6.2.2 Properties of X-Ray-Selected and Radio-Selected BL Lac Objects

The properties of X-ray-selected and radio-selected BL Lac objects are systematically different (Ledden and O'Dell 1985; Stocke et al. 1985; Maraschi et al. 1986; Padovani 1992a; Laurent-Muehleisen et al. 1993). On average, XBL have lower polarization, less variability, higher starlight fractions, and are less luminous and less core-dominated than RBLs (Stocke et al. 1985; Morris et al. 1991; Perlman and Stocke 1993; Jannuzi et al. 1994).

The two classes occupy different regions on the _ro - _ox plane, which is indicative of different broad-band energy distributions (Stocke et al. 1985; Ledden and O'Dell 1985). This is illustrated in Fig. 15 (Maraschi et al. 1994b), which shows the multiwavelength spectra of a radio-selected BL Lac (PKS 0537-441) and an X-ray-selected BL Lac (Mrk 421). The overall shape of the RBL spectrum, notably the wavelength of the peak of the synchrotron emission and the relative strength of the gamma-ray emission, is similar to that of the flat-spectrum radio quasar 3C 279 (Fig. 4), as is true for RBLs and FSRQ in general (Maraschi et al. 1994b). The peak of the synchrotron component is typically at infrared/optical wavelengths for the RBL (and FSRQ) and in the soft X-rays for the XBL. If the peak wavelength changes smoothly and continuously from short (ultraviolet/X-ray) to long (infrared/optical) wavelengths for BL Lac objects as a whole, the radio/optical/X-ray colors of RBL and XBL can be reproduced easily (Padovani and Giommi 1995a).

Figure 15. Multiwavelength spectra of a low-energy cutoff BL Lac (LBL) (PKS 0537-441; filled circles) and a high-energy cutoff BL Lac (HBL) (Mrk 421; open squares). The peak synchrotron luminosity occurs at infrared/optical wavelengths for the LBL and at soft X-ray wavelengths for the HBL. In this example, the HBL is both an XBL and an RBL, and (as in most cases) the LBL is an RBL. A value of q0 = 0.5 has been assumed. (Following Maraschi et al. 1994b; figure courtesy of Rita Sambruna.)

The 1 Jy sample of RBL shows a weak positive evolution, consistent at the 2 level with no evolution (< V/V_max > = 0.60 ± 0.05; Stickel et al. 1991), while the EMSS XBL display a negative evolution, appearing less abundant and/or less luminous at higher redshifts, (< V_e/V_a > = 0.36 ± 0.05; Wolter et al. 1994). For luminosity evolution of the form L_x (z) = L_x (0) exp[c_x T (z)], where T (z) is the look-back time, the best-fit evolution for the EMSS sample is c_x = -7.0, with a 2 range of -15.9 to -1.3 (Wolter et al. 1994), nearly consistent (at the ~ 2 level) with zero evolution. For the 1 Jy RBL sample, c_r = 3.1 with a 1 range of 1.7 to 4.2 (Stickel et al. 1991, with c = 1/), also consistent with zero evolution at the 2 level.

Both samples are still relatively small (30 and 34 objects for XBL and RBL, respectively), so it is premature to draw any conclusion regarding the supposedly different evolutionary behaviors of the two classes. This is confirmed by two additional results. First, for the fourteen RBL of the S4 survey (f_r 0.5 Jy at 5 GHz; Stickel and Kühr 1994), < V/V_max > = 0.44 ± 0.08 (Padovani and Giommi 1995a), more like the EMSS XBL. Second, for the EMSS, raising the flux limit only slightly, to 10^-12 erg cm^-2 s^-1, changes the mean V_e/V_a from 0.33 ± 0.06 to 0.48 ± 0.06 (Della Ceca 1993). These results for RBL and XBL are completely consistent with no evolution.