Seismic evidence indicates that the principal orogenic structure responsible for each of the great linear and curvilinear mountain ranges and oceanic trenches is a complex reverse fault. A study of eight regions in which orogenic activity is in progress reveals that these great faults occur in two basic types, here designated oceanic and marginal. Oceanic faults, situated within the oceanic domain, extend from the surface to depths of 550 to 700 km. They exhibit an average dip of 61°. Their elastic strain-rebound characteristics show that these faults are composed of two separate mechanical units—a shallow component extending from the ocean bottom to a depth of roughly 60 km, and a deeper component extending to the 700km crustal boundary. The marginal faults situated along the continental margins occur in dual and triple forms. The dual faults comprise a shallow member extending from the surface to a depth of approximately 60 km and an intermediate member extending to a depth of 200 to 300 km. The average dip is 33°. The marginal triple form is similar to the dual down to the 300 km level. At this depth the dip changes abruptly to 60° to form a third component extending down to the 650± km crustal boundary. The elastic strain-rebound characteristics of the marginal faults indicate that the components of these structures also move as separate units, although in South America the two lower elements exhibit some evidence for mechanical coupling. In the continental domain the 300 km level thus represents a tectonic discontinuity not as yet revealed by seismic wave-propagation studies but which is apparently the lower boundary of the continents. Since the oceanic faults and the deep components of the marginal faults have the same average dip (61°) it may be assumed that both are fractures in a single, continuous mechanical structure subject to a single stress system. The different average dip (33°) of the marginal intermediate fault components suggests that they occur in a structure mechanically distinct from the deep oceanic and continental layer and that they are activated by a different stress system. A hypothesis offered for the origin of the volcanoes associated with the faults assumes that the source of volcanic energy is heat produced in the fault rocks by the inelastic components of the repeated to-and-fro strains involved in the generation of the sequences of earthquakes and aftershocks.