Much has happened since the book first came out in 1997. Cosmology has been transformed by balloon and satellite studies of the microwave background and by studies of distant supernovae. Host galaxies of γ-ray burst sources have been identified and some of their properties revealed. Cosmological simulations have been very successful in accounting for the large-scale structure of the Universe, although they are still challenged by observed element:element ratios suggesting that the largest galaxies were formed rapidly a long time ago, limiting the time available for their formation by mergers. The coming of 10-metre class telescopes, supplementing the Hubble Space Telescope, has led to enormous advances in abundance determinations in stars of all kinds and in galaxies, notably at high redshifts. Some stellar atmospheres can now be modelled by ab initio hydrodynamical simulations which account for granulation and eliminate the need for ad hoc parameters describing ‘macro-turbulence’ and ‘micro-turbulence’, leading to increasingly sophisticated abundance determinations. Nevertheless, simple analytical treatments retain their usefulness because of the insight they provide into the essential ingredients of more elaborate numerical models, whether of stellar atmospheres or of galactic chemical evolution.

I thank Monica Grady, Chris Tout and Max Pettini for critically reading through the revised Chapters 3, 5 and 12 respectively, and Mike Edmunds for continued cooperation and enlightening discussions. I owe particular thanks to my wife Annabel Tuby Pagel for her loving care during difficult times.