8.2. Host Galaxy Morphologies
Gelderman (1994, 1996) obtained images through broadband B, V, R filters of a sample of 20 CSS sources. (10) He classified amounts of disturbance and interaction using the parameters DC, IC, Q defined by Dahari (1994, 1985a, 1985b). He finds that there is evidence that every CSS source in his sample is either disturbed or interacting according to Dahari's criteria. Gelderman also notes the trend for the CSS sources to show evidence for diffuse linear features, such as tidal tails, bridges, and shells. Since these features are produced in simulations of interactions involving disk systems, Gelderman suggests that the CSS interactions involve at least one gas-rich object.
The morphologies of the GPS host galaxies have been studied by
Biretta, Schneider, &
Gunn (1985),
O'Dea, Baum, & Morris
(1990a),
Stanghellini et
al. (1993),
de Vries, Barthel, &
Hes (1995),
and these results have been summarized and analyzed by
O'Dea et al. (1996a).
In the sample considered by
O'Dea et al. (1996a),
there are 30 objects for which there is sufficient signal-to-noise to
permit a determination of the basic morphology. The observations used
generally reach a sensitivity (i.e., lowest,
3 
, contour level) of
about 24.5 mag arcsec-2 (observed) in the r band.
O'Dea et al. (1996a)
find that 17 of these objects (57%) have distorted isophotes and that
four galaxies (13%) have a second nucleus in projection (within a few
arcsec or within 10 kpc).
O'Dea et al. (1996a) also determined the projected distance to the closest apparent companion or second nucleus of the GPS galaxies. The redshifts of the possible companions are unknown, so their relationship to the GPS source is currently unknown. The median angular separation is ~ 5", and the median projected linear separation is ~ 20 kpc. If these possible companions are confirmed, then their relative closeness to the GPS galaxy is consistent with the other evidence that these sources are interacting and/or merging. Thus, a large fraction of the CSS/GPS radio galaxies show evidence for interaction and/or mergers, suggesting that these processes must be relevant to or at least associated with the formation of CSS/GPS radio sources.
Other studies of powerful radio galaxies have found similar fractions of distorted objects (see, e.g., Heckman et al. 1986; Smith & Heckman 1989a, 1989b). Smith & Heckman (1989b) found similarly high fractions (54%) with distorted isophotes and (20%) with double nuclei in their sample of powerful radio galaxies at surface brightness levels (in the rest frame) brighter than 25 mag arcsec-2. Thus, the results for the morphologies of CSS/GPS radio galaxies are consistent with those found for the powerful radio galaxies with extended radio structure. We note that at low redshift, FR 2 galaxies tend to exhibit signs of interaction much more often than FR 1 galaxies (Heckman et al. 1986; Smith & Heckman 1989b; Baum & Heckman 1989; Ledlow & Owen 1995). Thus, the CSS/GPS hosts are more like FR 2 hosts than FR 1 hosts in this respect.
10 The three additional objects in the sample are now reclassified as large-scale doubles seen close to the radio axis. See section 3. Back.