Star Clusters as Tracers of Interactions in Stephan's Quintet (Hickson Compact Group 92)
K. Fedotov, S. C. Gallagher, I. S. Konstantopoulos, R. Chandar, N. Bastian, J. C. Charlton, B. Whitmore, G. Trancho
aa r X i v : . [ a s t r o - ph . C O ] M a y Star Clusters as Tracers of Interactions in Stephan’s Quintet (HicksonCompact Group 92) ∗ K. Fedotov , S. C. Gallagher , I. S. Konstantopoulos , R. Chandar , N. Bastian , J. C. Charlton B. Whitmore , & G. Trancho ABSTRACT
Stephan’s Quintet (SQ; also known as Hickson Compact Group 92) is a compactgroup of galaxies that exhibits numerous signs of interactions between its members.Using high resolution (0 . ′′
04 per pixel) images of Stephan’s Quintet in B , V , and I bands from the Early Release Science project obtained with the Wide Field Cam-era 3 on the Hubble Space Telescope , we identify 496 star cluster candidates (SCCs),located throughout the galaxies themselves as well as in intergalactic regions. Ourphotometry goes ∼ mag deeper and covers an additional three regions, the Old Tail,NGC 7317, and the Southern Debris Region, compared to previous work. Throughcomparison of the B − V and V − I colors of the star cluster candidates * Based on observations made with the NASA/ESA Hubble Space Telescope. Department of Physics & Astronomy, The University of Western Ontario, London, ON, N6A 3K7, CANADA Department of Astronomy and Astrophysics, The Pennsylvania State University, University Park, PA, 16802,USA University of Toledo, Toledo, OH, 43606, USA School of Physics, University of Exeter, Stocker Road, Exeter, EX4 4QL, UK Space Telescope Science Institute, Baltimore, MD, 21218, USA Gemini Observatory, Casilla 603, La Serena, Chile Myr. NGC 7318A/B also features a peculiar gap in the color distribution of the star clusters that canbe used to date the onset of the recent burst. The majority of the SCCs detected in theYoung Tail were formed — Myr ago whereas the tight distribution of star clus-ter colors in the Old Tail, allow us to constrain its age of formation to ∼ Myr ago.The star clusters in the Southern Debris region are seemingly divided into two groupswith ages of and ∼ Myr and virtually all of the SCCs detected in NGC 7317are over Gyr old. Based on these ages, we estimate time intervals for the interactionsbetween Stephan’s Quintet members that triggered the massive star cluster formation.
Subject headings: galaxies: clusters : individual : Stephan’s Quintet — galaxies:clusters : general — galaxies: evolution — galaxies: interactions — galaxies: starclusters
1. Introduction
Stephan’s Quintet (hereafter SQ; Stephan 1877), is one of the most studied galaxy groups.There are numerous signs of interactions between the galaxies in SQ, such as galaxy distortions,tidal tails, and active star formation in intergalactic regions, that have happened in the past or areongoing. However, we still lack a full understanding of the dynamical processes that led to the 3 –current spatial distribution of galaxies in SQ. In Figure 1, we present multi-band Hubble SpaceTelescope imaging of SQ, which covers four galaxies: two spirals (NGC 7319, NGC 7318B) andtwo ellipticals (NGC 7318A, NGC 7317). In Hickson’s designation those galaxies are HCG 92C,-B, -D and -E, respectively. Another spiral galaxy (lower left of the frame), NGC 7320, is known tobe a foreground galaxy (Allen & Sullivan III 1980, Moles et al. 1997), with a recessional velocity v R = 739 km s − (Falco et al. 1999). We do not examine this galaxy in our study. The core of thegroup consists of three galaxies NGC 7317, NGC 7319, and NGC 7318A, with approximately thesame v R (6637, 6652, and 6671 km s − , respectively; Table 1). The fourth galaxy, NGC 7318B,is a high-speed intruder (with v R = 5766 km s − ; Table 1) that is apparently interacting with thecore for the first time (Mendes de Oliveira & Hickson 1994; Moles et al. 1997).The notion that galaxy interactions trigger star cluster formation is the back-bone of thiswork. Note, however, that although mergers and interactions seem to be necessary for enhancedstar formation, they are not a sufficient condition (Bergvall et al. 2003). In the environment ofcompact groups of galaxies, which combines high densities (comparable to the number densitiesin the central regions of galaxy clusters) with low velocity dispersions ( σ ≈ — km s − ), wewould expect multiple interactions throughout the history of the group resulting in the productionof multiple populations of star clusters. Thus, the history of those interactions can be deduced fromdetailed studies of the populations of star clusters (Gallagher et al. 2001 and references within).Gallagher et al. (2001) WFPC2 study, which covered the Young Tail, NGC 7319, and most ofNGC 7318A/B system (Figure 1 of their paper), found 115 star cluster candidates. The majorityof the candidates were detected not in the central regions of the studied galaxies (NGC 7319,NGC 7318A/B), but in the tidal debris associated with those galaxies and in the Northern StarBurst Region (which we shall describe in greater detail subsequently). They also identified several 4 –epochs of recent star formation in SQ, spanning a large range of ages, from the ∼ — Myr oldclusters in the Northern Star Burst Region to the older population of — Gyr old clusters spreadover the entire field of view.The current study differs from the previous one in its use of a significantly wider field of view,focusing on all of the four related SQ galaxies (including NGC 7317) and their surrounding intra-group medium, and higher sensitivity and spatial resolution, owing to the new Wide Field Camera3 (WFC3) on the Hubble Space Telescope (
HST ). These new observations are approximately 2mags deeper in the V-band as compared to the previous data of Gallagher et al. (2001). The fainterdetection limit leads to a larger number of detections, as well as smaller photometric errors.In order to focus our study on a single interaction at a time, we divide the field into eightregions. Three of those regions are named after the galaxies they contain: NGC 7317, NGC 7319,and NGC 7318A/B. The last region contains two galaxies due to their proximity to each other inthe plane of the sky. The size of the regions is chosen to include the light from the galaxies as theyappear on the V image where the contrast has been scaled to emphasize low surface brightnessstructures. Although the gravitational pull of these galaxies stretches beyond the defined borders,we expect this procedure will suffice to identify star cluster candidate populations with their parentgalaxies. In their work on the star clusters in nearby starburst galaxies de Grijs et al. (2003) haverequired star clusters