G. Kauffmann et al., DETECTION OF STRONG EVOLUTION IN THE POPULATION OF EARLY-TYPE GALAXIES, Monthly Notices of the Royal Astronomical Society, 283(4), 1996, pp. 117-122
The standard picture holds that giant elliptical galaxies formed in a
single burst at high redshift. Ageing of their stellar populations sub
sequently caused them to fade and become redder. The Canada-France Red
shift Survey provides a sample of about 125 galaxies with the luminosi
ties and colours of passively evolving early-type galaxies and with 0.
1 < z < 1. This sample is inconsistent with the standard evolutionary
picture for elliptical galaxies with better than 99.9 per cent confide
nce. The standard Schmidt test gives [V/V-max] = 0.398 when restricted
to objects with no detected star formation, and [V/V-max] = 0.410 whe
n objects with emission lines are also included. A smaller sample of e
arly-type galaxies selected from the Hawaii Deep Survey gives equally
significant results. With increasing redshift, a larger and larger fra
ction of the nearby elliptical and SO population must drop out of the
sample, either because the galaxies are no longer single units or beca
use star formation alters their colours. If the remaining fraction is
modelled as F = (1 + z)(-gamma), the data imply that gamma = 1.5 +/- 0
.4. At z = 1, only about one-third of nearby bright E and S0 galaxies
were already assembled and had the colours of old passively evolving s
ystems. We discuss the sensitivity of these results to the incompleten
ess corrections and stellar population models that we have adopted. We
conclude that neither the corrections nor the models are uncertain en
ough to reconcile the observations with the standard picture. Hierarch
ical galaxy formation models suggest that both merging and recent star
formation play a role in the strong evolution that we have detected.