Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.mso.anu.edu.au/pfrancis/papers/2126-158.ps Äàòà èçìåíåíèÿ: Wed Feb 15 05:04:18 2006 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 03:05:22 2012 Êîäèðîâêà:

Mon. Not. R. Astron. Soc. 000, 000--000 (0000) Printed 31 January 2006 (MN L A T E X style file v2.2)
Multi­object spectroscopy of the field surrounding PKS
2126-158: Discovery of a z = 0.66 galaxy group
Matthew T. Whiting 1,2# and Rachel L. Webster 3 and Paul J. Francis 4
1 School of Physics, University of New South Wales, Sydney, NSW, 2052, Australia
2 present address: Australia Telescope National Facility, P.O. Box 76, Epping, NSW, 1710, Australia
3 School of Physics, University of Melbourne, VIC, 3010, Australia
4 Research School of Astronomy and Astrophysics, Australian National University, ACT, 0200, Australia
31 January 2006
ABSTRACT
The high­redshift radio­loud quasar PKS 2126-158 is found to have a large number
of red galaxies in close apparent proximity. We use the Gemini Multi­Object Spec­
trograph (GMOS) on Gemini South to obtain optical spectra for a large fraction of
these sources. We show that there is a group of galaxies at z # 0.66, coincident with a
metal­line absorption system seen in the quasar's optical spectrum. The multiplexing
capabilities of GMOS also allow us to measure redshifts of many foreground galaxies
in the field surrounding the quasar.
The galaxy group has five confirmed members, and a further four fainter galaxies
are possibly associated. All confirmed members exhibit early­type galaxy spectra, a
rare situation for a Mg ii absorbing system. We discuss the relationship of this group
to the absorbing gas, and the possibility of gravitational lensing of the quasar due to
the intervening galaxies.
Key words: quasars: individual: PKS 2126-158 -- quasars: absorption lines -- galax­
ies: general -- gravitational lensing
1 INTRODUCTION
Quasars located at high redshifts provide excellent probes of
the intervening universe. As their light travels through the
universe, it is partially absorbed by neutral hydrogen or met­
als such as magnesium -- an e#ect detectable in the quasars'
optical and UV spectra. These absorption lines provide an
e#ective way of probing the evolution of metal­bearing gas
over a large range of redshifts, as the sensitivity to absorp­
tion is largely independent of redshift, depending only on
the background source's surface brightness.
The Mg ii absorption systems in particular are most
likely associated with foreground galaxies (Bergeron &
Boisse 1991; Steidel et al. 1994, 2002), based on the veloc­
ity structure within the systems, the clustering properties
of the absorbers, and the presence of metals that have most
likely been produced locally. When identified, most of these
galaxies appear to be spirals (e.g. Steidel et al. 2002), al­
though absorbers have been found that span the range from
late­type spirals to galaxies resembling present­day ellipti­
cals (Steidel et al. 1994). The observational challenge then
is to correctly identify the absorbing galaxies and build up
a picture of the absorbing system. Such identifications al­
# E­mail: Matthew.Whiting@csiro.au
low one to track the evolution of metal­bearing gas and its
relationship to galactic environment.
The identification of the object responsible for an ab­
sorption system involves finding a galaxy close to the line­
of­sight to the quasar that has a redshift matching that of
the absorption system. As such galaxies are often faint, to
examine many in the vicinity of a given quasar requires rel­
atively large amounts of telescope time, or a large degree of
multiplexing. The advent of multi­object spectrographs on
large telescopes has enabled more systems and galaxies to
be examined, greatly assisting the identification of absorbing
systems.
The identification of nearby galaxies close on the sky to
a high­redshift quasar also allows us to address the question
of gravitational lensing. The high redshifts of these quasars
on face value imply very large luminosities, placing them at
the high end of the luminosity function, where the slope of
the function is steepest. If a small amount of magnification
due to gravitational lensing is taking place, the intrinsic lu­
minosity of the quasar will be reduced, changing the shape
of the luminosity function. This is particularly important
for the highest redshift quasars (such as those at z > 6, e.g.
Richards et al. (2004)), but also for those around z # 3 - 4.
In this paper we address these issues by targeting PKS
2126-158, a luminous high­redshift quasar that exhibits a
c
# 0000 RAS

2 M. T. Whiting et al
number of metal­line absorption systems and has a large
number of galaxies close to its line of sight, using the mul­
tiplexing capabilities of GMOS on the Gemini South tele­
scope. We use imaging and low­resolution spectroscopy to
obtain redshifts for many of the galaxies in the field, fo­
cussing on the z < 1 environment.