to have the maximum projected distance of ∼ kpc from the 3 σ sky contourin order to still be associated with the given galaxy. If we would apply the same approach in ourcase, the above mentioned distance of ∼ kpc from the 3 σ sky contour will be well within ourdefined regions for each galaxy.Following the naming convention of Gallagher et al. (2001) and Sulentic et al. (2001), we also 5 –define four regions containing extragalactic features: two tidal tails and two star formation regions.The Old Tail region is located southeast of the center of the group. We can observe only a small partof it emerging from behind NGC 7320. The other tidal tail region, which we call the Young Tail( § v R = 5985 km s − ,suggesting that this galaxy is responsible for the creation of the tail (Moles et al. 1997). Two tidalarms are found north of NGC 7318A/B. Due to the active star formation in this region, especially inthe eastern arm, we call it the Northern Star Burst Region (NSBR; Gallagher et al. 2001). It hostsSQ-A (Xu et al. 1999), a strong H α -emitting region, which apparently coincides with the overlapregion of the two tidal arms. Another star formation region is located between the two galaxiesNGC 7318A and NGC 7317. This region, which we call the Southern Debris Region (SDR), isseldom mentioned in the literature as hosting ongoing star formation, although a few studies haveshown the presence of H I gas in its vicinity (Williams et al. 2002; Sulentic et al. 2001). The lastregion, the X-Ray Shock Front, is defined as a contour of soft X-Ray emission ( . — . keV) withthe value of . count pix − from the Chandra image in the eastern part of NGC 7318B. This shockwave is the result of a high-speed collision between NGC 7318B, which is blueshifted by ∼ km s − (Mendes de Oliveira & Hickson 1994) with regards to the average radial group velocity,and the cold intergalactic medium seen in H I (Shostak et al. 1984). The latter was presumablydeposited in that location by previous interactions in the group (Moles et al. 1997; Xu et al. 1999).Positional, morphological, photometric and redshift information on individual galaxies hasbeen included in Table 1. 6 –
2. Observations and cluster candidate selection2.1. Data
The data were obtained with the new WFC3 on
HST , as a part of the Early Release Science(ERS; proposal 11502, PI K.S. Noll). The original observations were carried out in six filters ofwhich we use F438W ( B ), F606W ( V ), F814W ( I ), with the occasional use of the F657Nand F658N combined image. The exposure times were , , and seconds for the B , V , I filters, respectively. At the distance to SQ of . Mpc (adopting H = 70 km s − Mpc − ), equivalent to a distancemodulus of . mag, one pixel on the WFC3 corresponds to ∼ . pc (0 . ′′
04 per pixel). Becausemost individual stars are not luminous enough to be detected at the distance of SQ, we use starclusters as a tracer of star formation.In this paper we used PSF fitting as a suitable method of obtaining photometry. This is testedand widely accepted method of getting photometry for unresolved star clusters (e. g. Gallagher et al.2010, Konstantopoulos et al. 2010) and is very useful in the varying background and crowded fieldconditions present in SQ.According to the WFC3 Instrument Handbook the average FWHM of the PSF for BVI filtersis 1.7 pixels. After image manipulations the size of PSF only gets larger. Thus, assuming thebest case scenario of FWHM being 1.7 pixels, in order for a star cluster to be larger than the PSFit should have physical size of ∼ pc whereas the average size of a star cluster is currently 7 –accepted as 4 pc (e. g., Barmby et al. 2006; Scheepmaker et al. 2007). Thus, the star clusters of SQare expected to be unresolved and should appear as point sources, hence the detection and carefulselection of point sources are essential to our study.Point sources were detected with the DAOFIND (Stetson 1987) task in
IRAF , with a . σ threshold on a median-divided image (we used a 13 ×
13 pixel smoothing window and dividedthe original image by the smoothed one). The detection process closely follows that described inGallagher et al. (2010).The PSF was constructed from bright, isolated and unsaturated stars with point-like radialcurves of growth. There were 27, 14, and 48 of such stars in B , V , and I filters, respec-tively. This procedure was repeated for each filter. Using the DAOFIND.ALLSTAR package in
IRAF the photometry for all point sources was obtained. The aperture corrections between 3 to 10pixels were calculated for all filters, as an average of the difference between magnitudes obtainedfrom aperture photometry with a 10-pixel aperture and PSF-magnitudes (calculated at 3 pixels) forall the PSF stars ( ∆ B → = 0 . mag, ∆ V → = 0 . mag, ∆ I → = 0 . mag). Usingthe on-orbit enclosed energy (EE) curves for WFC3 UVIS we calculated the 10-pixels to infinitycorrection as the difference between unity and the enclosed energy in the given aperture and wave-length ( ∆ B →∞ = 0 . mag, ∆ V →∞ = 0 . mag, ∆ I →∞ = 0 . mag). The photometryin each filter was corrected for foreground extinction E ( B − V ) = 0 . mag (Schlegel et al.1998), with values of A = 0 . mag, A = 0 . mag, and A = 0 . mag. For the finalcatalog, we require sources to be detected in all three broad-band filters, which eliminates most IRAF is distributed by the National Optical Astronomy Observatory, which is operated by the Association ofUniversities for Research in Astronomy (AURA) under cooperative agreement with the National Science Foundation.
In order to select star cluster candidates (SCC), we require sources to pass several additionalcriteria. First, we reject sources with colors B − V > . mag and V − I > . mag, inorder to eliminate most foreground stars in the Galaxy. In order to select point sources, we followthe prescription described in Rejkuba et al. (2005), selecting sources which have: (1) magnitudeerror σ mag ≤ . mag in all three bands, (2) the sharpness of the source (a measure of the relativewidth of a source with respect to that of the PSF) between − . and . in all three bands, and (3)the goodness of fit factor from PSF-fitting χ ≤ in I band. The use of I band for the χ filteris dictated by the fact that the PSF model is best determined in that band and there is no expectedcontamination from emission lines ( § M V of − mag, which should eliminate nearly all individual luminous stars in SQ (Whitmore et al. 1999;2010).Based on these criteria, we select sources in NGC 7317 region, in NGC 7319, inNGC 7318A/B, in the regions of the X-ray Shock Front, in the Young Tail, in the OldTail, in the Northern Star Burst Region (NSBR), and SCCs in NGC 7317 (see Table 2).
