Astron. Astrophys. 323, 337-348 (1997)

7. Conclusions

7.1. Es and S0s

From all points of view, the early-type galaxies in Cl 0939+4713 share the properties of their local counterpart galaxies, from which they are indistinguishable, at least within our photometric and statistical errors and for the considered, mainly photometric, galaxy properties. Please note that E+A galaxies in Cl 0939+4713 morphologically are spirals and not ellipticals (Belloni et al. 1995 and Sect. 4).

The only difference known to date between the Cl 0939+4713 early-type galaxies and the local respective counterparts is their rest-frame colors (Stanford, Eisenhardt & Dickinson 1995). However, Stanford, Eisenhardt & Dickinson (1995) use Bower's (1992 and private communication) colors of Coma galaxies as a reference sample, which show a systematic difference with respect to Recillas-Cruz et al.'s (1990) .

Since there are no blue early-type galaxies (at least in the studied samples), the ancestors of early-types will first become red, which takes Gyr (Charlot & Silk 1994), and then will take the early-type morphological appearance. This implies that morphological components characterizing spiral galaxies, such as spiral arms, HII regions and dust in large quantity, need more than 1 Gyr to be made undetectable by whatever mechanism is operating in clusters if S galaxies change morphological type.

7.2. Ss

Given their distance from the cluster center, the spirals in Cl 0939+4713 have unusual properties. Their frequency with respect to the other morphological types is higher than in present day clusters, their colors are similar to those of field spirals, much more than to those of spirals in the center of Coma-like clusters, which are red. Their spatial distribution and mean surface brightness are not typical of red or infalling (blue) spirals in Coma, since the Cl 0939+4713 spirals are not overbright for their magnitude and do not avoid the cluster center. This suggests that, in Cl 0939+4713, the blue spirals are just spirals with a normal star formation rate, as suggested by the optical colors of blue galaxies in many other distant clusters (Rakos & Schombert 1995; Rakos, Maindl & Schombert 1996).

Why are blue spirals with a normal star formation rate present in the core of distant clusters, but absent in the core of nearby Coma-like clusters?

Many authors, starting with Butcher & Oemler (1984), suggest that some kind of evolution is responsible for the disappearance of this blue population, which becomes red or undetected. However, the fading and/or destruction of the distant blue population, to make them undetected in present day clusters, seems unreasonable (Rakos & Schombert 1995), since most galaxies have to disappear. Changing morphological type seems even more unreasonable, because of the homogeneity of the photometric properties of early-type galaxies which implies that they are at least 11 Gyrs old.

In present day clusters, the lack of blue galaxies in the cluster core, the color distribution of spiral galaxies and many of their properties can be explained if spirals falling in clusters have a starbust due to the ram pressure in the hot gas (Bothun & Dressler 1986) that consumes the galaxy's gas reservoir. During the burst, these galaxies become bluer and brighter in mean surface brightness, then, shortly after the burst, they become as red as ellipticals (Charlot & Silk 1994), thus explaining the presence of red spirals in cluster cores. Furthermore, the existence of galaxies exhibiting spectral signatures consistent with the presence of intermediate age stellar populations (Couch & Sharples 1987; Lavery & Henry 1988, Dressler & Gunn 1992), and the previously listed photometric evidence for the blue starburst spirals in Coma, support this scenario. In a simplified view, during the galaxy's motion through the intracluster hot gas, the galaxy's gas is removed by ram pressure (Gunn & Gott 1972) and therefore its star formation is truncated, thus accounting for the lack of blue galaxies in the cluster core.

This scenario can also account for the presence of many blue galaxies in the core of distant clusters if the conditions for galaxy gas stripping are not satisfied, for example if the intergalactic gas density is lower than the threshold for this stripping.

X-ray observations show that the gas density distribution in Cl 0939+4713 (Schindler & Wambsganss 1996) is quite different from that in Coma, and the conditions for triggering the starburst are perhaps not realized. In the distant cluster 3C295, which instead possesses a low fraction of blue galaxies (Butcher & Oemler 1984), the gas density is high (Henry et al. 1979). Furthermore, the existence of nearby clusters with a high fraction of blue galaxies (e.g. Abell 1367) and of distant clusters with a low fraction of blue spirals (e.g. Cl0016+16), even in the sample studied by Butcher & Oemler (1984) suggests that our present interpretation of the excess of blue galaxies in distant clusters is a possible alternative to the evolutive interpretation of the Butcher-Oemler effect.

In order to test this working hypothesis, we are now directly measuring the cluster gas brightness (which is a measure of the gas density) in a uniform way for a sample of clusters, nearby as well as distant, for which an accurate and uniform measure of the blue fraction of galaxies is available.

7.3. Morphological segregation

Sandage (1990) suggested testing whether the segregation is the result of a biased galaxy formation plus a current infall by measuring an evolution of the morphological segregation. In such a case, ellipticals form in high density regions whereas spirals in lower ones, which more recently detached from the Hubble flow and fall in the cluster. Therefore distant (younger) clusters have to be richer in ellipticals in their core than present day clusters, since their outer shells, richer in late type galaxies, have not yet fallen in.

However we have shown, in the detailed study of three cluster (Coma, Perseus and Cl 0939+4713), that the galaxies are more strongly segregated with respect to a privileged direction than in density or clustercentric distance, making a simple verification of the above prediction more difficult. Furthermore, the core of Cl 0939+4713 is poorer in early-type galaxies than Coma, showing that a more complex scenario is needed to explain the evolution of the galaxy populations in clusters.

A viable alternative for explaining why many distant clusters have higher spiral fractions than nearby ones is to remove the assumption that the former (or at least the observed ones) are the ancestors of the latter. This is reasonable, since it seems that distant and nearby clusters differ in richness and gas content. Therefore, the galaxy populations of distant clusters do not have to evolve necessarily into those of nearby ones. This resolves the problems encountered by evolutionary scenarios in accounting for the differences in galaxy populations observed in clusters at different redshifts.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de