Leg 209 was devoted to drilling mantle peridotites and associated gabbroic rocks along the Mid-Atlantic Ridge from 14° to 16°N. This area was identified at the 1996 Workshop on Oceanic Lithosphere and Scientific Drilling into the 21st Century (OL Workshop) as the ideal region for drilling of a strike line of short holes to sample the upper mantle in a magma-starved portion of a slow-spreading ridge (spreading rate = ~25 km/m.y.). In this area, igneous crust is locally absent and the structure and composition of the mantle can be determined at sites more than ~100 km apart along strike.

A central paradigm of Ridge Interdisciplinary Global Experiments (RIDGE) studies is the hypothesis that mantle flow, or melt extraction, or both, are focused in three dimensions toward the centers of magmatic ridge segments, at least at slow-spreading ridges such as the Mid-Atlantic Ridge. This hypothesis has essentially reached the status of accepted theory, but it has never been subject to a direct test. A strike line of oriented mantle peridotite samples extending for a significant distance within magmatic segments offers the possibility of directly testing this hypothesis. Continued dredging and submersible studies cannot provide the spatial information required to make such a test.

The primary aim of drilling was to characterize the spatial variation of mantle deformation patterns, residual peridotite composition, melt migration features, plutonic rocks, and hydrothermal alteration along axis. Hypotheses for focused solid or liquid upwelling beneath ridge segments make specific predictions regarding the spatial variation of mantle lineation or the distribution of melt migration features. These predictions were directly tested by drilling. We discovered that penetrative mantle deformation fabrics are weak at every site where mantle peridotite was sampled from 14°43'N to 15°39'N. Instead, at all of these sites, deformation was localized along high-temperature shear zones and later brittle faults. Intact blocks of peridotite with high-temperature, protogranular fabrics were preserved between these zones of localized deformation and underwent substantial tectonic rotation, perhaps as much as 90° around horizontal, ride-parallel rotation axes in some places.

At most sites, drilling recovered substantial proportions of gabbroic rocks intrusive into mantle peridotite. Some of these rocks have mineral assemblages that are probably indicative of crystallization at depths of 12 to 20 km beneath the Mid-Atlantic Ridge. Localized deformation at several of these sites occurred preferentially within contact zones between peridotite and these gabbroic intrusions. Abundant gabbroic intrusions were found close to the 15°20' Fracture Zone, at Site 1271, and far from the fracture zone at Sites 1270, 1268, and 1275. Conversely, some holes intersected very little gabbroic material; these were at Site 1272, very close to the fracture zone, and Site 1274, far from the fracture zone. Thus, there is little evidence from the results of this leg for focusing of melt distribution away from the fracture zone and toward the centers of volcanically active ridge segments.

Three new hypotheses may account for our observations:

1. Shallow mantle peridotites beneath the Mid-Atlantic Ridge in this region do not undergo penetrative deformation during "corner flow" associated with plate spreading. Instead, they rise passively until they reach the base of the thermal boundary layer at depths of 15 to 20 km below this slow-spreading ridge. There, they cool and become incorporated into the lithosphere. Subsequently, corner flow and ridge extension are accommodated along localized ductile shear zones which gradually evolve into brittle faults at shallower depths and lower temperatures. As some faults rotated to shallow dips and could no longer accommodate extension, new ones formed. Crosscutting generations of faults rotated nearly undeformed blocks of peridotite and associated gabbroic intrusions, with total rotations probably >60° or even 90° in some cases.
2. Textures in the relatively undeformed peridotite blocks suggest that many residual peridotites interacted with melts migrating by diffuse porous flow along grain boundaries at the base of the thermal boundary layer. In most peridotites, our qualitative observations of textures suggest that igneous spinel and pyroxene crystallized within a matrix of residual mantle olivine and orthopyroxene. More extensive crystallization of intergranular melt at slightly lower temperatures formed impregnated peridotites and hybrid troctolites, particularly abundant at Sites 1271 and 1275. Based on our limited sampling, it seems that focusing of melt transport into dunite conduits with sharp contacts against residual mantle peridotites, common in some ophiolites and perhaps beneath fast-spreading ridges, was not a very important process in the region we investigated. Instead, melts probably were in equilibrium with mantle peridotite up to the base of the thermal boundary layer, after which they probably ascended in brittle cracks.
3. We wonder if the "amagmatic" region between 14°40' and 15°40'N along the Mid-Atlantic Ridge is truly "magma starved" as has often been proposed or whether, instead, the relative lack of lava and gabbroic crust is offset by a relatively high proportion of gabbroic intrusions into peridotite, distributed over 15 to 20 km depth. In this view, many melts may crystallize 100% below the seafloor, with no magma rising to form lava flows. In keeping with this hypothesis, gabbroic rocks, particularly those at Site 1275, the top of Mt. Mike at 15°44'N, ~25 km west of the ridge axis, are generally very evolved. They cannot represent the refractory, primitive cumulates required to complement compositional variation in mid-ocean-ridge basalts, and instead must represent the crystal products of very evolved melts that rarely erupt. This may be heresy, but we wonder if, for example, gravity data for the Mid-Atlantic Ridge might be reconciled with such a theory, given that the widely distributed gabbroic rocks in such a lithospheric structure would generally be farther from the seafloor and therefore would have a smaller gravity signal than a thick gabbroic layer concentrated near the surface.

Very different hydrothermal alteration styles were observed at different Sites. In Hole 1268A, talc was particularly abundant in metaperidotites, accompanied by a dramatic metasomatic decrease in the (Mg + Fe)/Si ratio. Elsewhere, brucite was a prominent part of the alteration assemblage in peridotites, and rocks retain high (Mg + Fe)/Si. Gabbroic intrusions appear to have an important local control on serpentinization reactions in peridotite. Carbonate alteration of peridotites in some locales seems to be correlated with a metasomatic influx of calcium and may also substantially affect the trace element budget of serpentinites in some cases.