Ellis 



prominence in frame C, but frame D shows them finally waning. It seems most 

 likely that this phenomenon was an effect of the shock wave arising from the 

 first collapse of the main cavity. Presumably, after arriving at the free surface 

 as a compression wave, this was reflected as a wave of tension, which subse- 

 quently exploded an array of microscopic nuclei somehow accumulated in the 

 vicinity of the main cavity. The main evidence in favour of this explanation is 

 that the time from the end of the collapse to the first appearance of the effer- 

 vescence was approximately the same as the time of transit at the speed of 

 sound from the cavity to the free surface and back. The evident abundance of 

 nuclei near the rebounding cavity is not yet understood; but possible sources 

 may have been, first, electrolysis accompanying the spark discharge and, sec- 

 ond, the remains of needlelike protuberances of the cavity surface formed dur- 

 ing its initial, explosive expansion around the spark track. 



REPLY TO DISCUSSION 



A. T. Ellis 



REPLY TO PROFESSOR HAMMITT 



Professor Hammitt's view that the craters shown by Knapp (17) could be 

 caused by jets rather than shock waves is well taken. Reexamination of the 

 pit-to-bubble diameter ratios quoted in that paper confirm this. However, from 

 personal observations made at the time I would not like to exclude the possi- 

 bility of depressions caused by shock waves. This point is now being inves- 

 tigated. 



The author has believed for a long time that Monroe jets are very closely 

 akin to the jets which occur during bubble collapse. The details are different 

 from the usual shaped- charge theory because inertial forces more complicated 

 than those of a thin liner are involved and because one does not rely on a pass- 

 ing detonation shock front to determine initial speed of the interface. Current 

 work does show that the concave liquid surface and the sign of the acceleration 

 are important. Dr. Hammitt is certainly correct in suggesting that the simi- 

 larity is more than coincidental. 



It should be pointed out that there is a thin jet in Fig. 11 but not, for ex- 

 ample, in Fig. 2. The jet is thicker in Fig. 2 and moves toward the solid as the 

 Naude'- Ellis theory predicts, but there is no present theory which would describe 

 the thin jet in Fig. 11. 



During cavity collapse there are likely to be tensions generated locally by 

 high flow velocities and accelerations. Professor Hammitt's suggestion that 

 additional cavitation is generated by local flows appears quite reasonable. 



160 



