... formation?

The work presented here is based on data obtained at the ESO Very Large Telescopes, programs 68.A-0361(A) and 66.A-0660(A).

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

...Rigopoulou et al. 2002)

This lower limit will probably exclude the majority of bound galaxy satellites since most local spirals' satellites have velocity differences of less than 300 km s^-1, Zaritsky et al. (1997).

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

... estimation

Two N(H I $)=2\times10^{20}$ cm^-2 DLAs separated by $\Delta v =1~000$ km s^-1 will mimic a single system in $FWHM \sim5$ Å data if $S/N \la 20$, but for $\Delta v \sim 500$ km s^-1 they will do so unless $S/N \ga 50$.

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

... VPFIT

Available at http://www.ast.cam.ac.uk/~rfc/vpfit.html

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

... value

We treat the upper limit toward DLA B in Q2314-409 as a detection; a lower value than this does not affect the KS probability.

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

...$-2.4<\log(Z/Z_{\hbox{$\odot$ }})<-0.8$

Of course, uniformly distributed values of H I and Z do not represent the observed distribution of DLAs, this range is purely for the purpose of investigating the effects of ionization.

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

... transition

e.g., Q2206-19A is the only DLA in the sample that relies solely on Fe II $\lambda 1611$.

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