The study of quasars at high redshift has gained, in the last years, a remarkable position in cosmology because of its important contribution to the knowledge of the formation and evolution of cosmological structures. Besides the study of the quasar population itself, the analysis of the quasar absorption spectra allows us to probe the intervening matter up to the redshift of the emitting object.
Thanks to the improvement in astronomical instrumentation, a considerable amount of high resolution data has become available in recent years, greatly increasing the knowledge of the nature and evolution of the absorbers.
The absorption spectra of QSOs at high redshifts are characterized by a
"forest'' of lines shortward of the Ly
emission, first noted by
Lynds
(1971), who correctly interpreted them as Ly due to absorbers
distributed along the line of sight.
The large range in column densities of these lines suggests the presence of very different intervening structures, from fluctuations of the diffuse intergalactic medium to the interstellar medium in protogalactic disks.
Profile fitting techniques (e.g. Fontana & Ballester 1995; Hu et al. 1995) provide the few parameters (redshift z, Doppler width b and column density N) constituting the basis for the interpretation of the physical properties of the absorbers. The reliability of such parameters depends strongly upon the resolution and the signal to noise ratio of the spectral data. Controversies arisen in the past may be traced back to the difficulty of the observational problem, at the limit of the present technology (Pettini et al. 1990; Carswell et al. 1991; Rauch et al. 1993).
Remarkable results have been recently obtained in the study of
the clustering properties of Ly clouds.
The absence of clustering
for velocity separations between 300 and 30000 km s-1 has been
assessed (Sargent et al. 1980, 1982;
Bechtold et al. 1987; Webb & Barcons
1991).
On the other hand,
the presence of a significant non-zero correlation function
has been observed for Ly lines with 
km s-1 and 
(Chernomordik 1995;
Cristiani et al. 1995; Meiksin &
Bouchet 1995, Hu et al. 1995;
Fernández-Soto et al. 1996).
The correlation function for the present sample of Ly lines
is discussed in Cristiani et al. (1997).
Lanzetta et al. (1995) inferred from observational results
that, at 
, at least 32% (but it could be as high as 60%) of the
Ly absorption systems arise in luminous galaxies. This conclusion
is at variance with the longstanding belief that Ly-forest absorption
systems arise in intergalactic clouds. Moreover, recent spectra at very
high resolution have revealed the presence of C IV
absorptions associated with Ly lines with 
(Cowie et al. 1995; Tytler et al. 1995;
Womble et al. 1996; Songaila & Cowie
1996). The derived abundances seem to be similar to the ones derived
for the heavy elements absorptions originated in galactic halos, suggesting
continuity in their physical properties.
The clustering of C IV absorption systems has been investigated in the
past using large samples observed at low resolution - FWHM > 1 Å -
(33 QSOs in Young et al. 1982; 55 QSOs in Sargent et al. 1988 ).
From these data it has been possible to assess that C IV systems
do cluster on scales 
km s-1, but, due to the limited spectral resolution, the
characteristic clustering scale and the
clustering properties at smaller velocity separations have not been
established.
Petitjean & Bergeron (1994) analyzed a sample of 10 QSOs with a higher spectral resolution. They obtained as best fit of the two point correlation function in the range 30 - 1000 km s-1 a sum of two Gaussian components with dispersions of 109 and 525 km s-1, similar to the distribution observed for Mg II systems (Petitjean & Bergeron 1990). This result further confirms that metal systems as identified by C IV absorptions features arise in galactic halos.
More recent works by Womble et al. (1996) and
Songaila &
Cowie (1996) analyzed clustering of lower column
density samples detecting a lower signal on the same
scales studied by Petitjean & Bergeron (1994).
This trend with column density is consistent with what is observed for
Ly lines (see Cristiani et al. 1997).
In the present paper we address these issues on the basis of the absorption
spectrum of the quasar PKS 2126-158 (
).
This object is one of the brightest quasars
known (
) and for this reason its spectrum has been
studied in many previous
works (Jauncey et al. 1978; Young et al. 1979;
Sargent & Boksenberg 1983;
Meyer & York 1987; Sargent et al. 1988;
Sargent et al. 1989; Sargent et al. 1990;
Giallongo et al. 1993 hereafter GCFT).
Here we present new data, providing improved quality spectra in terms
of s/n ratio, resolution and wavelength range.
The paper is organized as follows.
In Sect. 2 (click here) we describe the observations and the data reduction,
in Sect. 3 (click here) the method of analysis.
Section 4 (click here) discusses the statistical properties of the Ly sample.
The description of the metal-line systems is given in Sect. 5 (click here).
Section 6 (click here) analyses some statistical properties of the C IV absorption systems.
In Sect. 7 (click here) the two point correlation function is computed for the C IV
lines and the correlation between clustering and column density is
investigated.
The conclusions of the paper are in Sect. 8 (click here).