1. determine the thermophysical characteristics of hydrothermal circulation in the upper oceanic crust in off-axis settings as influenced by crustal topography, sediment cover, and permeability;
2. determine the sensitivity of crustal fluid composition to the age, temperature, and degree of sediment burial of the igneous crust;
3. examine the nature and fundamental causes of physical consolidation and mineralogical and chemical alteration of the igneous crust as functions of age and degree of sediment burial; and
4. improve constraints on the fluxes of heat and elements between the permeable igneous crust and the overlying ocean.

In the following preliminary summary of results, individual interpretations are related to the above objectives where appropriate.

Leg 168 drilling sites were placed along a transect on the eastern flank of the Juan de Fuca Ridge, extending from about 20 km east of the ridge crest, where turbidite sediments begin to continuously blanket the 0.8-Ma igneous crust, to about 100 km from the crest, where the crust is 3.6 Ma in age. Coring and downhole measurements completed along this transect have provided unprecedented documentation of the efficiency with which water transports heat and solutes laterally in the upper oceanic crust and of the consequences of that circulation (Objectives 1, 2, 3, and 4). Additional information from long-term hydrologic observatories (CORKs) established at four of the sites will provide valuable quantitative constraints on the pressures that drive fluid flow, on basement-water composition, and on the details of the thermal structure in the uppermost oceanic crust as it is influenced by hydrothermal circulation (Objectives 1, 2, and 4).

A simple sedimentation history is inferred at each of the sites, with locally carbonate-enriched hemipelagic deposition beginning to accumulate some thousands of years after creation of the igneous crust, followed at the older sites by periods of accumulation of hemipelagic muds and fine turbidites, and finally by a Pleistocene sequence of variably thick silt- and sand-turbidites interbedded with hemipelagic muds. Despite the frequency of sandy layers in the upper part of the sections, this blanket of sediment that buries all primary topography of the igneous crust by up to 600 m is generally impermeable and prevents fluid flow at geochemically significant rates except where the section thins to less than about 100 m over local basement highs (Objective 2). At three such sites (Sites 1025, 1030, and 1031) there is clear evidence in the pore-fluid compositional profiles of discharge, although the seepage velocity is far too low to be thermally significant. The integrated effect of fluid seepage in the form of sediment alteration is virtually undetectable; at all sites, sediment alteration was evident only on a scale of up to a few meters above the basement contact (Objective 3).

Basalts were recovered at nearly all sites, and they represent a wide range of compositions and textures: normal tholeiitic pillows; massive, highly fractionated flows with extremely large (up to 1 cm) vesicles; volcanic breccias cemented with low-temperature hydrothermal mineralization; and a massive diabase unit erupted onto or intruded into the sediment section well off-axis. Alteration of the igneous rock varies in intensity from site to site along the transect, both as a consequence of increasing temperature and increasing degree of basement water isolation and as a result of variations in texture and lithology (Objectives 1 and 3). Fresh olivines present at Site 1023 indicate that these aphanitic samples-the youngest, coolest, and closest to a source of fresh seawater (about 3 km from the nearby region of outcropping basement)-have undergone the least amount of alteration. In contrast, practically all olivines in the massive, coarser-grained units at Site 1025 are completely replaced by a mixture of clay minerals (primarily saponite with some celadonite), minor carbonate, and rare talc and chlorite/smectite. Samples from sites throughout the transect contain mineral linings or complete mineral fillings of vesicles. At the oldest and hottest Sites 1026 and 1027, the vesicle filling involves two to four sequential layers. The observed distribution of the various vesicle fills is systematic, in detail linked to the proximity of the vesicles to veins and alteration haloes. Rocks at nearly all sites (except from the massive units) exhibit varying degrees of oxidative alteration that required significant open-seawater circulation. A subsequent stage characterized by carbonate, saponite, and sulfide alteration probably represents relatively closed hydrothermal circulation following sediment burial. No alteration, including that represented by vein-filling minerals, was observed that requires temperatures much higher than those prevailing today (Objective 3).

Active and highly efficient local advective heat and chemical transport is evident in the degree of homogenization of crustal fluid temperatures and compositions (Objectives 1 and 3). Sites 1026 and 1027, drilled ~2.2 km apart into a sediment-buried basement ridge and adjacent trough, respectively, encountered basement at virtually identical temperatures (~64°C), despite the large contrast in the thickness of sediment at the ridge and valley sites, 250 and 600 m, respectively (Objective 2). Basement-fluid compositions, inferred from pore waters squeezed from near-basal sediments and sampled directly from the discharging Site 1026, are also to a first order identical at the two sites, and identical to water sampled at a seafloor vent roughly 7 km away where water discharges through the seafloor at an isolated, small basement outcrop.

