Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.sai.msu.su/groups/cluster/efremov/arcs/ymc1.ps Äàòà èçìåíåíèÿ: Mon Jul 22 09:36:22 2002 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 07:18:25 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: m 87 jet

A&A manuscript no.
(will be inserted by hand later)
Your thesaurus codes are:
3 (11.09.1 11.16.1 11.19.2 11.19.4 )
ASTRONOMY
AND
ASTROPHYSICS
February 16, 1999
Young massive star clusters in nearby galaxies ?
I. Identification and general properties of the cluster systems
S.S. Larsen 1 and T. Richtler 2
1 Copenhagen University Astronomical Observatory, Juliane Maries Vej 32, 2100 Copenhagen ü, Denmark
email: soeren@astro.ku.dk
2 Sternwarte der Universit¨at Bonn, Auf dem H¨ugel 71, D­53121 Bonn, Germany
email: richtler@astro.uni­bonn.de
Received ...; accepted ...
Abstract. Using ground­based UBVRIHff CCD pho­
tometry we have been carrying out a search for young
massive star clusters (YMCs) in a sample consisting of 21
nearby spiral galaxies. We find a large variety concern­
ing the richness of the cluster systems, with some galaxies
containing no YMCs at all and others hosting very large
numbers of YMCs. Examples of galaxies with poor clus­
ter systems are NGC 300 and NGC 4395, while the richest
cluster systems are found in the galaxies NGC 5236 (M83),
NGC 2997 and NGC 1313. The age distributions of clus­
ters in these galaxies show no obvious peaks, indicating
that massive clusters are formed as an ongoing process
rather than in bursts. This is in contrast to what is ob­
served in starbursts and merger galaxies. The radial distri­
butions of clusters follow the Hff surface brightnesses. For
the galaxies in our sample there is no correlation between
the morphological type and the presence of YMCs.
Key words: Galaxies: individual -- photometry -- spiral
-- star clusters
1. Introduction
During the last decade many investigations have revealed
the presence of ``young massive star clusters'' (YMCs) or
``super star clusters'' in mergers and starburst galaxies,
and it has been speculated that these objects could be
young analogues of the globular clusters seen today in the
Milky Way. It is an intriguing idea that globular clusters
could still be formed today in some environments, because
the study of such objects would be expected to provide a
direct insight into the conditions that were present in the
Send offprint requests to: S.S. Larsen
? Based on observations made with the Nordic Optical Tele­
scope, operated on the island of La Palma jointly by Denmark,
Finland, Iceland, Norway, and Sweden, in the Spanish Observa­
torio del Roque de los Muchachos of the Instituto de Astrofisica
de Canarias, and with the Danish 1.5­m telescope at ESO, La
Silla, Chile.
early days of our own and other galaxies when the globular
clusters we see today in the halos were formed.
Probably the most famous example of a merger galaxy
hosting ``young massive clusters'' is the ``Antennae'',
NGC 4038/39 where Whitmore & Schweizer (1995) dis­
covered more than 700 blue point­like sources with ab­
solute visual magnitudes up to M V = \Gamma15. Other well­
known examples are NGC 7252 (Whitmore et. al. 1993),
NGC 3921 (Schweizer et. al. 1996) and NGC 1275 (Holtz­
man et. al. 1992). All of these galaxies are peculiar sys­
tems and obvious mergers. In fact, in all cases investigated
so far where star formation is associated with a merger,
YMCs have been identified (Ashman & Zepf 1998).
But YMCs exist not only in mergers. They have been
located also in starburst galaxies such as NGC 1569 and
NGC 1705 (O'Connell et. al. 1994), NGC 253 (Watson et.
al. 1996) and M82 (O'Connell et. al. 1995), in the nuclear
rings of NGC 1097 and NGC 6951 (Barth et. al. 1995),
and in the blue compact galaxy ESO­338­IG04 ( ¨
Ostlin et.
al. 1998). The magnitudes of YMCs reported in all these
galaxies range from M V ú \Gamma10 to \Gamma15, and the effective
radii R e (R e = the radius within which half of the light
is contained) have been estimated to be of the order of a
few parsec to about 20 pc, compatible with the objects
deserving the designation ``young globular clusters''.
All of the systems mentioned above are relatively dis­
tant, but in fact one does not have to go farther away
than the Local Group in order to find galaxies containing
rather similar star clusters. The Magellanic Clouds have
long been known to host star clusters of a type not seen
in the Milky Way, i.e. compact clusters that are much
more massive than Galactic open clusters (van den Bergh
1991, Richtler 1993), and in many respects resemble glob­
ular clusters more than open clusters. Some of the most
conspicuous examples are the 10 7 years old cluster in the
centre of the 30 Doradus nebula in the LMC (Brandl et.
al. 1996), shining at an absolute visual magnitude of about
­11, and the somewhat older object NGC 1866 (about
10 8 years), also in the LMC (Fischer et. al. 1992), which

2 Young massive star clusters..
has an absolute visual magnitude of M V ú \Gamma9:0. Even if
these clusters are not quite as spectacular as those found
in genuine starburst galaxies, they are still more massive
than any of the open clusters seen in the Milky Way to­
day. YMCs have been reported also in M33 (Christian &
Schommer 1988), and in the giant Sc spiral M101 (Bresolin
et. al. 1996).
Taking into account the spread in the ages of the
YMCs in the Antennae, Fritze ­ v. Alvensleben (1998)
recovered a luminosity function (LF) resembling that of
old globular clusters (GC's) to a very high degree when
evolving the present LF to an age of 12 Gyr. Elmegreen
& Efremov (1997) point out the interesting fact that the
upper end of the LF of old GC systems is very similar to
that observed for YMCs, open clusters in the Milky Way,
and even for HII regions in the Milky Way, and this is
one of their arguments in favour of the hypothesis that
the basic mechanism behind the formation of all these ob­
jects is the same. They argue that massive clusters are
formed whenever there is a high pressure in the interstel­
lar medium, due to starbursts or other reasons as e.g. a
high virial density (as in nuclear rings and dwarf galax­
ies). However, this doesn't seem to explain the presence of
YMCs in apparently undisturbed disk galaxies like M33
and M101.
So it remains a puzzling problem to understand why
YMCs exist in certain galaxies, but not in others. In this
paper we describe some first results from an investigation
aiming at addressing this question. It seems that YMC's
can exist in a wide variety of host galaxy environments,
and there are no clear systematics in the properties of the
galaxies in which YMC's have been identified. And just
like it is not clear how YMCs and old globular clusters are
related to each other, one can also ask if the very luminous
YMCs in mergers and starburst galaxies are basically the
same type of objects as those in the Magellanic Clouds,
M33 and M101.
We therefore decided to observe a number of nearby
galaxies and look for populations of YMC's. The galaxies
were mainly selected from the Carnegie Atlas (Sandage &
Bedke 1994), and in order to minimise the problems that
could arise from extinction internally in the galaxies we
selected galaxies that were more or less face­on. We tried
to cover as wide a range in morphological properties as
possible, although the requirement that the galaxies had
to be nearby (because we would rely on ground­based ob­
servations) restricted the available selection substantially.
The final sample consists of 21 galaxies out to a distance
modulus of m \Gamma M ú 30, for which basic data can be seen
in Table 1.
