Observed mass loss rates and Galactic stellar population distributions are used to estimate the rate of injection of stardust into the ISM. M stars and RLOH/IR Stars produce most of the silicates; most of the carbon and SiC comes from carbon stars. WR stars, novae, and supernovae may eject dust with chemical anomalies. There is little observational evidence for a major stellar source of hydrocarbon grains. The Galactic dust ecology considered by comparing stellar dust injection with depletion by star formation and supernova shocks suggests that dust grains are produced by accretion in molecular clouds at 1 to 5 times the stellar rate.

I review the observational evidence relevant to the composition of dust ejected from circumstellar shells. Emphasis is placed on those types of stars believed to be the principal sources of refractory stardust in the interstellar medium. In stars with normal abundances (O > C), the predominant species appears to be an amorphous silicate, characterised by smooth spectral features at 9.7 and 19μm, a far infrared emissivity close to unity, and a featureless ultraviolet extinction curve. Similarly, the characteristics of carbon stars (C > O) are consistent with the presence of amorphous carbon rather than graphite. Silicon carbide is a minor constituent of carbon-rich stardust and is therefore unlikely to be an important ingredient of interstellar grains. A major goal for future studies will be to establish the evolutionary link between amorphous carbon stardust and the carriers of the unidentified infrared emission features observed in planetary nebulae.

The basic concepts of astrophysical dust formation are surveyed. The potentials, the limitations and the shortcomings of the various approaches are discussed.

The need to replenish the reservoir of interstellar dust grains that are continuously being destroyed in the interstellar medium, and the presence of isotopic anomalies in meteorites, suggest that supernovae may be important sources of interstellar dust. Infrared observations of supernovae or their unmixed ejecta may provide the first direct evidence for newly-formed grains in this environment. The recently discovered supernova, SN 1987A, currently offers the best prospects for observing the actual process of dust formation in a supernova.

In contrast, supernova remnants constitute the most important grain destruction mechanism in the galaxy. Interstellar dust swept up by the expanding blast wave is predominantly heated and destroyed by collisions with a shocked x-ray emitting plasma. Infrared observations of remnants can therefore provide valuable information on the interaction between dust particles and a hot gas. This interaction can best be studied in supernova remnants detected with the Infrared Astronomical Satellite (IRAS). This paper reviews what we have learned so far on the subjects of grain formation and destruction by studying the infrared emission from supernovae and supernova remnants.

All participants at this meeting agree that it has been very exciting, this for three reasons. First, it has put together astrophysicists, laboratory and theoretical physicists and also chemists, thanks to the interdisciplinary nature of the field of interstellar dust. Second, we have had a very exciting confrontation between meteolitiscists, and specialists of interstellar, and cometary dust, and we saw growing evidence that interstellar dust is a pervasive solar-system material, although the history of this dust may have been peculiar. Third, last but not least, there has been in the recent years a wealth of new data on interstellar dust emerging from observations in various wavelength ranges, especially the infrared, and a lot of new laboratory data and sophisticated theories that allow us to interpret the observations. Clearly the study of interstellar dust, which had remained in its infancy for such a long time, is becoming a mature and much less speculative field.

These papers, presented at IAU Symposium No. 135 on Interstellar Dust, are published as a NASA Conference Publication (NASA CP 3036). Copies can be ordered from the National Technical Information Service, Springfield, Virginia, 22161, USA.