ar X iv : a st ro - p h / 95 06 10 6 v 2 2 1 Ju n 19 95 ARE GALAXIES OPTICALLY THIN TO THEIR OWN LYMAN CONTINUUM RADIATION ?

Previously published Hα data and UBV photometry of blue stars in the inner kpc of M33 are used to study the distribution of OB stars and HII regions in the galaxy and to determine whether individual regions of the galaxy are separately and/or collectively in a state of ionization balance. Based on the surface brightness of the Hα emission, we identify three distinct ionized gas environments (bright, halo and field). We find that ∼50% of the OB stars are located in the field, so that 1/2 of the lifetime of OB stars must be spent outside recognizable HII regions. If OB stars escape from bright HII regions by destroying their parent molecular clouds, this result would imply that molecular cloud lifetimes after forming OB stars could be as low as ∼5×10 yrs or 1/2 the typical lifetime of OB stars. We show that a possible origin for the large field OB population is that they were born in and subsequently percolated out of the ∼10 giant molecular clouds with masses ≥103M⊙ predicted to exist within the inner kpc of the galaxy. Using ionization models, we predict Hα fluxes in the bright, halo and field regions and compare them to those observed to find that the regions, separately as well as collectively, are not in ionization balance: predicted fluxes are a factor of ∼3-7 greater than observed. The heaviest loss of ionizing photons appears to be taking place in the field. Observed and predicted Hα luminosities in the field are in best agreement when Case A recombination is assumed. Therefore, our findings suggest that star formation rates obtained from Hα luminosities must underestimate the true star formation rate in these regions of M33. We have performed a similar analysis of an individual, isolated region with bright and halo Hα emission to find that comparable results apply and that the region, as a whole, is also not in ionization balance.