It was discovered in the early years of the Copernicus satellite operations that the stellar winds in early-type stars are often characterized by degrees of ionization inconsistent with the spectral types of the stars. The most notable early example was the discovery of strong OVI lines in the form of P Cygni profiles in stars too cool to have this ion in their photospheres, such as τ SCO (BOV; Rogerson and Lamers 1975). Systematic studies showed that excess ionization is a general property of the winds in OB stars (e.g. Lamers and Snow 1978).

Fell lines are frequently observed in the optical spectra of early type emission line stars. Space observations of their ultraviolet spectra led to the discovery of a considerable line absorption due to ionized metal lines formed in the stellar envelopes. We describe some examples and show that spectral synthesis is required especially for the analysis of the low resolution spectra, and may provide information about the structure of the envelopes and the mass loss rates. We also discuss the Felll lines in the UV spectra of a Herbig Be star and of β Lyrae.

We present high resolution IUE ultraviolet observations of multiple narrow absorption components seen in the SiIII, NV, SiIV, and CIV profiles of the B1 IIIe star, HD110432. Spectra taken during March 1986, spanning ∼ 11 days, are modelled using line profile fits. Central velocities and column densities of the discrete features are derived.

The 11 observations obtained during March 1986 (SWP24923–28031) exhibit only minimal line profile variability, but are substantially different from the only other available IUE spectrum of HD110432, taken in April 1981 (fig. 1). This previous image reveals just a single extremely narrow (FWHM ∼ 50 km/s) feature at ∼ 1360 km/s.

We present the results of a survey of stellar winds in 62 Be and 43 normal B stars in spectral types B0 to B5 and luminosity classes V-III. We find that in general the wind absorption seen in the resonance lines of C IV and Si IV in Be stars is the result of the superposed absorption from multiple shortward-shifted discrete absorption components. These components are observed in 63% of our Be stars, and are not observed in the normal B stars, confirming the result of Henrichs (1984).

High-resolution spectra of the Be star HD 174237 taken on September 9, 1980 will be presented. They show stellar absorption lines and superionized NV lines shifted redward equivalent to a velocity of 20 km s-1. This velocity when compared to the earlier measured velocity of - 20 km s-1 (cf. Hoffleit) suggests that the Be star HD 174237 is really a binary and the IUE and previous observations are in different phases of binary motion. The spectra also show strong A1III and SilV absorption lines together with strong NV lines but surprisingly the CIV lines are very weak. It appears that the A1III doublet is associated with a narrow blue-shifted component suggesting an outflow of cool gas from the star.

The initial results of a program to investigate the far-UV (500-1700 A) spectra of Be stars using the Voyager ultraviolet spectrometers axe presented. Observations of eleven Be stars, with spectral types from B0 through B7, obtained between March 1979 and July 1984 are discussed. The Voyager resolution (∼15 A) allows only limited spectroscopic studies; shortward of 1150 A C III λ977 and LyJβ λ1025 are the most prominent spectral features. Two stars, γ Cas and 59 Cyg, showed enhanced C III λ977 absorption. The flux distributions of the Be stars were intercompared with those of reference stars with low projected rotational velocities and, in a parallel study, to predictions from rotating star models. The model calculations suggest that the observable effects of a high rotational velocity (V≳ 0.9 of critical velocity) on the flux distribution are small except in the FUV (λ < 1150A). The Voyager flux distributions were found to be indistinguishable from those of the low v sin i reference stars. No evidence was found for critical rotation in Be stars. All observations were consistent with Be stars rotating at less than 0.85 of critical velocity.

Zeta Tau (HD 37202) is a well known Be/shell star of spectral type B1 IVe and vsin(i) = 220 km/sec (Slettebak 1982). Its visual and UV variability have been studied extensively (Heap 1975; Hubert-Delplace and van der Hucht 1978; Hubert-Delplace et al. 1983; Dawanas and Hirata 1984; Harmanec 1984). Zeta Tau has also been found to be a binary with an orbital period of 132.97 days (Harmanec 1984). Irregular light variations have been observed (Hoffleit 1982) and long term variations not associated with the 132.97 day period have also been seen (Hubert-Delplace et al. 1983).

