4. Discussion and conclusions
In the previous section we have discussed the possibility of a dependence on redshift of the AGN2/AGN1 ratio. In particular, this ratio remains constant or slightly increases up to 2, and then decreases. This would suggest that AGN2 are a later evolutionary stage of the AGN phenomenon, a possibility worth to be explored theoretically.
Another possibility is that the decreasing fraction of AGN2 for z2 is only apparent, and that in reality there is an increase of the fraction of sources with cm-2, i.e. completely hidden at all X-ray energies. This could be linked with the star formation rate history, which is observed to increase with the redshift up to z2, and than stays constant (Madau et al. 1996; Rowan-Robinson 1999). A high star formation rate would imply a large amount of dust and gas, and then a large absorption.
A different approach in fitting the XRB consists in a luminosity-dependent number ratio , as has already been done by Gilli et al. (1999b).
The direct way to discriminate between different evolutionary models is to study the AGN2 XLF, a task within the capabilities of the new generation X-ray missions (Chandra and XMM).
In Table 1 and Table 2 we report the AGN1 and AGN2 densities for different flux limits corresponding to the models without and with the inclusion of the 30% increase in the normalization of the XRB. The effect of shows up in an AGN2 percentage decreasing at lower fluxes, a consequence of the sampling at higher redshifts where the AGN2/AGN1 number ratio decreases. It is worth noting that erg cm-2 s- 1 and erg cm- 2 s-1 are the flux limits expected for the Deep observations of the Lockman Hole and the Hubble Deep Field scheduled for Chandra and XMM, respectively.
Table 1. AGN2 percentage prediction as a function of the sampling flux in the 5-10 keV band. The second and third column indicate the number density of AGN1 and AGN2.
© European Southern Observatory (ESO) 2000
Online publication: December 17, 1999