The Canada-France-Hawaii Telescope Imaging Survey of BL Lacertae Objects. I. Properties of the Host Galaxies

The results of an extensive imaging survey of 50 BL Lac objects conducted at the Canada-France-Hawaii 3.6 m Telescope (CFHT) are presented. This paper details the results pertinent to the host galaxies of BL Lac objects; a companion paper presents results on the clustering environments of BL Lac objects obtained from the same images. All images were obtained through a Gunn r-filter and most during subarcsecond seeing conditions. The sample selected for observations includes both X-ray and radio-selected BL Lac objects with known redshifts < 0.65. Because high-z and unknown redshift (i.e., optically very featureless) BL Lac objects are excluded from this study, these results may not be pertinent to all BL Lac objects. Forty-six of 50 BL Lac objects are resolved by these images, 36 of which were resolved well enough to allow a host galaxy classification based upon their surface brightness profiles. Twenty of these 36 are best-fit unambiguously by elliptical profile host galaxies, 10 are likely ellipticals, three are possible spirals (but with problems with their profiles that make this conclusion tentative), and three host galaxies are best-fit unambiguously by spiral galaxy profiles. One of the three best-fit by a spiral profile (MS 0205+351) is also the only point source which is significantly decentered relative to its "host galaxy" and the lowest luminosity galaxy in the sample. This one BL Lac object and the other two in spiral galaxies (PKS 1413+135 and OQ 530) are the only candidates for gravitationally lensed BL Lac objects to be identified from this study. Comparisons of host galaxy luminosities and morphology made internal to the sample find no evidence that any BL Lacertae subtype with distinct characteristics (e.g., high vs. low polarization; presence of weak emission lines; X-ray vs. radio-loud) has distinct host galaxy properties. This is particularly important for the X-ray versus radio- selected BL Lac objects which have been suspected of being different types of objects on the basis of their discrepant (V/V_max_) values, redshift distributions, and emission-line properties. Therefore, the results here support a picture for BL Lac objects in which the X-ray-loud and radio-loud objects are from the same parent population with the X- ray-loud BL Lac objects viewed further from the beaming axis than the radio-loud BL Lac objects. Comparisons between the CFHT BL Lacertae sample and large, nearby samples of rich cluster brightest cluster galaxies (BCGs) and Fanaroff- Riley (FR) type 1 and 2 radio galaxies find that the host galaxy luminosity distribution is most consistent with the FR 2's, not with the FR 1's as suspected previously using smaller BL Lacertae samples. Paper II in this series shows that the clustering environment of BL Lac objects is also more consistent with that of FR 2's than FR 1's, confirming with a larger sample the suggestion by Prestage & Peacock that BL lac objects avoid rich clusters. And yet it is well known that the extended radio emission in BL Lac objects is most consistent with FR 1's in luminosity and morphology. The resolution to this apparent contradiction is that the parent population of BL Lac objects is only a subset of FR 1's which excludes at least the BCGs in rich clusters and possibly all radio galaxies in rich clusters as well.