On the spectrum of H II regions

It is demonstrated that the validity of present models of H II regions is not well established. It is clear that simple models (and the usual formulas involving line ratios) are unable to provide a correct interpretation of the nebular spectrum. The importance of several geometrical and density effects which may account for some peculiar spectral features (such as bright forbidden S II lines and anomalous abundances) is emphasized. It is concluded that the best course to follow in H II region spectral analysis is the comparison of observations and sophisticated models on a statistical basis.