4 Discussion

The rapid, giant and persistent X-ray variability of 1ES 1927+654 detected with ROSAT and confirmed by Chandra point to a type 1 AGN classification, i.e. we have a direct view to the central region. The optical spectra, however, give a Seyfert 2 classification with some contribution from the host galaxy. There seems to be a significant discrepancy (factor $\sim$6) between the lower limit of the A_V value of 3.7 for the BLR, estimated from the optical spectrum, and the maximum A_V of 0.58 determined from the X-ray spectrum. We note that if the dust is optically thin for X-rays the intrinsic X-ray A_V can not greatly exceed the observed value of 0.58, otherwise we would observe much stronger absorption edges as well as a higher energy soft X-ray cutoff. In this section we will discuss the possible scenarios resulting in this unique combination of X-ray and optical properties.

4.1 An underluminous BLR

One possible explanation for the apparent disagreement in X-ray and optical properties is an extremely underluminous, or even absent, BLR. The luminosity derived for the putative broad H$\alpha$ component (see Fig. 10) is about $\rm 10^{38}~erg~s^{-1}$, whereas the total luminosity in the optical B-band is about  $\rm 10^{43}\ erg\ s^{-1}$. The typical ratio of emission line to total B-band luminosity in Seyfert galaxies is approximately 1 to 5 percent (Netzer, private communication), thus we would expect a line luminosity of about $\rm 10^{41}\ erg\ s^{-1}$ in 1ES 1927+654. This may be suggestive of an underluminous BLR.

4.2 An optically thick X-ray absorber and/or higher dust to gas ratios

The true optical extinction might be larger than that derived from the optical spectrum. It is possible that, for example, the dust grains in the absorbing matter could be individually thick to X-rays (cf. Fireman 1974). In this case, the soft X-ray spectrum cutoff does not change with increasing A_V values, i.e. it remains constant in the X-ray observations. In addition, we could have a large self-blanketing factor, in which case the absorbed luminosity from the BLR might be even larger than the value derived in the case where the dust is optically thin to X-rays. We also can not exclude that 1ES 1927+654 exhibits a higher dust to gas ratio than that of about 0.01 derived from interstellar dust to gas ratios (Gorenstein 1975).

4.3 Partial covering phenomena

Partial covering phenomena have recently been detected with XMM-Newton in broad- and narrow-line AGN (1H 0707-495, Boller et al. 2002; IRAS 13224-3809, Boller et al. in preparation; PG 1211+654, Reeves, in preparation, and PDS 654, O'Brien, in preparation). Compton thick material partially covers the central source, resulting in strong neutral or partially ionised Fe K absorption edges, Fe K re-emission features with fluorescent yields below 0.33, and strong soft X-ray emission. However, the limited spectral coverage of the Chandra LETG spectrum does not allow us to constrain partial coverer spectral components. XMM-Newton is required to test this hypothesis. Most probably, the absorbing material has to be Compton thick for X-ray and optical light (cf. Boller et al. 2002), which would not be in disagreement with the optical extinction derived for 1ES 1927+654.

4.4 Non-simultaneous X-ray and optical observations

We note that the X-ray observations were not taken simultaneously with the optical observations and it is possible that the obscuration is a strong function of time. However, this possibility is not very likely as the X-ray observations are spread over a time scale of 12 years which also holds for the optical spectroscopic data (see e.g., Perlman et al. 1996). Whenever we observe 1ES 1927+654 at X-rays, it displays strong and rapid variability. The optical spectrum also shows consistency over time.

4.5 Future observations

Near-infrared spectroscopy is required to further constrain the optical extinction for the BLR in 1ES 1927+654, e.g. observations of Pa$\rm\beta$, Pa$\rm\alpha$, Br$\rm\gamma$ and Br$\rm\alpha$. Assuming that we will detect broad components in these lines, this will allow us to determine upper limits for the A_V value up to about 68 (cf. Goodrich et al. 1994).

We have presented an extreme and unique combination of optical and X-ray properties in 1ES 1927+654. Clearly it is crucial to search for similar features in other galaxies. If such features are common in other objects, this will further constrain the range of suitable models.

Acknowledgements

We would like to thank the anonymous referee for many helpful and constructive comments. This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.