Comparing the results of Sections 3.1-3.4 reveals an apparent
contradiction. X-ray observations show that the column density of cold
(neutral) gas along the line of sight to the primary X-ray source is
. However, both the optical
continuum source and the BLR are highly reddened. On the basis of local
(Galactic) studies, we would expect a cold gas column density of 
to be associated with the dust
responsible for the reddening. This is more than an order of magnitude
above the X-ray limits. Since the X-ray emission is thought to occur
deeper within the central engine than either the optical continuum or
optical line emission, this result is somewhat surprising.
This discrepancy for MCG-6-30-15 was first hinted at by Pineda et al. (1980) and explicitly commented upon by Reynolds & Fabian (1995) on the basis of data drawn from the literature. A similar discrepancy has been found for IRAS 13349+2438 (Brandt, Fabian & Pounds 1996), an infrared luminous quasar which also displays a prominent warm absorber. These authors discuss various resolutions of this discrepancy. To summarize these discussions, it is found that the only way to reconcile this result with a plausible geometry and a physically reasonable gas-to-dust ratio is to suggest that the dust resides in the ionized gas that constitutes the warm absorber. The soft X-ray opacity of this material is less than that of cold material primarily due to the almost complete ionization of hydrogen and helium. The dusty warm absorber hypothesis has also been explored by Reynolds (1997) in the context of a sample of Seyfert galaxies. We now discuss dusty warm absorbers in more detail.