Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.atnf.csiro.au/people/rnorris/papers/n167.ps Äàòà èçìåíåíèÿ: Mon Dec 6 13:17:58 1999 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 11:06:00 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: m 101

Mon. Not. R. Astron. Soc. 000, 000--000 (0000) Printed 18 October 1999 (MN L A T E X style file v1.4)
Cold gas and star­formation in a merging galaxy
sequence
A. Georgakakis 1 , D. A. Forbes 1 , R. P. Norris 2
1 School of Physics and Astronomy, University of Birmingham, Edgbaston, B15 2TT, UK
2 Australia Telescope National Facility, CSIRO, Epping, NSW,Australia
18 October 1999
ABSTRACT
We explore the evolution of the cold gas (molecular and neutral hydrogen)
and the star­formation activity during galaxy interactions, using a merging
galaxy sequence comprising both pre­ and post­merger candidates. Firstly,
we find that the ratio of far­infrared luminosity to molecular hydrogen mass
(LFIR =M (H 2 ); star­formation efficiency) is, on average, increasing close to
nuclear coalescence. After the merging of the two nuclei there is evidence
that the star­formation efficiency declines again to values typical to ellipti­
cals. This trend is in agreement with previous studies and is attributed to
M (H 2 ) depletion due to interaction induced star­formation. However, there
is significant scatter, likely to arise from differences in the interaction details
(e.g. disk­to­bulge ratio, geometry) of individual systems. Secondly, we find
that the central molecular hydrogen surface density, \Sigma H2 , is also increasing
close to the final stages of the merging of the to nuclei. Such a trend, in­
dicating gas inflows due to gravitational instabilities during the interaction,
is also predicted by numerical simulations. Furthermore, there is evidence
for a decreasing fraction of cold gas mass from early interacting systems to
merger remnants, attributed to neutral hydrogen conversion into other forms
(e.g. stars, hot gas) and molecular hydrogen depletion due to on­going star­
formation. The evolution of the total­radio to blue­band luminosity ratio,
estimating the total (disk and nucleus) star­formation activity, is also investi­
gated. Although this ratio is on average higher than that of isolated spirals, we
find a marginal increase along the merging sequence, attributed to the relative
c
fl 0000 RAS

2 Georgakakis, Forbes & Norris
insensitivity of disk star­formation to interactions. However, a similar result
is also obtained for the nuclear radio emission, although galaxy interactions
are believed to significantly affect the activity (star­formation, AGN) in the
central galaxy regions. Nevertheless, the nuclear­radio to blue­band luminos­
ity ratio is significantly elevated compared to isolated spirals. Finally, we find
that the logarithmic FIR--radio flux ratio distribution of interacting galaxies
is consistent with star­formation being the main energising source.
Key words: Galaxies: mergers -- galaxies: starburst -- radio continuum:
galaxies
1 INTRODUCTION
Tidal interactions and mergers are believed to play a significant role in the evolution of
galaxies. Such phenomena can not only enhance the activity in and around the nuclear re­
gion (star­formation or AGN) but can also irreversibly alter the morphological appearance
of the participant galaxies. Toomre & Toomre (1972) were the first to demonstrate that
gravitational interactions can give rise to tidal features (e.g. bridges, tails) and also pro­
posed the merging of disk galaxies as a plausible formation scenario for ellipticals (called the
merging hypothesis). Indeed, recent more sophisticated numerical simulations have shown
that dynamical friction and violent relaxation during disk­galaxy interactions will disrupt
any pre­existing disks leading to relaxed r 1=4 --law light profiles similar to those of ellipticals
(Barnes 1988, 1992; Hernquist 1992, 1993). The same gravitational instabilities can pro­
duce significant gas inflows towards the centre of the galaxy, where enhanced star­formation
activity is likely to take place.
Indeed, high molecular gas densities have been observed in the central regions of the
IRAS starburst galaxies, thought to be gas rich systems close to the final stages of merging
(Kennicutt 1998; Planesas et al. 1997; Sanders & Mirabel 1996). The high molecular gas
density regions are also found to be associated with enhanced nuclear star­formation (and/or
AGN) activity as inferred from their far­infrared (FIR; Kennicutt 1998), radio (Hummel
1981; Hummel et al. 1990) and optical emission­line luminosity (Keel et al. 1985). A smaller
but systematic enhancement compared to isolated spirals is also seen in the disk radio power
(Hummel 1981) and disk Hff emission (Kennicutt et al. 1987) which is again attributed to
interaction induced star­formation activity. Additionally, the fraction of interacting systems
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 3
found in IRAS­selected samples increases with FIR luminosity (Lawrence 1989; Gallimore
& Keel 1993), suggesting that collisions play a major role in triggering powerful starbursts.
Evidence also exists linking merger remnants with elliptical galaxies. For example, merger
remnants tend to have optical and/or near--infrared light profiles that follow the r 1=4 --law
(Joseph & Wright 1985). Secondly, many, otherwise `normal' ellipticals, exhibit low surface
brightness loops and shells (Malin & Carter 1980; Schweizer & Seitzer 1988) that are likely
to be due to past disk­galaxy encounters. Recently, Forbes, Ponman & Brown (1998) showed
that late stage disk­disk mergers and ellipticals with young stellar populations deviate from
the fundamental plane of ellipticals. This can be understood in terms of a centrally located
starburst induced by a gaseous merger event.
The significance of tidal interactions in the evolution of galaxies has motivated a number
of studies aiming to explore the properties of interacting systems at different stages during
the encounter. Toomre (1977) first proposed a merging sequence of eleven peculiar galaxies
spanning a range of post­ to pre­mergers (the `Toomre sequence') and suggested that the
final product of the interaction is likely to resemble an elliptical galaxy. Keel & Wu (1995)
used morphological criteria to define a merging galaxy sequence by assigning a merger stage
number to each galaxy pair or merger remnant. They found that indicators of on­going
star­formation activity, such as the U \Gamma B, B \Gamma V colours and the FIR--to--blue­band lumi­
nosity ratio tend to peak close to the final stages of nuclear coalescence and then decrease
at post­merger stages to attain values typical of ellipticals. A similar result was obtained by
Casoli et al. (1991) who also studied the evolution of star­formation activity estimators (FIR
temperature, FIR--to--blue­band luminosity, FIR luminosity to molecular hydrogen mass) for
a small merging galaxy sequence defined by morphological criteria. More recently, Gao &
Solomon (1999) used the projected separation between the nuclei of merging FIR­selected
galaxies as an estimator of the interaction stage. They found clear evidence for increasing
star­formation efficiency (SFE; estimated by the ratio of FIR­luminosity to molecular hy­
drogen mass) with decreasing nuclei separation. They argue that this is primary due to the
depletion of the molecular gas reservoirs of these systems by on­going star­formation trig­
gered by interactions. Hibbard & van Gorkom (1996) studied the cold gas properties and
the dynamics of a small sample of pre­ and post­mergers from the Toomre sequence. They
find striking differences in the distribution of H I in pre­ and post­mergers, with increas­
ing fractions of H I outside the optical bodies at later stages. They argue that during the
interaction about half of the gas material is ejected in tidal features, whereas the atomic
c
fl 0000 RAS, MNRAS 000, 000--000

4 Georgakakis, Forbes & Norris
gas remaining within the original disks is either converted into stars or heated up to X­ray
temperatures. Read & Ponman (1998) investigated the X­ray evolution of a similarly defined
merging sequence. Although they also found a rise and fall in the X­ray--to--blue­band lumi­
nosity around nuclear coalescence, the increase is by a factor of ten smaller than that seen
in the FIR--to--blue­band luminosity. They argue that this is likely to be due to superwinds
blowing out the hot X­ray emitting gas. These studies clearly indicate that large changes
occur in the energetic, structural and kinematic properties of galaxies during interactions
and mergers.
The above mentioned studies either concentrated only on pre­mergers (e.g. Gao &
Solomon 1999) or investigated the properties of small samples of pre­ and post­merger galax­
ies (assumed to be representative), albeit in great detail (Hibbard & van Gorkom 1996; Casoli
et al. 1991; Read and Ponman 1998). In this paper we have compiled a large sample of in­
teracting systems from the literature spanning a wide range of pre­ and post­merger stages,
aiming to explore the evolution of both their star­formation and their cold gas (molecular
and neutral hydrogen) properties. Additionally, comparison of the galaxy properties along
the merger sequence with those of `normal' ellipticals and isolated spirals allows investigation
of the merging hypothesis for the formation of ellipticals.
In section 2 we discuss the sample selection, while section 3 describes the radio observa­
tions carried out for a selected sample of interacting galaxies. Section 4 presents the results
from our analysis. Finally, in section 6 we summarise our conclusions. Throughout this paper
we assume a value H o = 75 km s \Gamma1 Mpc \Gamma1 .
2 THE SAMPLE
Compiling a sample of interacting galaxies and merger remnants from the literature is prob­
lematic. Different authors have used different selection criteria (e.g. FIR, morphological
selection) that are likely to introduce biases against certain types of interactions. In this
study, we merged several interacting galaxy samples from the literature, with different se­
lection criteria in an attempt to minimise any selection biases. However, it should be noted
that most of the galaxies in this study are FIR luminous and are also biased against mergers
occurring along our line of sight. Our sample is largely culled from the following studies:
(i) Keel & Wu (1995) selected nearby pre­ and post­merger galaxies based on their optical
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 5
morphology and ordered them into a sequence by assigning a dynamical `stage' number to
each galaxy pair or merger remnant.
(ii) Gao & Solomon (1999) and Gao et al. (1998) compiled samples of FIR­luminous
and ultra­luminous galaxies with available CO(1­0) observations (providing an estimate of
the available molecular hydrogen mass, M(H 2 )). These samples consist exclusively of pre­
mergers.
(iii) Surace et al. (1993) presented a sample of merging galaxies, morphologically selected
from the 60¯m flux density limited IRAS Bright Galaxy Sample (Soifer et al. 1987).
We focus on interacting systems and merger remnants from the above mentioned samples
satisfying the following criteria:
(a) ffir . (D 1 +D 2 )=2, where ffir is the separation between the two nuclei of the merging
galaxies and D 1 , D 2 their major axis diameters.
(b) recessional velocities v . 13 000 km s \Gamma1 , corresponding to a distance . 170 Mpc.
(c) far­infrared luminosities L F IR ! 10 12 L fi .
(d) we only consider disk mergers by discarding pairs for which there is morphological
evidence that at least one of the components is elliptical.
(e) we attempt to restrict our sample to `major mergers'; i.e. mergers involving galaxies
of similar mass, in view of their relevance to the formation of elliptical galaxies. Therefore, we
only consider pairs in which the individual components have B­band magnitude difference of
less than 1.5 mag, corresponding to a mass ratio 1:4 (assuming the same mass­to­light ratio).
Mergers involving higher mass ratios may merely puff­up the disk and/or perhaps enlarge
the bulge, but will not completely rearrange the light profile of the galaxy. However, the
B­band magnitudes are affected by star­formation and do not provide a sensitive estimator
of the total mass of a system. Nevertheless, to the first approximation it should provide a
rough estimate of the galaxy mass that is sufficient for the purposes of this paper.
The sample employed in this study is presented in Table 1, which has the following
format
1. galaxy names
2. heliocentric distance, D, in Mpc, assuming H o = 75 km s \Gamma1 Mpc \Gamma1 . No correction for
the local group velocity or the Virgocentric flow has been applied. However, these corrections
are not expected to modify the estimated distances by more than 10%. Moreover, in our
analysis we consider ratios of observed quantities that are independent of distance.
c
fl 0000 RAS, MNRAS 000, 000--000

