1574 
propagation of the cavitation itself and similtaneous piezoelectric 
measurements outside of the region of cavitation. 
One interesting consequence of the assumptions (a), (b), 
and (c) is that there should be no cavitation vertically below any 
charge at eny depth, although the cavitation may extend deeper than 
the charge off to the side- Thus it should be possible to avoid the 
obscuration due to cavitation encountered in photography of shallow 
explosions@+ by observing from directly below the charge. This 
would have the additional advantage of giving a well lighted silhouett: 
picture. 
Sharpness of the negative wave fronte 
17-1 It is frequently asserted that negative shocks are unstable and 
would tend to spread "rapidly". (see reference 7, p- 25, and reference 
25, p- 181). It is interesting to examine the available experimental 
records in the light of this prediction, remembering, of course, that 
the minimum "fall time" which can be resolved is the transit time 
across the gaugee Examination of the records reproduced in Figures 
2 and 3 (and of the other existing records as well) shows that the 
"fall" of the negative wave is equally sharp at all ranges, i.e. of 
the order of magnitude of the gauge transit time of 20 to 30 micro- 
seconds, and that any spread of the front due to the combined effects 
of finite amplitude and viscous attenuation mst be smaller than this 
magnitude. 
It is possible to make a theoretical prediction of the ex- 
pected spread by application of the methods of section 9.3. In the 
case of the negative wave, the viscous and finite amplitude effects 
both tend to spread the front instead of opposing each other as in 
ee 