In order to test the completeness of our final list of SCCs, we used
DAOPHOT.ADDSTAR toadd 5000 artificial point sources to the image in the magnitude range of — mag, i.e. − . − . mag in absolute magnitude at the distance of SQ. We then followed the detectionalgorithm outlined in section 2.3. We found that for recovery rates of % and % the limitingmagnitudes were . mag and . mag in the B band, . mag and . mag in the V band, and . mag and . mag in the I band (after aperture and extinction corrections wereapplied).Though the quoted completeness levels are for the entire field, the star clusters are clearlyconcentrated within the galaxies, where the background can be high and variable. Nevertheless,the completeness levels within the regions defined in Figure 1 were ≥ % for a star cluster with M V = − , the cutoff for inclusion in our catalog. From a sister spectroscopic study to this that is being conducted in parallel (G. Trancho etal., in prep.), we obtained a list of objects that were found not to be associated with SQ, such asforeground stars and background quasars. The identifications of the objects were determined withspectra from GMOS on Gemini North (programs GN-2004B-DD-8 and GN-2006A-Q-38), basedon the measured v R . As can be seen from the M V vs. V − I color-magnitude diagram ofFigure 2, the contaminating objects have V − I colors redder than . mag. Although thespectroscopic sample of the study is luminosity-limited, the vast majority of contaminating objectswould only affect the old globular cluster bin (Figure 3). Also, the spatial distribution of the SCCsin our sample coincides with the host galaxies, and features in the intragroup medium (such astidal arms and star burst regions), as we would expect from legitimate star clusters. Additionally,we estimated the contamination from the Galactic foreground stars based on the results from the 10 –Besanc¸on models (Robin et al. 2003) in the direction of SQ up to the distance of 100 kpc. Themodel predicts ∼ foreground stars in the field of view of SQ ( . deg ) within the magnitudeand color ranges of SCCs. Hence, for the area of NGC 7319 region, which has the largest area ofall the regions, we expect to find about 5 foreground stars with apparent magnitudes ranging from to in all three bands.Taking into consideration all of the above, we are confident that our final list of star clustercandidates is not heavily contaminated by foreground/background objects. We compare the photometry of our SCCs with predictions from stellar synthesis models fromMarigo et al. (2008), assuming the Kroupa (1998) stellar initial mass function, with upper andlower masses of 0.15 and 100 M ⊙ , and Z ⊙ , as measured by Saracco & Ciliegi (1995) with theirX-ray study. However, these models only include stellar contributions, not contributions fromnebular emission lines. Nebular emission can be strong in regions with active star formation, andcan strongly affect measurements in the V filter (with H α , H β and [O III ] emission lines) andless so in the B filter (with the [O III ] and H β emission lines). Because of these contributions,the B − V colors appear redder and the V − I colors appear bluer (e. g. Vacca & Conti1992, Conti et al. 1996) than the colors from continuum emission alone. In order to account fornebular emission from the youngest star clusters, we include this contribution by calculating theexpected strength of the H α and H β emission lines from Starburst99 (Leitherer et al. 1999)for the same IMF and metallicity. The [O
III ] line strength is calculated as . × H β , which is the http://model.obs-besancon.fr/
11 –largest ratio observed for the ratio of [O
III ] to H α in the KISS sample of low-mass star forminggalaxies (Salzer et al. 2005).
3. Results and Discussion
Below we discuss the results obtained for each region of SQ, defined earlier in § M V < − mag) detected in SQ, and Figure 4 presentscolor-color plots for each of the defined regions, so that the reader can compare them side-by-side. To better understand the spatial distribution of star clusters with regard to their approximateages we also included zoomed-in images of the various regions in V band and their associatedcolor-color plots in Figures 5 - 9.Intrinsic reddening will mimic older cluster ages in broad-band colors, i.e. it will move thebroad-band colors redward along the model predictions. Since we do not have an extinction mapfor SQ we will not include the effects of extinction explicitly. Rather we will be making qualitativestatements based on the visual spatial distribution of the dust lanes. However, to indicate possibleeffects of intrinsic reddening we have included an A V = 1 mag Galactic reddening vector inall color-color and color-magnitude diagrams. This might not be appropriate for the SQ galaxiesconsidering their lower metallicities (Saracco & Ciliegi 1995), but is conservative and is widelyadopted in the literature. The derived ages therefore can be considered as upper limits; if localextinction is small the obtained ages are closer to the real ones, and if local extinction is largethen obtained ages are probably higher than they should be. The broken dot-dashed green linepresent on the color-color plots is drawn to roughly parallel the evolutionary tracks at a distance 12 –set to respect the width of the observed distribution of SCCs along the tracks. The spread in thedistribution includes both intrinsic and photometric scatter. The SCCs that are located to the leftof the broken dot-dashed green line are likely to be very young (less than Myr), due to nebularemission that affects their colors.For each of the regions we summarize the available literature trying to combine the theoreticalmodels and computer simulations, especially Renaud et al. (2010), the most recent and detailed dy-namical model to date, with our observations to construct a comprehensive history of interactionsin SQ during the last Gyr.
NGC 7317 is an elliptical galaxy and, as expected for a galaxy of that type, hosts mostly oldstar clusters. The distribution of these clusters in color space is concentrated near B − V = 1 . mag and V − I = 0 . mag (Figure 5). The minimal reddening that would be expectedin NGC 7317, as well as the spatial distribution of the clusters, which are concentrated around thecore of the galaxy, suggest that a large number of them can be considered globular clusters. The Old Tail is a low surface brightness tidal structure that originates presumably fromNGC 7319 (Renaud et al. 2010) and formed through the interaction with NGC 7320C (Moles et al.1997; Sulentic et al. 2001). The Old Tail is mostly obscured by NGC 7320 in our field of view,and as such it is heavily contaminated by the objects of the foreground galaxy NGC 7320. How- 13 –ever, a small unobstructed part of that tidal feature, which we are able to observe just southeast ofNGC 7320, appears to have minimal contamination from NGC 7320.Indeed, if star clusters of the Old Tail were part of the NGC 7320 then there should be someexplanation to such a high, asymmetric concentration of SCCs outside the galaxy (Figure 6). Themost plausible explanation would involve an interaction of NGC 7320 with another galaxy. How-ever, the H I study of SQ by Gutierrez et al. (2002) did not find any evidence (tidal tails, bridges,etc.) of ongoing or past interactions between NGC 7320 and NGC 7331, the closest galaxy (pro-jected separation is ∼ kpc) with similar recession velocity v R ∼ km s − . Moreover, theobservations of H I in NGC 7320 conducted by Williams et al. (2002) revealed the signature ofnormal disk rotation, without any signs of tidal distortion.To further test the membership of SCCs in the Old Tail region we looked at their size distri-bution. We used ISHAPE software (Larsen 1999) to determine the FWHM for all detected SCCsin the Old Tail and NGC 7320 with an observed magnitude range of V = 23 . — . mag. Thedistribution of radii for the Old Tail sources is quite narrow (from . — . pix), and is peaked atr ∼ . pix. In contrast, the NGC 7320 sources have a much broader distribution of radii rangingfrom . — . pix with a peak at r ∼ . pix. Given that the distance modulus of NGC 7320is . mag, the distribution likely contains a mix of stars and resolved clusters. A KolmogorovSmirnov test applied to the distributions of FWHM gave < . probability that both sets of SCCscame from the same distribution.All of the above suggests that the SCCs in question do belong to the Old Tail and contami-nation from NGC 7320 is minimal. Also, clusters in the Old Tail have a narrow range of colors(the color and spatial distributions are shown in Figure 6), with a mean B − V of . ± .
