ONE of the successes of the seafloor spreading hypothesis has been the coherent explanation it provides of world seismicity. Isacks et al.1 have given an account of seismic activity along mid-ocean ridges and island arcs in accordance with the concepts of plate tectonics, and McKenzie2 has shown that these concepts can be applied also to instantaneous motions across continental plate boundaries. It seems reasonable, therefore, to attempt an analysis of continental seismicity within plates in terms of the same hypothesis. Morgan3 and Le Pichon4 have shown that the assumption of perfectly rigid plates gives results which are in approximate accord with observation. Nevertheless, their results are unlikely to be adversely affected—and may well be improved—if inhomogeneous stresses within plates are postulated to explain patterns of earthquake activity in continental regions.