Preliminary calculations of the absorption profiles produced if multiple shells are ejected from the photosphere into a non-monotonic wind, suggest that this geometrical model can reproduce a wide variety of simple and complex profiles observed in the discrete components in OB and Be stars.

Spherical NLTE-model photospheres surrounded by envelopes are calculated in order to interpret classical Be-stars. The existence of the Hll-region can be understood by such models. The dependence of the calculated spectra on parameters corresponds to the observations. Applications to special objects show good agreement between the predicted and observed spectra in the visual and infrared wavelengh range.

With their rotating, magnetic, radiation-driven wind model, Friend & MacGregor (1984) found that rapid rotation and an open magnetic field could enhance the mass loss rate (ṁ) and terminal velocity (V∞) in an 0 star wind. The purpose of this paper is to see if this model could help explain the winds from Be stars. The following features of Be star winds need to be explained: 1) Be stars exhibit linear polarization (Coyne & McLean 1982), indicating an enhanced equatorial density. 2) There appears to be enhanced mass loss (at low velocity) in the equatorial plane, from IRAS observations of Waters (1986). 3) The width of the broad Balmer emission lines remains unexplained.

KX And (HD 218:193) is a peculiar interacting binary with the period of about 38.9 days. This study is a first attempt to model the system as a double-line binary. The main results of the He I - line analysis are summarized.

The intensity tracings of sixty blue-violet spectra were used. All the spectra were obtained in the coudé focus of the Ondřejov 2m telescope in the period 1972 – 1981. The original dispersion was 8 or 16 A mm -l. The phases of the observations were computed according to the ephemeris derived from the radial velocities of optical shell lines (Štefl et al. in preparation). Phase zero corresponds to the velocity maximum.

CX Draconis /HD 174237, B2.5Ve, v.sin i = 160 km.s-1/ is a typical Be star displaying most of the well known spectroscopic and photometric characteristics of these stars.

Koubský /1978/ disclosed that CX Dra is a SB1 /P=6.69603d, K = 35 km.s-1, e = 0/. Penrod /1985/ confirmed that the secondary is a F star. Peters and Polidan /1984/ reported strong, variable lines of NV, CIV and SilV from high-resolution spectra secured with IUE satellite. These features were interpreted to arise from gas stream and accretion close to the photosphere of the B-component of the system.

AX Mon is early indicated in the literature as a double system (Merrill 1952). Cowley (1964) noted that the system consists of a rapidly rotating B3 star and a somewhat fainter K giant.

We describe in this paper the main features of the far UV spectrum of HD 45910 observed with IUE at phase 0.5 (SWP 4762, LWR 4131). An analysis of this spectrum has been presented as part of a Ph.D. thesis (Danezis 1983). The far UV spectrum of AX Mon presents a large range of ionization stages going from CI, 01, NI, Mgl to SiIV, CIV, NV.

High resolution spectrograms of the two well known Be binary systems ζ Tau and KX And were obtained , in the IR ( 2-2.5µ ), with the Fourier Transform Spectrometer of the CFHT. They do not reveal evidence of a cool giant companion.

Semidetached close binary stars of the Algol type often have primary components of spectral type A0 or earlier and display emission at Hα (sometimes also at higher Balmer lines). They are therefore Be stars. Many binaries of this type are not eclipsing and must look like “ordinary” Be stars. We have discovered high-ionization emission lines of N V, C IV, Si IV, Fe III, etc. in the ultraviolet spectra of totally eclipsing Algols. They probably originate in circumstellar turbulent regions at fairly high electron temperatures, of the order of 100 000 K. They are not detectable in most non-eclipsing systems, but may be there and may play an important role in the dynamics of accretion and mass outflow from the systems.

Basic information about the star o and (HR 8762, HD 217675-6) is a bright (V=3.55m, B-V=-0.11m U-B=-0.50m) B6 III-Ve star. Its emission-line and shell spectrum is known to come and go on a time scale of years. In fact, the presence of a strong shell spectrum misled the authors of the HD catalogue to assign the star two HD numbers.

o and is a prototype of Be rapid light variables in the sense that it has the longest record of photoelectric monitoring which goes back to 1915! The long history of attempts to understand the variations of this star has recently been summarized by Harmanec (1983). Collecting and analyzing all historical photoelectric observations, Harmanec (1984) concluded that the light variations of o And can be understood as a superposition of a long-term cyclic (8.5y) variation with an amplitude over 0.1m (the star being faintest at the recorded shell maxima) and a rapid periodic variation with a period of 1.571272 days.