6 Georgakakis, Forbes & Norris
3. total radio flux density at 1.4 GHz (20 cm; S tot
1:4 ) in mJy. For most galaxies in the
present sample S tot
1:4 was obtained from Condon et al. (1991) and from the NRAO VLA Sky
Survey (NVSS) catalogue (Condon et al. 1998).
4. galaxy `age' parameter. Each galaxy is assigned an `age' parameter, relative to the time
of the merging of the two nuclei. Negative `ages' are for pre­mergers while positive `ages' cor­
respond to merger remnants. For pre­mergers the `age' is estimated by dividing the projected
separation of the two nuclei, ffir, by an (arbitrary) orbital decay velocity v = 30 km s \Gamma1 . It
is clear that the `age' parameter for pre­mergers is affected by projection effects or different
interaction geometries. However, to the first approximation, it provides an estimate of the
stage of the merging and allows plotting of pre­ and post­mergers on the same scale. For
post­mergers, the `age' is calculated by multiplying the dynamical stage number, defined by
Keel & Wu (1995), by the factor 4 \Theta 10 8 yr. This conversion factor is found to be appropriate
for the 3 merger remnants in the Keel & Wu sample with available spectroscopic estimates
(i.e. NGC 2865, NGC 3921, NGC 7252; Forbes, Ponman & Brown 1998).
5. far­infrared luminosity in solar units (L fi = 3:83 \Theta 10 26 W)
L F IR = 4ú D 2 \Theta 1:4 \Theta S F IR ; (1)
where S F IR is the FIR flux in Wm \Gamma2 between 42:5 and 122:5¯m (Sanders & Mirabel 1996)
S F IR (Wm \Gamma2 ) = 1:26 \Theta 10 \Gamma14 \Theta (2:58 \Theta f 60 + f 100 ); (2)
where f 60 and f 100 are the IRAS fluxes at 60 and 100¯m respectively in Jansky. The scale
factor 1.4 in equation (1) is the correction factor required to account principally for the
extrapolated flux longward of the IRAS 100¯m filter (Sanders & Mirabel 1996).
6. molecular hydrogen mass, M(H 2 ), estimated from the CO(1­0) emission. The sources
from which the CO(1­0) intensity measurements were obtained are given in Table 1. The
conversion factor N(H 2 )=I CO = 3 \Theta 10 20 cm \Gamma2 (Kkm s \Gamma1 ) \Gamma1 , appropriate for molecular clouds
in the Milky Way (Sanders, Solomon & Scoville 1984) was adopted. It should be noted that
use of this conversion factor assumes that the mean properties of the molecular gas in distant
galaxies (i.e. density, temperature and metalicity) are similar to those of the Milky Way.
However, the molecular clouds of the interacting systems studied here are likely to have both
higher densities and temperatures and different metalicities compared to those of the Milky
Way. These effects are expected to increase the CO­to­M(H 2 ) conversion factor for these
galaxies (Tinney et al. 1990). Nevertheless, to facilitate comparison of our results with other
studies we use the standard Galactic N(H 2 )=I CO conversion factor. In any case the results
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 7
can be interpreted in terms of CO luminosity (LCO ) rather than M(H 2 ), since a constant
scaling factor is used throughout.
7. neutral hydrogen mass, M(HI). The H I masses are related to the H I integrated
intensities, F (H I) (measured in Jy km s \Gamma1 ), by
MH I (M fi ) = 2:356 \Theta 10 5 \Theta F (H I) \Theta (D=Mpc) 2 : (3)
The sources from which the H I intensity measurements were obtained are also given in Table
1.
8. total B­band magnitude, B T . This has been corrected for Galactic extinction but not
for internal extinction. This is because the systems studied here have disturbed morphologies
and are likely to have more dust than normal spirals. Therefore, any correction for internal
extinction which is based on the galaxy morphology (like that introduced in the RC3 cat­
alogue by de Vaucouleurs et al. 1991) is expected to be unreliable. In few cases, RC3 did
not provide total B­band magnitudes and instead we used the total magnitudes from the
catalogue compiled by Garnier et al. (1996).
9. Central surface density of molecular hydrogen, \Sigma H 2
, in units of M fi pc \Gamma2 . This is
calculated from high resolution CO(1­0) observations by dividing the (unresolved) flux within
the synthesised beam by its area (in pc 2 ). A consequence of this definition is an increase of the
linear size the region enclosed by the synthesised beam with distance. However, to the first
approximation, \Sigma H 2
is representative of the molecular gas concentration of different systems.
The \Sigma H 2
for interacting systems and merger remnant candidates were mainly obtained from
Kennicutt (1998) and Planesas et al. (1997).
10­12. Central radio flux density at 1.49 GHz (20 cm), 4.79 GHz (6 cm) and 8.44 GHz
(3 cm) respectively, integrated within an aperture of ú 2 kpc diameter at the distance of the
galaxy. We use high resolution radio maps available on NED (1.49 GHz: Condon et al. 1990;
8.44 GHz: Condon et al. 1991) as well as our own radio observations at 4.79 and 8.64 GHz
(see section 3). Upper limits are 3oe estimates, where oe is the RMS noise within a beam.
13. Radio spectral index, ff, derived from the central 2 kpc diameter aperture flux den­
sities using the relation
ff = log(S š 1
=S š 2
)= log(š 1 =š 2 ) (4)
c
fl 0000 RAS, MNRAS 000, 000--000