We describe this quasar and our observations of it and
the surrounding field in Section 2, while the results of the
Gemini/GMOS observations are presented in Section 3. The
implications of these results for the absorption line systems
seen in the quasar's spectrum are discussed in Section 4,
and the possibility of magnification of this quasar due to
gravitational lensing is discussed in Section 5. A summary
of the results is found in Section 6. Note that throughout
this paper we assume a standard #CDM cosmology, with
H0 = 71 km s -1 Mpc -1
,# m = 0.27
and# # = 0.73 (Spergel
et al. 2003).
2 TARGET FIELD AND OBSERVATIONS
2.1 The quasar PKS 2126-158
The quasar PKS 2126-158 is a radio­loud, flat­spectrum
quasar at a redshift of z = 3.2663. It was first identified by
Condon, Hicks & Jauncey (1977) in optical follow­up of flat­
spectrum sources from the Parkes 2700 MHz surveys, and
at the time its redshift was measured it was only the fifth
radio quasar with z > 3 (Jauncey et al. 1978).
PKS 2126-158 is known to be a very bright object
at radio (S2.7 GHz = 1.17 Jy, Wright & Otrupcek (1990)),
optical/near­infrared (V = 16.92, H = 14.89, Francis, Whit­
ing & Webster (2000)) and X­ray (F0.1-2.4 keV = 2 â
10 -12 erg s -1 cm -2 , (Siebert et al. 1998)) frequencies. These
fluxes, combined with its relatively high redshift, mean it is
among the most luminous quasars known.
The optical/near­infrared spectrum of PKS 2126-158
has a blue power­law shape for wavelengths longer than V
band (Francis et al. 2000), where the Ly# line falls. The
flux drops o# at shorter wavelengths due to absorption by
the intervening gas of the Lyman alpha forest. Absorption
systems with strong metal lines are observed at the redshifts
indicated in Table 1. The relatively strong Mg ii system at
z = 0.663 is particularly relevant for this paper, and is dis­
cussed in more detail in Section 4.1.
The peculiar concentration of objects in the vicinity of
PKS 2126-158 was first noticed from Kn­band images taken
on 6 June 1993 by Drinkwater et al. (1997) with IRIS (Allen
et al. 1993) on the Anglo­Australian Telescope in the course
of identification of sources in the Parkes Half­Jansky Flat­
spectrum Sample (PHFS). There are # 16 sources visible
within 30 arcsec of the quasar, principally toward the east
and south.
Veron et al. (1990) obtained spectra of the two bright
sources to the west (designated therein as C1 and C2, at
distances of # 5 and # 10 arcsec west of the quasar respec­
tively). They were able to measure the redshift of C2 (an
emission line galaxy) as z = 0.210, while the spectrum of
C1 was inconclusive. This is the only redshift information
available from the literature for objects in the vicinity of
PKS 2126-158.
Table 1. Metal­line systems observed in the spectrum of PKS
2126-158 (D'Odorico et al. 1998; Giallongo et al. 1993).
z abs Velocity range Metal lines
0.6631 #v # 215km s -1 Mg ii, Mg i, Ca ii
2.3313 C iv, Si ii
2.3941 #v # 180km s -1 C iv, Si iv, Si ii, Al ii
2.4597 C iv, Fe ii
2.5537 C iv
2.6378 #v # 286km s -1 C iv, C ii, Al ii, Al iii, Si ii, Si iii,
Si iv, Mg i, O i, Fe ii
2.6788 C iv, Si iv, C ii, Fe ii, N v
2.7281 C iv, Si iv, Si ii, Fe ii
2.7689 #v # 350km s -1 C iv, Al ii, Si ii, C ii, O i, Si iv,
Al iii, Fe ii
2.8195 C iv, C ii, Si ii, O i
2.9071 C iv, Al ii, Si iii, Si iv
2.9675 Lyman limit system
3.2165 C iv, Si iv, Si ii
2.2 Gemini Observations
We used the Gemini Multi­Object Spectrograph (GMOS,
Hook et al. 2002) on the Gemini South telescope, in Multi­
Object Spectroscopy (MOS) mode, to obtain spectra for as
many of the nearby objects as possible. Pre­imaging of the
field was done with GMOS­South on 2003 July 15. The im­
age, shown in Fig. 1, was constructed from six 300 sec expo­
sures in the i # filter (Gemini filter i_G0327), and has image
quality of 0.7 arcsec (the image was binned 2 â 2 on­chip).
