Redshift Distribution

The distribution of redshifts among the DXRBS FSRQs identified so far (Figure 3) is quite similar to that of the 1 Jy and S4 samples. Not much more can be said at present. A more thorough comparison will have to await completion of our survey and the convolution of the redshift distribution with the WGACAT sky coverage. As has already been mentioned by Hook et al. (1996), selecting flat-spectrum sources is an efficient way of finding high-redshift, radio-loud quasars. We would therefore expect that the majority of the FSRQs would be at high redshifts (z > 1), a suspicion which the data confirm. The mean redshift of our FSRQ sample is $z\sim 1.2$, and the high-redshift tail extends to $z \sim 4$. Our redshift distribution cannot, however, be directly compared to that of Hook et al. (1996) because those authors required that the sources be particularly ``red,'' a criterion we do not impose.

While a significant fraction (7 of 33) of the BL Lacs so far included in our sample still lack redshifts, it is apparent that the redshift distribution for the DXRBS BL Lacs (Figure 4) is dominated by objects at z<0.4, and therefore more similar to that of the Slew Survey (Perlman et al. 1996a; or the EMSS survey: Stocke et al. 1991) than the 1 Jy sample (Stickel et al. 1991), which is less peaked at low redshifts and has a much more prominent high-z tail than either the DXRBS or Slew Survey samples. As above, a direct comparison between these distributions is not possible at present, given also the different survey methods and selection criteria. We only mention here that the difference between the DXRBS and 1 Jy redshift distributions appears to be striking considering that the radio flux limit of DXRBS is a factor 20 lower so that if anything, one would expect a higher fraction of high-z sources, rather than the dearth of high-z sources that we observe. The flatter distribution of the 1 Jy sample might be related to the fact that it misses some low redshift objects hidden within bright hosts (which might be misidentified as radio galaxies; see §7) and might also be contaminated by misidentified quasars at high redshifts (e.g., 1 Jy 1308+326), as suggested by Marchã & Browne (1995) and Perlman et al. (1996b). Some 1 Jy BL Lacs might be gravitationally lensed if superposed by chance onto a lower-redshift galaxy (as originally suggested by Ostriker & Vietri 1985). This last possibility now appears quite likely from the results of Stocke & Rector (1997), which show that the 1 Jy BL Lacs have a $2.5-3 \sigma$ excess of Mg II absorbers along their sight lines.