The radio flux value of the star HD 192163 (WN6) measured by Dickel et al. (1980) imposes strong restrictions on the possible mass outflow region models of this star.

A specific feature of WC stars is the presence in their spectra of strong subordinate series lines of carbon ions arising in transitions between comparatively high-lying states. The intensities of these lines suggest that carbon abundance must be comparatively high in WC star envelopes because subordinate series lines are already nearly transparent to radiation (escape probability coefficients βik are close to unity) and their intensities strongly depend on the abundance of C(Jki ∝ Nk). The resonance lines of carbon ions and the lines arising in the permitted transitions between the low-lying energy levels are in a wide abundance interval not very sensitive to abundance due to the fact that βik ≪ 1 for these transitions (Jki ∝ Nk/Ni). for 3 WC stars (HD 165763 (WC5), HD 192103 (WC8) and HD 164270 (WC9)) we carried out the same kind of model-fitting study as for the WN stars (reported earlier). We used the same empirical models, that is we assumed that also for them the wind reaches the terminal velocity at R ≃ 2 Ro (v = const for R ≥ 2 Ro) and that Te(R ≥ 2 Ro) = TIIe = 0.6 T∗, where T∗ denotes the core temperature. We took into account the ions (atoms) of He, H and C for determining the ionization structure in the envelope.

A computer program is described which calculates the structure and the radiation field (including high resolution spectra) of spherically extended and moving atmospheres. Some preliminary results for models of WN stars (L/L0= 105, M/M0= 10, Teff= 30 000 K, and M=3 · 10−5 MOy−1) are presented and discussed.

The star CPD −56°8032 is one of the few objects classified WC 11. First noted by Bidelman, MacConnell and Bond (1968), its blue-violet spectrum has been described by Cowley and Hiltner (1969), while the red/infrared region has been studied by Thackeray (1977). Infrared magnitudes have been measured by Webster and Glass (1974), and Aitken et al. (1980) have scanned the infrared spectrum from 8 to 13 µm. Cohen and Barlow (1980) have also published infrared fluxes from 3.45 to 20 µm. This star shows strong emission lines belonging to C II and C III, much sharper than the usual bands seen in WC stars. Bright lines of H, He I, O II, O III, Si III and Si IV are also present. O I λ 7774 triplet shows a P Cygni profile (Thackeray, 1977). Forbidden lines belong to [O I] (red lines), [O II] (violet and infrared doublets) [N II] (6548 and 6584 å), and [S II] (4068 and 6716 å).

In this review, three observationally accessible parameters of the winds of OB and Wolf-Rayet stars will be discussed:

1. (1) Terminal velocities

2. (2) Velocity laws

3. (3) Mass loss rates

The massive and fast winds of Wolf Rayet stars present a serious momentum problem for the line-driven wind theories that are commonly used to explain the fast winds of early type stars. It is perhaps possible for the winds to be driven by lines, if multiple scattering occurs and if there are a sufficient number of lines in the spectrum so that large fraction of the continuum is blocked by line opacity in the winds. Several other mechanisms are discussed, in particular two that rely on strong magnetic fields: a) Alfven wave driven wind models like those recently developed by Hartmann and MacGregor for late type supergiants and b) the “Fast Magnetic Rotator” model that was developed by Belcher and MacGregor for the winds from pre-main sequence stars. In either model, large magnetic fields (≈104 gauss) are required to drive the massive and fast winds of Wolf Rayet stars. Smaller fields might be possible if the multiple scattering line radiation force can be relied on to provide a final acceleration to terminal velocity.

Peter Conti has a tradition of always talking about 0-type stars at Wolf-Rayet symposia, and Wolf-Rayet stars at 0 star symposia. Since there is no well-developed theory for the origin of the winds of WR stars, it is my pleasure to join Peter's tradition, and to talk mainly about the theory of radiation driven winds in OB stars. The advantage of OB stars is that there exists a fairly complete wind theory, which agrees well with the available observations. The question is, can the mass loss observed from Wolf-Rayet stars be explained by a version of this wind theory which is scaled to the conditions found in the envelopes of Wolf-Rayet stars? The topics I consider are:

• — The calculated radiation pressure in OB stars, and its dependence on temperature, density, and chemical composition.