8 Georgakakis, Forbes & Norris
Table
1:
The
sample
NAMES
D
S
tot 1:4
age
LFIR
M(H2)
a
M(HI)
b
BT
\SigmaH
2
S
cen 1:49
S
cen 4:85
S
cen 8:44
ff
(Mpc)
(mJy)
(\Theta10
8
yr)
(10
9
\Theta
Lfi)
(10
8
\ThetaM
fi)
(10
8
\ThetaM
fi)
(mag)
(Mfi
pc
\Gamma2
)
(mJy)
(mJy)
(mJy)
NGC
520,
ARP
157
30.4
179.9
­1.92
65.0
102.0
1
70.0
1
12.19
3.81
NGC
1614,
ARP
186
63.7
115.0
­0.65
260.0
112.0
1
29.0
2
13.41
3.72
82.0
NGC
2418,
ARP
165
67.4
8.80
!2.9
13.05
NGC
2623,
ARP
243
73.8
98.5
1.60
260.0
55.4
1
?2.7
2
13.91
2.87
84.1
32.1
­0.55
NGC
2865
35.0
!0.4
12.00
!0.6
11.0
3
12.30
NGC
2911,
ARP
232
42.4
58.6
12.00
1.3
19.0
4
12.42
NGC
2914,
ARP
137
42.0
7.20
!0.9
5.2
5
14.03
NGC
3256
c;d
34.0
642.0
­0.29
220.0
117.0
2
11.60
3.13
65.0
39.0
­0.85
NGC
3256A
c;d
34.0
31.0
18.0
­0.91
NGC
3256B
c;d
34.0
34.0
21.0
­0.81
NGC
3303,
ARP
192
83.8
5.0
­1.27
5.5
!3.7
3
46.0
6
14.03
NGC
3414,
ARP
162
18.9
4.0
11.20
!0.3
1.0
7
11.96
NGC
3509,
ARP
335
102.7
27.9
3.20
36.0
210.0
8
13.35
NGC
3690,
MRK
171
41.6
658.0
­1.44
360.0
140.0
4
?45.0
9
11.98
4.05
250.2
80.6
­0.64
NGC
3690A
41.6
56.2
15.0
­0.75
NGC
3690B
41.6
194.0
80.6
­0.5
NGC
3921,
ARP
224
77.8
9.4
6.40
!13.0
35.0
5
51.0
1
13.06
2.13
NGC
4038/9,
ARP
244
21.9
572.0
­2.55
46.0
62.6
2
69.0
10
10.19
3.20
NGC
4038
21.9
255.0
10.86
NGC
4039
21.9
206.0
11.03
NGC
4194,
ARP
160
33.4
102.0
0.80
50.0
12.4
1
13.01
73.9
NGC
4676,
ARP
242
88.1
30.5
­3.76
52.0
87.2
6
68.0
1
12.87
2.53
NGC
4676A
88.1
13.82
NGC
4676B
88.1
13.46
IC
883,
ARP
193
93.3
103.6
­0.10
310.0
67.5
1
?61.0
11
14.40
3.95
70.5
28.6
­0.51
Mrk
273,
UGC
08696
151.0
143.0
­0.24
970.0
168.0
1
15.15
4.57
59.2
52.4
­0.06
NGC
6052,
ARP
209
62.9
94.0
­0.52
57.0
82.3
7
90.0
12
13.34
NGC
6052A
62.9
NGC
6052B
62.9
NGC
6090,
UGC
10267
117.0
46.4
­1.34
200.0
128.0
1
100.0
13
15.74
3.48
15.2
2.8
­0.97
NGC
6090A
117.0
11.2
2.8
­0.79
NGC
6090B
117.0
4.0
NGC
6240,
UGC
10592
97.9
427.2
­0.32
460.0
184.0
1
86.0
11
13.51
4.11
NGC
7252,
ARP
226
62.5
!128.7
9.60
37.0
36.7
8
35.0
1
12.54
2.61
NGC
7585,
ARP
223
46.0
13.60
0.7
!67.0
14
12.22
NGC
7592,
VV
731
97.7
69.2
­1.93
160.0
182.0
1
120.0
2
13.95
32.3
NGC
7592A
97.7
32.9
21.9
NGC
7592B
97.7
20.3
10.4
NGC
7727,
ARP
222
24.7
3.4
12.80
0.6
1.0
9
5.0
4
11.41
NGC
3597
46.5
67.8
­0.28
60.0
25.0
10
13.39
ESO
286­IG19
d
171.2
31.0
6.00
690.0
148.0
11
!35.0
15
14.65
10.5
8.2
­0.41
NGC
1487
d
11.3
­0.46
!1.0
0.1
12
13.0
15
12.34
1.0
!0.9
!­0.17
ESO
341­IG04
d
80.8
!2.5
12.00
19.0
24.3
12
40.0
15
13.27
0.2
!1.4
!3.29
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 9
Table
1:
continued
NAMES
D
S
tot 1:4
age
LFIR
M(H2)
a
M(HI)
b
BT
\SigmaH
2
S
cen 1:49
S
cen 4:85
S
cen 8:44
ff
(Mpc)
(mJy)
(\Theta10
8
yr)
(10
9
\Theta
Lfi)
(10
8
\ThetaM
fi)
(10
8
\ThetaM
fi)
(mag)
(Mfi
pc
\Gamma2
)
(mJy)
(mJy)
(mJy)
ESO
034­IG11
d
88.1
1.80
32.0
46.8
13
27.9
16
13.27
0.2
0.4
1.17
NGC
6769/70
d
50.8
­9.18
20.0
11.75
1.7
1.0
­0.89
NGC
6770
d
50.8
12.66
1.7
1.0
­0.89
NGC
6769
d
50.8
12.55
NGC
7764A
d
122.2
­9.28
30.33
NGC
7764A
NED2
d
122.2
15.35
2.5
2.0
­0.37
NGC
7764A
NED1
d
122.2
14.98
0.3
!1.5
!2.72
ESO
138­IG29
d
61.8
ú1.00
12.43
0.2
!1.5
!3.41
ESO
138­IG30
d
61.8
13.61
ARP
302,
UGC
09618
134.7
72.4
­8.85
310.0
768.0
14
?220.0
12
14.01
2.60
16.1
3.8
­0.82
ARP
302A
134.7
670.0
14
15.17
ARP
302B
134.7
110.0
14
14.37
NGC
6670,
UGC
11284
115.3
66.6
­4.87
267.0
550.0
4
130.0
13
14.98
2.48
UGC
2369
124.7
50.0
­4.28
268.0
340.0
4
?19.0
17
3.08
34.8
8.8
­0.78
UGC
2369A
124.7
28.5
UGC
2369B
124.7
6.3
ARP
055,
UGC
04881
157.1
31.6
­3.50
350.0
269.0
1
190.0
2
14.63
3.11
10.7
7.3
­0.21
VV
114,
IC
1623
80.2
221.0
­1.96
310.0
510.0
15
50.0
13
14.27
3.81
NGC
5256,
MRK
266
111.4
159.0
­1.57
210.0
210.0
16
13.89
?3.11
45.2
NGC
5256A
111.4
21.6
NGC
5256B
111.4
23.6
IRAS
01077­1707
133.9
41.3
­7.42
270.0
400.0
4
14.87
14.6
IRAS
01418+1651
109.7
41.7
­1.44
300.0
69.2
1
?17.0
12
15.97
32.8
16.3
­0.39
IRAS
02114­0456
120.8
47.4
­1.70
180.0
258.0
4
100.0
2
14.70
IRAS
03359+1523
141.5
19.4
­2.25
250.0
242.0
1
15.82
IC
2545
136.3
13.6
­1.08
320.0
105.0
17
110.0
18
14.76
IRAS
13001­2339
85.9
59.8
­0.85
210.0
73.2
11
14.49
UGC
8335,
ARP
238
123.2
53.1
­4.08
330.0
123.0
1
!29.0
2
14.20
39.4
16.7
­0.48
UGC
8335A
123.2
8.5
33.5
UGC
8335B
123.2
44.6
5.9
IC
4395,
UGC
09141
146.0
29.0
­1.37
150.0
191.0
4
65.0
13
14.67
MRK
848,
VV
705
160.7
46.8
­1.57
490.0
335.0
4
15.20
?3
19.0
10.6
­0.33
MRK
848A
160.7
29.9
6.9
2.4
­0.6
MRK
848B
160.7
18.0
12.1
8.2
­0.21
II
Zw
96
144.3
36.3
­2.52
520.0
287.0
4
120.0
19
NGC
835/3,
ARP
318
54.3
36.1
­5.16
45.0
75.7
17
32.0
13
12.47
NGC
835
54.3
60.2
18
12.91
NGC
833
54.3
15.5
18
13.69
NGC
2799/8,
ARP
283
24.1
71.6
­3.66
25.0
32.5
19
25.0
20
12.74
55.7
NGC
2798
24.1
28.6
19
12.0
20
13.04
NGC
2799
24.1
3.9
19
13.0
20
14.32
NGC
3395/6,
ARP
270
21.6
108.5
­2.87
9.7
10.2
20
30.0
20
11.74
14.0
NGC
3395
21.6
80.4
12.40
8.5
c
fl 0000 RAS, MNRAS 000, 000--000

