The BeppoSAX High Energy Large Area Survey ( HELLAS ) has surveyed several tens of square degrees of the sky in the 5–10 keV band down to a flux of about 5 \times 10 ^ { -14 } { erg~ { } cm } ^ { -2 } ~ { } { s } ^ { -1 } .
The source surface density of 16.9 \pm 6.4 deg ^ { -2 } at the survey limit corresponds to a resolved fraction of the 5–10 keV X–ray background ( XRB ) of the order of 20–30 % .
The extrapolation of the HELLAS logN–logS towards fainter fluxes with an euclidean slope is consistent with the first XMM– Newton measurements , in the same energy band , which are a factor 20 more sensitive .
The source counts in the hardest band so far surveyed by X–ray satellites are used to constrain XRB models .
It is shown that in order to reproduce the 5–10 keV counts over the range of fluxes covered by BeppoSAX and XMM– Newton a large fraction of highly absorbed ( log N _ { H } = 23–24 cm ^ { -2 } ) , luminous ( L _ { X } > 10 ^ { 44 } erg s ^ { -1 } ) AGN is needed .
A sizeable number of more heavily obscured , Compton thick , objects can not be ruled out but it is not required by the present data .
The model predicts an absorption distribution consistent with that found from the hardness ratios analysis of the so far identified HELLAS sources .
Interestingly enough , there is evidence of a decoupling between X–ray absorption and optical reddening indicators especially at high redshifts/luminosities where several broad line quasars show hardness ratios typical of absorbed power law models with log N _ { H } =22–24 cm ^ { -2 } .