The RGB Sample of Intermediate BL Lacertae Objects

Combining newly identified and previously known BL Lacertae objects from the ROSAT All-Sky Survey-Green Bank (RGB) catalog, we present a sample of 127 BL Lacertae objects, the largest ever derived from a single uniform survey. A complete sample of 33 objects brighter than O = 18.0 mag is also presented. These samples are compared to other known BL Lac samples and are generally found to exhibit properties intermediate between those of the previously disparate classes of high- and low-energy-peaked BL Lacertae objects (HBLs and LBLs, respectively). This result is most dramatic in the distribution of the X-ray to radio logarithmic flux ratios, where the RGB BL Lacertae objects are shown to peak precisely where the sharp dichotomy between the two subclasses was previously seen. The αro versus αox diagram also shows the RGB sample smoothly bridges the gap between the previously distinct subclasses of LBLs and HBLs. The range of broadband spectral energy distributions (SEDs) exhibited by the RGB objects also shows that, contrary to prior claims, searches based on relatively deep surveys cannot limit follow-up spectroscopy to targets with a narrow range of SEDs since BL Lacertae objects clearly constitute a homogeneous population with a wide range of SEDs. As in results based on the Einstein Extended Medium Sensitivity Survey (EMSS) and 1 Jy BL Lac samples, we find a weak but statistically significant correlation between the composite X-ray spectral index αxox and the radio-optical spectral index αro. This implies that the more LBL-like RGB BL Lacertae objects have secondary sources of X-ray emission, possibly from inverse Compton components. This result, in addition to other characteristics of the RGB sample, indicates that the simple unified scheme, which postulates that HBLs and LBLs differ solely by orientation, may be in need of revision. We also present both the X-ray and radio log N- log S distributions for which the competing HBL/LBL unification scenarios have differing predictions. The unknown effects of the triple flux limit inherent in the RGB Complete sample makes quantitative analysis uncertain, but the characteristics of the RGB sample compare well both with results obtained from previous samples and with general theoretical predictions based on a simple Monte Carlo simulation. Our analysis indicates that the unimodal distribution of BL Lac properties found in the RGB sample likely reliably reflects the underlying population, while the bimodal distribution found in earlier studies arose primarily from observational selection effects. The presence of not only intermediate but also extreme HBL and LBL objects is the RGB survey's unique strength and offers clear avenues for future studies that can undoubtedly address the question of how HBLs and LBLs are related.