The Fornax Cluster Spectroscopic Survey: A Sample of Confirmed Cluster Dwarfs
Extracted source text ยท arXiv astro-ph/0206510
% mnsample.tex
%
% v1.2 released 5th September 1994 (M. Reed)
% v1.1 released 18th July 1994
% v1.0 released 28th January 1994
\documentstyle[epsfig]{mn}
% If your system has the AMS fonts version 2.0 installed, MN.sty can be
% made to use them by uncommenting the line: %\AMStwofontstrue
%
% By doing this, you will be able to obtain upright Greek characters.
% e.g. \umu, \upi etc. See the section on "Upright Greek characters" in
% this guide for further information.
%
% If you are using AMS 2.0 fonts, bold math letters/symbols are available
% at a larger range of sizes for NFSS release 1 and 2 (using \boldmath or
% preferably \bmath).
\newif\ifAMStwofonts
%\AMStwofontstrue
%%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\ifoldfss
%
\newcommand{\rmn}[1] {{\rm #1}}
\newcommand{\itl}[1] {{\it #1}}
\newcommand{\bld}[1] {{\bf #1}}
%
\ifCUPmtlplainloaded \else
\NewTextAlphabet{textbfit} {cmbxti10} {}
\NewTextAlphabet{textbfss} {cmssbx10} {}
\NewMathAlphabet{mathbfit} {cmbxti10} {} % for math mode
\NewMathAlphabet{mathbfss} {cmssbx10} {} % " " "
\fi
%
\ifAMStwofonts
%
\ifCUPmtlplainloaded \else
\NewSymbolFont{upmath} {eurm10}
\NewSymbolFont{AMSa} {msam10}
\NewMathSymbol{\upi} {0}{upmath}{19}
\NewMathSymbol{\umu} {0}{upmath}{16}
\NewMathSymbol{\upartial}{0}{upmath}{40}
\NewMathSymbol{\leqslant}{3}{AMSa}{36}
\NewMathSymbol{\geqslant}{3}{AMSa}{3E}
\let\oldle=\le \let\oldleq=\leq
\let\oldge=\ge \let\oldgeq=\geq
\let\leq=\leqslant \let\le=\leqslant
\let\geq=\geqslant \let\ge=\geqslant
\fi
%
\fi
%
\fi % End of OFSS
\ifnfssone
%
\newmathalphabet{\mathit}
\addtoversion{normal}{\mathit}{cmr}{m}{it}
\addtoversion{bold}{\mathit}{cmr}{bx}{it}
%
\newcommand{\rmn}[1] {\mathrm{#1}}
\newcommand{\itl}[1] {\mathit{#1}}
\newcommand{\bld}[1] {\mathbf{#1}}
%
\def\textbfit{\protect\txtbfit}
\def\textbfss{\protect\txtbfss}
\long\def\txtbfit#1{{\fontfamily{cmr}\fontseries{bx}\fontshape{it}%
\selectfont #1}}
\long\def\txtbfss#1{{\fontfamily{cmss}\fontseries{bx}\fontshape{n}%
\selectfont #1}}
%
\newmathalphabet{\mathbfit} % math mode version of \textbfit{..}
\addtoversion{normal}{\mathbfit}{cmr}{bx}{it}
\addtoversion{bold}{\mathbfit}{cmr}{bx}{it}
%
\newmathalphabet{\mathbfss} % math mode version of \textbfss{..}
\addtoversion{normal}{\mathbfss}{cmss}{bx}{n}
\addtoversion{bold}{\mathbfss}{cmss}{bx}{n}
%
\ifAMStwofonts
%
\ifCUPmtlplainloaded \else
%
% Make NFSS 1 use the extra sizes available for bold math italic and
% bold math symbol. These definitions may already be loaded if your
% NFSS format was built with fontdef.max.
%
\UseAMStwoboldmath
%
\makeatletter
\new@mathgroup\upmath@group
\define@mathgroup\mv@normal\upmath@group{eur}{m}{n}
\define@mathgroup\mv@bold\upmath@group{eur}{b}{n}
\edef\UPM{\hexnumber\upmath@group}
%
\new@mathgroup\amsa@group
\define@mathgroup\mv@normal\amsa@group{msa}{m}{n}
\define@mathgroup\mv@bold\amsa@group{msa}{m}{n}
\edef\AMSa{\hexnumber\amsa@group}
\makeatother
%
\mathchardef\upi="0\UPM19
\mathchardef\umu="0\UPM16
\mathchardef\upartial="0\UPM40
\mathchardef\leqslant="3\AMSa36
\mathchardef\geqslant="3\AMSa3E
%
\let\oldle=\le \let\oldleq=\leq
\let\oldge=\ge \let\oldgeq=\geq
\let\leq=\leqslant \let\le=\leqslant
\let\geq=\geqslant \let\ge=\geqslant
%
\fi
\fi
%
\fi % End of NFSS release 1
\ifnfsstwo
%
\newcommand{\rmn}[1] {\mathrm{#1}}
\newcommand{\itl}[1] {\mathit{#1}}
\newcommand{\bld}[1] {\mathbf{#1}}
%
\def\textbfit{\protect\txtbfit}
\def\textbfss{\protect\txtbfss}
\long\def\txtbfit#1{{\fontfamily{cmr}\fontseries{bx}\fontshape{it}%
\selectfont #1}}
\long\def\txtbfss#1{{\fontfamily{cmss}\fontseries{bx}\fontshape{n}%
\selectfont #1}}
%
\DeclareMathAlphabet{\mathbfit}{OT1}{cmr}{bx}{it}
\SetMathAlphabet\mathbfit{bold}{OT1}{cmr}{bx}{it}
\DeclareMathAlphabet{\mathbfss}{OT1}{cmss}{bx}{n}
\SetMathAlphabet\mathbfss{bold}{OT1}{cmss}{bx}{n}
%
\ifAMStwofonts
%
\ifCUPmtlplainloaded \else
\DeclareSymbolFont{UPM}{U}{eur}{m}{n}
\SetSymbolFont{UPM}{bold}{U}{eur}{b}{n}
\DeclareSymbolFont{AMSa}{U}{msa}{m}{n}
\DeclareMathSymbol{\upi}{0}{UPM}{"19}
\DeclareMathSymbol{\umu}{0}{UPM}{"16}
\DeclareMathSymbol{\upartial}{0}{UPM}{"40}
\DeclareMathSymbol{\leqslant}{3}{AMSa}{"36}
\DeclareMathSymbol{\geqslant}{3}{AMSa}{"3E}
%
\let\oldle=\le \let\oldleq=\leq
\let\oldge=\ge \let\oldgeq=\geq
\let\leq=\leqslant \let\le=\leqslant
\let\geq=\geqslant \let\ge=\geqslant
%
\fi
\fi
%
\fi % End of NFSS release 2
\ifCUPmtlplainloaded \else
\ifAMStwofonts \else % If no AMS fonts
\def\upi{\pi}
\def\umu{\mu}
\def\upartial{\partial}
\fi
\fi
\title[Fornax Spectroscopic Survey]
{The Fornax Cluster Spectroscopic Survey: A Sample
of Confirmed Cluster Dwarfs}
\author[J.H. Deady et al.]
{J.H. Deady$^1$,
P.J. Boyce$^1$,
S. Phillipps$^1$,
M.J. Drinkwater$^2$,
A. Karick$^2$,\cr
J.B. Jones$^3$,
M.D. Gregg$^{4,5}$ and
R.M. Smith$^6$\\
$^1$Astrophysics Group, Department of Physics, University of Bristol, Bristol BS8 1TL\\
$^2$School of Physics, University of Melbourne, Victoria 3010, Australia\\
$^3$School of Physics and Astronomy, University of Nottingham, University Park,
Nottingham NG7 2RD\\
$^4$Department of Physics, University of California, Davis, CA 95616, USA\\
$^5$ Institute for Geophysics and Planetary Physics, Lawrence Livermore
National Laboratory, L-413 Livermore, CA 94550, USA\\
$^6$Department of Physics and Astronomy, University of Wales Cardiff,
PO Box 913, Cardiff CF2 3YB}
%\date{To appear in Mon.Not.R.Astron.Soc.}
\date{DRAFT version 0.1, 2002 February 11}
\pagerange{\pageref{firstpage}--\pageref{lastpage}}
\pubyear{2002}
\begin{document}
\maketitle
\label{firstpage}
\begin{abstract}
The Fornax Cluster Spectroscopic Survey (FCSS) project utilises the Two-degree
field (2dF) multi-object
spectrograph on the Anglo-Australian Telescope. Its aim is to obtain spectra
for a complete sample of all 14000 objects with $16.5\leq b_j \leq19.7$
{\em irrespective of their morphology}
in a 12~deg$^{2}$ area centred on the Fornax Cluster.
A sample of 24 Fornax Cluster members has been identified from
the first 2dF field (3.1~deg$^{2}$ in area) to be completed.
This is the first complete sample of cluster objects of known distance
with well defined selection limits.
19 of the galaxies (with --15.8$<M_{B}<$--12.7)
appear to be conventional dwarf elliptical (dE)
or dwarf S0 (dS0) galaxies. The other 5 objects
(with --13.6$<M_{B}<$--11.3) are those galaxies
which we described in Drinkwater et al. (2000b) and labelled
`Ultra-Compact Dwarfs' (UCDs). A major result
is that the conventional dwarfs all
have scale-sizes $\alpha$$\ga$3~arcsec ($\simeq$300~pc).
