Annu. Rev. Astron. Astrophys. 1999. 37: 127-189 Copyright © 1999 by Annual Reviews. All rights reserved

6.2. Spectral Content of Arc(let)s

The strong magnification of giant arcs also permits one to study the content and star formation rates of high-redshift galaxies. Preliminary studies started with Mellier et al (1991), Smail et al (1993) who explored the spectral content of some arcs. These samples do not show spectacular starburst galaxies and seem to be compatible with a continuous star formation rate. The HST images confirm that many of these galaxies have bright spots with ongoing star formation. The star formation rates inferred from new optical spectra of arcs in A2390 (Bézecourt & Soucail 1997), in A2218 (Ebbels et al 1996), in Cl1358+6245 (Franx et al 1997) or in Cl0939 (Trager et al 1997) range from 5 to 20 M / yr and are consistent with other observations (Bechtold et al 1998), but none of the rates computed for arclets are corrected from dust extinction.

It is only very recently that the material of magnified arcs has been studied in detail. Trager et al (1997) made the first attempt to estimate the metallicity of the arclets at z > 3 detected in Cl0939 with the Keck telescope, and found that they are metal-poor systems, having Z < 0.1 Z. The very first CO observations at IRAM of the giant arc in A370 [Casoli et al 1996: CO(J=2-1) detected] and at Nobeyama Observatory in MS1512-cB58 [Nakanishi et al 1997: CO(J=3-2) undetected] have demonstrated that (sub)millimeter observations are feasible thanks to the magnification and can provide useful diagnoses on the molecular and gas content of galaxies at high redshifts. If, as suggested by the cosmic infrared background (Puget et al 1996; see also Guiderdoni 1998 and references therein), a significant fraction of the UV emission of distant galaxies is released in the submillimeter range, the observations of lensed galaxies in the submillimeter and millimeter bands could be a major step in our understanding of the history of star formation in galaxies. Blain (1997) emphasized that the joint submillimeter flux-density/redshift relation and the steep slope counts make the observations of lensed distant galaxies in this waveband an optimum strategy, so a large number of bright (magnified) sources are expected. Both SCUBA (at 450 µm and 850 µm) and IRAM (at 1.3 mm) can therefore benefit from magnification of distant lensed galaxies. The large field of view and the wavelength range of SCUBA at JCMT seem perfectly suited for this program. Smail et al (1997b) are carrying out a long-term program of observations of lensing clusters with this instrument. They detected sources in A370 and Cl2244-02 with a success rate which implies that the number density of these galaxies is about 3 times higher than expected from the 60 µm IRAS count. Their observations of a new sample of 7 lensing clusters (Smail et al 1999) show that the energy emitted by these galaxies is much higher than the expectations from nearby galaxies. Most of these galaxies are at redshift larger than 1, and probably less than 5.5. The star formation activity of high-redshift galaxies is therefore important, and for those which have an optical counterpart, their morphology reveals signs of merging processes. Therefore, the star formation activity seems frequently triggered by interactions (Smail et al 1999), which corroborates the recent ISOCAM observations in some giant arcs, as in A2390 (Lémonon et al 1998). The star formation rates measured from the various fluxes have a very broad range, between 50 to 1000 M / yr, but they are difficult to estimate accurately, in particular for the hyperluminous system in A370 (Ivison et al 1998) because AGNs could contribute significantly to the flux.

The submillimeter observation is certainly one of the most promising tools for the future. The magnification and the shape of the continuum produced by dust make the "submillimeter gravitational telescope" perfect for studying the high-redshift Universe and the star formation history of galaxies.