Key points are not available for this paper at this time.
An empirical comparison of the observed H-R diagrams for the supergiants in our region of the Galaxy and the Large Magellanic Cloud reveals comparable distributions of spectral types and luminosities in the two galaxies. Supergiants of similar spectral types have the same luminosities, except for the A-type stars, where selection effects may be important. These results suggest that the same basic physical processes govern the evolution of the most massive stars in the two galaxies. Variations in the blue-to-red supergiant ratio with galactocentric distance and with luminosity involve chemical composition gradients and varying rates of mass loss. Since the relative numbers of the most luminous stars are more sensitive to mass loss, the B/R ratio from the less luminous supergiants may be a better indicator of galactic abundance gradients. The upper luminosity boundary for both the galactic and the LMC supergiants is characterized by (1) decreasing luminosity with decreasing temperature for the hottest stars and (2) an upper limit to the luminosities near Mbol - 9.5 to - 10 mag for stars cooler than 15,000 K. We suggest that the observed luminosity limits are due primarily to the effects of large mass loss on the evolution of the most massive stars. The examples of Car and P Cyg suggest that mass-loss rates can be very rapid and unsteady-higher on the average than presently observed for most of the hot supergiants. The evolution of stars greater than 60 to cooler temperatures is consequently limited by instabilities and the accompanying high mass loss. An initial mass near may be an empirical upper limit to the mass at which a star can evolve to the region of the M supergiants and probably accounts for the observed upper bound to the luminosities of the cooler supergiants. Subject headings: galaxies: Magellanic Clouds - galaxies: Milky Way - galaxies: stellar content - stars: evolution - stars: massive - stars: supergiants
Humphreys et al. (Sat,) studied this question.