This apparent minimum scale size implies an equivalent
minimum luminosity for a dwarf of a given surface
brightness. This produces a limit on their distribution
in the magnitude-surface brightness plane, such that
we do not observe dEs with high surface-brightnesses but faint
absolute magnitudes. Above this
observed minimum scale-size of 3~arcsec, the dEs and dS0s fill the whole area
of the magnitude--surface brightness plane sampled by our
selection limits. The observed correlation between
magnitude and surface brightness noted by several recent studies
of brighter galaxies is not seen with our fainter cluster sample.
A comparison of our results with the Fornax Cluster Catalog (FCC) of Ferguson
illustrates that attempts to determine cluster membership solely
on the basis of observed morphology can produce significant errors.
The FCC identified 17 of the
24 FCSS sample (i.e 71 per cent) as being `cluster' members,
in particular missing all 5 of the UCDs.
The FCC also suffers from significant contamination: within the FCSS's field
and selection limits, 23 per cent of those objects
described as cluster members by the FCC
are shown by the FCSS to be background objects.
\end{abstract}
\begin{keywords}
galaxies: luminosity function --
galaxies: statistics --
galaxies: clusters: individual: Fornax --
surveys --
techniques: spectroscopic
\end{keywords}
\section{Introduction}
It is generally perceived that dwarf galaxies are primarily of low
surface brightness, and that lower luminosity
galaxies have fainter surface brightnesses. In particular, low surface brightness
galaxies (LSBGs) seen towards a cluster are conventionally assumed to
be members, while apparently faint, but high surface brightness
galaxies (HSBGs) are presumed to be luminous objects in the background
(e.g. Sandage, Binggeli \& Tammann 1985).
Indeed, it has frequently been asserted that dwarf galaxies
(in particular) exhibit a strong surface brightness -- luminosity
relation (e.g. Sandage, Binggeli \& Tammann 1985; Bingelli, Sandage \&
Tammann 1985; Kormendy 1985; Ferguson \& Sandage 1988).
Binggeli (1994) has particularly persuasively
argued this point with reference to the Local Group. Nevertheless, other
authors have noted that a good deal of the evidence for the correlation
relies on the `eyeball' selection of dwarf galaxies `similar' to ones
already known, that any correlation breaks down at the low luminosity end,
and/or that there may be strong selection effects limiting the detected
galaxies' photometric parameters (e.g. Phillipps, Davies \& Disney 1988;
Irwin et al. 1990).
The failure of the relation,
i.e. the existence of large background LSBGs (such as the
serendipitously discovered Malin 1; Bothun et al.\ 1987), or of a
population of high surface brightness (compact) dwarfs in a cluster
(Drinkwater \& Gregg 1998), or indeed of a population of galaxies so compact
as to masquerade as stars and, hence, be missed
altogether from galaxy samples (e.g. Arp 1965; Drinkwater et al. 2000b; Phillipps et al.
2001) could have a dramatic effect on our
perception of the galaxy population as a whole (see Impey \& Bothun 1997).
The primary factor preventing a resolution of the controversy has been
the non-existence of suitable samples of confirmed dwarfs of known distance
(and hence luminosity) with well defined selection limits. Generally,
dwarf galaxy
samples have been either subjectively (for instance, morphologically)
selected sets of cluster galaxies (e.g. Ferguson 1989) or objectively
chosen sets of galaxies with unconfirmed cluster membership/unknown
distances (e.g. Irwin et al. 1990).
However, the development of a new generation of multi-object spectrographs,
exemplified by the `2 degree field', or 2dF, multi-fibre spectrograph
on the Anglo-Australian Telescope (AAT) (see e.g. Taylor, Canon \&
Parker 1998),
has made it possible to conduct a truly complete
spectroscopic survey of a given area on the sky, down to well
determined, faint limits, irrespective of image morphology or any
other pre-selection of target type, and thus obtain redshifts for
a large and unbiased sample of targets (Phillipps 1997).
Our Fornax Cluster Spectroscopic Survey (FCSS) aims to exploit
the huge multiplexing advantage of 2dF in this way by surveying, in total, a region of
12~deg$^{2}$ centred on the Fornax Cluster of galaxies.
The prime reason for choosing this area of sky area
was, of course, the presence of the Fornax Cluster itself.
Fornax, with Virgo, is the nearest reasonably rich cluster
(approximately Abell Richness Class 0 -- it is supplementary cluster
S0373 in Abell, Corwin \& Olowin 1989).
The targets of the FCSS
encompass both cluster galaxies, of a wide range of types and
magnitudes, and background galaxies (over a similarly
wide range of morphologies), as well as Galactic stars and QSOs
(Drinkwater et al. 2000a).
Our main motivation for an all-object survey was to
determine cluster membership for as complete a sample of objects as possible
across a wide range of surface brightnesses. We wished
to test, in particular, whether the usual assignment of LSBGs to the
cluster and high (or even `normal') surface brightness faint
galaxies to the background is justified. The situation regarding HSBGs
is even less clear than for LSBGs -- the question of the very existence
of normal or high surface brightness dwarf galaxies (excepting
possibly pathological cases like M32) has been highly controversial -- see,
for example, the contrasting views expressed in Ferguson \& Sandage
(1988) and Irwin et al.\ (1990). An up to date discussion is given in
Trentham \& Hodgkin (2002).
Drinkwater et al. (2000a) (hereafter Paper~I) outlined the aims
and methods of the FCSS and presented the results from the
first 2\degr\, diameter field surveyed. Drinkwater et al. (1999)
(Paper~II) described the nature of several compact galaxies found beyond
the cluster. Meyer et al. (2001) ( Paper~IV) discussed the quasars
found in the FCSS data. Drinkwater et al. (2000b) (Paper~III)
and Phillipps et al. (2001) (Paper~V) described the discovery of
5 extremely compact, high surface brightness dwarf galaxies
within the cluster. These objects (dubbed
`Ultra-Compact Dwarfs' -UCDs) have intrinsic sizes $\la$100~pc. They are
more compact and significantly less luminous than other known compact
dwarf galaxies, yet are much brighter than globular clusters.
The data from the first field (see Paper~I) revealed that
the Fornax Cluster is clearly defined kinematically.
A total of 24 galaxies from the data lie within the
cluster (this includes the 5 UCDs described in Papers~III and V).
This is, to our knowledge, the first complete sample of dwarf galaxies of known
distance (and hence luminosity) with well defined selection limits.
In this paper we study the nature of the objects in this complete sample.
Section 2 briefly describes the methods and present status of the FCSS.
Section~3 describes the contents of the complete Fornax Cluster sample and
presents $B$ and $R$ photometry,
radial intensity profiles and radial colour distributions for
all 24 confirmed cluster members. Using this data we are able to explore (in Section 4)
the distribution in luminosity and surface brightness
of the confirmed cluster members. Section 5 presents
our conclusions.
\section{The Fornax Cluster Spectroscopic Survey}
The FCSS survey-area will ultimately comprise four separate
2dF fields, covering the majority of the Fornax Cluster area
(see fig.~1 and table~1 of Paper~I). The results presented
in this paper are derived from the first of these four fields to be completed.
Field~1 is centred on the large galaxy NGC~1399 at the
centre of the cluster (R.A.=03$^{h}$38$^{m}$29.0$^{s}$,
Dec.=35\degr27\arcmin01\arcsec, J2000).
In common with the main 2dF Galaxy Redshift Survey
(2dFGRS; see Colless et al. 2001 and references therein), we
have chosen to select our targets from catalogues based on UKST Sky
Survey plates digitised by the Automatic Plate Measuring (APM) facility
at Cambridge (see Irwin, Maddox \& McMahon 1994). However,
unlike other galaxy surveys, which only
select resolved images for spectroscopic measurement, we avoid any
morphological pre-selection and include all objects, both resolved and
unresolved (i.e.\ `stars' and `galaxies'). This means that we can
include galaxies with the greatest possible range of surface
brightnesses: our only selection criterion is the (blue) magnitude
limit. Including objects normally classified as stars greatly
increases the size of our sample but it is the only way to ensure
completeness.
In order both to cover
a large number of targets and to go significantly deeper than previous
spectroscopic cluster surveys we chose to limit our survey at a $b_{j}$ magnitude
of 19.7. This is marginally deeper than the main
2dFGRS, which is limited at $b_{j} = 19.45$.
$b_{j}$ is the natural $B$ band photographic system of the UKST data.
This is related to the standard Cousins $B$ magnitude by
$b_j=B-0.28\times(B-V)$ (Blair
\& Gilmore 1982).
Targets are also chosen to be fainter than $b_{j} = 16.5$, since the
photometry of brighter objects is problematic with the UKST/APM data
(particularly because of image saturation).
Note that the selection limits of the overall APM
object catalogue (images larger than a certain area at a given isophotal detection
threshold) do not impinge on our target selection, since any galaxies
brighter than $b_{j}$ = 19.7 and with surface brightness high enough
to be observable with 2dF (see below) are well above the APM limit
(see also Paper I).
