The mineralogy of the igneous rocks recovered during Leg 185 from Hole 801C was estimated for several pieces per section by hand lens and binocular microscope identification. The results are noted in 801_MIN.XLS (see the "Supplementary Materials" contents list) and are summarized in the visual core description logs (see the "Core Descriptions" contents list). In addition, a total of 83 thin sections from Hole 801C were investigated during Leg 185, including seven from the Leg 129 basement cores. The thin sections from Leg 185 were chosen within representative lithologies, generally coinciding with X-ray fluorescence (XRF) samples, as well as within specific features, such as contacts, veins, phenocryst-rich intervals, and highly altered intervals. The approximate proportions and sizes of minerals (groundmass and phenocrysts) in thin section were estimated for both primary and secondary minerals. The thin-section descriptions are included in "Site 801 Thin Sections". The primary mineralogy is summarized in Table T6.
Generally, the igneous rocks recovered in Hole 801C during Leg 185 are aphyric to slightly phyric plagioclase- and/or olivine-bearing basalts. Euhedral plagioclase laths of an average size of 0.1-1.4 mm are present in each thin section as a major part (35%-60%) of the groundmass (Fig. F20). Subhedral pyroxenes ranging from 0.02 to 0.6 mm form the second major constituent (12%-45%) (also in Fig. F20). Small (<0.01-0.1 mm) dispersed oxides (mainly magnetite) are always present, typically with an euhedral or skeletal morphology (Fig. F21A, F21B). Olivine is rare (<10%) in the groundmass and when present is always altered (Fig. F22A, F22B). It is found in euhedral to subhedral forms ranging from 0.04 to 0.4 mm in size. The last major constituent of the groundmass is interstitial cryptocrystalline, brownish devitrified glass (Fig. F23). Basalts that consist of >40% interstitial material (or glass) are typically found in or near chilled pillow or flow margins and are, therefore, denoted as chilled basalts. Many of these chilled basalts contain fresh glass, but a complete spectrum from fresh (Fig. F24) to devitrified (Fig. F23) was observed; some basalts contain up to 30% devitrified glass (Fig. F25). Vesicles are ubiquitous in all thin sections but generally make up 1%, except in Sample 185-801C-37R-3 (Piece 1A, 37-40 cm) (thin section [TS] 62) where vesicle content approaches 10%.
The crystallinity of the basalts investigated ranges from holohyaline in chilled margins (e.g., TS 24; Sample 185-801C-16R-5, 107-110 cm) (Fig. F26) to holocrystalline in pillow and flow interiors (e.g., TS 63; Sample 185-801C-37R-5, 37-40 cm) (Fig. F27). Correspondingly, the grain size ranges up to medium grained. Most basalts show seriate textures of heterogranular plagioclase. In glassy pillow rims, plagioclase (Fig. F26), olivine, and sometimes pyroxene (Fig. F28A, F28B) are present as phenocrysts. Quenched textures also occur (Figs. F29, F30) but were not systematically distinguished. The proportion of the different groundmass minerals does not vary with depth (Fig. F31), although the size of the crystals in the groundmass reaches a maximum in Core 185-801C-37R at 815-820 mbsf (Fig. F31), which corresponds to the medium-grained interior of a massive flow (see "Lithologic Units").
Phenocrysts are rare within the basalts drilled in Hole 801C during Leg 185. Generally, plagioclase is the most common and largest phenocryst observed (Figs. F32, F33, F34; Table T6). Olivine and pyroxene phenocrysts are even rarer and are present only in the upper and lower parts of the Leg 185 section of Hole 801C (Fig. F32; Table T6). In crystalline samples, olivine phenocrysts are essentially always altered (Fig. F35) (see "Basement Alteration"), although an unaltered euhedral olivine of ~0.1 mm was found in one sample of fresh glass (TS 67; Sample 185-801C-42R-2 [Piece 1B, 126-128 cm]). This olivine contains a large fluid inclusion (Fig. F36).
Phenocryst abundances observed in thin section are somewhat different from those observed in hand specimen. For example, olivine phenocrysts were more commonly described in hand specimen than in thin section. One reason for this is that thin sections represent only a small part of the rock and may not necessarily represent the sample as a whole. On the other hand, the abundance in hand specimen may be overestimated because alteration leads to lustrous, black crystals, which are very obvious in hand specimen.
