Content: Elliptical galaxies, ?Spiral galaxies, ?Irregular galaxies
Adapted from The Astronomy and Astophysics Encyclopedia and
W.A. Baum


Ellipticals are galaxies that appear ellipsoidal in form, have no disks, and are devoid of features such as spiral arms, bars, or dust lanes. Because such features would be associated with recent or ongoing star formation, their absence indicates that nearly all of the stars in elliptical galaxies must be somewhat old.

The majority of bright galaxies in large clusters are ellipticals, but fewer than 15% of galaxies in the general field are ellipticals. Classical ellipticals (E galaxies), with their bright compact nuclei and steep brightness gradients, range from absolute visual magnitude MV ~ -23 down to MV ~ -16 mag. The very brightest of them, called cD galaxies, are found at the centers of clusters, and are among the most luminous galaxies in the Universe.

Dwarf ellipticals (dE galaxies) include an assortment of morphological types and range from MV ~ -19 down at least to MV ~ -12 mag, where their numbers are rising steeply and where surveys become seriously incomplete. They probably extend to (and may overlap) the range of globular star clusters, which commences at MV -10 mag. Dwarfs differ greatly from one another in compactness.

Seen on the plane of the sky, some E galaxies are quite round, and others are elongated. Although various classification schemes have been devised, the degree of elongation is commonly designated by Edwin P. Hubble's subclass, 10(1 - b/a), where a and b refer to major and minor axes. Thus, an EO galaxy is round, and an E7 (the most elongated subclass) has a projected axis ratio b/a ~ 0.3. In the elongated galaxies, the ellipticity is typically a function of the isophotal level. In some, the position angles of the major axes of the isophotes are also a function of the isophotal level; that is, such galaxies possess an isophotal twist. Moreover, isophotes sometimes depart from pure ellipses in the sense of being slightly rectangular ("boxy").

The three-dimensional shape of a galaxy has to do with the distribution of stellar velocities within it. In disk galaxies (spirals and their featureless SO cousins), the angular momentum due to Keplerian rotation dominates over random motions, and the resulting galaxy is an oblate spheroid. But in E galaxies the angular momentum is not dominant, and the three-dimensional shape of the system is maintained mainly by the dispersion of stellar velocities within it. In principle, the velocity dispersion can be anisotropic, so that an E galaxy can be a prolate ellipsoid or even a triaxial one. Unfortunately, the three-dimensional shapes cannot be directly observed.

Since E galaxies are not primarily supported by Keplerian rotation, it is not possible to calculate individual masses from their rotation curves in the manner used for tilted spirals. Internal velocity dispersions give only limited information. If one is willing to assume that clusters of galaxies (which often consist mainly of ellipticals) are in gravitational equilibrium, the mass of a whole cluster of galaxies can be estimated by application of the virial theorem to the observed dispersion of galactic velocities within the cluster, but the inferred mass tends to be puzzlingly large. The mass-luminosity ratio of the stellar population in giant ellipticals therefore remains quite uncertain. It might even be different in a cluster environment than in the general field.

Various formulas have been used to fit the observed radial distribution of brightness of E galaxies in the plane of the sky, but the one most often used today is the de Vaucouleurs law e (after Gerard de Vaucouleurs). It says that the logarithm of the surface brightness (usually expressed on a stellar magnitude scale) is an approximately linear function of the 1/4-power of the radial distance from the nucleus. This is a purely empirical finding that has no direct theoretical basis. Also, the goodness of fit varies somewhat from object to object.

At the very center of a large cluster of galaxies there is typically an unusually luminous elliptical surrounded by a faint but extensive envelope pervading the heart of the cluster. These cD galaxies are suspected of having grown by mergers and by the cannibalization of nearby dwarfs.

