arXiv:astro-ph/0308241v1 14 Aug 2003 elsie slniua S)glxe.Ti can This be to . galaxies (S0) E lenticular for become as trend thus reclassified has common Sahu, It somewhat 1994; a van 1996). Franx Kembhavi & & & Cinzano Jorgensen & Pandey, 1992; Vader 1994; al. 1990; Marel et White der & Nieto Rix 1991; Capaccioli 5; regarded Vigroux (e.g., section increas- once galaxies his steadily were (E) 1987, a elliptical that purely in objects as found of number been ing has disks lar Introduction 1. este o eerhi srnm,Ic,udrNS con- 5-26555. NASA NAS under Uni- tract Inc., of Astronomy, Association in Research the by for versities operated is which Institute, ence Telescope Space ble HST eerhSho fAtooyadAtohsc,Australian Astrophysics, and Astronomy of School Research 1 vdnefrebde,goerclyflt stel- geometrically-flat, embedded, for Evidence ae nosrain aewt h NASA/ESA the with made observations on Based eateto srnm,Uiest fFoia ..Bx11 Box P.O. Florida, of University Astronomy, of Department eateto srnm,Uiest fFoia ..Bx11 Box P.O. Florida, of University Astronomy, of Department prlglxe,priual ncutr,eov nodafelpia ga elliptical hav dwarf which headings: into evolution Subject evolve m We clusters, of b in theories depleted. the particularly in largely of galaxies, link” been faintness “missing spiral have the the may of the and disk example bulge is its an large that it cluste The suggests and Coma 3629 environment, cluster. the GMP cluster Coma to dense a belonging the in 3629) in detected GMP galaxy and such faintest 3292 (GMP galaxies dwarf aais niiul(M 22 M 69 aais spira galaxies: — 3629) GMP 3292, (GMP individual galaxies: EETO FSIA TUTR NTOCOMA TWO IN STRUCTURE SPIRAL OF DETECTION erpr h icvr fsia-iesrcuei ubeSaeTele Space Hubble in structure spiral-like of discovery the report We banda h pc eecp Sci- Telescope Space the at obtained , aais wr aais litcladlniua,c — cD lenticular, and elliptical galaxies: — dwarf galaxies: LSE WR GALAXIES DWARF CLUSTER re O C 61 abra Australia Canberra, 2611, ACT PO, Creek [email protected] lse .Graham W. Alister aalGuzm´anRafael ABSTRACT emtJerjen Helmut Hub- and 1 o erytet er htsm Elk galaxies known dE-like some been that has years are twenty it nearly galaxies Although for (dE) realized. elliptical being dwarf now in disks tected Free- & Carollo, Rix, 1998; 1999). al. Bender man & et Held Scorza Graham & 1994; 1995; Longo Capaccioli, & Nieto, Capaccioli 1987; 1988; behav- Carter dynamical (e.g., after bar-like signif- ior occurred reveals and/or also data rotation has, kinematical icant and of can, inspection It the 1998). al. Scorza et (e.g., profile ex- structure brightness closer component surface multiple a reveals galaxy’s as after a such or of disk, patterns, amination a spiral im- of and/or galaxy’s indicative bars a features of reveals masking age unsharp when happen iial,teeitneo rvosyunde- previously of existence the Similarly, ainlUiest,PiaeBg Weston Bag, Private University, National aais structure galaxies: — l 05 ansil,F 21,USA 32611, FL Gainesville, 2055, 05 ansil,F 21,USA 32611, FL Gainesville, 2055, laxies. rtsc aayobserved galaxy such first odsia-ieptenin pattern spiral-like road yteeoehv found have therefore ay rdce htdwarf that predicted e 1 aais omto — formation galaxies: cp mgso two of images scope .GP32 sthe is 3629 GMP r. do have disk-like morphologies (dS0; Sandage & galaxy selection criteria is described in the follow- Binggeli 1984; Binggeli & Cameron 1991), what ing section, as is the image reduction process and is new — in addition to the rising number of analysis. Section 3 provides a brief quantitative disk detections — is that some dE-like galaxies analysis of the disks, and Section 4 discusses pos- actually have (stellar) spiral structures in their sible evolutionary scenarios for dwarf disk galaxies disk. After the initial surprise announcement of a in clusters. tightly-wound, two-armed spiral structure in the We take Coma to be at a distance of 100 Mpc −1 −1 ′′ dE galaxy IC 3328 (Jerjen, Kalnajs, & Binggeli and use H0=70 km s Mpc , 0.1 therefore cor- 2000b), Jerjen, Kalnajs & Binggeli (2001) and responds to 47 pc. Barazza, Binggeli, & Jerjen (2002) reported pre- viously undetected spiral structure and bars (i.e., 2. Galaxy Sample and Image Analysis disks) in four more galaxies (2 dS0, 1 dE, and 1 low-luminosity E). De Rijcke et al. Galaxies meeting the following conditions — (2003) have additionally presented photometric discussed at length in Matcovi´c& Guzm´an (2003, and kinematic evidence for disks, and in one case in prep) — were selected from the Coma clus- spiral arms, in two edge-on Fornax cluster dS0 ter field catalog (Godwin, Metcalfe & Peach 1983; galaxies. The data to date suggests that up to hereafter GMP). All galaxies have positions within ′ ′ 20% of bright early-type dwarf galaxies in clusters the central 20 20 of the ; 17.5 < × − may have disks. We report here on the first