsible for the presence of rare, large, partially resorbed clinopyroxene xenocrysts (see Fig. F50).

We selected 11 samples for whole-rock analysis by inductively coupled plasma-atomic emission spectrometry (ICP-AES). Weight loss on ignition (LOI) data suggest that the degree of alteration decreases downhole in the basement sequence; LOI values decrease from 2.4 wt% in Units 1-3B to <0.6 wt% in the rest of the sequence (Table T11). All basalts analyzed are tholeiitic, and all but one (Sample 192-1183A-64R-2 [Piece 2, 15-17 cm]) are olivine normative (Table T11).

The sequence of basalts and limestone interbeds recovered at Site 1183 is similar to the sequences recovered at Sites 803 and 807, except that we did not encounter thick, massive basalt flows like those reported at Site 807 (e.g., Unit F, a 28-m-thick flow) (Kroenke, Berger, Janecek, et al., 1991). At Site 803, the basalt was described as fine grained and sparsely phyric with plagioclase, clinopyroxene, and olivine phenocrysts (Kroenke, Berger, Janecek, et al., 1991). In three instances we observed rare large phenocrysts of plagioclase and clinopyroxene (up to 1 cm). At Site 807, Unit A basalt was described as very fine grained with rare euhedral plagioclase phenocrysts up to 3 mm long. Basalt forming Units C-G was described as very fine grained to fine grained with olivine (pseudomorphs), plagioclase, and clinopyroxene phenocrysts (Kroenke, Berger, Janecek, et al., 1991) (see Fig. F61). At Site 1183, the phenocryst phases are olivine (altered to smectites, celadonite, and subordinate calcite) and plagioclase. Basalt from Site 807 seems to be slightly more evolved than that recovered from Site 1183. Basalt from both Sites 807 and 1183 contains clinopyroxene-plagioclase glomerocrysts (cf. Figs. F62, F49) and partially resorbed clinopyroxene xenocrysts (cf. Figs. F63, F50), although xenocrysts and glomerocrysts appear to be more abundant at Site 807. The one feature that sets the basalt recovered during Leg 192 apart from that recovered during previous drilling legs is the apparently ubiquitous presence of plagioclase-rich xenoliths (Fig. F36). However, small (3-5 mm) clusters of plagioclase crystals are present in one thin section from the Units C-G basalt of Site 807 (e.g., Sample 130-807C-93R-1 [Piece 39, 137-139 cm]; see Fig. F64), indicating that plagioclase-rich xenoliths are present, if not abundant, at Site 807.

Plagioclase-rich xenoliths are also present on the island of Malaita in the Solomon Islands, nearly 1000 km southeast of Site 1183. These xenoliths are present only in basalt of the Kwaimbaita Formation, with which the Units C-G basalt at Site 807 has been linked by Tejada et al. (in press). The xenoliths appear cumulate in nature, and the compositional zoning in the plagioclase (e.g., Fig. F56) indicates complex magma chamber processes. The presence of these xenoliths in petrographically similar basalt from widely separated areas of the plateau has implications for magma chamber size, processes, and/or the geographical extent of these flows.

The similarity between Site 1183 basalt and that from Units C-G at Site 807 is also evident from the geochemical data. The 11 basalt samples that we analyzed by shipboard ICP-AES are all low-K tholeiities (Fig. F65) and have slightly lower (on average) incompatible trace element abundances than the basalt of Units C-G (Fig. F66). Generally, however, the compositions of Site 1183 lava flows are virtually identical to those of Units C-G (Figs. F66, F67). Furthermore, the geochemical analyses (Table T11) show that lava flows comparable to Unit A basalts of Site 807 and the Singgalo Formation basalts of Malaita, both of which form the upper parts of their respective lava piles, are apparently absent from Site 1183.