Our input catalogue for the FCSS is a standard APM `catalogues' file
(Irwin et al. 1994) of field F358 from the UKST
Southern Sky Survey. The field is
centred at R.A.=03$^{h}$37$^{m}$55\fs9, Dec.=--34\degr50\arcmin14\arcsec\ (J2000) and
the region scanned for the catalogue file is 5.8$\times$5.8~deg$^{2}$,
approximately centered on the Fornax Cluster. The APM image catalogue lists image positions,
magnitudes and morphological classifications (as `star', `galaxy',
`noise', or `merged') measured from both the blue and red survey
plates. The `merged' image classification indicates two overlapping
images: at the magnitudes of interest for this project the merged
objects nearly always consisted of a star overlapping a much fainter
galaxy. All the positions are measured from the more recent red survey
plate (epoch 1991 September 13 compared to 1976 November 18 for the
blue plate) to minimise problems with proper motions. The APM
catalogue magnitudes are calibrated for unresolved (stellar) objects
only (see Bunclark \& Irwin 1984), so we supplemented these with total magnitudes for the galaxies
measured by direct analysis of the digitised
plate data. These magnitudes were
obtained by fitting exponentials to the APM galaxy surface brightness
profiles (see Phillipps et al. 1987 and Morshidi-Esslinger, Davies \& Smith 1999)
across the surface brightness range 22.7 to 25.7~B~mag~arcsec$^{-2}$ (i.e.
between the possibly saturated inner parts of the images and the
sky noise dominated outer parts). Central surface brightnesses and
exponential scale lengths could be determined in this way and
combined to give pseudo-total magnitudes (see
Paper I for further details); we refine
these measurements for the cluster objects in the present paper.
Our target selection then consisted simply of taking all objects from the
APM catalogue in each of our four 2dF fields with magnitudes in the
range $16.5 \leq b_j \leq 19.7$.
We take the distance to the
cluster to be 20~Mpc (a distance modulus of 31.5~mag) as derived
by Drinkwater, Gregg \& Colless (2001b). This leads to an absolute
magnitude range for cluster galaxies --14.8 $\la$ $M_{B}$ $\la$ --11.6, so
all the cluster galaxies from FCSS are dwarfs in terms of their luminosity.
We did not apply any morphological selection,
although the APM image classifications from the blue survey plate were
used to determine which photometry to use (i.e. the catalogued APM
magnitude in the case of unresolved images, or our fitted magnitudes for
resolved images).
Although our targets cover a wide range of surface brightnesses (see fig.~1
of Paper I),
our final spectroscopic sample inevitably suffers from
incompleteness at the low surface brightness end. It is not
possible to measure spectra for the faintest LSBGs catalogued
(at central surface brightness around 24.5~B~mag~arcsec$^{-2}$) in
reasonable exposure times even though the multiplex advantage of 2dF
enables us to profitably expose for longer times and hence
go fainter than most previous nearby cluster work. Two hour
exposures enabled us to obtain measurable spectra for
galaxies with central surface
brightnesses down to 23.2~B~mag~arcsec$^{-2}$
with a success rate of $\geq$80 per cent (see Paper I). In
order to try to extend our surface brightness limit, we have also
experimented with much longer exposures (around 4 hours) for the
lowest surface brightness subset and have obtained successful
spectra for some objects with surface brightnesses as faint as
23.7~B~mag~arcsec$^{-2}$. This forms the practical lower limit to our cluster
galaxy sample.
We observed all our targets with an identical observing setup for 2dF:
the 300B grating and a central wavelength setting of 5800\AA\ giving a
wavelength coverage of 3600--8010\AA\ at a resolution of 9\AA\ (a
dispersion of 4.3\AA\ per pixel).
The 2dF facility includes its own data reduction package, {\sc 2dfdr},
which permits fast, semi-automatic reduction of data direct from the
instrument. However, when we started the FCSS project
{\sc 2dFdr} was still under
development, so we chose instead to reduce the data with the {\sc
DOFIBERS} package in {\sc IRAF}\footnote{{\sc IRAF} is distributed by the National
Optical Astronomy Observatories, operated by the Association of
Universities for Research in Astronomy, Inc. under cooperative
agreement with the NSF.}, as described in Paper I.
We have since reduced the data independently
with {\sc 2dFdr}, obtaining entirely comparable results (Deady 2002).
In particular, 98 per cent of spectra gave the same velocity shift (see below)
to within the quoted errors; the remaining 2 per cent turned out, on further inspection, to have
spurious features which had not been removed by the cleaning of the original
data.
\begin{figure}
\epsfig{file=figure1.eps}
\caption{The unfluxed 2dF spectra for 2 of the Fornax Cluster sample.
The full set of spectra can be viewed at http://www.star.bris.ac.uk/pjb/2dfspectra.htm.}
\end{figure}
Redshifts are determined in a standard fashion
via cross-correlation of the spectra with templates,
using {\sc RVSAO} in {\sc IRAF} (Kurtz \& Mink 1998). However, as we required
an automated system capable of coping with a large number of stars as well
as galaxies, we have used stellar template spectra of various types, rather
than absorption line galaxy spectra. These were chosen from the library of
Jacoby, Hunter \& Christian (1984). In addition we use emission line
galaxy and quasar templates. For each spectrum we determined the
best-matching template, the Tonry-Davies (1979) R coefficient,
the redshift and its error.
(Again see Paper I for more details and illustrations of the template
spectra used).
The redshifts are measured as radial velocities in units of
$cz$ and are subsequently converted to heliocentric values. By
choosing the template giving the best $R$ coefficient we can determine
not only the redshift, but a first estimate of the object type. We
only accept identifications with $R\ge3$. Objects with redshifts of
$\approx$500~km\,s$^{-1}$ or less are Galactic stars for which the best
template indicates the stellar spectral type. At higher redshifts,
external galaxies are separated into absorption-line types if they
match one of the stellar spectra or emission-line types if they match
the emission-line galaxy template best.
Typical velocity errors are around 64~km~s$^{-1}$, while comparison with 44
galaxies with redshifts given in the NASA/IPAC Extragalactic Database (NED)
reveals a negligible external mean error of $7 \pm 17$ km~s$^{-1}$ and
rms scatter of 110~km~s$^{-1}$, consistent with our internal errors given
the heterogeneous data compiled in NED.
\begin{table*}
\begin{minipage}{140mm}
\caption{The confirmed members of the Fornax cluster sample}
\begin{tabular}{lrrrllrr} \hline
Name & & & & & FCC & $cz \; \; \; \;$ & \\
FCSS & FCC & FLSB & NG & Other & Type & km~s$^{-1}$ & Template \\ \hline
J033418.4-354740 & 132 & -- & -- & & dE2 & $1883 \pm 98$ & M0\\
J033612.9-360959 & 164 & 64 & -- & & dS0(5),N & $1485 \pm 45$ & K5\\
J033653.3-345618 & 181 & 71 & 56 & HPW19 & dE2,N & $1333 \pm 69$ & G6\\
J033703.3-353804 & -- & -- & -- & UCD1 & & $1491 \pm 39$ & K5\\
J033718.0-354157 & 194 & 252 & -- & & dE3 & $1237 \pm 84$ & G0\\
J033723.3-345400 & 195 & 80 & 55 & H14 & dE5,N & $1315 \pm 69$ & G6\\
J033734.1-354945 & 196 & 83 & 23 & & dE6 & $1797\pm 129$ & M0\\
J033754.7-345255 & 200 & 256 & -- & & dE2 & $1184 \pm 110$ & F6\\
J033806.3-352858 & -- & -- & -- & UCD2 & & $1312 \pm 57$ & K5\\
J033816.7-353027 & (B1241) & 95 & -- & & (dE3,pec?) & $2012 \pm 91$ & G6\\
J033821.5-351535 & 211 & -- & 71 & & dE2,N & $2192 \pm 49$ & F6\\
J033854.1-353333 & -- & -- & -- & UCD3 = CGF1-4 & & $1591 \pm 36$ & K5\\
J033905.8-360556 & 221 & 108 & 43 & & dE4,N & $1724 \pm 77$ & F6\\
J033919.5-354334 & 223 & 114 & 6 & HPW28 & dE0,N & $902 \pm 83$ & K5\\
J033935.9-352824 & -- & -- & -- & UCD4 = CGF5-4& & $1920 \pm 40$ & K5\\
J033952.5-350424 & -- & -- & -- & UCD5 & & $1321 \pm 66$ & F6\\
J034001.3-344529 & 230 & 120 & 76 & HPW29 & dE5,N & $1149 \pm 84$ & G0\\
J034023.5-351636 & 241 & 124 & 5 & HPW30 & dE0,N & $2045 \pm 107$ & F6\\
J034033.8-350122 & 245 & 126 & 3 & & dE0,N & $2265 \pm 43$ & K5\\
J034042.3-353940 & 247 & 127 & 40 & & dE3/Im? & $1097 \pm 108$ & F6\\
J034100.8-354433 & 254 & 129 & 22 & & dE0,N & $1517 \pm 94$ & G0\\
J034131.8-353521 & 264 & 135 & -- & & dS0(8),N & $2033 \pm 64$ & G6\\
J034159.4-352053 & (B1554) & -- & -- & CGF3-1 & (E) & $1667 \pm 59$ & K5\\
J034217.3-353226 & 274 & 141 & 15 & HPW39 & dE0,N & $977 \pm 85$ & K5\\ \hline
\end{tabular}
\end{minipage}
\end{table*}
In the second stage of the identification process we check each
identification interactively (cf. Colless et al. 2001),
using the {\sc RVSAO} package to
display the best cross-correlation and the object spectrum
with common spectral features plotted at the calculated redshift. In the
small number of cases
when the redshift is obviously wrong (e.g.\ with the calcium {\sc H} and
{\sc K}
lines clearly present but misidentified), it is flagged as being incorrect
or in some cases is recalculated interactively.
Objects still not dealt with successfully at this stage are flagged to be
re-observed.
