The gravitational fragmentation theory of star formation is reviewed. Theoretical arguments are presented which suggest that the lower stellar mass cut-off to the IMF in giant HII regions may be as high as 10 M ⊙. Mechanisms for bimodal star formation are described in the context of a coagulation model for formation of the giant molecular clouds, and application is made to starbursting galaxies.

The far reaching consequences of convective overshooting during the core H and He-burning phases of stars in the mass range 1.3 M⊙ to 100 M⊙ are discussed. In addition to this, the effects of mass loss in luminous stars of all spectral types, and in the red giant and asymptotic giant branch stars are briefly outlined. Furthermore, the effects of the novel 12C(α γ)16O reaction rate are also illustrated. The main purpose of this review resides however in lending convincing support to the idea that convective cores of real stars are greater than commonly supposed in classic models. To this aim, several observational embarrassments that could not be explained by classic models are reanalyzed in the light of the new ones. Since a much better agreement between theory and observations is now possible, we are inclined to conclude that convective overshooting may be of paramount importance in stellar structure theories and that convective cores in real stars ought to be larger by approximately one pressure scale height than predicted by classic models.

The importance of close binaries in the domain of luminous objects is examined. The group of unevolved massive stars of O- and B-type, the groups of Wolf-Rayet binaries and X-ray binaries are discussed, and limits on the mass ratios, masses and periods are listed, as well as the average values. The connections between these various groups are investigated, and relations among subgroups are established.

Recent investigations of the stellar content of giant H II regions and star formation bursts in nearby galaxies are summarized. The preliminary results of a spectroscopic survey of 78 H II regions in 15 galaxies are presented. A minimum frequency of 30 percent is found for the occurrence of luminous Wolf-Rayet stars in such bursts of star formation. Observations of the positions of historical type II supernovae revealed the presence of WC stars, indicating very high ZAMS masses for the SN progenitors. The application of population synthesis models to the analysis of the integrated stellar spectra in terms of IMF parameters is discussed. An example of such an analysis for a giant H II region in Cen A is presented.

This paper begins with an attempt to examine the problem of identifying stellar associations in galaxies in a consistent way, so that meaningful physical comparisons can be made for the population of stellar associations of different galaxies. A compilation of the existing data on associations in other galaxies is given and their properties compared. Questions relating to star formation in stellar associations are discussed, and then the issue of the initial mass function of core clusters, especially those located in giant HII regions, is briefly examined.

Massive X-ray binaries may be powered by two mechanisms generating X-radiation: accretion of stellar wind material of the O-star, or Roche lobe overflow (RLOF). The evolution of RLOF powered X-ray binaries has been studied by Savonije (1978). For massive binaries the duration of the RLOF powered stage is less than 100 years for binaries evolving through case B of mass transfer, and 5000 to 10 000 years for case A. On this basis Savonije concluded that the majority of the X-ray binaries should evolve through case A of mass transfer, in order to explain the observed number of active sources in massive binaries. In this study we compute the transition from compact+P systems into X-ray binaries, with emphasis on the wind powered stage.

We have examined further the consequences of mixing induced by rotation in massive stars (Pop I composition: X = 0.7; Y = 0.27) in the mass range 40–100 M⊙. The basic physics and a detailed examination for a 60 M⊙ model has recently been published (Sreenivasan and Wilson 1985b), as has a detailed examination of supergiants without such mixing (Sreenivasan and Wilson 1985a). We shall therefore not go into the background of our computation here, but merely summarize our present findings.

About twenty years ago van den Bergh (1964) recognized 188 OB associations in the Andromeda Nebula. He used plates taken by the 52-inches Tautenburgh Schmidt telescope in GDR. Later on Richter (1971) added 7 new associations in the south-western periphery of M31. Now we have the opportunity to continue the search of stellar associations in M31 with the 2m Ritchey-Chrétien (RC) telescope of the Bulgarian National Astronomical Observatory. The limiting magnitude of this telescope is approximately the same as that of the Tautenburg 2m Schmidt telescope but it possesses somewhat smaller field (1°x1°) with plate-scale 12.8 mm−1. That is why a new search of stellar associations in M31 by means of the 2m RC telescope is very efficient. Indeed, some previous inspections of the RC plates (Efremov, 1982) indicated some new associations in M31.

The associations OB 110 and OB 112 in M 33 galaxy (Humphreys and Sandage, 1980) form a powerful group of young stars. This rich star cloud is situated far (∼6 kpc) from M 33 nucleus. The abundance of hot stars and its huge dimensions (900 × 1500 pc) make it similar to the association OB 78 (400 × 800 pc) in the Andromeda galaxy. We use distance modulus (m-M)v = 25.35 given by Sandage and Carlson (1983) and EB-V = 0.03.

The evolution of expanding supershells produced by SN explosion in OB associations is considered. The expansion velocities of the shells and OB associations IMF spectral index are obtained. The results agree well with observational data.

