ARlogo Annu. Rev. Astron. Astrophys. 1982. 20: 517-45
Copyright © 1982 by Annual Reviews. All rights reserved

Next Contents Previous

4. MOLECULAR CLOUDS IN THE CENTRAL REGIONS OF GALAXIES


4.1. Is There a Distinct Central Concentration of Molecular Clouds?

The division of galaxies into those with and without bright central CO sources suggests that the central sources, when present, might represent an independent collection of molecular clouds distinct from those in the galactic disks. By this, we mean that there may exist a concentration of molecular material that is substantially greater than what would be expected from an extrapolation to small radii of the molecular distribution found in the disk, perhaps with rather different properties (e.g. mean temperature) for the constituent clouds. For our Galaxy, there is little molecular material inside 2 kpc except for the distinctive concentration (MH2 ltapprox 7 x 108 Msun) within 300 pc of the nucleus (Bania 1977). Is this true for other galaxies as well?

This question is difficult to answer directly because of limited spatial resolution. The currently available angular resolution of CO data is 0.'5 to 1.'0. For almost all the observed galaxies with distances > 3 Mpc, the corresponding linear resolution is of order 500 pc or worse. The exception is M82, for which VLA observations of OH absorption have been made with a resolution of a few arcsec (Weliachew et al., in preparation). In most cases, however, one cannot yet distinguish sources akin to the stellar nuclei.

Nevertheless there are some clues supporting the notion of a distinct central concentration. The first is that the velocity widths of the CO emission and OH absorption lines from the central sources are (barring inclination effects) quite large, up to 400 km s-1 (e.g. Figure 4). Even in the face-on galaxies, the central region is distinguished from the disk by its velocity width. These velocity spreads are comparable to those found within the tilted molecular disk proposed for the center of our Galaxy (Liszt & Burton 1978). Furthermore, if the central distribution of molecular clouds were not strongly concentrated toward the nucleus, then the widths of the OH absorption lines seen against compact nuclear sources would generally be smaller than the widths of CO emission lines from the central regions, because the OH absorption would be sampling a much smaller portion of the velocity field (see Section 2). That the OH and CO line widths are observed to be comparable implies that the OH absorption arises close to the nuclear continuum source where the velocity dispersion is largest. (This also implies that the velocity fields near these galactic nuclei include extensive noncircular motions.) A second clue is that the shapes of the OH line profiles often resemble those seen toward Sgr A (Gardner & Whiteoak 1975, Whiteoak & Gardner 1975a, b, Rickard et al. 1982), which is known to be dominated by the near-nuclear component. Finally the preferential detection of OH absorption toward galaxies for which the background flux is concentrated into a small nuclear source (Rickard et al. 1982) implies that the bulk of the absorbing gas is concentrated into the region within about 1 kpc of the nucleus.

Figure 4

Figure 4. The profiles of the four 18-cm OH transitions seen toward the nuclear continuum source of NGC 253, taken from Gardner & Whiteoak (1975). The axes are antenna temperature (with scales indicated for each profile) and heliocentric radial velocity.

It is difficult to decide whether a highly centralized molecular source is present on the basis of CO observations alone, unless high spatial resolution can be achieved. If the CO observations are made with telescope beams that include substantial portions of the surrounding disks, then one cannot distinguish classes 1, 2, and 3 even if the emission appears to peak at the nucleus. For example, Bieging et al. (1981) argue that because the CO profile shape in a few galaxies is so strongly reminiscent of "horned" global HI profiles, the CO must be mostly in the disk. However, it is not obvious that the data can distinguish horned profiles from the bicuspid profiles characteristic of central disks or rings, models of which have been successfully fit to the CO data of some galaxies (Rickard et al. 1977a). Furthermore, there is no indication that the CO profile characteristics are better correlated with those of the HI lines than, say, those of the (clearly central) OH lines.

In sum, the evidence favors there being a distinct central molecular source in many galaxies, though certainly not all. Gradations in the size and intensity of this source represent most of the differences among classes 1, 2, and 3.

Next Contents Previous