IGNEOUS PETROLOGY

Basement was encountered in Sections 191-1179D-10R-1 through 22R-6 (375-475 mbsf). Basement rocks recovered from Hole 1179D consist mostly of fresh aphyric basalts in massive flows, pillows, and breccia with a minor amount of interpillow sediments. The section is divided into 48 units based on lithology and flow and/or cooling unit criteria. All basalts are classified into three petrographic categories based on mineral occurrence: olivine-free basalt, olivine-poor basalt, and olivine-rich basalt. The basalt of the upper eight units is mostly olivine-poor basalt (Group I); the basalt of the middle 16 units is mostly olivine-free basalt (Group II); and all the basalt of the lower 24 units is olivine-rich basalt (Group III). The most distinct petrologic change among cored basalts can be seen between Units 24 and 25.

The basalts of the two upper groups are fine grained with subophitic texture; in the lowermost group, there is a tendency for medium grain size with ophitic fabric in thick massive lava flows. The glass rims from chilled margins on pillows and massive lava flows are mostly palagonitized; near-border parts are now hyalopilitic or cryptocrystalline to microcrystalline.

The groundmass of the basalts consists dominantly of lathlike plagioclase and clinopyroxene (Ti augite) and almost completely devitrified glass/palagonite as mesostasis. Very fine grained magnetite is concentrated in the former glass; Cr spinel, apatite, and very rare zircon are subordinate accessories. Olivine in the groundmass of the Group III basalts is nearly totally changed to iddingsite, as are the olivine phenocrysts in Groups I and III. Plagioclase phenocrysts are mostly fresh but in part corroded or replaced.

The primary mineralogy of the basalts from Site 1179 shows no features of an alkali-basaltic tendency.

The alteration of the basalts is surprisingly low. The alteration belongs to the low-grade zeolite facies in a possible temperature range between 10° and 30°C. Secondary alteration-related mineralogy includes calcite, celadonite, saponite, smectite, and zeolites that fill fissures, veins, and vesicles. Moreover, glass has been replaced in grains or patches, and, to a limited extent, plagioclase is affected. X-ray fluorescence (XRF) analyses show that all samples of basalt from Hole 1179D are geochemically characterized as tholeiitic basalt with mid-ocean-ridge basalt (MORB) affinity.

Prior to identification of lithostratigraphic units, petrographical descriptions based on hand-specimen examination of all basement cores from Hole 1179D were made using a hand lens and binocular microscope. All chilled or glassy contacts, decreases in grain size toward the margins of pillows or flows, and presence of breccia or hyaloclastite were described as indications of unit boundaries. Thin intervals (<10 cm) of sediment or interpillow material and breccia are noted as boundaries between units, whereas thicker intervals (>10 cm) are distinguished as individual units. Furthermore, 46 thin sections were examined to confirm the above descriptions. The VCD form for basement rocks includes all the igneous core description data for Hole 1179D (see "Site 1179 Visual Core Descriptions"): unit number, depth, interval thickness, lithology, texture, structure, color, presence or absence of chilled margins, comments, and the location of shipboard samples. The thin-section descriptions are summarized in "Site 1179 Thin Sections". The basement stratigraphy is summarized in Figure F39.

