Wolf-Rayet stars could be rapid rotators. We must ask if a rotationally flattened wind can lead to an observed continuous spectrum that depends on the angle of inclination. Schmid-Burgk (1982) showed that in the radio region the spectral shape is not changed. Here we consider the optical region, where both thermal emission and electron scattering are important. This study uses a simple, idealized model of a rotating wind in which v(r) is from Friend and Abbott (1986), but depends on polar angle in the manner of Poe et al. (1989). Equatorial densities are five times polar values at the same radial distance. The wind consists only of fully ionized He surrounding a spherical core and has a radial optical depth of 2.0 in the equator in the limit of pure electron opacity, essentially at wavelengths less than 2000A. Radiative transfer in the wind is fully treated by a second-order moment method based on the equation of transfer in general spherical coordinates (Doherty 1989). A constant flux at the core boundary is assumed. This flux corresponds to a T = 4×104K blackbody and the wind temperature is 2×104 . The shape of the emergent spectrum in the region computed (2500—7500A) changes little with inclination and resembles the spectrum of a spherical wind with intermediate density. Thus a rotating wind may masquerade as spherical but, in the present model at least, vary in flux by nearly a factor of two from pole to equator. This can affect the apparent magnitude, wind polarization, and possibly the line spectrum and wind acceleration.

We present 1 – 3.4 μm spectra of six Wolf-Rayet stars: WR 146 (WC4), WR 111 (WC5), WR 86 (WC7), WR 140 (WC7+O4), WR 135 (WC8) and WR 88 (WC9). Examination of the relative strengths of the emission lines has enabled us to make over 20 new line identifications. Recombination analyses of the carbon and helium lines yield C/He abundance ratios much in excess of the solar value and correlated with spectral subtype.

Hydrogen abundances in WN stars have been derived by Conti et al. (1983) from a “semiquantitative” study of the Balmer-Pickering decrement. No clear correlation between hydrogen detection and spectral subtype could be established: stars with hydrogen and stars without detectable hydrogen are found among both, the WNE (“early”) as well as the WNL (”late”).

High and low resolution UV and optical spectra of the four Population I WO stars originally classified by Barlow and Hummer (1982), Sanduleak 1, 2, 4 and 5, have been analyzed. Reddenings, terminal velocities and the relative abundances of He2+, C4+ and O6+ have been determined. The results are presented in Table 1.

A strong diffuse C IV (5801-12 A) emission has been discovered in Berkeley 87 cluster near to the WR (WO) star ST3 (Barlow and Hummer, 1982) using the Loiano 152 cm telescope equipped with a Boiler and Chivens 26767 spectrograph and a CCD RCA camera. The diffuse emission seems to interest all the core of the cluster Berkeley 87. In fact it is present around many cluster members sited near to the center.

Hitherto our quantitative analyses of WR spectra [2][5] have been based on pure-helium models [1][6]. Now we further improved our non-LTE calculations by including a complex model atom of nitrogen (90 energy levels, 351 line transitions; with low-temperature dielectronic recombination) into our model atmospheres in order to synthesize adequately the spectra of WN subtypes (Wessolowski et al., in preparation). Together with the nitrogen (the most important “metal” in WN atmospheres), we introduced an improved temperature structure into our model calculations [3], now accounting for non-grey radiative equilibrium instead of the grey approximation applied so far. Moreover we took into account the line overlap of the considered elements (here: helium, nitrogen) and also their blanketing effects on the continuous radiation field.

Evidence is mounting that the dominant random component of variability in single WR stars can be explained by one common phenomenon: stochastic formation, propagation and decay of density enhancements in the winds.

Many Wolf-Rayet stars display variations which are observable either spectroscopically or photometrically or both. These are not exclusively associated with binary systems. A study is made of the sporadic events in the context of extended atmospheric structure.

Twenty - two Wolf-Rayet stars (12 of type WN and 10 of type UC) were observed in the years 1986-1990 with the VBLUW photometer of Walraven. Eight (WC and WN) objects appeared to be constant. Five of the in total 14 variable objects will be discussed in the present paper: UR46 (WN3 pec), WR50 (WC6+a), WR55 (WN8), WR86 (WC7) and WR123 (WN8).