ever MB < 14.5; 0.2 < (U B) < 0.6 and 1.3 < − − detection of spiral-like structure in two dwarf, (B R) < 1.5; available HST WFPC2 images and − early-type galaxies residing in the densest clus- recessional velocities between 4,000 and 10,000 −1 ter environment studied so far, namely the Coma km s . The spectral analysis and recessional cluster. velocity derivation is also provided in Matcovi´c The presence, or at least detection, of (stellar) & Guzm´an (2003, in prep). The above require- spiral patterns in dwarf galaxies is a particularly ments were expected to result in the selection of rare phenomenon. Although dwarf versions of Sm Coma cluster dwarf elliptical galaxies, and we ob- and Irr galaxies have been known for a long time tained 18 such candidates. With the exception (e.g., van den Bergh 1960), referring to the early- of GMP 2960, there were no pre-existing mor- type Sa-Sc spiral galaxies, Ferguson & Sandage phological type classifications for these galaxies. (1991) wrote that “dwarf spiral galaxies do not ap- GMP 2960 (PGC 44707; Paturel et al. 1989) is pear to exist” (see also Sandage & Binggeli 1984 classified in NED as an S0 galaxy, and according and Sandage, Binggeli & Tammann 1985). They to the type-specific luminosity functions derived are a rare species; indeed, their very existence was from three clusters (Jerjen & Tammann 1997) we only recognized a few years ago (Schombert et al. conclude that GMP 2960 is either a low-luminosity 1995). Even then, Schombert et al. concluded that S0 or a bright dS0 galaxy. dwarf spiral galaxies only exist in the field. The The reduction process of the HST images is de- harsh environment within a — due scribed in Graham & Guzm´an (2003). Briefly, to galaxy mergers, with each other or the intra- we used the iraf task crrej to combine the cluster medium, and/or strong gravitational tidal HST-pipelined exposures, which we then further interactions — is commonly thought to have led cleaned of cosmic rays using LACOS (L.A.COSMIC, to the destruction of the delicate spiral patterns van Dokkum 2001). Due to the stellar halos of in dwarf galaxies. A comparison of the number nearby galaxies, we used the wavelet decompo- of such objects in low- and high-density environ- sition method of Vikhlinin et al. (1998) to si- ments may shed light on the nature of their exis- multaneously subtract this non-uniform light and tence. the sky background. Foreground stars and over- By searching for signs of apparent spiral struc- lapping background galaxies were searched for, ture and/or bars in the optical images from the and masked out, before we performed any image sample of 18 dE galaxy candidates presented in analysis or surface brightness fitting. Graham & Guzm´an (2003), we explore here which In order to search for non-symmetric structures galaxies may have embedded stellar disks. The in the images, we subtracted the

2 axisymmetric component of the galaxy light (see large-scale disk displaying no obvious spiral struc- Jerjen et al. 2000b), leaving a “residual image”. ture. Neither the light-profile analysis in Graham Following Barazza et al. (2002) and De Rijcke & Guzm´an (2003) nor the present residual image et al. (2003), we have additionally used an un- analysis provide evidence to suggest that the re- sharp masking technique to verify the presence maining 14 galaxies are anything but nucleated of features such as bars or spiral arms, which dwarf elliptical galaxies. would indicate the presence of a flattened stellar disk. Although the majority of galaxies showed no 3. Quantitative Results sign of non-axisymmetric structure, two galaxies (GMP 3292 and GMP 3629) were found to pos- The inwardly extrapolated exponential disk in GMP 3292 has a central surface brightness sess flocculent spiral arms (Figures 1-2). Their −2 basic properties are given in Table 1. of µ0,F 606W =20.68 mag arcsec (Graham & Guzm´an, their table 2). Correcting this value From our previous analysis of the radial light- for Galactic extinction ( 0.02 mag; Schlegel, profiles (Graham & Guzm´an 2003), we had al- Finkbeiner, & Davis 1998),− (1 + z)4 redshift ready identified GMP 3292 as a likely bulge/disk dimming ( 0.10 mag), and K-correction (0.02 system due to a clear break in its surface bright- mag; Poggianti− 1997) gives a value of 20.58 mag ness profile marking the bulge/disk transition. arcsec−2. Assuming a B F606W color of 1.08 With regard to GMP 3629, we had remarked upon (Fukugita, Shimasaku, &− Ichikawa 1994) yields2 the possibility of an outer disk — not dominating −2 ′′ µ0 =21.66 mag arcsec , very close indeed to until radii greater than 10 (4.7 kpc) — but ,B the canonical Freeman (1970) value of 21.65 B- we could not and did not∼ confirm this due to the mag arcsec−2. The disk scale-length is 2.12′′, low surface brightness levels at these outer radii. which translates to 1 kpc. This is at the small We can however now confirm that