Our field of survey corresponds to the field of the 20-inch du Pont plate
ID: CD14, taken on Eastman blue emulsion 103a-O, baked in forming gas and
exposed for one hour on the 100-inch du Pont reflector at Las Campanas.
The plate includes the central part of the Centaurus cluster covering an
area of 
bound by 
RA(1950) 
and 
Dec(1950) 
.
We carried out a visual search for all non-stellar objects with apparent
diameters larger than 10''=0.9mm on the high-quality film copy of the
Centaurus du Pont plate (
mm on a Schmidt plate). Using a magnifier
more than 1100 galaxies have been detected down to this diameter limit.
In order to find very low-surface-brightness galaxies and to rule out plate
flaws from our sample, we additionally employed film copies of the deeper,
but low-resolution Schmidt plates SRC 322J and 323J (GG395-IIIa-J, 70
minutes exposure time) from the European Southern Observatory/Science
Research Council (ESO/SRC) Southern Sky Survey.
The criteria we applied to subdivide the sample galaxies into Centaurus cluster members and other galaxies (background) are those worked out by, and described in, Binggeli et al. (1985). This means we had to take into account the surface brightness, the resolution of the galaxy in stellar content, and - if available - the radial velocity. Ferguson (1989) already applied successfully these criteria in his survey of the Fornax cluster. Meanwhile it was shown for the Virgo cluster (Binggeli et al. 1993; Drinkwater et al. 1996) that such morphological guesses are in excellent agreement with redshift data.
We estimate the likelihood of each "cluster member'' to be in the cluster by assigning a probability of membership in three steps of 50%, 75% or 100% depending on how strongly the galaxy fulfils the above mentioned criteria. Thus the initial survey sample ended in 4 different classes whose numbers are listed in Table1 (click here).
Table 1: Cluster membership code with numbers
The process of cluster/background subdivision and cluster membership assignment was
additionally facilitated by the fact that the surveyed region of the Centaurus cluster
is nearly free of field or nearby background cluster galaxies which could strongly
contaminate the Centaurus cluster sample (or even completely impede the survey). Besides
rather few foreground objects, a first galaxy concentration is located
behind Centaurus at a redshift of 
kms
(e.g.
Dickens
et al. 1986; Stein 1994) or roughly 3.5 times the distance to the
surveyed cluster. This fact makes us quite confident that our selection
of the 296 cluster members is stable against background contamination. All
uncertainties are expected to be covered by the given membership probabilities.
Galaxies of the final cluster sample were morphologically classified by the visual comparison of their images on the du Pont and Schmidt films with the reference galaxies of the extended Hubble classification system. These reference galaxies are illustrated in the Hubble Atlas (Sandage 1961), the Revised Shapley-Ames Catalogue (Sandage & Tammann 1981), the Atlas of Virgo Cluster Spiral Galaxies (Sandage et al. 1985a), and in the Atlas of Virgo Cluster Dwarf Galaxies (Sandage & Binggeli 1984). For the particular distinction between dwarf ellipticals (dE) and dwarf S0 (dS0) galaxies the explanation in Binggeli & Cameron (1991) was very helpful. Table2 (click here) lists the frequency of the main Hubble types in our cluster catalogue as non-weighted (Cols.2 and 3) and membership-weighted numbers (Cols.4 and 5).
In general, the detection of a galaxy as well as the accuracy of a morphological classification strongly depend on the Hubble type, the inclination and the angular size. The most crucial points are summarized here:
Low luminosity elliptical galaxies (exemplified by M32) are very compact and in
appearance comparable to the seeing disk of a star. Only the missing diffraction cross
typical for point sources gives evidence for a nonstellar object. The Centaurus du Pont plate
has a magnitude limit for crossed stars of about 
or 
in
absolute B-magnitude. On the other hand, a M32-type galaxy typically has
which explains why the detection of such objects in our survey is
a matter of chance. Nevertheless, our galaxy catalogue contains one strong candidate:
object CCC70 is a low luminosity elliptical with 
and a characteristic
high mean surface brightness. We will discuss its properties in PaperIV in more details
(Jerjen & Binggeli 1997).
Early-type dwarfs (dE&dS0) exhibit no active star formation and thus are detected
primarily based on their star density (= surface brightness). For this
galaxy type the surface brightness at a specific radius is correlated with the
total magnitude of the galaxy (Binggeli & Cameron 1991; PaperIV). As we demonstrate
in PaperIV the survey limit is 9'' in diameter at the isophote of 26.5Barcsec
,
which corresponds to a completeness limit of about 
for the dE&dS0 galaxies.
Table 2: Hubble type frequencies as non-weighted and membership-weighted numbers
In contrast to the early-type dwarfs, Im's exhibit a wide range in surface brightness at a given absolute magnitude due to the sensitivity to star-formation activity. This naturally biases detection against inactive Im's which should be taken into account if one works with Im samples. On the other hand, BCD's (Blue Compact Dwarfs) are compact extragalactic HII regions. Their bright irregular shape gives a high detection probability. But the appearance of a cluster BCD can be very similar to a star formation region of a more distant Im galaxy or a field BCD which makes the definite assignment of a BCD difficult in most cases. Nevertheless, only recently Drinkwater et al. (1996) confirmed with new redshift data that 97% of a morphological ground based sample of "background'' BCD's from the Virgo Cluster Catalogue (Binggeli et al.1985) are indeed not members of the Virgo cluster.
Figure 1: The survey area covered by the du Pont plate is shown. The
cluster members have been divided into five different symbols
(
, 
, 
, 
, 
) according to the
five main Hubble types E, S0, Sp, Im
BCD, and dE
dS0
Figure 2: Averaged density profiles of the northern and eastern
sensitometer marks of the SRC 323J Schmidt plate
Figure1 (click here) gives an overview of the surveyed cluster region with the cluster members subdivided into the five main Hubble types E, S0, Sp, Im&BCD, and dE&dS0. All cluster members, with their membership probability and detailed Hubble type, are listed in the Centaurus Cluster Catalogue (hereafter CCC).