Note that the
2dF spectra, though of low resolution and unfluxed, are still useful
for more detailed analysis than simple redshift measurements and
object classifications. However, we
defer any detailed analysis of the spectra in terms of, e.g.,
emission or absorption line equivalent widths, line ratios,
absorption line indices or rough metalicities
to later papers.
\section{The Fornax Cluster Sample}
\subsection{Composition of the sample}
Within Field 1, we found 24 galaxies with redshifts which place them in the
Fornax Cluster (mean $cz \simeq 1450$ km~s$^{-1}$).
(In fact, 26 of the input APM catalogue objects have such
velocities, but closer inspection reveals one to be an individual HII
region of the large irregular galaxy NGC~1427A and the other to be
a mistakenly demerged part of the image of the giant galaxy NGC~1381.)
Assignment to the cluster is unambiguous, as the cluster is well
isolated in redshift space (see Paper I). Our magnitude
selection limits place all the sample objects between roughly
$M_{B}$=--14.8 and $M_{B}$=--11.6; more precise magnitudes are determined
in Section~3.2 below. Five of the objects were originally observed as `stars', i.e.
they had unresolved or only marginally resolved images on the plate material
used for the APM catalogue.
The relatively small number of cluster galaxies is largely due to the
lower bound on the surface brightness at which successful fibre spectra
can be obtained. Many dwarfs at these faint magnitudes are expected to
have surface brightnesses well below our practical 2dF limit at
around 23.7~B~mag~arcsec$^{-2}$ (see, e.g., Irwin et al. 1990, Kambas et al. 2000).
Our 24 objects are listed in R.A. order in Table 1. We give their
official IAU designations along with their names in other recent Fornax catalogues.
FCC refers to Ferguson's (1989) Fornax Cluster Catalog, FLSBG numbers
are from the Fornax Low Surface Brightness Galaxy catalogue of Davies et al.
(1988), continued in Irwin et al. (1990) and the UCDs are from Phillipps
et al. (2001). NG refers to the galaxies listed by Caldwell (1987).
Note that the two non-compact galaxies without FCC numbers were given
background galaxy numbers by Ferguson (FCCB 1241 and 1554), but clearly
these designations are now inappropriate (see Section 4.2 for a full
discussion of this issue). The latter is also CGF3-1 from Hilker et al.
(1999), and UCD3 and UCD4 are CGF1-4 and CGF5-4 from the same source.
Finally
objects FCC 181, 195, 223, 230, 241 and 274 were earlier studied by Hodge
(1960) or Hodge, Pyper \& Webb (1965), where they were the dwarfs
HPW 19, H 14 and HPW 28, 29, 30 and 39, respectively (see Irwin et al. 1990).
As noted, 17 of the 19 non-compact galaxies appear as cluster galaxies
in Ferguson's FCC,
and we have included Ferguson's assessment of their morphological types
from his large plate scale Las Campanas plates. All are classified dE
or, in two cases, dS0; 12 were judged `definite members', 4 as `probable members'
and 1 as a `possible member' (see Section 4.2).
The remaining columns in Table 1 give our derived heliocentric redshift
(given as a `velocity', $cz$) and the best fitting stellar template (there
are no emission line objects in the sample once the NGC~1427A
HII region has been removed).
Example spectra are shown in Fig.~1 (all of them
can be accessed at http://www.star.bris.ac.uk/pjb/2dfspectra.htm). Remember that no correction
for throughput as a function of wavelength in each fibre has been attempted,
so the continuum shape in the plots is fairly arbitrary.
The majority (17) of the spectra are best matched by G (7) or K (10)
star templates, with 6 matches to F types and 2 to M types. The fact that
none of our objects displays any sign of star formation corroborates
Ferguson's dE and dS0 classifications.
The non-detection of any dwarf irregular (dI) galaxies within Field 1 is noteworthy.
Field 1 is centered on NGC1399 and covers the central part of the cluster.
Drinkwater et al. (2001a) obtained Flair-II
spectra for 108 brighter ($b_j<$18) members of the cluster over a
6\degr\, field. They found that among the dwarfs in
their sample ($M_{B}\ge$--16.5), the early-type dwarfs dominate in the cluster
centre (within a radius of 100 arcmin) with late-type dwarfs being concentrated
in the outer regions of the cluster.
Recent theoretical work on dynamical evolution
has suggested that, in rich clusters,
tidal forces from the cluster's tidal field
can result in significant morphological evolution, enough
to convert dIs to dEs (Moore, Lake \& Katz 1998). Boyce et al. (2001)
suggested that such forces may explain their observation that the core of
the rich cluster Abell~868 is devoid of dIs but that there are similar proportions
of dIs to dEs beyond the cluster centre. Mayer et al. (2001) have suggested
that a related process (`tidal stirring') may explain the distribution of dI and dE
galaxies in the Local Group. In this process tidal interactions between
the dwarfs and the larger member galaxies cause the tidal instabilities which induce
evolution from dI to dE. In a poor, but centrally concentrated cluster
like Fornax, both of these processes may be at work.
\subsection{Photometry of the sample}
\begin{table*}
\begin{minipage}{100mm}
\caption{Derived photometric parameters for the Fornax cluster sample}
\begin{tabular}{lrcrrrr} \hline
Name & $\alpha \; \;$ & $\mu_{0}$ & $B$ & $R$ & $M_B$ & $a \;$ \\
FCSS & arcsec & B mags arcsec$^{-2}$ & & & & kpc \\ \hline
J033418.4-354740 & 3.64 & 23.2 & 18.8 & 17.3 & --12.7 & 360 \\
J033612.9-360959 & 5.93 & 21.9 & 16.7 & 15.3 & --14.9 & 580 \\
J033653.3-345618 & 5.94 & 22.9 & 17.6 & 16.3 & --13.9 & 580 \\
J033703.3-353804$^{\star}$ & 0.89 & 21.5 & 20.2 & 18.7 & --11.4 & $<100$ \\
J033718.0-354157 & 4.13 & 23.3 & 18.6 & 17.3 & --12.9 & 410 \\
J033723.3-345400 & 7.27 & 22.8 & 17.1 & 15.7 & --14.4 & 720 \\
J033734.1-354945 & 7.90 & 23.7 & 18.2 & 16.5 & --13.3 & 790 \\
J033754.7-345255 & 4.62 & 23.0 & 17.8 & 16.4 & --13.7 & 450 \\
J033806.3-352858$^{\star}$ & 1.41 & 22.3 & 19.9 & 18.3 & --12.4 & $<140$ \\
J033816.7-353027 & 5.43 & 22.7 & 17.5 & 16.1 & --14.0 & 570 \\
J033821.5-351535 & 4.35 & 21.5 & 16.5 & 15.1 & --15.0 & 470 \\
J033854.1-353333$^{\star}$ & 0.98 & 20.7 & 18.9 & 17.2 & --13.6 & $<100$ \\
J033905.8-360556 & 3.86 & 22.6 & 18.0 & 16.5 & --13.5 & 380 \\
J033919.5-354334 & 9.21 & 23.2 & 16.6 & 15.3 & --14.9 & 900 \\
J033935.9-352824$^{\star}$ & 1.00 & 21.9 & 19.9 & 18.5 & --12.5 & $<100$ \\
J033952.5-350424$^{\star}$ & 0.97 & 21.8 & 20.1 & 18.7 & --11.6 & $<100$ \\
J034001.3-344529 & 5.30 & 22.6 & 17.6 & 16.4 & --13.9 & 520 \\
J034023.5-351636 & 8.76 & 23.7 & 17.1 & 16.0 & --14.4 & 860 \\
J034033.8-350122 & 6.85 & 22.3 & 16.3 & 15.0 & --15.2 & 670 \\
J034042.3-353940 & 6.02 & 23.4 & 18.1 & 17.1 & --13.4 & 590 \\
J034100.8-354433 & 9.42 & 24.6 & 18.1 & 16.6 & --13.4 & 920 \\
J034131.8-353521 & 6.08 & 21.9 & 17.0 & 15.7 & --14.5 & 600 \\
J034159.4-352053 & 3.23 & 22.1 & 17.8 & 16.2 & --13.8 & 320 \\
J034217.3-353226 & 6.30 & 22.8 & 16.9 & 15.8 & --14.7 & 620 \\ \hline
\end{tabular}
\end{minipage}
\end{table*}
As noted earlier, initial photometry for all the catalogue galaxies was
derived from the APM scan data in automated fashion, assuming pure
exponential radial profiles and circular isophotes (cf. Morshidi-Esslinger
et al. 1999). We have now improved on this photometry
by using
photographic data from a IIIaJ $B$ plate and a Tech Pan $R$ film
from the UKST, digitised by the SuperCOSMOS measuring machine at the
Royal Observatory, Edinburgh (e.g. Hambly, Irwin \& MacGillivray 2001).
Table~2 presents the results of this photometry.
The
data were calibrated (Deady 2002)
onto the standard Cousins $B$ and $R$
system via comparison of the surface brightness profiles of four suitable, lowish
surface brightness dwarfs (so as to avoid problems of saturation at
moderate to high surface brightness on the plates) with corresponding
CCD data from Davies, Phillipps \& Disney (1990).
The methodology was as in Phillipps \& Parker (1993).
Note that none of the calibrating galaxies are actually in the present cluster
galaxy sample, either being outside the 2dF area or of too low a surface
brightness for 2dF spectroscopy. The Davies et al. (1990) calibration is
closely similar (to 0.1~mag) to both that of the original APM low
surface brightness galaxy survey of Davies et al. (1988) and the Fornax
Cluster Catalog of Ferguson (1989). After independently calibrated
Curtis Schmidt data became available (Karick et al., in prep.)
we were able to confirm that our photographic and CCD $B$ magnitude
scales were in excellent agreement, too.
