We present a comparison of X-ray and optical luminosities and luminosity functions of cluster candidates from a joint optical/X-ray survey , the ROSAT Optical X-ray Survey ( ROXS ) .
Completely independent X-ray and optical catalogs of 23 ROSAT fields ( 4.8 square degrees ) were created by a matched-filter optical algorithm and by a wavelet technique in the X-ray .
We directly compare the results of the optical and X-ray selection techniques .
The matched-filter technique detected 74 % ( 26 out of 35 ) of the most reliable cluster candidates in the X-ray-selected sample ; the remainder could be either constellations of X-ray point sources or z > 1 clusters .
The matched-filter technique identified approximately 3 times the number of candidates ( 152 candidates ) found in the X-ray survey of the same sky ( 57 candidates ) .
While the estimated optical and X-ray luminosities of clusters of galaxies are correlated , the intrinsic scatter in this relationship is very large .
We can reproduce the number and distribution of optical clusters with a model defined by the X-ray luminosity function and by an L _ { x } - \Lambda _ { cl } relation if H _ { 0 } = 75 km s ^ { -1 } Mpc ^ { -1 } and if the L _ { x } - \Lambda _ { cl } relation is steeper than the expected L _ { x } \propto \Lambda _ { cl } ^ { 2 } .
On statistical grounds , a bimodal distribution of X-ray luminous and X-ray faint clusters is unnecessary to explain our observations .
Followup work is required to confirm whether the clusters without bright X-ray counterparts are simply X-ray faint for their optical luminosity because of their low mass or youth , or a distinct population of clusters which do not , for some reason , have dense intracluster media .
We suspect that these optical clusters are low-mass systems , with correspondingly low X-ray temperatures and luminosities , or that they are not yet completely virialized systems .