Basement of Hole 1179D consists mostly of microcrystalline to fine-grained basaltic flows, pillows, and breccia with subordinate amounts of sediment intercalations. The igneous section is divided into 48 units based on lithology and flow and/or cooling unit criteria (see "Site 1179 Visual Core Descriptions," Fig. F39). All basalts are classified into three petrographic categories based on mineral occurrence: olivine-free basalt, olivine-poor basalt, and olivine-rich basalt. The basalts of the upper eight units (Units 1-8) mostly belong to olivine-poor basalt, forming Group I, which contains very small amounts of olivine (<<1%) and plagioclase as phenocrysts. The basalts of the middle 16 units (Units 9-24) mostly belong to olivine-free basalt with a small amount of plagioclase (<1%) as phenocrysts, forming Group II. The basalts of the lower 24 units (Units 25-48) belong to olivine-rich basalt, forming Group III, which contains olivine as both phenocryst (<1.5%) and groundmass (±3%) phases. Additional basalts of the lower 24 units partly contain a small amount of Cr spinel as phenocrysts, indicative of their relatively primitive nature. There is a marked change in color between Units 24 and 25 in Hole 1179D. The upper 24 units show greenish color, whereas the lower 24 units show brownish color. This change appears to result from their original and secondary mineral compositions. The formation of the secondary minerals, such as celadonite, smectite, and saponite, which were formed in vesicles and veins or glassy parts in basalt during alteration process, makes the host rock change to a greenish color. In the lower 24 units, almost all of the olivine crystals have been changed into brownish iddingsite, resulting in the brownish color of olivine-rich basalts. The difference between those two parts is distinct not only in color but also in petrological and geochemical features. This will be referred to in the next two sections. The following paragraph summarizes some of the important characteristics of each lithostratigraphic unit and additional information for the VCD form.

Unit 1 consists of an aphyric basalt flow. The contact between the overlying sediments and basement was not recovered in a single core. Section 191-1179D-11R-1 contains fragments of dark brown chert in the upper 43 cm and aphyric basalt from 43 to 124 cm. The top of the basalt is made up of a highly altered chilled margin (Figs. F40, F41). We conclude that the loss of any igneous core in this interval is very small (i.e., this part might be assigned to the top or almost the top of the basement). The degree of alteration for the basalt throughout Unit 1 is moderate and dark gray in color with pronounced greenish gray and brownish gray concentric alteration halos forming Liesegang structures, which are indicative of an infiltration through the surrounding cracks (Fig. F42). Unit 2 is made up of rather thick (3.31 m) pillows of aphyric basalt with thin calcareous interpillow materials (Fig. F43). Several chilled pillow rims are present within Unit 2, and these define the individual pillows or cooling units. Basalt in the pillow interiors is fresh, and there are very few vesicles throughout Unit 2. Such features are more or less common in other pillow units above Unit 25, such as Units 7, 9, 12, 14, 17, and 23. Unit 3 is an interpillow or interflow volcaniclastic sediment up to 1.0 m thick, which consists of pillow breccia and hyaloclastite with a considerable amount of calcareous matrix (Fig. F44). Because the calcareous matrix contains a high density of calcified radiolarian fossils (see Fig. F45), the matrix is thought to have originally accumulated as radiolarian ooze, indicative of marked cessation of volcanism after the formation of Unit 4. Units 5, 16, 19, and 28 consist mainly of basaltic breccia. All these breccias are composed of moderate to highly altered glassy or microcrystalline basaltic fragments derived from pillow rims or flow margins. They contain considerable amounts of hyaloclastite and carbonate matrix (Fig. F46), with the exception of Unit 19, which has a very small amount of matrix materials. Unit 19 is rather thick (2.25 m), and for the most part, the basalt fragments are subangular to angular and vary from a few millimeters to 10 cm in size. Units 8 and 10 show typical thick (4.65 and 3.78 m, respectively) massive flow structure, although their flow margins are not recovered. Lithology of both units is quite similar, showing fine-grained aphyric basalt, gray to greenish gray in color, with fresh to slight alteration, very rare vesicles, and randomly oriented smectite, calcite, or zeolite veins. All these features are more or less common in other massive flow units above Unit 25, such as Units 4, 6, 11, 13, 15, 18, 20, 21, 22, and 24. There are some massive flow units rich in vesicles (up to 10% in volume) in the lower part, especially in Units 29, 33, 47, and 48 (Fig. F47). The abundance of vesicles in the other massive flow units in the lower part vary from <1% to a maximum of 5% from piece to piece. Unit 44 (3.46 m) consists of rather thick pillows of aphyric basalt with subordinate amounts of pillow breccia. Many chilled pillow rims occur within Unit 44, and thin intervals of hyaloclastite with altered glassy shards occur along pillow rims (Fig. F48). The degree of alteration in the pillow interior is slight, whereas it is moderate in the pillow rim. The amount of vesicles in Unit 44 varies from 1% to 5%. Other pillow units, such as Units 25, 32, 35, 37, 39, 42, and 46 in the lower part, show similar lithology to Unit 44. Fresh glass was found in Unit 28 (Section 191-1179-20R-2 [Piece 4]) and Unit 42 (Section 191-1179-21R-1 [Piece 4]).