In this paper we give an overview of our observations,
and we discuss the main properties of the populations of
YMCs in the galaxies in Table 1. In a subsequent paper
(Larsen et. al. 1999) we will discuss the correlations be­
tween the number of YMCs in a galaxy and various prop­
erties of the host galaxies in more detail, and compare our
data with data for starburst galaxies and mergers pub­
lished in the literature.
2. Observations and reductions
The observations were carried out partly with the Dan­
ish 1.54 m. telescope and DFOSC (Danish Faint Object
Spectrograph and Camera) at the European Southern Ob­
servatory (ESO) at La Silla, Chile, and partly with the
2.56 m. Nordic Optical Telescope (NOT) and ALFOSC
(a DFOSC twin instrument), situated at La Palma, Ca­
nary Islands. The data consists of CCD images in the fil­
ters U,B,V,R,I and Hff. In the filters BVRI and Hff we
typically made 3 exposures of 5 minutes each, and 3 ex­
posures of 20 minutes each in the U band. Both the AL­
FOSC and DFOSC were equipped with thinned, backside­
illuminated 2 K 2 Loral­Lesser CCDs. The pixel scale in the
ALFOSC is 0.189 00 /pixel and the scale in the DFOSC is
0.40 00 /pixel, and the fields covered by these two instru­
ments are 6:5 0 \Theta 6:5 0 and 13:7 0 \Theta 13:7 0 , respectively. All
observations used in this paper were conducted under pho­
tometric conditions, with typical seeing values (measured
on the CCD images) being 1.5 00 and 0.8 00 for the La Silla
and La Palma data, respectively.
During each observing run, photometric standard stars
in the Landolt (1992) fields were observed for calibra­
tion of the photometry. Some of the Landolt fields were
observed several times during the night at different air­
mass in order to measure the atmospheric extinction co­
efficients. For the flatfielding we used skyflats exposed to
about half the dynamic range of the CCD, and in general
each flatfield used in the reductions was constructed as an
average of about 5 individual integrations.
After bias subtraction and flatfielding, the three ex­
posures in each filter were combined to a single image,
and star clusters were identified using the daofind task
in DAOPHOT (Stetson 1987) on a background­subtracted
V ­band frame. Aperture photometry was then carried out
with the DAOPHOT phot task, using a small aperture
radius (4 pixels for colours and 8 pixels for the V ­band
magnitudes) in order to minimise errors arising from the
greatly varying background. Aperture corrections from the
standard star photometry (aperture radius = 20 pixels) to
the science data were derived from a few isolated, bright
stars in each frame. Because the star clusters are not true
point sources, no PSF photometry was attempted. A more
detailed description of the data reduction procedure will
be given in Larsen (1999).
The photometry was corrected for Galactic foreground
extinction using the AB values given in the Third Refer­
ence Catalogue of Bright Galaxies (de Vaucouleurs et. al.
(1991), hereafter RC3).

3
Table 1. The galaxies. In the first column each galaxy is identified by its NGC number, the second column gives the morpho­
logical classification taken from NED, right ascension and declination for equinox 2000.0 are in columns 3 and 4. Apparent blue
magnitude (from RC3) is in the 5th column, the distance modulus is given in column 6, and the absolute blue magnitude MB is
in column 7. The last column indicates which telescope was used for the observations (DK154 = Danish 1.54m. telescope, NOT
= Nordic Optical Telescope). The sources for the distances are as follows: 1 de Vaucouleurs 1963, 2 Carignan 1985, 3 Freedman
et. al. 1992, 4 Bottinelli et. al. 1985, 5 Nearby Galaxies Catalog (Tully 1988) 6 de Vaucouleurs 1979a, 7 de Vaucouleurs 1979b,
8 Shanks 1997, 9 Karachentsev & Drozdovsky 1998, 10 Freedman & Madore 1988 11 Karachentsev et. al. 1996.
Name Type ff (2000.0) ffi (2000.0) mB m \Gamma M MB Obs.
NGC 45 SA(s)dm 00:14:04 \Gamma23:10:52 11.32 28:42 \Sigma 0:41 4 ­17.13 DK154
NGC 247 SAB(s)d 00:47:08 \Gamma20:45:38 9.67 27:0 \Sigma 0:4 2 ­17.40 DK154
NGC 300 SA(s)d 00:54:53 \Gamma37:41:00 8.72 26:66 \Sigma 0:10 3 ­18.05 DK154
NGC 628 SA(s)c 01:36:42 +15:46:59 9.95 29:6 \Sigma 0:4 7 ­19.77 NOT
NGC 1156 IB(s)m 02:59:43 +25:14:15 12.32 29:46 \Sigma 0:15 11 ­17.84 NOT
NGC 1313 SB(s)d 03:18:15 \Gamma66:29:51 9.2 28:2 1 ­19.03 DK154
NGC 1493 SB(rs)cd 03:57:28 \Gamma46:12:38 11.78 30:3 5 ­18.62 DK154
NGC 2403 SAB(s)cd 07:36:54 +65:35:58 8.93 27:51 \Sigma 0:24 10 ­18.73 NOT
NGC 2835 SAB(rs)c 09:17:53 \Gamma22:21:20 11.01 28:93 \Sigma 0:42 4 ­18.30 DK154
NGC 2997 SA(s)c 09:45:39 \Gamma31:11:25 10.06 29:9 \Sigma 0:4 7 ­20.35 DK154
NGC 3184 SAB(rs)cd 10:18:17 +41:25:27 10.36 29:5 \Sigma 0:4 7 ­19.14 NOT
NGC 3621 SA(s)d 11:18:16 \Gamma32:48:42 10.28 29:1 \Sigma 0:18 8 ­19.21 DK154
NGC 4395 SA(s)m 12:25:49 +33:32:48 10.64 28:1 9 ­17.47 NOT
NGC 5204 SA(s)m 13:29:36 +58:25:04 11.73 28:4 5 ­16.68 NOT
NGC 5236 SAB(s)c 13:37:00 \Gamma29:51:58 8.20 27:84 \Sigma 0:15 6 ­19.78 DK154
NGC 5585 SAB(s)d 14:19:48 +56:43:44 11.20 29:2 5 ­18.00 NOT
NGC 6744 SAB(r)bc 19:09:45 \Gamma63:51:22 9.14 28:5 \Sigma 0:4 7 ­19.50 DK154
NGC 6946 SAB(rs)cd 20:34:52 +60:09:14 9.61 28:7 5 ­20.70 NOT
NGC 7424 SAB(rs)cd 22:57:18 \Gamma41:04:14 10.96 30:5 5 ­19.54 DK154
NGC 7741 SB(s)cd 23:43:53 +26:04:35 11.84 30:8 \Sigma 0:4 7 ­19.10 NOT
NGC 7793 SA(s)d 23:57:50 \Gamma32:35:21 9.63 27:6 \Sigma 0:20 2 ­18.04 DK154
2.1. Photometric errors
The largest formal photometric errors as estimated by
phot are those in the U band, amounting to around 0.05
mag. for the faintest clusters. However, these error esti­
mates are based on pure photon statistics and are not
very realistic in a case like ours. Other contributions to
the errors come from the standard transformation proce­
dure, from a varying background, and from the fact that
the clusters are not perfect point sources so that the aper­
ture corrections become uncertain.
