150 K) in front of a
rather strong continuum source. Then, there are mainly two explanations for
the large velocity range observed in absorption. The gas is either A) in
turbulent motion or B) in a regular Keplerian orbit around the nucleus.
Case A) means there are numerous cold gas clouds falling into and ejected out
of the nuclear region. To explain the similar velocity range for emission and
absorption the clouds are required to have maximum velocities similar to the
rotation amplitude of the galaxy. This model is, e.g., strongly favoured by
Mirabel & Sanders (1988) who in their paper conclude that ``... in the
central regions of the most luminous infrared galaxies there must be high
concentrations of turbulent atomic gas enshrouding the nuclear radio-continuum
source.''
Case B) requires rather high rotational velocities close to the nucleus and
was therefore often rejected. But now that interferometers and VLBI techniques
are used to resolve the nuclear continuum structure in galaxies those high
rotation velocities have been found in nearly all cases (see Section 4).
Most of the large absorption line widths can be reproduced if the rotation
amplitude in the nuclear region is similar to that in the outer region of the
galaxy. This requires either a rather flat rotation curve or two components
with similar amplitude as for example present in the Galactic rotation curve
(Dame et al. 1987). In some cases the inner part of the rotation curve has
to rise toward the nucleus to explain absorption line widths much larger
than the observed velocity range in emission. The extreme rotational velocities
of the nuclear maser emission in NGC 4258 shows what might be happening in the
centre of many starburst galaxies (see Section 4.3).
A more detailed discussion of the two cases is given by Koribalski, Dickey & Mebold (1993). In several galaxies both high rotational velocities and either infall or outflow of gas are observed.