— Stratigraphic and petrographic analysis of the Cretaceous to Eocene Tibetan sedimentary succession has allowed us to reinterpret in detail the sequence of events which led to closure of Neotethys and continental collision in the NW Himalaya. During the Early Cretaceous, the Indian passive margin recorded basaltic magmaüc activity. Albian volcanic arenites, probably related to a major extensional tectonic event, are unconformably overlain by an Upper Cretaceous to Paleocene carbonate sequence, with a major quartzarenite episode triggered by the global eustatic sea-level fall at the Cretaceous/Tertiary boundary. At the same time, Neotethyan oceanic crust was being subducted beneath Asia, as testified by calc-alkalic volcanism and forearc basin sedimentation in the Transhimalayan belt. Onset of collision and obduction of the Asian accretionary wedge onto the Indian continental rise was recorded by shoaling of the outer shelf at the Paleocene/Eocene boundary, related to flexural uplift of the passive margin. A few My later, foreland basin volcanic arenites derived from the uplifted Asian subduction complex onlapped onto the Indian continental terrace. All along the Himalaya, marine facies were rapidly replaced by continental redbeds in collisional basins on both sides of the ophiolitic suture. Next, foreland basin sedimentation was interrupted by fold-thrust deformation and final ophiolite emplacement. The observed sequence of events compares favourably with theoretical models of rifted margin to overthrust belt transition and shows that initial phases of continental collision and obduction were completed within 10 to 15 My, with formation of a proto-Himalayan chain by the end of the middle Eocene.

A new Early Triassic marine fauna is described from the Cen- tral Oman Mountains. The fauna is Griesbachian in age, on the basis of abundant conodonts and ammonoids, and was deposited in an oxygenated seamount setting off the Arabian platform mar- gin. It is the first Griesbachian assemblage from a well-oxygenated marine setting and thus provides a test for the hypothesis that widespread anoxia prevented rapid recovery. The earliest Gries- bachian (parvus zone) contains a low-diversity benthic fauna dom- inated by the bivalves Promyalina and Claraia. A similar level of recovery characterizes the immediate postextinction interval worldwide. However, the middle upper Griesbachian sedimentary rocks (isarcica and carinata zones) contain an incredibly diverse benthic fauna of bivalves, gastropods, articulate brachiopods, a new undescribed crinoid, echinoids, and ostracods. This fauna is more diverse and ecologically complex than the typical middle to late Griesbachian faunas described from oxygen-restricted settings worldwide. The level of postextinction recovery observed in the Oman fauna is not recorded elsewhere until the Spathian. These data support the hypothesis that the apparent delay in recovery after the end-Permian extinction event was due to widespread and prolonged benthic oxygen restriction: in the absence of anoxia, ma- rine recovery is much faster.