• — A comparison between predicted and observed mass loss rates and terminal velocities for OB stars.

• — The applicability of the standard radiation driven wind models to Wolf-Rayet stars.

• — Speculations on how Wolf-Rayet stars achieve their enormous mass loss rates within the context of the radiation pressure mechanism.

It is shown that the mass loss rate of the WN7 star HD 93162 decreases with larger values of the ratio of total to selective extinction R. for HD 93162 the mass loss rate will change one order of magnitude, only if ΔR ∼ 2. Mass loss rates are derived for nine other WR stars of which visual, red and near-infrared photometric observations were obtained.

From consideration of the observed properties of the envelopes produced by mass loss in WR stars we find that:

1. a) The velocity at the optical photosphere is in the range 200–800 km s

2. b) The effective photospheric radius for the continuous radiation capable to ionize helium twice (γ < 228 A) is typically 5 to 15 times the optical photospheric radius.

3. c) The radiation temperature in the Lyman continuum (γ < 912 Å) is around 5 × 104K. Therefore, most of the stellar radiation is emitted in the far UV and the total luminosity is considerably higher than currently estimated.

4. d) Multiple scattering (N ≃ 20) of radiation in the interval 228–504 Å can provide most of the momentum needed to accelerate the wind up to the observed terminal velocities.

Atmospheric models are presented for the outer layers of hot stars, O, of and Wolf-Rayet stars. The model is a two component hybrid model, consisting of a rapidly expanding component and a slower component. for the rapidly expanding component the energy sources are radiation pressure, a deposit of internally generated and stored energy, and radiation cooling. In this way a coronal layer with temperatures of the order of 4–7 million K is generated, with an extent of 1 to 2 stellar radii. The mass loss rates range between 10−6 and 10−4 Mo yr−1. The stellar wind velocity at infinity is of the order of 4000 km s−1. It is assumed that this rapid component interacts with the slow component, and gives rise to shocks. The corona as well as the shocks generated by the interaction of the two components can explain the observed X-rays.

The Wolf-Rayet phenomenon is defined by the striking emission line spectrum which is observed at UV, optical, and IR wavelengths. The usual interpretation is that this characteristic emission-line spectrum results from a stellar wind which has such a high density that even relatively weak lines are formed in the extended envelope formed by the outflow. This model leads to the expectation that the Wolf-Rayet phenomenon is linked to stars which have very high rates of mass loss. The purpose of this paper is to examine the relationship between the mass loss rate and the spectral appearance of O, Of, and WR stars.

In principle radio observations of WR stars offer the best possibility of determining the rate of mass loss, since for a simple model of the extended atmosphere the mass loss rate depends primarily on quantities—the flux density, the velocity, and the distance—which are observable (Barlow 1979). Until now, detections of Wolf-Rayet stars have been limited by the sensitivity and resolution of available telescopes. The advent of the Very Large Array makes a search for emission from a large number of these stars feasible.

The combination of IR data with radio, optical and UV observations has the potential to greatly increase our understanding of the WR phenomena. Consequently we are obtaining IR spectra from 1.1 to 4.2µ for a sample of southern WN stars (K < 7th magnitude) using a low resolution, R∼100, circular variable filter wheel (CVF) and a cooled grating spectrometer, (CGS) (Jones et al. 1981) with a maximum resolution of 2000. Previous IR spectral studies of WR stars have generally concentrated on line identification, and beyond 1.7µ they have been limited to low resolution (Barnes et al. 1974, Cohen et al. 1978, Williams et al. 1980). With the CGS it is now possible to obtain spectra at resolutions sufficiently high to split blends and reveal details of line profiles. Below we present preliminary results of our observations.

