CONCLUSIONS

The principal conclusions from the results of Leg 173, combined with independently acquired geophysical data, are as follows:

1. All fault blocks imaged on seismic reflection profiles are blocks of thinned continental (and not oceanic) crust. At least one block (Site 1065) was tilted during rifting.
2. The continental crust thins dramatically seaward (practically to zero) and is broken into blocks by low-angle detachment faults. These blocks are probably underlain by a tectonic crust/mantle boundary, which in the center of the transect on a basement ridge called Hobby High has been uplifted to within a few hundred meters of the top of acoustic basement.
3. All the mafic cores from Sites 900, 1067, and 1068 on Hobby High are of late Hercynian (270 Ma; early Permian) age and were emplaced in the lower continental crust or at the base of thinned crust. These mafic magmas were derived from melting of heterogeneous mantle sources during a late Hercynian extensional phase. The mafic magmas crystallized to form both cumulate and noncumulate gabbros, followed by ductile shearing at middle to lower continental crustal depths.
4. At the Hobby High sites, exhumation of lower crustal rocks proceeded over tens of millions of years and preceded the onset of seafloor spreading.
5. The Site 1068 and 1070 peridotites are not as depleted as typical abyssal (oceanic) peridotites and are more likely to be derived from suprasubduction zone or subcontinental mantle. The mantle rocks show geochemical and mineralogical evidence of heterogeneous partial melting (probably <10% melting). They were also locally and intensely percolated by melts during the last stage of high-temperature deformation that generated a porphyroclastic foliation. Available data indicate a range of compositions for these percolating melts, from depleted MORB-like trace element values to more enriched compositions that crystallized kaersutite and phlogopite.
6. The basement cores at Site 1070 show gabbro pegmatite overlying serpentinized peridotite veined by gabbro. On the other hand, geophysical observations indicate the presence of oceanic crust at this site but, remarkably, no rocks from the upper oceanic crust (neither basalt nor sheeted dikes) were encountered. The cores may indicate oceanic crust formed at a time when tectonism was more important than magmatism, but we cannot be sure that Site 1070 is representative of the surrounding crust.
7. The basement rocks at Site 1070 were exhumed at the seafloor at least 14 m.y. after the crustal age computed from seafloor-spreading magnetic anomalies. Either this exhumation was more rapid than at Hobby High, or more likely, the 250°-500°C isotherms were much closer together at Site 1070.
8. The surprising lack of synrift melt products in the basement cores from sites over thinned continental crust or the OCT zone can be explained qualitatively by the gradual post-breakup evolution of the margin toward steady-state seafloor spreading. Some indirect evidence exists for intrusive bodies within the top 6 km of the acoustic basement of the OCT, but their age of intrusion is unknown.
9. Strong and informative parallels were noted between the character and history of the Leg 173 (and Leg 149) cores and the character and history of the rifted margins and transition zones exposed today in the Alps.