Data resulting from the IRAS survey relating to the stars of the RV Tauri class have been examined. Of a total of 78 objects definitely assigned to this class in the 4th edition of the GCVS, 33 are to be found in the Point-Source Catalogue, and for six of these spectral data are available in the low-resolution Spectral Atlas (to 22.6 μm published by the IRAS Science Team (A & A Suppl. 65, 607, 1986). In the latter the pronounced difference in infrared emission between the carbon-rich star AC Her and most other RV Tauri stars, first noted by Gehrz in 1972, is striking, taking the form of a very marked continuation of that star's emission to at least 22 μm.

The Period-Luminosity (P-L) relation for the RV Tauri variables (DuPuy 1973) contradicts those of other variables in the instability strip as it predicts lower luminosity at longer pulsational periods. It is based primarily upon three globular cluster variables. DuPuy determined secular and statistical parallaxes from 23 field variables, defining a simple P-L relation with a slightly negative slope.

We present preliminary results obtained between March and August of 1987 of a two-year spectropolarimetric study of 10 RV Tauri stars. The observations were made at the University of Wisconsin's Pine Bluff Observatory using the 36 inch cassegrain telescope with a Lyot polarimeter. The spectral resolution of the instrument was 6 A and covered the range between 4350 A and 7500 A. The polarimetry covered the same wavelength range but is binned in four broad bands at 4828 A, 5361 A, 6025 A, and 6877 A.

A dynamic model of the LPV atmosphere which calculates the time-dependent variations of density, temperature, velocity, and composition with radius has been used to predict changes in absorption line profiles with phase. General results based on density variations are shown in figure 1. Similar graphs, based on concentrations of individual species, are presented for six metals: Fe, Si, Mg, Ca, Cr, and Ti. Those species which remain essentially proportional to the density (e.g., most neutral metals) exhibit the same type of variation. Ionized metals show more individual variations, including some doubling at certain phases, caused by enhanced ionization in the regions traversed by the upper shock. The basic variation predicted by this model has been observed for CO lines in Chi Cygni. (Hinkle, Hall, and Ridgway 1982, Ap. J. 252, 697)

An investigation of parameters of a sample of more than two hundred late-type long-period variable stars (LPVs) has been carried out. In more detail were studied 13 giants (Mira Ceti-type variables R Aql, RR Aql, RT Aql, R Leo, U Ori, U Her, R Cas, R Tau, Z Cyg, R Peg, U Aur and semiregular variables RT Vir and RX Boo) and 2 supergiants (S Per, PZ Cas). A considerable fraction of the sample stars (about one third) possess circumstellar maser emission in molecular spectral lines (OH, H20, SiO). Our aim was to elucidate the particularities of photometric characteristics of maser stars, such as period P, amplitude A, light curve asymmetry f = (M-m)/P, and, in prospect, to determine their status in course of their evolution on the asymptotic giant branch. An extensive comparison with non-maser LPVs was made.

Observations of the red, bright supergiant star a Orionis (Betelgeuse) during the past decade show significant variations in UV line fluxes and line shifts. Dupree et al. (1987) pointed out that the data from 1984-1987 indicates a periodicity of slightly more than one year in the B magnitude, UV continuum, and Mg II emission line fluxes. Our data shows that this conclusion applies to other strong and medium strong chromospheric lines as well. The variations seem to affect also the relative strength of lines within multiplets.

A review is given of the chromospheres of evolved stars with peculiar chemical abundances, emphasizing the observed dependence of chromospheric properties upon the evolutionary status of the stars. Some old and new physical processes which are potentially important in determining the observed chromospheric features are discussed.

The number of molecules detected at radio wavelengths in envelopes around red giants stands presently at 36. Among these OH and CO have proven to be the most useful for the study of the physical characteristics of a circumstellar envelope. The mass loss rate of the central star can be relatively accurately estimated and it appears possible to trace its evolution with time. Also fascinating objects in transition from the red giant phase to the planetary nebula phase are becoming observationally accessible.