The r.m.s. residuals of the standard transformations
were between 0.01 ­ 0.03 mags. in V, B­V and V­I, and
between 0.04 and 0.06 mags. in U­B.
The errors in aperture corrections arising from the fi­
nite cluster sizes were estimated by carrying out photom­
etry on artificially generated clusters with effective radii
in the range R e = 0 \Gamma 4 pixels (0 00 ­ 1.6 00 on the DFOSC
frames). 1 00 corresponds to a linear distance of about 20
pc at the distance of typical galaxies in our sample, such
as NGC 1313 and NGC 5236. The artificial clusters were
modeled by convolving the point­spread function (PSF)
with MOFFAT15 profiles.
The upper panel in Fig. 1 shows the errors in the aper­
ture corrections for V ­band photometry through aperture
radii R ap = 4 pixels and R ap = 8 pixels as a function of
R e , while the lower panel shows the errors in the colour in­
dices for R ap = 4 pixels. At R e ú 1 00 , the error in V ­band
magnitudes using R ap = 8 pixels amounts to about 0.15
magnitudes. For a given R e , the errors in the colours are
much smaller than the errors in the individual bandpasses,
so that accurate colours can be derived through the small
R e = 4 pixels aperture without problems. This convenient
fact has also been demonstrated by e.g. Holtzman et. al.
(1996).
The random errors, primarily arising due to back­
ground fluctuations, should in principle be evaluated in­
dividually for each cluster, since they depend on the local
environment of the cluster. Fig. 2 shows the random errors
for clusters in NGC 5236, estimated by adding artificial
objects of similar brightness and colour near each cluster
and remeasuring them using the same photometric proce­
dure as for the cluster photometry. Again it is found that
the errors in two different filters tend to cancel out when
colour indices are formed. The V ­band errors are quite
substantial, but we have chosen to accept the large ran­
dom errors associated with the use of an R ap = 8 pixels
aperture in order to keep the effect of systematic errors at
a low level.
3. Identification of star clusters
After the photometry had been obtained, the first step in
the analysis was to identify star cluster candidates, and to

4 Young massive star clusters..
Fig. 1. Aperture corrections as a function of intrinsic cluster
radius. Top: The errors in the V ­band aperture corrections for
aperture radii of 4 and 8 pixels. Bottom: The corresponding
errors in the colours for an aperture radius of 4 pixels.
Fig. 2. The random errors in V , B \Gamma V and U \Gamma B as a function
of V magnitude for clusters in NGC 5236.
make sure that they were really star clusters and not some
other type of objects. Possible sources of confusion could
be compact HII regions, foreground stars, and individual
luminous stars in the observed galaxies. However, each of
these objects can be eliminated by applying the following
selection criteriae:
-- HII regions: These can be easily identified due to their
Hff emission.
-- Foreground stars: Because our galaxies are located
at rather high galactic latitudes, practically all fore­
ground stars are redder than B \Gamma V ú 0:45, whereas
young massive star clusters will be bluer than this
limit. Hence, by applying a B \Gamma V limit of 0.45 we
sort away the foreground stars while retaining the
young massive cluster candidates. Remaining fore­
ground stars could in many cases be distinguished by
their position in two­colour diagrams, by their lack of
angular extent, and by being positioned outside the
galaxies.
-- Individual luminous stars in the galaxies: We apply a
brightness limit of M V = \Gamma8:5 for cluster candidates
with U \Gamma B ? \Gamma0:4 and M V = \Gamma9:5 for candidates
with U \Gamma B ! \Gamma0:4. The bluer objects are often found
inside or near star forming regions, but the magnitude
limit of M V = \Gamma9:5 should prevent confusion with
even very massive stars.
In addition to these selection criteriae it was found
very useful to generate colour­composite images using the
I, U and Hff exposures and identify all the cluster candi­
dates visually on these images. For the ``red'' channel we
used the Hff exposures, for the ``green'' channel we used
the I­band frames, and for the ``blue'' channel the U­band
frames. In images constructed like this, YMCs stand out
very clearly as compact blue objects, in contrast to HII
regions which are distinctly red, and foreground stars and
background galaxies which appear green.
Following the procedure outlined above, we ended up
with a list of star cluster candidates in each galaxy. The
cluster nature of the detected objects was further verified
by examining their positions in two­colour diagrams (U­
B,B­V and U­V,V­I), and compare with model predictions
for the colours of star clusters and individual stars. In ad­
dition, we have been able to obtain spectra of a few of the
brightest star cluster candidates. These will be discussed
in a subsequent paper.
The cluster samples may suffer from incompleteness
effects. In particular, we have deliberately excluded the
youngest clusters which are still embedded in giant HII
regions (corresponding to an age of less than about 10 7
years). Clusters which have intrinsic B \Gamma V ! 0:45 will
also slip out of the sample if their actual observed B \Gamma V
index is larger than 0.45 due to reddening internally in the
host galaxy.

5
Fig. 3. Histogram of (uncorrected) TN values.
3.1. Counting clusters
The specific frequency for old globular cluster systems has
traditionally been defined as (Harris & van den Bergh
1981):
SN = NGC \Theta 10 0:4\Theta(M V +15) (1)
where NGC is the total number of globular clusters belong­
ing to a galaxy of absolute visual magnitude M V . Such a
definition is a reasonable way to characterise old globular
cluster systems because NGC is a well­defined quantity,
which can be estimated with good accuracy due to the
gaussian­like luminosity function (LF) even if the faintest
clusters are not directly observable. In the case of young
clusters it is more complicated to define a useful measure
of the richness of the cluster systems, because the LF is
no longer gaussian and the number of young clusters that
one finds in a galaxy depends critically on the magnitude
limit applied in the survey. Nevertheless, we have defined
a quantity equivalent to SN for the young cluster systems:
TN = NYMC \Theta 10 0:4\Theta(M B+15) (2)
NYMC is the number of clusters NB + NR satisfying the
criteriae described in Sec. 3. We have chosen to normalise
TN to the B­band luminosity of the host galaxy because
it can be looked up directly in the RC3 catalogue.
4. Results
4.1. Specific frequencies
In Table 2 we give the number of clusters identified
in each of the observed galaxies. The columns labeled NB
and NR refer to the number of ``blue'' and ``red'' clusters
respectively, according to the definition that ``blue'' clus­
ters are clusters with U \Gamma B ! \Gamma0:4 (and hence M V !
\Gamma9:5) whereas the ``red'' clusters have U \Gamma B – \Gamma0:4 (and
M V ! \Gamma8:5). See also Sect. 3. The data for the LMC
Table 2. Number of clusters identified in each of the galaxies.
NB refer to the 'blue' clusters (i.e. clusters with U \Gamma B ! \Gamma0:4
and NR refer to the 'red' clusters (clusters with U \Gamma B – \Gamma0:4).
The fourth column is the total number of clusters, NB +NR .
The quantities TN and TN;C are defined in Sect. 3.1. 1 Only the
central parts of the galaxies were covered by our observations.