Since GMOS uses multiple slits in its multi­object
mode, one cannot have slits overlapping in the dispersion
direction (the horizontal direction in Fig. 1), particularly
when the target objects are close together on the sky. For
this project, this places a limit on the number of sources
that can be observed in a single pointing (we had time for
just a single mask setting). Fortunately, a judicious choice
of position angle (in this case, 215 # East of North) can en­
able a large fraction of the sources of interest to be placed
on a slit. The large field of view of GMOS (# 5.5 arcmin)
also enabled a number of potentially interesting sources fur­
ther from the quasar to be placed under a slit. These were
selected either on morphology, or on their location in the
field (i.e. whether a slit could be placed over them based
on previously allocated slits). The locations of the targeted
objects are shown in Fig. 1, while Fig. 8 shows each object
in more detail, as well as the location of each slit.
Note that the limiting magnitude of the image is ap­
proximately i # = 24.6. At this magnitude, an L# galaxy 1 is
visible out to a redshift of 5. However, at the magnitude of
the faintest spectroscopic target (i # = 23.33), an L# galaxy
can be seen only out to z = 0.73 (although a 3L# galaxy
survives the K­correction and can be seen out to z # 5).
The field was observed in queue mode on two nights --
2003 September 24 and 25 -- with the chosen slit configura­
tion, using the R150 grating and the GG455 long­pass filter,
for a total exposure time of 2 hours. This was divided into
sets of 3â1200 sec exposures at each of two grating settings,
1 We assume M #
i = 5 log 10 h - 21.00 (Bell et al. 2003), and cal­
culate K­corrections with template SEDs from Poggianti (1997)
c
# 0000 RAS, MNRAS 000, 000--000

Discovery of a z = 0.66 galaxy group 3
Figure 1. An image of the field taken in i # band with GMOS­South. The image has been cropped to a square region containing
all spectroscopic targets. The locations of each of the targets are indicated with their identification numbers used herein. The
orientation of the field is indicated by the arrows at top­left, and o#sets are measured with respect to the quasar PKS 2126-158.
one on each of the nights. The data was binned on­chip by
a factor of 2 in both spatial and dispersion directions.
The image quality, as measured from alignment images
(with the mask in place) taken prior to the spectral observa­
tions, was noticeably worse on the second night (0.5 arcsec
on night one, 0.95 arcsec on night two). This limited the
signal­to­noise achieved in the spectra, and consequently
several sources were unable to have redshifts measured (the
spectra where redshifts were measured typically had S/N
# 5, while these few were significantly less than that).
The data were reduced using standard procedures
from the Gemini IRAF package. The spectra were bias­
subtracted and flat­fielded using calibration frames from the
Gemini GCAL unit. Wavelength calibration using CuAr arc
spectra, as well as sky­subtraction, were performed on an in­
dividual slit basis, and the spectra were flux­calibrated using
observations of the standard star EG131.
3 RESULTS
The measured redshifts for galaxies in the field of PKS
2126-158 are shown in Table 2. The galaxies are num­
bered according to their position on the image, and sorted
in the table in order of decreasing redshift. The i # magni­
tudes shown are measured from the pre­imaging data (i.e.
c
# 0000 RAS, MNRAS 000, 000--000

4 M. T. Whiting et al
Figure 2. Distribution of measured redshifts of galaxies in the
field of PKS 2126-158.
the image shown in Fig 1), calibrated on the Landolt stan­
dard fields SA95 and TPhe, using the transformations from
Fukugita et al. (1996). We have converted these magnitudes
into absolute magnitudes, M i # . The K­corrections were cal­
culated using the template SEDs from Poggianti (1997) and
the filter transmission function for filter i_G0302 provided
on the Gemini web site. 2
The distribution of galaxy redshifts can be seen in Fig 2.
We find 8 galaxies with redshifts in the range 0.66 < z <
0.67, straddling the low redshift metal­line absorption sys­
tems seen by D'Odorico et al. (1998). The spectra of these
galaxies are shown in Fig. 5.
The velocity dispersion of these eight galaxies is # =
430 km s -1 , centred on a redshift of zg = 0.6660. Three of
these galaxies, however, are at large spatial spearations from
those close to the quasar. If we include just the four clos­
est to the quasar -- #17, #18, #19 and #20 -- as well as
#14, which is just close enough to be considered associ­
ated, we get # = 355 km s -1 and zg = 0.6662. This velocity
dispersion is characteristic of a group environment, rather
than a more dense cluster environment which would show
# # 500-1200 km s -1 . It is, in fact, a relatively poor group,
with only five members brighter than M i # = -21. As can
be seen in Fig. 3, there are at least four fainter galaxies for
which we have not obtained spectra that are also candidates
for inclusion in the group (see Sec. 4 for further discussion).
