22 249 



V ^^ Is here nearly equal to p^/pc, so that Vj^^ Just about cancels the last 

 term In Equation [3] and v^^ is small. It may safely be inferred that the 

 particle velocity in the cavitated region will grade from a small value at 

 the bottom to about 49 feet per second at the top. 



The whole cavitated layer, 9 feet thick, should therefore rise, 

 carrying a thin uncavitated sheet on top of it. This "solid" sheet will in- 

 crease in thickness as its lower boundary travels downward in the form of a 

 forced closing-front. The numerical values cited Indicate that the center 

 of gravity of the upper 10 feet of water will start upward with a velocity 

 of perhaps 25 feet per second; and there should be a downward acceleration of 

 g due to gravity and of {3'^/^0)g due to air pressure on the top, or a total 

 of U.U g. The center of gravity should rise, therefore, not over s = v^/2n = 

 25^/8.8 X 32 = 2.2 feet, during a time v/^g or 0.2 second. The surface of 

 the water will rise higher but certainly not more than twice as high or, at 

 the utmost, 5 feet. 



Now this picture as inferred from the analysis appears not to agree 

 too well with the facts. Hilliar's observations indicate that, in the case 

 considered, the dome v;ould certainly be less than 6o feet in radius but would 

 rise in a second or so to a height of 1 5 or 20 feet. The analytical estimate 

 would be changed considerably if a different breaking-pressure were assumed, 

 or if more recent values for the incident pressure were employed, but a large 

 disagreement with observation would remain. The large rise that is actually 

 observed could be explained only by supposing that the disintegration of the 

 water extends up to the surface and serves to admit atmospheric pressure to 

 the interior. Cavitation up to the surface might result from the initial 

 presence of air bubbles in the upper few feet of water, which would effec- 

 tively raise p^, perhaps up to p^. The whiteness observed in all explosions 

 of this kind does, in fact, extend to the very edge of the dome in the photo- 

 graphs; see Figures 12 and 15. It is not easy to believe, however, that air 

 can mix sufficiently rapidly with the cavitated v;ater to relieve the vacuum 

 effectively. 



The jaggedness of the edge of the dome, so clearly revealed by the 

 photographs, suggests a modified hypothesis. Perhaps the general mass of 

 water really does rise only a few feet, as the analysis suggests, and what 

 is seen as a white dome of considerable height is only an umbrella of spray 

 throvm up from the surface. 



The origin of the spray itself is perhaps to be found in an insta- 

 bility of the surface under impulsive pressure. The pressure gradient is 

 equivalent to a momentary increase of gravity by a factor of 100 to 1000, 

 followed by a reversal to similar values. If there are any small waves on 



