The purpose of this note is to demonstrate that recent Vp and Vs profiles are consistent with finite strain extrapolations of reasonable mineralogies. The difference in the size of the discontinuities in the Vp and Vs profiles is due to the relative pressure insensitivity of shear velocity in spinel structures. Most of the structure in the upper mantle can be attributed to the collapse of forsterite rich olivine to the β spinel and then the &ggr; spinel structures. The 640-km discontinuity on the basis of recent work appears to be due to the disproportionation of &ggr; spinel to a mixture of magnesiowustite (Mg,Fe)O and stishovite. The present results indicate that this postspinel phase in the lower mantle is enriched in SiO2 relative to the upper mantle. The 640-km discontinuity is therefore a chemical as well as a phase boundary.