General discussion

In all three powerful radio galaxies we have studied, H I absorption was found. In a larger sample of more northerly, on average weaker sources, van Gorkom et al. (1989) detected four radio galaxies with absorption in a sample of 29. It should be noted, however, that they did not have the sensitivity to detect very low optical depth absorption as in PKS 1934-638. Moreover seven of the sources in their list are now known to be BL Lac like and so should not be included in the comparison.

 

$$\tau$$ Name Position z logP_1.4 FWZI ref Table 3a:

3C 111 0415+379 0.0485 S1 FR II 26.2 <0.01 1

Hya A 0915-118 0.05414* S3 FR I 26.7 0.0015 110 2

3C 236 1003+351 0.0988 S3 FR II 26.1 0.03 370 1

4C 12.50 1345+125 0.1217 S2 CSO: 26.5 0.014A 950 3

PKS 1353-341 1353-341 0.2230 S2 Flat sp 26.1 0.122 500 4

3C 321 1529+242 0.0959 S2 FR II 26.2 0.018A 380 5

PKS 1814-637 1814-637 0.064: S2 CSO 26.3 0.214 300 4

PKS 1934-638 1934-638 0.1818 S2 CSO 27.4 0.0022 100 4

Cyg A 1957+405 0.05562* S2 FR II 28.4 0.06 420 6

3C 433 2121+248 0.1015 S2 FR I/II 26.7 0.0051A 800 3 Table 3b:

4C 31.04 0116+319 0.0598 S2 CSO 25.6 0.035 380 1,7

PKS 0625-354 0625-354 0.0553 FR I 25.6 <0.015 1

S5 1946+70 1946+708 0.1007* (S) CSO 25.6 0.05 350 8

PKS 2206-237 2206-237 0.0865 S2 25.7 <0.007 1

PKS 2322-123 2322-123 0.0822* S3 25.7 780 9

References: (1) van Gorkom et al. 1989 (2) Taylor 1996 (3) Mirabel 1989 (4) this paper (5) Mirabel 1990 (6) Conway & Blanco 1995 (7) Conway 1996a,b (8) Peck et al. 1999 (9) Taylor et al. 1999.

$$\tau \pm$$ Notes:

$$\Delta \pm$$

$$\pm$$

$$\pm$$

$$\pm \tau \alpha$$

To further investigate whether powerful radio galaxies are perhaps more prone to H I absorption, we have assembled in table 3a the available data for radio galaxies with log(P$_{1.4\rm GHz})\gt$ 26 (with P in W Hz^-1). We have found data for ten such objects and in all but one, H I absorption has been detected. In five objects with log(P$_{1.4\rm GHz}$)=25.5-26.0 presented in table 3b there are three positive detections. In table 3 we have excluded BL Lac like objects, defined as sources listed in table 2 of the Véron-Cetty & Véron (1998) catalogue. Before concluding from the results in table 3 that powerful radio galaxies are particularly prone to H I absorption we have to take into account that optical depth limits and angular resolution have been different in different studies and that it is not obvious that negative results have always been reported in the literature.

To construct a more unbiased sample we have searched the catalogue of Burbidge & Crowne (1979) and the 1 Jy catalogue of Stickel & Kühr (1993) for radio galaxies with z<0.1 and log(P$_{1.4\rm GHz})\gt$26. In this sample of 17 objects, the six sources with z<0.1 in table 3a are included, of which four have H I absorption with $\tau \gt$0.017, while two have $\tau <$0.017. For the other 11 sources ( 0106+130, 0349-278, 0404+035, 0518-458, 0945+076, 1549+202, 1717-009, 1733- 565, 18 42+455, 2243+394 and 2356-611) no data have been reported. The two most southerly of these could only have been observed with the AT and have not been. We conclude that four powerful radio galaxies have H I absorption with $\tau \gt$0.017 in a sample of at least six and at the very most 15, the latter corresponding to the unlikely case that for all others, upper limits would have been observed but not reported.

The measured optical depth may depend on the angular resolution. The values of the six low redshift sources in table 3a generally correspond to what the VLA would have measured, except for 3C 321 which was observed at Arecibo with lower resolution. Since H I absorption appears to occur mainly in the nuclear regions, the Arecibo optical depth should be a lower limit to what would result from an observation at VLA type resolution.

Turning now to lower luminosity sources, we find in the van Gorkom et al. (1989) sample, observed with the VLA, 16 radio galaxies with log(P$_{1.4\rm GHz})<$25.5 and with observations sufficiently accurate to establish upper limits of $\tau <$0.017. Only two were positively detected.

A sample of dominant cooling flow cluster galaxies was observed at Arecibo by McNamara et al. (1990). All had z<0.1 and log(P$_{1.4\rm GHz})<$25.5. Eight sources with accurate observations had $\tau <$0.01 (one in common with van Gorkom et al.) and there were no detection above this level. Had the four sources detected in our high luminosity sample been observed with Arecibo type resolution, three would still have had $\tau \gt$0.01.

Even though carefully selected samples are needed for a clear result, we conclude that there is probably a tendency for powerful radio galaxies to have a higher frequency of H I absorption. Since Compact Symmetric Objects (CSOs) and steep spectrum cores are generally high luminosity sources it may also be that, as suggested by Conway (1996b), these classes are particularly prone to H I absorption. However Cygnus A shows that also other types of luminous sources are involved.