We use the standard , adiabatic shell evolution to predict the differential size distribution N ( R ) for populations of OB superbubbles in a uniform ISM .
Assuming that shell growth stalls upon pressure equilibrium with the ambient ISM , we derive N ( R ) for simple cases of superbubble creation rate and mechanical luminosity function ( MLF ) .
For constant creation and an MLF \phi ( L ) \propto L ^ { - \beta } , we find that N ( R ) \propto R ^ { 1 - 2 \beta } for R < R _ { e } , and N ( R ) \propto R ^ { 4 - 5 \beta } for R > R _ { e } , where the characteristic radius R _ { e } \sim 1300 pc for typical ISM parameters .
For R < R _ { e } , N ( R ) is dominated by stalled objects , while for R > R _ { e } it is dominated by growing objects .
The relation N ( R ) \propto R ^ { 1 - 2 \beta } appears to be quite robust , and also results from momentum-conserving shell evolution .
We predict a peak in N ( R ) corresponding to individual SNRs , and suggest that the contribution of Type Ia SNRs should be apparent in the observed form of N ( R ) .
We present expressions for the porosity parameters , Q _ { 2 D } and Q _ { 3 D } , derived from our analysis . Q _ { 2 D } is dominated by the largest superbubbles for \beta < 2 and individual SNRs for \beta > 2 , whereas Q _ { 3 D } is normally dominated by the few largest shells .

We examine evolutionary effects on the H II region luminosity function ( H II LF ) , in order to estimate \beta .
We find that for a nebular luminosity fading with time t , { \cal L } \propto t ^ { - \eta } , there is a minimum observed slope a _ { min } for the H II LFs .
Empirical measurements all show a > a _ { min } , therefore implying that usually we may take \beta = a .
We also find that if nebular luminosity is instantaneously extinguished at some given age , rather than continuously fading , no a _ { min } will be observed .

Comparison with the largely complete H I hole catalog for the SMC shows surprising agreement in the predicted and observed slope of N ( R ) .
This suggests that no other fundamental process is needed to explain the size distribution of shells in the SMC .
Further comparison with largely incomplete H I data for M31 , M33 , and Holmberg II also shows agreement in the slopes , but perhaps hinting at systematic differences between spiral and Im galaxies .
We estimate porosities that are substantially < 1 for all of the galaxies except Holmberg II , for which we obtain values \ga 1 .
Most of these galaxies therefore may not be strongly dominated by a hot interstellar component .
However , porosity results for the Galaxy remain inconclusive with the available data .