10 Georgakakis, Forbes & Norris
Table
1:
continued
NAMES
D
S
tot 1:4
age
LFIR
M(H2)
a
M(HI)
b
BT
\SigmaH
2
S
cen 1:49
S
cen 4:85
S
cen 8:44
ff
(Mpc)
(mJy)
(\Theta10
8
yr)
(10
9
\Theta
Lfi)
(10
8
\ThetaM
fi)
(10
8
\ThetaM
fi)
(mag)
(Mfi
pc
\Gamma2
)
(mJy)
(mJy)
(mJy)
NGC
3396
21.6
28.1
12.60
5.5
NGC
4568/7,
VV
219
30.1
136.0
­3.72
46.0
166.0
19
54.0
21
11.09
NGC
4468
30.1
112.0
19
11.67
NGC
4467
30.1
53.5
19
12.05
NGC
5257/8,
ARP
240
90.6
75.5
­11.17
200.0
130.0
2
12.68
7.2
NGC
5257
90.6
13.35
2.7
NGC
5258
90.6
13.52
4.5
NGC
5427/6,
ARP
271
34.9
63.0
­7.97
34.0
47.0
20
110.0
6
11.56
NGC
5257
34.9
11.85
NGC
5258
34.9
12.60
NGC
5929/30,
ARP
90
33.2
108.0
­1.57
22.0
11.9
16
11.0
22
12.94
109.0
22.1
­1.36
NGC
5929
33.2
70.6
13.54
51.7
22.1
­0.72
NGC
5930
33.2
37.4
14.04
57.3
NGC
5953/4,
ARP
91
28.8
72.3
­1.93
21.0
27.0
16
15.0
23
12.49
40.2
NGC
5953
28.8
59.3
13.21
32.7
NGC
5954
28.8
13.0
13.61
7.5
NGC
4922,
UGC
08135
98.1
37.8
­3.72
120.0
!14.0
12
14.05
30.7
NGC
5331,
UGC
08774
132.1
19.3
­6.30
240.0
!220.0
12
14.00
9.4
NGC
5018
e
37.3
3.1
15.20
3.4
!6.0
21
6.2
24
11.47
a
CO
references
1:
Sanders
et
al.
1991;
2:
Aalto
et
al.
1995;
3:
Elfhag
et
al.
1996;
4:
Gao
&
Solomon
1999;
5:
Yun
&
Hibbard
1999;
6:
Casoli
et
al.
1991;
7:
Sage
et
al.
1993;
8:
Andreani
et
al.
1995;
9:
Crabtree
et
al.
1994;
10:
Wiklind
et
al.
1995;
11:
Mirabel
et
al.
1990;
12:
Horellou
et
al.
1997;
13:
Horellou
et
al.
1995;
14:
Lo
&
Gao
1997;
15:
Yun
et
al.
1994;
16:
Maiolino
et
al.
1997;
17:
Kazes
et
al.
1990;
18:
Boselli
et
al.
1996;
19:
Young
et
al.
1995;
20:
Solomon
&
Sage
1988;
21:
Huchtmeier
1995
b
HI
references
1:
Hibbard
et
al.
1996;
2:
Bushouse
et
al.
1987;
3:
Schiminovich
et
al.
1994;
4:
Roberts
et
al.
1991;
5:
Bottinelli
et
al.
1980;
6:
Theureau
et
al.
1998;
7:
Huchtmeier
1997;
8:
Heckman
et
al.
1983;
9:
Stanford
et
al.
1989;
10:
Huchtmeier
et
al.
1975;
11:
Richter
et
al.
1994;
12:
Mirabel
&
Sanders
1988;
13:
Martin
et
al.
1991;
14:
Eskridge
et
al.
1991;
15:
Horellou
et
al.
1997;
16:
Higdon
&
Wallin
1997;
17:
Haynes
et
al.
1997;
18:
Kazes
et
al.
1990;
19:
Giovanelli
&
Haynes,
1993;
20:
Davis
&
Seaquist
1983;
21:
Cayatte
et
al.
1990;
22:
Richter
et
al.
1991;
23:
Martin
1998;
24:
Huchtmeier
et
al.
1995
c
High
resolution
radio
data
from
Norris
&
Forbes
1995
d
The
central
radio
flux
densities
in
columns
11
and
12
are
estimated
at
frequencies
4.79
and
8.64GHz
respectively
e
The
age
assigned
to
NGC
5018
has
been
estimated
spectroscopically
(Terlevich
&
Forbes
1999)
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 11
3 NEW RADIO OBSERVATIONS
New radio observations of seven interacting systems (NGC 6769/70, ESO 286­IG19, NGC
1487, ESO 034­IG11, NGC 7764A, ESO 138­IG29, ESO 341­IG04) were carried out using the
Australia Telescope Compact Array at 4.79 GHz (6 cm) and 8.64 GHz (3 cm) simultaneously.
We alternated observations of the target galaxy and a phase calibrator throughout the
observing run. Our observational parameters are given in Table 2. The same amplitude
calibrator (1934--638) was observed with each galaxy, with an assumed flux density of 5.83 Jy
at 4.79 GHz and 2.84 Jy at 8.64 GHz. The data were edited, calibrated and CLEANed using
the AIPS software package. The typical half--power beam--width (HPBW) of the final images
are ú 2 arcsec at 4.79 GHz and ú 1 arcsec at 8.64 GHz.
Because these observations and their analysis were optimised for studying the nuclear
region, they are relatively insensitive to extended emission. We will therefore concentrate on
the emission within the central 10 arcsec. Details on individual galaxies are given in Appendix
A. Also shown in Appendix A are the radio contours at 4.79 GHz overlayed ? on the optical
images (from Digital Sky Survey) of the seven galaxies in Table 2 (Figures A1­A7).
It should be noted that ESO 034­IG11 and ESO 138­IG29 are both ring galaxies. In­
teractions that produce a ring system are somewhat different to the disk­galaxy interac­
tions/mergers studied here. In this case the intruder galaxy makes a rapid near perpendicu­
lar approach to the disk of the primary galaxy. Unlike more planar interactions, the disk is
little effected until the intruder passes through it. Such collisions produce merger remnants
that are different from those resulting from disk­galaxy encounters and therefore we consider
them separately from the main sample. In all the figures, ESO 034­IG11 and ESO 138­IG29
systems are plotted as post­mergers, although we differentiate them from the rest of the
merger remnants with different symbols.
4 RESULTS AND DISCUSSION
4.1 Star­formation efficiency
The star­formation efficiency (SFE) is defined as the ratio of FIR luminosity to molecular
hydrogen mass, L F IR =M(H 2 ) (Gao & Solomon 1999). The SFE is related to the number
of massive stars formed per molecular cloud and provides an estimate of the on­going star­
? These images were produced using KVIEW application from the Karma software suite (Gooch 1995)
c
fl 0000 RAS, MNRAS 000, 000--000

12 Georgakakis, Forbes & Norris
Table 2. Observational parameters
Galaxy Obs. Phase 3cm 3cm 3cm 6cm 6cm 6cm
Date Calibrator Beam P.A. Noise Beam P.A. Noise
(arcsec 2 ) (deg) (¯Jy) (arcsec 2 ) (deg) (¯Jy)
NGC 6769/70 1995 Dec. 10 1925--610 1.08 \Theta 0.96 29.0 30 2.02 \Theta 1.80 31.0 30
ESO286--IG19 1995 Dec. 10 2058--425 1.33 \Theta 1.03 13.4 50 2.47 \Theta 1.87 9.8 40
NGC 1487 1995 Dec. 11 0355--483 1.48 \Theta 0.93 --15.1 30 2.84 \Theta 1.71 --16.3 30
ESO034--IG011 1995 Dec. 12 0757--737 1.31 \Theta 0.87 --66.5 30 2.37 \Theta 1.55 --73.4 30
NGC 7764A 1998 Jan. 28 0008--421 1.80 \Theta 1.17 18.9 50 3.25 \Theta 2.11 18.0 42
ESO138--IG29 1998 Jan. 24 1718--649 1.44 \Theta 1.10 87.8 50 2.56 \Theta 1.99 87.1 46
ESO341--IG04 1998 Jan. 10 2106--413 1.84 \Theta 1.27 27.4 47 3.38 \Theta 2.22 23.4 37
formation efficiency. Figure 1 plots the SFE as a function of the galaxy `age' parameter. As
explained in section 2 negative `ages' correspond to pre­merger stages, zero corresponds to
nuclear coalescence, while positive `ages' are for merger remnants. It is clear from Figure
1 that there is a trend of increasing SFE as the interaction progresses towards the final
stages of nuclear coalescence. At later stages, the situation is less clear due to the small
number of systems with available CO measurements. Nevertheless, there is evidence that
throughout the merger process the SFE starts at a level comparable to isolated spirals, peaks
around nuclear coalescence and decreases at post­mergers to a level similar to that of normal
ellipticals.
It is also clear from Figure 1 that there is significant scatter in the SFE evolution of
pre­mergers. This is likely to be partly due to projection effects. We attempt to compensate
for these effects by averaging the data for pre­mergers in different `age' bins. The bins have
variable widths so that they all comprise similar number of points (ú 10). The mean SFE
and the standard error is then calculated within each bin. The results are also shown in
Figure 1. It is clear that the mean SFE increases toward the final coalescence of the two
nuclei.
However, the observed scatter in the SFE evolution of pre­mergers is also expected to be
due to differences in the details of individual interactions (e.g. geometry, initial conditions,
bulge­to­disk ratio). Mihos & Hernquist (1996) carried out numerical simulations to explore
the effect of galaxy structure on the gas dynamics and evolution of starburst activity in
mergers. They found that in the case of galaxies with dense central bulges significant gas
inflows occur close to the final stages of merging. On the contrary, gas inflows and thus, the
peak of star­formation in bulgless galaxies occur earlier in the interaction. As a result some
of the gas in these systems is depleted at early stages and only a relatively weak starburst
is expected during nuclear coalescence. As demonstrated in Figure 2, where we plot SFE
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 13
against \Sigma H 2
, gas inflows and the resulting high central molecular gas surface density (\Sigma H 2
),
appears to be associated with enhanced SFE. Unfortunately, morphological information for
the pre­merger systems in the present sample is sparse, with most of them classified as
peculiars or irregulars. Therefore, without further data it is difficult to explore trends in the
SFE evolution with bulge­to­disk ratio. Moreover, the orbital dynamics of the encounter also
play a role, albeit a modest one, in regulating the gas inflow and therefore the peak of star­
formation activity (Mihos & Hernquist 1996). In particular, prograde encounters produce gas
inflows at early stages, as opposed to retrograde ones, where the gas dissipation occurs close
to the final stages of the interaction. Since interacting galaxies may have a range of bulge­
to­disk ratios and different orbital dynamics, a scatter is expected in the evolution of their
SFEs. The fact that the SFE of the present sample peaks close to nuclear coalescence (despite
the scatter) indicates that systems with late gas inflows (i.e. bulge dominated galaxies in the
Mihos & Hernquist (1996) scenario) are likely to be over­represented in our sample. A similar
result is obtained by Gao & Solomon (1999) and Gao et al. (1998) who studied the SFE
as a function of nuclear separation for FIR luminous galaxies (some of which overlap with
the present sample). Therefore, it is likely that the FIR selection biases the sample towards
systems with late gas inflows. Although the present sample also comprises morphologically
selected galaxy pairs, most of the pre­merger galaxies are FIR­luminous.
The observed increase in the SFE is mainly due to the decreasing mass of available
M(H 2 ) as the interaction progresses to advanced stages (Gao & Solomon 1999). This is also
demonstrated in Figures 3 and 4 which show the FIR luminosity and molecular hydrogen
mass of interacting systems respectively as a function of the `age' parameter. It is clear from
these figures that there is no strong correlation between L F IR and `age' for pre­mergers,
although M(H 2 ) and `age' are roughly anti­correlated, albeit with significant scatter. The
molecular hydrogen depletion is mainly due to the increased star­formation rate triggered
off by the interaction process, converting the existing giant molecular clouds into stars (Gao
& Solomon 1999). At later, post­merger times, the FIR luminosity decreases as the star­
formation declines, while the molecular hydrogen is further depleted but at a slower rate
(Figures 3 and 4). Consequently, the SFE for merger remnants declines.
Also shown in Figure 4 are the typical SFEs for isolated spirals (Solomon & Sage 1988)
and normal ellipticals (Lees et al. 1991). For ellipticals, the presence of few systems with
high SFE in the Lees et al. (1991) sample biases the mean to large values. A more robust
estimator of the central value of a distribution with a long tail is the median value. Addi­
c
fl 0000 RAS, MNRAS 000, 000--000

