Previous catalogues of Virgo Cluster galaxies have been limited in several respects. First, the vast majority of their galaxy magnitudes were estimated by eye without independent zero-point calibrations, secondly, the galaxy samples themselves were selected by eye and thirdly, they were confined to a single pass band. The VPC does not suffer from any of these limitations. Also, the VPC does not discriminate against background galaxies (unlike the VCC which excludes non-Zwicky objects that were deemed to be in the background).
Figure 9: Map of the areas of the sky surveyed by the VPC (vertically shaded
area) and VCC (heavily shaded area VPC survey area)
The VPC however, covers a region of sky smaller in angular extent than those covered by de Vaucouleurs & Pence (1979) or by the VCC. The VPC survey area, which is centred on Cluster A, is compared with the VCC's in Fig. 9 (click here).
Most VPC-galaxy magnitudes have been derived by the numerical integration
of segmented plate-scan data and have been calibrated with several CCD frames
and/or several hundred aperture-photometry measurements. We therefore expect
these magnitudes to be a significant improvement on those found in previous
catalogues. For the highest surface-brightness
VPC galaxies however, the plate-scan data suffered from
saturation effects, and alternative measurements had to be sought from the
literature. As can be seen from Fig. 10 (click here), saturation was a problem
in the band for most galaxies brighter than
, though
fewer galaxies were saturated in the U band. For a very small minority of
saturated galaxies, no previous measurements could be found in the literature
and eye estimates were necessary.
Figure 10: Frequency distribution of VPC galaxies as a function of .
The heavily shaded areas represent objects saturated in
There has been much discussion concerning the completeness of galaxy surveys in the literature, and particularly of those conducted by eye; see e.g. Phillipps et al. (1988). As the VPC galaxy sample was selected primarily by means of an automated process, it was possible to quantify the observational limits to the sample with reasonable precision. These limits were:
(1) for the whole survey area
and
(2) profile-slope parameter (see Sect. 3.6 (click here)) .
However, Plate J4882 was found to be significantly desensitised over Fields C
and D (by mag) and the magnitudes for those fields had
therefore to be based on J9229 alone. Mean magnitudes (based on both the J4882
and J9229 values) were nevertheless adopted for Fields A and B. These
considerations were found to lead to the VPC-sample completeness limits
of:
(1a) for Fields A and B,
or
(1b) for Fields C and D,
and
(2) profile-slope parameter ,
which are derived at the end of this subsection.
Figure 11: Frequency distribution of VPC galaxies unsaturated in , as a
function of peak surface brightness; the heavily shaded areas representing
those galaxies brighter than the completeness limits
(
for Fields A and B;
for Fields C and D)
Should populations of ultra-low-surface-brightness galaxies exist (of peak
surface brightnesses fainter than 25.0 in
) they would
therefore remain uncatalogued regardless of their apparent magnitude because
isophotal magnitudes rather than total magnitudes
were used to select the galaxy sample.
Figure 11 (click here)
would however suggest that galaxies with peak surface brightnesses fainter than
do not normally have flat enough surface-brightness profiles to qualify
as 18th-magnitude (or brighter) objects, unless they possess very extensive
haloes of surface brightness slightly fainter than 25.0
.
Likewise, discrimination against ultra-low-surface-brightness galaxies on
account of using isophotal magnitudes cannot be a problem unless such galaxies
have very flat profiles and are quite large in angular extent (e.g. an object
of mean surface brightness =26.00
would have to subtend 28
arcsec in diameter in order to be a 19.0 magnitude object or 35 arcsec
in diameter in order to be an 18.5 magnitude object). Although giant
galaxies of very low surface brightness are known to exist (e.g. Malin 1,
which escaped detection in the VPC) it is unlikely that such objects could
be prolific enough to be a major hazard.
Figure 12: Frequency distribution of VPC galaxies as a function of the
profile-slope parameter (see Sect. 3.6 (click here)); the heavily shaded areas
representing those galaxies brighter than the completeness limits of
(
for Fields A and B;
for Fields C and D)
It would appear from Fig. 12 (click here) that brighter than the completeness
limit of galaxy loss due to the limit
is probably not
significant. However, from the same figure it can be seen that [due to this
criterion alone] the number of galaxies lost from the sample within the range
is probably quite significant. In other words,
for
, the galaxy profiles tend to be significantly more
starlike than for
. Unfortunately, it is rather difficult to
extrapolate the frequency distribution of galaxies as a function of
beyond the
cut off with great confidence, even though the
indications are that the frequency is already in sharp decline as
increases towards the cut off. An order of magnitude estimate based in
Fig. 12 (click here) would suggest that somewhere in the region of ten or
possibly twenty objects of
might have been
excluded by the limit in
alone.
