We present preliminary results on an extensive spectroscopical study of the WR Ring nebula NGC 6888. The observations combine high spatial (1.5 arcsec/pixel) and spectral resolution (30 to 50 km s–1) covering most of the optical range - λλ 3600 to 6800 å- at 3 different slit positions along the major axis of the nebula. The spectra of the central parts give an emission system with three different velocities: a) VLSR=-64 km s–1, b) VLSR=+18 km s–1, and c) VLSR=+78 km s–1. Assuming that NGC 6888 is in fact an expanding bubble of gas, as demonstrated by Marston and Meaburn (1988), we can identify components a) and c) as those moving towards us and receeding parts of the shell, while b) could be related to the ambient interstellar ionized gas outside the bubble. We have isolated the spectrum for each component in order to analyse the ionization structure and excitation mechanism, as well as to derive their physical conditions and chemical abundances. The use of diagnostic diagrams indicates that the nebula is basically photoionized, without any significant contribution of shock excitation in the zones studied. The ambient interstellar component shows a spectrum typical of an H II region with nearly solar abundances. In the case of the bubble components, their spectra produce [N II]/Hα line ratios which are outside the H II region box, and entering the extended planetary nebulae locus, suggesting a contribution of ejected nitrogen in the bubble. This fact is evident from the abundance analysis we have performed, finding that, with respect to the values quoted for the ambient gas, the O/H is deficient by a factor 4, N/H is 2.5 times higher, and helium is enhanced by a factor 2. These quoted values clearly indicate that a substantial fraction of the gas in the bubble is processed material ejected from the central massive star.

The existence of supernova remnants and strong stellar winds in giant extragalactic HII regions should be associated with the existence of gas at very high velocities (≈ 102 − 103 km s–1). The detection of large velocity/low intensity gas within the regions could be possible with the use of CCD detectors with long slit intermediate spectroscopy and long integration times. To test this observational approach we have conducted a survey in the brightest giant HII regions of M 101 and M 51. We report on the preliminary results of our programme, and discuss the implications for the energetics of the regions.

We have serendipitously discovered that the WN+O binary Br 40a in the LMC is surrounded by a remarkable high excitation nebula, showing extended, narrow emission lines of HeII. This prompted us to make a systematic search for similar high excitation nebulae around other WR stars in the Magellanic Clouds. This survey revealed a second even more extreme example surrounding the WN+O binary AB-7 in the SMC, and one other marginal detection AB-5 (SMC). The detection of nebular HeII emission implies that these WN stars emit a much harder UV spectrum than is traditionally expected. For each of these nebulae we have taken narrow band CCD images at the CTIO 0.9-m telescope, and have obtained spectrophotometry with the 2D-Frutti photon counting detector on the CTIO 1.0-m telescope.

BVRI CCD frames of NGC 604, the brightest Hll region in the nearby spiral galaxy M33, have been obtained at the Nordic Optical Telescope (Canaries Islands). The seeing was 0.6″ which corresponds to ~ 2 pc. at the distance of M33. The magnitude of the stars were obtained using the Capella stellar photometry package (Debray et ai., 1989).

Theoretical aspects of the modeling of observable evolutionary phases of massive single stars are reviewed. The SN 1987A progenitor evolution is considered in detail as an example for a star below the WR limit. Formation, structure, evolution, and mass loss of WR stars are discussed, and the impact of supernova research on stellar evolution theory is stressed.

The results of new grids of models of massive stars with metallicities Z = 0.002, 0.005, 0.020 and 0.040 and mass loss rates depending on Z are shown. When integrated over the mass spectrum, the models enable us to predict number ratios, such as WR/O, WC/WN, WNE/WR, WNL/WR, WCE/WR, WCL/WR, WO/WR as a function of Z in galaxies.

Comparisons between models and observations in galaxies are made and show, as was suggested by Maeder, Lequeux and Azzopardi (1980), that the effects of metallicity on the mass loss rates are the prime agent responsible for the different distributions of massive stars in galaxies.