14 Georgakakis, Forbes & Norris
tionally, the presence of M(H 2 ) upper limits in the Lees et al. (1991) sample requires the
use of survival analysis to estimate statistical quantities. The median SFE for ellipticals is
therefore, estimated using the ASURV Rev. 1.2 code (Feigelson, Isobe & LaValley 1992),
which implements the methods presented in Feigelson & Nelson (1985).
Early, well separated interacting systems have SFE comparable to that of isolated spi­
rals, suggesting that these systems are in a pre­starburst stage (Lo, Gao & Gruendl 1997).
Additionally, there is also evidence that the SFE of merger remnants and `normal' ellipticals
form a continuous decreasing sequence. However, CO(1­0) observations of a larger sample
of merger remnant candidates is needed to further explore their association with ellipticals.
Casoli et al. (1991) and Hibbard & van Gorkom (1996) also studied the evolution of the SFE
of a small sample of pre­ and post­merger. They found that post­mergers have SFE that
closely resembles that of `genuine' ellipticals, although they are relatively rich in cold gas
(molecular and neutral hydrogen) compared to E/S0s. Nevertheless, Hibbard & van Gorkom
(1996) argue that these merger remnants are likely to rid a large fraction of their gas within
few Gyr, mainly due to modest on­going star­formation.
In Figure 1 there is evidence that the ring galaxy ESO 034­IG11 has lower SFE com­
pared to merger remnant candidates produced by major disk­galaxy interactions. Horellou
et al. (1995) also found a mean SFE = 16 \Sigma 10 L fi =M fi for ring galaxies, suggesting that
the star­formation activity in ring systems declines faster after the close approach of the in­
truder galaxy compared to disk­galaxy mergers. Alternatively, this might indicate that the
encounters that give rise to ring systems do not produce as powerful starbursts as disk­galaxy
mergers. Indeed, the star­formation activity in ring systems is mainly restricted in the ring
(Higdon & Wallin 1997) due to the propagation of density waves, rather than the nuclear
region in the case of disk­galaxy mergers. Therefore, the density waves triggered by head­on
collisions are less likely to produce the high concentrations of molecular hydrogen found in
the nuclear regions of the systems resulting from disk­galaxy encounters. As demonstrated in
Figure 2, such high concentrations of molecular hydrogen are also associated with powerful
starbursts.
4.2 Molecular hydrogen surface density
The nuclear surface density of molecular hydrogen, \Sigma H 2
, is plotted as a function of the
galaxy `age' parameter in Figure 5. There is evidence for increasing \Sigma H 2
as the interaction
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 15
Figure 1. Star­formation efficiency (SFE = L FIR =M(H 2 )) as a function of the galaxy `age' parameter. Negative `ages' are for
pre­mergers while positive `ages' correspond to post­mergers. The dashed line signifies the time of nuclear coalescence (`age'=0)
and separates pre­ and post­merger systems. Open circles are the galaxies in the present sample. The filled circle represents
the ring galaxy ESO 034­IG11. Filled squares correspond to the mean SFE for (i) isolated spirals (left; Solomon & Sage 1988)
and (ii) ellipticals (right; Lees et al. 1991). The crosses signify the mean SFE for pre­mergers averaged within `age'­parameter
bins. The horizontal error bars represent the width of each bin, selected so that each bin comprises about 10 systems. The
vertical error bars are the standard error on the mean SFE within each bin. There is evidence that the SFE starts at a level
similar to that of isolated spirals, peaks at nuclear coalescence and then declines at post­merger stages to a level similar to that
of normal ellipticals.
Figure 2. Star­formation efficiency as a function of central molecular gas surface density. There is evidence that the SFE
and \Sigma H 2
are correlated, indicating that high central concentrations of molecular hydrogen are also associated with powerful
starbursts.
progresses towards the final nuclear coalescence. However, there is significant scatter and the
data are sparse, since there is still limited number of galaxies with high resolution CO(1­0)
observations. An additional caveat is that the synthesised beam (i.e. the minimum resolving
c
fl 0000 RAS, MNRAS 000, 000--000

16 Georgakakis, Forbes & Norris
Figure 3. FIR luminosity, L FIR , as function of the galaxy `age' parameter. The points are the same as in Figure 1. The dashed
line separates pre­ and post­mergers. For pre­mergers there is no obvious trend between L FIR and galaxy `age'. However, the
L FIR at post­merger stages steeply decreases as the the interaction induced starburst declines.
Figure 4. Molecular hydrogen mass, M(H 2 ), as function of the galaxy `age' parameter. The points are the same as in Figure 1.
The crosses signify the mean M(H 2 ) for pre­mergers averaged within `age'­parameter bins. The horizontal error bars represent
the width of the bin, selected so that each bin comprises the same number of systems (ú 10). The dashed line separates pre­
and post­mergers. There is evidence for a decrease in M(H 2 ) along the merger sequence, likely to be due to M(H 2 ) depletion
by the interaction induced starburst.
element) probes regions of different linear size for systems at different distances, contributing
to the observed scatter.
As explained in the previous section, we attempt to compensate for projection effects by
estimating the mean \Sigma H 2
within `age' bins of variable width. The results are also shown in
Figure 5, indicating an increase of the mean \Sigma H 2
for pre­mergers towards nuclear coalescence.
Also shown in Figure 5 is the mean \Sigma H 2
for isolated spirals (Kennicutt et al. 1998). It is
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 17
clear that the interacting systems have significantly higher \Sigma H 2
compared to isolated spirals.
However, it should be noted that the molecular gas surface densities of isolated spirals are
calculated by averaging the CO(1­0) emission over the optical radius of the galactic disk,
rather than the nuclear region. Therefore, the \Sigma H 2
of field spirals calculated by Kennicutt et
al. (1998) is expected to significantly underestimate their central surface density. We were
also unable to find a representative average \Sigma H 2
for ellipticals in the literature.
Numerical simulations (e.g. Mihos & Hernquist 1996) have demonstrated that tidal en­
counters trigger significant gas in­flows that lead to high central concentrations of gas. How­
ever, as discussed in the previous section, these models also predict that gas dissipation
occurs at different stages of the interaction depending on the galaxy internal structure and
interaction geometry.
For post­mergers the situation is less clear, since there are only three merger remnants
in our sample with available high resolution CO(1­0) observations and thus, any conclusions
are hampered by poor statistics. Nevertheless, there is evidence for decreasing \Sigma H 2 as the
system evolves after the nuclear coalescence. Observations of the molecular gas distribution
of a statistically complete sample of merger remnants are essential to further explore this
trend. Moreover, little is known about the molecular gas distribution of `normal' elliptical
galaxies. Comparison between the \Sigma H 2
for ellipticals and candidate merger remnants is
essential to test this aspect of the merging hypothesis.
4.3 Cold gas
The ratio of cold (molecular and neutral hydrogen) gas mass to the blue band luminosity
([M(HI) + M(H 2 )]=L B ) is plotted as a function of the `age' parameter in Figure 6. This
ratio estimates the fraction of cold gas mass in the system. There is evidence that (M(HI)+
M(H 2 ))=LB is, on average, decreasing along the merging sequence from early interacting
systems to late merger remnants, indicating cold gas depletion during the interaction.
As explained in the previous section, gravitational instabilities during the interaction
drive most of the gas into the centre of the system, where it is likely to be efficiently
converted into stars. Additionally, Hibbard & Van Gorkom (1996) found little evidence
for neutral hydrogen within remnant bodies, with most of it lying in the outer regions
(i.e. tidal features). Numerical N­body simulations have shown that the gravitational forces
experienced during the merger can force about half of the outer disc H I into a tail, the rest
c
fl 0000 RAS, MNRAS 000, 000--000

