Over the past decade a comprehensive, semiquantitative theoretical scenario for the final evolutionary stages of low and intermediate mass stars has been progressively elaborated and refined. It concerns the envelope ejection terminating the Asymptotic Giant Branch (AGB) phase, the AGB to Planetary Nebula (PN) transition, the fading and possible rejuvenation of PN nuclei, the formation processes of hydrogen-deficient stars, and the final production of white dwarfs (WD) of the DA and non-DA varieties (Renzini 1979, 1981a, 1981b, 1982, 1983, Iben & Renzini 1983, Iben et al. 1983, Iben 1984, 1985, 1987, Iben & Tutukov 1984, Iben & MacDonald 1985, 1986). In developing this scenario several important results of stellar evolution and hydrodynamical calculations have been incorporated, including in particular those of Paczynski (1971), Wood (1974), Härm & Schwarzschild (1975), Schönberner (1979, 1983), and Tuchman, Sack & Barkat (1979).

Over the past decade, we have come to realize that mass loss on the asymptotic giant branch (AGB) plays a significant role in the formation of planetary nebulae (PN). Mass ejected during the AGB can now be observed in haloes of PN and we believe that the main shell of PN is formed by the interaction of this material with a later-developed central-star wind. In this review, we show that the evolution from AGB to PN can be traced in a continuous infrared sequence. This sequence predicts properties of proto-PN which allow them to be identified.

According to the multiple winds model a planetary nebula forms as the result of the interaction of a fast wind from the central star with the superwind that had previously been emitted by the progenitor star. The basic theory which deals with the spherically symmetrical case is briefly summarised. Various improvements are then considered in turn. A better history is clearly needed of the way that the central star becomes hotter, it is unrealistic to make the assumption that the superwind is spherically symmetrical, and finally there are likely to be important instabilities at some of the interfaces in the PN, notably that between the shocked superwind and the HII layer. These changes in the theoretical description produce a better understanding of the conditions in the outer parts of a PN and of the nature of its general shape, and they should lead to an explanation for the occurrence of high speed motions, and of highly ionized species and high excitation spectral lines.

A knowledge of the formation rate of planetary nebulae is crucial to our understanding of a wide range of physical environments and processes, and as such has been fitfully investigated over the last forty or so years. The elemental composition of the interstellar medium, for instance, and in particular the seeding of the IS medium by CNO and s-process materials, turns out to be rather sensitively dependent upon the local rate of PN formation χ(PN) (Salpeter, 1978, Pottasch, 1984; Tinsley, 1978; and Wood et al, 1983). Similarly, and to varying degrees, the ionisation balance of the I.S. medium (cf. Pottasch, 1984; Salpeter, 1978), composition and number density of I.S. grains (Natta and Panagia, 1981), and ultimately the galactic star formation rate (cf. Miller and Scalo, 1979) can all (more or less directly) trace a dependency upon χ(PN). Finally, χ(PN) has proven historically important in establishing the evolutionary status of PN (cf. Abell and Goldreich, 1966), and is expected to be directly related to the rate of formation of white dwarfs, the death-rate of AGB-type stars (Mira variables, OH/IR sources), and the main-sequence turn-off rate in the mass range 2 ≲ M/M⊙ ≲ 10, some 1010 years ago.

High and low-resolution optical and near-IR spectroscopy of the candidate proto-planetary (or very young PN) Vy 2-2 (P-K 45 - 2°1) is reported. This object has associated OH maser emission and an angular diameter of only 0.4 arcsec, found from VLA and optical speckle interferometry. Empirical analysis gives the values Ne ≈3 × 105 cm−3, Te=11000(±1500)K. The electron temperature is quite uncertain because of the high density. Abundances of He, C, N, O, Ne and Ar are reported; the carbon abundance is uncertain as it relies on the C II λ4267Å line, since the object is too highly-reddened (c=1.8 ± 0.2) to be observed with IUE. We find He/H=0.10, O/H=4 × 10−4 and C/O=0.8. The HI Zanstra temperature is 38 000 K (for black-body). The spectrum shows broad stellar lines of He II λ4686, C III λ4647 and N III λ4640; the central star may be of type Of.

Vy 2-2 is, to our knowledge, the only planetary nebula exhibiting OH emission, indicating that it is a very young PN still enveloped by the neutral shell of the progenitor AGB star.

