Star-to-star variations in abundances of the light elements carbon, nitrogen,
oxygen, and sodium have been observed in stars of all evolutionary phases in
all Galactic globular clusters that have been thoroughly studied. The data
available for studying this phenomenon, and the hypotheses as to its origin,
have both co-evolved with observing technology; once high-resolution spectra
were available even for main-sequence stars in globular clusters, scenarios
involving multiple closely spaced stellar generations enriched by feedback from
moderate- and high-mass stars began to gain traction in the literature. This
paper briefly reviews the observational history of globular cluster abundance
inhomogeneities, discusses the presently favored models of their origin, and
considers several aspects of this problem that require further study.Comment: AN in press, 9 pages, Highlight talk from Astronomische Gesellschaft
2010 annual meetin