Modern ideas about the evolution of galaxies emphasize the role of accretion and mergers. What is the influence of environment on the chemical evolution of galaxies? Skillman et al. (1996) studied the abundances in spirals in the Virgo cluster, based on spectra of GEHRs. They found that galaxies in the cluster core, with marked H I deficiencies, were more metal rich than spirals with normal H I content located in the periphery of the cluster and in the field. They suggested that this results from the curtailment of metal-poor infall onto galaxies in the cluster core, where they are immersed in the hot cluster medium. In contrast, infall continues onto spirals in the cluster periphery and in the field, restraining the increase in abundances with time.
The question of infall onto the Milky Way relates to the proposal by Blitz et al. (1999) that some of the Galactic high velocity clouds (HVCs) observed in H I, and sometimes seen in heavy element absorption lines, are actually a population of dwarf galaxies or gas clouds belonging to the Local Group. The inferred accretion rate of these clouds onto the Milky Way is interesting in the context of chemical evolution. The Blitz model draws together a number of aspects, including cosmological predictions of dwarf galaxy counts and the dynamical evolution of the Local Group. However, the massive H I clouds involved in this model have not been found in other groups of galaxies (e.g., Zabludoff 2001; Zwaan 2001).
Also of interest is chemical evolution in low density environments. Peimbert & Torres-Peimbert (1992) obtained spectra of emission-line galaxies in the Boötes void. Most appear to be irregular galaxies with H II regions ionized by OB stars. Several of the objects have rather low N/O values, compared to a small sample of nearby H II regions with similar O/H, including the LMC. As noted by Garnett (1990), a low N/O value may occur in a young galaxy or in one whose oxygen has been enriched by a recent starburst, because N comes largely from lower mass stars with longer lifetimes than stars producing O. Thus Peimbert & Torres-Peimbert suggest that the Boötes objects with low N/O may be young galaxies; and they note that late collapse of density clumps to form galaxies may be a natural occurrence in a low density environment. However, the range of N/O values shown by Garnett (1990) encompasses the Boötes values, albeit for slightly lower O/H.
The study of abundances in galaxies in low and high density environments deserves more attention. These studies will require spectra of adequate sensitivity and wavelength resolution together with measurements of the gross properties of the galaxies, so that comparisons can be made in a way that isolates the effect of environment.