To determine total magnitudes for our galaxies, we used the {\sc SEXTRACTOR} package
(Bertin \& Arnouts 1996) to generate Kron magnitudes (Kron 1978). We selected
an aperture size of 3.5 Kron radii (cf. Metcalfe et al. 1995),
which contains 99.3 per cent of the light for an exponential profile. Our
magnitudes should therefore be total to within $\simeq$0.01~mag. A
more significant source of error will be the determination of the
background sky level. A realistic measure of the uncertainty in this
was judged to be $\sigma_{sky}/10$, where $\sigma_{sky}$ is the width
(standard deviation) of the histogram of local sky values. Repeating
the Kron radius and magnitude calculation with the background varied
by this amount gives our estimate of the magnitude errors from this
source. Allowing also for possible systematic zero point errors of $\simeq$0.05~mag
we find final photographic magnitude errors $\simeq$0.1--0.2~mag.
The $B$ and $R$ magnitudes for all our cluster galaxies are summarized in columns
4 and 5 of Table~2. Note, however, that the photographic data
of the UCDs (denoted
by an asterix in Table~2) suffer
from saturation, so the $B$ and $R$ values in Table~2 only provide
faint magnitude limits for these objects.
Column 6 presents $M_{B}$ values assuming
D=20~Mpc (see Section 2). In this column, we have avoided the problems of
the saturation of the photographic data for the UCDs,
by including, for these objects, the absolute magnitudes
derived from Curtis Schmidt CCD data by Phillipps et al. (2001), and corrected from $b_{j}$
to $B$ by assuming $B-V$$\simeq$0.7.
The $B-R$ colours are all rather similar (nearly all between 1.2 and 1.5),
and are consistent with the spectral
classifications as quiescent dE or dS0 galaxies. Aside from the UCDs, the
galaxies have a mean
colour of $<B-R>$=1.33$\pm$0.03, essentially identical
to the $<B-R>$=1.38 reported by Davies et al. (1990) for
their (presumed) cluster LSBGs. Although the numbers are small,
there is a (comforting) slight trend for redder colours for the objects
which match best to later type stellar spectra (in particular
the two M0 matches have an average $B-R$ of 1.54).
\begin{figure*}
\epsfig{file=figure2a.eps}
\caption{Radial intensity profiles and radial colour profiles
for the Fornax Cluster sample}
\end{figure*}
\addtocounter{figure}{-1}
\begin{figure*}
\epsfig{file=figure2b.eps}
\caption{continued}
\end{figure*}
\addtocounter{figure}{-1}
\begin{figure*}
\epsfig{file=figure2c.eps}
\caption{continued}
\end{figure*}
\addtocounter{figure}{-1}
\begin{figure*}
\epsfig{file=figure2d.eps}
\caption{continued}
\end{figure*}
\addtocounter{figure}{-1}
\begin{figure*}
\epsfig{file=figure2e.eps}
\caption{continued}
\end{figure*}
The {\sc STARLINK} routine {\sc ELPROF} was used to determine
the radial intensity profile, as a function of semi-major axis length,
for the cluster dwarfs (as well as the calibrating galaxies) from the
photographic $B$ and $R$ data. The radial intensity
profiles and radial colour distributions
are illustrated in
Fig.~2. Most of the radial intensity profiles are well fitted by a simple exponential law (straight
line in the plot), typical of dwarf galaxies, though some do show some
curvature. In some cases this is due merely to the saturation of the
$B$ band measurements at $\simeq$22~B~mag~arcsec$^{-2}$. This is less of a problem
with the $R$ profiles as the Tech Pan films have somewhat better
dynamic range. For completeness we have therefore also fitted the
more general Sersic (1968) law (cf. Davies et al. 1988). This
fitting confirms that 13 of the 19 normal (i.e. not ultra-compact)
dwarfs have best fitting indices $n = 1.0 \pm 0.2$ (in the $B$ band),
where $n = 1$ corresponds to a pure exponential. Four of the remainder have
$n$ between 0.4 and 0.8 (a pure de Vaucouleurs (1959) law, appropriate for
giant ellipticals, would have $n = 0.25$), while two have $n \simeq 1.4$.
Given that the differences in $n$ calculated for the $B$ and $R$ band
profiles are typically around 0.2 (even when there is no systematic
colour gradient), that is probably a reasonable estimate of the level of
uncertainty in $n$. For subsequent fitting we thus assume that the
standard exponential profile is a sufficiently good fit in all cases.
Columns 2 and 3 of Table 2 therefore contain
the scale sizes $\alpha$ (in arcsec) and extrapolated
central surface brightnesses $\mu_{0}$ of the exponential fit to the $B$ band
radial intensity profiles. Column 7 lists the scale size in kpc (assuming D=20~Mpc).
Fitting was carried out in the range between
22.5 and 26~B~mag~arcsec$^{-2}$. For the compact objects, the tabulated values
are merely limits on the intrinsic surface brightnesses and scale lengths
due to the effects of the point-spread function (as discussed in more
detail in Paper V). Errors in the scale sizes from the fitting
procedure are typically of order 10 per cent, and the differences between the
$B$ and $R$ scale lengths are at the same level in almost all cases
(though a couple show systematic colour gradients and therefore
differences up to $\sim$20 per cent). One galaxy, FCSS J034100.8-354433,
is difficult to treat as
simply as the others as it appears to have a two component structure.
In the table we give the parameters for its `outer disc' (which has a
very low surface brightness around 24.5~B~mag~arcsec$^{-2}$,
making it an outlier in the plots below),
but it also has a much more steeply rising `inner bulge' with actual central
surface brightness more like 23~B~mag~arcsec$^{-2}$ (hence the
successful spectroscopy of this object),
and a scale size of around 2~arcsec.
Given the small range of colours for most of the sample, unsurprisingly no
correlation between colour and surface brightness or scale length is
apparent.
\section{Discussion}
\subsection{The Bivariate Brightness Distribution (BBD) of the cluster sample}
Two necessary steps towards any full understanding of the processes
of galaxy formation and evolution are (i) the compilation of a
full and complete census of the local extragalactic
population and (ii) an accurate and detailed
quantification of this population. Traditionally
astronomers have sought to quantify the local galaxy population
by constructing galaxy luminosity functions (LFs).
The LF of a sample of galaxies is the space density of the
galaxies in that sample as a function of absolute luminosity.
In theory, if one could select a
complete sample of galaxies over a sufficiently large
section of the local Universe such that large-scale structure and
clustering effects could be ignored then the LF which
could be constructed would provide a definitive measure of the
local luminosity density. In practice the many studies of
the field population LF (Loveday et al. 1992;
Lin et al. 1997; Marzke et al. 1997; Bromley et al. 1998; Muriel, Valotto \& Lambas 1998;
Zucca et al. 1997; Folkes et al. 1999)
have produced significantly differing
results and have generally not been complete to very faint limits.
An alternative approach is to seek for clues about
galaxy formation and evolution by
constructing the LF for
particular environments and then studying the difference in the derived LFs between
environments.
The many recent studies of the LFs of particular clusters (e.g. Driver et al. 1994;
de Propris et al. 1995; Lobo et al. 1997; Wilson et al. 1997;
Valotto et al. 1997; Smith, Driver \& Phillipps 1997; Trentham 1997a,
1997b, 1998; Driver, Couch \& Phillipps 1998b; Garilli, Maccagni \& Andreon 1999)
fall into this
category.
\begin{figure*}
\epsfig{file=figure3.eps}
\caption{Plot of B-band extrapolated central surface brightness
against B-band scale-length
for the Fornax Cluster sample.
The Ultra Compact Galaxies (UCDs) are denoted by filled triangles, the other Fornax cluster galaxies by
filled squares. The open squares show the observed central surface brightness
for those 3 nucleated galaxies where this is significantly brighter than the extrapolated central surface brightness.}
\end{figure*}
However, it is now well established that
the inclusion of a galaxy in an optically
selected sample is critically dependent not only on its absolute
luminosity but also on its surface brightness (Disney 1976;
Phillipps, Davies \& Disney 1990; Impey \& Bothun 1997). It is also
well established that LSBGs exist with a wide range of total
luminosities. Such galaxies will have been excluded from most determinations
of the LF. A more thorough
and physically meaningful way to quantify the extragalactic population
is the construction of a Bivariate Brightness Distribution (BBD), i.e.
the space density of galaxies as a function of their total luminosity and
their surface brightness or, equivalently,
as a function of their size and surface brightness
(Boyce \& Phillipps 1995; Dalcanton 1998; Cross \& Driver 2002).
A consideration of any possible surface brightness--luminosity relation
is then encompassed as part of the wider determination of the BBD.
Past attempts at deriving
a BBD (e.g. Choloniewski 1985; van der Kruit 1987;
Sodre \& Lahav 1993) have generally suffered from poor statistics.
The advent of wide-field imaging surveys and dedicated spectroscopic surveys
has made the measurement of the BBD for statistically significant populations
possible for the first time (Driver 1999; Cross et al. 2001;
de Jong \& Lacey 2000; Blanton et al. 2001). These authors have all
studied the `field' population to relatively bright limits.
