Peterson 
stream velocity in the water tunnel. This conclusion assumes that 
the viscous effects, such as boundary layer separation, do not in- 
fluence inception. Further discussion on boundary layer separation 
will be deferred until later. Following the same reasoning, if the body 
size had been changed, then the number of freestream bubbles cavita- 
ting per unit time would vary directly with the square of the ratio of 
the body diameters. 
The key to all of the preceding discussion on bubble screening 
is whether in fact the free stream gas bubbles are responsible for 
cavitation. The remainder of this paper will be concerned with expe- 
riments specifically planned to extend our understanding of the role 
of gas content in water. 
VARIATION OF FREE GAS CONTENT AND BODY SURFACE QUALITY 
A. Criteria for Cavitation inception 
The commonly accepted criteria for the onset of vaporous ca- 
vitation is when a cavity grows "explosively", with the local pressu- 
re less than or equal to the vapor pressure (e.g. er and is ge- 
nerally considered the only true cavitation. On the other hand, gase- 
ous cavitation can occur at pressures either greater or lower than 
vapor pressure, with gas diffusion into the bubble possibly important 
and the growth rate of the bubble considered something less than 
"explosive''. However, these definitions are not specific enough for 
the purposes of the discussion here to delineate when in fact a cavity 
is growing "explosively", This problem was apparent to Hsieh [12] 
when he calculated bubble dynamics in the bubble trajectory. None of 
the bubbles he considered had what could be considered '"'explosive" 
growth, but in fact had a bounded maximum size. Therefore Hsieh 
arbitrarily defined a bubble to be cavitating when its diameter reach- 
ed a certain minimum ''visible''size. 
On the basis of the trajectory and bubble diameter calculations 
in the previous section for the NSRDC headform, it is felt that only 
bubbles actually striking the body could nucleate a vaporous cavity. 
This assumption is based on the observation through high speed pho- 
tography that the cavities appeared hemispherical in shape from the 
time of their first observation and translated along the body surface 
during both the growth and collapse. The "'visible'' size criteria is 
not applicable here since once a bubble strikes a body, the calcula- 
tions are no longer valid and in fact the visually observed cavity 
growth could correspond to gaseous cavitation. Since all of the cavi- 
ties observed in these studies were observed to be translating along 
1138 
