You are on page 1of 20

Draft 4, 04/24/06

A large population of ultra-compact dwarf galaxies in the Fornax Cluster


Michael D. Gregg, Arna M. Karick Physics Dept., U.C. Davis, and IGPP, Lawrence Livermore National Laboratory, L-413, Livermore, CA 94550, USA gregg,karick@igpp.ucllnl.org Michael J. Drinkwater, Ekaterina Evstigneeva, Russell Jurek Department of Physics, University of Queensland, QLD 4023, Australia mjd,katya,jurek@physics.uq.edu.au Steven Phillipps Department of Physics, University of Bristol, Tyndall Avenue, Bristol BS8 1TL, United Kingdom S.Phillipps@bristol.ac.uk Terry Bridges Department of Physics Queens University Kingston, Ontario, Canada K7L 3N6 tjb@astro.queensu.ca J. Bryn Jones Astronomy Unit, School of Mathematical Sciences, Queen Mary University of London, Mile End Road, London E1 4NS, United Kingdom bryn.jones@qmul.ac.uk and Kenji Bekki, Warrick J. Couch University of New South Wales

2 bekki,wjc@bat.phys.unsw.edu.au ABSTRACT The previously cataloged ultra-compact dwarf (UCD) galaxies in the Fornax and Virgo clusters all have 17.5 < bJ < 20. Using the 2dF spectrograph on the Anglo-Australian Telescope, we have carried out a search for fainter UCDs in the Fornax Cluster. In the magnitude interval 19.5 < bJ < 21.5, we have found 56 additional compact cluster members within a projected radius of 0. 9 (320 kpc) of the cluster center. There is luminosity and spatial overlap with objects classied as globular clusters belonging to the central cluster galaxy NGC 1399. The UCDs are a dynamically distinct population, however, from both the globular clusters and the nucleated dwarf ellipticals in Fornax. Correcting for incompleteness, we estimate that there are 70 10 UCDs to bJ < 21.5 in Fornax, and hence the UCDs outnumber other galaxy types in the central region of the Fornax Cluster. Subject headings: galaxies: clusters, clusters: globular

1.

Introduction

Galaxy transformation processes in dense environments should leave rich clusters littered with the remains of disrupted galaxies (Bassino et al. 1994; West et al. 1995; Bekki et al. 2001; Moore 2004). Observational evidence which might verify these processes has remained scant, limited to detections of diuse light trails and intra-cluster stars (Gregg & West 1998; Adami et al. 2005; Durrell et al. 2002; Ford et al. 2002; Feldmeier et al. 2004). A new class of extremely compact galaxy has recently been discovered in the Fornax and Virgo clusters; though a magnitude smaller in physical size than conventional dwarf galaxies, they are relatively easy to detect in nearby clusters. Hilker et al. (1999) found two compact objects in a spectroscopic survey of the Fornax cluster center, and our Fornax Cluster Spectroscopic Survey (FCSS, Drinkwater et al. 2000b; see also Phillipps et al. 2001, Drinkwater et al. 2003, Jones et al. 2006) identied six cluster members which were either unresolved or marginally resolved in ground-based imaging, including the two Hilker et al. objects. These objects have 13 MB 11, sizes 100 pc, and do not resemble any known type of galaxy (Phillipps et al. 2001), so were named ultra-compact dwarf (UCD) galaxies. High-resolution Hubble Space Telescope imaging and VLT echelle spectroscopy have since established that UCDs are a new type of low-luminosity compact galaxy distinct from both globular star clusters and all known types of dwarf galaxy (Drinkwater et al. 2003). In the Virgo cluster, Haegan et al. s

