New, NTT observations of the B[e] supergiant RMC4 in the Small Magellanic Cloud show that the star is embedded in a circumstellar nebula, whose velocity structure is typical of a bipolar outflow. The nebula is kinematically associated with RMC4 and expanding at 100 km/s on average. Its diameter as computed from the spatial extension of its spectral lines is of 8.6″ which corresponds to a linear size of 2.4 pc and gives a dynamical age of 12500 yr. The line ratio [NII]λ6584/[SII]λ6717 indicates that the nebula is nitrogen enriched as it is expected for ejected nebulae. This is the first bipolar outflow detected around a well-established B[e] supergiant which also belongs to a binary system. The nebular morphology and chemistry are consistent with the picture of a binary merger in which RMC4 was composed by a close pair and a third star (the observed companion). The close pair merged into a single star (the present B[e] supergiant) and produced a circumstellar nebula to be later shaped by the ensuing B star wind (cf. Langer & Heger 1998).

We use a phenomenological approach to derive new constraints on the mass loss mechanisms working in the classical Be and the Herbig Ae/Be stars, which represent two groups of stars located very nearby in the HR diagram. We discuss observational similarities of classical Be and Herbig Be stars on one side, and stress, on the other side, the major differences between Herbig Ae and Herbig Be stars. For further details and a detailed bibliography, see Balona & Böhm (2000). A more quantitative approach will be needed and is deferred to future work.

We present spectropolarimetric observations across Hα of a sample of Herbig Be objects. Roughly half of the observed Herbig Be stars show polarization changes across Hα, implying immediately that their ionized, hence inner, envelopes are not spherically symmetric. This pattern, if confirmed by observations of a larger sample, could indicate that the non-detection rate is simply a consequence of sampling randomly-oriented circumstellar disks able to scatter starlight within a few stellar radii, with as implication that the massive Herbig Be stars are still surrounded by accretion disks. In addition, we present preliminary modelling which will assist interpreting the kinematic structure of the electron-scattering region.

Assuming an extended spherical atmospheric model we analyze the continuum and line behaviors considering 3 different temperature laws and different density distributions. We show that - for the continuum intensity distribution - at some physical conditions, a limb brightening profile appears.

Absolute magnitudes of Be and B stars are computed for each spectral type and luminosity class V and IV, using the Hipparcos parallax measurements. Some simulations have been carried out in order to estimate the effects which could bias the mean absolute magnitude calculations. As a result, only stars with σπ/π < 15% have been used. A first result is that B stars are fainter than previous estimations by about 0.5 magnitude on average. We then observe that on average Be stars are brighter than B stars of the same spectral type and this over-luminosity increases with the spectral type. A possible interpretation is proposed based on the fact that the rotational velocity of the late Be stars is near the critical rotational velocity.

Interferometry in the visible now provides milliarcsecond spatial resolution and thus can be used for studying the circumstellar environment of Be stars. In this review I will first introduce the two basic quantities that an interferometer can access: the modulus and the phase of the visibility. I will illustrate how these two quantities can be used to understand the physics of Be disks through recent results from the VLA, the MkII and the GI2T interferometers. I will insist on the importance and the potential of coupling high angular resolution with high spectral resolution to the study of Be disks. Since the possibility of direct optical imaging of Be disks will be limited due to complexity and time consumption, I will present the role models can play in interpreting high angular resolution observations when direct imaging become difficult. Finally I will draw up a list of challenging objectives for the next generation of synthetic arrays (GI2T/REGAIN, Keck, ST3, VLTI) which will bring new understanding of active hot stars physics.

The ST3 mission will launch a space optical interferometer into heliocentric orbit, for observation of 50–100 sources on baselines of 40–200 m, during a time span of 3–4.5 months. The angular resolution in the 450–900 nm passband will be ≈ 500 microarcseconds. For Be stars, a velocity resolution of 50–100 km/s can be achieved by operating the interferometer in a Fourier Transform Spectrometer mode. This should allow spectrally resolved measurements of the brightness distribution in circumstellar disks, but will be limited to stars brighter than a visual magnitude of ≈ 5.

We discuss the infrared spectral energy distribution of Be stars, focusing on new results obtained with the Infrared Space Observatory. The 60-160 μm flux of some Be stars is higher than expected, which may be due to cold dust or an outer disk component with enhanced densities. The infrared spectrum of Be stars is dominated by numerous HI recombination lines, whose line strengths show a complex behaviour. The electron temperature in the disk of ϒ Cas was found to be about 9500 K, and evidence for an elevated temperature near the upper part of the disk is presented. Be stars may be recognized from their infrared spectrum on the basis of HI line flux ratios.

I discuss the X-ray observations of Be stars, and how their properties compare to non-emission B stars. I focus on several specific stars that show high flux levels and variability but also report on several interesting survey results. The Be X-ray properties are discussed in the context of wind-shock X-ray emission from normal OB stars as well as in the context of general mechanisms that have been proposed to explain the Be phenomenon. Finally, I conclude with a discussion of the spectral diagnostics that will be available from the new generation of X-ray telescopes.

Our program aims at observing selected bright multiple Be stars using the technique of long-baseline optical spectro-interferometry with the GI2T-REGAIN interferometer. These observations - combined with spectral and photometric observations and with appropriate modelling - will bring new constraints on the nature and evolutionary stage of multiple Be stars through an accurate determination of their basic physical properties.

