Old Galaxies at High Redshift

Abstract

The most passive galaxies at high redshift are unlikely to be identified by either narrow-band emission-line searches, or by Lyman limit searches (both techniques which have been highlighted at this meeting) simply because such selection methods rely on the presence of a strong ultraviolet component. Selection on the basis of extreme radio power has also proved to yield optically active objects with the majority of high-redshift objects studied to date displaying complex elongated optical/UV morphologies, relatively blue optical-ultraviolet continuum colours, and strong emission lines. These features, coupled with the failure to detect any spectral signatures of old stars at $z > 1$, has led to the suggestion that these galaxies are being observed close to or even during a general epoch of formation. However, we have recently demonstrated that radio selection at significantly fainter (mJy) flux densities can be used to identify apparently passively evolving elliptical galaxies at high redshift. Deep Keck spectra have now been obtained for two such objects yielding absorption line redshifts $z \simeq 1.5$; 53W091 at $z = 1.552$ (Dunlop et al. 1996) and most recently 53W069 at $z = 1.432$. The ultraviolet SEDs of these galaxies indicate minimum ages $> 3$ Gyr while, as stressed in this article, the strength of the reddenning-independent ultraviolet spectral breaks actually indicate a greater minimum age of 5 Gyr for both objects assuming solar metallicity. Since the spectra comprise the integrated light of each galaxy to radii greater than $r_e$, I argue that it is difficult to justify the adoption of significantly super-solar metallicity in interpreting these data. It thus seems hard to escape the conclusion that $Ω_0 < 1$ and that, irrespective of the adopted cosmology, at least some massive ellipitical galaxies were formed at high redshift ($z> 5$).