Annu. Rev. Astron. Astrophys. 1998. 36: 435-506 Copyright © 1998 by Annual Reviews. All rights reserved

8. INTERACTIONS IN THE LOCAL GROUP

The Local Group is a dangerous place for dwarf galaxies. NGC 205 and Sagittarius have wandered too close to their dominant parents and exhibit clear kinematic and structural signatures of tidal distortions (Hodge 1973, Bender et al 1991, Pryor 1996, Ibata et al 1997). Irwin & Hatzidimitriou (1995) noted that many nearby dSph systems show a strong correlation of tidal radius or ellipticity with the strength of the external tidal field. Bellazzini et al (1996) have shown convincingly that the central surface brightness, ₀, of dSph galaxies obey a bivariate relation in ₀, L_tot, and R_GC, where R_GC is the Galactocentric distance of the galaxy. They show that Sagittarius in particular appears to be unbound, even in its core (Mateo et al 1995c, but see Ibata et al 1992). A number of models have investigated what dwarfs look like before, during, and after strong tidal encounters (Allen & Richstone 1988, Moore & Davis 1994, Johnston et al 1995, Oh et al 1995, Piatek & Pryor 1995, Velázquez & White 1995, Kroupa 1997). At early times in a strong interaction or in the weak-interaction limit, stars are lost from the dwarf into leading and trailing orbits. These stars quickly fill a larger volume than that of the original galaxy, and if they were included in kinematic samples, they would reveal streaming motions that could be interpreted as rotation. Extratidal stars might be seen at this stage, even though the majority of the galaxy's stars remain bound and the central velocity dispersion is unaffected (e.g. Gould et al 1992, Kuhn et al 1996; in both cases many of the stars discussed are actually within recent estimates of the tidal radii of the respective galaxies). At later stages of strong interactions, the dwarfs become strongly elongated but not necessarily parallel to the orbital path of their center of mass. Alcock (1997a) claimed to see such a tilt in Sagittarius, but Ibata et al (1997) did not. At very late stages of a nearly complete tidal disruption event, the dwarf becomes a long strand that is stretched along its orbit, with a small clump (10% of the original mass) as the only remnant of the original galaxy. At no time except the very end of the tidal episode does the central velocity dispersion significantly exceed its virial value, even for models with no initial dark component.

These disrupted dwarfs should produce relatively long-lived streams in the halos of galaxies such as M31 and the Milky Way (1-2 Gyr; P Harding, private communication). Lynden-Bell & Lynden-Bell (1995) concluded that one possible stream can be traced out with the Magellanic Stream (Wakker & van Woerden 1997), Ursa Minor, Draco, and possibly Carina and Sculptor. A recent determination of the proper motion of Sculptor (Schweitzer et al 1995) suggests that this galaxy is not part of this putative (or any other proposed) stream. There have been many intriguing claims of halo substructure in recent years (e.g. Arnold & Gilmore 1992, Majewski 1992, Côté et al 1993, Kinman et al 1996) that could possibly be remnants of disrupted dwarfs.

More recently, Alcock et al (1997b), Zaritsky & Lin (1997) claimed to detect a possible signature of a foreground galaxy or galaxy tidal remnant towards the LMC. Gallart (1998) suggested instead that this new "galaxy" is in fact due to the signature of known but subtle stellar evolutionary phases that are becoming apparent in the large-scale photometric surveys carried out in the LMC. This is not the first time that a putative new galaxy has been detected directly in front of a known Local Group dwarf: Connolly (1985) identified a number of "foreground" RR Lyr stars toward the LMC that he concluded are members based on their photometric properties but are nonmembers kinematically. Saha et al (1986) also identified some anomalously bright RR Lyr-like stars, apparently in front of the Carina dSph galaxy, that could be either part of an extended halo of the LMC or possibly associated with a foreground system. A more natural explanation may be that these are instead anomalous Cepheids in Carina itself (Mateo et al 1995a). In none of these cases is the true nature of all of these "foreground" stars conclusively established, and in the case of the LMC, it is not unreasonable to suppose that a tidal tail is present (Zaritsky & Lin 1997). Nevertheless, it seems wise to treat claims of the existence of galaxies or tidal features directly in front of known Local Group systems with particular caution.

Mateo (1996), Unavane et al (1996) have discussed the possibility that a large fraction of the Galactic halo has been constructed from disrupted dSph systems. The latter considered Carina to be the template of such a system, while Mateo (1996) compared the properties of the ensemble of all Galactic dSph satellites with the halo. Neither approach is strictly correct. Carina arguably has the most unusual stellar population of any dSph system (Section 6.3; Figure 8); it is clearly not an appropriate choice as a template for the halo. On the other hand, present-day dSph systems are anomalous "survivors" that were able to form stars over a longer period than systems that were destroyed. They probably also follow orbits (relative to the Galaxy) that are quite distinct from the orbits of the galaxies that were consumed; thus, even an average of the stellar populations of all remaining dSph galaxies should not be expected to precisely match the current halo population. Given these differences in approach, the two studies nevertheless essentially agree: No more than 10% of the halo could have Carina-like progenitors (Preston et al 1994), but more than 50% of the halo could have formed from galaxies similar to the entire ensemble of Galactic dSph systems (though see van den Bergh 1994b).