[S-III] EMISSION AND EXTINCTION IN EXTRAGALACTIC H-II REGIONS

We critically examine the validity of the [S III] λ9532/Hα emission-line ratio as a quantitative extinction tracer in H II regions, and as a probe of highly obscured H II regions in infrared-luminous galaxies, as reported recently by Young, Kleinmann, and Allen. We have combined [S III] data from the literature with nebular models to determine the intrinsic error in [S III]-derived extinctions, and have obtained follow-up spectrophotometric observations of the "hidden" H II regions studied by those authors. Although we confirm the presence of extended [S III] emission in these galaxies, the actual [S III] intensities and [S III]/Hα ratios are 5-20 times lower than reported by Young, Kleinmann, and Allen. Analysis of our spectra demonstrates that most of the [S III] line emission reported by those authors is underlying stellar continuum radiation. The actual [S III]/Hα ratios in these regions are only slightly higher than average for normal disk H II regions. There are unquestionably regions of heavy extinction in these galaxies, and we cannot rule out the possibility of as-yet-undetected highly obscured H II regions. An analysis of the dispersion in intrinsic [S III]/Hα ratios in H II regions indicates that this method will yield extinctions with random errors of ± 2-3 mag, and larger systematic errors which correlate with galactic radius, abundance, and ionization mechanism. Several other techniques, using thermal radio emission or ratios of Brackett, Paschen, and/or Balmer recombination lines, are suggested as superior means of searching for and studying highly obscured H II regions in galaxies.