Name NB NR NYMC TN TN;C
NGC 45 1 2 3 0:42 \Sigma 0:29 ­
NGC 247 1 2 3 0:32 \Sigma 0:23 ­
NGC 300 1 1 2 1 3 1 0.18\Sigma0:11 1 ­
NGC 628 27 12 39 0:48 \Sigma 0:19 0:50 \Sigma 0:20
0:57 \Sigma 0:23
NGC 1156 13 9 22 1:61 \Sigma 0:41 1:81 \Sigma 0:46
3:09 \Sigma 0:92
NGC 1313 17 29 46 1:12 \Sigma 0:27 1:36 \Sigma 0:32
1:63 \Sigma 0:39
NGC 1493 0 0 0 0 ­
NGC 2403 4 1 10 1 14 1 0:45 \Sigma 0:16 1
NGC 2835 7 2 9 0:43 \Sigma 0:22 ­
NGC 2997 20 14 34 0:25 \Sigma 0:10 0:28 \Sigma 0:11
0:31 \Sigma 0:12
NGC 3184 3 10 13 0:28 \Sigma 0:13 ­
NGC 3621 22 23 45 0:93 \Sigma 0:21 1:14 \Sigma 0:26
1:40 \Sigma 0:32
NGC 4395 0 2 2 0:21 \Sigma 0:15 ­
NGC 5204 0 7 7 1:49 \Sigma 0:63 ­
NGC 5236 55 96 151 1:75 \Sigma 0:28 1:98 \Sigma 0:32
3:29 \Sigma 0:58
NGC 5585 1 8 9 0:57 \Sigma 0:22 ­
NGC 6744 12 6 18 0:43 \Sigma 0:22
NGC 6946 76 31 107 0:56 \Sigma 0:11 0:66 \Sigma 0:14
0:81 \Sigma 0:17
NGC 7424 7 3 10 0:15 \Sigma 0:09 ­
NGC 7741 0 0 0 0 ­
NGC 7793 12 8 20 1:21 \Sigma 0:35 1:33 \Sigma 0:39
1:62 \Sigma 0:47
(LMC) 1 7 8 0:57 ­
(M33) 0 1 1 0:04 ­
are from Bica et. al. (1996) and those for M33 are from
Christian & Schommer (1988).
The ``specific frequencies'' TN for the galaxies in our
sample are given in the fourth column of Table 2. The
number NYMC = NB +NR used to derive TN is the total
number of clusters, ``red'' and ``blue'', detected in each
galaxy.
The errors on TN were estimated taking into consider­
ation only the uncertainties of the absolute magnitudes
of the host galaxies resulting from the distance errors
as given in Table 1 and poisson statistics of the cluster
counts. However, it is clear that this is not a realistic esti­
mate of the total uncertainties of the TN values. Another
source of uncertainty arises from incompleteness effects,
particularly for the more distant galaxies. For all galax­
ies with more than 20 clusters we estimated the incom­
pleteness by adding artificial clusters with magnitudes of

6 Young massive star clusters..
Fig. 4. U­B vs. B­V diagrams for the clusters in six galaxies, compared to stellar models by Bertelli et. al. 1994 and evolutionary
synthesis models of stellar clusters.
18.0, 18.5 : : : 21.0, and testing how many of the artifi­
cially added clusters were detected by DAOFIND in all of
the filters U ,B and V . Because the completeness depends
critically on the size of the objects, we carried out com­
pleteness tests for artificial clusters with R e = 0 pc and
R e = 20 pc in each galaxy. The numbers of clusters actu­
ally detected in each of the magnitude bins [18.25 ­ 18.75],
[18.75 ­ 19.25] : : : [20.75 ­ 21.25] were then corrected by
the fraction of artificial clusters recovered in the corre­
sponding bin, and finally the ``corrected'' TN values were
derived. These are given in the last column of Table 2, la­
beled TN;C , for point sources (first line) and objects with
R e = 20 pc (second line). See Larsen (1999) for more de­
tails on the completeness corrections.
One additional source of errors affecting TN which re­
mains uncorrected, is the fact that an uncertainty in the

7
Fig. 5. Comparison between photometry of clusters in
NGC 1313 and M101. '+' markers indicate the HST photom­
etry of clusters in M101 by Bresolin et. al. 1996.
distance also affects the magnitude limit for detection of
star clusters . If a galaxy is more distant (or nearby) than
the value we have adopted, our limit corresponds to a
``too bright'' (too faint) absolute magnitude, and we have
underestimated (overestimated) the number of clusters.
Hence, the true TN errors are somewhat larger than those
given in Table 2, but they depend on the cluster luminosity
function. If the clusters follow a luminosity function of the
form OE(L)dL / L \Gamma1:78 dL (Whitmore & Schweizer 1995)
then a difference in the magnitude limit of \DeltaM V = 0:1
would lead to a difference in the cluster counts of about
7%.
A histogram of the uncorrected TN values (Fig. 3)
shows that a wide range of TN values are present within
our sample. Many of the galaxies in the lowest bins con­
tain only a few massive clusters or none at all, but a few
galaxies have much higher TN values than the average.
The most extreme TN values are found in NGC 1156,
NGC 1313, NGC 3621, NGC 5204 and NGC 5236. The
galaxy NGC 2997 also hosts a very rich cluster system,
but the TN value is probably severely underestimated be­
cause of the large distance of NGC 2997 which introduces
significant incompleteness problems. A similar remark ap­
plies to two other distant galaxies observed at the Danish
1.54 m. telescope, NGC 1493 and NGC 7424, while all the
remaining galaxies in Table 1 are either more nearby, or
have been observed at the NOT in better seeing condi­
tions, and hence their TN values are believed to be more
realistic.
4.2. Two­colour diagrams
In Fig. 4 we show the B \Gamma V; U \Gamma B diagrams for six
cluster­rich galaxies. These plots also include the so­called
``S'' sequence defined by Girardi et. al. (1995, see also
Elson & Fall (1985)), represented as a dashed line. The
``S'' sequence is essentially an age sequence, derived as a
fit to the average colours of bright LMC clusters in the
U \Gamma B; B \Gamma V diagram. The age increases as one moves
along the S­sequence from blue to red colours. The colours
of our cluster candidates are very much compatible with
those of the S­sequence, especially if one considers that
there is a considerable scatter around the S­sequence also
for Magellanic Cloud clusters (Girardi et. al. 1995). Also
included in the diagrams are stellar models by Bertelli et.
al. (1994) (dots), in order to demonstrate that the posi­
tion of clusters within such a diagram is distinctly different
from that of single stars. Already from Fig. 4 one can see
that there is a considerable age spread among the clusters
in each galaxy, with the red cut­off being due to our se­
lection criteriae. The reddening vector corresponding to a
reddening of E(B \Gamma V ) = 0:30 is shown in each plot as
an arrow, and it is quite clear that the spread along the
S­sequence cannot be entirely due to reddening effects.
Bresolin et. al. (1996) used HST data to carry out pho­
tometry for star clusters in the giant Sc­type spiral M101.
A comparison between their data and photometry for clus­
ters in one of our galaxies (NGC 1313) is shown in Fig. 5.
It is evident that the colours of clusters in the two galaxies
are very similar. NGC 1313 was chosen as an illustrative
example because it contains a relatively rich cluster sys­
tem, although not so rich that the diagram becomes too
crowded.
