We discuss the connection between the periodic light variations and the equatorial mass loss of Be stars. The observed properties of the short period (~ day) variations seem to indicate that they arise in the photosphere. An upper limit for the surface magnetic field of Be stars is derived from the rate of angular momentum loss expected from the typical mass loss-rate in Be stars. The upper limit suggests that surface magnetic fields of Be stars are too weak to make a spot. We argue that the periodic variations of Be stars are explained by nonradial pulsations whose periods on the stellar surface are much longer than the rotation period. They transport angular momentum from the core to the envelope to accelerate the surface regions. If this mechanism works sufficiently well, the rotation speed near the surface will reach to the critical velocity and an excretion disk will be formed around the star. A simple model for a steady-state excretion disk around a Be star is found to be consistent with the density structure inferred from the IR fluxes.

Be stars are hot stars which have shown emission in at least one Balmer line on at least one occasion. As the definition implies, the Be phenomenon can be variable with time: on time scales of days to decades as the circumstellar disc develops and disperses; on time scales of days to months in a few Be stars which are interacting binaries; on time scales of 0.2 to 2 days due to non-radial pulsation or possibly rotation. The Be stars are worthy of photometric study because they are bright and numerous; the nature of the short-term variability is not yet agreed upon; the cause of the development of the disc - and its relationship to the short-term variability - is also not yet known.

Since their discovery (by Father Secchi in 1866) until the end of sixties, Be stars were not a subject of any systematic studies of their possible light and colour variations. Already at that time, the astronomical literature contained ample evidence showing that a number of Be stars were light variables. However, almost all such findings resulted as by-products of studies of different or wider groups of objects. Feinstein (1968) was probably the first who pointed out explicitly that many Be stars are light variables. A pioneering study which was aimed at the detection of light variations of a large group of Be stars by means of differential photoelectric photometry was carried out by Haupt & Schroll (1974).

This is part of a series which has the purpose of examining the nature of the stars belonging to open clusters. The aim of this series is, among others things, to study short period pulsating stars, mainly of the Delta Scuti type, by first establishing the membership of each star to the cluster, to determine the abundance of the Be and Ap phenomena and blue stragglers in open clusters for clusters of different ages and metalicities and, eventually, to study the chemical enrichment of the galaxy when age, dynamics and metalicity are known for a fair number of clusters.

Many previous surveys of the rapid variability of Be stars (e.g., Balona et al. 1987, Cuypers et al. 1989, Balona et al 1992) were limited to the Strömgren b-band only. We have undertaken a new observational effort with the specific aim of studying colour variations. A search for new rapidly variable stars should provide us with statistical material needed to generalize properties of the phenomenon with respect to the stellar parameters. Some preliminary results of these items are presented here.

The reality of the previously reported triple-wave light curve (Cuypers et al. 1989, A&A Suppl. 81, 151), and relations between rapid line profile and brightness variations of the B1-2III-Ve star η Cen (HD 127972) were investigated by means of simultaneous high-resolution spectroscopy (4 nights, 86 Si III 455.2622 nm profiles) and two-station uvby photometry (14 nights) secured in 1992. To this database were added the uvby photometry by Cuypers et al. and unpublished b observations by L.A.B. from 1988–1991.

The usual methods of spectral clasification, equivalent widths of Balmer lines or photometric calibrations are not suitable for the determination of the astrophysical parameters of the underlying star in Be-type objects. The spectrum is distorted by the circumstellar envelope lines, while the contribution of the envelope continuum radiation contaminates the photometric indices.

A large fraction of Be stars show periodic light and line profile variations with a timescale of about one day. The mechanism which causes these periodic variations has been attributed to nonradial pulsation (NRP) or rotational modulation (RM). The authors present arguments supporting the two opposing points of view with the purpose of stimulating subsequent discussion by the Symposium participants.

ɛ Per (45 Per, HR 1220, HD 24760, ADS 2888A) is a star which was used for years as a spectrophotometric and MKK classification standard. Nowadays it is known as an archetype of early-type line-profile variables (LPV). The variations are so pronounced that they were misinterpreted for a composite spectrum (B0.5+A2) and the star was even reported to be a double-lined spectroscopic binary (Petrie 1958). Since 1983, the characteristics of line-profile variability in the spectrum of ɛ Per have been intensively studied (Bolton 1983, Smith 1985, Gies & Kullavanijaya 1988). The recent investigations of RV variations of ɛ Per led to the conclusion that ɛ Per is a spectroscopic binary with a period of about 14.05 days and a rather eccentric orbit (Harmanec & Tarasov 1990). The issue of the binary nature (with a very eccentric orbit) is of utmost importance since it could be causally related to the extremely large observed line-profile variations (e.g., Polfliet & Smeyers 1990, Tassoul & Tassoul 1992).