14 –mag (where the error is the standard deviation in the mean) and a mean V − I of . ± . mag. This is another sign of minimal contamination from the foreground galaxy, since variableextinction that is expected in spiral galaxies would tend to spread out the colors along the ex-tinction vector. The uniform and relatively young age distribution is also suggesting in situ starformation, meaning that the Old Tail was formed with a large amount of gas. Because the tidaltail environment is usually associated with little extinction, the predicted age that best matches themean colors is Myr. This is consistent with ages obtained by Xu et al. (2005): t ≃ . ± . yr,corresponding to the age range of — Myr, from their extinction-corrected UV and opticalcolors, although somewhat less consistent with ages & Myr, the estimate based on dynamicalarguments (Xu et al. 2005).We also tried to detect the Old Tail SCCs situated behind the foreground galaxy NGC 7320.Although we found a number of sources with a consistent spatial distribution and with similar col-ors to our uncontaminated Old Tail region, no definite results could be drawn because the projectedwidth of NGC 7320 is almost identical to the apparent width of the Old Tail and contamination bystar clusters or stars with the same colors in the foreground galaxy cannot be reliably determined.
The Young Tail is a tidal structure that is parallel to the Old Tail but has higher surface bright-ness. Clusters in the Young Tail tend to have bluer colors than those in the Old Tail, although thereare a few redder SCCs as well. The concentration of blue clusters (Figure 6) have mean colors of B − V = 0 . ± . mag and V − I = 0 . ± . mag. These mean colors most closelymatch a model age of Myr, indicating that the bulk of the clusters formed more recently than 15 –in the Old Tail. This age estimate is in good agreement with the values of ∼ Myr, obtained byGallagher et al. (2001) from their WFPC2 imaging study. Also, the sister spectroscopy study ledby G. Trancho (in preparation) constrained the ages of clusters in the Young Tail to . Myr.The redder, and hence presumably older clusters (3 clusters with ages ∼ Myr and 5 clusters & Gyr old), present in the Young Tail were plausibly deposited there from the disks or halosof the galaxies by the interaction responsible for the creation of this tidal feature; alternatively,they are a line-of-sight projection of a few old clusters that surround the nearby galaxies. Anotherpossibility is that these clusters are young and highly extincted, although this explanation is lesslikely as tidal tails typically have low extinction.Previous studies (Moles et al. 1997; Sulentic et al. 2001) credit a second pass of NGC 7320Cthrough NGC 7319 as the event that formed the Young Tail. However, there are a few problemswith this hypothesis. Figure 5 in Williams et al. (2002) presents the total H I column density dis-tribution in SQ. Based on that figure, the distribution of H I in the Young Tail is not continuous.Approximately two thirds of the optical tail that is closer to NGC 7319 is H I free; the last thirdcontains H I and the density contours of the gas follow the optical tail. However, at the end of theoptical tail the H I contours change direction very rapidly and align themselves along the North-South line. The authors point out a common tendency that is seen in numerical studies of galaxyinteractions: tidal tails are usually pointing to the source of disturbance. This happens due to theexchange of momentum between the intruder-galaxy and the tidal debris. From this observationand the H I distribution, they conclude that the H I associated with the last third of the optical partof the Young Tail cannot be primarily driven by NGC 7320C which lies to the East.In Sulentic et al. (2001), the observed lack of H I in the region of the tail closest to NGC 7319and the length of the feature are explained by a low-velocity and low inclination interaction be- 16 –tween NGC 7319 and NGC 7320C. However, Renaud et al. (2010) argue that in order to strip sucha large amount of material, both gaseous and stellar, the low mass NGC 7320C had to pass veryclose to the center of NGC 7319 where the density is high. Furthermore, they argue that becauseof dynamical friction, orbital decay after two encounters would be too large to allow NGC 7320Cto occupy its current, distant position. In order to verify this hypothesis, they conducted 96 simu-lations of double-encounters between NGC 7319 and NGC 7320C, varying the orbital parameterswithin reasonable values, and no model gave the observed separation between these galaxies.A similar point is argued by Xu et al. (2005). Based on the measured redshifts of NGC 7319and NGC 7320C they estimated that it would take more than Myr for NCG 7320C to get to itscurrent position after the encounter with NGC 7319. This is consistent with the Old Tail age but istoo old for the Young Tail. Instead, they propose the scenario where the Young Tail is formed by aclose encounter between NGC 7319 and NGC 7318A.Another possible argument against NGC 7320C being responsible for the formation of theYoung Tail is the appearance of that galaxy. Assuming this is the second interaction betweenthese galaxies, and taking into account the difference between the masses of these two galaxies(NGC 7319 is ∼ — times more massive; Renaud et al. 2010), and the disturbed appearanceof NGC 7319, we would expect NCG 7320C to look at least as disturbed as NGC 7319 or, in allprobability, to be severely tidally distorted. Although NGC 7320C does exhibit signs of past inter-actions (e. g., strong spiral arms) they are not as strong as in the case of NGC 7319. This, however,could be an effect of particular orbit/spin orientations. The models of the encounter between thetwo galaxies (mass ratio 4:1; not 30:1 as in the case of NGC 7319 and NGC 7320C) showed that,in the case of prograde (for the larger galaxy) and retrograde (for the smaller galaxy) encounter, thelarger galaxy can develop a tidal feature while the smaller one would be left seemingly unaffected 17 –(Renaud, 2010, private communication).Based on the reasons mentioned above, we agree with the conclusion of Renaud et al. (2010),from collisionless N-body simulations of SQ, that the Young Tail was formed by the interactionbetween NGC 7319 and NGC 7318A, and we further advance their scenario with a more robustage restriction for this event from the SCC population to — Myr ago.