In Fig. 5 we have also included a curve showing
the colours of star clusters according to the popula­
tion synthesis models of Bruzual & Charlot (1993, here­
after BC93). The agreement between the synthetic and
observed colours is very good for U­B ? \Gamma0:3, but for
U­B ! \Gamma0:3 the B­V colours of the BC93 models are sys­
tematically too blue compared to our data and the S se­
quence. The ``red loop'' that extends out to B­Vú0.3 and
U­Bú­0.5 is due to the appearance of red supergiants at
an age of about 10 7 years (Girardi & Bica 1993) and
is strongly metallicity dependent. Girardi et. al. (1995)
constructed population synthesis models based on a set
of isochrones by Bertelli et. al. (1994) and found very
good agreement between the S­sequence and their syn­
thetic colours. The models (solar metallicity) are included
in Fig. 5 as a solid line. In these models the ``red loop''
is not as pronounced as in the BC93 models, and the
youngest models are in general not as blue as those of
BC93, resulting in a much better fit to the observed clus­
ter colours.
4.3. Ages and masses
A direct determination of the mass of an unresolved
star cluster requires a knowledge of the M/L ratio, which
in turn depends on many other quantities, in particular
the age and the IMF of the cluster. However, if one as­
sumes that the IMF does not vary too much from one

8 Young massive star clusters..
Fig. 6. Absolute visual magnitudes versus ages for the clusters in six galaxies, compared with data for open clusters in the
Milky Way (Lyngša 1982). Population synthesis models by Bruzual & Charlot (1993) for three different IMFs, corresponding to
a mass of 10 5 M fi , have been included in each plot.
star cluster to another, then the luminosities alone should
facilitate a comparison of star clusters with similar ages.
Applying the S­sequence age calibration to the clusters
in our sample, the luminosities of each cluster can then be
directly compared to Milky Way clusters of similar age, as
shown in Fig. 6. Ages and absolute visual magnitudes for
Milky Way open clusters are from the Lyngša (1982) cata­
logue, and are represented in each plot as small crosses. In
the diagrams in Fig. 6 we have also indicated the effect of a
reddening of E(B­V) = 0.30. In these plots the ``reddening
vector'' depends in principle on the original position of the
cluster within the (U­B,B­V) diagram from which the age
was derived, but we have included two typical reddening
vectors, corresponding to two different ages.

9
In all of the galaxies in Fig. 6 but NGC 2403, the ab­
solute visual magnitudes of the brightest clusters are 2 ­
3 magnitudes brighter than the upper limit of Milky Way
open clusters of similar ages. Accordingly they should be
nearly 10 times more massive. In the case of NGC 2403,
the most massive clusters are not significantly more mas­
sive than open clusters found in the Milky Way. Fig. 6 also
confirms the suspicion that the cluster data in NGC 2997
are incomplete, particularly for M V ? \Gamma10.
We have included population synthesis models for the
luminosity evolution of single­burst stellar populations of
solar metallicity by BC93 in Fig. 6, scaled to a total mass
of 10 5 M fi . Models for three different IMFs are plotted:
Salpeter (1955), Miller­Scalo (1979) and Scalo (1986), all
covering a mass range from 0.1 ­ 65 M fi . The different
assumptions about the shape of the IMF obviously affect
the evolution of the M V magnitude per unit mass quite
strongly, and unfortunately the effect is most severe just
in the age interval we are interested in. The difference
between the Miller­Scalo and the Scalo IMF amounts to
almost 2 magnitudes, but in any case the most massive
clusters appear to have masses around 10 5 M fi .
In Fig. 6 we have also indicated the location of a ``typ­
ical'' old globular cluster system with an error bar cen­
tered on the coordinates 15 Gyr, M V = \Gamma7:4 and with
oe V = 1:2 mags. Although the comparison of masses at
high and low age based on population synthesis models
is extremely sensitive to the exact shape of the IMF, it
seems that the masses of the young massive star clusters
are at least within the range of ``true'' globular clusters.
Reddening effects alone are unlikely to affect the de­
rived ages to a high degree, as a scatter along the ``redden­
ing vectors'' in Fig. 6 would then be expected. Basically
this would mean that one would expect a much steeper
rate of decrease in M V vs. the derived age, while the ob­
served relation between age and the upper luminosity limit
is in fact remarkably compatible with that predicted by
the models. The comparison with model calculations im­
plies that the upper mass limit for clusters must have re­
mained relatively unchanged over the entire period during
which clusters have been formed in each galaxy.
5. Notes on individual galaxies
5.1. NGC 1156
This is a Magellanic­type irregular galaxy, currently un­
dergoing an episode of intense star formation. Ho et. al.
1995 noted that the spectrum of NGC 1156 resembles that
of the ``W­R galaxy'' NGC 4214. NGC 1156 is a completely
isolated galaxy, so the starburst could not have been trig­
gered by interaction with other galaxies. We have found a
number of massive star clusters in NGC 1156.
5.2. NGC 1313
This is an SB(s)d galaxy of absolute B magnitude MB =
\Gamma18:9. de Vaucouleurs 1963 found a distance modulus
of m \Gamma M = 28:2, which we adopt. The morphology of
NGC 1313 is peculiar in the sense that many detached
sections of star formation are found, particularly in the
south­western part of the galaxy. There is also a ``loop''
extending about 1.5 Kpc (projected) to the east of the
bar with a number of HII regions and massive star clus­
ters. Another interesting feature is that one can see an
extended, elongated diffuse envelope of optical light, with
the major axis rotated 45 ffi relative to the central bar of
NGC 1313, embedding the whole galaxy. It has been sug­
gested by Ryder et. al. (1995) that the diffuse envelope
surrounding NGC 1313 is associated with galactic cirrus
known to exist in this part of the sky (Wang & Yu 1995),
but this explanation does not seem likely since it would
require a very perfect alignment of the centre of NGC 1313
with the diffuse light. Also, the outer boundary of the ac­
tive star­forming parts of galaxy coincide quite well with
the borders of the more luminous parts of the envelope. In
our opinion the most likely explanation is that the diffuse
envelope is indeed physically associated with NGC 1313
itself.
Walsh & Roy (1997) determined O/H abundances for
33 HII regions in NGC 1313, and found no radial gradient.
This makes NGC 1313 the most massive known barred
spiral without any radial abundance gradient.
NGC 1313 hosts a rich population of massive star clus­
ters. When looking at the plot in Fig. 6 it seems that there
is a concentration of clusters at log(Age) ú 8.3 or roughly
200 Myr. We emphasize that this should be confirmed
by a more thorough study of the cluster population in
this galaxy, and in particular it would be very useful to
be able to detect fainter clusters in order to improve the
statistics. If this is real it could imply that some kind of
event stimulated the formation of massive star clusters in
NGC 1313 a few hundred Myr ago, perhaps the accretion
of a companion galaxy. A second ``burst'' of cluster forma­
tion seems to have been taking place very recently, and is
maybe going on even today.