The mass loss from evolved red giants is considered. It seems that red giants on the Asymptotic Giant Branch (AGB) are losing between 3 and 6 10-4 MΘ kpc-2 yr-1 in the solar neighborhood. If all the main sequence stars between 1 and 5 MΘ ultimately evolve into white dwarfs with masses of 0.7 MΘ the predicted mass loss rate in the solar neighborhood from these stars is 8 10-4 MΘ kpc-2 yr-1. Although there are still uncertainties, it appears that there is no strong disagreement between theory and observation. However, it could also be that we have not yet identified much of the source of the mass-loss from pre-white dwarf stars.

With the advent of the IRAS satellite a completely new class of stellar objects became evident; viz., objects which are obviously in a rapid transition from the very tip of the asymptotic giant branch (AGB) into the planetary-nebulae region. These so-called pre-planetary nebulae (PPN) are characterized by stellar spectra indicative of supergiants mainly of F and G spectral types in conjunction with a large infrared excess due to a cool circumstellar dust shell. At the same time theoretical calculations through the AGB, with the inclusion of mass loss, and also through the following evolutionary stages down to the white-dwarf sequence, became available (Schonberner 1979, 1983; Wood and Faulkner 1986). These calculations predict evolutionary lifetimes of several 1000 years in the transition region between the tip of the AGB and the planetary-nebulae region. Thus a direct comparison between theory and observation now appears possible in the very early phase of this post-AGB evolution.

Ten new carbon stars with oxygen-rich circumstellar shells have been discovered using IRAS LRS spectral class. The origin of these objects have been briefly discussed. It is likely that they may be a consequence of release of oxygen due to carbon condensation, though other scenarios cannot be ruled out.

A characteristic feature of carbon star spectra is the dramatic decrease in energy flux toward the violet. This flux decreases faster toward shorter wavelengths in N-type carbon stars than in either M giants or blackbodies of the same usual color temperatures, and an unknown violet opacity source in these carbon stars has consequently been the subject of speculation for the past 60 years. We investigate this longstanding problem of the violet flux falloff in cool carbon stars by directly testing, through synthetic spectra, the effects of many new and previously suggested opacity sources, based on currently available model atmospheres for carbon stars and M-giant stars. While several bound-free edges of neutral metals are important opacities, those of Na I at 2413 Å, Mg I at 2514 Å, and Ca I at 2940 Å are especially significant. Collectively, thousands of atomic lines, and in particular, the neutral metal resonance lines are the primary cause of the violet flux falloff. The severe violet flux decrease begins in cool carbon stars shortward of 4500 Å, roughly coincident with the redward wing of the extremely strong Ca I resonance line 4227 Å. The enormous line of Mg I at 2852 Å is one of the largest opacity sources in the near-ultraviolet region of these stars and influences the spectrum well into the visible. The pseudo-continuum of C3 and the photodissociation continuum of CH both play noticeable but secondary roles. Other opacities that may affect the violet flux include the photodissociation of other hydrides, particularly NH and SiH, which have dissociation energies similar to CH. Synthetic spectra from the carbon-star models computed both with and without polyatomic molecules fit nicely to the collected observations of the well-observed carbon star TX Pse.

We present near IR (H, K and L band) medium resolution (ƛ/Δƛ ∼ 600) spectra for a selection of 9 red giants which have previously been shown to exhibit anomalous dust emission as characterised by their IRAS LRS spectra. The objects observed (during UKIRT and AAT service time) include Carbon stars whose LRS spectra show the 9.7μm silicate feature and also M stars whose LRS spectra display an 11.3μm feature similar to that usually associated with emission from SiC dust grains.

Observations of CO in long period variables have been widely used to determine mass loss rates by applying models for CO line formation (e.g. Knapp and Morris, 1985) which use a simple method to take the impact from infrared radiation into account. Recent 00(2-1) and (1-0) observations of some more evolved OH/IR stars yielded much too low mass loss rates using these simple models, thus indicating that they cannot be extrapolated to far evolved AGB stars with optically thick circumstellar envelopes.

