35 



alone does not account for the lowest pressures. 40 Thus, it appears that 

 liquid phase is present throughout the cavity. That vapor phase is not pres- 

 ent throughout the entire wake, even with the pressures lower than vapor pres- 

 sure, may then be explained by establishment of a system in which equilibrium 

 is maintained between the liquid and vapor phases in such a way that the evap- 

 orization and entrainment of vapor is just balanced by reentrant liquid. Be- 

 cause of the large volume ratio between vapor and liquid, the presence of un- 

 evaporated liquid at pressures somewhat smaller than vapor pressure is quite 

 possible despite the high rate of evaporization, and, as a matter of fact, 

 there is photographic evidence that this conclusion is valid. 40 Furthermore, 

 because of the high rate of condensation, there are no difficulties in account- 

 ing for the vapor. 



It seems clear, then, that the actual composition of the cavities in 

 the experiments of Reference kO is a mixture of liquid and vapor phases and 

 that the surface oscillations are associated with a high rate of turbulent mix- 

 ing at the boundary of the cavitating zone and the general instability of 

 wakes. The resultant motion when considered as a wake is then a function not 

 only of the cavitation numbers but also the Reynolds number of the flow. It 

 was also pointed out in Reference U-0 that the finite angle of separation of 

 the cavities from the surfaces of the discs, hemispheres, semi-ellipsoids and 

 ogive section used in the experiments are, on the basis of the above descrip- 

 tions, merely a manifestation of the bounding streamline of a separated bound- 

 ary layer. It is interesting that the same conclusions as to treatment of the 

 cavitation as a wake phenomenon and as to the appearance of the separation 

 angles were reached by Wayland and White 58 in experiments on spinning spheres, 

 although the dependence on Reynolds number of pressure distributions on 

 spheres was already well known. 



Subsequent to the experiments described in Reference kO , high speed 

 motion pictures were taken of the cavity behind a disc. These showed that not 

 only does the surface of the cavity oscillate, but that the entire cavity ap- 

 pears to grow and collapse although it is difficult to show definite periodic- 

 ity or simultaneous collapse over the entire surface. Such oscillations were 

 also observed in the experiments of Hunsaker and Spannhake 59 in their experi- 

 ments in venturi tubes. The latter experiments are discussed in somewhat more 

 detail in a later section. 



Measurements of the drag of cavitating bodies were made by Reichardt 

 and also by the writer and Mr. Pond. These measurements will be discussed in 

 connection with the results of analytical descriptions of steady-state 

 cavities. 



