Over the past few years a number of strongly-lensed distant galaxies have been found where the magnification is high enough that the source object can be studied at a level of detail that is impractical for typical galaxies at that redshift. An excellent recent example is the very good paper by Pettini et al. (2000) on the z = 2.73 strongly-lensed object MS 1512-cB58. Another example is the strongly-lensed object CL 1358+62_G1 in the z = 0.33 cluster CL 1358+62 (Franx et al. 1997). This object is at z = 4.92. The large magnification (~ 10X) makes this a particularly important pathfinder for assessing the structure of galaxies at high redshifts. While it would be valuable to have other examples to enlarge the sample, this remains the best known at such a high redshift.
The cluster and its arc are shown in Figure 7, along with the arc prior to reconstruction, and after reconstruction into the source plane. The resolution in the reconstructed image is better than 20 milliarcsec, comparable to what one will get with adaptive optical systems on 8-10 m telescopes in the near-IR, though the sensitivity to low surface brightness extended structures will be much less in such ground-based data. This resolution corresponds to about 200 pc.
The object in the reconstructed image covers about 7 kpc, with several star forming knots interspersed throughout the galaxy. The knots dominate in that they contribute about 75% of the flux in the object, with the brightest knot alone contributing about half the flux. This corresponds to more than 1011 L in a single region with a half-light radius r1/2 ~ 20 mas or a FWHM of about 300-400 pc. The SFR implied by the observed UV flux, under the assumption of no extinction, is about 50 M yr-1. However, it is clear from the combined HST optical and Keck near-IR data (Soifer et al. 1998) that the source is reddened with E(B - V) ~ 0.3, indicating that the SFR needs to be increased significantly to > 102 M yr-1, possibly >> 102 M yr-1, depending on the reddening law and the dust distribution. The value of the IR data can be seen in Soifer et al. (1998) in the fits of the Bruzual and Charlot models, with Calzetti extinctions, to the measured HST WFPC2 F606W and F814W fluxes, and the Keck NIRC near-IR J, H, and K fluxes. Both the instantaneous burst and continuous star formation models need extinctions E(B - V) ~ 0.3 to match the observed fluxes, given the constraint that the burst must contain stars less than 107 yr in age so as to ensure the UV flux required for the strong Ly.
The knot mass can be estimated from the IR luminosity, assuming a Salpeter IMF, and is ~ 5 x 109 M. Given the scale size of the knot, this implies a velocity dispersion of ~ 200 km s-1. The size of the region, and such a velocity dispersion, is typical of many present day bulges.
Another interesting aspect of this object is that there are multiple star-forming knots within a more extended structure. This morphology is characteristic of the ``christmas tree'' model of star-formation in distant galaxies where different knots may ``turn on'' at different times, as was discussed by Lowenthal et al. (1997). A more detailed analysis of the implications for star formation at high redshift is to be found in Illingworth et al. (2000).