The role of differential and solid body rotation in the evolution of massive stars undergoing mass loss is discussed. The implications for Of, WR, β Cephei stars and shell stars are brought out.

Evolutionary tracks for a 30 M⊙ star with mass loss rates (0.0, 1.0, 2.5, 5.0, 10.0)x10−7 M⊙/yr have been calculated. The effect of the different rates on the main sequence lifetime and on the effective temperature of the core He burning is discussed.

The rapid mass loss observed in O stars can affect their evolution dramatically if it takes place during core hydrogen burning on the main sequence. Conti has suggested that this is the case for the brightest Of stars, and that these stars evolve into Wolf-Rayet (WR) stars. Support for this scenario comes from the similarity in the spectra of the Of and WR stars, and from the observed helium enrichment of Wolf-Rayet WN stars along with the correlation of the latter with the most luminous O stars. If the most massive Of stars evolve to WN stars of relatively low mass, one needs to determine the range of zero-age stellar masses where such an evolution occurs. This communication abstracts some observational evidence that bears on the question of the minimum zero-age mass for rapid mass loss on the main sequence. It then summarizes the author's investigation of a model for mass loss where photospheric acoustic waves control the flow rate.

Three situations involving mass loss from binary systems are discussed. (1) Non-conservative mass exchange in semi-detached binaries. No quantitative estimate of this mechanism is possible at present. (2) Common envelope binaries. There are both theoretical and observational indications that this phase of evolution happens to many systems, even to some that are not very close initially (orbital periods ~ years). (3) Stellar winds in binaries. Observational evidence suggests that stellar winds from components of close binaries (especially semi-detached) are significantly stronger than from single stars at the same location in the H-R diagram. Theoretical arguments indicate that in some cases stellar wind may stabilize the component of a binary against the Roche lobe overflow. In some cases there is weak evidence of an anisotropy in the stellar wind.

A way of treatment of evolution of common envelope binaries based only on the laws of conservation of energy and angular momentum is suggested. It is shown that the final configuration depends on masses of components and initial period of the system, and on parameters describing friction in the envelope, and mass loss by the system. Possible final stages for massive binaries are either a Thorne-Zytkow type object for initially close binaries or a Wolf-Rayet star in pair with a relativistic compact remnant for wider ones. In the course of disruption of the latter system with orbital periods up to several hours very high space velocity (up to 500 km/s) pulsars can arise.

It is generally accepted that massive (and thus luminous) stars lose mass by stellar wind, driven by radiation force (Lucy and Solomon, 1970; Castor et al. 1975). For the components of massive binary systems, rotational and gravitational effects may act together with the radiation force so as to increase the mass loss rate. Our intention here is to discuss the influence of a stellar wind mass loss on the evolution of massive close binaries. During the Roche lobe overflow phase, mass and angular momentum can leave the system. Possible reasons for mass loss from the system are for example the expansion of the companion due to accretion of the material lost by the mass losing star (Kippenhahn and Meyer-Hofmeister, 1977) or the fact that due to the influence of the radiation force in luminous stars, mass will be lost over the whole surface of the star and not any longer through a possible Lagrangian point as in the case of classical Roche lobe overflow (Vanbeveren, 1978). We have therefore investigated the influence of both processes on binary evolution. Our results are applied to 5 massive X-ray binaries with a possible implication for the existence of massive Wolf Rayet stars with a very close invisible compact companion. A more extended version of this talk is published in Astronomy and Astrophysics (Vanbeveren et al. 1978; Vanbeveren and De Grève, 1978). Their results will be briefly reviewed.

In this work we calculate the evolution of a binary system with a common envelope, which consists of a blue supergiant and a neutron star. We consider as a free parameter the effectivity with which the energy liberated at the orbit produces mass loss from the system.

The evolutionary calculations were made, using various values of this parameter, for a system with mass ratio 25:1. As initial state we choose a model in the phase of Hydrogen-shell burning, before and after the begin of Helium-burning in the core.

We found that, under certain conditions, it is possible for the radius of the orbit and the period of the system to increase; the time scale for the “spiral-in” would be of the order of 104-105 years. Mass loss rates are between 10−3 M⊙/y and 10−4 M⊙/y.

