2.3. Some Early Ideas about Ring Formation

Although now thought to be an unlikely interpretation of ring galaxy formation, it is of interest to review briefly the ideas of Freeman and de Vaucouleurs (1974). Their model was essentially a collisional one, involving the gas dynamical interaction of a cool intergalactic hydrogen cloud (IGC) with a pre-existing galaxy containing an external HI ring. This collision was expected to strip away the HI ring from the underlying galaxy, leaving behind the bulge which would lie on the minor axis of the HI ring. Subsequent star formation in the HI was believed to provide the characteristic ring galaxy appearance. De Vaucouleurs and de Vaucouleurs (1975) also suggested a similar mechanism for the formation of the ring in NGC 985 (a Seyfert 1 galaxy). Although many ring galaxy candidates were mentioned by Freeman and De Vaucouleurs, two particular galaxies were discussed in detail. They were Arp 144 and NGC 2444 / 5 = Arp 143. The case of Arp 144 was particularly compelling, since the morphology of the ring seemed to suggest that the intruding cloud was caught in the act of peeling the ring away from the disk of the target galaxy. (The ring has a peculiar "folded" appearance on some optical photographs). The case for Arp 143 was less secure, since Burbidge and Burbidge (1959) had already demonstrated that the "ring" in this case seemed to contain a nuclear "knot" in addition to the supposed displaced bulge. However, the overall morphology was approximately consistent with an IGC-galaxy collision. Burbidge and Burbidge suggested themselves that NGC 2445 might have "formed in the wake" of the elliptical-like component, although no specific details were given.

Later searches for galactic-scale HI clouds have in almost all cases failed (Lo and Sargent 1976), although a few large HI complexes are known. Some of the known clouds are probably an extreme form of low-surface-brightness galaxy (see Briggs 1990; van Gorkom 1993). It is now believed that the space density of truly isolated neutral hydrogen clouds is probably too low to account for the ring galaxies.

Secondly, the case for an interaction between a pure HI cloud and a galactic disk is almost certainly ruled out in two out of three of the above prototypical cases presented by Freeman and de Vaucouleurs. In the case of Arp 144, Joy et al. (1989) have made near-IR observations of the system and conclude that it contains not one, but two nuclei: the second one heavily obscured by dust in the visible part of the spectrum. It would appear that Arp 144 can now be explained in terms of an ongoing merger between two galaxies, without needing to invoke a separate HI cloud (see also Higdon 1988). The same is true for NGC 985. There is now strong evidence that this system is also a double galaxy system (See Section 3.5).

The case of Arp 143 is not quite so easy to dismiss as simply a two-galaxy interaction. At first glance, the huge HI streamer (Appleton et al. 1987a, b) extending away from it may be taken as evidence for the IGC model (see Figure 5). Appleton et al. (1987a, b) suggested that the ring-like morphology of NGC 2445 (especially in HI) and the 150 kpc long HI streamer may be the result of a head-on collision between an HI rich low-surface-brightness galaxy and the disk of NGC 2445. Evidently the plume is not a pure HI cloud, since faint traces of starlight are found associated with the plume, although it seems deficient in molecules (Smith and Higdon 1994). No completely convincing model has yet been put forward which explains all the known facts (Appleton, Ghigo and van Gorkom, in preparation).

Figure 5. A contour map of the neutral hydrogen plume extending from the peculiar galaxy pair Arp 143 (From Ghigo, Appleton and van Gorkom in preparation, using the D array of the VLA). The plume extends over an angular scale of more than 1/3 the diameter of the full moon. At the distance of Arp 143, this corresponds to more than 200kpc in linear scale.