The last few years have seen a revival of interest in colliding ring galaxies and these new studies are providing an important stimulus for future work. We therefore conclude our review with some general recommendations for future work in this area.
7.1. Observational Directions
The discovery of large radial color gradients in ring galaxies provides an important stimulus for obtaining observations of ring galaxies with high spatial resolution and high sensitivity. Clearly of great importance is the detailed examination of the cause of the radial color gradients. As we have discussed, the most likely explanation is in terms of an evolving starburst population in the wake of the expanding rings. However, as with most color diagnostics, the effects of both reddening by dust and variations in metallicity are always factors to be explored. High resolution observations with the Hubble Space Telescope and with ground based telescopes equipped with adaptive optics provide the best hope of understanding the details of this process. At the time of writing, HST images have been made of the Cartwheel and IIZw28 by K. Borne and collaborators (including the authors of this review) and these data are currently being analyzed. These data have the potential of providing information about the time evolution and stellar content of the young star clusters formed in the rings and how they change with time downstream of the wave. In addition, new informations about the stellar make-up of the spokes will soon be forthcoming. Spectroscopic follow-up of these kinds of observation are vital. In particular, large ground-based telescopes will be required to obtain observations of stellar absorption lines of the faint material inside the "centrally-smooth" ring galaxies like VIIZw466 and Arp 147. Because these galaxies lack a complicating central bulge, the faint stellar material contained in their centers represents virgin territory in terms of following the evolution of the starburst from the outer ring to the center. These galaxies should be prime targets for future studies. Work on other less well studied centrally smooth southern rings (such as AM2145-543 and AM0417-391 in the south) should also be undertaken to build up a larger sample of "centrally-smooth" rings.
As the tantalizing work of Charmandaris et al. (1993) has shown, the HII regions in ring galaxies may hold some clues about different modes of star formation in galaxies. However, very little work has been done on the neutral and molecular gas content of these galaxies with the arcsecond resolution needed to begin to exploit the rings as star formation laboratories. Useful parallel studies on the distribution of dust (perhaps with ISO or focal plane arrays such as SCUBA on the JCMT) will help to answer questions about the importance of dust in the observed color gradients as well as its importance in density wave triggered star formation. Questions such as whether galaxies like Arp 10 are showing evidence for "threshold" star formation behavior can only be properly tested if we can map the distribution of molecules with high spatial resolution. As the work of Horellou et al. (1994) has shown, very few of the ring galaxies are strong CO emitters, and so mapping the fainter ones will require more sensitivity than is available with current CO interferometers.
The recent models of ring galaxies using hydrodynamics with realistic heating and cooling terms incorporated into the dynamics shows that the ring will be the site of major vertical outflows which should be detectable with AXAF. Mapping of the soft X-ray emission from outflows associated with the rings should be possible with AXAF's HRC instrument, and such observations will provide the next important tests of the models. The models suggest that the degree of heating in the ring is crucial to both the formation of spokes (too much heating blows them apart) and the formation of a reasonably narrow dense ring. The combination of both high resolution optical and uv observations with 0.5 arcsec X-ray imaging from AXAF will strongly constrain the range of possible energetic behavior in the young stellar associations and this will ultimately lead to a better understanding of the importance of ISM heating from O/B association winds as compared with the selfgravity of the dense regions in the ring.
Additional information on the nature of the stellar evolution in and behind the ring waves will be obtained with high sensitivity radio continuum observations at a number of frequencies. These are well within the reach of the current capabilities of the VLA and some work is being done in this area by one of us (PNA). High resolution observations, particularly those sensitive to the polarization of the radio emission, will allow for a more detailed comparison to be made between the star formation regions and the structure of the magnetic field in the wake of the ring wave. It has been suggested (R. Allen, personal communication) that the outgoing density wave may lead to an ordering of the magnetic field lines behind the wave if most magnetic fields are generated locally within clouds on a scale smaller than the wave itself. Further interesting effects might be expected as the disk puffs up behind the ring, increasing the vertical scale-height for the cosmic-ray electrons. The spectral steepening in the radio continuum found in Arp 10 behind the main ring (Ghigo and Appleton, in preparation) may also relate to an aging of the relativistic electron population as a function of time. Further observations are needed of a much larger sample of ring galaxies to determine how universal this effect may be.
Finally, observations of ring galaxies at high redshift would be of great interest. The majority of known ring galaxies have been found through inspection of photographic plate material. Observations with high angular resolution, such as those made with the Hubble Space Telescope, are yielding many background galaxies, often serendipitously. The large number of WFPC-2 images now being accumulated will be an important source of new high redshift ring galaxies and we urge efforts to be made to find new candidates.
High redshift ring galaxies are expected to be interesting for a number of reasons. Firstly, galactic collisions are likely to be more frequent in the past than at the present epoch and so HST images may yield a dramatic increase in the number of new ring galaxies known. Secondly, high redshift ring galaxies offer the possibility of exploring the evolution of the galactic potential as a function of redshift. We have discussed earlier how the existence of multiple rings is particularly useful in determining the form of the gravitational potential in the target galaxy. In particular, broad widely spaced rings suggest a rising rotation curve whereas more closely spaced double rings with narrow ring structure suggest a declining rotation curve. Relatively simple morphological studies of high redshift rings will provide an interesting test of models of galaxy evolution that lead to the growth of dark matters halos by mergers. As we have emphasized in this review, care has to be taken to find gas-poor ring galaxies if this approach is to yield meaningful results, since the ring morphology is heavily influenced by triggered star formation. High redshift galaxies that are similar to systems like AM 1724-622 (which seem to be an example of a mainly stellar dynamical orbit crowding process - the classical Toomre process) would be the ideal candidates for tests of galactic structure.
On the other hand, if the recent models of Gerber are correct, then the relationship between the distribution of neutral gas and the underlying stellar disk may also provide information about the dominance of dark matter in high redshift ring galaxies (see Section 4.4 for further discussion). Since HI and CO observations are unlikely to be able to compete for some time with the spatial resolution of HST, it would seem that measuring the gas distributions in high redshift ring galaxies is a project for future generations. On the other hand, as has been shown for nearby ring galaxies, the distribution of ionized gas in such galaxies can provide some information about the distribution of gas versus old stars.
If the large color gradients found in nearby ring galaxies turn out to be ubiquitous through the universe, and if they can be understood in terms of the evolution of stars in the wake of the ring, then one might be able to determine the distance to the ring galaxies, even at high redshift. The method would rely on our ability to determine the age of the ring based on the radial color gradient alone, assuming that (see Section 4.4) the gradient is the result of the evolution of starburst populations in the wake of the expanding ring. Then, having a measurement of the angular diameter of the ring and an independent measurement of the expansion velocity in the ring (through detailed modeling and spectroscopy), the distance to the ring could be estimated. Of course, the fundamental uncertainty would be the age of the stellar population and the corrections one might make for internal reddening within the galaxy. A crucial test of this method may soon come with the new HST observations of the Cartwheel. Here it will soon be possible to determine how well models of the color evolution of a stellar population fit with the observed distribution of star clusters and their colors in the region between the first and second rings.