SYSTEMSI am often asked this question, and the truthful answer is: "We do not know". Spectroscopically, DLAs are straightforward to identify (see Figure 2). The large equivalent widths and characteristic damping wings which signal column densities of absorbing neutral hydrogen in excess of N(H I) = 2 × 1020 cm-2 are easy to recognise even in spectra of moderate resolution and signal-to-noise ratio (significantly worse than those of the spectrum reproduced in Figure 2).
 |
Figure 2. The strong absorption feature
centred near 3375Å in the near-ultraviolet (UV) spectrum of the
bright QSO Q1331+170 is a good example of a damped
Ly |
The galaxies producing DLAs, however, have proved difficult to pin down. Wolfe and collaborators, who were the first to recognise DLAs as a class of QSO absorbers of special significance for the study of the high redshift universe (e.g. Wolfe et al. 1986), proposed from the outset that they are the progenitors of present-day spiral galaxies, observed at a time when most of their baryonic mass was still in gaseous form. The evidence supporting this scenario, however, is mostly indirect. For example, Prochaska & Wolfe (1998) showed that the profiles of the metal absorption lines in DLAs are consistent with the kinematics expected from large, rotating, thick disks, but others have claimed that this interpretation is not unique (Haehnelt, Steinmetz, & Rauch 1998; Ledoux et al. 1998).
Imaging studies at high redshift are only now beginning to identify some of the absorbers (Prochaska et al. 2002; Møller et al. 2002). At z < 1, where the imaging is easier (an example is reproduced in Figure 3), it appears that DLA galaxies are a very `mixed bag', which includes a relatively high proportion of low surface brightness and low luminosity galaxies (some so faint that they remain undetected in their stellar populations, e.g. Steidel et al. 1997; Bouché et al. 2001), as well as more `normal' spirals (Boissier, Péroux, & Pettini 2002).
 |
Figure 3. (Reproduced from
Pettini et al. 2000a).
WFPC2 F702W exposure of the field of Q0058+019 (PHL 938).
A model point spread function has
been subtracted from the QSO image (labelled `Q'),
revealing the presence of a galaxy (labelled `G')
approximately 1.2 arcseconds to the NE of the QSO position.
Given its proximity, this is likely to be the damped
Ly |
Much was made in the early 1990s of the apparent
correspondence between the neutral gas mass traced by DLAs
at high redshift
(
DLA
h 
1.2 ×
10-3,
when expressed as a fraction of the critical density)
and today's luminous stellar mass, leading to suggestions
that the former are the material out of which the latter formed (e.g.
Lanzetta 1993).
However, the apparent decrease in
DLA from
z = 3 to 0, upon which
this picture was based, has not been confirmed
with more recent and more extensive samples
(Pettini 2001).
As can be seen from Figure 4,
current data are consistent with an approximately constant value
DLA
over most of the Hubble time, and this includes
the most recent estimates of
H I
in the local universe from 21cm surveys
(Rosenberg &
Schneider 2002;
not shown in Figure 4).
Perhaps DLAs pick out a particular stage in the evolution
of galaxies, when their dimensions in high surface
density of neutral gas are largest, and
it may be the case that different
populations of galaxies pass through this stage at
different cosmic epochs.
 |
Figure 4. Recent estimates of the
mass density of neutral gas traced by damped
Ly |
65°, at a projected
separation of 6 h-1 kpc from the QSO sightline.