The classical detection of magnetic fields in Be stars remains a challenge due to the sensitivity threshold and geometrical cancelation of the field effects. We propose to study the Zeeman effect using Spectro-Polarimetric INterferometry (SPIN) which consists of the simultaneous use of polarimetry and very high angular resolution provided by long baseline interferometers. As monitoring of the instrumental polarisation is mandatory in order to calibrate interferometric observations in any case, the polarised signal is a natural by-product of interferometers. This method will be tested on the GI2T interferometer thanks to its high spectral resolution and its polarimetric capabilities.

The wide range of periodic and cyclic variabilities commonly observed in Be stars and their circumstellar disks is reviewed. Many of them are related to the basic nonradial pulsation, the effects of which may range from the photosphere to the disk. Through pulsation-triggered outbursts, some stars may owe their Be characteristics largely to their intrinsic variability. However, because late-type Be stars do not normally show detectable short-periodic variability, this is probably not true of all Be stars. Comparisons are made with other variable early-type stars.

In this study we present an overview of the search for multiple frequencies in variable Be stars from surveys. We review the recent ground-based photometric surveys and conclude from them that multiperiodicity is not detected. The photometric experiment on board Hipparcos, on the other hand, does allow us to derive multiple periods for a few Be stars. We present the results of our frequency searches in the Hipparcos data for 50 λ Eri stars.

Further we review the recent spectroscopic campaigns dedicated to some stars. It seems that all of these campaigns point to multiperiodic phenomena, although for most of them the time base is insufficient to derive the periods with high accuracy. Moreover, the spectroscopic periods are often not the same as those found from the photometry. This clearly demonstrates the necessity of long-term simultaneous photometric and spectroscopic campaigns. We present the preliminary results of such a campaign for the λ Eri star HD 105382.

We close with reflections on the cause of the variability observed in the λ Eri stars and on the best viewpoint for further research in this area.

Photometric observations of o And collected from 1994 to 1998 are presented. Neither multimodal pulsations nor rotational modulation seems able to explain completely the complex light curves, but a model based on the pulsations together with some type of magnetic activity could be applicable.

The Be stars HR 1960 and HR 2968 have been monitored from 1978 to 1998 in Geneva photometry and from November 1989 to March 1993 by the Hipparcos satellite. A spectroscopic survey of these stars started in September 1998. Both stars show luminosity variations with a period of 395 d and 371 d for HR 1960 and HR 2968 respectively.

Variations in the winds of the Be stars λ Eri and 66 Oph during the lifetime of the IUE spacecraft and the response of the Hα-emitting disk to these changes are briefly discussed. Common behavior is identified and evidence for cyclical variability on a time scale of ∼1 yr is presented.

Do optical/UV variations of spectra of Be stars arise from pulsations or structures hanging over them? There is a need for a temperature-sensitive discriminator for circumstellar structures above Be stars to determine whether their UV flux variations are due to occulting structures or photospheric pulsations.

To address this question, we acquired far-UV (SWP camera) echellograms covering λλ1150–2100 from the IUE final archive. For selected B stars we have selected suitably exposed spectra observed within a short timespan, typically 1–3 days. We summed all the valid net fluxes within each of 60 echelle orders and placed them into an order vs. time array. We identify those times corresponding to high and low fluxes over several reference orders and form the quantity 1 - Fmin(λ)/Fmax(λ), for all echelle orders. The result is a low-resolution spectrum sensitive to absorptions at certain wavelengths, either strong lines such as Lyman α and several UV resonance lines or aggregates of metallic lines. All these are potentially sensitive to temperature variations.

Several periods close to 0.5 d have been reported for μ Cen. We obtained over 300 echelle spectra and simultaneous multicolour photometry over a two-week period. We find that the period is 0.66 d and is, in fact, singly periodic. Greyscale images of many consecutive nights show moving features which do not repeat nightly, as may be expected for P = 0.5 d, but repeat approximately every alternate night. In spite of this, periodogram analysis of the radial velocity, as measured at the point of minimum intensity in the spectral line, does indeed show a 0.5-d period. The amplitude of this periodicity decreases during the course of the run. This seeming contradiction, as well as the apparent multiperiodicity, is due to the use of period finding techniques which exclude singly-periodic solutions of a complex nature. We interpret the variations as due to two, co-rotating, circumstellar clouds distributed around the star, such that they are separated by a time interval of 0.5 d. It appears that μ Cen is another example of the growing list of Be stars in which the periodicity is due to co-rotating photospheric clouds.

We re-discuss the reality of the four periods near 0.5 d and reject the claim (Balona et al., these proceedings) that μ Cen exhibits a single ∼0.66-d period caused by co-rotating clouds. The alleged 0.66-d period is not genuine but results from the simultaneous proximity of this value to the beat period between the true 0.28- and 0.5-d periods as well as to a 2-c/d alias of the 0.28-d period.

It is shown that the line profile variability (lpv) of μ Cen can be modeled as non-radial pulsations (nrp). The use of the most recent available modeling technique for the nrp and carefully computed stellar observables (intrinsic line profile and flux) are, however, mandatory to take full advantage of present high-quality data.