3 (2005) and Jones et al. (2006) have identied 15 20 objects which are morphologically indistinguishable from the Fornax UCDs, suggesting that such objects are ubiquitous in clusters. Explanations for the origin and nature of ultra-compact dwarfs include unusually luminous globular clusters (e.g. Hilker et al. 1999; Drinkwater et al. 2000a; Phillipps et al. 2001; Mieske et al. 2002), evolved extremely luminous star clusters formed in galaxy interactions (Fellhauer & Kroupa 2002; Maraston et al. 2004), and low luminosity analogs of M32 (Drinkwater et al. 2000a). Some theories have argued that highly compact galaxies might have formed in the early Universe (Blanchard, Valls-Gabaud & Mamon 1992; Tegmark et al. 1997), and UCDs may yet prove to be such objects. A favored hypothesis for the formation of UCDs, however, is that they are the remnant nuclei of dwarf elliptical galaxies that have been tidally disrupted during passages close to the central cluster galaxy; we refer to this process as galaxy threshing (Bekki et al. 2001, 2003). Recently, a deeper survey of the central region of the Fornax Cluster found 54 new globular cluster-like objects within 20 of NGC1399 down to V = 21.0 mag (to MB 9.8 mag) (Mieske et al. 2004). They conclude that the brighter (V < 20) objects are consistent with UCDs formed by the threshing process but that most of the fainter objects are genuine globular clusters. The UCDs may also be the remnant nuclei of or giant star clusters from late type galaxies destroyed by the cluster potential, as has been suggested for the giant globular G1 in Andromeda (Meylan et al. 2001). Whatever their nature, these compact objects are an important constituent of galaxy clusters, and determining their origin and evolution will help in understanding the formation of galaxy clusters. We present the results of observations which extend the search for fainter UCDs in the Fornax Cluster over the much larger eld accessible to 2dF. Based on the luminosity function of dE,N nuclei in the Virgo Cluster, searching 1.5 magnitudes deeper for UCDs in the Fornax Cluster should approximately triple the original sample of 6 UCDs. In Section 3 we present our surprising result that 56 new compact objects were found, many more than expected. Preliminary results of these observations were presented in ?; here we present further observations and discuss the properties of the UCDs in detail, and compare them to objects classied as globular clusters. We adopt a distance of 20 Mpc to the Fornax Cluster (Drinkwater et al. 2001).

4 2. Observations

The original FCSS observations in the central eld of the Fornax Cluster produced 6 UCDs in the range 16.5 < bJ < 20. Here we dene UCDs as objects which were classied as stellar (unresolved) in the photographic APM catalog but were found to have redshifts consistent with membership of the Fornax Cluster (600 < cz < 2500km/s); see Drinkwater et al. (2000b). Allowing for incompleteness, we might expect one more UCD in this magnitude range (Drinkwater et al. 2000a; Jones et al. 2006). If UCDs arise by galaxy threshing of nucleated dwarf elliptical (dE,N), then the UCD luminosity distribution should follow that of the dE,N nuclei (Binggeli & Cameron 1991), and extension of our search 1.5 magnitudes fainter (to bJ 21.5) than the original discovery observations should triple the UCD sample size to 20. In 2003 October and 2004 November we made new 2dF observations in Fornax to test this prediction. As in our search of the Virgo cluster, we restricted the observations to a limited color range (bJ r < 1.7) and also to slightly less than the whole 2dF eld (radius < 0. 9), selecting targets from the APM Catalog of bJ and r photographic survey plates. In the extension to fainter magnitudes for this current work we became limited by the depth of the r plate data which reaches only r < 20.4 mag. For objects that were not detected on the r plate we therefore did not apply a color selection, but observed all objects with 18 < bJ < 21.5; for these objects, we have only an upper limit on their bJ r colors, but all are bluer than bJ r = 1.7 (Fig. 1). The 2dF observations (Table 1) and reductions were carried out in the standard fashion as outlined in Drinkwater et al. (2000b), except that longer exposure times were used to reach the fainter magnitude limits. In 4 nights we observed 2500 unresolved stellar targets of which 56 proved to be cluster members. Combined with our rst sample of 6 UCDs, this brings the total of UCDs in Fornax to 62 (Table 2). Their color, magnitude, and spatial distributions are shown in Fig. 1 and Fig. 2. Contained in our full sample of 62 UCDs, there are 29 objects previously listed by Mieske et al. (2004) in their study of the central 20 region. The sample we present here extends almost to the full 1 radius of the 2dF system. In Table 2, we identify objects already found in the lists of Mieske et al. (2002) and Dirsch et al. (2004). There are now at least three distinct naming conventions for UCDs in Fornax; with the large numbers now being found throughout the cluster (and we anticipate also in Virgo), we suggest that UCDs be designated by their J2000 coordinates to avoid confusion as much as possible. Our nal samples were 96% complete (dened as the fraction of the input targets with well-measured redshifts) in the magnitude range 16 < bJ < 20.5 (21 UCDs found), 82%

