The ideal sites to study galaxy encounters are pairs and groups of galaxies. The pioneers in the studies of pairs of galaxies are Holmberg (1937, 1958) and Page (1952, 1961) and more recently Karachentsev (1972). Major effort has been focused into assembly complete samples of isolated pairs of galaxies for the Southern Hemisphere (Reduzzi & Rampazzo 1995; Soares et al. 1995; see Sulentic 1992 for a review on the subject). Several studies have analyzed the effect of interaction in binary galaxies, e.g. Rampazzo & Sulentic (1992); Rampazzo et al. (1995); de Mello et al. (1996); Marquez & Moles 1996; Reduzzi & Rampazzo (1996), and de Souza et al. (1997).
One simple approach in studying tidal effects is to analyze galaxy morphology in binary galaxies. A standard morphological classification of binaries is based on their components morphology. For instance, early-type binaries, usually called EE, are formed by two early-type galaxies and late-type binaries (SS) by two spiral galaxies. Mixed morphology binaries, ES, are formed by an early and a late-type galaxy. The observed fraction of the three types is 0.60 for SS, 0.30 for for ES and 0.10 for EE (Sulentic 1992). The existence of many isolated mixed binaries raises problems for models of galaxy formation that envision local environment as a major determinant of morphology. However, there are a few evidences, like morphological distortion in the late-type galaxy and young stellar population in the early-type showing that mixed binaries are physically interacting (de Mello et al. 1996; Rampazzo & Sulentic 1992). Therefore, if mergers of spiral galaxies are the mechanism for the formation of ellipticals then binaries containing an elliptical would be the product of interaction between at least 3 galaxies. To understand whether galaxy transformation is a common phenomenon in galaxy evolution more studies of galaxy morphologies in interacting systems are needed. In this work we present the results of such analysis for a sample of binary galaxies.
One of the most important goals of this study is to build a photometric and spectroscopic database in order to study interacting galaxy evolution. For this, the bright and faint pairs should constitute a carefully matched sample with projected separations such that the objects would be expected to merge in much less than a Hubble time.
The faint (19<mR<22) sample was selected from automatic compiled catalogs of existing (de Mello et al. 1997a,b) CTIO 4 m images at the equator; the bright sample (16<mV<18) was selected from the catalog of isolated binary galaxies (CPG) compiled by Karachentsev (1972), which is the best available catalog that meets the local control sample requirements.
Pairs of faint galaxies (separation 
) were
chosen such that 
.
Here 
is the minimum separation at which
pairs can be reliably separated (nominally 2'');
corresponds to some physical separation, 
,
chosen so that: (i) physical pairs are doomed to merge
in < 109 yr (on the basis of both empirical studies and
conventional dynamics); and (ii) most pairs in the
faint sample are physically associated. These conditions are
satisfied by 
kpc, which corresponds to
at z = 0.35.
What fraction of the faint pairs is expected to be physical
pairs and merge? Infante et al. (1996) have calculated the
angular correlation function, 
for pairs
at mR < 21.5 and 
. It follows that
of faint pairs will be physically associated.
Furthermore, a cut in 
(say 
km
s-1) should isolate most of the physical pairs.
Following Carlberg et al. (1994) the fraction of galaxies
that will merge is 
. Thus, it is expected that
of galaxies in pairs will eventually merge.
The two samples should be affected by identical
selection effects. The CPG contains 603 binaries brighter
than 
, is the most complete catalog of binaries
and has complete redshift information. In order to match the
properties of z=0.35 pairs we looked at all Karachentsev
pairs with characteristics similar to what we expected for
pairs at high z. We have selected all pairs with projected
linear separation 
kpc
and radial
velocity difference 
.