|Annu. Rev. Astron. Astrophys. 1981. 19:
Copyright © 1981 by . All rights reserved
In the foregoing discussion we have reported many facts and not a few speculations. What are the main conclusions that one can draw from these? Broadly speaking, a number of expectations from stellar evolution and nucleosynthesis theory are confirmed: the abundance of heavy elements in the ISM and in the corresponding stellar populations increases with time at a given place (the solar neighborhood) and with the degree of conversion of gas into stars at different places at a given time (the present), though with a "prompt initial enrichment" in our own neighborhood that can be accounted for by taking into account the slow formation of the galactic disk (Lynden-Bell 1975, Larson 1976, Tinsley & Larson 1978, Chiosi & Matteucci 1980). The differential behavior of elements and isotopes is qualitatively in accordance with this trend, and some of the quantitative details - e.g. those concerning helium and C,N,O isotopes - permit interesting constraints to be placed on cosmology, initial mass functions, and the end products of stellar evolution, although many of the details are still not satisfactorily explained.
The existence of differences in heavy-element composition between and across galaxies is now well established, although the absolute scale of the variations remains somewhat uncertain. Whether the mechanism causing differences between galaxies is the same as that which causes gradients across galaxies is unknown. The most obvious candidate for the first effect is mass-dependent gas loss (e.g. Hartwick 1980), but accretion and/or inward gas flows may be important for the gradient effects (e.g. Tinsley 1980, Mayor & Vigroux 1981).
The influence of gas loss is certainly evident in clusters of galaxies, since elliptical galaxies probably contribute a significant amount of heavy-element-rich gas to the intracluster medium (Larson & Dinerstein 1975) where it is visible in X rays (Culhane 1978). The correlation with shape and velocity dispersion (Terlevich et al. 1981) suggests that such processes may be influenced by the local gravitational potential.
More data on a wide range of Hubble types are undoubtedly required. More accurate photometric colors for ellipticals and SO's, and the correlation of color gradients with other properties, would be useful - particularly to find out if a significant fraction of ellipticals shows no color gradient. Further HII region analysis of spirals, particularly of early types, is needed. This should be quite possible using direct imaging techniques if the methods of analysis using only strong-line intensities can be unambiguously established as good abundance indicators. With the advent of the Space Telescope (e.g. Pagel 1979c) examination of the population metallicity structure of some nearby systems will become feasible and population synthesis placed on a broader base.
Perhaps the most surprising fact about abundances in galaxies is not their dissimilarities, but the considerable similarities of abundance distribution as a function of mass of the system, and the observation of gradients in both spirals and ellipticals. The differences between galaxies of different types, masses, or environments undoubtedly have much to tell us about galaxy formation and evolution. The similarities (including QSO abundances) argue that all galactic systems share the same basic nucleosynthetic processes.