A comprehensive multicolor CCD photometric study has been made for a sample of forty-five low-redshift (0.04 <= z <= 0.18) rich clusters of galaxies, in order to investigate some global properties of cluster galaxies. Accurate photometric measurements in the Kron-Cousins R,,B and I filters were generated, to sufficiently faint levels that permitted the contribution of dwarf galaxies to be included. In thesis we have investigated the color magnitude relation (CMR) for early type cluster galaxies and the cluster galaxy luminosity function (LF).

Through this work we have addressed some unresolved problems regarding the CMR and the LF. We believe that the analysis presented here has revealed that the universality of CMR for early type cluster galaxies is sustainable. However, the universality of the LF is in general unsustainable, but we find evidence to support the LF universality in its weak form; that is, the property can be considered as universal when a large sample is considered, but allows variations at the individual level. In the case of the galaxy LF we have found that the existence cD galaxies is correlated with variations both at the bright and faint end of the LF. This gives evidence that the presence of cD galaxies is affecting the global properties of the cluster. Below we present this thesis main conclusions.

We found through the analysis of the color distribution in our cluster sample that the CMR for low redshift clusters is universal. The universality of the CMR within the redshift interval that we have sampled indicates that no significant star formation has occurred in early-type galaxies during the past 3 Gyr. The observed behavior of the CMR in the observer's frame is consistent with the results of a recent spectral synthesis model that explains the CMR in terms of metallicity effects and passive evolution (Kodama and Arimoto 1997). Because of the universality of the CMR, this model also suggests that early-type galaxies in clusters were formed more than 11 Gyr ago. Moreover, it was found that that the change of the CMR slope with redshift provides us with a powerful way to constrain galaxy evolution models. By estimating the slope of the CMR in the observer's frame and plotting it against redshift one can developed a test that is insensitive to the absolute calibration of the magnitudes. Applying this test at high redshift, one could constrain directly the epoch of stellar formation cessation in elliptical galaxies. The models assume that star formation is abruptly stopped when supernova-driven winds inject enough energy to the galaxy that causes the expulsion of the gas that was not locked up in stars (the wind age, t_gw, in the models of Kodama and Arimoto 1997). Moreover, the close match between our observations and the models supports the idea that cluster early type galaxies are coeval and are evolving passively. We have also explored the applications of the CMR as a redshift indicator and as tool for cluster detection.

We have detected clear variations at both the bright end and the faint end of the cluster galaxy LF. However, such variations are not erratic. A further investigation on the nature of this variations show that at the bright end, the M* for cD clusters is correlated with cluster richness; in the fashion that poor cD cluster have dimmer M*. We interpreted that this effect is due, partially, to the effects of dynamical friction. On the other hand, non-cD cluster seem to have a universal LF because the dispersion of their M* is consistent with the errors in the determination of the LF. We found < M* > = -22.32 mag and a dispersion [sigma] = 0.26 mag.

In our survey we have been able to include fainter galaxies than in previous studies. We have seen that variations at the faint-end of the luminosity function are more dramatic than at the bright end; as the variations of the faint-end slope for the contribution due to dwarf galaxies is in the range -1.9 <= [alpha] <= -0.90. Rich cD clusters seem to have effectively depleted a large number of their dwarf galaxies. There is a significant correlation that suggests that the brightness of the brightest cluster galaxy is inversely proportional to the dwarf to giant ratio (D/G). These variations have allowed us to single out a group of seven clusters that we have named flat-LF clusters. It turned out that flat-LF clusters are a homogeneous class of dynamically evolved cluster, as they can be characterized by their low Bautz-Morgan type, Rood-Sastry cD class, high richness, and high X-ray luminosity. These clusters have provided us with an important piece of evidence that motivated us to propose an alternative formation scenario for cD galaxies. In our heuristic model, we assumed that cD luminous halos are produced by the disruption of a large fraction of dwarf galaxies. With this model we can account for all the light in cD halos and a large fraction of the intra-cluster gas.