both of these end of the range from∼ 1.0 to 2.5 kpc found in galaxies possess stellar disks as indicated by the Schombert et al.’s (1995) sample of dwarf spiral presence of a spiral pattern. Using the velocity galaxies3. The total apparent galaxy magnitude catalog of Edwards et al. (2002), Guti´errez et al. is 16.74 F606W-mag (Graham & Guzm´an 2003, (2003) derive a mean recessional velocity of 6862 −1 their table 2). Assuming a distance modulus of km s , and a velocity dispersion of 1273 145 −1 35.0 gives a corrected absolute B-band magni- km s , for Coma dwarf galaxies fainter± than tude of -17.28 B-mag. Based on the luminosity MB = 17.5 mag. The recessional velocities of −1 functions of individual Hubble types (Jerjen & GMP 3292− (4955 km s ) and GMP 3629 (5219 −1 Tammann 1997, their Fig.3), GMP 3292 is likely km s ) are therefore consistent with membership to be a small late-type (Sc-Sm) . The in the Coma cluster. Although, these velocities are placement of its disk in the µ0 log(h) diagram 1.5 and 1.3 sigma from the mean cluster value, one (Figure 3) is also consistent with− this assignment, may actually expect to find such an offset given although it should be kept in mind that this region the “infalling group” feeding mechanism for clus- of parameter space containing small faint disks is ters (see, e.g., Zabludoff & Franx 1993; Conselice, far from fully explored. Gallagher, & Wyse 2001; Drinkwater, Gregg, & Colless 2001a). The second galaxy, GMP 3629, is modeled here in an identical fashion to Graham & Guzm´an’s From the full sample of 18 dwarf galaxies, Gra- treatment of GMP 3292. A Moffat function was ham & Guzm´an (2003) had additionally found fitted to nearby stars and/or globular clusters on that neither GMP 2960 nor GMP 3486 could be the HST WFPC chip containing the image of described with a single S´ersic model; that is, their GMP 3629. The best-fitting Moffat function was surface brightness profiles suggested the presence then used to convolve the central Gaussian, S´ersic of more than one component (aside from nucle- ation). However these two galaxies display no ev- 2A small but unknown inclination correction for dust may idence of spiral, or in fact any asymmetric, struc- be required; the outer galaxy isophotes have an ellipticity ture in their residual images. GMP 3486 may of only 0.20. therefore be, like the previously mentioned classi- 3The disk scale-length range given in Schombert et al. (1995) H −1 −1 fication for GMP 2960, a with a increases to 1.2-3.0 kpc when using 0=70 km s Mpc .

3 R1/n, and exponential models which were simulta- of gas and dust. More specifically, they observed neously fitted to GMP 3629’s nuclear star cluster, a flocculent nature to the spiral arms and they bulge, and disk respectively (see Figure 4). The measured “double horned” shapes in the HI line resultant central disk surface brightness of 25.58 profiles of their galaxies. Their dwarf spiral galax- F606W-mag arcsec−2 translates into a corrected ies therefore contain significant gas and their disks B-band value of 26.56 mag arcsec−2 — 5 mag do rotate. Unfortunately we have no rotational arcsec−2 fainter than the Freeman value, sugges- data for GMP 3629. As for the presence of gas tive of depletion. This value is however somewhat and dust, there is no obvious signs of this either. poorly constrained and a deeper exposure would However, if GMP 3629 has been somewhat ha- be of great value for better quantifying the disk. rassed by the cluster environment, then we would The total apparent F606W magnitude of indeed expect to find that the gas and dust have GMP 3629 is 18.02 mag — derived from the been stripped away from this galaxy’s disk, or to best-fitting star-cluster/bulge/disk models extrap- have sank to the center of the galaxy where new olated to infinity. This agrees well with the appar- stars may have formed. We may also expect the ent F606W magnitude of 18.18 mag obtained in stars in the original disk to have been heated up Graham & Guzm´an (2003)4 and the apparent blue to create the bulge-like structure we see, and the magnitude 19.03 B-mag obtained by Godwin et al. spiral arms to have simultaneously broadened to (1983) within the isophote b26.5. The value 18.02 their present state. It does therefore seem plau- F 606W -mag translates into a corrected absolute sible that GMP 3629 may be one of the first dS blue magnitude MB=-16.00 mag, which happens galaxies detected in a cluster, although, undergo- to mark the boundary between bright and faint ing a metamorphism. Technically, it is no longer a dwarf galaxies (Ferguson & Binggeli 1994). This dwarf spiral galaxy, but it is by no means (yet?) a is also the limiting magnitude5 where Sandage dwarf either. From the decompo- et al. (1985) claimed fainter (Virgo cluster) spi- sition of the light-profile (Figure 4), the disk has ral galaxies did not exist (see also Ferguson & only 38% the luminosity of the bulge — a very Sandage 1991). Assuming a B V color of 0.9, low value which may be indicative that much of GMP 3629’s magnitude is equivalent− to the mag- the disk has either been removed from the galaxy nitude of three of the six “dwarf spiral” galax- or redistributed. The disk-to-bulge luminosity ra- ies presented in Schombert et al. (1995). Unlike tios for the early-type dS galaxies in Schombert et GMP 3292, GMP 3629 is faint enough to meet al. (1995) were such that their disks were, relative to their bulges, some 200-600% more luminous. Schombert et al.’s selection criteria for“dwarf spi- ∼ ral” galaxies. 