Next Contents Previous

6.2. Dwarf companions in the field

Holmberg's (1969) classical, statistical study of satellites around bright spiral galaxies has only recently been superseded by Vader and Sandage (1991), who identify dwarf companions individually by morphology, and by Zaritsky et al. (1993), who identify them by kinematics. A spectroscopic follow-up of Vader and Sandage (1991) has been started by Vader and Chaboyer (1992). Vader and Sandage (1991) present first results for E and S0 primaries, most satellites of which turn out to be early types as well, i.e. dwarf ellipticals. The projected distribution of the companions is shown to follow a surface density law close to Sigma (r) approx r-1, in accord with Zaritsky et al.'s (1993) finding for spiral primaries. Interestingly, this is also the profile of the dark halos of spiral galaxies as derived from rotation curve studies. The correlation function of dE's on the smallest scale (around massive giants) is steeper than for any other type of galaxy.

Vader and Sandage (1991) confirm the trend of increasing early-type dwarf frequency per giant galaxy with increasing richness 1 of the aggregate (absolute number of giants) found by Ferguson and Sandage (1991). Note that this environmental variation is distinctly different from the morphology-density relation: there is no significant variation of the dwarf/giant ratio within a given cluster, while the spiral/elliptical ratio changes dramatically with radius. In addition, the percentage of dwarfs that are bound to giants decreases with increasing richness: virtually all dE's in the field are bound to giants, while in the Virgo cluster only a few percent of them are (Ferguson 1992b). Tidal stripping is a possible explanation of the low frequency of bound satellites in a cluster environment. In the field, galaxies that start off within about 50 kpc of the primary will be accreted in less than a Hubble time, while in a cluster such galaxies could be ``liberated'' by the tidal field of the cluster, leading to a higher d/g ratio in clusters. However, it is not yet clear whether such a process could produce the observed correlation of d/g with richness. Other possibilities are discussed in Sect. 7.2.

In examining the spatial distribution of nearby swarms of dwarf companions around larger galaxies Einasto et al. (1974; 1975) noticed an intesting segregation of dwarf ellipticals from dwarf irregulars. At fixed luminosity, dwarf ellipticals are found at smaller separations from the primary galaxy than dwarf irregulars. The radius that divides dE's from Im's increases with companion-galaxy luminosity: luminous Im's are found at small separations, while low-luminosity Im's are not. Einasto et al. (1974) used this result to argue for the existence of massive gaseous coronae around galaxies such as the Milky Way that would act to strip irregulars of their gas (Sect. 7.6). However, while the general tendency of dE's to cluster around giants, in contrast to irregulars, is quite evident (e.g., Binggeli et al. 1990), the specific segregation line of Einasto et al. (1974) has not been confirmed by Zaritsky (1992) [who lacks the high morphological resolution of the Vader and Sandage (1991) study, however], or by Ferguson (1992b) for satellites within the Virgo cluster.

The only satellite system where we have the full 3D information on the distribution is of course our own, i.e. the two Magellanic Clouds plus the eight local dE's (or ``local dwarf spheroidals''), which all lie within 250 kpc of the Milky Way Galaxy. The possible correlations of M / L and star-formation history with galactocentric distance have been mentioned previously. Another interesting phenomenon is that the distribution of the local dE's around our Galaxy is not isotropic. This is evident from their sky-projected distribution. Lynden-Bell (1976) and, independently, Kunkel and Demers (1977) noticed that several globular clusters and dwarf spheroidals lie within only a few degrees of the great polar circle defined by the Magellanic Stream and some of the High Velocity Clouds. The most closely associated dwarf systems are Sculptor, Draco, Ursa Minor, and Sextans. It is significant that Sextans was only recently discovered (Irwin et al. 1990b) and that no other dwarf has thereafter been found in an extensive survey of the southern sky (Irwin 1994), although one has since turned up serendipitously (Ibata et al. 1994). The causal connection between these dwarfs and the Magellanic system remains unclear. It has been suggested that the dwarfs, like the Magellanic Stream, are tidal debris from a recent close encounter of the Clouds (Kunkel 1979; Murai and Fujimoto 1980; Lin and Lynden-Bell 1982; Gerola et al. 1983). However, with respect to their structure and content, the local dE's are indistinguishable from the dE satellites of M31 (Armandroff et al. 1993; Armandroff 1994) where no phenomenon like the Magellanic Stream is known (although the M31 dwarfs also lie roughly in a plane).


1 Note that the quantity ``richness'' used here is not the same (nor as well defined) as that used by Abell (1958), as his definition is not useful for very poor clusters or loose groups. The entire discussion of environmental variations in the morphological mix would benefit from a more precisely defined measure of richness, or a more physically motivated environmental parameter. Back

Next Contents Previous