5 Discussion

The data presented in the previous sections indicate a systematic alignment, and possibly a multiply-imaged object in the field. The easiest interpretation of these findings is the existence of a mass in the line of sight of Slens1 which could produce both strong (one candidate multiple image galaxy but with a puzzling parity problem) and weak (a coherent tangential alignment which can be quantified through the $M_{\rm ap}$-statistics as a significant peak together with several candidate arclets distributed tangentially around the same position) gravitational lensing effects. However, with our current STIS data, it is not possible to proceed further with the analysis and to confirm that a massive dark halo is present in this field. The significance of the $M_{\rm ap}$ peak is too low to discard the possibility that it is a statistical fluctuation, even if we would not expect a better significance given the size of the field and the number of objects in it. We therefore considered the result of the $M_{\rm ap}$-analysis only to define a "center'' of alignment, relative to which the orientations of the arclets were measured. The mean of these orientations is small and is expected to occur randomly with a probability of only $0.3\%$. The arclets show a pattern strongly resembling those of lensing clusters.

If these image features are produced by a gravitational lens, it does not seem to be associated with any luminous objects visible in the deep exposure obtained with STIS. Only the location of the $M_{\rm ap}$ peak could suggest that this mass may somehow be associated with the group of bright elliptical galaxies on the center-top part of the field. If we suppose that this is the case, the mass necessary to produce the candidate multiply imaged galaxy like G1 and G2 within a 20'' radius would be similar to the mass of a rich cluster of galaxies. If the lens is really associated with those bright galaxies located close to the peak, then the M/L ratio is at least 2 orders of magnitude above the normal values for groups of elliptical galaxies. In any case, if this is a lens, the mass we find is unacounted for in terms of light for what is normally regarded as the M/L ratio in groups or clusters of galaxies. We tested also the light profile of the brightest elliptical in this group. This galaxy could be at a redshift of 0.5-0.6 based on similar galaxies in terms of magnitude and apparent size for which we have spectroscopic redshifts from the CFRS (Lilly et al. 1995). If this galaxy had a cD profile, it would strongly suggest the presence of an underlying cluster, but this galaxy clearly has a deVaucouleurs profile, which makes the test inconclusive.

In conventional models of structure formation, the most massive halos detected by aperture mass techniques like the one described in Sect. 3 would be easily observed because they should have accreted gas which, after subsequent cooling, would form stars. Nevertheless, it is possible that one could find a number of dark lenses, where the dark matter would have collapsed in structures massive enough to create a strong lens effect, while prohibiting baryonic matter to settle in the dark matter halos and initiating the star formation. So far, two other candidates have been found (Erben et al. 2000; Umetsu & Futamase 2000) but neither of them has been clearly confirmed. Also, a class of X-ray Over-Luminous Elliptical Galaxies (OLEGs, Vikhlinin et al. 1999) show similar characteristics of a dark halo, being compatible in X-ray with the mass of a poor cluster but presenting a very low optical luminosity. Recently White et al. (2001) have suggested that those detections could be explained by projection effects of several mass sheets along the line of sight. Another explanation could be that most of the objects observed in this field are at the same redshift and belong to the same gravitational structure, the preferred orientation of the galaxies being then just a product of their infall into the gravitational potential.

To confirm the reality and the nature of the $M_{\rm ap}$ peak, wider deep exposures centered on the position of the peak are necessary. At the same time, wider deep multicolor imaging would allow us to obtain photometric redshifts that would give an indication for the redshift of all the objects, including G1 and G2, in the field and its surroundings. If G1 and G2 present similar photometric redshifts then, if this is a lens, a third image should be detected in those images unless this is a very unexpected configuration. The combination of weak-lensing analysis and photometric redshift information can then be used to unveil the true nature of the objects in this field. Wittman et al. (2001) have recently used the combination of these techniques for the discovery of a new cluster of galaxies. The detection and confirmation of only a single dark-lens would already have important consequences for the standard picture of structure formation (Trentham et al. 2001). It would also show the importance of using gravitational lensing as a tool for surveys, selecting objects by mass instead of by light as in traditional methods.

Acknowledgements

We would like to thank Yannick Mellier and Ludovic Van Waerbeke for long and helpful discussions and for first pointing out at us the double image candidate. This work was supported by the TMR Network "Gravitational Lensing: New Constraints on Cosmology and the Distribution of Dark Matter'' of the EC under contract No. ERBFMRX-CT97-0172, by the DLR grant 50 OR 0106 and the Deutsche Forschungsgemeinschaft.