5 complete in the magnitude range 20.5 < bJ < 21 (22 UCDs found), and 36% complete in the fainter range 21 < bJ < 21.5 (19 UCDs found). The completeness limits are also indicated in Fig. 1. Correcting for completeness we therefore expect the total number of UCDs in this region (radius <0. 9), 16 < bJ < 21.5 is 105 13, although we must consider the possible overlap with objects previously classied as globular clusters, as discussed below.

3.

Properties of faint UCDs

The most remarkable feature of the new, fainter, UCDs is the sheer number of objects, many more than the 20 expected by extrapolating from the number of bright UCDs. The large size of the sample now permits a statistical analysis of the distribution of their luminosities, velocities and positions within the cluster. In Fig. 3 we compare the luminosities of the combined UCD sample to those of conrmed globular clusters around NGC 1399 (Dirsch et al. 2004) and the nuclei of dE,N galaxies in the Virgo cluster (Binggeli & Cameron 1993). Whereas the rst small UCD sample was arguably an extension of the bright tail of the GC distribution (Mieske et al. 2002), the UCDs now outnumber the GCs. The current GC sample, though, is limited to objects within 10 of the cluster center, while the UCD sample extends to 1 . There is clearly an overlap between our UCD sample and objects which have previously been classied as globular clusters associated with the central cluster galaxy NGC 1399. The overlap in luminosity and velocity space between the UCD and globular samples is shown in Fig. 4 where we compare the radial distribution of published globular clusters (Dirsch et al. 2004) to our UCD sample: the samples overlap at radii below 10 and in the magnitude interval 20.5 < bJ < 21.5. Fig. 3 shows the limits of our survey in cluster radius and magnitude in the upper panel. In the inner 54 we are complete to bJ 21; at larger radii (out to 2. 5) where we have observed only two 2dF elds, we are complete to bJ = 19.8. The distribution of the detected UCDs indicates that we would expect the fainter UCDs to exist at still larger radii, and the existence of UCDs right at our magnitude limit suggest that fainter examples exist. Twenty Nine objects in our full sample of 62 UCDs are also listed by Dirsch et al. (2004), Mieske et al. (2002), or Mieske et al. (2004); these are identied in Table 2. At present, the only way to distinguish UCDs and GCs is through high-resolution spectroscopy or imaging (Drinkwater et al. 2003), and even with such data, it is not obvious that the two classes are distinct (Haegan et al. 2005). For the time being, then, we shall refer to the Dirsch et al. s (2004) objects as globular clusters; this sample is so large and so near the center of NGC1399

