Annu. Rev. Astron. Astrophys. 1984. 22: 37-74 Copyright © 1984 by Annual Reviews. All rights reserved

3.2 Distributions of Star-Forming Regions

Irregular galaxies, particularly high surface brightness Irrs, are noted for and defined by the rather chaotic spatial distributions of their star-forming regions (cf. 164). Correlations in the distributions and spacings of HII regions seem to be lacking, but there do seem to be a few significant chains of HII regions that are probably coeval (188). This indicates, as does the presence of ill-defined ``spiral arms'' in some systems, that the star formation mechanism is not entirely random. In Irrs with bars, for example, large star-forming regions often occur at one end of the bar, posing the possibility that gas flows due to the bar may be important there (98, 106).

In spite of the apparent chaos, star-forming regions are not uniformly distributed over the disks of the Irr galaxies; instead, they seem to be asymmetrical and clumped on large scales. Hodge (159) interpreted the clumping in terms of localized star formation bursts on a scale of ~ 1 kpc, and the major star-forming regions then must migrate around a galaxy with time. This effect can also be seen from the distributions of star clusters in NGC 6822 and IC 1613 (167) and of the Cepheids in the LMC (92, 269); the main star-forming centers of the recent past (~ 10⁷-10⁸ yr) are in different locations than currently active regions. Large-aperture spectrophotometry centered on the most active areas of Irrs also shows that the star formation rates are too high and the metallicity and gas content too low for star formation to have always continued in that region at the current rate (191). Furthermore, in galaxies such as NGC 3738, NGC 4214, and NGC 4449, one can identify large complexes that obviously recently supported star formation but that are now in decline (188). In fact, the HII complexes typically cover less than 4% of the optical areas of Irrs; thus each position in a galaxy probably experiences a major star formation event once every 10⁸-10⁹ yr. This shifting of the major centers of star formation seems physically reasonable, since the star formation process must certainly deplete the local cool gas. But why star formation would necessarily clump on the observed scales and whether it migrates in a systematic manner are not known.

In addition, we do not find bright HII regions out to the optical ``edges'' of Irrs. That is, typically the current activity is within the inner 60% of the blue optical dimensions given in the Uppsala General Catalogue (UGC, 254a). A few extreme galaxies, such as NGC 5253, seem to be forming stars mostly in their central cores, although star formation obviously did occur in the outer parts of NGC 5253 at some time in the past (370). It is possible that areas outside of the active central zones no longer have gas above the critical density necessary for star formation, and so these galaxies will never form stars in their outer regions again (see below). However, an alternative possibility is that star formation in these less dense outer regions of Irrs is continuing but in a more diffuse manner. This is analogous to the problem of the dwarf Irrs, which also lack the giant HII complexes and large OB associations: Is star formation temporarily absent, or instead is the process of star formation different (i.e. higher mass stars and/or clusters are not formed)?