The idea that planetary nebulae (PN) originated from outer layers of red giants goes back to Shlovskii (1956). This hypothesis was supported by Abell and Goldreich (1966) who argued convincingly that red giants are the most likely progenitors of PN. Although this is generally accepted today, the details of the transition from red giants to PN remain in controversy. It was pointed out by Paczyński (1971 a) that PN progenitors must have similar luminosities to central stars of PN, and therefore are likely to be late-type supergiants undergoing double-shell burning. The advent of infrared astronomy led to the discovery that most, if not all, late-type giants and supergiants are losing mass at rates of 10-6 - 10-5 Mʘyr-1 (Gehrz and Woolf 1971). Such mass loss rates greatly exceed the nuclear burning rate (6 x 10-8 - 5 x 10-7 Mʘyr-1 for stars with core masses between 0.6 and 1.2 Mʘ) and must be the dominant factor in the late evolution of intermediate mass stars. The observation of white dwarfs in open clusters implies that up to 6 Mʘ can be lost during the red-giant phase (Romanishin and Angel 1980). Since the observed masses of PN are no more than a few tenths of a solar mass, the existence of massive circumstellar envelopes formed by steady mass loss must have an effect on the formation of PN.