IUE low resolution spectra are an excellent way to determine the temperatures of the hot companions of binary Cepheids. The spectral types of the companions are derived by comparing the spectra with the spectra of standard stars. Absolute magnitudes are calculated from the magnitude difference between the two stars and the absolute magnitude of the Cepheid. In this study, eight binaries containing a Cepheid and a hot companion evolved beyond the ZAMS are discussed.

In the past several Wolf–Rayet stars have been investigated photometrically by different authors to search for fast pulsations. So far all results were negative. However, in 1992 Blecha, Schaller & Maeder published a 627sec period with a semi–amplitude of about 2.5mmag for the Southern Wolf–Rayet star WR40 (HD96548).

Over thirty years ago Baker & Kippenhahn (1962) demonstrated that an instability driven by the opacity mechanism is the cause of Cepheid pulsations. Recently it has been shown that the same mechanism is responsible for oscillations observed in β Cephei, SPB and perhaps in other variable B-type stars. The search for the driving mechanism in hot stars began in the late sixties with no success until the opacities calculated with the OPAL code (Iglesias, Rogers and Wilson 1990, 1992) became available. The crucial new feature in the opacity is the local maximum at T ≈ 2 × 105 K caused by iron lines which was ignored in earlier calculations. Recently, stellar opacity data from an independent project (OP) became available (Seaton et al., 1993). The agreement between the two opacity data is satisfactory.

In B stars the opacity mechanism drives two distinct categories of normal modes. The one relevant to β Cep stars encompasses low order p- and g-modes with periods 0.1–0.3 d. The other includes high-order g-modes with periods ranging up above 4 d. Excitation of such modes may explain most of the slow variability observed in B stars. The theoretical instability domain in the H-R diagram is very sensitive to metal abundance. For the standard value of Z = 0.02, the total instability domain in the main sequence band extends from spectral type O9.5 to B9. In types later than B2 only high-order g-modes are unstable.

In the calculation of stellar models with the Cox–Stewart opacities no convective zones in the outer layers of massive stars appear. The new OPAL opacities (Rogers & Iglesias, 1992) show a significant bump in the opacity near temperatures of log T = 5.2. This opacity effect results in a small convective zone in the envelope of stars with mass ranging from 15 M⊙ to 150 M⊙, apart from possible convective zones caused by ionization. This was also briefly mentioned by Glatzel & Kiriakidis (1993). For stars on the main sequence this zone is small, about 1% of its radius on the zero age main sequence up to 7% at the onset of the core helium burning and contains a negligible amount of mass. For helium burning stars, however, this convective zone moves inward, keeping the same size but containing more and more mass.

We show that the OPAL opacities, in addition to explaining the origin of β Cep stars pulsations, also predict existence of a large region in the Main Sequence band at lower luminosities, where high-order g-modes of low harmonic degrees, l, are unstable. The excitation mechanism remains the same and is due to the usual κ-effect acting in the metal opacity bump (T ≈ 2 × 105K). The new instability domain nearly bridges the gap in spectral types between δ Set and β Cep stars. Periods of unstable modes are in the range 0.4–3.5 days for l = 1 and l = 2. We propose that this excitation mechanism causes photometric variability in the Slowly Pulsating B-type stars (SPB stars, Waelkens 1991) and perhaps in other B stars whose variability in the same period range has been reported.

Radiative Rosseland-mean opacities for stellar envelopes from The Opacity Project, hereafter OP, became available (Seaton et al. 1993). Here we present a preliminary survey of pulsation properties of B-star models, obtained with this new opacity data and compare these properties with those of models built with the OPAL opacities (Iglesias, Rogers and Wilson 1992). The two opacity projects employ very different approximations in treatment of the plasma and atomic physics. In addition the OP tables are calculated including absorbtion due to three iron group elements ignored in OPAL tables, namely Cr, Mn and Ni.

An investigation of the stability properties of stellar models describing massive stars is motivated observationally by the necessity to explain the observed Humphreys - Davidson (HD) limit and the variability of the most massive stars known, i.e. the existence of luminous blue variables (LBVs). Theoretically, a determination of the upper mass limit for stable stellar objects together with its physical explanation and interpretation is of fundamental interest.

