$\begin{figure} \par\includegraphics[width=7cm,clip]{H3208F4.ps} \end{figure}$

Figure 4: Relative abundances of $\alpha$ /Fe-peak elements. In all panels, open stars are literature values (Molaro et al. 2001; Outram et al. 1999; Lu et al. 1998; Ge et al. 2001; Pettini et al., in prep; Lu et al. 1996; Centurión et al. 2000; Prochaska et al. 2001), the filled triangle is for the DLA found towards Q0201+1220 in a galaxy group at $z \sim 3.4$ (Ellison et al. 2001), filled circles and squares represent values determined for DLAs A and B respectively. The typical error bar is shown in the lower right corner of the bottom panel.

We have presented high resolution spectroscopic observations of a pair of DLAs separated by $\sim$ 2000 km s^-1 ( $\sim$ 11 Mpc h₇₀^-1) at $z_{\rm abs}$ $\sim 2$ and propose the possibility that these absorbers may form part of an extended structure. Absorbing structures of similar size at high z have previously been inferred from multiple lines of sight, for example the posited super-cluster at $z \sim 2$ towards Tol 1037-2704 (Sargent & Steidel 1987). However, this is the first time that the abundances of two proximate DLAs in a single line of sight have been studied in detail. One other multiple system, a triple DLA towards CTQ 247 (Lopez et al. 2000), has been recorded in the literature and a program has been initiated to study its abundances.

Although we have found that the metallicity of both DLAs is typical of that measured in other such systems at the same redshift, there is evidence from their unusual abundance ratios that these absorbers may have experienced somewhat different star formation histories than most other documented DLAs. This evidence comes from the generally low $\alpha$ /Fe-peak abundances of both DLAs, and the DLA in the group towards Q0201+1120 (Ellison et al. 2001), suggestive that their global environment may have played some role in their evolution. Despite the possible effects of saturation (in the case of O I $\lambda$ 1302) and atypically low amounts of dust (which may explain [S/Fe]), the low [Si/Fe] compared to the large number of literature values remains convincing evidence that star formation history may be responsible for these ratios. Such low $\alpha$ /Fe-peak abundances are usually interpreted as the signatures of low star formation rates, or of an ISM enriched by an early generation of stars that is now evolving quiescently. Therefore, the abundances observed here are reminiscent of the truncated activity observed in lower z clusters.

At intermediate redshifts ( $z \sim 0.5$ ), observations indicate that star formation is suppressed in rich cluster galaxies (possibly after an initial burst of enhanced activity) by processes such as ram pressure and tidal stripping (Couch et al. 2001 and references therein). Just as this process manifests itself as an excess of post-starburst (E + A) galaxies with no strong emission lines but strong Balmer absorption, so we may expect to see the chemical signature of star formation truncation. However, it would be somewhat surprising if such processes already have such an effect on star formation at $z \sim 2$ where environments are relatively poor and canonical rich clusters have yet to form. In addition, although the kinematics of these DLAs, determined from the unsaturated metal lines, extend over $\sim$ 100 km s^-1, this is not atypical of the range exhibited by other DLAs, i.e. there is no evidence for significant disturbance of the ISM.

Follow-up observations of this field, by either multi-colour (to determine photometric redshifts) or narrow band Lyman $\alpha$ imaging are clearly of great interest to confirm the richness of the environment around these DLAs. In addition, it is important to identify other multiple DLAs and follow them up with high resolution spectroscopy in order to ascertain whether DLAs in groups exhibit distinct abundances from their "field'' counterparts.

5 Discussion