6 that it is unquestionably dominated by ordinary globulars, even though the brightest objects in it may be UCDs indeed, as noted above, many objects overlap with our sample, but are too few in number to make a dierence in the statistical analysis below. Since all objects in our sample of 62 UCDs have been selected in a uniform manner, for the sake of analysis, we will consider all of them to be UCDs, even though some of the fainter objects, especially close to the center of NGC1399, may be ordinary globulars. We compare the velocity distributions of the UCDs to both the NGC 1399 globular clusters and the Fornax population of nucleated dwarf elliptical (dE,N) galaxies in the middle panel of Fig. 4. To explore the trends with cluster position, we have formed running means of the velocities and dispersions within each of the three samples (Fig. 5). The UCD and dE sample means have been computed for subsamples of 11 objects as a function of cluster radial position; the subsample size for the noisier velocity data but much larger GC sample is 41 objects. In the radius range 6 < R < 10, the GCs have a mean velocity of 1432 km s1 , almost exactly that of NGC1399 (1415 km s1 ), while the 10 UCDs in this interval have a much larger mean velocity of 1679 km s1 , indicating that these UCDs are not part of the dynamical system of NGC 1399. A Students t-test shows that these velocity means dier with 98% signicance. The UCDs and dE population velocities merge seamlessly, conrming that the UCDs as a group belong to the cluster potential and not to NGC 1399 alone. There are 19 UCDs within 6 of the center of NGC 1399; their mean velocity is 1417 km s1 , so these objects are probably truly associated with the NGC 1399 dynamical system, as can be expected from their proximity. The running means of the velocity dispersions also reveal some interesting dierences (bottom panel, Fig. 4). The UCDs have a lower overall velocity dispersion (26220kms) than either the dE,N galaxies within 60 (41763 km s1 ) or the GCs (33411 km s1 ), at greater than 99.5% signicance level (F-test). This indicates that the UCDs form a more relaxed or lower-energy population, which, if they started life as dE,N galaxies, would facilitate threshing to convert them to UCDs: dEs with low energies relative to the cluster will be most aected and more quickly stripped of their halos, while objects with high energies will minimize their time spent near the cluster center, increasing the likelihood that they will remain as dEs over the life of the cluster. To test that the velocity dispersion dierences are not spurious and result from systematics between the 2dF UCD ber spectroscopy versus the long slit data for the GCs, we compared the measured velocities for the 25 objects within 20 of the Fornax center (Fig. 5). The scatter is consistent with the typical errors of 100 km s1 , though there are four or ve outliers. The 2dF spectra of these ve are reasonably good and we have con-

7 dence in the derived velocities. There is no systematic trend with cluster position of 2dF and literature velocity dierences for the UCDs, except possibly over a restricted interval < 8 . Using the tted line shown, a velocity correction can be derived to apply to the 2dF results; this, however, produces an even lower dispersion for the UCDs in this interval. We conclude that the dierences in the velocity means and, especially, velocity dispersions for the three samples are real. The only selection criterion which separates the GCs and UCDs, however, is apparent magnitude, so the dynamical dierence strongly suggest that the UCDs form a distinct population of objects and are not merely the bright end of the GC luminosity function. Even though the GC surveys are not complete to the large radii over which we have searched for UCD candidates, it is instructive to compare the radial distributions of the dierent populations. In Fig. 6 we show the cumulative radial distributions of the populations within the 54 radius of the complete UCD sample. Setting aside the GCs, this plot again emphasizes that the UCDs are much more centrally concentrated than the normal dwarf galaxies. A two-sample Kolmogorov-Smirnov test shows that the UCD and dE,N galaxy radial distributions dier at the 99.9% condence level. This is not unexpected in the threshing model for formation of UCDs because the objects on lower energy orbits would be preferentially threshed; in fact, the dearth of dE galaxies relative to UCDs in the central region of Fornax could be interpreted as support for tidal destruction of dwarf galaxy halos.

4.

Discussion

Our observations have extended the Fornax Cluster Spectrographic Survey results to search for fainter ultra-compact dwarfs in the central 1. 8 diameter region of the Fornax Cluster, revealing an unexpectedly large number of compact objects spread throughout intracluster space. The size of our sample allows us to make statistical comparisons of this population with globular clusters associated with the central galaxy and the general population of nucleated dwarf galaxies. Although there is some overlap in our sample with cataloged globular clusters, we show that the UCDs are dynamically distinct from both the NGC 1399 globular clusters and the Fornax dE,N galaxies, having higher mean velocity and a lower velocity dispersion. The UCDs are much more widely distributed across Fornax than the GCs. The spatial distribution of UCDs is signicantly dierent from that of existing nucleated dwarf galaxies in the same part of the cluster: the UCDs are much more centrally concentrated, as would be expected from the signicantly smaller velocity dispersion. These results lead us to revise our original hypothesis for the formation of UCDs, that they represent a small subset of the current distribution of cluster dE,N galaxies which have

