1490 
would not afford a good, critical method of selecting the velocity 
ratio.) 
3.5 The striking measure of agreement between theoretical and 
experimental curves in Figure & affords added confirmation of 
the predicted effects of the phase distortion. It must be 
remembered that the theoretical analysis assumed (1) plane 
rather than spherical waves, (2) smooth, plane, horizontal, fluid 
bottom rather than the actual sloping, undulating semi-solid 
one, (3) purely exponential decay of the primary shock wave, 
whereas it is known that the decay is exponential only for an 
interval equal to the time constant and then proceeds much more 
slowly. 
It is not surprising, therefore, that various 
discrepancies can be detected upon closer examination of 
Figure 2. Some reflections seem to disappear on the experimental 
records; others differ appreciably from the predicted shape; in 
some cases there are appreciable discrepancies in arrival times. 
All of these discrepancies could reasonably stem from errors 
in our knowledge of depths and ranges and from deviations from 
the idealized model assumed in the theoretical analysis. As 
would be expected, agreement between experimental and theoretical 
curves becomes poorer with increasing order of reflection. 
Many records were obtained in which the peak of the 
first bottom reflection was considerably higher than that of 
the incident shock, and much of the true peak may actually be 
- 1b. 