Cataclysmic-variable stars (CVs) are technically Be stars, since Balmer emission lines appear in their spectra. In general, CV spectra are so unusual that they are easily recognized. The main features are extremely broad, shallow hydrogen lines with broad, faint emission cores. The Balmer emission decrement is very slow, more like that of an H II region than that of a normal Be star. The Balmer decrement in cataclysmic variable stars has been discussed by Elitzur, et al.

If a CV were viewed pole-on, the spectrum might be quite different. It could exhibit a smooth continuum or may even resemble a normal Be star, except for the Balmer decrement. In this case, there may be a bright cataclysmic variable lurking in the Be star class. A possible example is the star NS 274-67, an 03e star described by Conti et al. (1986), which has Balmer emission from H-beta to H-ep-silon, but no nebular emission at [0 II] or [0 III].

Over the last three years I have obtained about 2000 spectra of a sample of 25 rapidly rotating Bn and Be stars. All but two of the program stars show obvious line-profile variations due to non-radial oscillations. The non-emission stars are each pulsating in one or two short-period high-degree (l = 4 to 10) modes, while the Be stars are in all cases pulsating in a long-period % = 2 mode, and often in a short-period high-Z mode as well. The amplitude of the pulsations in several stars (λ Eri, o And, ζ 0ph, and 2 Vul) is correlated with the occurrence of Be outbursts. The amplitude of the pulsations is largest before the outbursts, declines slowly during the emission phases to a fraction of its previous amplitude, and then slowly recovers to its previous amplitude, a few months before the onset of the next outburst. The correspondence between the presence of a long-period % = 2 mode and Ha emission in rapidly rotating B stars strongly suggests that non-radial pulsation and rapid rotation are the essential components which enable single early B stars to become Be stars. The time scale between Be outbursts probably reflects the relaxation oscillation cycle of the I = 2 mode excitation and damping.

Recently we have monitored the absorption line profile variations of a number of rapidly rotating B stars. These variations are observed as alternating quasi-absorption and -emission bumps that travel blue-to-red across the profile with a local period defined by the time to move one bump spacing. This type of pattern can be modelled successfully by assuming the presence of one or two active nonradial pulsation (NRP) modes in the photosphere of the star. Many observations show significant departures from the simple expectations of the models: successive bumps can have different amplitudes (which may vary on timescales of hours), and individual bumps can occasionally appear just ahead or behind the position predicted from a mean period. Nevertheless, we believe these anomalies can be treated as perturbations from a time-averaged pulsation model.

The locations of 143 B0e-B7e III-V stars listed in the Bright Star Catalogue are similar to those of the early-B Gould Belt stars. Therefore the Be stars have roughly the same ages and origins as early B stars. The frequency of runaway Be stars must be less than several percent. The radial velocities of Be stars show primarily the reflex of solar motion and show no evidence for systematically negative velocities attributable to mass-loss effects upon the absorption profiles. The scatter in the residual velocities is such that there is unlikely to be many undetected binaries with orbital amplitudes greater than 10 km s-l. We are unable to state whether or not Be stars tend to occur during the overall contraction stage, but we do observe Be stars in roughly constant frequency in clusters of all ages. About 18% of the field B0-B7 III-V stars are Be stars. Clusters show both lower and higher frequencies that may be real or may be due to different observational techniques. The frequencies and distribution of binary periods for Be stars is the same as for non-emission B stars except for the lack of periods less than 10-1 yr. Statistically the 12 classical Be stars with known orbital elements have mass functions indicating that their secondaries are more massive than neutron stars and their secondary mass distribution is like that of normal B stars. We observe 35 companions for 100 Be primaries, so after correction for undetected companions, it seems likely that most Be stars have companions, mostly with periods of years.

Be stars are located in or near the main-sequence band for non-rotating stars. Although this stage of evolution is relatively well understood, there are two main effects that make it impossible to say whether all Be stars are in the same stage of evolution and, if so, what that stage is. One effect is the spread in observed magnitude and colour as a result of rotation. Correction for rotation is not possible because of the unknown inclination of any particular star's rotation axis to the line of sight and because it is not clear what the internal angular momentum distribution is or how it changes as a result of evolution. The other effect is that there are uncertainties in the theoretical evolutionary tracks because the amount of convective overshooting is unclear. Other mixing mechanisms that might in principle also confuse the tracks seem to be small near the main sequence. If Be stars are related to the β Cephei and 53 Persei stars in the same part of the HR diagram, then the non-radial pulsation properties of Be stars may give a clue to their evolutionary state. The existence of a circumstellar disc or ring, however, tells us very little about the evolutionary stage of the underlying star. A useful way forward may be to try to understand individual stars in as much detail as possible.

