Welcome to this Joint Discussion on stellar atmospheres and interiors. For stellar structure, the atmospheres are of importance for two reasons, first as a source of information on effective temperature, gravity (or mass: luminosity ratio), chemical composition and dynamical effects, and secondly as an outer boundary condition for stellar models.

In the last few years there have been substantial advances in stellar atmospheres as a result of both improved theories and new measurements. In the particular field of abundance determination, non-LTE theory has at last come of age and provided a certain number of corrections that are both realistic and significant, for example in reconciling the neon abundance in B stars with that in nebulae and solar flare particles. Solar abundances have been improved by the provision of better oscillator strengths and better treatment of line broadening, generally with the effect of leading to still closer agreement with carbonaceous meteorites. Nevertheless, he would be a bold man who claimed to know the initial solar abundance parameters Y and Z to better than, say 25 per cent, even if we grant that the photosphere is a true sample of the initial interior abundances - an assumption that is now being questioned (again) in view of the solar neutrino problem.