GALAXIES, STELLAR CONTENT ROBERT W. O'CONNELL Stars can have lifetimes that exceed the age of the universe. Therefore, the stellar content of a galaxy represents a fossilized record of its entire evolutionary history, and this is the main motivation for studying the populations of stars in other galaxies. Thanks to the enormous progress made in stellar physics since the 1920s, stars are the best understood constituent of galaxies. They are also responsible for the vast majority of the observational characteristics of galaxies. So, even if it should turn out that the "dark matter" is a more important constituent in terms of mass, the analysis of the stellar component will remain fundamental to understanding galaxies. The stellar content of galaxies is studied in two distinct ways. In nearby galaxies, the properties of individual luminous stars can be studied with large telescopes. The statistical properties of large samples of individual stars can also be analyzed in the Hertzsprung- Russell (H-P) diagram (see Fig. 1). To date, this approach has been limited mainly to the galaxies of the Local Group, but new generations of ground-and space-based telescopes are rapidly expanding its scope. The study of individual stars also plays a fundamental role in measuring the distances to galaxies. In fact, observations of Cepheid variable stars in Local Group galaxies were used by Edwin Hubble to demonstrate the very existence of external galaxies and to provide the first good estimate of the scale of the universe. In more distant galaxies, individual stars are too difficult to observe, and one must study instead the integrated, or composite, light of the stellar component. Although there are limitations to this technique, nonetheless it can be applied in principle to any galaxy in the observable universe-even to very distant protogalaxies still in the process of formation. Data obtained across the entire electromagnetic spectrum, from x-ray to radio wavelengths, are now used in this kind of analysis. A galaxy may consist of many generations of stars. To clarify their measurable properties, consider the evolution of a single such generation. Stars form over a large range of masses, typically from 0.1-100 M*. The distribution over mass is defined by the initial mass function, which may vary with physical conditions. The longest-lived stage of stellar evolution is the main sequence phase, in which the star burns hydrogen into helium. In this phase, more massive stars are brighter and hotter (and therefore bluer) than less massive stars. Stars over 10 M* are luminous enough to be detected individually throughout the Local Group and beyond. They also emit copious amounts of far-ultraviolet radiation, which is capable of ionizing the interstellar hydrogen gas around them. The ionized, or HII, regions are readily recognizable by their emission lines. However, massive stars evolve quickly off the main sequence, in times of only about 10 million years, to become even shorter-lived giants or supergiants. Most will explode as supernovae, leaving neutron star or black hole remnants. low mass stars like the sun are initially much fainter and cooler and have main sequence lifetimes of 10 billion years or more. However, once they leave the main sequence, they brighten to become cool, red giants before they fade to white dwarf remnants. Thus, the progression of a stellar generation is from a bright, blue phase at early times, with many individually bright stars and much ionizing radiation, to a red phase at later times, dominated by the light of cool, low mass giant stars. Evolution in the integrated properties of the generation is much more rapid at early times. Chemical composition also plays a second- order role: for a given age, stars with a smaller abundance of "metals" (elements heavier than helium) will be slightly hotter. STELLAR POPULATIONS AND GALAXY MORPHOLOGY Early photographic studies revealed a remarkable correlation between the stellar content of galaxies and their morphology. The disks of spiral galaxies are blue and exhibit considerable structure in the form of individual bright stars, star clusters or associations, and HII regions. On the other hand, elliptical and SO galaxies and the bulges of spirals are red and amorphous in appearance, with little evidence of resolution into stars and usually no HII regions or dust clouds. Spectroscopic studies by Milton Humason, William W. Morgan, and others also showed a systematic trend in the integrated spectrum of galaxies from one end of the Hubble classification sequence to the other: Elliptical galaxies and large-bulge spirals have cool, red giant spectra whereas Sc and irregular have the blue spectra typical of hot stars and also exhibit emission lines. The concept of stellar populations was crystallized by Walter Baade in 1944. At Mt. Wilson Observatory, taking advantage of the dark skies provided by the wartime blackout, he was able to resolve the brightest stars in several Local Group elliptical galaxies and the bulge of the Andromeda Galaxy. He found them to be cool, low mass, red giants, apparently of the same kind that are found in the globular star clusters in our galaxy. He designated this type of system Population II, to be contrasted with the blue, easily resolved Population I that is characteristic of spiral arms and open star clusters in our galaxy. Subsequent development of the theory of stellar evolution (described above) led to the realization that the distinction between the population types is primarily one of age: Population I represents young, recently formed generations, whereas Population II consists of old generations whose main sequences have burned down to low mass stars. Parameters other than age are also included in the modern concept of stellar populations. Metal abundance is second in importance. Baade's original system was oversimplified in grouping the elliptical galaxies and globular clusters together. Both may be old, but it has since been demonstrated that the globular clusters in our galaxy are predominantly metal-poor systems (with abundances as low as 1% of those in the Sun), whereas galaxy centers can have metal abundances 2-4 times higher than that of the Sun. Kinematics is also an important population parameter, although observational difficulties limit its application for stars in other