A thorough attempt, therefore, to catalogue and quantify the contents of a cluster,
requires, as a first stage, the detection of a complete sample of
cluster objects to well defined (preferable very faint) limits in absolute luminosity and
surface brightness. Secondly, the morphologies of the various objects found need
to be determined. Finally, the BBD for each morphological type
can be constructed. Even this could be further refined since, for example,
one might expect the morphological composition of the cluster core to differ from its outer
regions. Hence, one may wish to study the BBD as a function of cluster radius.
Whilst we do not yet have a large enough sample of Fornax Cluster objects to undertake a
full determination of the cluster BBD, we do have a complete sample close to the cluster centre
with well defined luminosity and surface brightness limits and can, therefore,
draw some general conclusions about the luminosity--surface brightness or
size--surface brightness
distribution for the first time.
Fig.~3 shows the size--surface brightness distribution of the Fornax
Cluster sample using the B-band data from Table~2 (i.e. plotting
$\mu_{0}$ against $\alpha$).
The limits imposed by the APM magnitude selection limits
($16.5 \leq b_{j} \leq 19.7$) are noted.
These boundaries become slightly
fuzzy in terms
of the final photometry since the target selection was in terms of
APM $b_j$ magnitudes.
The actual APM catalogue limit (on isophotal size at a threshold surface
brightness)
hardly impinges on our sample, as noted earlier,
since the low surface brightness galaxies which we are biased against
including because of their small isophotal sizes
(see e.g. Davies et al. 1988) have too low a surface brightness
to be successfully observed by 2dF anyway.
Also noted in Fig.~3 is the position of
our absolute surface brightness limit at 23.7~B~mag~arcsec$^{-2}$.
The surface brightness effective limit
is a fuzzy boundary between 23.2 and 23.7~B~mag~arcsec$^{-2}$ where the 2dF
success rate falls from $\simeq$80 per cent to zero.
Since we are dealing with fibre spectra, it is only the central
region of the object which is observed. This can have a positive effect
on the inclusion of low surface brightness objects in
the case where they are nucleated (i.e. dE,Ns)
since the nucleus may be bright enough to obtain a successful spectrum
even if it sits in a very low surface brightness galaxy.
For example, we noted
in Section 3.2 that J034100.8-354433 has a prominent nuclear bulge
and that the values of $\mu{o}$ and $\alpha$ included in Table~2 refer only to
the outer disk component of this galaxy. In fact,
according to the FCC, 12 of our sample are nucleated (see Table 1). However, in
our imaging
data only 3 (J034023.5-351636, J034100.8-354433 and J034217.3-353226)
show an observed central surface brightness significantly
brighter than the extrapolated value listed in Table 2. For these 3
objects, we show on Fig.~3 the observed as well as the extrapolated central surface
brightness. This makes it clear how J034100.8-354433 in particular
has been included in the sample
despite having an extrapolated central surface brightness of 24.6~B~mags~arcsec$^{-2}$.
From Fig.~3 it can be clearly seen that the Fornax sample
falls into two distinct areas of the size--surface brightness
plane.
Firstly,
the UCDs can be seen around $\alpha$$\simeq$1~arcsec ($\simeq$100~pc at the
distance of Fornax) between
$\mu_{o}$=20.5--22.3~B~mag~arcsec$^{-2}$.
Secondly, the rest of the (non-compact) dwarfs have $\alpha>$3~arcsec.
Above $\alpha$=3~arcsec ($\simeq$300~pc), the cluster sample essentially fills all of the
available ($\alpha$,$\mu_{o}$) parameter space and therefore
shows no correlation between the parameters. This non-correlation
was claimed previously
on purely photometric grounds by Davies et al. (1988) and Irwin et al. (1990).
However, the area of parameter space between $\alpha$=1--3~arcsec
contains no galaxies. There appears to be a minimum scale size which
a Fornax dE can have.
Whilst Fig.~3 re-emphasises that the UCDs appear to form a distinct population
from the larger `normal' dwarfs, one cannot extrapolate
too much about the relative properties of the UCDs and the
non-compact objects from
this figure alone.
As noted above, due
to the unresolved nature of the UCDs in the photographic data,
the plotted values of
$\mu_{o}$ and $\alpha$ are merely limits: the UCDs may be smaller and have
higher surface brightness than we measure.
In fact a preliminary analysis of HST imaging of these objects
(Drinkwater et al, in prep.) suggests that the UCDs actually have scale
sizes as small as 0.1~arcsec and central surface brightnesses up to 3~mag
brighter than measured from our photographic data. The bright central surface
brightnesses of the UCDs will make them detectable to fainter magnitudes than
the non-compact objects and hence the relative number densities of the two populations
are not clearly illustrated by Fig. 3. The nature of the UCDs
is currently a subject of some discussion (Drinkwater et al. 2000b; Hilker
et al. 1999; Mieske, Hilker \& Infante 2002), the debate focussing
on whether the UCDs form a distinct population of galaxies not linked to globular clusters
or whether there is a smooth
transition between both populations. We defer further consideration of this issue
until presentation of our HST results.
From Fig.~3, we can say little about the area with surface brightness fainter than 23.7~B$\mu$.
However, it is noteworthy that we have detected 6 objects with
surface brightness between 23.2--23.7~B~mag~arcsec$^{-2}$, despite the completeness of the
sample as whole falling from 80 to 0 per cent in this
range.
We might reasonably expect further dwarfs to lie within this region.
\begin{figure*}
\epsfig{file=figure4.eps}
\caption{Plot of B-band
extrapolated central surface brightness against
B-band absolute magnitude for the Fornax Cluster sample.
The Ultra Compact Galaxies (UCDs) are denoted by filled triangles, the other Fornax cluster galaxies by
filled squares.}
\end{figure*}
Fig.~4 shows the relationship in its alternative form, showing
$B$ central surface brightness against
absolute $B$ magnitude. The upper and lower
$b_{j}$ magnitude selection limits are once again shown along with
the absolute surface brightness limit at 23.7~B~mag~arcsec$^{-2}$.
The outlier FCSS~J034100.8-354433
is again seen apparently defying these limits at
$\mu_{o}$$\simeq$24.5~B~mag~arcsec$^{-2}$,
$M_{B}$$\simeq$--13.4.
Also shown in Fig.~4 is the locus
for a galaxy with a pure exponential profile with a
constant scale size of 3~arcsec ($\simeq$300~pc at the distance of
Fornax), assuming
a circular image. There is a large area of
parameter space to the upper right of this line within which the FCSS
would have detected galaxies,
but the only galaxies
within it are the UCDs. However, as noted above, the UCDs may have
central surface brightness up to 3~mag brighter than derived from
our photometry. If so, then the plotted area of parameter space to the upper right of the
$\alpha$=3~arcsec locus would be completely empty.
Below the $\alpha$=3~arcsec locus the sample fills the entirety of
available parameter space.
It is notable that this limit in scale-size imposes on the sample
the upper-bound of an apparent surface brightness--magnitude `correlation'.
However, to our limits of $M_{B}$ and $\mu_{o}$
there is no evidence for a lower bound to any such correlation.
If such a surface brightness--magnitude correlation does exist
for the Fornax dwarfs it must be fairly broad. For example,
at $M_{B}$=--15.0 the surface brightness distribution is $\ga$2.5~mag in width.
\begin{figure*}
\epsfig{file=figure5.eps}
\caption{Plot of B-band
extrapolated central surface brightness against B-band absolute magnitude
for the Fornax Cluster sample (filled squares), for the
Local Group (open symbols -- see key) and for the faint end of Cross et al.'s (2001) 2dFGRS sample.}
\end{figure*}
In Fig.~5 we again present the Fornax dwarf sample on a plot of
central surface brightness against absolute luminosity. We also show on this figure the
positions of Local Group galaxies taken from Mateo (1998). One needs to be wary
of reading too much into this diagram.
Whilst we might reasonably
expect that Local Group galaxies form the most complete sample of galaxies at
faint luminosities and low surface brightness available, the Local
Group data does not form a sample to strictly defined completeness limits.
We are also, of course, comparing
galaxies from two different environments (i.e. the centre of the Fornax cluster and
the much more diffuse Local Group). The presence of several dIs in the Local Group, where
none was found in Field 1 of the FCSS, is testimony to this. In common with the
Fornax dEs, all of the Local Group dEs also have $\alpha$$\ga$300~pc.
Most of the Local Group dIs are also constrained by this scale size limit, apart
from two outliers, both of which lie in the region of ($M_{B}$,$\mu_{o}$) space
occupied by the UCDs. However, as noted above, the UCDs
have central surface brightnesses up to 3~mag brighter than derived from our photometry,
in which case the Local Group dIs do not in reality lie in a similar region
of ($M_{B}$,$\mu_{o}$) space.
The dSphs of the Local Group
do appear to show a correlation between surface brightness and absolute magnitude, although
these objects are generally considered to form a distinct population from the
brighter dEs. Many of the Local
Group dSphs have scale-sizes less than 300~pc.
Our FCSS data are not
sensitive enough to detect any similar objects in Fornax.
We also note on Fig.~5 the region at bright absolute luminosities ($M_{B}<$--16.0)
within which the galaxies from the faint end
of Cross et al.'s (2001) 2dFGRS sample lie. This is
one of the largest samples of bright galaxies yet assembled
and formed the basis of
Cross et al.'s determination of the BBD
(see also Cross \& Driver 2002). It should be noted, however, that Cross et al.'s
sample is not differentiated
by morphological type (it includes ellipticals, bulge dominated spirals, disc dominated
spirals and irregulars). In order to compare the Cross et al. sample with our data,
we have translated their effective surface brightnesses to central surface brightness
by assuming a shift of $\simeq$1.8~mag appropriate for an exponential profile.
