Conclusions

1. The presence of a prominent transverse ridge is typical of several fracture zone ( i.e., Romanche FZ, Kane FZ, etc.). A number of factors may have contributed to their formation. These factors, as discussed in details by several authors BONATTI1978, BONATTI1982, PARMENTIER1986, RUTTER1987, CHEN1988, MORGAN1988, BONATTI1994, POCKALNY1996, include thermal rejuvenation of old lithosphere near the RTI, frictional shear heating along the transform, lithospheric flexure due to thermal stresses, nonlinear viscoelastic deformation of the lithosphere, hydration/dehydration reaction of upper mantle rocks near the transform igneous activity and transform-related transpression transtension. Compressional or extensional stresses, due to reorientation of the transform fault and ridge axis following changes in the direction of the plate motions, are able to build up corrugation or flexural uplifted blocks, respectively, with vertical motions comparable to those observed in transverse ridges BONATTI1978, POCKALNY1996).

2. The Vema transverse ridge is unique in so far as it exposes a nearly undisturbed and complete section of oceanic lithosphere with lateral continuity, it gives us the opportunity to study the signature of temporal and spatial variations in the activity of a slow spreading ridge such as MAR. Following this main objective we sampled ultramafic rocks, at sites 5 km apart, along a 143 km long stretch of VTR northern wall, in a depth range of 4300-5000 m. The VTR can be interpreted as a continuous sector of oceanic lithosphere, but characterized by sub-domains coherent in lithology. Preliminary observations on these samples show systematic lateral variations in the structure and composition of the upper mantle, caused probably by temporal variations in the processes of accretion at ridge axis. Some intervals are made prevalently by coarse-grained, porphyroclastic peridotites, while in other intervals strongly deformed, mylonitic peridotites are prevalent. In addition, amphibole-rich ultramafics are com- mon in some intervals.

3. Slide and slumping deposits are visible at the base of the southern scarp of the VTR (Fig. 6). Slope instability probably contributes, in this area, to the symmetric profile of the VTR. The seismic reflection data reveal the presence of two listric slide planes at the base of the northern wall of VTR. The origin of the mid-scarp terrace on the northern side of the VTR is probably related to selective erosion between massive or slightly tectonized gabbros and the dyke complex. There is no evidence of tectonic control on the terrace.

4. The PTDZ has been subjected to recent compressional movements.

5. The sedimentary rocks sampled along the lower VTR scarp are mostly products of the physical and chemical disgregation of the transverse ridge rock units. The upper VTR slope (Fig 3a e 3b) is a typical young active slope ($>$20°) where intense mass wasting has occurred. It is cut by scoop-shaped scarps leading to a distinct gully and canyon topography. These gullies and canyons cut into the basalt and dyke complex unit, channeling debris transport downslope toward a gentler slope ($<$20°). At the base of these slopes or where the terrace is present, the gully and canyon network leads to coalescing debris fans. The sedimentary samples that we collected at various locations along the VTR include sedimentary breccias, sandstones and limestones: their matrix is rich of pelagic components such as foraminifera that will allow age determinations.