5.3. NGC 2403
NGC 2403 is a nearby spiral, morphologically very sim­
ilar to M33 apart from the fact that NGC 2403 lacks a
distinct nucleus. It is a textbook example of an Sc­type
spiral, and it is very well resolved on our NOT images. A
photographic survey of star clusters in NGC 2403 was al­
ready carried out by Battistini et. al. 1984, who succeeded
in finding a few YMC candidates. NGC 2403 spans more
than 20 \Theta 20 arcminutes in the sky, so we have been able
to cover only the central parts using the ALFOSC. Within
the central 6:5 0 \Theta 6:5 0 (about 6\Theta6 Kpc) we have located
14 clusters altogether, but the real number of clusters in

10 Young massive star clusters..
NGC 2403 should be significantly higher, taking into ac­
count the large fraction of the galaxy that we haven't cov­
ered, and considering the fact that in the other galaxies
we have studied, many clusters are located at considerable
distances from the centre.
5.4. NGC 2997
NGC 2997 is an example of a ``hot spot'' galaxy (Meaburn
et. al. 1982) with a number of UV luminous knots near
the centre. Walsh et. al. (1986) studied the knots and con­
cluded that they are in fact very massive star clusters, and
Maoz et. al. (1996) further investigated the central region
of NGC 2997 using the HST. On an image taken with
the repaired HST through the F606W filter they identi­
fied 155 compact sources, all with diameters of a few pc.
Of 24 clusters detected in the F606W filter as well as in an
earlier F220W image, all have colours implying ages less
than 100 Myr and masses – 10 4 M fi . Maoz et. al. (1996)
conclude that the clusters in the centre of NGC 2997 will
eventually evolve into objects resembling globular clusters
as we know them in the Milky Way today.
In our study we have found a number of massive star
clusters also outside the centre of NGC 2997. Taking the
numbers at face value, the cluster system does not appear
to be as rich as that of NGC 5236, but with better and
more complete data we would expect to see a number of
YMCs in NGC 2997 that could rival that in NGC 5236.
5.5. NGC 3621
This galaxy is at first sight a quite ordinary late­type spi­
ral, and has not received much attention. It was observed
with the HST by Rawson et. al. (1997) as part of the Ex­
tragalactic Distance Scale Key Project, and cepheids were
discovered and used to derive a distance modulus of 29.1.
Our data show that NGC 3621 contains a surprisingly
high number of massive star clusters. The galaxy is rather
inclined (i = 51 ffi , Rawson et. al. 1997), and nearly all the
clusters are seen projected on the near side of the galaxy,
so a number of clusters on the far side may be hidden
from our view. Ryder & Dopita (1993) noted a lack of HII
regions on the far side of the galaxy, and pointed out that
there is also a quite prominent spiral arm on the near side
that doesn't appear to have a counterpart on the far side.
So it remains possible that the excess of young clusters
and HII regions on the near side is real.
5.6. NGC 5204
NGC 5204 is a companion to the giant Sc spiral M101.
The structure of the HI in this galaxy is that of a strongly
warped disk (Sicotte & Carignan 1997), and one could
speculate that this is related to tidal interaction effects
with M101. Sicotte & Carignan (1997) also find that the
dark matter halo of NGC 5204 contributes significantly to
the mass even in the inner parts.
The high TN value of this galaxy is a consequence of
its very low MB rather than a high absolute number of
clusters ­ we found only 7 clusters in this galaxy. Curiously,
all of the 7 clusters belong to the ``red'' class, suggesting
that no new clusters are being formed in NGC 5204 at the
moment.
5.7. NGC 5236
NGC 5236 (M83) is a grand­design barred spiral of type
SBc, striking by its regularity and its very high surface
brightness ­ the highest among the galaxies in our sample.
The absolute visual magnitude is M V = \Gamma20:0 (de Vau­
couleurs et. al. 1983). NGC 5236 is currently undergoing
a burst of star formation in the nucleus as well as in the
spiral arms.
A study in the rocket UV (Bohlin et. al. 1990) has
already revealed the presence of a number of very young
massive star clusters inside the HII regions of NGC 5236,
and HST observations of the nucleus (Heap et. al. 1993)
showed an arc of numerous OB clusters near the centre
of the galaxy. These clusters were found to have abso­
lute visual magnitudes in the range from M V = \Gamma10:4 to
M V = \Gamma13:4. and typical radii of the order of 4 pc. Masses
were estimated to be between 10 4 and 10 5 M fi .
Our investigation adds a large number of massive star
clusters in NGC 5236 also outside the centre and the HII
regions. In terms of absolute numbers the cluster system
of NGC 5236 is by far the richest in our sample, and in
particular there is a large number of clusters in the ``red''
group. This may be partly due to reddening effects al­
though Fig. 6 shows that there is in fact a large intrinsic
age spread among the clusters in NGC 5236.
5.8. NGC 6946
The study of NGC 6946 is complicated by the fact that it
is located at low galactic latitude (b = 12 ffi ), and there is
an interstellar absorption of AB = 1:6 magnitudes and a
large number of field stars towards this galaxy. NGC 6946
is nevertheless a well­studied galaxy, and we also chose
to include it in our sample, reasoning that star clusters
should be recognizable as extended objects on the NOT
data.
The chemical abundances of HII regions in NGC 6946
were studied by Ferguson et. al. (1998), who concluded
that their data were consistent with a single log­linear
dependence on the radius. At 1.5­2 optical radii (defined
by the B­band 25th magnitude isophote) they measured
abundances of O/H of about 10%­15% of the solar value,
and N/O of about 20% ­ 25% of the solar value.
Among the approximately 100 clusters we have identi­
fied in NGC 6946, one stands out as particularly striking
(Fig. 7). This cluster is apparently a very young object,

11
Fig. 7. A V­band image of NGC 6946. The cluster discussed
in the text is the luminous object located to the lower left of
the centre of the bubble­like structure.
located in one of the spiral arms at a distance of 4.4 Kpc
from the centre, and with an impressive visual luminos­
ity of M V = \Gamma13. Using a deconvolution­like algorithm
(Larsen 1999), the effective radius was estimated to be
about 15 pc. The cluster is located within a bubble­like
structure with a diameter of about 550 pc, containing nu­
merous bright stars and perhaps some less massive clus­
ters. On optical images this structure is very conspicuous,
but it is not visible on the mid­IR ISOCAM maps by Mal­
hotra et. al. (1996). There are no traces of Hff emission
either, except for a small patch at the very centre of the
structure.
5.9. LMC and M33
For these galaxies, we have adopted data from the litera­
ture.
As mentioned in the introduction, both the LMC and
M33 contain young star clusters that are more massive
than the open clusters seen in the Milky Way. However,
as is evident from Table 2, only one cluster in M33 is a
YMC according to our criteriae. The LMC, on the other
hand, contains a relatively rich cluster population, with
7 clusters in the ``red'' group and 1 cluster in the ``blue''
group. The cluster R136 in the 30 Doradus nebula of the
LMC has not been included in the data for Table 2 because
of its location within a giant HII region. Compared to
the other galaxies in our sample, the LMC ranks among
the relatively cluster­rich ones, but it is also clear that a
cluster population like the one of the LMC is by no means
unusual.
Because the LMC is so nearby, the limiting magnitude
for detection of clusters is obviously much fainter than
in the other galaxies in our sample, and the Bica et. al.