During the past twenty years, different methods have been developed to identify the modes in non-radially pulsating stars. Before the introduction of high-resolution spectrographs with sensitive detectors, identifications were obtained from photometric observations. More recently, mode identification is obtained by means of spectroscopic methods. In this paper, we present an overview of the different mode-identification techniques currently used and we describe their accuracy to identify the modes present in different kinds of pulsating stars. By means of some applications of the moment method, we show that this method deserves far more attention than it has received until now.

We have extended a formalism that was developed for radial pulsators (Buchler & Goupil 1984) to the nonradial ones (Goupil & Buchler 1993). For an ℓ = 1 triplet, for example, the modal amplitudes satisfy the nonlinear amplitude equations

The periodic variations in the photospheric lines of Be stars could result from rotation or nonradial pulsations (NRP). Here I review the spectroscopic results on ten Be stars which have been targeted in recent multiwavelength campaigns. Most of the stars display variability consistent with an l = –m = 2 NRP mode with periods on the order of a day, and the associated light curves are consistent with the pulsation periods, amplitudes, and phases derived from spectroscopy. The variability is strongest in the line wings which only agrees with the NRP prediction that horizontal motions should exceed vertical motions for g-mode pulsations. High frequency, non-periodic variations are also observed, but I argue that these are of circumstellar origin and do not represent high-order NRP. The NRP periods are also found in variations of the Hβ emission line but these probably result from photospheric and not disk variations. However, the same periods occur in the C IV λ1550 P Cygni line, which strongly suggests that NRP modulates the local stellar wind.

We present the results of time series observations of He I λ4921 in the spectra of 10 Be stars. The spectra were obtained during multi-wavelength campaigns of IUE spectroscopy, optical photometry and polarimetry, and spectroscopy undertaken to study the connection (if any) between nonradial pulsation (NRP) and mass loss in Be stars. Power spectra have been calculated for the time series of intensity variations within the profiles. These were used to determine periodic variability and phase content to analyze the variations in the context of the NRP model. The observations were made at three observatories (Table 1). All the spectra record at least Hβ and He I λ4921 with a S/N > 200 per pixel and a resolution of ≈ 0.2 Å. We have concentrated the analysis on He I λ4921 as a probe of photospheric variability. This line is generally free of emission (unlike He I λ6678), but in some cases an Fe II λ4924 shell component appears in the red wing.

28 Cyg (V1624 Cyg, HD 191610, HR 7708; B2e, v sin i = 310 km s-1) has been the target of several observational projects, and in 1988 of a large international campaign. This attention was inspired by several photometric studies and especially by the 1985 nearly simultaneous optical and UV spectroscopic monitoring by Peters & Penrod (1988). They found that the line-profile variations were controlled by two frequencies, 1.45 c/d, and 7.43 c/d, which they identified with sectorial pulsations of modes l = 2, m = +2 and l = 10, m = +10. Rapid changes (0.5 to 1 hr) of the CIV wind profile were found; its equivalent width appeared to correlate with the phase of the l = 2 mode. Pavlovski & Ružić (1990) - who independently analysed Hvar 1985 UBV photometry of 28 Cyg - found periodic light variations with a double-wave light curve and a frequency of 1.54 c/d. However – because of the residual scatter around the mean light–curve – the authors tentatively suggested possible multiperiodicity (1.54, 1.33, and 0.95 c/d).

We have started in 1990 a search for moving bumps in the HeI λ 667.8 nm of mainly southern, bright Be stars. The objects of our sample have been selected on the basis of photometric variability (Cuypers et al., 1989). High resolution (R≥ 30,000), high signal-to-noise ratio (S/R≥ 300) spectroscopic observations have been performed at the brazilian Laboratório Nacional de Astrofísica with a CCD camera attached to the coudé spectrograph of the 1.60 m telescope (e.g. Table I). Several hundred spectra have been taken during the last three years. Photometric observations simultaneous with spectroscopy were made on the same site in July 1992 with a two-channel photometer (Stromgren b filter) and a CCD camera (Johnson B filter) installed at two 0.60 m telescopes. The idea is try to disentangle the controversy between NRP and RM models with the help of simultaneous spectroscopy and photometry.

We present here a preliminary report on the analysis of simultaneous photometric and spectroscopic observations of the Be star α Eri, taken during 1992 at the Laboratório Nacional de Astrofísica. α Eri (Archenar, HR472, HD10144, B3–4 III-IV) is the brightest Be star in the sky. It shows activity cycles in a time scale of a few years. Balona et al. (1987) presented a comprehensive study of the star based on photometric and spectroscopic observations. They found a periodicity of 1.26 days in both radial velocity and light variations and argued that a spotted star rotating at this period is the simplest working model that explains the observations.

