Previous
abstract 
Next
abstract
The well known correlations of emission-line widths (from the NLR of Seyfert galaxies) with the ionization potential and/or critical density of the emitting species have ambiguous interpretation. The emission lines from highly ionized atoms tend to have high critical density. By comparing line-widths of species with similar ionization structure but different critical density, we argue that observed correlations of line-width with ionization potential and/or critical density are best interpreted as correlations with ionization potential alone.
\vskip 0.15truecm We suggest an interpretation of line-width correlations with ionization potential as a column density effect. To emit the lines of low ionization ([S II] $\lambda$$\lambda$ 6716,6731, [O II] $\lambda$$\lambda$3726,3729, and [O I] $\lambda$6300 in particular), a minimum column density, $N_{min}$, is required. Lines of high ionization (e.g. [Fe VII] $\lambda$6087) are emitted for any column density provided the ionization parameter is sufficient. Therefore, if the ratio, ${N_{col} \over N_{min}}$, where $N_{col}$ is the cloud column density, increases in the direction of decreasing velocity, correlations similar to those observed would be expected. Spherical clouds of constant mass and constant ionization parameter have ${N_{col} \over N_{min}}$ increasing inward so that observed line-width correlations can be reconciled with cloud velocity decreasing inward.
\vskip 0.15truecm One well known object which is particularly relevant to the above discussion is Mkn 463E. We present line profiles of Mkn 463E with blue wings extending out to 2,000 $km$ $s^{-1}$ in the lines of lowest critical density. Combined with the presence of strong [O III] $\lambda$4363 emission, which requires the presence of high density gas somewhere in the NLR, the presence of extended wings argues strongly for velocity increasing outward in this object. We also point out that NGC 5548 is anomalous in that the line widths appear to correlate with critical density rather than ionization potential. This work has been supported by NASA grant NAG 5-1630.