4. Discussion In addition to having faint magnitudes (MB & 16 mag), and faint central surface brightness − −2 The nomenclature for dwarf galaxies identifies a values (23-24 B-mag arcsec ), Schombert et al. number of different species. At least structurally, (1995) identified additional characteristics of their the dwarf elliptical galaxies ( 13 . MB . 18) dwarf spiral galaxies. They found that the op- are now known to be the low luminosity− extension− tical radii were small, with R26 < 5-6 kpc (see of ordinary, bright (MB . 18) elliptical galaxies their table 1)6. The GMP catalog gives an isopho- − ′′ (Jerjen & Binggeli 1997; Jerjen, Binggeli, & Free- tal radius for GMP 3629 of b26.5 = 10.1 (4.7 man 2000a; Graham & Guzm´an 2003)7. The gas- kpc); although the R26 radius is expected to be deficient8 (e.g., Skillman & Bender 1995; Young slightly larger, this is nonetheless a small galaxy. 2000), low surface brightness (µ0 & 23 B-mag Schombert et al. also remarked on the presence arcsec−2) dwarf spheroidal (dSph) galaxies are 4 Graham & Guzm´an (2003) derived the total galaxy magni- 7 The term “dwarf elliptical” should not to be confused tude for GMP 3629 from an R1/n-bulge (plus star-cluster) with the rare class of “compact elliptical” galaxy (de Vau- fit to the surface brightness profile; no disk was fitted. 5 couleurs 1961), whose very existence has recently been Sandage et al. (1985) used a Virgo cluster distance modulus questioned (Graham 2002). of m − M=31.7. 8 6 Gas has, however, been detected surrounding the optical These values increase to 6-7 kpc when using H0=70 km −1 −1 component of dSph galaxies (e.g., Carignan 1999; Blitz & s Mpc Robishaw 2000).

4 9 fainter still (MB . 13, Grebel 2001) . Their lo- galaxy and galaxy-cluster interactions is expected cation in the magnitude–central− surface brightness to thicken a galaxy’s disk (e.g., T´oth & Ostriker diagram reveals a continuous extension with the 1992) and suppress spiral features while ram- dE and E galaxies and hence the division of galaxy pressure stripping can remove the gas and prevent classes is somewhat artificial. Reflecting this is the further star formation — producing S0 and dS0 fact that many Authors don’t even bother to make galaxies which are characterized by their thick fea- the distinction between dE and dSph galaxies; al- tureless disks. Additionally, “galaxy threshing”, though, the latter may have a greater range of an extreme example of tidal forces (Bekki, Couch, formation histories (e.g., Conselice 2002) They are & Drinkwater 2001a), has been invoked to explain also generally recognized as increasingly dark mat- the formation of M32-like objects (Bekki et al. ter dominated (e.g., Carignan & Freeman 1988; 2001b; Graham 2002), and the recently discov- Irwin & Hatzidimitriou 1995; Mateo 1997; Kleyna ered “ultra-compact” dwarf galaxies (Drinkwater et al. 2002), but see Miligrom (1995), Klessen & et al. 2003) — purportedly the nuclear remnants Kroupa (1998), Klessen & Zhao (2002) and, in the of nucleated dE/dSph galaxies which have been case of Ursa Minor, Gomez-Flechoso & Martinez- severely tidally stripped, Lastly, “galaxy starva- Delgado (2003). tion” — the removal of halo gas, as opposed to disk The gas-rich, clumpy (both optically and in HI gas, from spiral galaxies falling into a galaxy clus- gas) dwarf irregular (dIrr) galaxies may be the ter — can also result in the diminished prominence progenitors of some dSph, dE, and dS0 galaxies. of spiral arms and the eventual transformation to Although dIrr galaxies are considered to be rotat- an S0 galaxy (Larson, Tinsley, & Caldwell 1980; ing disk galaxies, such a morphological transfor- Bekki, Couch, & Shioya 2001, 2002; Conselice et mation may occur via galaxy merging (Toomre & al. 2001, 2003). Most spiral galaxies would use Toomre 1972; Toomre 1974; Barnes & Hernquist up their disk gas within a few Gyrs (Gallagher, 1996; Bekki 1998; Burkert & Naab 2003), or via Bushouse, & Hunter 1989). the somewhat less severe ram-pressure (and tur- The paucity of spiral galaxies in the central re- bulent and viscous) stripping of gas from a galaxy gions of nearby clusters, and the relative abun- as it moves through the (e.g., dance of S0 galaxies relative to the field population Gunn & Gott 1972; Nulsen 1982; Lin & Faber (e.g., Michard & Marchal 1994), has been taken 1983; Cayatte et al. 1994; Sofue 1994; van den as evidence of the transformation of spirals into Bergh 1994; Quilis, Moore, & Bower 2000; Fujita lenticular galaxies (e.g., Dressler et al. 1997). In- 2001; Toniazzo & Schindler 2001; Grebel, Gal- deed, observations at intermediate redshifts have lagher, & Harbeck 2003; Lee, McHall, & Richer revealed a higher percentage of spirals in clusters 2003). Although, this latter scenario can not be than is observed locally. The presence of spiral applicable to the brighter dEs, which are more galaxies in the periphery of massive clusters (Oem- massive than the dIrr galaxies (Bothun et al. ler 1974; Melnick & Sargent 1977; Dressler 1980) 1986). has also been interpreted as testimony to their in- Another scenario is “galaxy harassment” (Moore creased survivability in this lower density environ- et al. 1996; Moore, Lake, & Katz 1998; Mao & Mo ment. But what about the dwarf galaxies? 1998; Mayer et al. 2001) which not only consid- The lower gravitational potential of dwarf ers the perturbing influence of the entire cluster’s galaxies is suspected to make them particularly gravitational field (e.g., Byrd & Valtonen 1990) susceptible to ram-pressure stripping, although but also tidal effects from repeated, fast flybys other factors, such as the velocity of the galaxy of massive galaxies. Tidal heating from galaxy- through the ICM and the density and tempera- ture of the ICM, are also likely to be important. 9 For a long time dSph galaxies were largely only observed What ever the case, if the disruption process is in the Local Group (e.g., Da Costa 1998; Mateo 1998; Ar- mandroff, Jacoby, & Davies 1999; Caldwell 1999; van den not too severe, at least for the stars, it may leave Bergh 1999, 2000), but have now been detected in other behind the spiral structure. Once the gas is re- groups (e.g., Jerjen et al. 2000a; Zabludoff & Mulchaey moved, spiral patterns are expected to disappear 2000; Carrasco et al. 2001) and also in the Virgo and For- in less than 10 galactic rotations (Sellwood & nax clusters (Phillipps et al. 1998; Drinkwater et al. 2001b; Hilker, Mieske, & Infante 2003). Carlberg 1984); which would suggest GMP 3629

5 was stripped within only the last couple of Gyrs. sinks to the very center of the galaxy and the The growing recognition of disks in intermediate- stellar distribution is heated to the extent that luminosity ( 18 & MB & 20) elliptical galaxies it closely resembles a , al- may be evidence− of their increased− ability to retain though some may retain a very thick stellar disk their disks, as compared with the fate of the lower- and would have the appearance of a dwarf lentic- mass/lower-luminosity galaxies (M & 18). ular. B − If, in addition to the dIrr galaxies observed The remarkably faint spiral/disk in GMP 3629 in nearby clusters, dwarf spiral galaxies were to suggests that we may indeed have an example of have entered and/or resided within the hazardous an incomplete morphological transformation of a environment of a galaxy cluster, they too would dwarf spiral into a dE or dS0 galaxy. Although be subject to the above mentioned processes and Jerjen et al. (2000b) stated that IC 3328 was not therefore likely be transformed into early-type a dwarf spiral galaxy, the possible detection of spi- galaxies. But is there evidence that dwarf spi- ral arms in the dS0 galaxy FCC 288 (De Rijcke et rals were once a predominant population in clus- al. 2003), and the spiral patterns seen by Barazza ters? And if so, is there evidence that they are the et al. (2002) in three Virgo dwarf galaxies is strong progenitors of today’s cluster population of dwarf evidence for this scenario. Although our observa- ellipticals? tions do not provide conclusive evidence in sup- For over 25 years a substantial fraction of the port of the galaxy harassment model, there are galaxies in clusters at intermediate redshifts (z > additional hints in our data that make this model 0.3) were seen only as fuzzy blobs in ground-based particularly attractive. For instance, there is the 10 images (Butcher & Oemler 1978, 1984). HST hint of a bar-like structure in the residual im- images revealed that these objects are low lumi- ages of GMP 3629. We also remark that the large nosity spiral galaxies, often exhibiting disturbed amount of gas predicted to sink to the galaxy cen- morphologies (Couch et al. 1994). Oemler et al. ter may, in principle, provide the fuel for forming (1997) concluded that merging is an implausible the large stellar clusters seen in nucleated dwarf mechanism (see also Ostriker 1980) to explain the ellipticals. disturbed morphologies as the blue galaxy frac- Due to the deficit of galaxies with spiral struc- tion is large and the merging probability is low. ture in dense clusters, in such environments it is They also noticed that disturbed spirals were ob- more