A sample was selected from the IRAS Point Source Catalogue based on the following selection criteria: very red (“cold”) IRAS-colours: roughly F25/F12 > 2.5 and F60/F25 < 1.2; and low IR-variability: VAR < 30. These non-variable IR-sources may be stars that have evolved beyond the AGB (Asymptotic Giant Branch); a large fraction (40%) is associated with known planetary nebulae (Van der Veen and Habing, 1987, Astron. Astrophys., in press). To determine the nature of the other 60% additional observations were made mainly in the infrared: 1–13 μm, during 4 observing runs: ESO (La Silla, Chile) in July 1986 and June 1987; UKIRT (Hawaii) in August 1986 and June 1987. A total number of 58 sources was observed. A summary of the observations: -IR broad band photometry at 1.2, 1.6, 2.2, 3.8 and 4,6 μm for all 58 sources. -IR broad band photometry at 8.4, 9.7 and 12.8 μm for 19 sources. -IR small band photometry for 4 sources in the ranges 2–2.5 μm and 3–3.5 μm. -IR spectroscopy for 10 sources in the ranges 2–2.5 μm and 3–3.5 μm, -V, R, I observations (0.55, 0.7 and 0.9 μm) for 5 sources associated with a star of visual magnitude 8–9. These observations were carried out by D. de Winter (Amsterdam) with the 0.5-m ESO telescope at La Silla (Chile). -Walraven photometry (0.32, 0.36, 0.38, 0.43 and 0.54 μm) for 21 stars brighter than V = 15 and within 10“ from the IRAS position. These observations were carried out by M. van Haarlem (Leiden) with the 0.9-m Dutch telescope at La Silla (Chile).

IRAS 17516-2525 is a cool object at infrared wavelengths between 1 and 60μm. Spectroscopy in the wavelength ranges 2 - 2.5μm and 3 - 3.5μm shows the presence of Brα, Brγ, Pfγ emission lines and of a weak C2H2 absorption band, all clearly associated with the IRAS source. At the infrared position a faint star with a visual magnitude of about 20 is found.

A small number of bipolar planetaries -including M2-9, M1-91, GL 618, and M2-56- exhibit very strong emission lines of low ionization species such as [O I], [N I], [N II], and [S II]. Most previous authors have attempted to analyze the spectra of these objects assuming that they are photoionized by their central stars. Closer examination, however, suggests that a different excitation mechanism may be at work -that of shock heating.

The IRAS infrared colors, (60 - 25)μm and (25 - 12)μm, allow efficient identification of Type II OE/IR stars. We present Arecibo1 OH (1612 MHz) observations of 474 IRAS point sources chosen to define the exact regions of the two-color diagram occupied by OH/IR stars. Our observations are complete within the boundary regions of the two-color locus and within the region, 16h right ascension < 22h, 0° < declination < 37°. The sensitivity of the Arecibo telescope allows the identification of many weak sources that would not have been detected by previous surveys and the weak end of the masing phenomenon has been studied for the first time.

We have initiated a near infrared photometric study of OH/IR stars and proposed protoplanetary nebulae, using the 2.12 m telescope and the IR photometric system at the Observatorio Astronómico Nacional in Baja California, México.

Eder, Lewis, and Terzian (1987) examined ∼ 400 sources from the IRAS Point Source Catalogue with colors appropriate to OH/IR stars, for the presence of 1612 MHz emission. We examined a proportion of these objects at Effelsberg for the presence of water-maser emission. In sources with |bII| > 10° which are therefore relatively local, we find a 68% detection rate for water-masers among objects associated with 1612 MHz masers, as opposed to a 17% detection rate among sources with similar colors but without 1612 MHz emission. Those conditions in a circumstellar shell that favor the presence of water-masers also favor the presence of a 1612 MHz maser. These results are consistent with most Type II masers being associated with water-masers. Since Cooke and Elitzur (1985) show that water-masers are collisionally excited, this result excludes stirring of the envelope by a companion star with an associated loss of velocity coherence, as the primary cause for the existence of the color-analogue sources without 1612 MHz masers. We discuss an alternative scenario.

Some dependences between the parameters for approximately 150 stars, of which 81 are sources of maser emission in molecular lines, are constructed. The following parameters are considered: period P, asymmetry (M-m) of the visual light curve, visual amplitude A, color index (I-K). We use the data of the General Catalogue of Variable Stars. For the stars Z Cyg, R Tau, R Peg, RT Vir, RX Boo, PZ Cas, U Her, and R Cas, some parameters were determined by the authors.

