Old Massive Ellipticals and S0 in the Hubble Deep Field Vanish from View at z>1.3: Possible Solutions of the Enigma

We have investigated the properties of a bright K-band selected sample of early-type galaxies in the Hubble Deep Field, as representative of the field galaxy population. This dataset is unique as for the morphological information on faint high-z sources, and for complete photometric and spectroscopic coverage. The majority of bright early-type galaxies in this field are found at redshifts $z \leq 1.3$ to share common properties with those of high-z cluster samples, as for the age and mass of the dominant stellar population -- which are found to be as old as 3-5 Gyr and as massive as $10^{11} M_\odot$ already at $z\simeq 1$. There is no evidence of a steepening of the mass function from $z$=0 to $z$=1, as inferred by some authors from analyses of optically-selected samples and favoured by hierarchical clustering models forming most of the E/S0s at $z<1$. What distinguishes this sample is a remarkable absence of objects at $z>1.3$, which would be expected as clearly detectable above the flux limits, given the aged properties of the lower redshift counterparts. So, something hide them at high redshifts. Merging could be an explanation, but it would require that already formed and old stellar systems would assemble on short timescales of 1 Gyr or less. Only a negligible fraction of the galactic mass in young stars has to be added, a rather contrived situation from the dynamical point of view. An alternative interpretation could be that a dust-polluted ISM obscures the first 3 to 4 Gyr of the evolution of these massive systems, after which a galactic wind and gas consumption makes them transparent. The presence of dust would have relevant (and testable) implications for the evaluation of the global energetics from galaxy formation and for the visibility of the early phases by current and future infrared