Some form of regional-scale flow is also inferred on the basis of thermal and geochemical observations made across the full length of the transect (Objectives 1, 2, and 4). A comparison of seafloor heat-flow measurements with the total expected from the cooling lithosphere beneath shows a clear deficit on this sediment-sealed eastern ridge flank out to a distance of at least 20-30 km from the nearest basement outcrop; it has been inferred that this deficit reflects the heat lost advectively from the upper igneous crust through regions of outcropping basement through lateral fluid flow. Observations at the first three sites of the Leg 168 transect substantiate this inference. Clear gradients in basement temperature and basement-water chemistry were documented, with temperatures at the sediment/basement contact increasing systematically with distance from the region of outcropping basement, from approximately 15°C at Hole 1023A, to 22°C at Site 1024, and to 38°C at Site 1025 (Objectives 1, 3, and 4). This trend is opposite to that expected, given the increasing age and decreasing sediment thickness across this part of the drilling transect; it is best explained by rapid fluid flow and heat transport in the upper basement rocks beneath the sediment section, with cool seawater recharge supplied in the area of extensive igneous outcrop near the ridge crest. Basement-water compositions revealed systematic changes with increasing temperature and distance from the outcrop in many elemental concentrations, including magnesium and calcium, which are particularly reactive in basement. Elemental concentrations at Site 1023, situated 3.5 km away from the line of abrupt basement outcrop to the west, were changed only slightly from seawater, whereas water at Site 1025, nearly 16 km away, showed significant reaction with basement (Objective 2). Basement-water chlorinity values also showed clear signs of lateral flow in basement, with the influence of post-Pleistocene seawater, only a few thousand years old, seen beneath the sediment cover also as far away from the basement outcrop as Site 1025.

Perhaps the most remarkable conclusion from the results of the leg concerns the scale over which significant lateral heat and chemical transport must take place (Objectives 1 and 4). Over most of the length of the transect the local average heat flow through the sediment section does not exceed about 80% of the amount expected from the lithosphere beneath. Basement temperatures simply continue to increase systematically with distance from the region of basement outcrop near the ridge to the maximum temperature at Sites 1026 and 1027 noted previously. This disagreement with lithospheric cooling theory may point to a fundamental lack of understanding of lithospheric heat loss, but it is more likely that a significant quantity of heat is lost advectively from beneath the sediment seal over distances of many tens of kilometers. This conclusion is supported by the level of sulfate observed in basement waters, which falls systematically with distance from the region of basement outcrop but to a remarkably elevated minimum of 18 mmol/kg at the most distant Sites 1026 and 1027 (Objectives 3 and 4). At all sites, the sediment section is seen to serve as a very efficient sink of sulfate; the only source for replenishment is seawater supplied via purely-basement pathways. Very large distances of lateral transport (up to 80 km) are implied.

The primary scientific objectives of this phase of ridge-flank drilling focused on lateral thermal and chemical gradients associated with hydrothermal circulation, and they were achieved with highly limited basement penetration. Virtually all observations can probably be accounted for with a hydrothermal system that is dominated by fluid flow in only the upper tens to few hundreds of meters of the igneous oceanic crust. However, some hints of deeper levels of fluid flow are suggested at Sites 1030/31, which penetrated a basement ridge where anomalous basement fluids appear to be present (Objectives 1, 2, and 4). Chlorinity and the concentrations of Mg and Ca in basement water are anomalous with respect to the trends defined by other sites of the transect and suggest reactions at a temperature much higher than that at the top of the crust; basement fluids here may be "contaminated" by water flow locally up a fault zone from a depth where temperatures and fluid residence times are greater.

Although ODP Leg 168 was highly successful in setting strong constraints on the rates of fluid flow through the uppermost igneous crust and in beginning to identify the surprisingly large lateral scale over which hydrothermal circulation can operate, the need for additional work in this area is clear. Determining the routes of flow, including the hydrologic role of normal faults and the depth of penetration of significant circulation, will require deeper sampling and downhole experiments possible only through a second leg of drilling. Effort must also be given to better identifying the locations and details of seawater recharge and crustal water discharge, where very large chemical and heat fluxes must occur.