Circumstellar envelopes of cool giants and supergiants contain atomic and molecular gas, and dust. The charateristic spectral features of these different components can be observed at optical (atoms), at radio (molecules) and at infrared wavelengths (dust). Since the detection of circumstellar matter around giants and supergiants most studies concentrated on detemining mass loss rates from observations of a single component assuming steady mass loss during late stellar evolution. Nevertheless, the IRAS colour colour diagramme and evolutionary models rather point to a non steady evolution during the mass loss phase of the star.

IRAS 09371+1212 (sometimes called the Frosty Leo nebula) is up to now an absolutely unique object. Its stellar nature has been proved by the detection of circumstellar CO (J = 1-0) by Forveille et al. (1987), and by its stellar M giant spectral type, although its infrared colors and its visible appearance could have suggested a galaxy (Condon and Broderick 1986). It displays unique IRAS far infra-red colors which have been attributed by Forveille et al. (1987) to a strong emission in the 40-70 micron bands of ice. The presence of an extremely large quantity of ice is confirmed by extraordinary strength of its 3.1 micron absorption band (Rouan et al. 1988; Hodapp et al. 1988; Geballe et al. 1988).

ER Del = BD+8:4506 is a faint (m ≈ 10) irregularly variable PRG star classified spectral type S5.5/2.5 (Ake 1979). Spectra show H alpha in emission and a strong UV continuum with emission lines of C IV, Si III], and C III] — characteristic of symbiotic-like stars. Although a few other MS and S stars have hot companions, this is the first to show hydrogen emission lines in the optical region. It has been suggested (Smith & Lambert 1986; Little et al. 1987) that all Tc-deficient PRG stars are accidental; that is, they arose by mass transfer when the currently degenerate companion was itself an AGB star. Indeed, the Tc-deficient S stars HR 1105 (Ake, Johnson and Peery 1988) and HD 35155 (Ake and Johnson 1988) and the R8 star HD 59643 (Johnson et al. 1988) not only have hot, subluminous companions, but also are interacting systems - a sign the components are close enough for mass transfer to have occurred. An accretion disk is probably present.

The outer atmosphere of a cool red giant star is an ideal locale for the operation of line fluorescence processes. Low plasma densities imply low rates of collisional de-excitation and thus allow radiative decay of levels populated by selective radiative pumping. There are many strong sources of line radiation (i.e. possible pumps) and numerous possible upward transitions from highly populated low-lying levels of abundant elements such as Fe II, thus providing many chance coincidences between potential pumps and lines to be pumped. These conditions ensure that many of the chromospheric emission features observed in the UV spectrum of such a star are affected by fluorescence. Many of the observed emission features originate from energy levels populated solely by radiative fluorescent excitation, including strong lines of S I, O I, CO, Ni II, Si I, Fe I and Fe II, as well as weaker lines from Cr II and Co II. Important pumps active in these atmospheres include hydrogen Lyman alpha, and individual lines of 0 I, C I, Si II, Fe II, and Mg II. In the case of Fe II, there are many additional features arising from upper levels whose populations, although primarily maintained by collisions, are also significantly affected by radiative fluorescent excitation. In fact, there may be virtually no level in Fe II not affected to one degree or another by direct decays or cascades down from levels populated by fluorescence, driven either by Lyman alpha or, in some cases, by lines of Fe II itself (“self-fluorescence“).

We present a brief description of the main results obtained on the following topics: (1) Sulfur-bearing molecules in O-rich envelopes: SO2(104 and 160 GHz), H2S (169 and 217 GHz) in 15 stars, SO (86 GHz) in 7 stars (all new detections except H2S in 0H231.8+4.2). Excitation temperatures and line profiles could suggest in some stars a main contribution from external layers for SO2 and internal layers for H2S.

The bulk of meteorites consist of material that condensed out of the primitive solar nebula, and their isotopic composition is therefore similar to that of the sun. Unfractionated meteorites (carbonaceous chondrites), nevertheless, contain a small fraction of inclusions that have a distinctly different isotopie composition of carbon, xenon, krypton and other elements (Zinner et al., 1987). These inclusions are therefore believed to be of pre-solar origin, and their composition and mineralogical structure carries information about the place where they were born.