18 Georgakakis, Forbes & Norris
Figure 5. Central nuclear surface density of molecular hydrogen, \Sigma H 2
, as a function of the galaxy `age' parameter. The points
are the same as in Figure 1. The filled squares corresponds to the mean \Sigma H 2
of isolated spirals (Kennicutt 1998; however see
discussion in text). The crosses signify the mean \Sigma H 2
for pre­mergers averaged within `age'­parameter bins. The horizontal
error bars represent the width of each bin, selected so that each bin comprises about 5 points. The vertical error bars are the
standard error around the mean \Sigma H 2
within each bin. There is evidence that the \Sigma H 2
for pre­mergers peaks close to nuclear
coalescence, indicating gas inflows arising from gravitational instabilities. At post­merger stages the \Sigma H 2
seems to decline but
the poor statistics do not allow any firm conclusions to be drawn.
of the H I being forced into the inner regions (Hibbard & Mihos 1995). As less than a quarter
of the total H I is found within these regions, Hibbard & Van Gorkom (1996) concluded that
most of the centrally forced H I gas is converted during the merger into some other form.
They propose that the gas has been turned into molecular gas, stars or has been heated
up to X­ray temperatures, either through compression leading to cloud­cloud collisions or
through energy input from massive stars and supernovae. The presence of Balmer absorption
lines in the merger remnants NGC 7252 and NGC 3921 (Dressler & Gunn 1983; Schweizer
1996) is direct evidence that some of the original atomic gas is ultimately converted into
stars. Searches for molecular hydrogen in these same galaxies have revealed that although
they are gas rich compared to ellipticals or S0s, they have below average molecular gas
content for their spirals progenitors (Solomon & Sage 1988; Young & Knezek 1989; Hibbard
& van Gorkom 1996), indicating that any net conversion of atomic to molecular hydrogen is
relatively inefficient. This is also demonstrated in Figure 7, where we plot the ratio of neutral
to molecular hydrogen mass as a function of the `age' parameter. It is clear that there is
significant scatter without any obvious correlation, suggesting that the net conversion from
H I to H 2 is not large during the merging. However, it should be noted that any conversion
from H I to H 2 is likely to take place in the central galaxy regions, whereas Figure 7 plots
the global gas properties of interacting systems.
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 19
Regarding the X­ray properties of mergers, Read & Ponman (1998) found an increase in
the X­ray luminosity of galaxies close to the nuclear coalescence indicative of the presence
of hot gas. However, they found little evidence for the presence of hot X­ray emitting gas in
merger remnants. They concluded that this is likely to be due to galactic winds, similar to
those observed in the nearby starburst M 82, that blow the hot gas out of the system.
Elliptical galaxies are also known to have little cold gas. This is demonstrated in Figure
6 showing the mean [M(HI) +M(H 2 )]=L B ratio for ellipticals (Bregman, Hogg & Roberts
1992). Although merger remnants are gas rich compared to normal ellipticals, they seem to
form a sequence that supports the merger scenario as a possible formation mechanism for
elliptical galaxies. Hibbard et al. (1994) studied the gas properties of the merger remnant
NGC 7252 and found that although it is gas rich compared to ellipticals and S0s, it is likely
to resemble these galaxies in few Gyrs. In particular, the H I is mostly found in the tidal tails,
while the atomic gas content of the remnant body is typical to that of E/S0s. There is also
some evidence for on­going conversion of the returning tidal tail H I into stars. Moreover,
the molecular gas in the NGC 7252 is also likely to be depleted within the next few Gyrs,
due to modest on­going star­formation.
The ring galaxy ESO 034­IG04 in Figure 1 has cold gas mass fraction similar to that
of merger remnant candidates resulting from major disk­galaxy encounters. A similar result
was obtained by Horellou et al. (1995), who found a mean [M(HI) + M(H 2 )]=L B ratio of
0:22 \Sigma 0:17 for ring systems.
4.4 Radio flux density
We define the ratio, R, between total radio (1.4 GHz) flux density, S total
1:4 and B­band lumi­
nosity (Hummel 1981)
R = log(S total
1:4 ) + 0:4 \Theta (B T \Gamma 12:5): (5)
The R parameter is plotted against the galaxy `age' in Figure 8. The mean R within different
`age' bins (of variable width) is also shown in the same figure. It is clear that the R parameter
for pre­mergers, although on average higher than that of isolated spirals (ú 0:8 dex; Klein
1982), hardly increases along the merging sequence towards the final stages of the tidal
encounter. This is a surprising result since the radio flux is related to star­formation activity
in galaxies (Condon et al. 1992 and references therein). This can be partly attributed to
projection effects and differences in the details of individual interactions. Moreover, it has
c
fl 0000 RAS, MNRAS 000, 000--000

20 Georgakakis, Forbes & Norris
Figure 6. Total mass of neutral and molecular hydrogen normalised to the B­band luminosity as a function of galaxy `age'.
The points are the same as in Figure 1. Points connected with a line represent the upper and lower [M(H 2 ) + M(HI)]=LB
limits for that system. For isolated spirals and ellipticals (filled squares) the mean [M(H 2 ) +M(HI)]=LB is taken from Young
& Knezek (1989) and Bregman, Hogg & Roberts (1992) respectively. The filled circle is the ring system ESO 034­IG11. There
is evidence for a decrease in [M(H 2 ) + M(HI)]=LB from pre­ to post­mergers, likely to be due to cold gas depletion.
Figure 7. Neutral to molecular hydrogen mass, M(HI)=M(H 2 ), as a function of galaxy `age'. The points are the same as in
Figure 1. For isolated spirals and ellipticals (filled squares) the mean M(HI)=M(H 2 ) is taken from Young & Knezek (1989)
and Wiklind et al. (1995) respectively. The filled circle is the ring system ESO 034­IG11. There is no obvious trend, implying
little net conversion from H I to H 2 during the interaction (however see discussion in text).
been shown that tidal encounters primary act to increase the nuclear galaxy star­formation
within the central kpc (Keel et al. 1985; Kennicutt et al. 1987) and only moderately affect the
activity (i.e. star­formation) in the disk. In particular, Hummel (1981) found little difference
between the disk radio power (normalised to the blue­band luminosity) of interacting pairs
and isolated spirals. On the contrary an increase by a factor of 2.5 was found for the central
radio power of the two samples. Also shown in Figure 8 is the mean R for ellipticals (Sadler
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 21
Figure 8. Total radio to B­band flux ratio (R = logfS tot
1:4 + 0:4 \Theta [B T \Gamma 12:5]g) as a function of galaxy `age'. The points are
the same as in Figure 1. The filled squares are the mean R for isolated spirals (Klein 1982) and ellipticals (Sadler 1984). The
crosses signify the mean R for pre­mergers averaged within `age'­parameter bins. The horizontal error bars represent the width
of the bin, selected so that each bin comprises about 10 systems. The vertical error bars are the standard error on the mean.
Although the mean R for pre­mergers is higher that that of isolated spirals, it marginally increases along the merger sequence
towards the nuclear coalescence. The R parameter at post­merger stages declines steeply, to values typical of normal ellipticals.
1984). It is clear that the radio properties of merger remnant candidates and ellipticals are
in fair agreement.
To further explore changes in the central radio activity along the merging sequence, we
estimate the 1.4 GHz radio flux density of the galaxies in the present sample within the cen­
tral ú 2 kpc diameter region. For that purpose we employ high resolution radio data available
on NED, mostly taken from Condon et al. (1991). For galaxies without available 1.4 GHz
data, S cen
1:4 is estimated from high resolution observations at other frequencies (8.4 GHz or
4.9 GHz), if available, assuming a power­law spectral energy distribution S š / š \Gammaff . The
radio spectral index is either taken to be 0.8 or calculated from the 8.4 GHz and 4.9 GHz
radio flux densities, if available. The results are presented in Figure 9, plotting the central
radio flux density to blue band luminosity, RC , against galaxy `age'. Again, the increase in
the mean RC along the merger sequence for pre­mergers is marginal with significant scatter,
although the poor statistics do not allow any firm conclusions to be drawn. Nevertheless,
the mean RC of interacting systems is significantly elevated compared to that of isolated
spirals (ú 1:5 dex), in agreement with previous studies (Hummel 1981).
The ring galaxies ESO 034­IG11 and ESO 138­IG29 in Figure 9 have central radio to
the B­band flux ratios significantly lower than merger remnants resulting from major disk­
galaxy encounters. This is likely to be due to the fact that the star­formation in ring systems
c
fl 0000 RAS, MNRAS 000, 000--000

22 Georgakakis, Forbes & Norris
Figure 9. Central radio to the B­band flux ratio (RC = logfS cen
1:4 +0:4 \Theta [B T \Gamma 12:5]g) as a function of galaxy `age'. The points
are the same as in Figure 1. The filled square is the mean RC for isolated spirals (Hummel 1981). The filled circles are the ring
galaxies ESO 034­IG04 and ESO 138­IG29. The crosses signify the mean R for pre­mergers averaged within `age'­parameter
bins. The horizontal error bars represent the width of the bin, selected so that each bin comprises about 5 points. The standard
error around the mean is shown by the vertical error bars. Although the mean RC for pre­mergers is significantly higher that
that of isolated spirals, it marginally increases along the merger sequence towards the nuclear coalescence. At post­merger
stages there is evidence for a decline in the RC but the poor statistics hamper any interpretation.
is concentrated in the ring rather than the nuclear region, due to density waves triggered by
the nearly head­on collision.
5 THE FIR--RADIO CORRELATION
The logarithmic FIR--to--radio flux density ratio is defined by (Condon et al. 1991)
q = log[(S F IR =3:75 \Theta 10 12 )=S tot
1:4 ]; (6)
where S F IR is the FIR flux density (section 2, equation 2) and S tot
1:4 is the total radio flux
density (section 2) in units of Wm \Gamma2 Hz \Gamma1 . Radio and FIR selected starbursts as well as
optically selected spiral and irregular galaxies exhibit a very narrow q­distribution (oe q ú 0:2)
centred on ! q ?ú 2:34. This tight distribution is attributed to star­formation activity
resulting in both FIR emission and supernovae explosions, whose remnants emit at radio
wavelengths via sychnotron radiation. Additionally, galaxies with radio emission powered by
an AGN have, on average, q ! 2, implying that the FIR--radio flux ratio can be employed
to constrain the nature of the energising source (AGN/star­formation; Condon 1992 and
references therein).
Figure 10 plots q as a function of the galaxy age parameter. Also shown in this figure is
the region occupied by starbursts and normal spirals. The interacting systems in the present
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 23
sample (both pre­ and post­mergers) have a mean FIR--radio flux ratio ! q ?= 2:36 \Sigma 0:25,
in fair agreement with the canonical value of q = 2:34 \Sigma 0:20. Therefore, the q parameters of
most of these systems are consistent with star­formation activity being the main energising
source. There also few pre­mergers in the sample with q ! 2, which might be indicating the
presence of an AGN contributing to the observed activity. Interestingly, all these systems
have nuclear separation ffir ! 5 kpc implying that they are likely to be at an advanced
interaction stage, close to the final merger event. Nevertheless, the majority of the very
close galaxy pairs (ffir ! 5 kpc) have FIR--radio flux ratios consistent with star­formation
activity.
Smith et al. (1993) also investigated the FIR--radio correlation of interacting galaxies and
concluded that star­formation is likely to be the main source responsible for the observed FIR
and radio activity. Similarly, Bushouse et al. (1988) studied the FIR properties of interacting
galaxies and concluded that it is not necessary to invoke mechanisms other than starbursts
to account for their activity. Dahari (1985) used optical spectra to determine the nature
of the energising source in interacting systems and found little evidence for an excess of
Seyfert nuclei in paired galaxies compared to isolated spirals. Moreover, he found no Seyfert
type spectra in a sub­sample of extremely distorted spirals (e.g. similar the majority of
pre­mergers studied here).
Merger remnant candidates in Figure 10 exhibit significant scatter, with many of them
deviating from the expected relation for starbursts, although the poor statistics do not allow
any firm conclusions to be drawn. Moreover, little is known about the FIR­radio flux ratio
distribution of merger remnants. Nevertheless, elliptical galaxies, with which post­mergers
are most likely associated, follow the FIR--radio relation for star­forming galaxies (Wrobel
& Heeschen 1991). Nevertheless, a number of ellipticals are also found to deviate from
this relation having low q values, indicating the presence of an AGN. Moreover, Wrobel &
Heeschen (1991) also found that a fraction of the elliptical galaxies in their sample lied well
above the canonical FIR­radio relation (high q values). They argue that these systems are
likely to have extended low­surface brightness radio emission, associated with star­formation,
that might remain undetected by the existing observations.
c
fl 0000 RAS, MNRAS 000, 000--000