The following discussion of petrology at Site 1179 is the result of visual core description and optical microscopy. Problems with the inductively coupled plasma-atomic emission spectrometer during Leg 191 prevented us from generating geochemical data. The 48 units, varying between massive lava flows and pillow lavas, can be categorized by their olivine contents into three major groups (Table T7).

The uppermost Group I consists of fine-grained aphyric basalt with sparse plagioclase and olivine. Less than 1% of tiny olivine phenocrysts is present in massive flows and pillows, and the groundmass is olivine free. Fine-grained to microcrystalline Group II basalts are aphyric to sparsely plagioclase phyric; olivine is not present. All fine- to medium-grained basalts of the lowermost Group III contain olivine not only as phenocrysts (up to 1.8%) but also in the groundmass (~3%). They are sparsely olivine- and olivine and plagioclase-phyric basalts without differences between massive and pillow lavas. Table T8 shows the phenocryst contents in the three groups. In spite of the low percentage of phenocrysts, the differences between the groups are unequivocal. This subdivision is essential for petrologic discussion.

Generally, plagioclase phenocrysts in all three groups are labradorite, normally euhedral, rarely subhedral, simply twinned (most likely after Manebach's law), and partly fresh (Fig. F49A). Some crystals are zoned with a large core and a small rim poorer in anorthite. The majority of phenocrysts are, however, more or less corroded, and a few are totally replaced by smectite, calcite, and zeolite. Some plagioclase phenocrysts contain numerous tiny groundmass inclusions, of which the composition can not be resolved by microscope (Fig. F49B). The lengths of the crystals differ between the three groups and between flow and pillow basalts (except in the chilled margins), as seen above.

Olivine phenocrysts are generally almost totally altered to iddingsite (Fig. F49C), although in many cases, euhedral olivine phenocrysts are observed. No grain-size distinctions could be observed between flow and pillow basalts, but olivines in Group III are three times larger than those from Group I.

A third mineral forming phenocrysts is Cr spinel. However, it was only observed as a very rare accessory phase in four thin sections from Units 30, 41, 45, and 47, which correspond to massive lava flows of Group III. The maximum spinel grain size is 0.25 mm (Fig. F49D).

Two additional rare accessory phenocrysts are apatite and zircon. In all thin sections, only five zircon crystals (10-30 µm) were found within clinopyroxene. Figure F50A shows an example with a small halo and plagioclase. Apatite, in contrast, is clearly more common as 10- to 75-µm large crystals within plagioclase and clinopyroxene (Fig. F50B). Whereas zircon is limited to the massive basalts, apatite is observed in massive and pillow basalts.

The groundmass of pillow and flow basalts in all three groups has the same features. Glassy quenched margins show a wide devitrification spectrum as an expression of different cooling velocities. They are classified into four textural categories in ascending order of degree of devitrification as follows: (1) glass/palagonite radial clusters of initial, impure clinopyroxene ± magnetite (Fig. F50C, F50D); (2) a carpetlike pattern of clinopyroxene bunches with sporadic plagioclase (Fig. F51A); (3) variolitic plumose plagioclase crystallites (Fig. F51B) and variolitic branching plumes of plagioclase or clinopyroxene, the latter commonly with plagioclase crystal nuclei; and (4) plagioclase clusters with associated clinopyroxene and magnetite grains (Fig. F51C) and skeletal crystals.