(1996) catalogue should certainly be complete down to our
limit of M V = \Gamma8:5, corresponding to V = 10:25 (taking
into account an absorption of about 0.25 mags. towards
the LMC). If the LMC was located at the distance of most
of the galaxies in our sample we would probably not have
detected 8 clusters, but a somewhat smaller number, and
the TN value would have been correspondingly lower. This
should be kept in mind when comparing the data for the
LMC with data for the rest of the galaxies in the sample.
6. Radial density profiles of cluster systems
As an attempt to investigate how cluster formation
correlates with the general characteristics of galaxies, we
have compared the surface densities of YMCs (number
of clusters per unit area) as a function of galactocentric
radius with the surface brigthness in U , V , I and Hff.
Obviously, such a comparison only makes sense for rela­
tively rich cluster systems, and is shown in Fig. 8 for four
of the most cluster­rich galaxies in our sample. We did not
include data for the apparently quite cluster­rich galaxy
NGC 6946 in Fig. 8 because of the numerous Galactic
foreground stars in the field of this galaxy which make
the cluster identifications less certain.
The surface brightnesses were measured directly on
our CCD images using the phot task in DAOPHOT. In
the case of Hff we used continuum­subtracted images, ob­
tained by scaling an R­band frame so that the flux for
stellar sources was the same in the R­band and Hff im­
ages, and subtracting the scaled R­band image from the
Hff image. The flux was measured through a number of
apertures with radii of 50, 100, 150 : : : pixels, centered on
the galaxies, and the background was measured in an an­
nulus with an inner radius of 850 pixels and a width of 100
pixels. The flux through the i'th annular ring was then cal­
culated as the flux through the i'th aperture minus the flux
through the (i­1)'th aperture, and the surface brightness
was finally derived by dividing with the area of the i'th
annular ring. No attempt was made to standard calibrate
the surface brightnesses, so the y­axis units in Fig. 8 are
arbitrary. The cluster ``surface densities'' were obtained
by normalising the number of clusters within each annu­
lar ring to the area of the respective rings. Finally, all
profiles were normalised to the V­band surface brightness
profile.
For all the galaxies in Fig. 8 the similarity between the
surface brightness profiles and the cluster surface densities
is quite striking. In the cases of NGC 2997 and NGC 5236,
where the Hff profiles are markedly different from the
broad­band profiles, the cluster surface densities seem to
follow the Hff profiles rather than the broad­band pro­
files. Accordingly the presence of massive clusters must be
closely linked with the process of star formation in gen­

12 Young massive star clusters..
Fig. 8. Radial cluster distributions compared with surface brightness profiles in U , V , I and Hff. The dots with error bars show
the ``surface density'' of clusters, and length of the error bars correspond to poisson statistics.
eral in those galaxies where YMCs are present. In order
to get a complete picture one should include the clusters
in the central starbursts of NGC 2997 and NGC 5236, but
this would, in any case, affect the conclusions only for the
innermost bin.
7. Discussion
Perhaps the most striking fact about the cluster­rich
galaxies in our sample is that they do not appear to have
a lot of other properties in common. Fig. 9 shows the spe­
cific frequency TN as a function of the ``T''­type (Table 1),
and does not support the suggestion by Kennicutt & Chu
(1988) that the presence of YMCs in galaxies increases
along the Hubble sequence. Instead, at wide range of TN
values is seen independently of Hubble type, so even if
YMCs might be absent in galaxies of even earlier types
than we have studied here the phenomenon cannot be en­
tirely related to morphology.
However, what characterises all these cluster systems
is that they do not seem to have been formed during one
intense burst of star formation. Instead, their age distribu­
tions as inferred from the ``S'' sequence are quite smooth
Fig. 9. TN values as a function of ``T''­type.
(possibly with the exception of NGC 1313), so in contrast
to starburst galaxies like the Antennae or M82, the rather
``normal'' galaxies in our sample have been able to main­
tain a ``production'' of clusters over a longer timescale, at
least several hundred Myr, in a more quiescent mode than
that of the starburst galaxies. The most luminous clusters

13
we have found have absolute visual magnitudes of about
M V = \Gamma12, about three magnitudes brighter than the
brightest open clusters in the Milky Way, but still some­
what fainter than the M V = \Gamma13 to M V = \Gamma15 clusters
in the Antennae and certain starburst galaxies.
One notable exception is NGC 6946 which is form­
ing such a ``super star cluster'' just before our eyes. That
cluster is located far away from the centre of the galaxy,
something which is not unusual at all. Also in NGC 1313
the most massive cluster is located far from the centre of
the host galaxy, at a projected galactocentric distance of
about 3.7 kpc, and in the Milky Way a number of high­
mass (old) open clusters are found in the anticentre di­
rection, e.g. M67. It can of course not be excluded that a
massive cluster like the one in NGC 6946 could be located
in a region of the Galactic disk hidden from our view,
but in any case the Milky Way does not seem to contain
any large number of young massive clusters as seen e.g. in
NGC 5236 or NGC 1313.
In general we find, however, that the distribution of
YMCs follows the Hff surface brightness profile, at least
for those galaxies where the statistics allow such a compar­
ison. Taking Hff as an indicator of star formation, it then
appears that in certain galaxies the formation of YMCs
occurs whenever stars are formed. This raises the ques­
tion whether the presence of massive cluster formation is
correlated with global star formation indicators, such as
Hff luminosity or other parameters. These questions will
be addressed in more detail in a subsequent paper (Larsen
et. al. 1999).
Two of the galaxies in our sample, NGC 5236 and
NGC 2997, have a lot of properties in common. Both
galaxies are grand­design, high surface­brightness spirals
although NGC 2997 lacks the impressive bar of NGC 5236,
and both were known to contain massive star clusters near
their centres also before this study. We have identified rich
cluster system throughout the disks of these two galaxies.
In our opinion it is becoming clearer and clearer that
a whole continuum of cluster properties (age, mass, size)
must exist, one just has to look in the right places. For
some reason the Milky Way and many other galaxies were
only able to form very massive, compact star clusters dur­
ing the early phases of their evolution, these clusters are
today seen as globular clusters in the halos of these galax­
ies. Other galaxies such as the Magellanic Clouds, M33
and NGC 2403 are able to form substantially larger num­
ber of massive clusters than the Milky Way even today,
and in our sample of galaxies we have at least 5 galax­
ies that are able to form clusters whose masses reach well
into the interval defined by the globular clusters of the
Milky Way. Still more massive clusters are being formed
today in genuine starburst and merger galaxies such as
the Antennae, NGC 7252, M82 and others, and it seems
that the masses of these clusters can easily compete with
those of ``high­end'' globular clusters in the Milky Way.
Whether YMCs will survive long enough to one day be
regarded as ``true'' globular clusters is still a somewhat
controversial question, whose definitive answer requires a
detailed knowledge of the internal structure of the indi­
vidual clusters and a better theoretical understanding of
the dynamical evolution of star clusters in general.
One could also ask if the LF of star clusters really has
an upper cut­off that varies from galaxy to galaxy, or if the
presence of massive clusters is merely a statistical effect
that follows from a generally rich cluster system. In order
to investigate this question it is necessary to obtain data
with a sufficiently high resolution that the search for star
clusters can be extended to much fainter magnitudes than
we have been able to do in our study.