The current ideas concerning the evolution of close binary systems (van den Heuvel, 1976), accepting the hypothesis that the system is not disrupted by the first supernova (SN) explosion, predict that the Wolf-Rayet phase can occur twice. The first time the companion of the WR star is a normal OB star and the second time it is a collapsed object. In this context, the importance of searching for binary systems with collapsed companions among the “single” WR stars is evident. Due to its large distance from the galactic plane, z = 280 pc (Smith, 1968a), when compared with the average height (z = 60 pc, Cruz-González et al., 1974) of extreme Population I stars, HD 50896 was considered to be a likely candidate to this type of system.

The stellar wind mass loss rates of at least some single Of type stars appear to be sufficient to remove much if not all of the hydrogen-rich envelope such that nuclear processed material is observed at the surface. This highly evolved state can then be naturally associated with classic Population I WR stars that have properties of high luminosity for their mass, helium enriched composition, and nitrogen or carbon enhanced abundances. If stellar wind mass loss is the dominant process involved in this evolutionary scenario, then stars with properties intermediate between Of and WR types should exist. The stellar parameters of luminosity, temperature, mass and composition are briefly reviewed for both types. All late WN stars so far observed are relatively luminous like Of stars, and also contain hydrogen. All early WN stars, and WC stars, are relatively faint and contain little or no hydrogen. The late WN stars seem to have the intermediate properties required if a stellar wind is the dominant mass loss mechanism that transforms an Of star to a WR type.

It is shown from various observational contraints that WN7/WN8 stars tend to occupy a class of their own set apart from other Wolf-Rayet subclasses. They probably evolved by stellar-wind mass loss from Of stars with original masses greater than ~35 m⊙, whether single or component of a binary. Two WN7/WN8 stars may be runaways in the second WR phase. One, HD 197406, is situated over 1000 pc from the galactic plane and has a low-mass, overluminous companion which is eclipsed by the WN7 component.

Much interest has been shown in recent years concerning the possible link in stellar evolutionary terms between the O-type (particularly Of) and the WR stars, arising from both observational studies of O stars (Conti 1976, Bohannon and Conti 1977) and theoretical studies of the evolution of massive hot stars with high mass loss (Chiosi and Nasi 1974, de Loore et al. 1977, 1978). These investigations and ideas coincided with a renewed attack on the abundance problem in the WR stars based on the acquisition of new ultraviolet observations and the application of recently developed techniques for treating line transfer in rapidly expanding atmospheres (Willis and Wilson 1978a). The aim of this paper is to summarize some of the salient aspects of this latter reference and to examine the evolutionary status of the WR stars in the light of these new results.

Because of the growing evidence (Conti 1976, Snow and Morton 1976, Chiosi et al. 1978, de Loore etal. 1977) that some Of stars are evolving into WN stars, we reexamine the classification of WN stars. There are four schemes for classifying WN stars, those of the IAU (1938), Hiltner and Schild (1966), Smith (1968), and Walborn (1974). All of these schemes involve estimating by eye the ratio of the nitrogen line strengths. Walborn's scheme is the most useful because he avoids problems that may arise from the use of helium lines. In other respects Walborn's scheme gives WN classifications similar to those found by Smith.

Systematic wavelength shifts of series of spectral line centers observed in many early type stars, generally interpreted as due to large scale motions, can give us information about the velocity gradients in stellar atmospheres. However, it should be borne in mind that the velocity gradients inferred from the observed displacements of spectral lines may not correspond to a unique alternative (e.g. see Karp 1978). Also, and especially when we are dealing with stars which have emission lines in their spectra, the structure of the velocity field depends on the assumed temperature structure of the atmosphere, i.e. in which atmospheric region do the lines originate.

As John Hutchings summarized at the opening of this symposium the most easily detected symptom of mass loss in hot stars is emission at Hα. Some years ago I made a survey of the Large Magellanic Cloud for stars with Hα in emission (Bohannan and Epps 1974) and since then have obtained low resolution spectra of some of these stars to establish their spectral identity. I would like to talk today of one of these objects, BE 381, which displays spectroscopic features of both the extreme Of and low excitation WN Wolf-Rayet classifications.

We consider the evolution of massive close binaries. The reader is referred to the first contribution by the same authors during this conference for the general ideas. The present contribution deals with the implications for the Wolf-Rayet phase, i.e. that part of the evolution where the primary is liable to show W-R characteristics (as usual the primary is defined as the originally more massive component). We refer to the evolution after the exchange of mass between the components, or even during the exchange phase (see section 3).