24 Georgakakis, Forbes & Norris
Figure 10. Logarithmic FIR--radio flux ratio, q, as a function of the galaxy age parameter. Open circles are the interacting
galaxies studied here. The continuous line is the mean q for starbursts, while the dashed lines signify the 1oe envelope around
the mean. The dotted line separates pre­ from post­mergers. Most of the interacting galaxies are consistent with star­formation
being their energising source.
6 CONCLUSIONS
In this paper we explore the evolution of the gas properties and star­formation along a
galaxy merger sequence. Our conclusions are summarised below
(i) The SFE of on­going mergers in the sample peaks close to the final stages of nuclear
coalescence. Nevertheless, there is significant scatter, attributed to both projection effects
and differences in the interaction details of individual systems. The observed trend is likely
to be due to M(H 2 ) depletion by star­formation. At post­merger stages, despite the poor
statistics, there is evidence that the SFE declines to values typical to ellipticals, in agreement
with the merger hypothesis.
(ii) The central surface density of molecular hydrogen increases, on average, close to
nuclear coalescence, indicating gas dissipation due to gravitational instabilities. However,
projection effects and differences in the interaction details of individual systems contribute
to the observed scatter.
(iii) There is evidence for a decrease in cold gas mass fraction (neutral and molecular
hydrogen) along the merging sequence. This attributed to H I conversion into other forms
within the body of the system during the interaction and M(H 2 ) depletion due to resid­
ual star­formation activity. This trend also seems to support the merging scenario for the
formation of ellipticals.
(iv) The total radio power normalised to the blue­band luminosity, although higher than
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 25
that of isolated spirals, marginally increases along the merger sequence. This is attributed to
the fact that interactions mainly affect the nuclear galaxy activity, whereas there is moderate
enhancement in the disk star­formation rate. However, a similar result is obtained for the
central radio power of interacting systems (normalised to the blue­band luminosity). Never­
theless, the nuclear radio to blue­band luminosity ratio of interacting systems is significantly
elevated compared to isolated spirals.
(v) The FIR--radio flux ratio distribution of interacting galaxies is consistent with star­
formation being the main energising source. However, there is evidence that some systems
might have an AGN contributing to the observed activity.
7 ACKNOWLEDGEMENTS
This research has made use of the NASA/IPAC Extragalactic Database (NED), which is
operated by the Jet Propulsion Laboratory, Caltech, under contract with the National Aero­
nautics and Space Administration.
REFERENCES
Aalto S., Booth R. S., Black J. H., Johansson L. E. B., 1995, A&A, 300, 369
Andreani P., Casoli F., Gerin M., 1995, A&A, 300, 43
Barnes J. E., 1988, ApJ, 331, 699
Barnes J. E., 1992, ApJ, 393, 484
Becker R. H., White R., L., Edwards A. L., 1991, ApJS, 75, 1
Bregman J. N., Hogg D. E., Roberts M. S., 1992, ApJ, 387, 484
Bergvall N., Johansson L., 1995, A&AS, 113, 499
Bergvall N., Ronnback J., Johansson L., 1989, A&A, 222, 49
Boselli A., Mendes De Oliveira C., Balkowski C., Cayatte V., Casoli F., 1996, A&A, 314, 738
Bottinelli L., Gouguenheim L., Paturel G., 1980, A&A, 88, 32
Bushouse H. A., 1987, ApJ, 320, 49
Bushouse A. H., Lamb S. A., Werner W. M., 1988, ApJ, 335, 74
Casoli F., Dupraz C., Combes F., Kazes I., 1991, A&A, 251, 32
Cayatte V., Balkowski C., van Gorkom J. H., Kotanyi C., 1990, AJ, 100, 604
Condon J. J., 1992, ARA&A, 30, 575
Condon J. J., Huang Z. P., Yin Q. F., Thuan T. X., 1991, ApJ, 378, 65
Condon J. J., Helou G., Sanders D. B., Soifer B. T., 1990, ApJS, 73, 359
Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F., Perley, R. A., Taylor, G. B., & Broderick, J. J. 1998, AJ, 115, 1693
Crabtree D. R., Smecker­Hane T., 1994, AAS, 185, 107.14
Dahari O., 1985, ApJS, 57, 643
Davis L. E., Seaquist E. R., ApJS, 53, 269
c
fl 0000 RAS, MNRAS 000, 000--000

26 Georgakakis, Forbes & Norris
de Vaucouleurs G., de Vaucouleurs A., Corwin H.G., Buta R.J., Paturel G., Fouque P., 1991, Third Reference Catalogue of
Bright Galaxies (Springer­Verlag: New York) (RC3)
Dressler A., Gunn J. E., 1983, ApJ, 270, 7
Elfhag T., Booth R. S., Hoeglund B., Johansson L. E. B., Sandqvist A., 1996, A&AS, 115, 439
Eskridge P. B., Pogge R. W., 1991, AJ, 101, 2056
Feigelson E. D., Nelson P. I., 1985, ApJ, 293, 192
Feigelson E., Isobe T., La Valley M., 1992, Dept. Astron. & Astrophys. Pennsylvania State Univ. Pub. ASURV.
Forbes D. A., Ponman T. J., Brown R. J. N., 1998, 508, L43
Gallimore J. F., Keel W. C., 1993, AJ, 106, 1337
Gao Y., Solomon P. M., 1999, ApJ, 512, 99
Gao Y., Gruendl R. A., Hwang C. Y., Lo K. Y., astro­ph/9812459
Garnier R., Paturel G., Petit C., Marthinet M. C., Rousseau J., 1996, A&AS, 117, 467
Giovanelli R., Haynes M. P., 1993, AJ, 105, 1251
Gooch R. E., ``Karma: a Visualisation Test­Bed'', in Astronomical Data Analysis Software and Systems V, ASP Conf. series
vol. 101, ed. G. H. Jacoby and J. Barnes, ASP, San Francisco, p. 80, ISSN 1080­7926
Haynes M. P., et al. 1997, AJ, 113, 1197
Heckman T. M., Balick B., van Breugel W. J. W., Miley G. K., 1983, AJ, 88, 583
Hernquist L., ApJ, 1992, 400, 460
Hernquist L., ApJ, 1993, 409, 548
Hibbard J. E., Guhathakurta P., van Gorkom J. H., Schweizer F., 1994, AJ, 107, 67
Hibbard J. E., van Gorkom J. H., AJ, 1996, 111, 655
Hibbard J. E., Mihos J. C., 1995, AJ, 110, 140
Higdon J. L., Wallin J. F., 1997, ApJ, 474, 686
Horellou C., Booth R., 1997, A&AS, 126, 3
Horellou C., Casoli F., Combes F., Dupraz C., 1995, A&A, 298, 743
Hummel E., Van der Hulst J. M., Kennicutt R. C., Keel W. C., 1990, A&A, 236, 333
Hummel E., 1981, A&A, 96, 111
Huchtmeier W. K., Sage L. J., Henkel C., 1995, A&A, 3000, 675
Huchtmeier W. K., 1997, A&A, 319, 401
Huchtmeier W. K. Bohnenstengel H. D., 1975, A&A, 44, 479
Joseph R. D., Wright G. S., 1985, MNRAS, 214, 87
Kazes I., Proust D., Mirabel L. F., Combes F., Balkowski C., Martin J. M., 1990, A&A, 237, 1
Keel W. C., Kennicutt R. C., Hummel E., Van der Hulst J. M., 1985, AJ, 90, 708
Keel W. C., Wu W., 1995, AJ, 110, 129
Kennicutt R. C., ApJ, 1998, 498, 541
Kennicutt R. C., Roettiger K. A., Keel W. C., Van der Hulst J. M., Hummel E., AJ, 1987, 93, 1011
Klein U., 1982, A&A, 116, 175
Lawrence A., Rowan­Robinson M., Leech K., Jones D. H., Wall J. V., 1989, MNRAS, 240, 329
Lees J. F., Knapp G. R., Rupen M. P., Phillips T. G., 1991, ApJ, 379, 177
Lo K. Y., Gao Y., Gruendl R. A., 1997, ApJ, 475L, 103
Malin D. F., Carter D., 1980, Natur., 285, 643
Martin J. M., Bottinelli L., Gouguenheim L., Dennefeld M., 1991, A&A, 245, 393
Martin M. C., 1998, A&AS, 131, 77
Maiolino R., Ruiz M., Rieke G. H., Papadopoulos P., 1997, ApJ, 485, 552
Mihos J. C., Hernquist L., 1996, ApJ, 464, 641
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 27
Mirabel I. F., Booth R. S., Johansson L. E. B., Garay G., Sanders D. B., 1990, A&A, 236, 327
Mirabel I. F., Sanders D. B., 1988, ApJ, 335, 104
Neff S. G., Hutchings J. B., Standord S. A., Unger S. W., 1990, AJ, 99, 1715
Norris R. P., Forbes D. A., 1995, ApJ, 446, 594
Planesas P., Colina L., Perez­Olea D., 1997, A&A, 325, 81
Read A. M., Ponman T. J., 1998, MNRAS, 297, 143
Richter O. G., Sackett P. D., Sparke L. S., 1994, AJ, 107, 99
Richter O. G., Huchtmeier W. K., 1991, A&AS, 87, 425
Roberts M., Hogg D., Bregman J. N., Forman W. R., Jones C., 1991, ApJS, 75, 751
Sadler E. M., 1984, AJ, 89, 34
Sage L. J., Loose H. H., Salzer J. J., 1993, A&A, 273, 6
Sanders D. B., Scoviller N. Z., Soifer B. T., 1991, ApJ, 370, 158
Sanders D. B., Solomon P. M., Scoville N. Z., 1984, ApJ, 276, 182
Sanders D. B., Mirabel I. F., 1996, ARA&A, 34, 749
Smith D. A., Herter T., Haynes M. P., 1998, ApJ, 494, 150
Smith E. P., Kassim N. E., 1993, AJ, 105, 46
Soifer B. T., Sanders D. B., Madore B. F., Neugebauer G., Danielson G. E., Elias J. H., Lonsdale C. J., Rice W. L., 1987, ApJ,
320, 238
Solomon P. M., Sage L. J., 1988, ApJ, 334, 613
Solomon P. M., Downes D., Radford S. J. E., Barrett J. W., 1997, ApJ, 478, 144
Storchi B. T., Patroriza M. G., 1986, ApJ, 304, 305
Surace J. A., Mazzarealla J., Soifer B. T., Wehrle A. E., 1993, AJ, 105, 864
Schiminovich D., van Gorkom J. H., van der Hulst J. M., Kasow S., 1994, ApJ, 423, 101
Schweizer W., 1996, AJ, 111, 109
Schweizer W., Seitzer P., 1988, ApJ, 328, 88
Stanford S. A., Wood, Douglas O. S., 1989, ApJ, 346, 712
Terlevich A. I., Forbes D. A., 1999, in preparation
Theureau G., et al. 1998, A&AS, 130, 333
Tinney C. G., Scoville N. Z., Sanders D. B., Soifer B. T., 1990, ApJ, 362, 473
Toomre A., Toomre J., 1972, ApJ, 178, 623
Toomre A., 1977, `The Evolution of Galaxies and Stellar Populations', edited by B. M. Tinsley and R. B. Larson (Yale University,
New Haven), p. 401
Young J. S., Knezek P. M., 1989, ApJ, 347, L55
Young, J. S., et al. 1995, ApJS, 98, 219
Yun M. S., Hibbrand J. E., 1999, astro­ph/9903463
Yun M. S., Scoville N. Z., Knop R. A., 1994, ApJ, 430, 109
Wallin J. L., Struck­Marcell C., 1994, ApJ, 433, 631
Wiklind T., Combes F., Henkel C., 1995, A&A, 297, 643
Wrobel J. M., & Heeschen D. S., 1991, AJ, 101, 148
c
fl 0000 RAS, MNRAS 000, 000--000

