Key points are not available for this paper at this time.
Measurements of the velocity of compressional waves in silicates and oxides having a range of density from 2.6 to 5g/cm3 suggest a simple dependence of velocity upon density and mean atomic weight. Consequences of the assumption that this relation remains valid at high pressures are examined with reference to the composition of the mantle, especially of the transition layer. It appears probable that the abnormally high rate of increase of velocity with depth in the transition layer may be accounted for principally in terms of phase change, with little change of composition. Recent studies with strong shock waves are examined in connection with the composition of the core; the evidence is unfavourable to the hypothesis of a metallic core of light elements, but is consistent with a core of iron alloy.
Francis Birch (Fri,) studied this question.