Star formation in the dust lane of NGC 5128 (Cen A) has been studied using digitized CTIO 4m plates. The digital images were calibrated to the standard U, B and V passbands using photoelectric photometry (van den Bergh 1976). Ages for the dominant stellar component were derived by de-reddening each pixel along an R = 3.3 reddening law to its intercept with a theoretical cluster evolution track in the U-B, B-V plane (Davis 1979). The HII regions of Hodge and Kennicutt (1983) were used as tracers for the most recent star formation.

As a consequence of positive feedback effects in interstellar chemical and star forming processes, a star formation region may undergo nonlinear oscillations.

The evolution of massive close binaries is altered if the effects of mass loss through a stellar wind and overshooting from the convective core are taken into account. The occurrence of mass transfer as well as the extent of the mass transfer itself differs from the classical case (Doom and De Greve 1983, Doom 1984). The main-sequence widening due to the enlargement of the convective core results in an enhancement of the frequency of case A of mass transfer. It occurs for initial periods much longer than in the classical case. A cosiderable number of early-type systems will therefore undergo this type of mass transfer (Sybesma, 1985b). Systems with primary masses larger than 35 M will most likely not undergo mass transfer at all as these stars do not form red supergiants if overshooting is coupled to the effects of mass loss through stellar wind. These systems will only be able to undergo case A of mass transfer, and then only for initial periods below 2–3 days.

The nebular and stellar observations were carried out by G. T. at the Cassegrain focus of the Danish 1.50m telescope at the ESO La Silla Observatory in August and September 1984, using a RCA CCD (512 × 320 pixels) with a pixel size of 0.47”. The frames were reduced using the VAX/MIDAS software at ESO Garching and Meudon Observatory (INVENTORY automatic program).

“Wolf-Rayet stars in the LMC: how faint do they go?” That was the title of a recent paper by Massey and Conti (1983) who found that the faintest WN star in Breysacher's (1981) catalogue of LMC WR stars has mv = 17.1. The faintest WC star in that catalogue is Br74 which has mv = 15.6. We have now detected two faint WC stars at mv 18.6, three magnitudes fainter than Br74.

The bright stars for 15 SMC clusters were classified in order to derive the distribution of various spectral types. The studied clusters represent all evolutionary ages (disk, intermediate and halo) and are located at various places of the parent galaxy. The spectal classification of the stars was carried out using film copies of the 1.2 m Schmidt telescope objective prism plates. Low dispersion (2440 Å at Hγ) and medium dispersion (830 Å at Hγ) unwidewed UJ and RI spectra were examined by means of a binocular microscope. Short exposure plates were used as well for the most bright stars and particularly for the stars at the central areas where crowding is more severe. More details about the used material and the criteria used for the classification are described by Kontizas et al (1985). For each cluster a circular area was examined inside its tidal radius. (Kontizas, 1984). The stars in the innermost part of the populous clusters were not classified because of the overlapped images. Stars of fields in the vicinity of each cluster were also classified to find the contribution of field stars in the cluster area. The magnitude range of the studied stars is 14.5<V<17.50.

Masses and tidal radii of star clusters in the LMC have been derived by means of star counts from U.K. Schmidt plates. Two groups of clusters according to their distance from the rotation centre of the LMC were measured. The tidal radii of the central clusters vary from 58 to 85 pc and those of the most distant clusters from 33 to 86 pc whereas masses were found to vary from 105 to 4×105M⊙ and from 104 to 2×105M⊙ respectively.

Star counts can be used to investigate radial distribution of stars of different mass. Relaxation through stellar encounters is a mechanism that does make a distrinction between stellar masses, so systems that have undergone such relaxation should show differences in distribution between stars of high and low mass. That does not happen for systems that have undergone an initial violent relaxation since this type of relaxation treats all masses equally.

A supernova explosion in the galactic disk creates a hot, low density cavity which persists much longer than typical lifetimes of supernova remnants. McCray and Snow (1979), Bruhweiler et al. (1980), Tomisaka, Habe and Ikeuchi (1980) and Kafatos et al. (1980) pointed out that repeated supernovae from a stellar association will produce an expanding shell of gas (R ≳ 100 pc) and offered extensive evidence for such shells -primarily in the form of H I shells.

The massive stars of the Magellanic Clouds are of considerable current interest with regard to questions of initial mass function, star formation mechanisms, stellar evolution with mass loss and the chemical evolution of galaxies. The effective temperatures, surface gravities and helium abundances of 6 main sequence O-type stars, obtained by fitting non-LTE model atmospheres to high quality spectra, are presented here; these are the first results from a long-term program to determine accurately the parameters and chemical abundances of massive stars in the Magellanic Clouds. The program stars were selected to be main sequence objects, according to the classification of Conti et al. (1985, in prep.), with He II λ 4686 Å in absorption, and to have minimal reddening and nebular emission. Spectra were obtained in 1984 December with the Cassegrain echelle spectrograph (CASPEC) and a CCD detector at the ESO 3.6 m telescope. A preliminary analysis of these spectra has been carried out by fitting the equivalent widths of He I λ 4471 Å and the profiles of Hγ and the Pickering lines (for details of the technique, see Kudritzki, 1980). The resulting values of effective temperature and gravity are given in Table I, along with the identification, spectral type and mv, of each star.