Some investigators have attributed the photometric and spectral line profile variations (lpv) that are common among the B stars to nonradial pulsation while others have attempted to explain them by rotation of photospheric or circumstellar structures with respect to our line of sight. One of the problems in resolving this debate has been our lack of knowledge of how these variations depend on fundamental stellar characteristics. Surveys of lpv have covered the Be and a few Bn stars (D. Penrod, unpublished), the O stars (Fullerton 1990), and B8-B9.5 main sequence stars (Baade 1989), but noone has carried out a systematic search for lpv among the near main sequence, non-emission line early- and mid-B stars. This paper describes preliminary results from such a survey.

MUSICOS (MUlti-SIte COntinous Spectroscopy) is an international collaboration interested in areas of astronomical research requiring continuous, high-resolution spectroscopy. The MUSICOS stategy has been 1) to organize international campaigns (Catala et al 1993), 2) to build a prototype fibre-fed, high-resolution, echelle spectrograph (Baudrand & Böhm 1992), and 3) to install duplicates of the MUSICOS spectrograph on 2m class telescopes around the world. Stages 1 and 2 have been completed. Preliminary results on θ2 Tau from the second MUSICOS campaign are presented here.

Preliminary results of a time-series analysis on a sequence of high dispersion, high-signal-to-noise, optical echelle spectra of the O4 supergiant, ζ Puppis (HD 66811), are presented. Lines of Hei λ5876Å, Heii λ5411Å, Niv λ6381Å and Civ λλ5801, 5812Å are analysed; they show variations in line shape and equivalent width, with ‘bumps’ and ‘dips’ seen moving from blue to red within the line profile. The qualitative pattern of variability seen in all these lines is similar.

A Fourier technique, incorporating the iterative clean algorithm, was used on the individual wavelength samples. A periodicity of 8.5-hrs was detected in Heii, and strongly suspected in the remaining absorption lines. This periodicity seems to rule out wind variability as the principal origin for the optical line profile variations seen in ζ Puppis.

HD 93521 is an O9.5 V star with several well-documented peculiarities, including: an extremely large projected rotational velocity (400 km/s according to Conti & Ebbets 1977); absorption line profile variations (Fullerton 1990); weak emission at Hα (Irvine 1989); evidence for a rotationally-modified stellar wind in its UV resonance lines (Massa 1992; Howarth & Reid 1993); and stellar wind variability in the form of discrete absorption components (Prinja & Howarth 1986). Thus, HD 93521 provides an ideal laboratory for studying the interaction between rapid rotation and variability in both photospheric layers and the circumstellar environment.

The line-profile variables observed on the upper main sequence have been interpreted by some astronomers to be the manifestation of nonaxisymmetric oscillations. More specifically, most of these variables can be modelled by prograde or corotating equatorial waves. In the absence of rotation, these waves have surface velocity distributions which are given simply by spherical harmonics. Unfortunately, the corresponding velocity fields in the presence of rotation are much more difficult to calculate. In this paper, I will summarize what is known about the effect of rapid rotation on the normal mode eigenfunctions of main sequence stars. The principal conclusions are as follows: Low-order, axisymmetric modes couple very strongly to rotation and their velocity distributions are very much different from those of their zero-rotation counterparts. On the other hand, higher-order (shorter wavelength), nonaxisymmetric modes couple only weakly to rotation and, therefore, retain many of the spherical harmonic properties that they possess in the absence of rotation.

We present the result of nonadiabatic analysis for nonradial pulsations in uniformly rotating main sequence models.

The angular dependence of the amplitude of a nonradial pulsation mode with an azimuthal order m in a rotating star is represented by a sum of terms proportional to spherical harmonics Ylm(θ, ø) with l = |m|, |m| + 2,… (even mode) or l = |m| + 1, |m| + 3,… (odd mode; see e.g. Saio and Lee 1991 for detail). (In this paper we consider only even modes.) This property makes the analysis complex compared with the case without rotation, in which a single Ylm expresses the angular dependence of a given mode. In our numerical analysis the summation is truncated, in which only first two terms are taken into account. Lee and Saio (1987) give the differential equations for nonadiabatic nonradial pulsations in a uniformly rotating star. Treating the angular frequency of rotation as a free parameter, we applied the nonadiabatic analysis to a main-sequence evolutionary model, for which the effect of rotation is neglected.