common to think about how spiral density served throughout the cluster. This would ar- waves have been destroyed than to contemplate gue against ram pressure stripping being respon- how spiral structure may form. If created, such sible for the disturbed spiral structure since this spiral patterns must be a transient phenomenon mechanism is expected to only operate efficiently — or at least they are destined to remain a faint near the cluster center. The “galaxy harassment” feature — otherwise they would have already been scenario (Moore et al. 1996), however, does pro- observed. The bulge/disk decomposition in Fig- vide a plausible explanation to the disturbed mor- ure 4 suggests that most of GMP 3629’s light phology of the low luminosity spirals seen in in- is not from the disk component, therefore rul- termediate cluster and their transformation into ing out the option that swing amplification of dwarf early-type systems by the present epoch. noise, or clumps of material within the disk, may Although direct mergers are extremely rare in the have caused the spiral-like pattern (Toomre 1981; cluster environment, every galaxy experiences a Toomre & Kalnajs 1991). Non-symmetric gravi- high speed close encounter with a bright galaxy tational perturbations may invoke torques capable approximately once every Gyr. Moore et al. (1999) of generating spiral patterns in galaxy disks (e.g., show that these fly-by collisions have dramatic ef- Kormendy & Norman 1979). Rotating galactic fects on the morphologies of the dwarf galaxy pop- bars (e.g., Schwarz 1981), triaxial bulges (e.g. Tru- ulation. The first encounter leads to a pronounced jillo et al. 2002), and/or triaxial dark halos (e.g., bar instability. After several strong encounters, 10 the loss of angular momentum combined with im- The spiral arms in GMP 3629 do not appear to sprout from pulsive heating leads to a prolate shape supported the galaxy center, but instead from the ends of a very faint bar — although poor contrast makes this claim tentative. equally by random motions and rotation. The gas

6 Bureau et al. 1999; Bekki & Freeman 2002; Masset We thank Bruno Binggeli and Agris Kalnajs & Bureau 2003) have been proposed as the culprit, for kindly reviewing this paper. We also thank at least when it comes to modifying the gas dis- Chris Conselice and Jim Schombert whose com- tribution. With regard to bars, while the gas can ments helped improve this work. R.G. acknowl- be forced about, Sellwood & Sparke (1988) con- edges funding from NASA grant AR-08750.02-A cluded that only the strongest of bars are likely and NRA-01-01-LTSA-080. to have a significant influence on the distribution This research has made use of the NASA/IPAC of stars, a result confirmed by the structural anal- Extragalactic Database (NED) which is operated ysis of bar and arm strength in Seigar & James by the Jet Propulsion Laboratory, California In- (1998; see also Seigar, Chorney, & James 2003). stitute of Technology, under contract with the Na- Other mechanisms may be “grooves” in the dis- tional Aeronautics and Space Administration. tribution of the angular momentum density (Sell- wood & Kahn 1991), or gravitational tides from the passage of nearby objects (e.g., Toomre 1974 and references therein; Noguchi 1987; Sundelius et al. 1987). At least in projection, all of the 18 galaxies reside close to luminous elliptical galaxies - indeed, all of these dwarf galaxies reside in the same field as the HST pointings which were actu- ally directed at luminous ellipticals in the Coma cluster. Simulations, however, have shown that such “harassment” from massive neighbors within a galaxy cluster destroys, rather than creates, spi- ral patterns (Moore et al. 1996, 1998; Gnedin 2003). Lastly, we remark that intermediate mass black holes, having masses somewhere between those of stellar-mass black holes and supermassive black holes, may have formed in young compact star clusters via the runaway merging of massive stars (Rees 1984; Kochanek, Shapiro & Teukolsky 1987; Matsushita et al. 2000; Ebisuzaki et al. 2001; Portegies Zwart & McMillan 2002; Marconi et al. 2003). If so, one might expect some of the dwarf ellipticals containing central star clusters to har- bor black holes. On the other hand, perhaps the star clusters are the result of the adiabatic growth of a pre-existing black hole (e.g., van der Marel 1999). If this is the case, then tests of the adia- batic growth model that have excluded the nuclear component should be reconsidered (e.g., Ravin- dranath, Ho, & Filippenko 2002). Clearly, more data are needed to provide con- clusive evidence in support of or against the galaxy harassment model. In particular, HST/STIS spatially-resolved spectroscopy of our sample of Coma dEs may yield a critical test of this model by measuring the amount of rotation and anisotropy in these galaxies, and also the ages of their nuclear clusters.

