In order to investigate the statistical properties of C IV systems, an enlarged sample has been created by merging the present data with those recently obtained at similar resolution for Q0000-26 (Savaglio et al. 1997) and Q0055 - 269 (Cristiani et al. 1995). All the spectra have been analyzed in a homogeneous way according to the same procedures described in Sects. 2 and 3.
The distribution of C IV column densities is
shown in Fig. 7 (click here).
It can be described as a single power-law distribution, 
, with 
(in agreement with Petitjean & Bergeron 1994), down to a
completeness limit 
.
Figure 7: Histogram of the logarithmic number of C IV absorptions as a
function of the column density. The solid line represents
a power-law: 
, (Petitjean & Bergeron 1994)
Figure 8: Rest equivalent width of single C IV system components as
a function of the column density (in units of
). The
overplotted curves of growth have been computed for values b = 6, 14, 22 km
s-1 of the Doppler parameter (Spitzer 1978;
Press & Rybicki 1993)
The distribution of the Doppler parameters for the same C IV lines has a
mode of 
km s-1 and a standard deviation 
km s-1.
A simple derivation of the expected 
from the
corresponding 
, with the hypothesis of pure
thermal broadening with a single temperature would provide
km s-1.
This indicates either that turbulent motions are present or
some of the lines are blends of subcomponents, which would
be separated only by observations with higher spectral resolution
and s/n.
Figure 8 (click here) shows the equivalent widths, estimated from the b
and 
values derived for each subcomponent by profile fitting,
versus the column density. It can be seen that most of the lines are
on the linear part of the curve of growth, at least for 
cm-2.
Curves of growth have been drawn for values of the Doppler parameter
km s-1.