Optical light variability appears to be commonplace in Be stars, on time scales from days to decades, but the phenomenon has yet to be satisfactorily explained. Some possible causes include changes in 1) the star's photospheric temperature or radius, or 2) the envelope's opacity or continuous emission. To investigate the nature of the photometric variability in Be stars, we have been monitoring the FUV flux of selected objects with the ultraviolet spectrometers (UVS) on the Voyager 1 and 2 spacecrafts (cf. Broadfoot, et al. 1977, and Holberg, et al. 1982, respectively, for information on the instrumentation and calibration, and Drilling, et al. 1984 and Polidan, et al. 1987 for a discussion of data reduction techniques), and report the initial results from these observations in this paper.

In this paper we report preliminary results on the first search for short term photometric variations in the far-UV in Be stars. Results for mu Cen and 28 Cyg indicate no significant variations.

Studies of line profile variability of the ultraviolet 1550 Angstrom resonance transitions of C IV in Be stars (Sonneborn et al. 1986; Grady, et al. 1986a,b) have prompted an investigation into the short- and long-term behavior of the C IV lines in other types of B stars. We present examples of two well-studied Be stars, Omega Orionis and 66 Ophiuchi, and two non-Be stars, Beta Cephei and the standard star Zeta Cassiopeiae. Zeta Cas is also known to be a 53 Per variable (see Cox 1983). The IUE SWP high-dispersion spectra of Beta Cep and Zeta Cas have been obtained from the IUE archives. It has been known for some years that the C IV line profiles in Beta Cep vary in a time scale of several days (Fishel and Sparks, 1980). However, it came as a surprise to discover C IV variability in Zeta Cas. Available data allow us to set an upper limit of several months for the time scale of Zeta Cas C IV variability. The principal difference between the C IV variability in Be and non-Be stars appears to be the magnitude and velocity range of the effect.

The spectral type of HD 166596 (HR 6804) is given as B2.5III in The Bright Star Catalogue (Hoffleit and Jaschek 1982) and as B2HIII by Houk (1978). Its rotational velocity, Vsin i, is 255 km s-1 (Uesugi and Fukuda 1982). Neubauer (1930) noted that its radial velocity is probably variable. Bidelman and MacConnell (1973) designated this star as a silicon star, and thereafter, Renson and Manfroid (1980) found its photometric period, P = 1.67 ± 0.01 days (or 0.83 days) with the amplitudes of 0ṃ04 ∼ 0ṃ05 from their four-color photometry. To the best of our knowledge, there is no report on the presence of any emission line in the foregoing literature.

Most evidence on X-ray emission from the vicinity of Be stars concerns the Be/X-ray binaries. Presently some 20 of these systems are known, making them the most numerous class of massive X-ray binaries. Evidence for the binary nature of these systems comes from (i) Doppler modulation of X-ray pulse periods, (ii) periodic X-ray flaring behavior, and (iii) correlated optical and X-ray variability. The correlation between X-ray pulse period and orbital period found by Corbet (1984) can potentially provide important information on the densities and velocities in the circumstellar disks of Be stars.

Evolutionary models indicate that the Be/X-ray binaries represent a later stage in the evolution of normal close binaries with initial primary masses predominantly in the the range 8 to 15 M⊙ . These models indicate that also a class of slightly less massive Be star binaries should exist in which the compact companions are white dwarfs. Be-type blue stragglers in galactic clusters may be such systems.

The principal aim is to report on the sudden appearance and disappearance of (mainly) Fe II emission lines in the Be/X-ray system HDE 24-5770/ A0535+26, about 24 days before an X-ray outburst was observed. The system is characterized as an 09-7 Ille star of mass 20-25 M⊙, accompanied by a neutron star (X-ray pulse period ≃104 s), with orbital period -110 d, eccentricity > 0.3; for a more complete description see Giovannelli et al. (1985) or de Loore et al. (1984). X-ray outbursts have been observed recurrently, obviously connected with periastron passage of the neutron star triggering gas infall onto it. The observed X-ray outbursts are of varied strength, from barely detectabl to 3 Crab units at peak (Nagase et al. (1982), Giovannelli et al. (1984)). A weak X-ray event took place on December 13, 1981.