The remaining galaxies at z # 0.66 (#24, #03 and #01)
are part of the same large­scale structure that the group is
embedded in, but not necessarily directly associated with
the group itself.
2 Note that this transmission function is for the equivalent i #
filter at Gemini North, believed to be similar to the one used at
Gemini South. The transmission function for the Gemini South
filter was not available at time of writing.
In addition to this structure, we find several other fea­
tures in redshift­space. There are two other main structures
seen in the field, at z # 0.21 and z # 0.46 (and possibly
z # 0.51), although these are not as tightly grouped on the
sky as the main group at z # 0.66. There are a number of
field galaxies at other redshifts as well. We are thus seeing
several features in the large­scale structure along the line­
of­sight to PKS 2126-158. The spectra of the remaining
galaxies with measured redshifts are shown in Figs. 6 & 7.
Of the remainder, six objects were identified as stars,
and a further five had spectra of too low quality for a redshift
measurement. Of these latter objects, two (#11 and #15)
have extended morphologies in the i # GMOS image, and
i # -K colours redder than all the confirmed group members
(i # - K # 4.7 and 4.0 respectively, cf. i # - K # 3 for the
group members). They are thus either red galaxies in the z #
0.66 group (more likely in the case of #15), or background
galaxies (and redder due to their greater redshift).
4 ASSOCIATION WITH ABSORBING
SYSTEM
4.1 Is one of the identified galaxies the absorber?
As seen in Section 3, we find a concentration of galaxies
in the redshift range 0.66 < z < 0.67. This coincides in
redshift­space with a metal­line absorption system detected
by Giallongo et al. (1993) and D'Odorico et al. (1998). The
system consists of four components in velocity, with three
species detected -- Mg ii and Ca ii, plus Mg i in the two high­
est column density components. The redshifts of the compo­
nents are z = 0.6625, 0.6628, 0.6634 and 0.6637 (D'Odorico
et al. 1998). This system is quite strong: the two strongest
Mg ii components (at z = 0.6628 and z = 0.6625) have
column densities of 4.1 â 10 17 cm -2 and 5.8 â 10 16 cm -2
respectively (Ryabinkov et al. 2003; D'Odorico et al. 1998).
Using the Doppler parameters quoted by D'Odorico et al.
(1998), we can convert these to equivalent widths of 0.44 š A
and 0.41 š A respectively (for the Mg ii #2798 line in the rest
frame).
A key question is whether this system can be associ­
ated with the group of galaxies detected in our GMOS ob­
servations. The nearest of the galaxies to PKS 2126-158
is Galaxy #17, which lies at a distance of 51.2 kpc to the
quasar's line­of­sight (hereinafter QLOS). This places the
QLOS in the outer parts of the galaxy's halo. This separa­
tion is comparable to those seen in other absorption systems
(see, for example, Steidel et al. (1997)). This galaxy, how­
ever, does show a large velocity o#set from the absorbing
system, with velocity di#erences #v # v gal - vabs with re­
spect to the four components ranging from 430-610 km s -1 .
Galaxy #18 provides a much closer match to the velocities
(#v = 74-253 km s -1 ), and is only 61.2 kpc from the QLOS.
It is also the brightest galaxy in the group, with i # = 20.34.
A relationship has been noted by Lanzetta & Bowen
(1990) between the equivalent width W of the Mg ii absorp­
tion system and the impact parameter # of the associated
galaxy, where W # # -0.92±0.16 . By converting the column
density and Doppler parameters of the absorption lines, we
can obtain the equivalent widths for the system: the four
Mg ii #2796 lines from D'Odorico et al. (1998) yield rest­
c
# 0000 RAS, MNRAS 000, 000--000

Discovery of a z = 0.66 galaxy group 5
Table 2. Observational data for galaxies in the field of PKS 2126-158, numbered according to Fig. 1. The i # magnitude is measured
from the GMOS pre­imaging data. The positional o#sets #RA, #Dec and ## are the angular distances from PKS 2126-158, while
the impact parameter # is the linear distance from the galaxy to the line­of­sight to PKS 2126-158. For the comments on the
spectra: PEG = Passive Elliptical Galaxy; ELG = Emission Line Galaxy; AGN = Active Galactic Nucleus.
