It is well known that the ocean floor is not everywhere made of mafic rocks but also consists locally of serpentinized peridotites. Proposed models for the presence of the peridotites at, or close to, the seafloor are still debated. Many questions remain pertaining to the emplacement mechanisms of the peridotites. In general, peridotite occurrences can result from original mantle upwelling processes or from crystal fractionation in the layered gabbro zone that is immediately above the upper mantle. In places where the oceanic crust is very thin, such as along transform faults or along ridge valleys of slow-spreading ridges, it is quite possible that peridotites can be carried directly onto the oceanic floor by motion along normal faults (Bonatti and Honnorez, 1976; Girardeau et al., 1988). Such a mechanism has been proposed to explain the uplift of the peridotites recovered along ridges or transform fault segments of ridges. On the other hand, hot mantle uplift during continental rifting or asthenospheric vertical diapirism has been proposed to explain the original ascent of peridotites exposed on Zabargad Island (Nicolas et al., 1985, 1987; Bonatti et al., 1986), in the Alps (Nicolas, 1984), in the Betics (Obata, 1977; Tubia and Cuevas, 1977), and in the Pyrenees (Vielzeuf and Kornprobst, 1984). To explain the peridotites cropping out at the boundary between continents and oceans, Boillot et al. (1987) proposed that the rise of peridotites to the surface can result simply from tectonic denudation of the mantle as a consequence of stretching of the lithosphere during rifting. Beslier et al. (1990) further suggested that under the tectonic denudation model peridotite emplacement is strongly influenced by the extensional stress field at the end of continental rifting. At present, evidence from petrologic and fabrics studies of the peridotites alone cannot discriminate among various models. Additional geological and geophysical data are needed.
Another inherent problem in the investigation of peridotites is the determination of their precise ages. This stems largely from the absence of biostratigraphic dating and correlation in peridotites. An added problem is that most recovered peridotite has suffered extensive low-temperature and late-stage alteration, as typically evidenced by calcite veining as well as replacement of the serpentine by calcite, which makes it difficult to apply radiometric dating methods to these peridotites.
Paleomagnetism has proved to be a powerful tool in studying stratigraphic, tectonic, paleoclimatic, and paleoceanographic problems, and it continues to play a pivotal role as a standard for age-dating and correlation in Ocean Drilling Program (ODP) studies. Paleomagnetic dating is based on the facts that the Earth's magnetic field occasionally reverses polarity and that many rocks retain a magnetic imprint of the field at the time they were formed or altered. Under favorable circumstances, paleomagnetic dating can furnish highly resolved numerical ages by identifying the polarity patterns and fitting the polarity patterns into biostratigraphically identified zones, the geomagnetic polarity time scale, or other geochrological framework.
In this paper, I present a detailed account of paleomagnetic and rock magnetic results from the peridotite samples from Holes 897D and 899B drilled during Leg 149. By combining the paleomagnetic results with other available geological data, useful constraints can be placed on the timing of emplacement and the postemplacement alteration of the recovered peridotites. This information may further our understanding of the processes that accompanied continental breakup and the onset of steady-state seafloor spreading.