In the 1940s, elliptical galaxies were assigned by Walter Baade to his Population II, by which he meant specifically that the color-magnitude diagram of their stellar population should be essentially the same as that for globular star clusters of the Milky Way. Owing to their high random velocities and their nearly spherical distribution in the Milky Way, globular star clusters were known to have formed during a very short interval before the Milky Way had settled into a disk and before the interstellar medium had become enriched with heavy elements ("metals"). It was therefore assumed that elliptical galaxies had similarly formed during an early burst of star formation.

But integrated photoelectric photometry soon showed (and spectroscopy later confirmed) that giant elliptical galaxies differ greatly in spectral energy distribution from globular star clusters. Therefore, the dominant stellar population of giant ellipticals cannot be similar to that of globular clusters. There was thus already good evidence in the 1950s that star formation in giant ellipticals must have continued long enough for a high degree of metal enrichment to permeate the stellar population. It is only with the accumulation of further evidence in recent years, however, that the concept of temporally distributed star formation in ellipticals has finally gained general acceptance.

Judged from integrated colors and spectra, some dwarf ellipticals must be at least partly similar in stellar content to giant ellipticals, while other dwarfs consist more of stars like those in globular star clusters. There are also a few dwarf ellipticals near enough for telescopic resolution of the brightest stars, but the resolved stars do not necessarily belong to the population that dominates the integrated light or the mass. Such is clearly the case for M32, the compact metal-rich companion of the Andromeda galaxy. On the other hand, several of the very tenuous metal-poor galaxies classed as dwarf spheroidals are near enough that the color-magnitude diagrams of their stellar populations can be definitively identified.

Owing to the fact that no giant elliptical happens to lie near us, none has ever been resolved well enough to permit photometry of individual stars. Incipient resolution of the nearest ellipticals can barely be achieved under the very best Earth-based observing conditions. Some improvement can be expected with telescopes in space. Using a form of noise analysis on such images, one can infer the nature of the brightest stars, the number of them Per unit surface brightness, and the relative distance of nearby ellipticals. Noise parameters, taken together with known properties of the integrated light, should enhance our knowledge of the population mix.

It has been known for a long time that giant ellipticals tend to be slightly redder in their inner regions than in their outskirts, suggesting a gradient in the metallicity of the stellar population. Gradients have now been measured spectroscopically for several elements, and the gradient for the prominent magnesium feature around 5175 Å has been extended to faint levels in the outskirts by narrow-band CCD photometry. On reasonable assumptions, those data indicate the inner regions to be more metal rich than stars in the solar neighborhood, while the halos are only a little less so. In other words, the halo population of ellipticals is much more metal rich than that of the Milky Way. On the other hand, globular star clusters that have been detected in the halos of nearby ellipticals do not seem (from their colors) to be so metal rich; so the globulars must have formed at an earlier time than many of the individual stars that now populate the same halo regions.

It should be noted that the strength of an absorption feature such as the 5175-Å magnesium band is dependent upon stellar surface gravity as well as metallicity. However, the temperature of the giant branch of the H-R diagram (or color-magnitude diagram) is largely controlled by metallicity, and it is the giant branch that would be expected to dominate the integrated light. For subgiants just above the main-sequence turnoff, surface gravity differences (which are correlated with age differences) play a role, but the turnoff stars probably do not contribute strongly to the integrated light.

No existing theoretical model for the formation and evolution of giant elliptical galaxies is able to explain all of their observed properties, but some form of inhomogeneous dissipative collapse appears to be indicated, and star formation (though not active today) was evidently not limited to a single early burst. It presumably required a long time to build up the high observed metallicity. Mergers may also have played a role. Any successful theory of giant ellipticals must take account of the following observed properties: (1) They have high metallicities but low metallicity gradients, resulting in relatively metal-rich halos. (2) They are inferred to have triaxial figures supported by anisotropic velocity dispersion, rather than by rotation. (3) Globular clusters in giant ellipticals have lower metallicities than halo stars in the same regions. (4) The distribution of globular clusters in giant ellipticals is less centrally concentrated than is the main body of stars.