Object i # M i # z # RA # Dec ## # Comments
[mag] [mag] [arcsec] [arcsec] [arcsec] [kpc]
35 22.46 ­21.33 0.8335 42.1 136.7 143.0 1090.7 Emission line at 6830 š A, ID as [O ii]
20 21.82 ­21.58 0.6704 10.2 4.6 11.2 78.5 PEG
24 21.48 ­21.92 0.6694 47.4 18.3 50.8 356.3 PEG + strong [O ii]
17 20.79 ­22.60 0.6668 6.3 ­3.7 7.3 51.2 PEG
03 20.91 ­22.48 0.6666 ­127.2 ­59.7 140.5 983.3 PEG + weak [O ii]
14 20.88 ­22.51 0.6648 ­14.5 ­19.4 24.2 169.3 PEG
19 22.08 ­21.30 0.6647 12.6 ­2.4 12.8 89.7 PEG
18 20.34 ­23.04 0.6643 8.2 ­3.2 8.8 61.2 PEG
01 22.74 ­20.63 0.6601 ­103.2 ­93.8 139.4 971.9 PEG + weak [O ii]
32 22.11 ­20.75 0.5415 24.4 114.8 117.3 744.4 [O ii], with weak H# and [O iii]
04 21.88 ­20.89 0.5245 ­87.4 ­79.2 118.0 736.0 Strong, extended [O ii] emission
08 21.71 ­21.00 0.5124 ­36.5 ­59.1 69.5 428.5 PEG + [O ii]
09 20.43 ­22.28 0.5122 ­71.3 ­21.3 74.4 458.7 PEG
33 21.88 ­20.63 0.4761 19.3 140.6 141.9 839.9 [O ii]
16 21.60 ­20.90 0.4746 2.2 ­12.7 12.8 75.9 [O ii]
13 20.38 ­22.08 0.4657 ­8.3 ­26.2 27.5 160.7 Starburst (higher­order Balmer lines + [O ii])
06 20.02 ­22.33 0.4478 ­35.5 ­86.9 93.9 536.8 ELG (H#, [O iii], [O ii])
26 23.33 ­18.74 0.4045 ­7.4 78.7 79.0 425.3 ELG (H#, [O iii], [O ii])
07 22.75 ­18.51 0.2994 ­57.8 ­53.7 78.9 348.3 ELG (H#, [O iii], [O ii])
30 19.21 ­21.35 0.2311 92.1 48.1 103.9 379.8 PEG + weak H# & [O iii]
27 18.14 ­22.19 0.2124 ­4.8 82.8 82.9 284.0 Barred spiral
12 19.15 ­21.15 0.2104 ­44.2 ­10.3 45.4 154.4 PEG
29 19.98 ­20.30 0.2086 ­0.4 98.0 98.0 331.1 PEG + weak AGN?
28 17.95 ­20.84 0.1160 ­13.3 98.2 99.1 205.8 PEG
25 18.87 --- 0.0000 ­0.0 57.4 57.4 --- Star
23 17.98 --- 0.0000 55.3 ­0.9 55.3 --- Star
22 22.17 --- 0.0000 36.8 3.1 36.9 --- Star
21 20.49 --- 0.0000 13.2 6.6 14.7 --- Star
05 16.44 --- 0.0000 ­103.1 ­53.9 116.3 --- Star
10 20.66 --- 0.0000 ­74.9 ­7.6 75.3 --- Star
34 22.14 --- --- 39.8 132.7 138.5 --- No redshift measured
31 21.42 --- --- 8.0 114.5 114.8 --- No redshift measured
15 22.29 --- --- 5.6 ­22.1 22.8 --- No redshift measured
11 22.58 --- --- ­20.3 ­36.7 41.9 --- No redshift measured
02 21.61 --- --- ­59.2 ­116.0 130.2 --- No redshift measured
frame values of W = 0.41, 0.44, 0.13 and 0.13 š A (in order of
increasing redshift).
Comparing with the Lanzetta & Bowen (1990) relation­
ship, we find that the two stronger systems have W values
corresponding to # # 30 kpc, while the two weaker systems
correspond to # # 110 kpc. The galaxies we have observed
to be in the group all lie at impact parameters between these
two values. There is, however, an object closer than the clos­
est observed galaxy to the QLOS (# 17), which, should it
prove to be at the same redshift, may be a counterpart to one
of the stronger systems (see Section 4.2 for more discussion
on this source).
If we focus just on the confirmed group members, the
Lanzetta & Bowen (1990) relationship taken at face value
indicates that none of these galaxies are suitable counter­
parts for the absorption system(s). Recent studies, however,
such as that presented in Churchill et al. (2005), indicate
that there is much more scatter in the W - # relationship
than the Lanzetta & Bowen (1990) relationship would sug­
gest. This calls into doubt its utility in making predictions
about the likelihood of potential absorbers. The Churchill
et al. (2005) results suggest that the observed galaxies can
indeed be considered potential hosts of the absorber, even
with their relatively large impact parameters. The identity
of the absorber then is still an open question.