Using theoretical models of WR binary evolution, and adopting bolometric corrections for the O-type companions, the B.C. of a sample of WN stars are derived using the luminosity difference and the mass ratio. From the present characteristics of observed WN binaries the initial parameters are estimated assuming conservative case B of mass transfer and stellar wind as mass modifying processes. The influence of nonconservative interaction is discussed.

A theoretical model of the ensemble of galactic Wolf-Rayet stars is constructed, assuming that all of them are members of either close or wide binaries. The model provides a reasonable explanation of the observed number of WR stars, their distribution over masses, mass ratios of components in binary systems, and spatial velocities. It predicts that up to 10 % of the apparently single WR stars have relativistic companions hidden inside thick stellar winds.

Observational evidence strongly favors an evolutionary relation between Wolf-Rayet (WR) stars and stars with masses above ~ 10 M⊙. The galactic distribution of WR stars closely resembles the distribution of O stars both in the z-direction as well as in the spiral-arm pattern (Hidayat, Admiranto, and van der Hucht 1984, van der Hucht et al. 1988, Conti and Vacca 1990). Lundström and Stenholm (1984) found that the surface density of W R stars increases with decreasing distance from OB associations as well as from young open clusters. The fraction of WR stars in clusters turns out to be about the same as that of O stars. A large number of extragalactic WR stars is situated in giant H II regions, which also contain high numbers of very massive 0 stars (see, e.g., Walborn 1990 for 30 Doradus; Drissen, Moffat, and Shara 1990 for M33). Many WR stars are members of binary systems with well established parameters so that their masses can be determined directly. In several cases, WR-stax masses of M ≥ 10 M⊙ have been derived (Massey 1981, St.-Louis et al. 1987, Schulte-Ladbeck 1989).

We observed four OB associations in the Large Magellanic Cloud which contain Wolf-Rayet stars. The associations are located in the neighbourhood of 30 Dor. They contain 15 WR stars of WCE and most WN subtypes. The observations include UBV photometry for several thousand objects and spectroscopy in the range from 3500 to 5000 Å for about 250 stars. Two new extreme Of stars and a new WR star as well as many hitherto unknown O stars were identified. Cluster HR diagrams reveal the age structure of the association stars and yield tentative conclusions about evolutionary connections between Wolf-Rayet stars in the LMC.

The physical characteristics and behavior of evolved massive stars in three different mass ranges are reviewed with application to whether they may eventually evolve to the WR stage 1. >40-50 M⊙ as LBV's, 2. ∼30-40 M⊙ as cool hypergiant-OH/IR stars and 3. ∼10-30 M⊙ as red supergiant-OH/IR stars. I emphasize the importance of the relatively short but high mass loss phases as LBV's and as OH/IR stars in determining the fate of massive stars from 10 to 100 M⊙.

AG Car is a Luminous Blue Variable which recently evolved from AIe to Ofpe/WN9 in four years at about constant bolometric luminosity, while in the visual the star faded by two magnitudes. This change is probably associated with variable opacity of an unstable massive expanding envelope of a hot star. We discuss the main spectral features of the star and of its ring nebula, and the spectral variations.

Two observational programs which provide new information about particular LBVs through investigations of their immediate surroundings are described. (1) Digital spectral classification of OB supergiants in compact groups apparently associated with Radcliffe 127 and S Doradus has revealed several interesting objects and indicates which of them are likely to be generically related to the LBVs. (2) Velocity-resolved images of the Eta Carinae shell show qualitatively new features, which will contribute substantially to the interpretation of its complex spatial/kinematical structure.