Somewhat farther from the contact with the older underlying flow or overlying seawater, hyalopilitic and microcrystalline textures may be present (Figs. F51D, F52A). Rosettelike bundles of plagioclase crystals and sub- to anhedral dirty clinopyroxene are observed in this matrix (Fig. F52B, F52C). Within the main internal part of the massive flows and pillows of Groups I and II, subophitic textures dominate (Fig. F52D). In Group III, with distinctly larger grain size, a strongly ophitic fabric is present (Fig. F53A, F53B). Within this sequence, the portion of crystallized phases increases and that of glass/palagonite decreases toward zero: the transition from a hypocrystalline to a holocrystalline basalt has taken place.

Table T9 shows a simplified summary of the distribution of glass/palagonite. The data in this table distinctly reflect the higher portion of glass within the pillows against the massive basalts with a maximum in Group II and a tendency of regression in the massive lava flows with increasing depth. We consider this to be a primary feature and not a consequence of aging or superposition.

Within the groundmass, euhedral to subhedral plagioclase is the dominant mineral phase in all textural varieties. Long, stretched lathlike needles are simply twinned and fresh and sometimes skeletal (Fig. F53C), even if some show partial corrosion (Fig. F53D). Anhedral clinopyroxene and glass/palagonite fill the space between the plagioclases as a kind of mesostasis. Based on their optical properties, the clinopyroxenes belong essentially to the Ti augites, especially in the larger grain-sized Group III. The clinopyroxenes are the essential bearers of the Ti content of the basalts from Hole 1179D. No ilmenite was observed, and the Ti content of the magnetites awaits microprobe examination. Very fine grained partly skeletal magnetite, frequently in single lines or nestlike, is restricted to the glass or its later devitrified products, mainly clinopyroxene.

The average amount of vesicles is generally low, 1% in Groups I and II and ~2% in Group III. The small diameter of the vesicles, commonly between 0.1 and 1 mm and only rarely >3 mm, is clear evidence for a deep-water emplacement of the massive and pillow lava flows. The basalts must have been emplaced above ~4000 m, the depth of the CCD, because on and between pillows and at the top of some massive flows, an interpillow material is embedded. It consists of very fine grained carbonatic sediment (pure calcite after XRD) with peeled-off crumbs and shards of volcanic glass, forming hyaloclastite. As the possible depth for the emplacement of the lava flows from Hole 1179D, <2000 m seems to be realistic after consideration of the vesicle sizes. From all these facts and data, the following conclusions for the petrogenesis can be drawn:

1. Based on their mineralogy, all basalts from Hole 1179D are more or less primitive. An alkali-basaltic tendency can be excluded, but the melts from Group II were somewhat more fractionated than those of Group I and especially those of Group III. For the geochemistry this would mean, for instance, somewhat higher contents of MgO, Cr, and Ni in Group III.
2. Rapidly extruded lava flows have generated the massive basalts from Hole 1179D. For the pillow lavas, a somewhat lower effusion speed can be surmised. The lack of large amounts of phenocrysts in each group suggests a slow degree of fractionation. This may be the consequence of relatively small and short-lived magma chambers.
3. The basalts from Hole 1179D contain a low percentage of phenocrysts. Within the erupted basalt melts, plagioclase and subordinated olivine ± Cr spinel appear to be the first solidus phases, followed by clinopyroxene, and lastly, interstitial glass. Magnetite has crystallized almost exclusively within the glass. Altogether, solidification and crystallization took place rapidly.
4. The quenched glass at the chilled margins has undergone, for the most part, palagonitization and devitrification. The glassy mesostasis is more or less crystallized to microcrystalline clinopyroxene, plagioclase, and magnetite.
5. All in all, alteration does not play a considerable role. The basalts from Hole 1179D are surprisingly fresh. Alteration scheme is discussed in more detail in the following section.
6. Site 1179 is situated near Shatsky Rise, off its northwestern flank on magnetic lineation M8, indicating an Early Cretaceous age for the basement. Shatsky Rise, together with such huge oceanic plateaus as Ontong Java, Manihiki, Mid-Pacific Mountains, and Hess Rise, are assumed to have formed over a mantle plume (Neal et al., 1997; for the general question of the global plume activity also see Wilson, 1992). These extensive submarine volcanic accumulations can be compared with continental flood basalts and should be designated as oceanic flood basalts (OFBs). From geochemistry and isotopic characteristics, it differs clearly from MORB and shows a certain affinity to oceanic island basalt.

