The delay obtained for the central component with respect to the UV
continuum of is similar to that obtained for the K and
L bands, where the emission is caused by heated dust (Clavel et al.\
1989; Sect. 4.3.3). Therefore the dust, and the gas producing the central
component, seem to be coexisting in the same region. Even, if the central gas is
situated at
(more compatible with the
photoionization models) and the dust at
, it
would appear that the outer part of the gas would be (within the errors)
associated with the inner part of the dust zone. An upper limit to the dust mass
at 400 light-days has been estimated by Clavel et al. in
and we
will estimate the mass-to-dust ratio from the maximum luminosity of both
\
and
.
Assuming a typical
and that at the CIV
emission maximum all carbon is
and all ions are in the lower level,
, with a maximum of
,
. For a solar carbon abundance of
this gives
and
(Aller 1984). Then, the ratio
, similar to the galactic value of 200.
Using the
luminosity,
(Aller 1984), for the Case B of recombination
(
), with
and the maximum
luminosity of
(Wamsteker et al. 1985), we find
. Then
and the ratio
, four
times larger than the galactic value.
Note that the mass of the central gas is in the range of from the different lines and the subsequent mass(central gas)/mass(dust)
ratio extends from 100 to 750 could indicate that the abundance C/H is less
than the solar one. If we take
from the first
calculus and
from
, an
abundance
is obtained, an order of magnitude lower
than the solar abundance assumed previously. Another explanation to the difference
in the results from both lines might be that also the
mixture of gas-to-dust is stratified, as had been previously found
for the gas only. This is a consequence of the significantly different
derived gas
mass depending on the ionization state of the ion used for the calculation,
while the dust can not survive in the inner regions. This would be
consistent with the general concept from previous results suggesting stratification in
the BLR (e.g. Clavel et al. 1991).
However at the same time the coexistence of the gas and dust, and the
difference in line delays with respect to the continuum, seem to indicate that
we are most likely dealing with a situation where the mixture with
dust and the ionization stratification are all taking place in a region at
approximately
from the central source in an
anisotropic radiation field.