8 have been tidally disrupted by the cluster potential (galaxy threshing) to leave just the nuclei remaining. In the galaxy threshing process the remaining nuclei keep the same orbits as the parent galaxies, so would have the same radial distributions and velocity dispersion: there is no signicant loss of energy that would allow them to fall into closer orbits around the central galaxy. Only a subset of dwarf galaxies, however, are on orbits that will lead to threshing, so a dierence in velocity dispersion between the objects threshed and those not (yet) threshed is expected. In a future paper, we will model this in detail to ascertain whether the observed distribution and velocities of UCDs in Fornax are consistent with threshing of an initial population of dE,N galaxies. The results presented here suggest that if UCDs are created by threshing dE,Ns, then a substantial fraction of UCDs must be from a parent population which formed on closer, low energy orbits around the cluster center, objects predetermined to become UCDs. In future work we plan further high-resolution spectroscopic observations to distinguish UCDs and globular clusters using their internal dynamics. It is vital to determine if the intra-cluster space between the giant galaxies is occupied by a population of faint compact objects: true intra-cluster globular clusters perhaps (West et al. 1995). We also plan to use cosmological simulations of cluster formation to determine if there might have been populations of dwarf galaxies formed on low-energy orbits more readily susceptible to galaxy threshing, facilitating the formation of UCDs. We are grateful to the Anglo-Australian Observatory for continued support of our observing sessions with the 2dF spectrograph and, in particular, for allowing the nal observations to be performed in service mode at short notice. We thank Peter Thomas for helpful discussions about the dynamical simulations. This work has been supported by a Discovery Project grant from the Australian Research Council and grant No. 0407445 from the National Science Foundation. Part of the work reported here was done at the Institute of Geophysics and Planetary Physics, under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48.

REFERENCES Adami, C., Slezak, E., Durret, F., Conselice, C. J., Cuillandre, J. C., Gallagher, J. S., Mazure, A., Pell, R., Picat, J. P., & Ulmer, M. P. 2005, A&A, 429, 39 o Bassino, L. P., Muzzio, J. C., & Rabolli, M. 1994, ApJ, 431, 634 Bekki, K., Couch, W. J., & Drinkwater, M. J. 2001, ApJ, 552, L105

9 Bekki, K., Couch, W. J., Drinkwater, M. J., & Shioya, Y. 2003, MNRAS, 344, 399 Binggeli, B. & Cameron, L. M. 1991, A&A, 252, 27 . 1993, A&AS, 98, 297 Burkert, A. 1995, ApJ, 447, L25 Dirsch, B., Richtler, T., Geisler, D., Gebhardt, K., Hilker, M., Alonso, M. V., Forte, J. C., Grebel, E. K., Infante, L., Larsen, S., Minniti, D., & Rejkuba, M. 2004, AJ, 127, 2114 Drinkwater, M. J., Gregg, M. D., & Colless, M. 2001, ApJ, 548, L139 Drinkwater, M. J., Gregg, M. D., Hilker, M., Bekki, K., Couch, W. J., Ferguson, H. C., Jones, J. B., & Phillipps, S. 2003, Nature, 423, 519 Drinkwater, M. J., Jones, J. B., Gregg, M. D., & Phillipps, S. 2000a, Publications of the Astronomical Society of Australia, 17, 227 Drinkwater, M. J., Phillipps, S., Jones, J. B., Gregg, M. D., Deady, J. H., Davies, J. I., Parker, Q. A., Sadler, E. M., & Smith, R. M. 2000b, A&A, 355, 900 Durrell, P. R., Ciardullo, R., Feldmeier, J. J., Jacoby, G. H., & Sigurdsson, S. 2002, ApJ, 570, 119 Feldmeier, J. J., Ciardullo, R., Jacoby, G. H., & Durrell, P. R. 2004, ApJ, 615, 196 Fellhauer, M. & Kroupa, P. 2002, MNRAS, 330, 642 Ford, H., Peng, E., & Freeman, K. 2002, in ASP Conf. Ser. 273: The Dynamics, Structure & History of Galaxies: A Workshop in Honour of Professor Ken Freeman, 41+ Ghigna, S., Moore, B., Governato, F., Lake, G., Quinn, T., & Stadel, J. 1998, MNRAS, 300, 146 Gregg, M. D. & West, M. J. 1998, Nature, 396, 549 Haegan, M., Jordn, A., Ct, P., Djorgovski, S. G., McLaughlin, D. E., Blakeslee, J. P., s a oe Mei, S., West, M. J., Peng, E. W., Ferrarese, L., Milosavljevi, M., Tonry, J. L., & c Merritt, D. 2005, ApJ, 627, 203 Hilker, M., Infante, L., Vieira, G., Kissler-Patig, M., & Richtler, T. 1999, A&AS, 134, 75