4.2 An alternative absorbing galaxy.
There are three objects closer to PKS 2126-158 than
Galaxy #17. These can be seen in Fig. 3, along with some of
the group members. A spectrum of the object west of PKS
2126-158 was observed by Veron et al. (1990) (their C1).
They were unable to measure its redshift, but its optical
and near­infrared colours are similar to their C2 (the next
galaxy to the west), which they found to be at z = 0.21. It is
therefore likely that this galaxy is foreground to the group
at z # 0.66.
The remaining two nearby objects were not able to be
observed in our GMOS observations (due to the limitations
of slit placement). Object X1 appears to be stellar in mor­
c
# 0000 RAS, MNRAS 000, 000--000

6 M. T. Whiting et al
Figure 3. A 30 ## â30 ## subset of the GMOS i # image surrounding
PKS 2126-158, marked by 'Q'. Objects mentioned in the text are
labeled. North and East are indicated by the arrows. The 5 arcsec
scale bar at bottom is equivalent to 34.9 kpc at the redshift of
the galaxy group.
phology, and has colours i # - K = 1.78 -- not typical of a
galaxy at the same redshift as the group. Similarly, the more
distant (from the QLOS) object X3 has a stellar morphol­
ogy, and even bluer colours than X1 (i # - K = 1.49), and
thus is likely to be a star.
The closest object to the quasar, X2, has an extended,
non­stellar morphology, and redder colours than X1 or X3:
i # -K = 2.97, similar to the colours of the galaxies in the
group at z # 0.66. It may thus be a galaxy in the same sys­
tem. This putative galaxy would then lie at a projected dis­
tance of # 23 kpc from the QLOS, making it a possible can­
didate as a host of the absorbing system. Its impact param­
eter provides a good match to the stronger absorption sys­
tems, if one uses the W-# relationship of Lanzetta & Bowen
(1990). A further point in its favour is its extended, some­
what elongated morphology -- suggestive of a disc galaxy,
matching the morphology of many Mg ii­absorbing galaxies.
4.3 Intra­group gas as an absorber.
Is the absorbing gas necessarily associated with a galaxy?
The fact that we have a group environment means that the
velocity dispersion is comparable to the rotational veloci­
ties of the individual galaxies. Galaxy interactions are then
more likely to strip gas from galaxies and leave it in the
intra­group medium. Such interactions could date from the
formation epoch of the group, or from more contemporary
encounters. An potential example of such an interaction,
based on imaging data, is the pair #17 and #18. The veloc­
ity di#erence between the two (358 km s -1 ), is comparable
to the velocity dispersion of the group, and so is feasible for
an interaction.
A close­up of this pair of galaxies is shown in Fig. 4.
There is evidence for a connection in flux between the two,
Figure 4. Close­up of Galaxies #18 (top) and #17. Orientation
is the same as in Fig. 1. Contour levels are at 4#, 5#, 7#, 11#, 19#
above the mean background. Field of view is 6.13arcsec, and the
direction to PKS 2126-158 is indicated.
with the flux in the ``bridge'' region reaching > 7# above
the background. There is certainly no spectroscopic sign of
any strong interaction taking place, with both galaxies ex­
hibiting simple passive elliptical spectra. However, sensitive
narrow­band imaging may be able to detect H# emission
from ejected gas in the vicinity of the galaxies, and would
prove a good test of this interaction hypothesis.
4.4 Summary and Discussion
PKS 2126-158 is observed to have a strong Mg ii absorption
system at the same redshift as the group of galaxies detected
through our GMOS spectroscopy. All the nearby galaxies
that have spectroscopy are observed to be ellipticals: each
shows a passive elliptical galaxy spectrum, and none shows
evidence for on­going star­formation (for instance, through
the presence of an [O ii] emission line), which would indicate
the presence of large amounts of gas.
Many Mg ii absorbers are identified with spiral galaxies
(e.g. Steidel et al. 2002), although this is by no means an ex­
clusive statement. Steidel et al. (1994) found that Mg ii ab­
sorption could be found in galaxies resembling present­day
ellipticals as well as late­type spirals, Jenkins et al. (2005)
found Mg ii absorption associated with two S0 galaxies, and
recent studies (Churchill et al. 2005) have found a small frac­
tion of elliptical absorbers. The latter studies also indicate
that the absorption is also relatively patchy, and so one (or
more) of the galaxies could have isolated areas of absorbing
gas that happen to fall across the QLOS, but are not them­
selves an important overall part of the galaxy. This may be
similar to the situation observed by Bowen et al. (1995),
who failed to find absorption associated with two elliptical
galaxies despite their small impact parameters.
