The problem of determining the past relative positions of continents now situated around the North Atlantic has been a challenge to geologists and geophysicists for many years. Yet a solution acceptable to all does not exist. Since the acceptance of the idea of continental drift, a number of reconstructions has been published. The earlier reconstructions, prior to the birth of the plate tectonic theory, were done qualitatively in order to match morphological features of the continental margins (Choubert, 1935) or geological boundaries on land. The first quantitative fit of the North Atlantic was published by Bullard and others (1965) who described the motion between continents or plates as angular rotation about a set of poles. Since then a number of papers have been published showing reassembly of continents during various stages of evolution of the North Atlantic using magnetic anomalies in the oceanic regions as isochrons or plate boundaries and fracture zones as the direction of motion between plates. The most notable of these are by Pitman and Talwani (1972), Laughton (1971, 1972), Williams (1975), LePichon and others (1977), Sclater and others (1977), Kristoffersen and Talwani (1977) , Talwani and Eldholm (1977), Kristoffersen (1978), Srivastava (1978, 1985), Olivet and others (1981), Unternehr (1982), Nunns (1983), and Vink (1982, 1984). Most of these reconstructions are based on a limited amount of geophysical information from one or more regions of the North Atlantic available at the time of publication. Thus different criteria and assumptions had to be made

No abstract

The boundaries of three major plates (Africa, India, and Antarctica) meet in a triple junction in the Indian Ocean near 25°S, 70°E. Using observed bathymetry and magnetic anomalies, we locate the junction to within 5 km and show that it is a ridge‐ridge‐ridge type. Relative plate motion is N60°E at 50 mm/yr (full rate) across the Central Indian Ridge, N47°E at 60 mm/yr across the Southeast Indian Ridge, and N3°W at 15 mm/yr across the Southwest Indian Ridge; the observed velocity triangle is closed. Poles of instantaneous relative plate motion are determined for all plate pairs. The data in the South Atlantic and Indian oceans are consistent with a rigid African plate without significant internal deformation. Two of the ridges at the triple junction are normal midocean spreading centers with well‐defined median valleys. The Southwest Indian Ridge, however, has a peculiar morphology near the triple junction, that of an elongate triangular deep, with the triple junction at its apex. The floor of the deep represents crust formed at the Southwest Indian Ridge, and the morphology is a consequence of the evolution of the triple junction and is similar to that at the Galapagos Triple Junction. Though one cannot determine with precision the stability conditions at the triple junction, the development of the junction over the last 10 m.y. can be mapped, and the topographic expressions of the triple junction traces may be detected on the three plates.

Christ o p he r Taps c 0 t t * Departmeit t of k.arth and Planetary Scicn ces.