Due to the spatial proximity of NGC 7318A and B, we have decided to analyze them as a unitrather than as individual galaxies.From the color-color diagram in Figure 7, we conclude that the NGC 7318A/B region exhibitsongoing star formation for the last
Myr. In the literature NGC 7318A is identified as anelliptical galaxy (e. g. Nilson 1973; de Vaucouleurs et al. 1991; Moles et al. 1998), which wouldexplain the relatively large number of old SCCs observed in this region. Another interesting featurepresent in the color-color plot for NGC 7318A/B (Figure 7) is a lack of detected SCCs between theages of
Myr and Gyr. Five out of the seven objects that are located in the box of . < V − I < . and . < B − V < . (i.e., approximately 1 Gyr old) are either coincident withbright sources in the H α images or located in the actively star-forming regions, making them verylikely to be reddened young clusters. Thus, there appears to be a well-pronounced gap in colorspace between the ages of ∼ Myr and Gyr. Moreover, the presence of this gap is somewhatunusual as recent studies (e. g. Gallagher et al. 2010, Konstantopoulos et al. 2010) have shown thatthe distribution of SCCs for spiral galaxies tends to be more continuous, spanning ages from a few s Myr up to Gyr. The above mentioned gap could be explained by the following reasoning. 18 –Firstly, as mentioned above, NGC 7318A is considered an elliptical galaxy so we would not expectit to contribute many younger clusters. In any case, the SCCs which are just under Gyr couldhave faded below the detection limit at such a distance. For a given mass, they are fainter thanthe young star clusters ( . a few tens Myr) by ∼ mags and even a small amount of extinctionmay either push them below our detection limit or redden them, and so they will appear as olderclusters ( > Gyr). Thirdly, previous work (e. g. Moles et al. 1997) has found that NGC 7318B isinteracting with SQ for the first time. They draw this conclusion mainly from two observations:the v R of NGC 7318B with respect to SQ ( ∆ v ∼ km s − ) and its mostly unperturbed spiralstructure, a sign that this galaxy has not experienced many interactions in its recent history. Thus,considering that galaxy interactions usually trigger star formation, without recent interactions withother galaxies, NGC 7318B will be expected to have a small number of SCCs younger than Gyr above our luminosity limit, consistent with a galaxy with continuous star formation typicalof an isolated spiral. The gap can therefore be used to age-date the interaction from the onset ofenhanced star formation.
The analysis of Figure 8 shows that star formation in NGC 7319 has been going on throughoutits history, reflected in the continuous distribution of SCCs along the evolutionary track. However,some peculiarities are observed. For example, the fraction of SCCs younger than
Myr is lowerthan found in NGC 7318A/B and NSBR. Our analysis of Figure 10 (which gives us the approx-imate values of the masses and ages of SCCs above the 50% completeness level) shows that weshould be able to detect a
Myr old clusters down to . M ⊙ , relatively modest-sized clusters 19 –that are massive enough to have a fully sampled initial mass function and therefore ’normal’ colors(as investigated by Silva-Villa & Larsen 2011). Thus, the low number of . Myr old clusters islikely to be real. Another observable is a somewhat tighter concentration of SCCs correspondingto the ages between approximately and
Myr. In light of the models of interaction found inthe literature, our results are consistent with the following scenario: the first traceable interactionfor NGC 7319 was with NGC 7320C and it happened ∼ Myr ago. It was the cause of the initialincrease in the star formation rate. At the same time, this interaction stripped a large quantity ofthe gas from NGC 7319 and deposited it into the intergalactic medium (IGM), between galaxiesNGC 7319 and NGC 7318A (Moles et al. 1998). Since the mass of the H I gas in that region is esti-mated to be . × h − M ⊙ (Moles et al. 1997), more than what would be expected in a galaxywith the mass of NGC 7319 ( . × M ⊙ ; Renaud et al. 2010), gas from other galaxies may alsocontribute. The most likely candidates are NGC 7320C, which has no detectable H I (Sulentic et al.2001), and NGC 7318A. The models by Renaud et al. (2010) suggest that approximately Myrago NGC 7319 had another interaction, that time with NGC 7318A (see § . Myr) star formation as can be seen in Figure 8.
The Southern Debris Region (SDR) is located between the NGC 7317 and NGC 7318A galax-ies. It hosts low surface brightness structures, with sizes between 2 to 4 kpc, that might qualifyas tidal-dwarf galaxies. As can be observed in Figure 5, SCCs in the SDR are divided into two 20 –groups, the old star clusters ( & Gyr) and cluster candidates with ages between and ∼ Myr. The high number of old star clusters can be explained by the proximity of elliptical galaxyNGC 7317, since the expected radius of ∼ kpc for its old globular cluster system would overlapwith the SDR. The radial extent of the GC system was calculated with the recipe from Rhode et al.(2007), which relates the mass of a galaxy (derived from the mass-to-light ratio specific to thattype of galaxies given its V -band luminosity) to the radial extent of the GC system.However, the origin of the Southern Debris region is not yet clear. On the one hand, the H I inthe SDR has the same velocity (in the range of — km s − ; Williams et al. 2002) as theshock front of NGC 7318B (Sulentic et al. 2001), suggesting that this region might be related toNGC 7318B. On the other hand, the age distribution of younger SCCs is consistent with the agesof the Young Tail. In Renaud et al. (2010), the preferred model for the formation of the YoungTail is the interaction of NGC 7319 and NGC 7318A ( § . Myr, could fit into this scenario as a partof the west pointing tail. However, there is a caveat. In order for this model to be plausible priorto the interaction, NGC 7318A is assumed to have been a spiral galaxy with a cold gaseous disk,which is still questionable as mentioned in § As previously mentioned, NGC 7318B is undergoing a collision with NGC 7318A and theIGM, deposited here by previous interactions involving