Astron. Astrophys. 355, L31-L33 (2000)

3. Discussion

The proposed modification, while retaining the basic and most successful characteristic of the unification model (i.e. that all intermediate and type 2 Seyferts harbour an obscured type 1 nucleus), explains, at least qualitatively, the observed differences between the average properties of Seyfert 1s and Seyfert 2s and between the column density distributions of intermediate and strict type 2 Seyferts. This model is somewhat different (even if on a similar line of thought) from that proposed by Maiolino & Rieke (1995), and takes into account recent observational results. The two `flavours' of Seyferts, with and without the thick torus, may represent either two different branches in the AGN evolution, or a different evolutionary stage in the life of any (or most) source.

Testing the proposed model would require further studies of the statistical properties of Seyferts, which caution in separating intermediate and strict type 2 sources. For instance, as the torus is expected to be axially symmetric (while the dust lanes are probably more randomly distributed), a correlation between Compton-thick absorption and presence of ionization cones and large polarization is expected. Another test is to search for the presence of the torus in Seyfert 1s, as we predict that many sources of this class do not have it. This may be done either by searching for strong IR emission (which however may be related to starburst rather than reprocessing of UV/X-rays from thick matter), of by searching for signatures of X-ray reprocessing (e.g. Ghisellini, Haardt & Matt 1994; Krolik, Madau & Zycki 1994). The latter way have already produced a clear case of a Seyfert 1 with the torus: NGC 4051 was caught by BeppoSAX when the nuclear emission was switched-off, and clear evidence for reprocessing by large amount of optically thick distant matter was present (e.g. a 600 eV equivalent width iron line and a cold reflection continuum, Guainazzi et al. 1998). However, this kind of observations based on variability requires rather extreme behaviours of the X-ray emission, and it is not clear how common a switching-off of the nucleus can be. More promising is to search for narrow (i.e. unresolved) iron K lines in addition to the relativistically broadened component from the accretion disc. When the nuclear X-ray emission is directly visible, equivalent widths of the order of 50-100 eV are expected from the torus (Ghisellini, Haardt & Matt 1994). While ASCA and BeppoSAX results have been rather ambiguous in this respect, such a search is certainly within the capabilities of XMM.

© European Southern Observatory (ESO) 2000

Online publication: March 9, 2000
helpdesk.link@springer.de