It is possible that there is intra­group gas responsible
for the absorption, although there is little direct evidence
c
# 0000 RAS, MNRAS 000, 000--000

Discovery of a z = 0.66 galaxy group 7
for this at this stage. The best candidate, however, for the
absorber is probably galaxy X2. It is closer to the QLOS,
and shows a much less uniform morphology than the identi­
fied galaxies. Ultimately, a spectroscopic confirmation of its
redshift is the only way to be sure -- future observations will
hopefully resolve this question.
5 THE LIKELIHOOD OF GRAVITATIONAL
LENSING.
PKS 2126-158 is rather luminous at all frequencies from ra­
dio to X­ray. We would like to know whether it is intrinsically
luminous, or has been magnified through gravitational lens­
ing by some intermediate mass distribution. Strong lensing
can be ruled out, as PKS 2126-158 is not multiply­imaged
-- even at VLBI resolution (# 1 mas), it shows a distinct
core­jet structure (Stanghellini et al. 2001).
Magnification due to gravitational lensing is still possi­
ble without multiple images being detected. There are two
contexts in which this can take place: either a single im­
age is produced, or there are indeed multiple images but
at either faint levels or unresolved separations. Since VLBI
observations place quite strong limits on the existence of
optically­unresolved images, for lensing to be taking place
we would require the production of a single image only.
However, Keeton, Kuhlen & Haiman (2005) find that
this is possible only when the lensing is due to massive
cluster­sized haloes (# 10 13.5 M
# ). Our observations show
that the largest mass concentration along the line­of­sight is
merely a galaxy group. The total luminosity of all galaxies
in the group is # 3 â 10 11 L # (assuming an absolute magni­
tude of the Sun of M i # = 4.45, and taking into account all
possible members, including X2 and the fainter objects in
Fig. 3). For reasonable mass­to­light ratios, the group mass
will be much less than the threshold given by Keeton et al.
(2005). This is supported by the non­detection by Crawford
& Fabian (2003) of any extended X­ray emission in Chandra
images of PKS 2126-158, indicating the lack of hot cluster
gas around or in the foreground of the quasar. We can thus
rule out the possibility that PKS 2126-158 is being magni­
fied by any significant amount. Its large apparent luminosity
is then indeed indicative of its intrinsic power.
6 SUMMARY
We have observed the field surrounding PKS 2126-158 with
Gemini South + GMOS in multi­object spectroscopy mode,
measuring the redshifts of most of the nearby galaxies. We
find a group of galaxies at z # 0.66, at a similar velocity to a
metal­line absorption system seen in the quasar's spectrum.
The group has five confirmed members, and a further four
fainter galaxies are seen nearby and are possibly associated.
There are a further three galaxies close in redshift but at
large separations on the sky -- while not part of the group,
they are certainly part of the same large­scale structure.
We have also made use of the multiplexing capabilities
of GMOS to measure the redshifts of many other galaxies
in the field, and we see several distinct features in redshift­
space foreground to the group. These observations, in show­
ing the lack of a large cluster in front of PKS 2126-158,
indicate that the likelihood the quasar is lensed is small.
While the group as a whole appears to be associated
with the absorption system, the fact that the redshifts of a
number of nearby candidates have not been measured means
some doubt remains as to the identity of the absorber. If the
absorber is associated with the identified galaxies however,
this would be a rare example of an Mg ii absorption system
associated with an early­type galaxy.
ACKNOWLEDGMENTS
We wish to thank the sta# at Gemini South for their help
in obtaining the spectroscopic data, particularly acknowl­
edging the help provided by Michael Ledlow, our contact
scientist until his tragic passing. MTW acknowledges the
assistance provided by a NewSouth Global Fellowship from
the University of New South Wales.
This work is based on observations obtained at the
Gemini Observatory under Gemini Program GS­2003B­Q­
15. The Gemini Observatory is operated by the Associa­
tion of Universities for Research in Astronomy, Inc., un­
der a cooperative agreement with the NSF on behalf of
the Gemini partnership: the National Science Foundation
(United States), the Particle Physics and Astronomy Re­
search Council (United Kingdom), the National Research
Council (Canada), CONICYT (Chile), the Australian Re­
search Council (Australia), CNPq (Brazil), and CONICET
(Argentina).
REFERENCES
Allen D. et al. 1993, PASA, 10, 298
Bell E., McIntosh D., Katz N., Weinberg M., 2003, ApJS,
149, 289
Bergeron J., Boisse P., 1991, A&A, 243, 344
Bowen D., Blades J., Pettini M., 1995, ApJ, 448, 634
Churchill C., Kacprzak G., Steidel C., 2005, in IAU Colloq.
