experiments with sea water, the fresh water must be super- 

 saturated with air (about 3,0% supersaturation)o This is again 

 compatible with the expected behavior of nuclei in that sea 

 water which is saturated near the surface also certainly con- 

 tains a large number (relatively) of gas- bubbles whether 

 entrained or stabilized on particles (plankton 3, etCo)o Thus, 

 supersaturated fresh water insures the growth of undissolved 

 nuclei to provide the foci for cavitation inception^ However. 

 it is not likely that the question will be completely settled 

 until the behavior and role of such nuclei can be studied in 

 detail In the sense suggested in the text of the paper,. 



With regard to Dro Todd's question concerning the damage 

 problem, the writer's views are essentially the same as were 

 expressed in the earlier paper (TMB Report 712 )« ioe,j that 

 damage is associated with the pressures arising ^upon collapse 

 of transient cavities whether resulting directly in fatigue 

 failures or in producing deformations in the crystal structure 

 of a metal which result in internal corrosive electric currents > 

 The absolute pressures which can be developea in the liquid 

 depend upon the permanent gas in the bubble, the rate of 

 condensation of vapor, the viscosity of the liquid ^ the sur» 

 face tension^ compressibility, etc, and no computations have as 

 yet been carried out including all these variables which 

 enable firm statements to be made concerning the absolute pres- 

 sures in real liquids » It will be noted that except for sur- 

 face tension all these variables tend to retard the rate of 

 collapse o Furthermore, accurate measurements of the peak 

 values have not as yet been made since these occur in only 

 the last few microseconds of the collapse cycle and such 

 resolution with presently available instrumentation is diffi- 

 cult to attain o The problem is further complicated by the 

 fact that bubbles near a surface or in pressure gradients 

 do not collapse spherically and the pressures in such cases 

 are less than for spherical collapse o Nevertheless ^ that 

 the pressures developed are sufficient to cause damage 

 has been convincingly demonstrated by direct experiments as 

 well as indirectly by such experiments as those in which air 

 is introduced into the cavitated region to '"cushion" the 

 collapse 6 Whether the hydrodynamic pressures are alwayg 

 large enough to cause damage is still an open question, 

 however,, In TMB Report 712,, the writer suggested that another 

 possible mechanism may be associated with impacts produced 

 by jets which might arise as a result of non- spherical collapse 

 of the bubbles, it being known that the impact of drops and 

 jets cause the same type of damage as that observed in cavita- 

 tion damage o 