A list of 124 Be-type stars belonging to 52 open clusters has been compiled. All of them have photometric UBV data, and many of them spectral classification in the MK system. Besides, the cluster's distance modulus and the mean color excess from the member stars are known. Then, we have computed the absolute magnitude and the intrinsic colors of each Be star.

The observed color-color diagram (Figure 1) shows clearly that nearly all Be stars appear to be reddened. In Figure 2 is given the color-magnitude diagram, (B-V)o versus Mv, in whicn the ZAMS (Blaauw, 1963) is also included. A few stars to the left of the main sequence belong to the young open clusters: NGC 869, 6167, 6530 and 6611. This effect may be due to wrong correction of the color indices because of abnormal values of the reddening law.

Knowing that especially young clusters can contain a considerable high fraction of Be stars (Mermilliod 1982, Feast 1972) we made a spectroscopic study of Hα on the ratio of B/Be stars. The aim was to select normal B stars for chemical abundance analyses.

It is well known that a promising way to investigate the evolutionary status of Be stars is to study their frequency and positions in the H-R diagram of open clusters. However, these studies have not yet provided satisfactory results (cf. e.g. Slettebak 1985); observations concerning the B/Be star ratio are badly needed. Therefore we present our observations although they are only a by-product from a different program.

The double cluster h and χ Persei has since long been known to be rich in Be stars (Trumpler, 1926). It belongs to an age group in which various types of variable stars occur (Mermilliod, 1981). Other clusters of this age group are NGC 3293, the cluster that is richest in β Cephei variables (Balona and Engelbrecht, 1983), IC 2581, NGC 4755, and NGC 6871. The comparison of the characteristics of variable stars in all these clusters and in clusters of other age groups could reveal to which extent the cluster stellar content is determined by age and by other parameters, such as galactic location, and so metallicity. Since h and χ Persei are very rich clusters that have not yet been studied intensively as far as stellar variability is concerned (Percy, 1972), we have undertaken a study of the photometric variability of the brightest stars in the double cluster. Our first interest was to detect β Cephei variables, but it rapidly turned out that the double cluster contains a considerable amount of highly variable Be stars.

Rappaport & Van den Heuvel (1982) suggested the following formation scenario for B emission (Be)/X-ray binaries: the progenitor of the neutron star is the initial primary star with mass M1 = 10-20 M⊙, which during hydrogen-shell burning transfers mass to the companion (with M2 < M1). The secondary is spun up due to disk-accretion (e.g. Packet 1981) and has become a rapidly rotating Be star (M ≃ 10-20 M⊙).

With this scenario the observed orbital periods (Porb > 15 d) of the Be/X-ray binaries can be explained. This was shown by van den Heuvel (1983) by adopting an idealized period distribution for unevolved O- and B-type close binaries with Porb, < 30 d (see Fig. 1) and assuming conservative evolution, i.e. that mass and angular momentum are conserved during mass transfer.

Be X-ray binaries have the following general characteristics of interest for evolutionary scenarios: the spectral type of the non-compact star is B2 - 09 (e), the orbital periods are large, ranging from 16 to 581 days, the orbital eccentricities can be large, which could be related to the fact that the X-ray stages are transient; the masses of the optical components range from 10 to 20 M⊙ and their radii are relatively small (5 - 10 R⊙). The optical luminosity is lower than 3 x 10 4L⊙.

Several Be stars are identified with bright X-ray sources. (Rappaport and Van den Heuvel, 1982). The bright X-ray emission and observed periodicities indicate the existence of compact objects (white dwarfs, neutron stars or black holes) near the Be stars. A prime example is the brightest X-ray source A0538-66 in LMC, which contains a neutron star with a rotation period of 59 ms. Apparao (1985) explained the X-ray emission, which occurs in periodic flares, by considering an inclined eccentric orbit for the neutron star around the assumed Be-star. The neutron star when it enters a gas ring (around the Be-star) accreting matter giving out X-rays.

The X-ray emission from the compact objects, when the gas ring from the Be-star envelopes the objects, has interesting consequences. The X-ray emission produces an ionized region (compact object Stromgren sphere or COSS) in the gas surrounding the compact object (CO).