Copyright © 1984 by Annual Reviews. All rights reserved
|
| Annu. Rev. Astron. Astrophys. 1984. 22:
37-74 Copyright © 1984 by Annual Reviews. All rights reserved |
Even though Irrs are less evolved than most types of luminous galaxies, they are normally far from being in primordial states. Intermediate metallicity levels, modest gas-to-stellar-mass ratios, and strongly composite stellar populations found in the average Irr are indicative of maturity levels that have required billions of years to achieve. Similarly, stars and gas in all but the smallest dwarfs are in flattened disks and are largely supported by rotation. Thus considerable dissipation of energy has likely occurred since a purely gaseous protogalactic state, and sizes of present-day Irrs therefore reflect the angular momentum and density distributions as well as the extent and mean density of their pregalactic progenitors (101, 117, 327, 328). The smooth rotation properties in the outer disks of many (but not all) Irrs further indicate that a few rotation periods (which are in some cases (~ 109 yr) have elapsed since outer HI disk formation. Most irregulars therefore are probably old systems that formed as gravitationally bound entities ~ 1010 yr ago, but we should be aware of possible exceptions. Metal-poor extragalactic HII regions are potential young galaxy candidates, and in some instances they have the disorganized, multiple gas cloud structures that are expected to characterize newly formed galaxies (21, 225, 227, 377). These small systems, however, could be old, but due to their low masses they have simply needed very long times to become dynamically organized and produce stars (cf. the multiple HI cloud structure of the SMC; 242).
Other exceptions to the above ``late-bloomer'' model are found primarily among the most luminous of the Irr family, i.e. giant blue compacts and clumpy irregulars. These systems have moderate-to-solar metallicities (36, 120, 342; Gallagher, Hunter & Bushouse, in preparation) and rotate at velocities similar to spirals, but they produce OB stars on incredible scales and in chaotic fashions that lead to Irr morphologies (66, 273, 274). The processes that cause such extreme (and probably transitory) evolutionary events in spirallike galaxies are not known, but they could include interactions with other galaxies or external agents (e.g. 220; many luminous Irrs are in binary pairs), major internal instabilities, such as relaxation of nonaligned gas disk components (90, 322), or delayed galaxy formation (e.g. 344). In any case, the existence of luminous Irrs and possibly related bursting Irr galaxies serves to remind us of the point stressed by van den Bergh (367): The evolution of galaxies may not be smooth in time, but instead may proceed by leaps and bounds in the form of postformation eruptions of star formation that dramatically and quickly alter the states of galaxies.
Most Irrs, however, are not in star formation burst phases, but rather they are evolving at nearly constant rates. Even though the mechanisms that produce such equilibrium behavior are not understood, it is clear that local processes play an extremely important role in the evolution of Irrs, probably through cumulative effects of many independent star-forming events or collective interactions between star-forming cells as envisioned in SSPSF theories. In the absence of strong differential rotation or dynamical forcing by spiral arms, these processes naturally lead to galaxies with Irr characteristics, i.e. astration is globally comparatively inefficient and star-forming complexes are distributed with a high degree of randomness. These features, furthermore, are not sensitive to details of galactic structure; it is then understandable that Irrs are found with remarkably uniform properties among slowly rotating galaxies over a range of ~ 103 in mass and that they predominate among low-mass disk galaxies.
The common dwarf Irr systems thus ultimately stem from the tendency
for density and degree of central concentration to decrease in tandem
with mass in disk galaxies, i.e. low-mass disk systems are slow,
near-rigid-body rotators. A similar strong correlation between stellar
mass and central density is found in diffuse dwarf elliptical galaxies
(160,
293).
Evidently the link between mass and density is nearly a
universal property among galaxies with M
1010
M
, independent of
morphological type, level of star-forming activity, or environmental
factors (e.g. cluster vs field locations). This implies that initial
conditions are crucial in determining fundamental properties of less
massive galaxies, and that later environmentally induced modifications
either have not been very common or have failed to produce major
structural modifications.
Furthermore, as dwarf and luminous galaxies apparently have similar, clumpy spatial distributions within the Local Supercluster (362, 380), most regions of space must have produced a large range of galaxy masses, i.e. dwarfs do not originate from special initial conditions. Indeed, some would view the extreme dwarf Irrs as being representative of the types of individual bound fragments from which all galaxies initially arose (see 327). In this regard it is interesting that bound systems consisting only of cool gas (i.e. objects that contain HI but have undergone little or no evolution) are evidently extraordinarily rare (if they exist at all), even among the least massive galaxies (230). The characteristics of extreme dwarf Irrs may also provide useful tests for models in which galaxy formation is explosively induced (194, 260), since this mechanism naturally yields a lower cutoff mass for galaxies, which should be observationally accessible (261).
On the other hand, Irrs are fragile and thus easily influenced by external factors. Indeed, they are often observed to have been affected by close passages near giant galaxies. It is also possible that the seas of diffuse dwarf ellipticals that populate regular clusters of galaxies (278, 293, 385) represent systems that initially formed as dwarf Irrs but were later transformed into stellar fossils as a result of gas removal by the hot intracluster medium via processes such as stripping or thermal evaporation (see 39, 384). Alternatively, initial conditions at or near the time of formation may have separated low-mass galaxies over a spectrum of star formation rates (125, 293, 368), in which case initial conditions were simply not the same in regions destined to become regular clusters and in the much broader general field.
Interactions with the environment, however, need not always be destructive. Violent disruptions suffered by massive galaxies or their companions during deeply interpenetrating collisions or mergers yield sizable fragments that later can become independent dwarf galaxies, perhaps even gas-rich ones of the Irr type (129, 311). Populations and structural-type distributions of small galaxies therefore will be in a constant state of flux, and observational programs directed toward studies of Irrs in a wide range of settings are of considerable interest. The still unsolved problem of the relative importance of environment vs initial conditions (``genetics'') in determining traits of galaxies thus extends to the Irrs. Since these galaxies are vulnerable and relatively uniform in their properties when in undisturbed states, comparative studies of Irrs in a range of environments have the potential to allow an empirical solution of this sticky issue.
By combining this information with an improved understanding of internal evolutionary processes, it should also prove possible to develop models for the initial states of Irrs. There are grounds for optimism about this task, since these low-density systems are probably closer to their protogalactic predecessors than are their dense spiral relatives. Thus somewhat paradoxically, Irr galaxies, which are dominated by young stars, are likely to be excellent stepping stones to the epochs of galaxy formation.
We wish to thank our many colleagues who have provided stimulating discussions and preprints of their work. We are grateful to Bill Bagnuolo, Craig Foltz, Paul Hodge, Jim Kaler, and Tom Kinman for reading and commenting on preliminary versions of this manuscript. The task of typing was efficiently dispatched by the office of the Department of Astronomy at the University of Illinois, and we wish to acknowledge Sandie Osterbur, Deana Griffin, Jan Wehmer, and especially Carol Stickrod for their help.
We similarly are indebted to the photographic and drafting personnel at Kitt Peak, who provided their prompt services. DAH acknowledges a research associateship at Kitt Peak National Observatory and JSG research grants from the National Science Foundation.