2 Observed Properties of 0840+3633 and 1556+3517

2.1 Optical Spectra and Photometry

Optical spectra obtained with the Lick Observatory 3m telescope for FIRST J084044.5+363328 (0840+3633) and FIRST J155633.8+351758 (1556+3517) are displayed in Figure 1, shifted to the rest frame, along with the spectrum of 0059-2735 (from Weymann et al. 1991). Unlike most quasar spectra, which are characterized by strong, broad emission lines, these BAL QSO spectra are completely dominated by their absorption features; emission lines are relatively weak or absent. In Figure 1, the distinguishing absorption features are indicated; an asterisk denotes a metastable iron feature. In the spectrum of 0840+3633, the unabsorbed remnant of the Mg II 2800Å emission line yields a redshift of 1.22. Its spectral features are very similar to those of 0059-2735, albeit with deeper absorption troughs, and it is unmistakably a more extreme example of this class of BAL QSO. The spectrum of 1556+3517 has no prominent emission lines, but the absorption features are so similar to 0840+3633 and 0059-2735 that there is again no doubt that it is the same kind of object with even higher absorption. Not only are the broad absorption lines nearly saturated, but the continuum is heavily absorbed. The absorption features indicate a redshift of 1.48. In the two new BALs, the broad lines of Al II, Al III, and Mg II have nearly complete absorption in their cores. The absorption due to metastable excited lines of Fe II and Fe III (Hazard et al. 1987) is also deeper in the two new objects than in 0059-2735.

  
Figure 1: Lick Kast 3m spectra of 0840+3633 and 1556+3517 compared to the spectrum of 0059-2715 from Weymann et al. (1991). Prominent absorption features are marked; an asterisk denotes metastable states of iron.

For 0840+3633, the APM POSS I magnitudes are O = 17.3 and E = 15.9, while for 1556+3517, O = 21.2 and E = 18.7. Both objects are redder than typical intermediate redshift quasars, for which O - E 0.5 (FBQS). For 0059-2735, Hazard et al. (1987) estimate R = 17.1 and Hewett et al. (LBQS, 1995) list = 18.13. We adopt H = 50 km s^-1Mpc^-1, q = 0.1, and an optical spectral index = -1.0 to compute the absolute R-band magnitudes. The pertinent data for the three iron Lo-BALs are listed in Table 1.

Table 1: BAL Quasar Vital Statistics^a

   ID         RA         Dec      z     B     R      MR   S1400b  S5000   Log(L1400)  Log(R*)
                 (J2000)                                  (mJy)  (mJy) (ergs s-1Hz-1)

0059-2735  01 02 17.1 -27 19 51  1.62  18.1  17.1  -28.9   ...  <0.36c   <31.71e   <0.04
0840+3633  08 40 44.5 +36 33 28  1.22  17.3  15.9  -29.3   1.6   ...      32.03     0.37
1556+3517  15 56 33.8 +35 17 58  1.48  21.2  18.7  -27.0  30.6  <27.0d    33.50     3.18

^a for H_o = 50, q_o = 0.1, Alpha(Radio) = -0.1, Alpha(Optical) = -1.0
^b FIRST Survey, Becker et al. 1995
^c Stocke et al. 1992
^d GB 5GHz Survey, Becker et al. 1991
^e Estimated from S₅₀₀₀ upper limit

2.2 Radio Properties

Both new BAL QSOs are point sources in the FIRST catalog. The 1.4 GHz flux densities for 0840+3633 and 1556+3517 are 1.3 and 30 mJy, respectively. The Greenbank 5 GHz catalog (Becker, White, & Edwards 1991) lists a flux density of 27 mJy for 1556+3517, giving a spectral index of -0.1, assuming no variability. Stocke et al. (1992) provide an upper limit of 0.36 mJy for the 5 GHz flux density of 0059-2735. The flat radio spectrum for 1556+3517 could be indicative of relativistic beaming, making a calculation of radio luminosity problematic, but we have computed L for all three sources assuming isotropic radiation (Table 1) and adopting a spectral index of -0.1.

We have computed the ratio of 5 GHz radio to 2500Å optical flux, log(R), using equations 1-3 of Stocke et al. (1992). No K-corrections have been applied. The BAL 0059-2735 is radio-quiet with log(R) 0, 0840+3633 is ``radio-moderate'' with log(R) = 0.4, 1556+3517 has log(R) > 3. The usual dividing line between radio-loud and radio-quiet is log(R) 1 (e.g. Stocke et al.), so, even with the uncertainties in the optical photometry, 1556+3517 is the first known radio-loud BAL QSO.

Could the conclusion that 1556+3517 is radio-loud be a consequence of its large optical absorption? Its absolute B magnitude is 3.4 magnitudes fainter than 0059-2735 and 0840+3633; if this difference is attributed entirely to attenuation, then log(R) drops to 1.8, still a radio-loud object. Additionally, it has L = 10^33.5 ergs s^-1Hz^-1, well above the usual divide between radio-loud and radio-quiet of 10^32.5 based on radio power alone (Schneider et al. 1992; Stocke et al. 1992).