7 REFERENCES Cinzano, P., & van der Marel, R.P. 1994, MNRAS, 270, 325 Armandroff, T.E., Jacoby, G.H., & Davies, J.E. 1999, AJ, 118, 1220 Conselice, C.J. 2002, ApJ, 573, L5 Barazza, F.D., Binggeli, B., & Jerjen, H. 2002, Conselice, C.J., Gallagher, J.S., & Wyse, R.F.G. A&A, 391, 823 2001, ApJ, 559, 791 Barnes, J.E., & Hernquist, L. 1996, ApJ, 471, 115 Conselice, C.J., O’Neil, K., Gallagher, J.S., & Wyse, R.F.G. 2003, ApJ, in press (astro- Bekki, K. 1998, ApJ, 502, L133 ph/0303185) Bekki, K., Couch, W.J., & Drinkwater, M.J. Couch, W.J., Ellis, R.S., Sharples, R.M., & Smail, 2001a, ApJ, 552, L105 I. 1994, ApJ, 430, 121 Bekki, K., Couch, W.J., Drinkwater, M.J., & Da Costa, G.S. 1998, in Stellar astrophysics for the Gregg, M.D. 2001b, ApJ, 557, L39 Local Group, VII Canary Islands Winter School Bekki, K., Couch, W.J., & Shioya, Y. 2001, PASJ, of Astrophysics, eds. A.Aparicio, A.Herrero, 53, 395 F.Sanchez (Cambridge: CUP), 351 Bekki, K., Couch, W.J., & Shioya, Y. 2002, ApJ, De Rijcke, S., Dejonghe, H., Zeilinger, W.W., & 577, 651 Hau, G.K.T. 2003, A&A, 400, 119 Bekki, K., & Freeman, K.C. 2002, ApJ, 574, L21 de Vaucouleurs, G., 1961, ApJS, 5, 233 Binggeli, B., & Cameron, L.M. 1991, A&A, 252, Dressler, A. 1980, ApJ, 236, 351 27 Dressler, A., et al. 1997, ApJ, 490, 577 Blitz, L., & Robishaw, T. 2000, ApJ 541, 675 Drinkwater, M.J., Gregg, M.D., & Colless, M.M. Bothun, G.D., Mould, J.R., Nelson, C., & 2001a, ApJ, 548, L139 MacGillivray, H.T. 1986, AJ, 92, 1007 Drinkwater, M.J., Gregg, M.D., Hilker, M., Bekki, Bureau, M., Freeman, K.C., Pfitzner, D., & K., Couch, W.J., Ferguson, H.C., Jones, J.B., Meurer, G.R. 1999, AJ, 118, 2158 & Phillipps, S. 2003, Nature, 423, 519 Burkert, A., & Naab, T. 2003, in Galaxies Drinkwater, M.J., Gregg, M.D., Holman, B.A., & and Chaos, eds. G.Contopoulos and N.Voglis Brown, M.J.I. 2001b, MNRAS, 326, 1076 (Springer), (astro-ph/0301385) Ebisuzaki, T., et al. 2001, ApJ, 562, L19 Butcher, H., & Oemler, A. 1978, ApJ, 219, 18 Edwards, S.A., Colless, M.M., Bridges, T.J., Butcher, H., & Oemler, A. 1984, ApJ, 285, 426 Carter, D., Mobasher, B., & Poggianti, B.M. Byrd, G., & Valtonen, M. 1990, ApJ, 350, 89 2002, ApJ, 567, 178 Caldwell, N. 1999, AJ, 118, 1230 Ferguson, H.C., & Binggeli, B. 1994, A&ARv, 6, 67 Capaccioli, M. 1987, , IAU Symp. No. 127, ed. T.deZeeuw, Kluwer: Dordrecht, p.47 Ferguson, H.C., & Sandage, A. 1991, AJ, 101, 765 Capaccioli, M., & Longo, G. 1994, A&ARv 5, 293 Freeman, K.C. 1970, ApJ, 160, 811 Carignan, C. 1999, PASA 16, 18 Fujita, Y. 2001, ApJ, 550, 612 Carignan, C., & Freeman, K.C. 1988, ApJ, 332, Fukugita, M., Shimasaku, K, & Ichikawa, T. 1994, L33 PASP, 107, 945 Carrasco, E.R., Mendes de Oliveira, C., Infante, Gallagher, J.S., Bushouse, H., & Hunter, D.A. L., & Bolte, M. 2001, AJ, 121, 148 1989, AJ, 97, 700 Carter, D. 1987, ApJ, 312, 514 Gnedin, O.Y. 2003, ApJ, 582, 141 Cayatte, V., Kotanyi, C., Balkowski, C., & van Godwin, J.G., Metcalfe, N., & Peach, J.V. 1983, Gorkom, J.H. 1994, AJ, 107, 1003 MNRAS, 202, 113 (GMP)