Within the limits of their sample,
Cross et al. found a correlation between surface brightness and absolute magnitude
such that $\mu_{e}$ scales as 0.42$M_{B}$. The relationship
is around 2.5~mags wide in surface brightness at each absolute magnitude.
It is clear from the Fornax and Local Group data that this relationship
does not extend to fainter magnitudes. As noted, the lower limit to the
scale size of the dEs (at $\alpha$$\simeq$300~pc) creates
an apparent upper-bound to a surface brightness-magnitude `correlation'.
Obviously the $\mu_{e}$ of this upper-bound scales as $M_{B}$, i.e.
much steeper than the Cross et al. correlation. There is, though, no evidence that
there is a low surface brightness bound to any such correlation for dwarfs.
The correlation between surface brightness and absolute magnitude for Local Group dSphs
is also much steeper than that from Cross et al.'s bright galaxy sample.
\subsection{Morphology as an indicator of cluster membership}
A prime aim of the FCSS is to test the reliability of
using optical morphology as a basis for determining cluster membership.
Firstly, it is worth noting that only $\simeq$50 per cent of our complete cluster
sample galaxies actually have $\mu_{0}$$>$22.5~B~mag~arcsec$^{-2}$, the criterion often used
to define a low surface brightness galaxy. In other words, though all
our objects are dwarfs, many of them extreme dwarfs, in terms of luminosity,
half of them are {\em not} low surface brightness galaxies. So, any selection
based on the assumption that all cluster dwarfs are going to be LSBGs
will be flawed. Of course, many
other LSB dwarfs which are really present will be too faint for us to obtain
a spectroscopic redshift, so the true fraction of non-LSBG dwarfs will
be considerably smaller than 50 per cent, but the important point as regards
morphological (pre)selection is that such dwarfs do exist.
In addition we find 15 background objects with $\mu_{0}$$>$22.5~B~mag~arcsec$^{-2}$
which would morphologically be likely to be assigned to the cluster.
Our best point of comparison is with
Ferguson's (1989) FCC.
Ferguson catalogued 2678 objects from photographic plates of an area
40~deg$^{2}$ centered on the Fornax cluster. He gave each object
one of five classifications: `class 1 - definite cluster member';
`class 2 - probable cluster member'; `class 3 - possible
cluster member'; `class 4 - likely background object';
and `class 5 - definite background object'.
Of our complete
sample of 24 Fornax Cluster dwarfs, a total of 17 (71 per cent of our cluster
sample) were
ascribed by Ferguson to the cluster: 12 as class 1;
4 as class 2; and 1 as class 3.
One of our cluster sample (FCSS~J034159.4-352053) was described as
class 5 (`definite background object').
This object has the smallest scale size (3.23~arcsec) of any of the
non-UCDs in our sample, although its surface brightness and absolute magnitude
are not remarkable compared to the rest of the cluster sample.
Another of our sample (FCSS~J033816.7-353027) was listed by
Ferguson as a `class 4' (`possible background object').
This does not have a remarkable surface brightness or absolute magnitude although
it has one of the smaller scale lengths of the non-UCD objects
(5.43~arcsec). Ferguson did not include any of the UCDs within the FCC either as background galaxies
or cluster members. Presumably these were so compact as to be considered
foreground stars.
In addition, within the FCSS survey of Field 1 we found that
a further 5 objects which Ferguson had ascribed to the cluster (1 as class 1;
4 as class 2) are actually background objects. So within our field and
our selection limits, Ferguson ascribed 22 objects to the cluster. FCSS has found that 5
of these objects (i.e. 23 per cent of Ferguson's cluster objects) are actually background objects.
A further 9 objects which were listed by Ferguson as class 3 (`possible cluster
member') but included in his table of background galaxies, are confirmed
by FCSS as background objects.
To summarise: Ferguson did not include any of the FCSS's UCDs within
the FCC (either as 'cluster' or 'background' objects),
presumably assuming them to be stars. Ferguson did include within the FCC all 19 of the
non-compact cluster dwarfs from the FCSS: however, he mis-classified 2 of these as `background' objects.
Hence, the FCC contains 89 per cent of the non-compact
dwarfs within the FCSS cluster sample (but only
71 per cent if one considers all dwarfs in the FCSS). However, the FCC contains a further
5 objects as `cluster' members which have been shown by the FCSS to be background objects:
i.e. within the FCSS's field
and selection limits, 23 per cent of those objects
described as cluster members by the FCC
are shown by the FCSS to be background objects.
This clearly illustrates that attempts to determine cluster membership solely
on the basis of observed morphology can produce significant errors: firstly because many compact galaxies will
be mis-classified as stars; and secondly because many non-compact background objects will be
mis-classified as cluster members and a smaller number of cluster members will be
mis-classified as background objects.
\section{Conclusions}
We have obtained spectra for a morphologically unbiased set of cluster
dwarf galaxies as part of the Fornax Cluster Spectroscopic Survey.
The present paper provides the photometric study of these confirmed
cluster dwarfs. We find that, despite their low luminosities, not all
the dwarfs also have low surface brightness. There is therefore a wide
spread of parameters in the space of surface brightness versus luminosity
or scale size. The only correlation evident in these plots is the one
in the luminosity--surface brightness plane caused by the absence of
`normal' dwarfs with scale sizes below about 300~pc. Even then, a separate
class of Ultra-Compact Dwarfs fills some of this otherwise empty area of parameter
space. Morphological
selection of cluster dwarfs solely on the grounds
of their low surface brightness may therefore overlook a significant
number of dwarf galaxies which are virtually impossible to discriminate from
background, larger galaxies on appearance alone.
The data presented within this paper result from the completion of only one of the
intended 4 FCSS fields. We now have full data for the second field and are in the process
of reducing these. The completion of all 4 fields will not only improve the statistics of
our results but will also enable us to study the distribution
of dwarf galaxies
of different morphologies (i.e. dE, dI, UCD)
as a function of position within the cluster. Such studies will
be assisted by the
deep CCD g,r,i,z imaging of the cluster region which we have now obtained
using the 4.0-m Blanco Telescope at the
Cerro Tololo Inter-American Observatory. These data are currently being reduced.
Important scientific advances could also potentially be made by improving the
sensitivity of our spectroscopy to fainter luminosities and/or surface brightnesses.
This can be clearly seen by studying Figs.~4 and 5.
Using the CCD data we should be able to define a sample to a considerably fainter
$B$ magnitude limit. The standard 2dF set-up would be adequate to obtain spectra of
the higher surface brightness objects in such a sample. This would include many potential
UCDs. However, even bigger advances will require us to obtain spectra for objects
at fainter surface brightness limits than the present limit of 23.7~B~mag~arcsec$^{-2}$.
This possibility now exists with the commissioning of multi-object spectrographs on
the new generation of 8-m telescopes.
The number of cluster objects found in Field 1 is dwarfed by the number of confirmed background
objects (1175). Such a sample is potentially of great value to studies of the BBD of field
galaxies. Although the present magnitude limit is only slightly lower than that of the 2dFGRS
used by Cross et al. (2001) for their determination of the BBD, the FCSS background sample
has the advantage that the FCSS sources were not pre-selected as galaxies.
Hence, our sample will contain a fairer representation of compact objects in the background
than Cross et al.'s (see Paper~II). For the lower surface brightness objects
in our sample we have also used longer exposure times than the 2dFGRS, so FCSS extends to fainter
surface brightness than 2dFGRS.
However, if the FCSS sample could be pushed to even fainter luminosity
and surface brightness limits then, not only will this enable us to study the cluster BBD better, but
the resulting sample will be ideal for studying the field BBD to fainter surface brightness
limits than yet obtained. For example, Cross et al.'s correlation between luminosity
and surface brightness has only been tested between --24$<M_{B}<$--16.0 and 18.0$<\mu_{e}<$24.5.
At present the FSCC is $\simeq$80 per cent complete to a surface brightness about 0.5~mag fainter than this.
Further improvements in the sensitivity of the spectroscopy will enable us to sample whole
new regions of ($M_{B}$,$\mu_{o}$) space.
\section*{Acknowledgments}
This project would not have been possible without the superb 2dF facility
provided by the AAO and the generous allocations of observing time from
PATT and ATAC. We wish to thank the other members of the FCSS Team:
Jonathon Davies, Elaine Sadler and
Quentin Parker. JBJ, PJB and JHD acknowledge support
from the UK Particle Physics and Astronomy Research Council, and MJD
that from the Australian Research Council. Part of this work was
done at the Institute of Geophysics and Planetary Physics, under the
auspices of the U.S. Department of Energy by Lawrence Livermore
National Laboratory under contract No.~W-7405-Eng-48.