Simultaneous spectroscopy is made for WR46 and WR50. These results will be also discussed in short.

Time-resolved spectroscopy has revealed small, systematically varying features superimposed on the broad emission lines of WR stars. We believe that these structures are due to inhomogeneities of emitting wind material propagating with the general wind. Here we suggest a common, but still unknown, origin for the intrinsic variations observed in WR stars.

The winds of the hot, luminous, O, B, and WR stars are driven by the line-scattering of the star's continuum radiation flux. Several kinds of observational evidence indicate that such winds are highly structured and variable, and it seems likely that a root cause of this variability is the known strong instability of the line-driving mechanism. Initial dynamical models of the nonlinear evolution of this instability confirm that the wind indeed becomes highly structured, with large amplitude (~500 – 1000 km/s) shocks that separate high-speed rarefied flow from lower speed, dense shells. Remarkably, such variability can often have an intrinsic character, persisting even in the absence of explicit perturbation, and it now appears that this is a direct consequence of a degeneracy in the steady-state solutions for such models. However, recent work indicates that including scattering effects, which have so far been ignored in these pure-absorption models, might reduce or even break this steady-state solution degeneracy; through the “line-drag” effect, scattering can also reduce the strength of the instability, possibly rendering it an advective character for which wind variability now requires explicit perturbation from below. This review will examine the consequences of these ideas for understanding the likely nature of wind variability among the various kinds of early-type stars.

The instabilities of WR stars are discussed in the framework of both the quasi-adiabatic and non-adiabatic methods. We show that a very careful treatment of stellar opacities is necessary. Both methods show that WR stars, with little or no H left, are vibrationally unstable in the fundamental radial mode due to the ε-mechanism, confirming earlier results by Maeder (1985). Thus, from the point of view of stability, WR stars are fundametally different from O stars. Our new analysis also supports our previous view that WR stars evolve, keeping at the edge of vibrational instability.

The phase dependent profile variations due to wind eclipses in WR + O binary systems are shown to be a means of establishing the WR wind structure. The results of two model calculations are presented which indicate that there are important qualitative differences between the profile variatons which will be observed in slowly -and rapidly- accelerating winds.

The WN 5 star HD 50896 has been the subject of many variability studies in the optical. A period of 3.766 days is frequently associated with this variability although it is by no means clear whether it reflects, for example, binary motions or rotation.

EZ CMa (HD 50896, WN5) is an enigmatic object. New photometry and polarimetry of EZ CMa are presented in the figure. Again the 3.77 day period is found but, as observed at previous epochs (e.g. Drissen et ai. 1989, Ap. J., 343, 426), the shapes of the curves change. The new photometry can also be interpreted in terms of a shorter period, of 1.254 days. A period of about one day is also claimed in other sets of photometric data (e.g. van der Hucht et ai., 1990, A. A., 228, 108) and in the IUE spectra of St.-Louis et ai. (1990, this symposium). However, despite the complex nature of the light curve, the 3.77 day period is strongly supported by the polarimetry, which shows no evidence for the shorter period.

In this paper, we present results of an intensive search for periodic variations in the strength of the Hell emission line of �2 Vel relative to the nearby continuum. The data were obtained in 1986 and 1988 using an automated optical system located at the Amundsen-Scott South Pole Station. Ten of the data sets have a time resolution between 2 and 3 minutes and consist of 2.3 - 20.4 hours of continuous photometry (Taylor, AJ, in press). Three of the data sets have a time resolution of about 1 minute and span between 1.2 and 6.5 hours.

Recent observational efforts have presented new results concerning (linear) polarization variations as a function of time as well as wavelength that have considerable impact on WR star models:

Polarimetry throughout orbital phase has provided inclinations of double-line spectroscopic binaries containing WR stars with O star companions, thus improving our understanding of the masses and evolution of WR binaries (Schulte-Ladbeck 1989, A.J. 97, 1471). New binary-model calculations were presented at the meeting and continuing comparison with observations should prove productive.

Light-traces of a light bulb fixed on top of the stem of a spinning top were recorded to simulate rotation. It was found that with dissipation of energy the rotational frequency did not decrease monotonically, but showed ‘quasi-periodic’ sequences.