The energy observed in the seismic waves P and pP in a large number of earthquakes is compared, by means of two partly independent methods, with the theoretical energy calculated from a standard equation. The results are analyzed, depth of focus, distance from epicenter to observing station, geographical location of epicenter, and azimuth from station being used as variables. When compared with the theoretical ratio, the ratio of the energy in pP to the energy in P averaged over a distance range 60°-90° is observed to decrease with depth, by 0.5 on a logarithmic scale of energy between 100 and 600 km. depth of focus. The results for the two waves are compared separately with theoretical values, and the observed effect appears to be about equally due to an increase of P energy and a decrease of pP energy with depth. The theoretical formula is reëxamined to determine if permissible changes in the assumptions or numerical values can account for the results. No such changes are found; certain assumptions with respect to increased absorption of energy near the surface of the earth provide a partial qualitative explanation, but quantitatively they cannot be reconciled with data from other sources. The variation with distance between observed and calculated energies is not large enough to be treated quantitatively; slight changes in the slope of the accepted velocity vs. depth curve are tentatively suggested on the basis of it. The energy ratio pP/P compared to the theoretical ratio is significantly too large in shallow earthquakes occurring in the Aleutian region and can be attributed to pP rather than P. For shallow shocks in the New Hebrides region and very deep shocks of the Southwest Pacific there is some indication that the energy ratio is smaller than for comparable shocks elsewhere. Data on observed wave periods are given. The effect of differences between instruments is considered.