10 Ikebe, Y., Ezawa, H., Fukazawa, Y., Hirayama, M., Izhisaki, Y., Kikuchi, K., Kubo, H., Makishima, K., Matsushita, K., Ohashi, T., Takahashi, T., & Tamura, T. 1996, Nature, 379, 427 Jones, J. B., Drinkwater, M. J., Jurek, R., Phillipps, S., Gregg, M. D., Bekki, K., Couch, W. J., Karick, A., Parker, Q. A., & Smith, R. M. 2006, AJ, 131, 312 Lotz, J. M., Telford, R., Ferguson, H. C., Miller, B. W., Stiavelli, M., & Mack, J. 2001, ApJ, 552, 572 Maraston, C., Bastian, N., Saglia, R. P., Kissler-Patig, M., Schweizer, F., & Goudfrooij, P. 2004, A&A, 416, 467 Mieske, S., Hilker, M., & Infante, L. 2002, A&A, 383, 823 . 2004, A&A, 418, 445 Moore, B. 2004, in Clusters of Galaxies: Probes of Cosmological Structure and Galaxy Evolution, 296+ Navarro, J. F., Frenk, C. S., & White, S. D. M. 1996, ApJ, 462, 563 Phillipps, S., Drinkwater, M. J., Gregg, M. D., & Jones, J. B. 2001, ApJ, 560, 201 Salucci, P. & Burkert, A. 2000, ApJ, 537, L9 West, M. J., Cote, P., Jones, C., Forman, W., & Marzke, R. O. 1995, ApJ, 453, L77

A This preprint was prepared with the AAS L TEX macros v5.0.

11

Table 1: Observations. Date Set exp(h) seeing (arcsec) 2003 October 21 15 2.0 2.0 2003 October 21 16 2.0 1.7 2003 October 22 14 1.5 1.7 2003 October 22 17 4.5 2.3 2003 October 23 18 2.5 1.5 2003 October 23 19 4.0 1.5 2004 November 12 20 4.0 1.4

12

Fig. 1. Color-magnitude distribution of conrmed stars (dots) and UCDs (triangles) in our survey eld. The bJ magnitudes are plotted against bJ r color. The sample was selected by limiting bJ magnitude and the completeness limits (fraction of targets with measured redshifts) are shown for the three bJ ranges indicated by horizontal dotted lines. The r < 20.4 detection limit of the r plate is shown by the angled dotted line: UCDs satisfying our bJ limits but not detected on the r plate are plotted here as open triangles with arbitrary bJ r = 0 colors. All the UCDs, including these, satisfy our color selection bJ r < 1.7 shown by the vertical dashed line.

13

Fig. 2. Distribution of all 62 UCDs overlaid on a two-degree wide bJ Digitized Sky Survey image of the center of the Fornax Cluster. The circle indicates the 0.9 degree radius within which the UCD search is complete.

14

Fig. 3. Luminosities of UCDs (this paper) compared to N1399 GCs (Dirsch et al. 2004) and Virgo dE,N nuclei (Binggeli & Cameron 1991).