galaxies. HISTORY OF STAR FORMATION IN GALAXIES Much of the subsequent work on the stellar content of galaxies has concentrated on inferring the detailed time dependence of star formation and metal abundance in galaxies. Ellipticals and spiral bulges evidently completed most of their star formation long ago. Either because of the efficiency of this process or of mechanisms like galactic winds, which sweep out residual gas, little cool interstellar gas remained to form more recent generations of stars. Although it seems clear these systems are old (over 5 billion years), it is a matter of controversy whether they completed star formation as long ago as the globular star clusters (15 billion years). By contrast, appreciable amounts of cool interstellar matter (typically 10% of the total mass) remain in irregular galaxies and spiral disks, and this fuels ongoing star formation. (It is not completely understood why different galaxy types should differ so much in their current cool gas content.) Studies of these objects indicate that most experienced constant or declining mean star formation rates during the last 10 billion years and probably have an initial mass function like that in the galactic disk. Again, it is unclear exactly how old the disks are; some studies of white dwarf statistics and chemical abundances suggest an age well below 15 billion years for the disk of our galaxy. Most galaxies should consume all their remaining cool gas within another 10 billion years. Some information on the chemical history of galaxies has also been gleaned. Globular star clusters are found in all types of bright galaxies, and many are very old and have very low metal abundances. Both spiral and elliptical galaxies also exhibit internal gradients in metal abundance, with the central regions having higher metallicities. There is also a dependence of the mean abundance of a galaxy on its mass: more massive galaxies have higher mean metallicities. All these features can be understood in a picture where there is self-enrichment of the gas in protogalaxies by star formation. Globular clusters are representative of the earliest generations of stars formed directly from primordial gas. Massive stars rapidly synthesize heavier elements in their interiors and release them through stellar winds or supernova explosions. The enriched gas collapses inward, so that stars formed later or nearer the center of the protogalaxy have higher abundances. Enrichment ceases when supernova explosions become so frequent that the gas is heated to high temperatures and escapes (a galactic wind). This would happen at an earlier stage in lower mass galaxies, leading to the observed mass-abundance correlation. RECENT DEVELOPMENTS New observational techniques are permitting better study of stellar remnant populations (white dwarfs, neutron stars, and black holes) and their relationship to their parent galaxies. For example, compact binary systems including neutron stars or black holes can be copious emitters of x-rays, and such objects can be detected individually in other galaxies. Far-ultraviolet observations of elliptical galaxies have apparently detected the "post-asymptotic-giant-branch" phase of evolution that immediately precedes the formation of white dwarf stars. The relationship between a galaxy's far-UV spectrum and its metal abundance is an important clue to the little-understood physics of advanced stellar evolution. The possibility that the "dark matter" consists of remnants or very low mass stars has not yet been ruled out, so these advances in our ability to study intrinsically faint stars are welcome from that standpoint as well. It is becoming evident that the history of galaxies can be punctuated by periods of intense star formation, often called starbursts. Interactions between galaxies, for instance tidal encounters, often play an important role in triggering starbursts. The best known nearby example is the unusual galaxy M82, in which a burst converting about 10 M* of gas into stars each year was probably initiated by a close encounter with its spiral neighbor, M81. The most spectacular instances are distinct interacting systems identified by the Infrared Astronomical Satellite, where the star formation rate may reach 1000 M* yr-1. The supernova rate in such objects (including M82) appears to be high enough to drive hot gas plumes out of the galaxies and into the intergalactic medium. Unlike most areas of science, where earlier history can only be inferred, astronomers can directly witness evolutionary processes as they occurred billions of years ago simply by observing distant galaxies at large "lookback" times. In the last 10 years important progress has been made in techniques for such studies. Observations of distant clusters of galaxies indicate that the stellar populations of galaxies were undergoing rapid evolution as recently as a few billion years ago. This unexpectedly dramatic change is named the Butcher-Oemler effect, after its discoverers, Harvey Butcher and Augustus Oemler. Once again, interactions between galaxies or with the surrounding gaseous medium are apparently largely responsible. Galaxy mergers, in particular, can have drastic consequences, producing wholesale transformations of the stellar component and morphology of galaxies. A merger could convert a pair of spirals to an elliptical galaxy, for example. It is unclear if mergers play a role in the evolution of most galaxies, but they certainly influence some. The most distant galaxies detected to date have redshifts higher than 3, meaning that they are seen at over 75% of the look-back time to the Big Bang, and appear to be still in the protogalaxy phase. Research on these remarkable systems holds the promise of providing, at last, an understanding of the processes by which galaxies form. Additional Reading Hodge, P.(1981). The Andromeda galaxy. Scientific American 244 (No. 1) 89. Mihalas, D. and Binney, J.(1981). Galactic Astronomy. W.H. Freeman, San Francisco. Morgan, W. and Osterbrock, D.(1969). On the classification of the forms and the stellar content of galaxies. Astron. J. 74 515. Norman, C., Renzini, A., and Tosi, M.(1986). Stellar Populations. Cambridge University Press, Cambridge. Sandage, A.(1986). The population concept, globular clusters, subdwarfs, ages, and the collapse of the galaxy. Ann. Rev. Astron. Ap. 24 421.