Interaction of hot stars and of the interstellar medium. VII. The rate and fate of stellar ultraviolet photons.

Summary. We use the new set of observational data gathered by Y. M. Georgelin (1975) on exciting stars of galactic nebulae to reanalyze the calibration of spectral-type into effective-temperature. We used the available model-atmospheres which appear to be the closest to an ideal blanketed NLTE spherical model, as discussed by Panagia (t973), and a sample of nebulae best fitted for the use of the Zanstra method. We obtain an empirical calibration of (the ratio of the ionizing photons flux to the energy emitted in the photometric V-band) as a function of the spectral MK types in the range 0 6B 0, and then the total photon luminosity and excitation-parameter In Part II, we show that additional nebulae not used for the calibration fall symmetrically on each side of the mean curve previously obtained in the , spectral type plot. Thus, density-boundness is not the rule for optical nebulae, contrary to a suggestion by Churchwell (1974) for stars hotter than 0 8. In Part III, we make a detailed comparison of the effective-temperature and excitation-parameter scales obtained by different authors. Our effective-temperature scale is in excellent agreement with the one obtained by the first application of this method from fewer data by Morton (t969). It is in good agreement with the results of Conti (1973) using a quite independent spectroscopic method and NLTE model-atmospheres. The effective-temperature, spectral-type scale obtained from LTE model-atmospheres is definitely higher than the Zanstra scale, and we feel this is an additional argument to dismiss LTE model-atmospheres for hot stars. We do not think that the validity of the Zanstra method may be questioned by the presence of dust in the nebula as far as our calibration sample of nebulae is concerned. We emphasize the advantages of using the ratio NJitFv rather -than an absolute flux (like (photons s -1) or U ) to describe the excitation of the nebula, because this ratio is quite insensitive to gravity, model-atmosphere, absolute magnitude scale and distance. Key words: hot stars effective-temperature Lyman continuum luminosity - H ii regions dust