NGC 7319, NGC 7320C, and perhaps 21 –NGC 7318A. This resulted in the formation of numerous star clusters, of which we were able todetect 110, a rather large number given this region is located outside of NGC 7318A & B. The mostnoticeable features of this region are two tidal tails observed to the north of NGC 7318A/B thatseemingly cross each other. Renaud et al. (2010) suspect that those tidal tails are physically over-lapping, triggering intense star formation at the intersection. This is supported by the discoveriesof strong H α , IR (Xu et al. 1999), and UV (Xu et al. 2005) emission at that location.The hypothesis of the intersecting tidal tails is also supported by the H I maps of SQ presentedby Williams et al. (2002) who measure a velocity of ∼ km s − for the H I at the part of the tailclosest to NGC 7318A. The velocity is increasing with the distance from NGC 7318A, reaching ∼ km s − at the point of intersection of the two tails, and then continuing to increase alongthe eastern tail towards NGC 7318B up to to a value of v ≈ km s − . The velocity of H I inthe intersection region ( ∼ km s − ) is consistent with its being situated between NGC 7318A( km s − ) and NGC 7318B ( km s − ). That, in turn, suggests that the two tidal tails are aproduct of the interaction between those galaxies.From Figure 4, we can see that the NSBR region has many young SCCs. Also, the gap incolor-color plots between the intermediate-aged clusters (few Gyr) and younger clusters (few Myr) seen in other regions of SQ (e. g. NGC 7318A/B, SDR) is not as pronounced here. Mostprobably, this is due to reddened young clusters filling any gaps, if present. Overall, the ageestimates for this region are somewhat uncertain due to the presence of significant amounts of dustand gas. However, based on Figure 9 we see that the NSBR is a region with ongoing active starformation as we can observe a number of star clusters as young as a few Myr, consistent with G01. 22 –
4. History of interactions in SQ
At the present moment, we cannot say anything certain about interactions that might havehappened earlier ( > Myr ago. Moreover, this encounter isalso responsible for stripping the ISM from NGC 7319 and depositing it to the west of the galaxy(Moles et al. 1998). Since NGC 7320C has no detectable H I (Sulentic et al. 2001) and assumingthat there was only one interaction between those galaxies (see § § — Myr ago, based on the age estimates obtained in this paper,NGC 7319 experienced a close approach by NGC 7318A coming from the northeast (Renaud et al.2010). If we assume that NGC 7318A used to be a spiral galaxy (as assumed in Renaud et al.2010), then the results of the interaction with NGC 7319 were pronounced. NGC 7319 formed asecond tidal tail (the Young Tail; Xu et al. 2005), and star formation was triggered in both NGC 23 –7319 itself and in the tail. This is seen in the color-color plots of Figures 6 and 7 as an over-densityof star clusters corresponding to those ages. Specifically, the bulk of the younger star clusters inthe Young Tail were formed at approximately the same time ( ∼ Myr ago). The low surfacebrightness features located south of NGC 7318A/B, and that currently do not belong to any of theearlier defined regions, might represent parts of the stripped spiral arms of NGC 7318A. We inferthis by incorporating the hydrodynamic modelling of Renaud et al. (2010) with our age-estimatesfor the young clusters in the SDR. Most of the ISM from NGC 7318A and NGC 7319, along withthe ISM from NGC 7320C, is then located between NGC 7318A and NGC 7319 (Renaud et al.2010), in what becomes the NSBR.More recently, the collision of NGC 7318B with the core occurred (Sulentic et al. 2001,Xu et al. 2003). According to Renaud et al. (2010), the collision was twofold; first NGC 7318Binteracted with NGC 7318A, and then it collided with the ISM. It is very difficult to determine theapproximate time of the collision between NGC 7318B and NGC 7318A since our results fromBVI data are distorted by the presence of gas and dust in that region (i.e., making SCCs likely toappear older than they are). Nevertheless, a crude estimate can be obtained from the color-colorplot for the NGC 7318A/B region (Figure 7). Considering the gap in the distribution of SCCs dis-cussed in § log( t ) = 8 . — . yr, weconclude that the upper limit on age of the first event is ∼ — Myr. In all probability, the realage of the event is younger since the NGC 7318B collision with the SQ’s core would have likelyhappened after the formation of the Young Tail ( ∼ Myr), purely by a kinematical argument.Assuming the radial velocity difference between the SQ core and NGC 7318B ( ∼ km s − )to be constant for the last Myr, and the observational fact that NGC 7318B is in the midst ofthe collision with the IGM (assumed to be in the same plane with NGC 7319 and NGC 7318A), it 24 –would imply that NGC 7318B started at a distance of more than kpc (along the line-of-sight)from the present position. This means that
Myr ago NGC 7318B and NGC 7318A were too farapart for an interaction between them to be responsible for the increase of the star cluster numberswe observe. Additionally, NGC 7318B still possesses its spiral structure, another argument for arecent collision. At the same time, based on Figures 7 and 9, we can state with certainty that thecollision of NGC 7318B with the core is still ongoing because we can detect a number of veryyoung star clusters < Myr old.The future of the group is strongly dependent on the real spatial velocity dispersion betweenits members. However, based on the v R , some predictions can be made. Since NGC 7319 andNGC 7318A have almost identical radial velocities (see Table 2) and close spatial proximity it isreasonable to conclude that these galaxies will eventually merge (Renaud et al. 2010). No clearhypothesis can be drawn for the future of the rest of the galaxies. NGC 7320C is relatively farfrom the group (at ∼ kpc in projection), NGC 7318B has a very large velocity difference withregards to the rest of the group, and we do not know the real trajectory of NGC 7317. It is possiblethat if NGC 7317 merges with NGC 7319 and NGC 7318A, resulting in a deeper gravitationalwell, NGC 7320C and/or NGC 7318B might not be able to escape.
