Massive stars provide the major source of chemical enrichment of the Galaxy. Their structure and evolution, and in particular their mass-loss rates, depend on metallicity, Z, helium abundance, Y, and the degree of chemical segregation that has taken place within the star. This we know by comparing H-R diagrams of supergiants in the Small Magellanic Cloud (Z below solar by 0.6 dex), the Large Magellanic Cloud (Z down by 0.3 dex) and the solar vicinity, which indicate that nucleosynthesis is different in the three cases. Similar trends exist between carbon stars, M stars and red giants. They have a crucial bearing on whether the stars will become Wolf-Rayet stars, or supernovae. Nobody yet knows how the envelopes are formed around Be stars; detailed studies of pulsations of Cephei stars will enable constraints to be put on the structures of those stars and their dependence on Y and Z. It will also be important to calibrate models of asymptotic giant stars. The present state of stellar modelling is such that there is an uncertainty of a factor of at least 2 in the rate of heavy-element production. Seismic calibration will reduce that uncertainty substantially.
Another important question we propose to address concerns the cosmological issue of the evolution of early galaxies. Radio observations in the K-band can be adequately interpreted only with a better knowledge of the evolution of the massive stars that dominate the light output (Longair, 1993). These stars have low Z. Most low-Z stars in our Galaxy also have low mass, which is of major concern. However, beyond 1.5 kpc outside the solar orbit, the metallicity gradient steepens, and at 5 kpc there are young stars forming with a metallicity substantially below solar. After accounting for interstellar reddening, one might expect to find some stars with ( would be 7.5 if unreddened) that would be accessible to detailed seismological study; but we need to investigate whether or not the fields are sufficiently sparse to isolate individual target stars. UV spectral measurements of all such stars are important for measuring their winds.