Episodes in the Discovery of Variations in the Chemical Composition of Stars and Galaxies

Five episodes from the 1950s are reviewed that led to a fundamental change in prevalent hypotheses of how, when, and where the chemical elements were formed. The episodes are (1) the discovery that all stars do not have the same abundance ratios of heavy elements to hydrogen, and the inventions of methods to detect and measure the variations, culminating in the fundamental paper by Chamberlain and Aller; (2) Roman's discovery of the ultraviolet excess in 17 high-velocity F subdwarfs, and the subsequent discovery of the excess in globular cluster stars; (3) the development of the understanding of the physics of the H-R diagram and the evolution of stars, explaining the various sequences in that diagram; (4) Hoyle's idea of the recyclying of the products of nucleosynthesis from stellar interiors of evolving stars through the interstellar medium by stellar mass loss and supernovae explosions; and (5) the development of the theory of the various nuclear processes in stellar interiors by a small number of different buildup mechanisms (equilibrium burning, slow neutron capture using normally produced neutrons, proton capture, and fast neutron capture in supernova explosions) that can produce all the chemical elements. This Essay is one of a series of invited contributions which will appear in the PASP throughout the year 2000 to mark the upcoming millennium. (Eds.)