A SPECTROSCOPIC SEARCH FOR RED SUPERGIANTS IN THE M33 GIANT H-II REGION NGC-604

We present high signal-to-noise ratio spectrophotometry obtained in th e optical and near-infrared with the double spectrograph ISIS on the W HT, of the M33 giant H Ir region NGC 604. Our main aim was to deduce g lobal properties of a nearby giant H II region, The near-infrared spec tral range, central to our study, comprises the gravity-sensitive Ca I I triplet (CaT) stellar absorption features. The data were obtained us ing two different techniques: (1) scanning with a 1.75-arcmin-long sli t that provided integrated spectra of the central part of the nebula, and (2) long-slit spectra of the brightest continuum knots. A single r ed supergiant (RSC) star was detected, judging from the measured stren gth of the stellar CaT absorption lines, visible in one region and onl y after carefully subtracting the hydrogen Paschen emission lines that dominate the near-infrared spectra. The feature is not observed in th e integrated spectrum. The observed Paschen discontinuity in emission allows us to determine an electron temperature, which is similar to th e one obtained from the ratio of forbidden line ([O III]) intensities, suggesting that temperature fluctuations are not present in the nebul a. Wolf-Rayet (WR) features have been found in several positions; the observed He II lambda 4686-Angstrom line intensity is found to be larg er than in galactic WR stars, by a factor of similar to 2. Exceptional ly broad components of permitted lines of hydrogen and helium (FWHM si milar to 2500km s(-1)) are observed in one of the brightest stars in N GC 604, This object shows a large overabundance of He and strong spect ral variability on time-scales of similar to 10yr, We classify it as a n LBV-WR transition object. We identify the core of the cluster ionizi ng the nebula, which is probably younger than 5 Myr. Implications of t hese results for the evolutionary state of NGC 604 are discussed. Also discussed are the difficulties involved in the techniques for using R SGs as tracers of starburst activity in galactic nuclei.