As for stellar contamination, this is unlikely to be a serious problem as
objects for which exceeded -0.065 were excluded. Only in a couple of
cases was there any uncertainty as to whether an object was a star or a galaxy,
as all galaxy candidates were visually inspected. Overlapping stellar images
were occasionally difficult to distinguish from elliptical galaxies, but it is
extremely unlikely that stellar contamination could account for even
0.25
of the objects catalogued (i.e. about three objects in total).
From Fig. 13 (click here) it is evident that the VPC galaxy sample is complete to
for Fields A and B (assuming
). It should be noted that the
slight asymmetry in the distribution of data points about
is due to the sensitivity of J4882
being less consistent than that of J9229. As a consequence, the most wayward
data-points tend to correspond to images that yielded higher magnitudes on
J4882 than on J9229. For Fields C and D, there is almost certainly completeness
at
, but galaxy loss probably only becomes significant beyond
, as can be seen from Fig. 14 (click here).
Figure 13: The disparity between magnitudes generated from different
plates as a function of mean magnitude [as quoted in the VPC] for Fields A and
B combined. The completeness limit (dashed line) and the observational
selection criterion,
, (solid diagonal line) are
shown for reference
Figure 14: The completeness limit
(dashed line) to the galaxy sample in the cases of Fields C and
D. Note that due to pronounced de-sensitisation of Plate J4882's emulsion over
Fields C and D, a mean zero-point shift of 0.31 mag was observed between
those magnitudes from Fields C and D of Plate J9229 [which were included in the
VPC] and those due to J4882 for the same fields [which were discarded]
Equal-area and total () and (
) colours have been computed for most
VPC galaxies; the main exceptions generally being those
whose images were saturated in one or both relevant pass bands, and/or those that
appeared heavily merged with adjacent images (whether galaxies or stars).
For the equal-area colours,
the areas were generally defined by the
or the
isophote.
The total colours were based on the differences between
extrapolated
,
and
values, and as a whole are probably less susceptible
to systematic effects even though they exhibit a considerably larger scatter.
However, for any individual galaxy, the equal-area colours are generally more reliable
than the total ones (see Sects. 7.2 (click here) and 7.3 (click here)).
It was also possible to
estimate very approximate total (
) colours for many of the saturated
objects by means of Eq. (5 (click here)) (which is an approximation based on
Eq. (1)) from values of
and
whenever listed in the
RC3. Unfortunately though, published (
) colours could not be found
for those objects whose images were saturated.
Figure 15: Frequency distribution of VPC galaxies still lacking published velocities
(unshaded areas), those measured for the first time by
Drinkwater
et al. (1996) (vertically shaded areas) and those measured previously (unshaded);
all as a function of
Only a minority of galaxies fainter than have published
radial velocities, but the situation has improved since
Drinkwater et al.'s
(1996) study. The present situation is depicted in Fig. 15 (click here).
Morphological types are essentially complete to , but
sporadic beyond this limit as shown by Fig. 16 (click here). As the Du Pont
plates of Binggeli et al. (1985) had considerably larger plate scales than
did our UKST plates, VCC types have been
adopted for almost all of the galaxies in common between the VPC and the VCC
galaxy samples. Those galaxies which were brighter than
yet
not included in the VCC (generally because they were deemed to be background
objects) were classified by visual inspection of a copy plate of UKST plate
J9229. Due to the small scale of this copy plate and the problem of image
saturation in many cases, no attempt has been made to provide detailed
morphological classifications (e.g. spiral subclasses or luminosity classes)
in the VPC.
Figure 16: Frequency distribution of VPC galaxies that have been typed (heavily
shaded areas) and untyped (unshaded areas) as a function of
. Note that any unshaded area brightward of
represents galaxies which defied typing
Unambiguous identification of the vast majority of galaxies within the central
Virgo field (background objects included) brighter than
should now be possible, as the use of a measuring machine has enabled improved
positions to be computed. Only for a subset of objects saturated in
, was
it necessary to extract the coordinates from the RC3 or occasionally from the
VCC, though these objects are generally so extended that the positional
accuracy required to identify them unambiguously is not that high. The VPC also
contains orientation and ellipticity information.