5. Summary
Based on the sensitive, high S/N photometry that was obtained from ERS with WFC3 on
HST we were able to make robust age estimates of recent events in the history of SQ by studying thestar cluster populations. Thus, we found the following: • The star cluster population of NGC 7317 consists of clusters that are over Gyr old, indi- 25 –cating no recent epochs of significant star formation. From their luminosities, the majorityof them can be considered globular clusters. • The Old Tail formed approximately
Myr ago, based on the distribution of ages amongits star clusters. The remarkably tight distribution and relatively young age of star clusterssuggests in situ star formation, meaning that the Old Tail was formed with a large amount ofgas. • Most of the star cluster candidates detected in the Young Tail are in the age range between and
Myr suggesting this tidal feature formed . Myr ago. • NGC 7319 exhibits continuous star formation throughout its history, albeit at a lower rateover the past few tens of Myr presumably because of gas stripping from the most recentinteraction. The SCC color distribution shows an over-density around
Myr and a possibleclump between
Myr and Gyr. The first feature corresponds to the time of formation ofthe Young Tail, and the second feature could correspond to the time of formation of the OldTail. • NGC 7318 A/B shows ongoing active star formation; there are a number of star clusters thatare younger than Myr. This region also features a pronounced gap in the SCCs distributionin color space between ∼ Myr and Gyr, thus dating the onset of the interaction-inducedstar formation to . Myr. • We have detected 110 SCCs in the Northern Star Burst region, a large number given the inter-galactic location of this region. The age estimates for this region are somewhat uncertain dueto the presence of significant columns of dust and gas. However, we can conclude that starformation is ongoing in this region, given the large number of clusters younger than Myr. 26 – • The star cluster population of the Southern Debris region consists of two age groups: SCCswith ages between and ∼ Myr,dating perhaps to the interaction between NGC 7318Aand NGC 7319, and & Gyr old clusters, likely stripped from NGC 7318A. It hosts lowsurface brightness structures, with sizes between 2 and 4 kpc, which might qualify as tidal-dwarf galaxies.The combination of results obtained through our new optical photometry and empirical resultsand dynamical modelling in the literature has advanced our understanding of the history of interac-tions in this fascinating group of galaxies. However, there are still some unanswered questions. Forexample, the morphological type of NGC 7318A is still a mystery. Although some of the observedproperties correspond well with it being an elliptical galaxy, the presence of tidal arms as well asregions with debris of low surface brightness would rather suggest that this galaxy was a spiral notthat long ago. It is possible that NGC 7318A is in fact the stripped core of a spiral galaxy.Another unclear point in SQ’s history is the formation of the SDR region. The present imagesobtained with the ERS are not deep enough to see if there are some low surface brightness struc-tures that might be associated with the SDR. The possibility that we adopted in this paper is thatthis region originated at the same time as the Young Tail and is a direct result of the interactionbetween NGC 7319 and NGC 7318A. The same scenario is proposed by Renaud et al. (2010).Additional high resolution photometry in the U -band would improve the age-estimates ofclusters in the — Myr age range, and thus tighten the constraints on the time scale ofinteraction-induced star formation, particularly along the X-ray shock region. 27 –We thank Florent Renaud for his useful feedback. Support for this work was provided by theNational Science and Engineering Research Council of Canada and the Ontario Early ResearcherAward (KF & SCG). SCG also thanks the Rotman Institute for its hospitality during Summer2010. R. C. is grateful for support from NSF through CAREER award 0847467. J. C. C. and I.S. K. acknowledge their support by the NSF through award 0908984. This research has made useof SAOImage DS9, developed by Smithsonian Astrophysical Observatory, and of the NASA/IPACExtragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California In-stitute of Technology, under contract with the National Aeronautics and Space Administration.
REFERENCES
Allen, R. J., & Sullivan III, W. T. 1980, Astronomy and Astrophysics, 84, 181Barmby, P., Kuntz, K. D., Huchra, J. P., & Brodie, J. P. 2006, The Astronomical Journal, 132, 883Bergvall, N., Laurikainen, E., & Aalto, S. 2003, Astronomy and Astrophysics, 405, 31Conti, P. S., Leitherer, C., & Vacca, W. D. 1996, The Astrophysical Journal, 461, L87+de Grijs, R., Anders, P., Bastian, N., Lynds, R., Lamers, H. J. G. L. M., & O’Neil, E. J. 2003,Monthly Notices of the Royal Astronomical Society, 343, 1285de Vaucouleurs, G., de Vaucouleurs, A., Corwin Jr., H. G., Buta, R. J., Paturel, G., & Fouqu´e,P. 1991, Third Reference Catalogue of Bright Galaxies. Volume I: Explanations and ref-erences. Volume II: Data for galaxies between 0 h and 12 h . Volume III: Data for galaxiesbetween 12 h and 24 h . (New York, NY: Springer)Falco, E. E., et al. 1999, Publications of the Astronomical Society of the Pacific, 111, 438 28 –Gallagher, S. C., Charlton, J. C., Hunsberger, S. D., Zaritsky, D., & Whitmore, B. C. 2001, TheAstronomical Journal, 122, 163Gallagher, S. C., et al. 2010, The Astronomical Journal, 139, 545Gutierrez, C. M., LopezCorredoira, M., Prada, F., & Eliche, M. C. 2002, The Astrophysical Jour-nal, 579, 592Hickson, P. 1982, The Astrophysical Journal, 255, 382Konstantopoulos, I. S., et al. 2010, The Astrophysical Journal, 723, 197Kroupa, P. 1998, in Astronomical Society of the Pacific Conference Series, Vol. 134, BrownDwarfs and Extrasolar Planets, ed. . M. R. Z. O. R. Rebolo, E. L. Martin, 483Larsen, S. r. S. 1999, Astronomy and Astrophysics Supplement Series, 139, 393Leitherer, C., et al. 1999, The Astrophysical Journal Supplement Series, 123, 3Marigo, P., Girardi, L., Bressan, A., Groenewegen, M. A. T., Silva, L., & Granato, G. L. 2008,Astronomy and Astrophysics, 482, 883Mendes de Oliveira, C., & Hickson, P. 1994, The Astrophysical Journal, 427, 684Moles, M., Marquez, I., & Sulentic, J. W. 1998, Astronomy & Astrophysics, 334, 8Moles, M., Sulentic, J. W., & M´arquez, I. 1997, The Astrophysical Journal, 485, L69Nilson, P. 1973, Nova Acta Regiae Soc. Sci. Upsaliensis Ser. VRejkuba, M., Greggio, L., Harris, W. E., Harris, G. L. H., & Peng, E. W. 2005, The AstrophysicalJournal, 631, 262 29 –Renaud, F., Appleton, P. N., & Xu, C. K. 2010, The Astrophysical Journal, 724, 80Rhode, K. L., Zepf, S. E., Kundu, A., & Larner, A. N. 2007, The Astronomical Journal, 134, 1403Robin, A. C., Reyl´e, C., Derri`ere, S., & Picaud, S. 2003, Astronomy and Astrophysics, 409, 523Salzer, J. J., Jangren, A., Gronwall, C., Werk, J. K., Chomiuk, L. B., Caperton, K. A., Melbourne,J., & McKinstry, K. 2005, The Astronomical Journal, 130, 2584Saracco, P., & Ciliegi, P. 1995, A&A, 301, 348Scheepmaker, R. A., Haas, M. R., Gieles, M., Bastian, N., Larsen, S. r. S., & Lamers, H. J. G.L. M. 2007, Astronomy and Astrophysics, 469, 925Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, The Astrophysical Journal, 500, 525Shostak, G. S., Allen, R. J., & Sullivan III, W. T. 1984, Astronomy and Astrophysics, 139, 15Silva-Villa, E., & Larsen, S. S. 2011, Astronomy & Astrophysics, 529, A25Stephan, M. 1877, Monthly Notices of the Royal Astronomical Society, 37Stetson, P. B. 1987, Publications of the Astronomical Society of the Pacific, 99, 191Sulentic, J. W., Rosado, M., Dultzin-Hacyan, D., Verdes-Montenegro, L., Trinchieri, G., Xu, C.,& Pietsch, W. 2001, The Astronomical Journal, 122, 2993Vacca, W. D., & Conti, P. S. 1992, The Astrophysical Journal, 401, 543Whitmore, B. C., Zhang, Q., Leitherer, C., Fall, S. M., Schweizer, F., & Miller, B. W. 1999, TheAstronomical Journal, 118, 1551 30 –Williams, B. A., Yun, M. S., & Verdes-Montenegro, L. 2002, The Astronomical Journal, 123, 2417Xu, C., Sulentic, J. W., & Tuffs, R. 1999, The Astrophysical Journal, 512, 178Xu, C. K., Lu, N., Condon, J. J., Dopita, M., & Tuffs, R. J. 2003, The Astrophysical Journal, 595,665Xu, C. K., et al. 2005, The Astrophysical Journal, 619, L95 This preprint was prepared with the AAS L A TEX macros v5.2.
