The estimates made in Reference 3 were based on results of Taylor"^ and Green^ who 

 found from computations based on models of isotropic turbulence that the pressure fluctuation 

 p'may be written 



^ 2 ^ 



where p is the mass density of the fluid and w' is the velocity fluctuation. The value of K 

 given by Taylor and by Green for a number of theoretical models is of the order of one. In 

 1950, Batchelor^ considered this problem in the light of modern theories of turbulence and was 

 able to derive the following relation for the fluctuating pressures in homogeneous, isotropic 

 turbulence: 



7^ = 0.34p2(^)2 



In terms of Taylor's formula, this corresponds to a value of K of 0.39. Since this magnitude 

 is essentially of the order assumed in estimating the effects on cavitation, the previous com- 

 putation need not be modified. Batchelor found further that the pressure scale is of the order 

 of half the velocity scale. In addition to such computations, actual measurements of pressure 

 fluctuations will be required.* Such measurements should include the temporal as well as 

 spatial correlations with the ultimate aim of estimating the time available for formation of 

 cavities. 



The average conditions in both laminar and turbulent boundary layers are also of some 

 interest with respect to analysis of behavior of nuclei within the boundary layer. It was 

 shown in Reference 3 that, aside from low pressures associated with turbulence which occur 

 at a small distance from the boundary, the minimum pressures may be expected to occur at the 

 boundary rather than elsewhere in the boundary layer. Furthermore, nuclei which are exposed 

 to low pressures in the slowly moving fluid of the boundary layer have a longer time in which 

 to grow beyond a critical size and start the cavitation process. 



Further work on boundary layer and wake flows will, of course, be necessary before 

 the imiportance of the processes in such flows in relation to cavitation can be fully evaluated. 

 Recent studies by Townsend' have shown that the outer regions of the turbulent boundary 

 layer resemble wake flows with a rather well-defined boundary between the turbulent and non- 

 turbulent regions but that at any point in this outer region, the turbulence is intermittent. 

 The picture presented is that of a wake with jets or fingers of turbulent fluid extending into 

 the nonturbulent fluid. This phenomenon will be mentioned again in connection with the 

 discussion of steady-state cavities in real liquids. 



*Such measurements are being taken as part of a program of studies of jet cavitation at the Iowa Institute of 

 Hydraulic Research. Preliminary results were reported by Dr. H. Rouse at the Eighth International Congress on 

 Theoretical and Applied Mechanics, Istanbul, Turkey, August, 1952. However, published results, which will be 

 of considerable interest in relation to these questions, are not, as yet, generally available. 



