SiO2
is <1.0. Where both thermometers were applied to clearly silica-saturated
samples, they gave virtually identical temperatures. Thus, all reported
temperatures are based on the edenite-richterite geothermometer.Estimating the temperatures of formation of the veins in Hole 735B is very difficult because few, if any, veins contain equilibrium mineral assemblages. Most of the amphibole veins are monomineralic, and all of the felsic and plagioclase-rich veins contain mixtures of minerals ranging from relatively high-temperature plagioclase + quartz assemblages to low-temperature mixtures of clay minerals, zeolites, and carbonate. An even greater problem is that nearly all of the relatively high-temperature minerals are moderately to strongly zoned, making it extremely difficult to identify equilibrium pairs. Plagioclase in the veins varies from a maximum of about An43 to a minimum of An0.5, and the full range of variation is commonly present within individual crystals. Although not as strongly zoned, amphiboles exhibit color variations from brown or dark brown in the core to green or light green on the rim, corresponding to varying contents of alumina and iron.
The monomineralic amphibole veins are restricted to ductilely deformed metagabbros that were formed under the equivalent to granulite and amphibolite facies conditions. The amphibole veins both crosscut and are cut by the gneissic and mylonitic foliation in these rocks, suggesting that ductile and brittle deformation occurred penecontemporaneously (Dick et al., 1991; Stakes et al., 1991).
In those rocks with neoblasts of both clinopyroxene and orthopyroxene, equilibrium temperature calculations suggest that the earliest deformation and metamorphism took place under anhydrous granulite facies conditions between 849° and 908°C (Stakes et al., 1991). The amphibole gneisses mark the start of hydrous metamorphism and formed at temperatures between ~590° and 720°C (Stakes et al., 1991), roughly the same temperature range calculated for sparse amphibole + plagioclase veins in this interval (see below). We infer, based on their textural relations, that the monomineralic amphibole veins formed essentially at the brittle-ductile transition, which is believed to lie everywhere in the ocean basins between 700° and 800°C (Phipps Morgan and Chen, 1993). The temperature estimates for the amphibole veins in Hole 735B are in good agreement with those of Manning et al. (1999) for amphibole + plagioclase veins in gabbros from Hess Deep (687°-745°C).
For amphibole + plagioclase veins we applied the geothermometers of Holland and Blundy (1994) with some success, despite the problems of identifying equilibrium assemblages. Holland and Blundy (1994) present two amphibole-plagioclase geothermometers: an edenite-tremolite thermometer that requires silica saturation in the system and an edenite-richterite thermometer that can be applied to either silica-saturated or unsaturated samples.
Most of the amphibole +
plagioclase veins in the section also contain quartz, suggesting a
silica-saturated environment. However, petrographic examination indicates that
the quartz is almost invariably present as a late-stage mineral filling
interstices between zoned and pitted plagioclase crystals and thus is not part
of an equilibrium assemblage. Only in a few veins with well-developed myrmekitic
textures is the quartz clearly intergrown with plagioclase. To test this
interpretation, we applied both geothermometers to the same data and compared
the results. In nearly every case the edenite-tremolite thermometer, which
requires silica saturation, produced temperatures 50°-60°C higher than the
edenite-richterite thermometer even though quartz is present in the veins.
Holland and Blundy (1994) predict such a discrepancy for samples lacking silica
saturation. In this situation, the lower temperature is taken as the equilibrium
temperature and the higher temperature is offset because SiO2
is <1.0. Where both thermometers were applied to clearly silica-saturated
samples, they gave virtually identical temperatures. Thus, all reported
temperatures are based on the edenite-richterite geothermometer.
In order to deal with compositional variations in the vein plagioclase and amphibole, we applied the edenite-richterite thermometer in two ways. First, we attempted to match the Fe-rich amphibole rims with sodic plagioclase rims and the Fe-poor amphiboles with calcic plagioclase cores. Second, we calculated the range of temperature based on all amphibole and plagioclase vein compositions and took the mean value. The resulting temperatures, which are estimated to be accurate to ±40°C, range from a high of 834°C to a low of 525°C (Fig. F12). Pressure effects are assumed to be negligible since this crustal section was probably never more than 1-2 km below the seafloor (Dick et al., 1992). The highest temperatures are from veins with the most calcic plagioclase, assumed to be the closest to the original composition. The lowest temperatures are based on the most sodic plagioclase rims, clearly formed by hydrothermal processes. The mean temperature for all the veins is 674°C (standard deviation = 77°C). These temperatures are slightly higher than those estimated by Vanko and Stakes (1991) for the formation of amphibole + plagioclase veins in the upper part of the section. Using the geothermometers of Spear (1980) and Plyusnina (1982), they obtained temperatures between 520° and 640°C for andesine-bearing (An31) veins very similar to those described here. They reported temperatures of <500°C for albite and oligoclase assemblages in brecciated horizons. A small trondhjemite intrusion registered temperatures between ~500° and 550°C (Vanko and Stakes, 1991).
The temperatures of formation for the diopside and diopside + plagioclase veins can be estimated in two ways. The diopside compositions plot in a tight cluster on the pyroxene quadrilateral (Fig. F6) and lie well below the 500°C temperature contour of Bird et al. (1986). Homogenization temperatures of fluid inclusions in diopside range from 310° to 420°C for three diopside + plagioclase veins in the upper part of Hole 735B (Vanko and Stakes, 1991). Interestingly, homogenization temperatures for plagioclase from the same veins are significantly lower (200°-290°C). The temperatures obtained for the diopside + plagioclase veins are significantly lower than those obtained from the amphibole + plagioclase veins, and are clearly in the hydrothermal range.
Temperatures for the other veins can only be estimated from their mineral assemblages. Greenschist minerals such as epidote, chlorite, and actinolite probably formed at temperatures of 300°-450°C. These are volumetrically insignificant and are present chiefly as replacements of higher-temperature vein minerals. Smectite and zeolite (thomsonite and natrolite) veins presumably formed at temperatures <50°C, and the carbonate veins probably formed at temperatures close to ambient seawater. When these phases are present in relatively high-temperature veins, they always fill late-stage fractures or replace the higher-temperature minerals. Prehnite is closely associated with smectite and carbonate probably formed at similar temperatures.