We report here on preliminary infrared photometry of six WN stars in the Large Magellanic Cloud. Besides R136, the central object of the 30 Doradus Nebula, the sample includes three stars (R139, R140, R145) located near the center of the region (within ∼ 1 arcmin) and two more stars (R144, R147) at a distance of ∼ 5 arcmin from R136.

The observations reported here have been obtained with the ESO Reticon system attached to the Boller and Chivens spectrograph at the 3.6 m telescope. The reciprocal dispersion was 228 A mm−1 with a useful spectral range extending from 5800 A to 11000 A. The slit was 200 µm wide which corresponds to 1.4 arcsec on the sky and to about 33.5 µm(∼ 7.6 A) in the plane of the detector. Sky substraction was performed by means of observations at 93 arcsec. East and West of the object. The correction for atmospheric extinction was based on a mean extinction curve of La Silla (Tug, 1977). As a consequence the strongest atmospheric bands are still present on the final spectra. The data were also corrected for the instrument sensitivity response by using calibrated stars selected from Breger's catalog (1976). As these reference data are from filter photometry, they are not well suited to the resolution of our spectra, at least in some regions. This problem can be significant, as clearly illustrated by Johnson (1980).

Ultraviolet spectra from the “Copernicus” satellite of the Wolf-Rayet star HD 50896 (Johnson, 1978; Cassinelli and Rumpl, 1982) show several P-Cygni profiles whose emission component has a larger equivalent width than the absorption component (see Figure 1). for lines with saturated absorption components, such as the N V (λλ1239–1243) line, the excess emission is due to thermal emission from the wind. However, for unsaturated lines which are due to scattering, such as the P V (λλ1118–1128) and O VI (λλ1032–1038) lines, the excess emission suggests that the scattering is not spherically symmetric.

The study of Wolf-Rayet binaries is important for the information we can gleam about the most fundamental property of any star: its mass. In this regard we are very fortunate, since the masses of WR stars can be determined with far greater confidence than can their luminosities, compositions, colors, effective temperatures, etc. By simply measuring the velocities of both components in a WR+0 system, we can find the minimum masses and the mass ratio of the two stars; if we can get some further handle on the orbital inclination, we then know the mass of the WR star directly.

Evidence is accumulating in favour of a heretofore missing link in the evolution of massive binary stars — a second WR phase in which the companion to the WR star is a neutron star of mass 1–2 M⊙, or in some cases a more massive black hole. The discovery of such objects is also important in assessing the role of duplicity in the formation of WR stars.

High resolution (Δλ ≈ 0.1å) IUE spectra have been obtained of the two WC9 stars HD 164270 and HD 136488, covering the wavelength range λλ1150–2050. The former star shows P Cygni profiles indicating a stellar wind terminal velocity of v∞ ≈ −1400 km s−1, and the latter v∞ ≈ −1800 km s−1. A common feature in the spectra of both stars is narrow displaced absorptions due to Fe III (UV34) transitions arising from a metastable lower level. These features are displaced at subterminal velocities (−830 km s−1 for HD 164270 and −1030 km s−1 for Hd 136488) and are believed to be formed in the deceleration region of their stellar winds. The properties of these inferred Fe III circumstellar shells derived from these data are discussed.

We have analysed new spectroscopic observations of 22 population I WR stars in a sample defined by b<12.5 mag, 18h<α<6h, δ > −30°. This completes the search for orbital duplicity among all of the 30 pop I WR stars contained within these limits. We thus have an unbiased data base in a volume segment bounded by a distance of 3–4 kpc for WNE and WC stars , and 4–5 kpc for WNL stars Although more observations are still needed in some cases, we have found several new binary systems. Some of these are low-amplitude, single-line binaries, often with runaway velocities and/or large separations from the galactic plane. We consider them to be candidates for a WR + collapsar stage, as originally predicted by Tutukov and Yungelson (1973) and de Loore and De Grève (1975).