Theoretical models of supernova explosions of helium stars with various masses are reviewed to examine possible connections between Wolf-Rayet stars and Type Ib/Ic/IIb supernovae. Nucleosynthesis, Rayleigh-Taylor instabilities, and light curves are compared with observations. Maximum brightness and the fast decline of the light curves of typical SNe Ib/Ic can be well accounted for by the helium star models if the helium star mass is as low as 3-5 M⊙. These low mass helium stars can form from stars of 12-18 M⊙ after Roche-lobe overflow in close binary systems. Probably progenitors of typical SNe Ib/Ic are not classified as Wolf-Rayet stars.

I discuss evidence for and against the hypothesis that Type Ib and Type Ic supernovae (SNe) are produced by core collapse in massive, evolved progenitors. A key object is SN 1987K, whose spectroscopic classification changed from Type II to Type Ib/Ic as it aged. The progenitor of SN 1987K may well have been a massive star which experienced incomplete mass loss, leaving a thin outer envelope of hydrogen. However, several arguments are used to conclude that in most SNe Ib/Ic, the pre-supernova mass loss cannot be caused entirely by strong winds as in Wolf-Rayet stars. Mass transfer in close binary systems is probably important, but in such cases the supernova progenitor is not necessarily a Wolf-Rayet star; instead, it may be a relatively quiescent, hot, low-mass helium star that explodes via core collapse. For example, the rapid decline of the light curve of the Type Ic SN 1987M, and its seemingly low ejected mass, are consistent with this idea. It is also possible that some, but not all, SNe Ib/Ic arise from deflagrations or detonations of white dwarfs.

It is widely speculated that a Type Ib supernova is the explosion of a Wolf-Rayet star. We calculate the X-ray and gamma-ray signatures of models of that type, assuming all hard photons to have originated with Ni decay chains, in hopes of providing diagnostics of the exposed-core models of massive stars, which constitute one model of the Wolf-Rayet stars, calculated by Ensman and Woosley (1988). These provide the characteristic luminosity peak and light curve of Type Ib supernovae for helium-core masses between 4 and 6 M⊙. We compute gamma-ray line shapes and fluxes and the Comptonized X-ray continuum resulting from the decay of the radioactive 56Co and 57Co isotopes that are synthesized by the explosion of the presupernova star (the suggested Wolf-Rayet or post-Wolf-Rayet star) with a Monte Carlo transport code. The expansion velocity, the total mass of the ejecta, the radial mixing of radioactivity in that ejecta, and the 56Ni yield effect both the strength and the evolution of the hard radiation. With the anticipated launch of Gamma Ray Observatory, we can hope to detect Type Ib supernovae to distances of 3 Mpc and utilize the characteristics of the gamma lines and X-ray spectrum to distinguish between differing Type Ib supernova models and to address their suggested relationship to Wolf-Rayet stars.

The influence of wind mass loss during core hydrogen burning on the evolution of massive stars can be investigated using a relation between Ṁ-L-M-Teff based on

We calculated the evolution of both components of a massive binary system with masses 26 MO and 23.4 MO, initial period 27.4 days, through a case B of mass transfer, up to the end of the core helium burning phase of the primary (and further to the end of the main sequence of the secondary). For the WC phase we adopted Langer's (1989) mass dependent mass loss rate Ṁ ~ M2.5). The luminosities were transformed to visual magnitudes, applying bolometric corrections for O stars of Code et al. (1976) and Flower (1977) and BC = − 4.5 for the WR star (Smith and Maeder, 1989). If conservative mass transfer is adopted, the WR star is about 1.2 to 2.0 magnitude less brighter in the visual than the O star during the WN phase (lasting 1.5 105 yr). The brightness difference reaches almost 3 magnitudes during the WC + 08 phase (lasting 4.3 105 yr), showing the system as a single-lined O-type (!) spectroscopic binary. During the WR phase the mass function of this system varies from 0.67 sin3i to 0.05 sin3i! Nonconservative mass transfer and (or) a smaller initial mass ratio may result in the inverse situation, a single-lined WR binary with a large mass function (De Greve et al., 1988). Some of the results are shown in table 1 below.