Janney and Castillo (1996, 1997) have postulated that a mantle plume formed Shatsky Rise. Unfortunately, none of the numerous holes drilled on Shatsky Rise has recovered basalt. If the basalt recovered from Site 1179 is influenced by Shatsky Rise, or is even a part of this magmatic province, then indications of the postulated mantle plume would be substantiated and it may be geochemically distinct from MORB.

The petrographic examination of the Site 1179 thin sections does not indicate an alkali tendency for the basalts. A comparison of our results with those from the few nearby holes in the vicinity of the Shatsky Rise that have reached basement may follow. At Site 581 north of Shatsky Rise, a very similar profile was drilled: 343 m of Miocene-Pleistocene sediment followed by 9.5 m of chert over basalts of probable Early Cretaceous age. These basalts were unfortunately only very briefly described by Fountain et al. (1985) as MORB in two subgroups. One is somewhat more fractionated; the other is somewhat less. Mineralogy and degree of alteration are comparable to basalts from Hole 1179D.

To the west of Shatsky Rise are DSDP Sites 303 and 304, and to the southwest, DSDP Site 307. The basement is assumed to be of Early Cretaceous age. All the basalts are classified as MORB, but they are extremely altered (Marshall, 1975).

Far from Shatsky Rise, another extensive OFB province of Cretaceous age lies in the Nauru Basin in the western Pacific. About 600 m of basalt was drilled during DSDP Legs 61 and 89 at Site 462. The mineralogy and alteration style of these rocks (Floyd, 1986; Floyd and Rowbotham, 1986) resemble Hole 1179D basalts. Their geochemical features are similar to mildly depleted (transitional) MORB, comparable with the Reykjanes Ridge volcanites (Castillo et al., 1986; Floyd, 1986; Saunders, 1986). These findings suggest that two geochemically different types of OFB provinces may exist. The Shatsky Rise volcanic complex could belong to the Nauru (MORB) type.

With all requisite prudence for such a comparison, we can state that Hole 1179D basalts are tholeiitic (= MORB) by petrography.

Twenty-one samples of basaltic rocks from Hole 1179D were analyzed in a shore-based laboratory by XRF for major and trace elements (Table T10). The analytical technique, conditions, and accuracy are described in "Igneous Petrology" in the "Explanatory Notes" chapter. The samples analyzed represent the least-altered basalt. As previously described, basalt from Hole 1179D was classified into three petrographic groups: Group I is composed of olivine-poor basalt, Group II of olivine-free basalt, and Group III of olivine-rich basalt. The CIPW norm compositions show fairly good agreement with this classification (Table T11). Samples belonging to Group I have a small amount of olivine in the norm except for one sample (191-1179D-11R-1, 86-88 cm). The amount of normative olivine is also very small or zero for samples belonging to Group II except for one sample (191-1179D-14R-2, 63-65 cm). Additionally, samples from Group II, Units 18 to 23, show the presence of normative quartz. All samples in Group III carry olivine in the norm up to 7.7%. None of them provide normative nepheline, which is thus an alkaline-basaltic tendency that can be excluded. Samples in Groups I and III show distinctive composition ranges in the TiO₂ vs. FeO*/MgO field (Fig. F54). Those from Group III show higher Ti content relative to samples from Group I. On the other hand, samples from Group II show a wider compositional range. As for trace element composition, Group III basalt is distinguished by its high Cr, Ni, and Sr content. This feature shows good agreement with its specific mineral composition, such as an abundance of olivine and the presence of Cr spinel (Fig. F49D). The abundance of rare earth elements (REEs) is nearly homogeneous throughout the samples, showing moderate depletion in the light REEs. In contrast, Masuda and Nagasawa (1975) reported results from an altered basalt dredged from Shatsky Rise that shows strong enrichment in the light REEs. This difference suggests Site 1179 basalts are distinct from those of Shatsky Rise.

