3.3. The increasing fraction of irregular and multiple component systems

The 1980s also saw the first deep redshift surveys which provided quantitative evidence for galaxy evolution brighter than the break points in Fig 1. Progress followed the new technology of multi-object spectrographs, from plug-plate fibre systems (Hill et al. 1980), robot positioners (Parry & Sharples 1988) to multislit spectrographs such as the Cryogenic Camera (Butcher 1982), LDSS-1/2 (Colless et al. 1993, Allington-Smith et al. 1994) and MOSIS (LeFevre et al. 1994). Such surveys revealed a rapidly declining population of star-forming galaxies over 0 < z < 1 which seem to be responsible for the bright count excess. What are these rapidly-evolving galaxies?

Figure 3. A mosaic of HST images for galaxies with irregular morphology selected from the sample of Brinchmann et al. (1997). Labels refer to rest-frame [O II] equivalent widths and redshifts.

The Medium Deep Survey (Griffiths et al. 1994, Windhorst et al. 1995) analyzed parallel WFPC-2 data for 30 fields and presented the morphological mixture as a function of apparent magnitude. Although normal galaxies are seen in numbers consistent with approximately constant co-moving space densities, these authors were struck by the rising fraction of faint system with irregular morphology; many are suggestive of merging systems. Glazebrook et al. (1995b) and Driver et al. (1995) argued that rapid evolution was almost exclusively occurring in the 'irregular/peculiar/merger' category - admittedly rather a catch-all for non-regular systems whose physical nature remains unclear. Abraham et al. (1996a, b) introduced a more quantitative basis for faint galaxy morphologies and discussed how to allow for redshift-dependent distortions in extended the analysis, using the HDF, to I = 25. They claimed few of the faintest galaxies could be shoe-horned into the classical Hubble sequence.

The linking of redshifts and HST morphologies has been slow to emerge. An international group, based on the original CFRS and LDSS teams, have now amassed over 300 faint galaxies to I < 22 or B < 24 for which extended WFPC-2 imaging and spectroscopic redshifts are available. This was not feasible to construct with the Medium Deep Survey because of the mismatch between the small WFPC-2 field and the larger ones of the ground-based multiobject spectrographs. Brinchmann et al. (1997) examined in some detail the possible effects of redshift on the perceived morphology. They found a rapid rise with redshift in the fraction of galaxies with irregular morphologies which they claim cannot be due to k-correction or surface brightness effects. This trend represents a major component of the evolution seen over 0 < z < 1. However, it is clear the evolution in blue luminosity density need not come entirely from this population (Lilly, this volume) and, moreover, it is not obvious that the declining fraction of galaxies with irregular sources is matched by a compensating growth of disks or spheroidal galaxies over the same period as might be expected in simple hierarchical pictures (Baugh et al. 1998).