Right Ascension Offset Aa (arcsec)
Figure 19.15 Optical image of the ultraluminous infrared galaxy Arp 220. The inset is a 2.2 |am image of the nuclear region only. Offsets in the main figure are relative to the nominal IRAS position of the galaxy. Offsets within the inset are relative to the peak emission at 2.2 |am.
Figure 19.16 (left panel:) Optical image of The Antennae, spanning 18'. (right panel:) Image in 2.12 pm of the innermost 2'5, or 14 kpc at the galactic distance of 19 Mpc.
even two galaxies locked in a binary orbit cannot undergo many interactions. The ejection of stars during each encounter robs energy and angular momentum from the large-scale, orbital motion. After only a few close passages, a pair of gravitationally bound galaxies spirals together and truly merges, creating a system quite different from the original ones. The presence of dark halos accelerates this process, both by increasing the geometric cross section and through halo spinup, which further diminishes the orbital angular momentum of the visible component.
Each spiral galaxy of an interacting pair contains some 1011 stars. These are so thinly distributed, however, that there is no chance for direct collisions, even when the systems partially or wholly merge. The two stellar populations simply interpenetrate. During all subsequent motions, including violent ejection of stars, total energy is conserved. Similar statements cannot be made of the gas. This component, relatively small in mass, responds to the changing gravitational field created by the stars. Fluid streamlines may then cross, even when the galaxies themselves are still some distance apart. Streamline crossing creates shocks, which radiate energy.
This energy loss within the circulating gas of each galaxy allows that material to spiral rapidly inward, deeper into the potential well. If the gas shocks repeatedly, the spiraling occurs over just a few galactic orbital periods, or a total interval of order 108 yr. Such a time scale accords with that deduced empirically for the starburst phenomenon. We can thus understand, at least in broad terms, how circumnuclear gas accumulates. But how, in detail, do stars form within this new medium? And what is the nature of the perturbation leading to crossed streamlines and shocks?
The best-resolved examples of starburst regions show formation activity occurring in discrete clusters. The youngest of these do not widely differ within the system or from one galaxy to another. While their internal stellar densities may be a 100 times those in even the most vigorous Galactic sites, the inferred cluster formation times of order 107 yr are similar. The indication, then, is that localized regions are making stars through the usual process, albeit in an unusually dense and massive environment.
We have noted, in fact, that the molecular clouds producing individual clusters in starbursts are probably more akin to Sgr B2 than to anything in the solar neighborhood. There is a lingering uncertainty regarding the mass distribution of the stars being formed. The best-studied case is 30 Doradus. Here, the color-magnitude diagram yields a distribution that falls off at high M* in a manner similar to the Galactic initial mass function. In the other direction, the flattening and subsequent turnover may occur at a larger M*-value. If so, the region is forming a greater proportion of relatively massive stars. We must bear in mind, however, such technical issues as the proper dereddening procedure in an inhomogeneous, dusty environment. High extinction effectively removes objects from a magnitude-limited sample. The corrections are problematic but important, since they alter the low-mass end of the distribution.
Returning once more to the large-scale picture, we have noted that gas transport is facilitated by energy dissipation. The material also needs to lose angular momentum if it is to spiral inward. Any nonaxisymmetric perturbation to the stellar surface density will torque the gas gravitationally, and a trailing spiral wave induces mass flow in the proper direction. (Recall our discussion of the Galactic center.) To generate a very high mass flow, the perturbation itself must be exceptionally strong. Numerical simulations have demonstrated two fundamental results in this context. First, a trailing, spiral disturbance can be enhanced in a galaxy through tidal interaction with a neighbor. Second, a sufficiently large-amplitude disturbance becomes a linear bar in the central region.
Figure 19.17 illustrates both points through a representative calculation. This sequence shows the response of a face-on spiral galaxy to another, passing spiral. The latter, which is inclined 71° to our line of sight, only appears in the second frame, at its point of closest approach. Both systems are modeled as containing stars and gas, with the latter comprising 10 percent of the disk mass. There are also massive, dark halos, not shown. The stars (upper panels) are treated as point masses, interacting only through their mutual gravity. Gas elements (lower panels) have an extra, repulsive force to mimic pressure, following the technique of SPH (Chapter 12). Each unit of nondimensional time corresponds to 2.5 x 108 yr.
Before the companion galaxy appears, both the gas and stars exhibit a clear spiral pattern whose amplitude is modest, especially for the stellar component. The patterns are greatly disturbed in the second pair of frames, when the center of the neighboring galaxy reaches 8 kpc from the target nucleus. Following this close passage, the two-armed spiral is much more open and of greater amplitude. A central bar develops and is dominant in the last frames shown. The rotation sense of the face-on galaxy is clockwise. Hence, the gaseous bar slightly leads the stellar one. The stars exert a braking torque on the gas, draining its angular momentum.
Was this article helpful?