The number of measured redshifts per cluster is often
quite small (from 5 to 20 galaxies).
It is therefore difficult to have reliable estimates of the velocity
dispersion.
Using the NED database
we searched for other galaxies with measured redshifts
within a radius of one degree
from the center of each target cluster. This allowed us to add one more
velocity to A1035 (Batuski et al. 1991), three to A1413
(Stocke et al. 1991;
Allen et al. 1992), one to A1045 (Allen et al. 1992),
sixteen to A1831 (NED 1992 and Owen et al. 1995),
two to A2034 (Crawford
et al. 1995),
four to A195 (Giovanelli & Haynes 1993), four to A2245
(from Rhee & Katgert (1988) and NED 1992),
six to A 2457 (Quintana & Ramirez 1995;
Hewett et al. 1995;
Watson et al. 1991).
Velocity histograms are shown in Fig. 2 (click here).
Figure 2: The line-of-sight velocity histograms for our target clusters
Table 2: Estimates of the redshift and velocity dispersion of the clusters
We have applied to our cluster data set the updated ROSTAT package
(version 1.2) developed by Beers et al. (1990). We chose the most
appropriate estimators according to Beers et al. (1990),
taking into account
the number of available velocities 
. The results are listed
in Table 2 (click here). In the "tiny case'', when 
,
we have used the median 
and
the bi-weight 
as location estimators.
In the small-intermediate case (
), we used the median
and the bi-weight 
as estimators of the location, and
the canonical standard deviation 
, the bi-weight 
,
the gapper 
.
The mean estimator of the location has also been listed, although it is known
to be a poor estimator in the case of a non-Gaussian distribution.
Table 2 lists in Col. (1), the Abell cluster number, in Col. (2) and (3)
the right ascension and declination of the cluster center
(Abell, et al.
1989)
precessed at J2000.0, in Cols. 4, 5 and 6, bi-weight, median and mean
estimates of the redshift of the cluster from our analysis,
in Col. 7 the previous measurement of the cluster redshift from literature
(referenced in Col. 10) and the
number of galaxies used for this estimate, in Cols. 11, 12 and 13, the
standard deviation, bi-weight, and gap estimates of the velocity dispersion
within the cluster, and in Col. 13 the number of cluster members used
for the previous determinations.
The velocity analysis is indeed limited by the small number of galaxies with available
redshift by cluster. The various estimates of
velocity dispersions are listed in Table 2 so that the degree of self-consistency
can easily be checked. Determinations for clusters with less that 10 redshift measured
are listed between braces.
Note on individual clusters
Some clusters of our sub-sample show a complex velocity distribution.
For instance, it is clear from Fig. 2 (click here) that the velocity distribution of
A1035 is bimodal, with a group of
9 galaxies clustered around 
km s
with small dispersion
(
km s
)
and a group
of 10 galaxies belonging a to a more massive structure
around 
km s
(
km s
).
A1781 shows a complex structure that cannot be well sampled with
our data.
The field of this cluster was recently examined in detail by Ramella et al.
(1995; RGHT), as a loose group of galaxies at 
is found in that
region of the sky.
This results in a projection of the loose group on
the more distant Abell cluster (at 
from
Postman et al. 1992), of which galaxies 2, 3, 4, and 6 are
probably genuine members.
In Fig. 3 (click here)), the chain of four galaxies in the north corresponds to a
foreground structure, the redshifts of these galaxies being respectively,
from south to north, 11019 km s
(RGHT),
13699 km s
(our galaxy No.1), 11928 km s
(RGHT), and 11940 km s
(our galaxy No. 7).
A1831 shows two peaks in the velocity histogram,
at 
km s
and 
km s
, with
velocity dispersions of respectively 
km s
and 
km s
.
As mentioned before, we will not pursue the analysis of the three previous
clusters.
A1413 shows a main concentration of 9 galaxies at a mean velocity of
km s
(including the central cD galaxy) identified as the
main cluster. The estimate of its velocity dispersion (listed in Table 2) is
1460 km s
. This value is quite different from the estimate with
the bi-weighted method (
km s
). This can mean that the number
of galaxies is still not sufficient to use the bi-weighted technique.
Previous estimates of
the velocity dispersion in A1413 have to be taken with care until
more redshift measurements are available.
Finally, we note the presence of
a foreground group of 4 galaxies around v=30000 km s
.
In A2457, the spectrum of galaxy 7 (Fig. 3 (click here)) shows three very broad systems
of emission lines. The first complex includes a blend of 
and
[O III], with a FWHM of about 89 Å, the second one (with a FWHM
of 
includes H
and the [OIII] doublet in the tail of
H
, the third one (FWHM 
includes a
blend of 
and [N II]. Accurate fits with a gaussian could be done
when deblending the [OIII] doublet and the 
and [N II]
components, giving a mean velocity of 
km s
.
The cross-correlation technique for the absorption lines gives 
km s
. This object has been included for its
and [SII] lines in the catalog of the Palomar Transit Grism
Survey (Schneider et al. 1994), which gives a velocity
km s
. The difference with our estimate
is easily explained by the fact that in the case of the PTGS
and [N II] could not be deblended, given the
low-resolution of that survey; as a consequence,
the wavelength of the 
line -and the redshift- is overestimated.
The existence of very broad permitted emission lines and narrower forbidden ones suggests that this object is a Seyfert 1 galaxy.
Figure 3: The sky-subtracted spectrum of galaxy 7 in A2457.
This object shows emission lines characteristic of a Seyfert I galaxy