31 –Table 1: Basic information on HCG 92 (members a through e ). a Identifier Hickson Coordinates Type B v R designation b (J2000) (mag) ( km s − )NGC 7317 92e 22 35 51.9 +33 56 41 E1 15.3 6637NGC 7318A 92d 22 35 56.8 +33 57 57 E2(pec) 14.4 6671NGC 7318B 92b 22 35 58.4 +33 57 57 Sbc 14.9 5766NGC 7319 92c 22 36 03.7 +33 58 35 Sbc 14.8 6652NGC 7320 (foreground)
92a 22 36 03.4 +33 56 54 Sd 13.8 739 a From The Updated Zwicky Catalog (UZC), Falco et al. (1999) b Hickson (1982)
Table 2: Properties of Star Cluster and Globular Cluster Candidates. a Regions a The number of Star Cluster Candidates (SCCs; M V < − ) detected in SQ. The algorithm of detecting SCCs isdescribed in §
32 –Fig. 1.—
An inverted black and white image (in V ) with defined regions. The contrast of the image has beenscaled to emphasize low surface brightness structures. The 8-shape objects observed in the upper left corner and onthe bottom of the image, as well as in Figures 5 and 8, are ghost images caused by reflections off the CCD and returnreflections from the CCD housing entrance window in WFC3.
33 –Fig. 2.— V versus V − I color-magnitude diagram for SCCs (black filled circles) in all regions. Thesolid lines are evolutionary tracks from Marigo et al. (2008) with a metallicity of Z ⊙ (Saracco & Ciliegi 1995) fordifferent cluster masses (in Solar masses). The dashed lines represent evolutionary tracks that account for nebularemission (SB99), as suited to the very youngest clusters that have not yet expelled their natal gas (see § § V is shown in the topright corner.
34 –Fig. 3.— B − V versus V − I colors for SCCs in all regions, plotted on top of an evolutionary track(solid red line; Marigo et al. 2008) with a metallicity of Z ⊙ (Saracco & Ciliegi 1995). SCCs are denoted by filledblack circles. The objects represented by purple triangles were spectroscopically confirmed to not be members of SQ( §
35 –Fig. 4.—
Color-color plots for Star Cluster Candidates (SCCs) in individual regions, as defined in Figure 1. Thenumber in parentheses gives detected SCCs for that particular region. A typical photometric error bar, based on themedian errors, is located in the lower left corner of each panel. The fainter the population, the larger the typical errors.Larger versions of the color-color diagrams are presented in Figures 5 — 9, along with the spatial distribution of theSSCs.
36 –Fig. 5.—
Top : The spatial distribution of SCCs in the SDR and NGC 7317 regions marked according to their ages:crosses represent very young star clusters ( . Myr), squares represent star clusters from ∼ — Myr to 2 Gyr old,and circles represent star clusters that are & Gyr old.
Bottom : Color-color plots of the same regions. In parenthesesare the numbers of detected SCCs in that regions. SCCs in the SDR (bottom left panel) are divided into two groups,the old star clusters and cluster candidates with ages between to Myr, with virtually no SCCs between thosetwo groups. NGC 7317 (bottom right panel) hosts mostly old star clusters as expected for an elliptical galaxy.
37 –Fig. 6.—
Top : The spatial distribution of SCCs in the Young and Old Tail regions marked according to their ages(symbol definitions follow Figure 5).
Bottom : Color-color plots of the same regions. For the Young Tail (bottomleft panel) the concentration of blue clusters with mean colors of B − V = 0 . ± . mag and V − I = 0 . ± . mag is observed. These mean colors most closely match a model age of Myr. Clusters in the OldTail (bottom right panel) have a narrow range of colors, with a mean B − V = 0 . ± . mag and a mean V − I = 0 . ± . mag, corresponding to a model age of Myr.
38 –Fig. 7.—
Top : The spatial distribution of SCCs in the NGC 7318A/B region marked according to their ages(symbol definitions follow Figure 5).
Bottom : Color-color plots of the same region. A lack of SCCs between theolder ( & Gyr) and younger ( . Myr) populations of SCCs is observed. Five out of the seven objects that arelocated in the box of . mag < V − I < . mag and . mag < B − V < . mag very likely to bereddened young clusters (see § s Myr up to Gyr.
39 –Fig. 8.—
Top : The spatial distribution of SCCs in the NGC 7319 region marked according to their ages (symboldefinitions follow Figure 5).
Bottom : Color-color plots of the same region. There are a number of intermediate-aged clusters along the northeast spiral arm. The number of SCCs younger than
Myr is lower in comparison toNGC 7318A/B and NSBR regions. It would appear that star formation in this region was truncated ∼ Myr agoconsistent with the observed lack of neutral hydrogen in NGC 7319.
40 –Fig. 9.—
Top : The spatial distribution of SCCs in the NSBR region marked according to their ages (symboldefinitions follow Figure 5).
Bottom : Color-color plots of the same region. A large number of very young SCCs( < Myr) is observed. A gap in color-color plots between clusters of a few Gyr and a few
Myr seen in otherregions of SQ (e. g. NGC 7318A/B, SDR) is not as prominent here, likely from reddened young clusters populatingthis region.
41 –Fig. 10.—
Plot of predicted cluster masses over time for 90% completeness limits in the B , V , and I bands. For the intermediate-aged star clusters ( < log( t ) < ) we are capable of detecting star clusters with massesas low as ∼ . M ⊙ . For comparison, the dashed-dotted line represents the estimated cluster masses detectable inGallagher et al. (2001) based on 90% completeness limit in the V569