A number of high-galactic latitude supergiant stars of intermediate spectral types have been suggested to be proto-planetary nebulae (PPN), such as the 89 Her and R CrB stars. A number of these stars along with some IRAS sources expected to be PPN -either sources with IRAS low Resolution Spectra (LRS) showing features which may indicate unusually cool dust shells or unusually red IRAS sources for which CO emission from a circumstellar envelope has been observed- where chosen for study. The IRAS 12/25 and 25/60 μm colours of 32 such stars from 3 groups on a colour-colour diagram. Class I show colours similar to ordinary stars; Class II have a 60 μm excess but have normal 12/25 μm colours; Class III are much redder than ordinary stars.

Maser emission has been observed at the frequency of the (0,2,0) J - 1 - 0 line of HCN (89088 MHz, energy 2050 K above the ground state) in 7 C-rich circumstellar envelopes: CIT 6, S Cep, IRC+ 50096, IRC+30374, FX Ser, AFGL 2513 and IRAS 17581-1744. This is the second molecule showing strong maser emission in millimeter lines, and the first strong maser ever observed in a C-rich circumstellar envelope. The emission is particularly strong (70 Jy) in CIT 6 because of its proximity (Guilloteau, Omont, and Lucas 1987). The masers in the stars have luminosities and linewidths (FWHM about 1 km/s) similar to CIT 6; they are also blueshifted by a few km/s with respect to the velocity of the star given by the central velocity of the ground-state HCN emission (Lucas, Guilloteau, and Omont 1987). These properties are somewhat similar to those of SiO masers in O-rich stars, although the intensities of HCN masers are weaker.

The lowest mass observed for a nucleus of a planetary nebula (NPN) is about 0.55 M⊙ (Weidemann and Koester, 1983, Schonberner, 1983). Hence, Lower mass WD's should have been produced without going through the phase of a visible PN ejection. Recently, Harpaz et al. (1987), have shown that very low mass WD's (up to 0.45 M⊙) can be formed by a single star evolution from red giant branch (RGB) stars, due to mass loss along the RGB. It turns out that WD's in mass range of 0.46–0.55 M⊙ formed by a single star evolution should be formed from the AGB, without an observable PN.

Polarization mapping of some reflection nebulae (e.g., NGC 6729) reveals parallel bands of polarization vectors across pre-main sequence stars (Ward-Thomson et al. 1985, Mon. Not. R. Astr. Soc., 215). These authors have suggested a model in which the bands would be. Explained by dust discs, the grains being aligned by toroidal magnetic fields ∼ 102 μG.

Spectra of the nuclei of two planetaries show what appear to be features caused by fast stellar winds as they encounter the surrounding nebulae. Superimposed upon the high velocity (3670 km s−1) ultraviolet P Cygni profiles of Abell 78 are low velocity absorption lines that likely arise from a density enhancement in the wind as it brakes and builds up against the inner edge of the nebula. The deepest portions of these narrow absorptions fall at −78 km s−1 for N V and O V and −26 km s−1 for C IV, which implies a gradient in the decelerating wind, as does the profile of the strongest C IV line. The lower value may be related to the expansion velocity of the inner helium-rich nebulosity, which we associate with a sharp absorption feature. Another density enhancement, evidenced only by absorption lines, appears at −250 km s−1, and may be caused by a rebound shock of the sort envisioned by Kahn (IAU Symposium No. 103, Planetary Nebulae, 1983, 305) and Okorokov et al. (Astr. Ap., 1985, 142, 441).

Self-consistent mass-loss solutions are computed for post-AGB stars, and the evolution of the central star is followed by utilizing the sequences of the steady mass-loss and the static solutions. The mass loss is driven by the radiation pressure gradient and the matter is accelerated in the inner region to the photosphere. The optically thick wind occurs when the H-burning luminosity is larger than the Eddington limit at the surface region. Such a situation is realized in the low-temperature region of the H-R diagram. Therefore no wind solution exists when the surface temperature becomes high enough.

Very high spectral resolution slit spectra have been used to investigate the kinematic structure of the double shell planetary nebula NGC 2392, the Eskimo Nebula. Each shell produces a spectrum corresponding to the projected radial velocity, a pattern which varies according to the position angle of the spectrograph's entrance slit. Multiple slit observations were used to determine that the inner shell is an incomplete prolate spheroid pointed almost at the observer and missing the tip ends. The outer shell is nearly spherical and more uniform.