1563 
ideal behavior may not be realized at the receiving gauge owing 
to: (a) non-linear reflection (b) transmission to the air (c) 
sea-surface roughness (d) cavitation near the sea surface (e) 
propagation with finite amplitude (f) viscous attenuation. 
15-2 Keil has shown)? that for any charge weight and charge 
depth, if the angle of incidence at the surface is sufficiently 
Oblique, the reflected wave propagates along the initial wave 
front, producing a range of distortion in both waves. In all 
shots of the series reported here, at ranges less than 5000 ft., 
the direct and reflected waves are clearly resolved on the records, 
so that non-linear reflection plays no part in these cases. 
However, the few shots at extreme ranges (~10,000 ft.) exhibit 
gross attenuation of the direct shock wave and show nothing 
directly identifiable as a reflected wave. (See Figure 2e.) 
If a half-pound charge of pentolite is detonated at a depth of 
40 feet, distortion should begin to occur(according to reference 
15) at the surface approximately 8000 feet from the charge, and 
will then travel along the initial wave front, influencing it 
for example, to a depth of about 7 feet from the surface at a 
range of 20,000 feet. In the case of Figure 2e where the 
receiving gauges were at a depth of 25 feet, and at a range of 
9100 ft. the distortion already appears quite pronounced, while 
the theory indicates that it should not have set in. We are, 
however, close enough to the border line so that small changes 
in the parameters could reverse this prediction, and we can 
probably consider this as within the margin of error of the 
theory. Thus it is not unreasonable to ascribe tentatively, at 
ee 
