Based on observations obtained at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Incorporated, under NASA contract NAS 5-26555.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

However, we investigated the possibility that lines did not share common kinematics, but we did not find evidence of this.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... panel

The different trend (high increase) of $\chi^2_{r}[scaled]$ at small angles with respect to NGC 3998 (Paper I, Fig. 11) is determined by the a priori fixing of an upper limit for ${\it\Upsilon _V}$ values.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...Ferrarese & Ford (2005)

We normalized the central velocity dispersion to an aperture of radius equal to 1/8 of $R_{\rm e}$ and derived $\sigma_{\rm R_{\rm e}/8} = 255 \pm 17$ $\rm {~km ~s}^{-1}$ .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... form

A similar trend is obtained using the correlation in the Ferrarese et al. form.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.