Compared to the Leg 129 basalts in the upper parts of Hole 801C (Shipboard Scientific Party, 1990), which contain as much as 10% plagioclase phenocrysts, the Leg 185 basalts are phenocryst poor.
Characterizing the basaltic geochemistry at Site 801 was a key step toward quantifying and refining the igneous lithologic units defined by hand-specimen and thin-section observations. Geochemical data were used to define units that were not evident in hand specimen descriptions. Shipboard geochemical analyses also provided high-precision data, which aided in determining magmatic crystallization sequences and parent magma characteristics. As a means of identifying the geochemical features of the basement rocks, selected samples from Hole 801C were analyzed for major and trace elements by shipboard XRF (Table T7). In most cases, samples were chosen to identify downhole trends in the least-altered basalt, although a few samples were taken to characterize specific styles of alteration (e.g., pale green alteration; Sample 185-801C-15R-7, 72-74 cm) or different varieties of interpillow material (e.g., recrystallized interpillow sediment; Sample 185-801C-16R-3, 60-62 cm).
These new data expand the existing geochemical data set for rocks recovered from Holes 801B and 801C during Leg 129 (Castillo et al., 1992), which includes fresh and altered basalt samples from 466 to 591 mbsf (0-125 m into basement). Floyd and Castillo (1992) divided the basalts from Leg 129 Holes 801B and 801C into two distinct geochemical categories, an upper alkali basalt sequence (~60 m thick) and a lower tholeiitic sequence (through Core 129-801C-12R), separated by ~20 m of ocherous hydrothermal deposit. The new Leg 185 data consist of 51 XRF major element analyses and 50 trace element analyses, 3 of which are of interpillow material and the remaining 48 of fresh and altered basalts, spanning a depth range of 605-920 mbsf. The data acquired from Core 185-801C-14R correlate well with the existing Site 801 data set, continuing the chemical trends observed in the tholeiites of Core 129-801C-12R.
Alteration plays a substantial role in the observed geochemistry of the lavas drilled at Site 801. Commonly, the degree of alteration is most readily gauged by a few sensitive geochemical parameters such as the loss on ignition (LOI), which is roughly equivalent to the total H2O and CO2 of the rock, or the K2O content of the basalt. For the purpose of determining the primary igneous character of the basalts, samples with >2 wt% LOI or >1 wt% K2O were therefore excluded from consideration. No observable relationship between LOI and elements such as Mg, Zr, or Cr exists, suggesting that alteration has not significantly affected these elements in most rocks and that these abundances are primary in the basalts.
On a broad scale, a simple geochemical trend dominates most of the sequence drilled during Leg 185. MgO decreases upward (Fig. F37A) through Core 185-801C-16R, illustrating that the lavas at this site are more primitive at the base and become more evolved (less MgO rich) with time. Trace elements also show a consistent trend. The incompatible element Zr is enriched in magmatic liquids with increasing degrees of crystal fractionation. Consistent with the trend in MgO, Zr shows a systematic increase uphole (Fig. F37B), indicating more primitive magma at depth in the hole. In contrast, the lower hydrothermal deposit in Core 185-801C-16R marks an abrupt change in the course of magmatic evolution. The upper tholeiites recovered in Cores 185-801C-14R through 16R, and in Cores 129-801C-5R through 12R, show a reversal of the geochemical trends from those deeper in the section. MgO values in the shallower section steadily increase, whereas Zr values decrease toward the top of the tholeiitic sequence.
Within these larger geochemical trends, however, excursions from the general pattern take place over smaller depth ranges and define more extreme chemical compositions. Small-scale breaks and trends within the data identify more subtle and localized changes in basalt chemistry, which can be used to segregate the entire sequence of basalt from Site 801 into 18 distinct geochemical units. These units are most distinctive on a plot of Zr vs. depth and are reinforced by consistent changes in MgO and Cr (Fig. F37B), and by changes in the character of the downhole logs.