XRF analyses show that all samples of basalt from Hole 1179D are geochemically characterized as tholeiitic basalt with MORB affinity. About half of them show elemental composition close to primitive normal-MORB reported from the East Pacific Rise (Humphris et al., 1980; Srivastava et al., 1980; Thompson et al., 1989), and the rest show a wider range in composition. However, tholeiitic basalts of similar composition to MORB are reported from Nauru Basin and/or Ontong Java Plateau, where so-called oceanic plateau basalt has been widely erupted (Castillo et al., 1986; Floyd, 1986; Saunders, 1986; Mahoney et al., 1993). Actually, some of samples from Hole 1179D have similar composition to those from Nauru Basin and/or Ontong Java Plateau (see Fig. F54). A study on why there is wider compositional variety in only one hole should be conducted in the context of the influence of the igneous activity that formed nearby Shatsky Rise. Such a study would require a more detailed analysis of isotope compositions.

All basaltic basement cores from Hole 1179D are surprisingly fresh by both visual core description and microscopic examination. Only ~20% of the pillow and lava flow material shows more than slight alteration. The maximum alteration is observed in very limited parts of Units 33 through 48 (~455-475 mbsf) and may amount to not more than 25%-30%. In Units 1 through 32, the degree of alteration usually amounts to <5%-10%. The minerals originating during the alteration process are restricted to very low temperature formation, such as calcite, celadonite, smectite/saponite, and zeolite(s). This secondary mineral association is more or less constant over the whole profile. Phases such as prehnite, pumpellyite, epidote, K feldspar (adularia), and quartz could not be detected. Even the presence of chlorite was not confirmed reliably.

Four main types of alteration features can be distinguished:

1. Hydration of volcanic glass from small pillow and flow rims to palagonite. This alteration is by no means complete, because considerable parts of the original glass remain nearly unchanged. Palagonitization converts the primary black and nearly opaque glass into more or less transparent, cryptocrystalline or amorphous products of brown to yellow color. The optical properties of palagonitized glass range from isotropic to weakly or moderately birefringent. Later crystallization starts with fine crystal skeletons at the borders and beside fissures. Palagonite shards are predominantly conserved within breccias and interpillow materials (Fig. F55A). The groundmass glass is also affected by palagonitization.
2. Fissures are predominantly filled with calcite. Typically the veins are 1-2 mm wide but may be as wide as 8-10 mm. The veins may include green to light green minerals, commonly at the borders, seldom over the whole width, that are part celadonite and part smectite. Figure F55B shows a sector of a calcite-celadonite vein in subophitic basalt partly replaced by smectite before the fissure was filled. In rare cases, vein calcite is associated with later zeolite.
3. Vesicles (usual diameter = <1 mm; maximum = 3 mm) are filled with one or more of four minerals: calcite, celadonite, smectite/saponite, and zeolite(s). Most calcite-filled vesicles are monomineralic. When subordinate minerals are present, they have a small initial layer of celadonite or smectite. In one thin section, some larger vesicles filled with calcite show some colorless clear grains of zeolite in the center (Fig. F55C). Otherwise, zeolites are clearly a subordinate mineral against the dominant phyllosilicates.

   Pale yellow and yellow-brown to ochre but also pale to yellowish green extremely fine grained smectite/saponite can fill the whole vesicle or form the first outer layer, followed by celadonite in the center, or vice versa (Fig. F55D). The proportions of these two minerals vary widely within one concentrically filled vesicle.

   Smectite and saponite cannot be distinguished microscopically. X-ray diffraction shows that saponite could be a possible phase in the alteration system under discussion. Celadonite was also proved in some samples. In most cases, the latter can be distinguished from smectite/saponite under the microscope in spite of the slight differences in their optical properties. Celadonite is bright to dark green, also bluish green, and in small vesicles is predominantly spherulitic or otherwise extremely fine grained (cryptocrystalline).

