The following remarks cannot possibly be a summary of the wealth of data and ideas which were presented and discussed during this symposium. May I rather indicate a broad-brush picture for late stellar evolution as it seems to me to emerge from the many diverse investigations we have heard about. Before doing so, however, I would like to touch on one subject that is not directly connected with late evolution phases but may turn out to be relevant to it. A. MISSING SOLAR NEUTRINOS I consider the

more »

... negative result of the solar neutrino test a sufficiently serious matter that I would not want simply to ignore it. It is true that our present theory of stellar evolution has so many substantial contacts with observational data that it is hard to believe that any required corrections would entirely alter the picture. Nevertheless, we surely cannot feel safe as long as a test as fundamental as the solar neutrino test is strongly discordant with our predictions. Resolution of this discrepancy may lie in any one of three fields: the solar neutrino detection experiment itself, the nuclear physics built into our stellar structure theory, or the rest of the physics and mathe matics that goes into our model stars. For each of us it is easy to persuade himself that the resolution of the discrepancy must lie in one of the two fields other than the one in which our expertise lies. However, if all of us follow this natural reaction and in consequence do nothing about the neutrino discrepancy, its resolution is not likely to come forth soon. Accordingly, it would seem to me more effective if we all accepted the working hypothesis that the actual problem lies in our own field of expertise, kept an active watch for any new ideas relevant to this critical discrepancy, and tried to work them out whenever they lie in our field of specialty. B. EVOLUTION CLASSES OF STARS Now to the broad-brush picture of advanced stellar evolution. To sort out the great variety of phenomena we have heard about and have discussed during the past three days it would seem to me useful to consider all stars in terms of four evolution classes. Even though these classes will divide all stars more or less according to their initial mass, it would seem to me more useful to base the definition of the classes on the nuclear processes dominating the entire life of a star, most specifically the late part of its life, rather than on fixed mass limits. C. FEATHERWEIGHT STARS Under this name we might understand that class of stars which during their entire life never burn nuclear fuel -except possibly such minor fuels as deuterium and lithium. Such a definition implies masses of less than about 0.07 A/©. This class of stars has Tayler (ed.), Late Stages of Stellar Evolution, 256-258. All Rights Reserved.