The results of experimental investigations on turbidity currents of high density are described. Hydrodynamical properties are deduced which permit tentative extrapolation to the larger dimensions which may be expected in nature. The results are applied to the problem of submarine canyons. The reasons, given in an earlier paper, why the deposits tend to be graded are confirmed and the possibility of erosion is discussed. Settling velocities of said grains in clay and clay-sand suspensions are given and the influence of salinity is investigated. Viscosity measurements of clay suspensions are recorded. An attempt is made to find properties by which the deposits of ancient turbidity currents can be distinguished from those of normal processes of sedimentation. Filially, the problem is reviewed as to whether slumping or turbidity currents have played the major role in forming submarine canyons.