15

Fig. 4. Comparison of the UCD, GC, and dE populations as a function of cluster position. The upper panel shows that the three samples are relatively separate in magnitude, though there is a little overlap between the GCs and UCDs. In the bottom two panels, running means of the velocity and dispersion for the UCDs (solid) compared to those for globular clusters (dotted) (Dirsch et al. 2004) and nucleated dwarf elliptical galaxies (dashed) (Drinkwater et al. 2001) reveals dierences in dynamical properties. Although the UCDs and dE,N samples form a continuous distribution in velocity space, there appears to be a real dierence in the mean velocities of the UCDs and GCs in the 6 10 radius interval. The velocity dispersions of the three classes dier markedly as a function of cluster radius.

16

Fig. 5. Comparison of the 2dF UCD velocities with literature values for 25 objects within 21 of the cluster center. There is no systematic trend with either velocity or with cluster position, except possibly for the 18 objects inside 8 . The dotted line is a t to these objects, excluding the three outliers. If this trend is applied to the 2dF UCD velocities, the UCD sample dispersion drops to even lower values in this radial interval.

17

Fig. 6. Cumulative radial distributions of UCDs compared to globular clusters (Dirsch et al. 2004) and nucleated dwarf elliptical galaxies (Drinkwater et al. 2001).

18

Table 2. Ultra-compact dwarfs in Fornax


N (2000) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 34 35 36 36 36 36 36 36 36 36 37 37 37 37 37 37 37 37 37 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 38 51.51 12.34 22.28 26.72 27.74 28.70 34.36 47.65 47.74 51.68 03.30 24.91 27.61 38.29 41.85 43.06 43.56 43.60 45.13 05.08 06.33 06.53 09.27 10.39 10.78 12.02 14.25 16.54 16.70 17.61 18.48 21.73 21.84 23.27 23.78 25.08 25.56 35 35 35 35 35 34 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 44 12 36 22 14 56 19 29 48 30 38 36 30 20 41 22 15 22 29 24 28 23 35 24 25 39 26 26 20 33 27 26 25 20 13 29 37 02.8 59.4 34.3 01.6 13.9 30.7 32.5 36.9 34.1 38.9 04.6 09.7 12.6 20.6 22.8 11.9 09.6 51.8 01.6 09.6 58.8 04.0 07.0 06.1 46.0 57.2 43.8 19.7 23.1 02.8 39.8 16.5 13.8 00.8 49.5 25.3 42.8 R mag 20.3 0.0 19.7 19.1 18.9 0.0 19.9 19.9 20.3 20.2 18.6 19.6 19.4 20.4 20.2 20.2 20.4 19.2 19.9 18.4 18.7 19.6 20.2 18.9 19.4 19.7 20.3 19.7 19.5 19.8 20.3 0.0 0.0 19.8 19.3 20.2 19.6 bJ mag 21.0 20.8 20.4 20.1 20.2 21.1 21.0 20.9 20.9 21.3 19.9 20.3 20.5 21.3 21.3 20.7 18.9 20.1 20.9 19.4 115.1 20.5 21.4 20.1 20.5 21.2 21.5 20.6 20.5 20.8 21.3 21.0 20.5 20.7 20.2 21.1 20.5 vrad km s1 1368.5 1446.2 1461.6 1498.8 1296.9 1657.7 1816.8 1445.9 1340.4 1374.9 1491.4 1496.1 1828.2 2225.6 1175.2 1146.4 1640.5 1326.1 1640.7 1211.8 1311.8 1510.3 1805.5 1549.3 1764.1 1307.2 1376.6 1124.7 1552.5 1505.4 1332.1 1402.5 1410.8 1370.3 1992.5 1157.5 1697.9 vrad km s1 67.5 49.4 75.9 116.8 45.3 65.2 103.8 88.9 90.5 45.7 38.7 55.0 77.2 87.2 60.8 86.3 99.0 82.2 62.9 31.8 57.1 64.3 57.5 64.0 55.2 62.7 95.6 100.5 87.6 92.3 62.9 75.5 170.4 63.5 199.1 63.9 52.3 Notes

