We present results of steady jet calculations in which the cooling and compression behind internal shocks leads to optical emission with an intensity pattern similar to the regular well-aligned emission knots characteristic of stellar jets.

The IRAS Point Source Catalog contains tens of thousands of stars with circumstellar shells. Most are cool stars surrounded by dust-rich envelopes, but many B and A stars are present with ionized circumstellar envelopes. IRAS and its main products are briefly discussed.

Eight of thirty-four previously unobserved IRAS sources were found to be relatively strong 1612 MHz OH emitters. Five of these emit at 1667 MHz. Of the eight half are high velocity range, population I type stars, the other half are low velocity range, population II type stars. The pump efficiencies are in the range 0.018 ≤ e ≤ 0.163.

Most of the sources noted by the Infrared Astronomical Satellite are, as expected, more or less normal cool giant stars, but a very large number of objects of earlier spectral type were also picked up. These cases must involve, in a variety of ways, circumstellar gas and dust. From a large-scale program of identification of the IRAS objects, some systematics emerge.

The contribution from the line emission of ions to the radiation at four IRAS bands has been estimated. The dust grains are likely to be mixed with ionized gas.

Some M Miras show the 9.7 and 20μm silicate emission. Recently, Vardya, de Jong & Willems (1986; hereafter VDJW) discovered using IRAS LRS spectra a weak broad emission feature ~12μm; this new feature may be silicate (VDJW). Here we investigate whether this can be due to H2O ice.

Using the IRAS catalog of low resolution spectra (LRS), we have analyzed the silicate features in emission and absorption, the carbon-rich emission features and the featureless spectra, in the range 8–22 μm. The sample sizes are large enough to allow average properties to be established (e.g. energy distributions), as well as correlations between luminosities, excesses, colors and coordinates, and histograms and galactic distributions, all with a good degree of confidence.

Recent measurements in far-infrared (12–200 μm) by IRAS (Aumann et al, 1984) and KAO (Harper et al, 1984) have shown evidence of infrared excess in the AO star Alpha Lyrae. This excess in the infrared radiation was interpreted as due to circumstellar dust constituted by grains much larger (≳10 μm) than those of interstellar origin. However, there appears to be a difficulty in interpreting these observations in terms of a single or a combination of several temperatures of dust. We have formulated a simple model to explain the far-infrared data in 12–200 μm region available with IRAS and KAO observations in terms of multiple dust shells of different temperatures. We have shown that dust grains of size ~10 μm can explain the excess emission in the entire region 12–200 μm with temperatures of 130 ± 15 K and 40 ± 5 K. The far-infrared fluxes seem to vary as λ−β with μ = 3.0–3.8, in quite contrast with small sized grains. This work lends further support to the interpretation of far-infrared excess in ∝ Lyrae as emission from circumstellar shell with large (∼10 μm) size grains. Our calculations show, however, that grains of size larger than 10 μm cannot explain the fluxes.

IRAS observations of sources identified with cepheid variables are used to give estimates of observed mass-loss rates for those stars.

This article reviews recent radio observations of maser emission from OH, H2O and SiO molecules in the circumstellar envelopes of OH-IR sources. The different radio lines require different conditions for their excitation, and each therefore probes different regions in the circumstellar envelope. For some stars radio interferometer maps of several maser lines are now available, and a consistent picture of the envelope structure is beginning to emerge.

A search has been made for 1612 MHz OH maser emission from OH/IR stars in the Magellanic Clouds. Candidate objects were selected mainly on the basis of their 25μm flux densities and the 25 to 12μm flux ratio as given in the IRAS point source catalog; two known long-period variables and two HII regions (30 Doradus and N159) were also examined. One OH source (IRAS 04553-6825) was detected, this being the first OH/IR star found in the Magellanic Clouds. Upper limits were placed on the flux for 17 other sources. The expansion velocity of the circumstellar material surrounding IRAS 04553-6825, as indicated by the OH peak separation of 11 km s−1, is surprisingly small compared to Galactic sources of similar bolometric and OH luminosity. The OH intensity of IRAS 04553-6825, and the upper flux limits placed on many of the other objects examined, indicate that Magellanic Cloud OH/IR stars do not emit OH as strongly as their Galactic counterparts of similar 25μm/12μm flux ratio. Both the low expansion velocity of IRAS 04553-6825 and the low OH intensity of the Magellanic Cloud infrared sources may be explained by the low metal abundance in the Clouds.

