The measurements of the velocity of compressional waves up to 10 kilobars for some 250 specimens of rock, reported in part 1, are discussed with respect to the effects of porosity, alteration, anisotropy, and composition. The relations of isotropic elasticity are shown to be approximately valid for a number of examples. Reasonable agreement with theoretical values for quasi-isotropic aggregates is demonstrated where comparison is possible. At pressures above a few kilobars, the principal factors determining velocity are density and mean atomic weight; oxides and silicates conform to the same general relations, with a few exceptions. Details of symmetry or crystal structure appear to be of secondary importance. Velocity is an approximately linear function of density for materials having a common mean atomic weight, but is in general not a single-valued function of density alone.