28 Georgakakis, Forbes & Norris
APPENDIX A: NOTES ON INDIVIDUAL GALAXIES WITH NEW RADIO
DATA
NGC 6769/70
This galaxy is strongly interacting with an equal mass spiral NGC 6770 with a bridge of
stars connecting them. The optical centres of the two galaxies lie at RA = 19 h 18 m 22.6 s ,
Dec.= --60 ffi 30 0 03 00 and RA = 19 h 18 m 37.6 s , Dec.= --60 ffi 29 0 50 00 (J2000) respectively. Surface
photometry of the galaxies is discussed in Storchi & Patroriza (1986). The radio contours
at 4.79 GHz overlayed on the optical image of NGC 6770 are shown in Figure A1.
ESO 286--IG19
Imaging of ESO286--IG19 by Johansson (1991) reveals two tidal tails and a single r 1=4 like
nucleus. This would tend to suggest a late stage merger. Comparison with the sequence of
Keel & Wu (1995) suggests a dynamical stage of 1.5 (i.e ú 6 \Theta 10 8 yrs). The radio contours
at 4.79 GHz overlayed on the optical image of ESO 286--IG19 are shown in Figure A2.
NGC 1487
This system is in an early stage merger revealing two clear tails and two nuclei but suffi­
ciently advanced to be one galactic body (e.g. Bergvall & Johansson 1995). It is slightly more
evolved than NGC 4676 (``The Mice'') but less so than NGC 4038/9 (``The Antennae''). The
radio contours at 4.79 GHz overlayed on the optical image of NGC 1487 are shown in Figure
A3.
ESO 034--IG11
Also known as the Lindsay--Shapley Ring (Lindsay & Shapley 1960). The asymmetric ring
suggests an off--centre collision. It has been studied by Higdon & Wallin (1997). They found
an optical bridge from the ring to the `intruder' galaxy at RA = 06 h 43 m 26 s , Dec. = --
74 ffi 15 0 29 00 (J2000). The optical nucleus of the perturbed galaxy is at RA = 06 h 43 m 06.7 s ,
Dec.= --74 ffi 14 0 16 00 (J2000). From the expansion rate of the ring in ESO34--IG11 the interac­
tion occurred about 1:8 \Theta 10 8 yrs ago (H 0 = 75 km s \Gamma1 Mpc \Gamma1 ; Higdon & Wallin 1997). The
radio contours at 4.79 GHz overlayed on the optical image of ESO 034--IG11 are shown in
Figure A4.
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 29
NGC 7764A
The NGC 7764A system is an interacting triple system containing NGC 7764A NED2
(AM2350--410), 7764A NED3 to the south--east and 7764A NED1 to the north--west. The
largest galaxy, NGC 7764A NED2, appears to be in the process of merging with NGC 7764A
NED1 Borchkhadze et al. (1977) note the tidal material between these two galaxies. The
smallest galaxy, NGC 7764A NED3 on the other hand shows only faint asymmetries in its
outer optical isophotes. In this study, the NGC 7764A NED1/NED2 system is treated as a
pre--merger, with two distinct galaxies present. The radio contours at 4.79 GHz overlayed on
the optical image of NGC 7764A are shown in Figure A5.
ESO 138--IG29
Also known as the ``Sacred Mushroom'', this galaxy appears to be a young ring system
formed by ESO138--IG30 as it passed through the disk of ESO138--IG29 less than 10 8 yrs
ago (Wallin & Struck--Marcell 1994). Optical imaging and dynamical models of this system
have been carried out Wallin & Struck--Marcell (1994) which suggest that ESO138--IG29
was originally an S0 galaxy. The radio contours at 4.79 GHz overlayed on the optical image
of ESO 138--IG29 are shown in Figure A6.
ESO 341--IG04
This galaxy is probably at the very last stages of a merger. Its optical appearance is close to
that of an elliptical galaxy (its surface brightness profile follows an r 1=4 law out to 5 effective
radii), although it still contains a large amount of HI gas (Bergvall et al. 1989). It has one
nucleus and one tail or plume, in addition to the prominent south--west loop. It appears to
be more evolved than NGC 7252 but not quite at the end of the Keel & Wu (1995) merger
sequence. The optical nucleus is at RA =20 h 41 m 14.3 s , Dec. = --38 ffi 11 0 40 00 (J2000). The radio
contours at 4.79 GHz overlayed on the optical image of ESO 341--IG04 are shown in Figure
A7.
c
fl 0000 RAS, MNRAS 000, 000--000

30 Georgakakis, Forbes & Norris
Figure A1. NGC 6770. Radio contours at 4.79 GHz (6 cm) overlayed on the Digital Sky Survey optical image (linear scale).
The radio contours are logarithmically spaced between 0.05 mJy/beam and 1.35 mJy/beam using a logarithmic step of 0.48.
Figure A2. ESO 286­IG19. Same as in Figure A1. The radio (4.79GHz) contours are logarithmically spaced between
0.06 mJy/beam and 3.84 mJy/beam using a logarithmic step of 0.60.
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 31
Figure A3. NGC 1487. Same as in Figure A1. The radio (4.79GHz) contours are logarithmically spaced between
0.05 mJy/beam and 0.20 mJy/beam using a logarithmic step of 0.30.
Figure A4. ESO 034­IG11. Same as in Figure A1. The radio (4.79GHz) contours are logarithmically spaced between
0.05 mJy/beam and 0.20 mJy/beam using a logarithmic step of 0.30.
c
fl 0000 RAS, MNRAS 000, 000--000

32 Georgakakis, Forbes & Norris
Figure A5. NGC 7764A. Same as in Figure A1. The radio (4.79 GHz) contours are logarithmically spaced between
0.07 mJy/beam and 1.89 mJy/beam using a logarithmic step of 0.48.
Figure A6. ESO 138­IG29. Same as in Figure A1. The radio (4.79GHz) contours are logarithmically spaced between
0.07 mJy/beam and 0.21 mJy/beam using a logarithmic step of 0.60.
c
fl 0000 RAS, MNRAS 000, 000--000

Cold gas and star­formation in a merging galaxy sequence 33
Figure A7. ESO 341­IG04. Same as in Figure A1. The radio (4.79GHz) contours are logarithmically spaced between
0.05 mJy/beam and 0.40 mJy/beam using a logarithmic step of 0.30.
c
fl 0000 RAS, MNRAS 000, 000--000