This research has made use of the NASA/IPAC Extragalactic Database (NED) which is
operated by the Jet Propulsion Laboratory, Caltech, under agreement with the
National Aeronautics and Space Administration.
\begin{thebibliography}{99}
\bibitem{b1} Abell G.O., Corwin H.G. (Jr), Olowin R.P., 1989, ApJS, 70, 1
\bibitem{b2} Arp, H., 1965, ApJ, 142, 402
\bibitem{b3} Bertin E., Arnouts S., 1996, A\&AS, 117, 393
\bibitem{b4} Binggeli B., 1994, in Meylan G., Prigniel P., eds,
Proc ESO/OHP Workshop, Dwarf Galaxies. ESO, Garching, p.13
\bibitem{b5} Binggeli B., Sandage A., Tammann G.A., 1985, AJ, 90, 1681
\bibitem{b6} Blanton M.R. et al., 2001, AJ, 121, 2358
\bibitem{b7} Blair M., Gilmore G., 1982, PASP, 94, 742
\bibitem{b8} Bothun G.D., Impey C.D., Malin D.F., Mould J.R., 1987, AJ, 94, 23
\bibitem{b9} Boyce P.J., Phillipps S., 1995, A\&A, 296, 26
\bibitem{b10} Boyce P.J., Phillipps S., Jones J.B., Driver S.P., Smith R.M., Couch W.J., 2001, MNRAS, 328, 277
\bibitem{b11} Bromley B.C., Press W.H., Lin H., Kirshner R.P., 1998, ApJ, 505, 25
\bibitem{b12} Bunclark P, Irwin M.J., 1984, in Capaccioli, M., ed., Proc. IAU Colloq. 78, Astronomy with Schmidt-type Telescopes. Reidel, Dordrecht, p. 147
\bibitem{b13} Caldwell, N., 1987, AJ, 94, 1116
\bibitem{b15} Choliniewski J., 1985, MNRAS, 214, 197
\bibitem{b16} Colless M.M. et al., 2001, MNRAS, 328, 1039
\bibitem{b17} Cross N.M., Driver S.P., 2002, MNRAS, 329, 579
\bibitem{b18} Cross N.M. et al., 2001, MNRAS, 324, 825
\bibitem{b19} Dalcanton J.J., 1998, ApJ, 495, 251
\bibitem{b20} Davies J.I., Phillipps S., Cawson M.G.M., Disney M.J.,
Kibblewhite E.J., 1988, MNRAS, 232, 239
\bibitem{b21} Davies J.I., Phillipps S., Disney M.J., 1990, MNRAS, 244, 385
\bibitem{b22} Deady J.H., 2002, PhD Thesis, Univ. Bristol
\bibitem{b23} de Jong R., Lacey C., 2000, ApJ, 545, 781
\bibitem{b24} De Propris R., Pritchet C.J., Harris W.E., McClure R.D., 1995, ApJ, 450, 534
\bibitem{b25} de Vaucouleurs G., 1959, Handbk. Phys., 53, 120
\bibitem{b26} Disney M.J., 1976, Nature, 263, 573
\bibitem{b31} Drinkwater M.J., Gregg M.D., 1998 MNRAS., 296, L15
\bibitem{b32} Drinkwater M.J., Phillipps S., Gregg M.D., Parker Q.A.,
Smith R.M., Davies J.I., Jones J.B., Sadler E.M., 1999, ApJ, 511, L97 (Paper II)
\bibitem{b33} Drinkwater M.J. et al., 2000a, A\&A, 355, 900 (Paper I)
\bibitem{b34} Drinkwater M.J., Jones J.B., Gregg M.D., Phillipps S., 2000b, Pub. Astron. Soc. Australia, 17, 227 (Paper III)
\bibitem{b35} Drinkwater M.J., Gregg M.D., Holman B.A., Brown M., 2001a, MNRAS, 326, 1076
\bibitem{b36} Drinkwater M.J., Gregg M.D., Colless M.M., 2001b, ApJ, 548, 139
\bibitem{b27} Driver S.P., 1999, ApJ, 526, L69
\bibitem{b28} Driver S.P., Phillipps S., Davies J.I., Morgan I., Disney M.J., 1994, MNRAS, 268, 393
\bibitem{b29} Driver S.P., Couch W.J., Phillipps S., Smith R.M., 1998a, MNRAS, 301, 357
\bibitem{b30} Driver S.P, Couch W.J., Phillipps S., 1998b, MNRAS, 301, 369
\bibitem{b37} Ferguson H.C., 1989, AJ, 98, 367 (FCC)
\bibitem{b38} Ferguson H.C., Sandage A., 1988, AJ, 96, 1520
\bibitem{b39} Folkes S. et. al., 1999, MNRAS, 308, 459
\bibitem{b40} Garilli B., Maccagni D., Andreon S., 1999, A\&A, 342, 408
\bibitem{b41} Hambly N.C., Irwin M.J., MacGillivray H.T., 2001, MNRAS, 326, 1295
\bibitem{b44} Hilker M., Infante L., Vieira G., Kissler-Patig M., Richtler T., 1999, A\&AS, 134, 75
\bibitem{b42} Hodge P.W., 1960, PASA, 72, 188
\bibitem{b43} Hodge P.W., Pyper D.M., Webb C.J., 1965, AJ, 70, 559
\bibitem{b45} Impey C.D., Bothun G.D., 1997, ARA\&A, 35, 267
\bibitem{b46} Irwin M.J., Davies J.I., Disney M.J., Phillipps S., 1990, MNRAS, 245, 289
\bibitem{b47} Irwin M.J., Maddox S., McMahon R., 1994, Spectrum, 2, 14
\bibitem{b48} Jacoby G.H., Hunter D.A., Christian C.A., 1984, ApJS, 56, 257
\bibitem{b50} Kambas A., Davies J.I., Smith R.M., Bianchi S., Haynes J.A., 2000, AJ, 120, 1316
\bibitem{b51} Kormendy J., 1985, ApJ, 295, 73
\bibitem{b52} Kron R.G., 1978, PhD Thesis, Univ. California, Berkeley
\bibitem{b53} Kurtz M.J., Mink D.J., 1998, PASP, 110, 934
\bibitem{b54} Lin H., Yee H.K.C., Carlberg R.G., Ellingson E., 1997, ApJ, 475, 494
\bibitem{b55} Lobo C., Biviano A., Durret F., Gerbal D., Le Fevre O., Mazure, A.,
Slezek E., 1997, A\&A, 317, 385
\bibitem{b56} Loveday J., Peterson B.A., Efstathiou G., Maddox S.J., 1992, ApJ, 390, 338
\bibitem{b57} Marzke R.O., Da Costa L.N., Pellegrini P.S., Willmer C.N.A.,
Geller M.J., 1997, ApJ, 503, 617\bibitem{b} Mayer L., Governato F., Colpi M., Moore B., Quinn T., Wadsley J., Stadel J., Lake G., 2001, ApJ, 559, 754
\bibitem{b58} Mateo M., 1998, AR\&A, 36, 435
\bibitem{b59} Metcalfe N., Shanks T., Fong R., Roche N.,1995, MNRAS, 273, 257
\bibitem{b60} Meyer M.J., Drinkwater M.J., Phillipps S., Couch W.J., 2001, MNRAS, 324, 343 (Paper IV)
\bibitem{b61} Miekse S., Hilker M., Infante L., 2002, A\&A, 383, 823
\bibitem{b62} Moore B., Lake G., Katz N., 1998, ApJ, 495, 139
\bibitem{b63} Morshidi-Esslinger Z., Davies J., Smith R., 1999, MNRAS, 304, 311
\bibitem{b64} Muriel H., Valotto C.A., Lambas D.G., 1998, ApJ, 506, 540
\bibitem{b65} Phillipps S., 1997, in Kontizas E., Konizas M., Morgan D.H., Vettolani G.P.,
eds, Proc Working Group IAU Commission 9, Wide-Field Imaging. Kluwer, Dordrecht, p. 281
\bibitem{b66} Phillipps S., Parker Q.A., 1993, MNRAS, 265, 385
\bibitem{b67} Phillipps S., Driver S.P., 1995, MNRAS, 274, 832
\bibitem{b68} Phillipps S., Disney M.J., Kibblewhite E.J., Cawson M.G.M., 1987, MNRAS, 229, 505
\bibitem{b69} Phillipps S., Davies J.I., Disney M.J., 1988, MNRAS, 233, 485
\bibitem{b70} Phillipps S., Davies J.I., Disney M.J., 1990, MNRAS, 242, 235
\bibitem{b71} Phillipps S., Drinkwater M.J., Gregg M.D., Jones J.B.,
2001, ApJ, 560, 201 (Paper V)
\bibitem{b72} Sandage A., Binggeli B., Tammann G., 1985, AJ, 90, 385
\bibitem{b73} Sersic J.L., 1968, Atlas de Galaxies Australes, Observatorio Astronomico, Cordoba
\bibitem{b80} Sodre L. (Jr), Lahav O., 1993, MNRAS, 260, 285
\bibitem{b81} Smith R.M., Driver S.P., Phillipps S., 1997, MNRAS, 287, 415
\bibitem{b74} Taylor K., Canon R.D., Parker Q.A., 1998, in McLean B.J., Golombek D.A., Hayes J.J.E.,
Payne H.E., eds, Proc. IAUS 179, New Horizons from Multi-Wavelength Sky Surveys. Kluwer, Dordrecht,
p.135
\bibitem{b79} Tonry J.L., Davis M., 1979, AJ, 84, 1511
\bibitem{b75} Trentham N., 1997a, MNRAS, 286, 133
\bibitem{b76} Trentham N., 1997b, MNRAS, 290, 334
\bibitem{b77} Trentham N., 1998, MNRAS, 295, 360
\bibitem{a78} Trentham N., Hodgkin S., 2002, MNRAS, 333, 423
\bibitem{b82} Valotto C.A., Nicotra M.A., Muriel H., Lambas D.G., 1997, ApJ, 479, 90
\bibitem{b83} van der Kruit P., 1987, A\&A, 173, 59
\bibitem{b84} Wilson G., Smail I., Ellis R.S., Couch W.J., 1997, MNRAS, 284, 915
\bibitem{b85} Zucca E. et al., 1997, A\&A, 326, 477
\end{thebibliography}
\label{lastpage}
\end{document}