There are several pieces of evidence that suggest that the host galaxies of the GPS and CSS galaxies may contain significant amounts of dense gas. (1) The radio sources are strongly depolarized and some sources show very large Faraday rotation measures (section 4). (2) The emission-line nebula in some sources show evidence for strong reddening and high densities (section 8.9). And some GPS galaxies seem underluminous in emission lines (section 8.8). (3) The CSS sources show the alignment effect (section 8.3). (4) Some GPS quasars show evidence for large columns of X-ray absorbing gas and some GPS galaxies may have large columns that obscure the soft X-rays (section 9). (5) The radio morphologies are often distorted, suggesting interaction with dense clouds (section 3).
The presence of dense gas in the host galaxies has implications for how the radio sources will evolve (see, e.g., Bicknell et al. 1997). Interaction of the radio source with dense gas can slow the propagation of the jets (see, e.g., Balsara 1991; O'Dea et al. 1991; De Young 1993; Carvalho 1994, 1998) or possibly even "frustrate" them so they are unable to escape the subkiloparsec scales (see, e.g., Wilkinson et al. 1984a; van Breugel et al. 1984b; O'Dea et al. 1991). Thus, I review here the evidence for direct detection of dense gas in the host galaxies of GPS and CSS sources.
For comparison, "normal" elliptical galaxies tend to have masses of cold gas in the range 107-108 M (see, e.g., Lees et al. 1991), while the very gas-rich ultraluminous IRAS galaxies (ULIRGs) have gas masses in the range 109-5 × 1010 M (see, e.g., Sanders, Scoville, & Soifer 1991). Powerful radio galaxies tend to have cold gas masses that overlap both those ranges at about 108-1010 M (see, e.g., Knapp, Bies, & van Gorkom 1990; Mazzarella et al. 1993; O'Dea et al. 1994b; Evans et al. 1996).
Estimates (or limits) on the mass of cold atomic or molecular gas in GPS and CSS sources are given in Table 9. Estimates of molecular gas mass from either CO or thermal IR are very high for the few sources that have been detected and are in the range 1010-1011 M. These are close to the upper end of the distribution of gas masses. Given the small numbers and the likelihood of strong selection effects, this result should be viewed with caution. However, it does indicate that at least some GPS and CSS sources are in very gas-rich host galaxies. As a counterexample, the CSO 2352+495 seems relatively gas poor, though this could be due to radiative excitation of the gas as discussed by Readhead et al. (1996b).
|0116+319||CSS||G||2×1010||dust at 1 mm||1||1|
|0134+329||CSS||Q||3×1010||CO (1 0)||4|
|1345+125||GPS||G||7×1010||CO (1 0)||5||4|
|1404+286||GPS||Q||1×1011||dust at IR||7||1|
|1404+286||GPS||Q||< 2×107||HI absorption||2||2,3|
|1934-638||GPS||G||< 5×1010||CO (1 0)||9|
|2352+495||GPS||G||< 3×108||CO (1 0) absorption||10||2,5|
|2352+495||GPS||G||< 4×109||HI absorption||10||2|
NOTES. - Estimates of the mass of cold gas in the host galaxies, taken from the literature and scaled to our distance scale. Comments are as follows. (1). Assumes a gas to dust ratio of 100. (2). Assumes that the gas with the estimated column density is within 1 kpc of the nucleus, similar to ULIRGs. (3) Assumes a spin temperature of 100. (4) In 1345+125 the CO may be associated with the Seyfert nucleus instead of the radio nucleus (e.g., Dickey et al. 1990). (5) The non detection of CO absorption may be affected by radiative excitation (Maloney et al. 1994).
REFERENCES. - (1)Knapp & Patten 1991; (2) van Gorkom et al. 1989; (3) Mirabel 1990, Conway 1996; (4) Scoville et al. 1993, Wink et al. 1997; (5) Mirabel et al. 1989; (6) Mirabel 1989; (7) Knapp et al. 1990; (8)Véron-Cetty et al. 1995, Fosbury et al. 1997; (9) O'Dea et al. 1994b; (10) Readhead et al. 1996b.
Conway (1996) notes that four out of 28 large-scale radio galaxies observed by van Gorkom et al. (1989), and three out of five CSOs show HI absorption. The much higher detection fraction in the CSOs could be due to them having much more gas in their nuclei than the large-scale sources. An alternate explanation favored by Conway is that the CSOs are better suited to probe HI disks in the nuclei because of their symmetric and high surface brightness.
Curiously, the estimates/limits of atomic hydrogen are significantly less than those for molecular hydrogen. This suggests that (1) the gas is predominantly molecular, (2) most of the atomic gas avoids the line of sight to the radio source, or (3) the mass of molecular gas is overestimated.