Results of programs to monitor Ha emission line variability in two Be star X-ray binaries are presented. These systems provide a means of investigating the influence of a binary companion on the circumstellar envelope of a Be star.

The rotationally-enhanced stellar wind model for Be stars represents one attempt to understand many of the phenomena related to Be stars in terms of a stellar wind distorted and perhaps also enhanced by rapid stellar rotation. This review will concentrate exclusively on this particular approach; the current status of other attempts to model the circumstellar envelopes (CE) of Be stars are described in other reviews in this volume. It is assumed that Be stars are single stars; if a member of a multiple system, the separation of all components from the Be star is assumed to be sufficiently large that interactions due to mass transfer from any companion to the Be star are negligible.

A general binary hypothesis of the emission-line phenomenon in early-type stars was published eleven years ago (Kříž & Harmanec 1975, Harmanec & Kříž 1976; for a reformulation see Harmanec 1982). Since then, new observational techniques and theoretical concepts have led to a number of exciting findings which have made the emission-line phenomenon even more puzzling and challenging than before. It is therefore of interest to ask whether the binary hypothesis can withstand all these developments or not.

The observational status of the ubiquitous rapid variability of Be stars is summarized. The most comprehensive interpretation is obtained with a traveling velocity field and associated temperature variations. But neither the available observations nor theoretical predictions presently allow an unambiguous mode determination of these nonradial pulsations. In addition to rapid rotation, an NRP mode of low azimuthal order, m ≈ 2, seems another prerequisite for a B star becoming a Be star. The amplitudes of these modes are variable and have been observed to decrease with some delay after an outburst. For low-order sectorial modes, NRP's therefore have the potential of explaining the three oldest unsolved problems of Be stars: the difference between Be and Bn stars (the latter do not usually have a low-order mode), the episodic component of the mass loss from Be stars, and the equatorial density maximum of their winds. Mechanisms for the transfer of energy to the atmosphere and its transformation into the kinetic energy of an outburst are discussed.

The proposed model is empirical; it is based on analysis of the available data on Be stars obtained in all the observable spectral regions, and it is required to be thermodynamically self-consistent. Rather than trying to answer the question: “What is the origin of the Be-phenomenon?” We ask: First. “What phenomena characterize empirically Be stars?” Second, “What thermodynamic characteristics are implied by the existence of such phenomena?” Third. “What inferences may be made an the atmospheric structure of a Be star from these empirical and thermodynamical characteristics?” The observed phenomena, their thermodynamical implications and the resulting model have each two aspects. On the one hand, the observational evidence for a nonradiatively heated, expanding chromosphere-corona implies the existence of both a nonradiative energy flux and a mass outflow from the photosphere. This first aspect is common to both Be and normal B stars, at least for the earliest subtypes.

The magnetic-loop model for a stellar mantle is described and it is shown how this model may be used to understand the differences in spectrum and change of spectrum which are observed for main-sequence stars haying effective.temperatures in the range from 2.5 x 104 to 3.0 x 104 K.

It is suggested that the difference in spectrum between a shell star such as Gamma Cas when it is in an active state and a Wolf-Rayet star may be due to a difference in environment.

Profiles of emission lines contain key information on the dynamics, the physical condition, and the geometry of the cool envelopes of Be stars. Despite much observational evidence, doubts persist about the rotating, equatorial ring model first proposed by Struve (1931) to explain the gross features of the profiles and later developed by Huang (1972) in the two-dimensional, transparent limit. The basic rotating model is now reconsidered and profile formation studied from a physical point of view.

The excitation structure of a disk rotating around a B star in circular, Keplerian orbits is computed by means of a new, iterative approach. The approach takes full account of both stellar and diffuse radiation at line and continuum frequencies and makes no assumptions about the optical properties of the disk. It is applied to an isothermal model with parameters based on observations of the shell star 48 Librae.