UCD 1, 2-2031

2-078 4-2028

UCD 6, 2-2143 UCD 2, 2-2111, 91:93 2-2153 81:47 2-2134 2-073 0-2024 80:115 89:22 0-2062, 89:107 80:12 80:30 0-2066 3-2027 2-2106 1-2024

19

Table 2Continued
N (2000) 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 38 38 38 38 38 38 38 38 38 38 38 38 38 38 39 39 39 39 39 39 39 40 40 40 41 26.76 28.83 29.04 29.07 36.86 36.99 39.37 40.23 41.98 43.14 45.81 47.49 50.73 54.10 17.72 20.56 34.78 35.95 37.21 43.56 52.58 19.94 24.98 37.11 35.88 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 34 35 30 28 22 25 28 25 27 27 33 28 34 37 33 33 25 19 53 28 15 26 04 15 06 58 54 07.7 47.1 56.5 00.3 09.5 44.2 05.8 03.1 13.4 01.5 27.4 13.5 48.3 33.6 30.2 14.6 44.2 24.5 21.7 59.5 24.1 29.8 37.6 40.0 57.8 R mag 20.1 0.0 19.5 0.0 20.2 0.0 20.4 0.0 19.2 20.2 20.1 20.0 20.4 17.0 19.9 18.9 20.0 18.4 20.0 19.6 19.0 20.1 19.8 20.1 19.8 bJ mag 21.1 21.1 20.5 20.7 20.8 21.0 21.4 20.8 19.8 21.0 20.6 21.1 20.4 17.7 20.8 20.2 20.7 18.8 20.9 20.0 19.7 21.1 20.5 21.3 21.0 vrad km s1 1474.9 1460.0 1720.4 1491.0 1364.7 1322.1 1644.4 1230.3 2079.7 1574.1 1844.5 1893.3 1886.5 1590.7 1021.7 1420.2 1528.0 1920.0 1800.2 1447.5 1355.0 1649.5 1432.5 1811.1 1628.8 vrad km s1 72.2 76.5 73.8 72.6 55.6 87.4 135.2 111.9 147.7 116.6 86.6 68.3 105.1 35.8 46.1 63.9 73.5 39.9 92.5 101.2 72.3 73.5 59.1 159.3 56.5 Notes

0-2069 0-2031 78:117

1-058 1-021

UCD 3, 1-2053 1-060 3-2004 UCD 4, 1-2083 3-2019 UCD 5

Note. Photometry is from the APM digitized sky survey database. Original 6 UCDs from Phillipps et al. (2001) and Jones et al. (2006) are identied. Also identied are objects in common with Mieske et al. (2002), Mieske et al. (2004) (hyphenated) or Dirsch et al. (2004) (colons).

20

Table 3: Velocity distributions. Sample Globular clusters (all) Globular clusters (< 6 ) Globular clusters (> 6 10) UCDs (all) UCDs (< 6 ) UCDs (6 10) dE,Ns dE,Ns < 60 N 442 331 112 62 19 10 48 23 v v GC t, F-test dE,N t, F-test 1 1 ( km s ) ( km s ) v, v v, v 144616 33411 145116 33711 143216 32411 152735 24625 93.2 99.6 99.98 141740 17929 99.7 167940 21729 98.0 80.8 150257 39641 142387 41763

Note. The nal 2 columns show the results of Students t-tests or F-tests that the UCD mean velocity v or velocity dispersions v of the samples in each row are dierent from the corresponding GC and dE,N samples. The percentage values indicate the statistical signicance that the dierences seen are real. No data indicates that the signicance is low. For comparison, the velocity of NGC 1399 is 1415 58 km s1 and the mean cluster galaxy velocity is 1493 36 km s1 with v = 374 26 km s1 . There is a known dierence between the giant (v = 308 30 km s1 ) and dwarf (v = 429 41 km s1 ) galaxy velocity dispersions (Drinkwater et al. 2001).

You might also like