   Rarely, in vesicles filled dominantly with celadonite, collomorphic precipitations of goethite/hematite are observed (Fig. F56A) together with smectite/saponite and euhedral pyrite crystals.

4. Replacement of glassy groundmass of the basalt to palagonite is ubiquitous but with considerable local variation. To a lesser degree, preexisting minerals may also be affected. Whereas calcite forms amorphous patches up to 1-2 mm, phyllosilicates replace in most cases only very small spots of glassy mesostasis between plagioclase and clinopyroxene (Fig. F56B). As an exception, plagioclase can be replaced in part by smectite or celadonite. Olivine phenocrysts are always transformed into iddingsite, consisting of goethite/hematite and some other phases (Fig. F56C). Clinopyroxene and the very small magnetite grains and crystal skeletons are not affected by replacement.

   The replacement of groundmass occurs in part irregularly, but next to veins there is commonly developed a more intensive accumulation of small blastic grains or patches of phyllosilicates (Fig. F56B). Smectite and celadonite exclude one another not only in one single grain but also in a limited "district of grains," and altogether smectite seems to be somewhat more common as a neogenic mineral than celadonite (Fig. F56D). Zeolite, as part of the alteration, could not be observed within the groundmass. There are no differences between pillow and flow basalts with respect to the occurrence of alteration minerals.

   Generally, the top and bottom walls of the lava flows are oxidized to a certain extent as a very early event in the alteration process. Somewhat later, during alteration by seawater along fissures but in some places without visible feeder channels, a slight infiltration with Fe and resulting impregnation of the whole groundmass by FeOOH took place. In some cases, the secondary mineral association is also affected and smectite and celadonite are intensely red colored.

The restricted mineral paragenesis of the alteration with calcite, celadonite, smectite/saponite, and zeolite(s) indicates a very low temperature regime. According to Staudigel and Hart (1983), Floyd (1986), Floyd and Rowbotham (1986), and Alt (1993), this is reflected by moderate to high levels of water, low oxidation potential, and relatively low CO₂ content. The current visible hydration of glass to palagonite is restricted to the borders of different massive and/or pillow lava flows and onto interpillow material. The necessary high water flow for palagonitization was enabled by the clastic structure of these brecciated zones and their high permeability.

Smectite/saponite formation needs only low seawater/rock ratios, but the stability conditions of celadonite reflect higher seawater/rock ratios to provide a sufficient supply of potassium. A system of fissures, resulting from cooling of massive flows and pillows, has promoted the penetration of seawater. The compositional change of zoned vesicle fillings with smectite rims and celadonite cores, or vice versa, indicates a clear variability in the chemistry of penetrating seawater, especially with respect to the K and Fe contents and the redox potential.

The nearly total absence of sulfides, especially pyrite, in the secondary paragenesis is remarkable. This is a clear expression of S^2- deficiency within the percolating seawater and/or a redox potential that was too high for reducing SO₄^2- but not high enough for constant and more than partial oxidation of Fe²⁺.

For an estimation of the alteration temperatures, we can only compare the existing secondary mineral paragenesis with data from former DSDP and ODP Legs (Muehlenbachs, 1980; Honnorez et al., 1983; Alt et al., 1992; Alt, 1993; Janney and Baker, 1995). From these data, especially ¹⁸O results, we can deduce that for the smectite/saponite formation, a temperature span between 0°-15°C is realistic. For celadonite, a somewhat higher temperature can be assumed and comparable data for calcite (also based on ¹⁸O) imply a temperature range of 10°-45°C.

Therefore, using these data, we can conclude reliably that the subseafloor metamorphism of the basalts from Hole 1179D took place at temperatures <50°C, with a certain probability for a temperature range of 10°-30°C. The absence of chlorite is of concern with such a low-temperature regime and corresponds to Miyashiro's "low to middle zeolite facies" (Miyashiro, 1981). This very low temperature alteration of the basalts can be described also as a kind of "submarine weathering."