8. Conclusions
The data presented in this paper demonstrate that mas­
sive star clusters are formed not only in starburst galaxies,
but also in rather normal galaxies. None of the galaxies
in our sample show obvious signs of having been involved
in interaction processes, yet we find that there is a large
variation in the specific frequency TN of massive clusters
from one galaxy to another. Some of the galaxies in our
sample (notably NGC 1313 and NGC 5236) have consider­
ably higher TN than the LMC, while other galaxies which
at first glance could seem in many ways morphologically
similar to the LMC (e.g. NGC 300 and NGC 4395) turn
out to contain no rich cluster systems. In general there
is no correlation between the morphological type of the
galaxies in our sample and their TN values. Whether a
galaxy contains massive star clusters or not is therefore
not only a question of its morphology (as suggested by
Kennicutt & Chu 1988), so one has to search for correla­
tions between other parameters and the TN values. Within
each of the galaxies that contain populations of YMCs,
the number of clusters as a function of radius follows the
Hff surface brightness more closely than the broad­band
surface brightness, which implies that the formation of
massive clusters in a given galaxy is closely linked to star
formation in general.
Acknowledgements. This research was supported by the Dan­
ish Natural Science Research Council through its Centre for
Ground­Based Observational Astronomy. This research has
made use of the NASA/IPAC Extragalactic Database (NED)
which is operated by the Jet Propulsion Laboratory, Califor­
nia Institute of Technology, under contract with the National
Aeronautics and Space Administration. We are grateful to J.V.
Clausen for having read several versions of this manuscript, and
the DFG Graduierten Kolleg ``Das Magellansche System und
andere Zwerggalaxien'' is thanked for covering travel costs to
S.S. Larsen.
References
Ashman K.M., Zepf S.E. 1998, ''Globular Cluster Systems''
Cambridge University Press

14 Young massive star clusters..
Barth A.J., Ho L.C., Filippenko A.V., Sargent W.L.W. 1995,
AJ 110, 1009
Battistini P., B`onoli F., Federici L., Fusi Pecci F., Kron R.G.
1984, A&A, 130, 162
Bertelli G. et. al., 1994, A&AS 106, 275
Bica E., Clari'a J.J., Dottori H., Santos J.F.C. Jr., Piatti A.E.
1996, ApJS 102, 57
Bohlin R.C. et. al. 1990, ApJ 363, 154
Bottinelli L., Gouguenheim L., Paturel G., de Vaucouleurs G.,
1985, ApJS 59, 293
Brandl B. et. al. 1996, ApJ 466, 254
Bresolin F., Kennicutt R.C., Stetson P.B., 1996, AJ 112, 1009
Bruzual G.A., Charlot S. 1993, ApJ 405, 538
Carignan C. 1985, ApJS 58, 107
Christian C.A. & Schommer R.A. 1988, AJ 95, 704
Elmegreen B.G., Efremov Y.N. 1997, ApJ 480, 235
Elson R.A.W., Fall S.M. 1985, ApJ 299, 211
Ferguson A.M.N., Gallagher J.S., Wyse R.F.G. 1998, AJ 116,
673
Filippenko A.V., Sargent W.L.W. 1989, ApJ 342, L11
Fischer P., Welch D.L., Cote P., Mateo M., Madore B.F. 1992,
AJ 103, 857
Freedman W.L., Madore B.F. 1988, ApJ 332, L63
Freedman K.C. et. al. 1992, ApJ 396, 80
Fritze­v. Alvensleben U. 1998, A&A 336, 83
Girardi L., Bica F. 1993, A&A 274, 279
Girardi L., Chiosi C., Bertelli G., Bressan A. 1995, A&A 298,
87
Gnedin O.Y., Ostriker J.P. 1997, ApJ 474, 223
Harris W.E., van den Bergh S. 1981, AJ 86, 1627
Heap S.R., Holbrook J., Malumuth E., Shore S., Waller W.
1993, BAAS 182
Ho L.C., Filippenko A.V., Sargent W.L.W. 1995, ApJS 98, 477
Holtzman J.A. et. al. 1992, AJ 103, 691
Holtzman J.A. et. al. 1996, AJ 112, 416
Karachentsev I.D., Drozdovsky I.O. 1998, A&AS 131, 1
Karachentsev I., Musella I., Grimaldi A. 1996, A&A 310, 722
Kennicutt R.C., Chu Y­H. 1988, AJ 95, 720
Landolt A.U. 1992, AJ 104, 340
Larsen S.S. et. al. 1999, A&A, in preparation
Larsen S.S. 1999, A&A, in preparation
Lyngša G. 1982, A&A 109, 213
Malhotra S. et. al., 1996, A&A 315, L161
Maoz D. et. al. 1996, AJ 111, 2248
Meaburn J., Terrett D.L. 1982, MNRAS 200, 1
Miller G.E. & Scalo J.M. 1979, PASP 90, 506
O'Connell R.W., Gallagher III J.S., Hunter D.A. 1994, AJ 433,
65
O'Connell R.W., Gallagher III J.S., Hunter D.A., Colley W.N.
1995: ApJ 446, L1
Rawson D.M. et. al. 1997, ApJ 490, 517
Richtler T. 1993, ``The Globular Cluster­Galaxy Connection''
eds. G.H. Smith & J.P. Brodie, ASP p. 375
Ryder S.D. & Dopita M.A. 1993, ApJS 88, 415
Ryder S.D., Staveley­Smith L., Malin D., Walsh W. 1995, AJ
109, 1592
Salpeter E.E. 1955, ApJ 121, 161
Sandage A. & Bedke J. 1994, ``The Carnegie Atlas of Galaxies''
Carnegie Institution of Washington
Scalo J.M. 1986, Fundamentals of cosmic physics 11, 1
Schweizer F., Miller B.W., Whitmore B.C., Fall M.S. 1996, AJ
112, 1839
Shanks T. 1997, MNRAS 290, L77
Sicotte V., Carignan C. 1997, AJ 113, 609
van den Bergh S. 1991, ApJ 369, 1
van den Bergh S. 1995, ApJ 450, 27
de Vaucouleurs G., 1963, ApJ 137, 720
de Vaucouleurs G., 1979a, AJ 84, 1270
de Vaucouleurs G., 1979b, ApJ 227, 729
de Vaucouleurs G., Pence W.D., da Voust E. 1983, ApJS 53,
17
de Vaucouleurs G., de Vaucouleurs A., Corwin H.G., Buta R.J.,
Paturel G., Fouqu'e P. 1991, ``Third Reference Catalogue of
Bright Galaxies'', Springer­Verlag New York
Walsh J.R., Nandy K., Thompson G.I., Meaburn J., 1986, MN­
RAS 220, 453
Walsh J.R., Roy J.­R. 1997, MNRAS 288, 726
Wang O.D., Yu K.C. 1995, AJ 109, 698
Watson et. al. 1996, AJ 112, 534
Whitmore B.C., Schweizer F., Leitherer C., Borne K., Robert
C., 1993, AJ 106, 1354
Whitmore B.C., Schweizer F., 1995, AJ 109, 960
Worthey G. 1994, ApJS 95, 107
¨
Ostlin G., Bergvall N., R¨onnback J. 1998, A&A 335, 85