The subset of least-altered basalts recovered during Leg 185 forms a trend consistent with crystallization of olivine + plagioclase + clinopyroxene. The trend in MgO vs. Al2O3 (Fig. F38), for example, follows a calculated liquid line of descent (Weaver and Langmuir, 1990), which predicts the changing composition of a magma as crystals are fractionated from a parent liquid. Sample 185-801C-46R-2, 75-77 cm, was chosen as the parent liquid, and the resultant calculation predicted co-crystallization of olivine and plagioclase over the temperature range of 1195°-1150°C, with a third phase (clinopyroxene) saturating at 1150°C. Samples deviating significantly from the trend of the line may have accumulated as much as 1% olivine or 5% plagioclase phenocrysts. One percent of olivine accumulation is consistent with proportions observed deeper in the section. Basalt with as much as 5% plagioclase phenocrysts, however, was not encountered in any of the basement drilled during Leg 185. The high-Al samples are interspersed throughout the sequence and, thus, cannot be explained by either a separate liquid line of descent or by the simple addition of plagioclase phenocrysts to the basalt. When examined in thin section, however, plagioclase phenocrysts in some of these samples show slight evidence of resorption, which could explain an increase of Al2O3 in the liquid.
Two basalts (Samples 185-801C-48R-1, 42-44 cm, and 48R-3, 14-16 cm) deviate significantly from the trend predicted by the liquid line of descent. These two samples represent the basalt deepest in the drill hole for which shipboard XRF data are presently available. For these samples to conform to the predicted trend, >2% olivine must be included with the basalt. In hand specimen, however, they do not display a significant increase in phenocryst abundance. It is possible that these samples are related to a different parent magma and fall along a separate liquid line of descent, but the available data are inconclusive and the question of their origin remains open.
Relating these rocks by crystal fractionation is important for determining the magmatic history of the basalts at Site 801, but identifying mantle characteristics and understanding the origin of the parent magma are equally critical tasks. The Zr/Y of a magma is not affected by crystal fractionation, but rather the ratio reflects the nature of the mantle source (enriched vs. depleted) and the degree of partial melting. The downhole trend in Zr/Y for the tholeiites of Site 801 is remarkably consistent (Fig. F37D), varying only slightly from a value of 3, suggesting that the mantle and melting processes at this site changed very little over time.
Shipboard XRF analyses provide a continuous geochemical record of basalt recovered during Legs 129 and 185. Downhole trends in MgO and Zr indicate a primitive magma at the bottom of Hole 801C, which evolved toward more fractionated lavas in Core 185-801C-16R then generally back toward more primitive compositions in Core 129-801C-5R.
Smaller trends within the data, however, show short periods of fractionation and intervals of injection of more primitive magma. The break between geochemical Units 14 and 15 (Fig. F37) is a prime example of a significantly fractionated magma that experienced a simultaneous increase in MgO and decrease in Zr at 735 mbsf. This abrupt change in chemistry is consistent with the addition of more primitive, high MgO-low Zr material into the magma chamber.
The least-altered basalt samples fall largely along a single liquid line of descent, indicating that the entire sequence of basalt recovered during Leg 185 could have originated from the same parent magma. Data from the superfast spreading segment of the southern East Pacific Rise (EPR at 15.00°-18.72°S) (Sinton, 1991) fall along a similar trend (Fig. F38) and provide a meaningful comparison, because this segment of the EPR represents a possible modern-day analog for the ridge that produced the eruptive sequence at Site 801.
The Zr/Y value changes only slightly downhole, showing little change in either the mantle source or the degree of melting. The data from Hole 801C are comparable to, although more fractionated than, MORB from the modern Mid-Atlantic Ridge (MAR) (Fig. F39), with Zr/Y falling close to a value of 3. When compared to the modern EPR (8°-15°N), however, Hole 801C Zr/Y ratios are significantly lower, indicating that the modern EPR is either tapping a more enriched mantle or experiencing a lower degree of partial melting.
The basement drilled during Leg 185 consists mainly of aphyric basalts. Plagioclase, pyroxene, and olivine phenocrysts are rare but are present throughout Hole 801C. The groundmass phases are plagioclase, pyroxene, magnetite, interstitial devitrified glass, and rarely olivine.
Shipboard geochemical analyses reveal simple evolutionary trends within Hole 801C, corresponding to events of primitive magma injection and subsequent fractionation. An overall trend is observed toward more evolved lavas in Core 185-801C-16R and back to more primitive compositions in Core 129-801C-5R. These geochemical trends have been used to divide the core into 18 geochemical units. Basalts analyzed during Leg 185 are all related to the same parent magma and vary only slightly from a dominant mantle signature, which is similar to, although more depleted than, modern EPR MORB.