Two studies of OH maser emission in envelopes of late type stars -miras (3) and OH/IR objects- have been performed with the Nancay radio-telescope. Mainly, the OH miras are found with thicker dust envelopes than the non OH ones. A sample of unidentified IRAS point sources selected on their colors has been observed. We have detected 46 new OH sources. The OH detection rate is a function of the galactic longitude and of the IRAS spectral classification.

We have begun a co-ordinated programme of high angular-resolution radio and infrared measurements to study the physical structure of the circumstellar envelopes surrounding high mass-loss OH-stars. Here we give near-infrared (NIR) angular diameters for 5 stars. For each of these stars the spatial distribution of the OH maser emission at 1612 MHz or 1665 MHz has been previously mapped.

Two large bipolar bubbles emanating from the OH/IR star OH0739-14 have been discovered. Interference filter images show that the bubbles are emission-line objects, and longslit spectra reveal that the bubbles are expanding, with a radial velocity-difference between the bubblefronts of over 200 km/sec.

The profiles of maser lines from spherically symmetric expanding shells are computed with the help of a simple model. It is shown that a beaming effect must be taken into account in order to correctly account for the OH profiles of type II OH stars.

Circumstellar envelopes of cool giants and supergiants are composed of different parts which can be observed through characteristic spectral features. One is SiO-maseremission which is typically found in Miras and some semi-regular variable supergiants. These masers are thought to originate in those parts of the upper atmosphere or lower circumstellar envelope where still no mass-loss takes place. Location and trigger mechanism was still not definitely clear.

After a comparison of conventional photospheres with circumstellar envelopes the radiation in spherical shells is considered. We discuss the transfer equation and a new quasi-exact solution in term of the transition matrix-Various methods used for the numerical evaluation of the specific intensities are summarized. The general properties of the radiation fields and some recent detailed model calculations are briefly reviewed.

From March 1980 to December 1983, the author took part in regular observations of variability of maser radio emission in the H2O line at 22 GHz. The observations were carried out at the 22-meter radio telescope of the P. Lebedev Physical Institute (USSR Academy of Sciences) in Pushchino (Moscow Region). The interval between consecutive observational sessions was usually 1.5–2 months. The observational program included 21 late-type variable stars (Miras and SRs): R Aql, RR Aql, RT Aql, SY Aql, U Aur, NV Aur, RX Boo, VY CMa, S CrB, KY Cyg, NML Cyg, U Her, W Hya, X Hya, R Leo, U Lyn, U Ori, UU Peg, VX Sgr, RS Vir, RT Vir. The results for eight stars ending June 1982 were published by Berulis et alt (1983). A comparison was made between the time dependences of the H2O line radio flux F and the curves of visual and near-infrared brightness of the stars. Miras (R Aql, R Leo, U Ori, U Aur), as a rule, have a rise in F connected with the visual maximum (phase 0), the maximum F occurring at phases 0.1–0.2 (see figure for an example). Not all visual maxima (only one out of each two or three) are accompanied by H2O flares. This Miras! behaviour was also noted earlier in the H2O line by Berulis et al. (1984), Gómez Balboa and Lépine (1986), as well as in the SiO maser line v=1, J=2−1 by Nyman and Olofsson (1986).

In this review I consider what clues the data are providing us concerning the mass loss from late-type stars. I consider in turn the major classes of mass-loss mechanisms (thermally-driven winds, radiatively-driven winds, and wave-driven winds), and consider whether the empirical mass loss rates and other data are consistent with any of these mechanisms acting alone. It is likely that several mechanisms act together to produce the large mass loss rates in the Mira and non-pulsating M supergiants. Studies of the solar atmosphere suggest that thermal bifurcation driven by molecular condensation instabilities may play a critical role in cooling the atmospheres of luminous cool stars and forming silicate dust. It is possible that several metastable modes of atmospheric structure may exist for a given set of stellar parameters.

The goal of this paper is to provide a framework for thinking about the various physical processes that may play significant roles in driving the massive winds of cool, low-gravity stars. First, some general theoretical considerations involving mass, momentum, and energy balance are discussed. Next, the value of the solar wind as an analog for these late-type stellar winds and for related astrophysical flows is briefly examined. Finally, four specific mass-loss mechanisms are discussed, and the possible importance of each of these mechanisms for massive winds from cool, low-gravity stars is evaluated.