199: Probing Galaxies through Quasar Absorption Lines.
pp 24--41
Condon J., Hicks P., Jauncey D., 1977, AJ, 82, 692
Crawford C., Fabian A., 2003, MNRAS, 339, 1163
D'Odorico V., Cristiani S., D'Odorico S., Fontana A., Gi­
allongo E., 1998, A&AS, 127, 217
Drinkwater M. et al. 1997, MNRAS, 284, 85
Francis P., Whiting M., Webster R., 2000, PASA, 17, 56
Fukugita M., Ichikawa T., Gunn J., Doi M., Shimasaku K.,
Schneider D., 1996, AJ, 111, 1748
Giallongo E., Cristiani S., Fontana A., Trevese D., 1993,
ApJ, 416, 137
Hook I. et al. 2002, Proc. SPIE, 4841, 1645
Jauncey D., Wright A., Peterson B., Condon J., 1978, ApJ,
223, L1
Jenkins E., Bowen D., Tripp T., Sembach K., 2005, ApJ,
623, 767
Keeton C., Kuhlen M., Haiman Z., 2005, ApJ, 621, 559
Lanzetta K., Bowen D., 1990, ApJ, 357, 321
Poggianti B., 1997, A&AS, 122, 399
Richards G. et al. 2004, AJ, 127, 1305
c
# 0000 RAS, MNRAS 000, 000--000

8 M. T. Whiting et al
Figure 5. Reduced GMOS­South spectra of the eight galaxies at redshifts of z # 0.66, listed in order of decreasing redshift. Likely
members of the group are indicated. The flux scale is F # [10 -18 erg s -1 cm -2 š A -1 ], plotted as a function of observed wavelength. Note
that the varying flux scale from spectrum to spectrum. The identities of certain spectral features are indicated, and residuals from
night­sky emission lines at 5577 š A and 6300/6363 š A have been masked out. The noticeable spike at # 5890 š A in the spectrum of Obj#19
is a sky­subtraction artifact, caused by the location of this object close to the edge of a tilted slit (see the image in Fig. 8).
Ryabinkov A., Kaminker A., Varshalovich D., 2003, A&A,
412, 707
Siebert J., Brinkmann W., Drinkwater M., Yuan W., Fran­
cis P., Peterson B., Webster R., 1998, MNRAS, 301, 261
Spergel D. et al. 2003, ApJS, 148, 175
Stanghellini C., Dallacasa D., O'Dea C., Baum S., Fanti
R., Fanti C., 2001, A&A, 377, 377
Steidel C., Dickinson M., Meyer D., Adelberger K., Sem­
bach K., 1997, ApJ, 480, 568
Steidel C., Dickinson M., Persson S., 1994, ApJ, 437, L75
Steidel C., Kollmeier J., Shapley A., Churchill C., Dickin­
son M., Pettini M., 2002, ApJ, 570, 526
Veron P., Veron­Cetty M.­P., Djorgovski S., Magain P.,
Meylan G., Surdej J., 1990, A&AS, 86, 543
Wright A., Otrupcek R., 1990, The Parkes Catalogue
(PKSCAT90). Australia Telescope National Facility, Ep­
ping
c
# 0000 RAS, MNRAS 000, 000--000

Discovery of a z = 0.66 galaxy group 9
Figure 6. Reduced GMOS­South spectra of confirmed galaxies in the field of PKS 2126-158 with z > 0.4, listed in order of decreasing
redshift. The flux scale is F # [10 -18 erg s -1 cm -2 š A -1 ], plotted as a function of observed wavelength. The identities of certain spectral
features are indicated, and residuals from night­sky emission lines at 5577 š A and 6300/6363 š A have been masked out.
c
# 0000 RAS, MNRAS 000, 000--000

10 M. T. Whiting et al
Figure 7. Reduced GMOS­South spectra of confirmed galaxies in the field of PKS 2126-158 with z < 0.4, listed in order of decreasing
redshift. The flux scale is F # [10 -18 erg s -1 cm -2 š A -1 ], plotted as a function of observed wavelength. The identities of certain spectral
features are indicated, and residuals from night­sky emission lines at 5577 š A and 6300/6363 š A have been masked out.
c
# 0000 RAS, MNRAS 000, 000--000

Discovery of a z = 0.66 galaxy group 11
Figure 8. GMOS i # images of individual targets, listed in the same order as in Table 2. The location of the slit used is indicated. Each
image is centred on the object indicated at the top­right. The angular scale, which is the same in each image, is shown in the image of
#35, at top­left.
c
# 0000 RAS, MNRAS 000, 000--000