PRELIMINARY SCIENTIFIC ASSESSMENT

The overall success of Expedition 304 is due to a combination factors.

A disappointing aspect of Expedition 304 is that we did not succeed in drilling more than ~20 mbsf in the hanging wall. The operational strategies employed, which were restricted because of loss of gear while setting the reentry system at the footwall site, were not adequate. Conceptually, it seems that the ring-bit spline drive hammer-in casing assembly would have provided a better start if it had been available. However, there is no guarantee that unstable conditions that might be dealt with by casing the upper 20–30 mbsf would not be just as problematic at any depth below the base of casing, because the hole would still be in fractured basalt. The minimal amount of rock recovered during a series of failed attempts to start a reentry hole in the hanging wall provides an intriguing glimpse of a compositional aspect of those rocks (see "Site U1310" and "Site U1311"), but not a conclusive test of the nature of the eastern volcanic block. Without significant recovery from the hanging wall, we cannot address petrogenetic relationships with the footwall rocks or the magnitude of hanging wall rotation relative to the footwall. In retrospect, this aspect of hanging wall studies might have been better served by simply attempting a series of single-bit holes, some of which might have been successful at achieving a few tens of meters penetration and partial recovery. The goal of penetrating an unexposed portion of the detachment fault beneath the hanging wall was always a high-risk endeavor for Expedition 304. Once we recognized how difficult setting casing for the reentry system was in this environment, we adjusted the site location (from Site U1310 to Site U1311), in hopes that a small klippe covering the detachment could provide access to this second hanging wall objective with much less penetration. We (the engineers, drillers, staff scientist, and co-chiefs) worked hard to come up with an approach that would provide a workable reentry site in unsedimented, young basalt that characterized the sites; after 11 days it was clear that we could advance the objectives of the overall project more by returning to the footwall site.

Return to the footwall site allowed penetration to continue to over three times the original planned depth for Expedition 304. This bodes well for Expedition 305 objectives. Shipboard results are preliminary, but it is already clear that the structural and paleomagnetic data from Holes U1309B and U1309D will provide key constraints on the evolution of the footwall during formation of the OCC. At this (early!) stage, it is not clear how the classic "rolling hinge" model can explain the observations we have made on the core. Petrologic and geochemical data are equally intriguing, although for different reasons.

The relatively high proportion of gabbroic rocks recovered at Site U1309 on the central dome contrasts with recovery by dredge and submersible dives on the south wall of Atlantis Massif, which are dominated by serpentinized peridotite with gabbro making up only ~30% of those samples (Blackman et al., 1998, 2004). This discrepancy may simply reflect the difficulty of seafloor sampling from massive, undeformed exposures of possible gabbro, where subvertical outcrops up to 200 m high are cut by few fractures. Attempts to sample these sections with the Alvin were unsuccessful, so samples collected in 2000 were biased toward more fractured rocks, including those with well-developed foliation and/or fault rocks. Alternatively, it is possible that there are differences in the overall structure/composition between the southern ridge and the central dome, in which case the drill core provides important comparative data. Despite the high percentage of gabbroic rocks in Holes U1309B and U1309D being somewhat surprising, the recovered section provides an exceedingly rich data set for investigations on the petrogenesis and accretion history of ultramafic and mafic rocks generated at slow-spreading ridges. Because the petrology and alteration of the sequence varies on the scale of meters to a few tens of meters, relationships are available for study here that could not be addressed with prior ODP hard rock data.