φ Per (HD 10516) is a spectroscopic binary with a 126.699-day period (Ludendorff 1910, Cannon 1910, Harmanec 1985). However, most of the published RV curves of the primary are based on H I shell lines and exhibit a typical distortion with a sharp maximum, shallow minimum and a bump at 0.P4 after the RV maximum (c.f., e.g., Harmanec 1985). There has been controversy on the nature of the secondary. Hynek (1940) and Hendry (1976) concluded that the binary was composed of two B stars. Peters (1976) suggested that the secondary of φ Per is a Roche-lobe filling K giant. Poeckert (1981) reported the discovery of a weak He II 4686 emission which moved in antiphase to the Be primary and suggested that it originated in the disk around the secondary. In his interpretation, the secondary is a helium star, a remnant of an originally more massive star which in the past transferred its mass to the present Be star. There is now no mass tranfer in the system according to Poeckert. Using RVs of the broad absorptions for primary, and of the He II 4686 emission for the secondary, Poeckert obtained two roughly sinusoidal RV curves and M1sin3i = 21.1 m⊙ and M2sin3i = 3.4 m⊙.

The hydrogen-deficient B stars (often known as extreme-helium stars: EHes) show weak or absent Balmer lines and the following properties (Heber 1986a): Teff ~ 12 − 25kK, log g ~ 2 − 3, log nH ~ −3 − −4, nHe ≈ 0.99, nC ~ 0.01. Their galactic positions and apparent magnitudes indicate that they are low-mass supergiants belonging to the Galactic bulge (Jeffery et al. 1987).

A few hydrogen-deficient B stars (or extreme helium stars: EHes) exhibit Be-type behaviour (the HdBe stars). The principal class members are BD −9°4395, LSE 78 and DY Cen. The hot RCrB stars MV Sgr and V348 Sgr and the hot EHes HD 160641 and LSS 5121 are related to these objects. Since there are wide abundance variations amongst the EHes, as well as wide variations in Be-type behaviour, the principal HdBes have been analysed for atmospheric parameters, including Teff, log g and composition.

We report the discovery of a systematic and large calcium abundance stratification in cold Ap stars. These detections are in very good agreement with diffusion theory and set stringent upper limits on turbulent processes.

Important astrophysical applications of Stark broadening of spectral lines of multiply charged ions are in the physics of stellar interiors (Seaton 1987). In subphotospheric layers, the modelling of energy transport requires radiative opacities and thus, certain atomic processes must be known accurately. At these high temperatures (105 K or more) and densities (1017 − 1022 cm-3) Stark broadening of strong multicharged ionic lines plays a non-negligible role in the calculation of the opacities, especially in the UV. Moreover, with the development of spectroscopic investigations from space, UV and extreme UV spectral line research has been further stimulated.

Lines of Pt II have been discovered in Hg Mn stars by Dworetsky (1969). The analysis of a few strong Pt II transitions, which are also observed in IUE spectra of stars, has shown (Dworetsky et al., 1984) “that Pt is, like Hg, among the most overabundant elements in the atmospheres of Hg Mn stars, with enhancements of the order of 104 to 105 over the solar system abundances”. Dworetsky et al. (1984) selected also the four Pt II lines which might be used for astrophysical applications. Moreover, they determined the corresponding theoretical gf values. The aim of this contribution is to investigate the Stark broadening of these Pt II lines and to provide the corresponding Stark widths.

A progress report will be given on recent studies of circumstellar variability of Be stars, their implications on the formation process, the dynamics, the structure and the morphology of the envelope. The following points are discussed: recurrence of outbursts, simultaneous infall and outflow events, variability in hot regions versus variability in the cool envelope, the establishment and vanishing of Be and Be-shell phases, periodic medium-term and quasi-periodic long-term phenomena, and constraints on models.

Based on spectroscopic observations of more than 100 Be stars, we present the results of the phase changes of Be stars.

We examine Hα equivalent widths versus spectral-types for 41 Be stars. Although the Wα value is scattered for a given spectral-type, a well-defined upper limit exists. Most shell stars are located in the upper limit region. The shell stars all have v sin i/vcri larger than the normal Be stars in the upper limit region. This strongly suggests that the distinction between shell stars and normal Be stars of high equivalent width is a result of variation in inclination angle i. Polarization data also support this hypothesis.

It is very important for the understanding of the Be phenomenon, and particularly for locating a possible Be phase in the evolutionary track of B stars, to accurately determine the proportion of Be stars among all B stars. This type of study was already made several times in the past. Results obtained generally show a maximum Be frequency around spectral type B2 then a decrease towards late spectral types. Actually Be stars do not have the same characteristics as “normal” B stars and we have to take this into account in the determination of the ratio : number of Be stars / number of B stars. We use the Bright Star Catalogue (Hoffleit & Jaschek 1982) and the Supplement to the Bright Star Catalogue (Hoffleit, Saladyga & Wlasuk 1983) containing stars V= 7.10 and brighter. This study needed to be made separately for the different spectral types because:

- Physical parameters of B stars are very different from B0 to B9;

- Emission characteristics of Be stars vary very much, with a decrease from B0e to B9e.

We successively consider three effects which can influence the frequency of Be stars:

- The over-luminosity of Be stars as compared with B stars;

- Spectral type changes during constant mass evolution;

- Spectral type changes due to the fast rotation of Be stars.

Although very different in shape, Be-type and Be-shell-type spectral lines are thought to arise from the same kind of circumstellar environment, a disklike envelope. The only discrimination between these two types of spectra is the inclination angle: If sufficiently high, a large column depth of circumstellar material causes partial obscuration of the stellar disk, i.e. shell absorption; otherwise a pure emission line is observed. A shell line can be considered as equivalent to P Cygni-type absorption trough for different geometry (disk instead of sphere) and kinematics (rotation instead of outflow; see Hanuschik, these proceedings).