8 Gomez-Flechoso, M.A., & Martinez-Delgado, D. Lee, H., McHall, M.L., & Richer, M.G. 2003, AJ, 2003, ApJ, 586, L123 125, 2975 Graham, A.W. 2002, ApJ, 568, L13 Lin, D.N.C., & Faber, S.M. 1983, ApJ, 266, L21 Graham, A.W., Colless, M.M., Busarello, G., Zag- Mao, S., & Mo, H.J. 1998, MNRAS, 296, 877 gia, S., & Longo, G. 1998, A&AS, 133, 325 Marconi, A., et al. 2003, ApJ, 586, 868 Graham, A.W., & de Blok, W.J.G. 2001, ApJ, Masset, F.S., & Bureau, M. 2003, ApJ, 586, 152 556, 177 Mateo, M. 1997, in Arnaboldi M., Da Costa G.S., Graham, A.W., & Guzm´an, R. 2003, AJ, 125, Saha P., eds, ASP Conf. Ser. Vol. 116, The Na- 2936 ture of Elliptical Galaxies. ASP, San Francisco, Grebel, E.K. 2001, ApSSS, 277, 231 p.259 Grebel, E.K., Gallagher, J.S. III, & Harbeck, D. Mateo, M. 1998, ARA&A, 36, 435 2003, AJ, 125, 1926 Matsushita, S., Kawabe, R., Matsumoto, H., Gunn, J.E., & Gott, J.R. 1972, ApJ, 176, 1 Tsuru, T.G., Kohno, K., Morita, K.-I., Oku- Guti´errez, C.M., Trujillo, I., Aguerri, J.A.L., Gra- mura, S.K., Vila-Vilar´o, B. 2000, ApJ, 545, ham, A.W., & Caon, N. 2003, ApJ, submitted L107 Hilker, M., Mieske, S., & Infante, L. 2003, A&A, Mayer, L., Governato, F., Colpi, M., Moore, B., 397, L9 Quinn, T., Wadsley, J., Stadel, J., & Lake, G. 2001, ApJ, 559, 754 Irwin, M., & Hatzidimitriou, D. 1995, MNRAS, 277, 1354 Melnick, J., & Sargent, W.L.W. 1977, ApJ, 215, 401 Jerjen, H., & Binggeli, B. 1997, in The Nature of Elliptical Galaxies; The Second Stromlo Sym- Michard, R., & Marchal, J. 1994, A&A, 283, 779 posium, ASP Conf. Ser., 116, 239 Miligrom, M. 1995, ApJ, 455, 439 Jerjen, H., Binggeli, B., & Freeman, K.C. 2000a, Mobasher, B., et al. 2001, ApJS, 137, 279 AJ, 119, 593 Moore, B., Lake, G., Quinn, T., & Stadel, J. 1999, Jerjen, H., Kalnajs, A., & Binggeli, B. 2000b, MNRAS, 304, 465 A&A, 358, 845 Moore, B., Lake, G., & Katz, N. 1998, ApJ, 495, Jerjen, H., Kalnajs, A., & Binggeli, B. 2001, 139 in Galaxy Disks and Disk Galaxies, eds. Moore, B., Katz, N., Lake, G., Dressler, A., & J.G.Funes, S.J. and E.M. Corsini, ASP Con- Oemler, A. 1996, Nature, 379, 613 ference Series, Vol. 230, p. 239 Nieto, J.-L., Bender, R., Poulain, P., & Surma, P. Jerjen, H., & Tammann, G.A. 1997, A&A, 321, 1992, A&A, 257, 97 713 Nieto, J.-L., Capaccioli M., & Held, E.V. 1988, Jorgensen, I., & Franx, M. 1994, ApJ, 433, 553 A&A, 195, L1 Klessen, R.S., & Zhao, H. 2002, ApJ, 566, 838 Noguchi, M. 1987, MNRAS, 228, 635 Klessen, R.S., & Kroupa, P. 1998, ApJ, 498, 143 Nulsen, P.E.J. 1982, MNRAS, 198, 1007 Kleyna, J., Wilkinson, M.I., Evans, N.W., Gilmore, Oemler, A.Jr. 1974, ApJ, 194, 10 G., & Frayn, C. 2002, MNRAS, 330, 792 Oemler, A.Jr., Dressler, A., & Butcher, H.R. 1997, Kochanek, C.S., Shapiro, S.L. & Teukolsky, S.A. ApJ, 474, 561 1987, ApJ, 320, 73 Ostriker, J.P. 1980, Comments on Astrophysics, 8, Kormendy, J., & Norman, C.A. 1979, ApJ, 233, 177 539 Paturel, G., Fouque, P., Bottinelli, L., & Gouguen- Larson, R.B., Tinsley, B.M., & Caldwell, C.N. heim, L. 1989, A&AS, 80, 299 (PGC) 1980, ApJ, 237, 692

9 Phillipps, S., Parker, Q.A., Schwartzenberg, J.M., Toomre, A. 1974, IAU Symp. No. 58, 347 & Jones, J.B. 1998, ApJ, 493, L59 Toomre, A. 1981, in Structure and Evolution of Poggianti, B.M. 1997, A&AS, 122, 399 Normal Galaxies, eds. S.M.Fall & D.Lynden- Portegies Zwart, S.F., & McMillan, S.L.W. 2002, Bell (Cambridge: Cambridge University Press) ApJ, 576, 899 p.111 Quilis, V., Moore, B., & Bower, R. 2000, Science, Toomre, A., & Kalnajs, A.J. 1991, in Dynamics 288, 1617 of Disc Galaxies, ed. B.Sundelius (Gothenburg: G¨oteborgs University) p.341 Ravindranath, S., Ho, L.C., & Filippenko, A.V. 2002, ApJ, 566, 801 Toomre, A., & Toomre, J. 1972, ApJ, 178, 623 Rees, M. 1984, ARA&A, 22, 471 T´oth, G., & Ostriker, J.P. 1992, ApJ, 389, 5 Rix, H.-W., & White, S.D.M. 1990, ApJ, 362, 52 Trujillo, I., Asensio Ramos, A., Rubino-Martin, J.A., Graham, A.W., Aguerri, J.A.L., Cepa, J., Rix, H.-W., Carollo, C.M., & Freeman, K. 1999, & Gutierrez, C.M. 2002, MNRAS, 333, 510 ApJ, 513, L25 Vader, J.P., & Vigroux, L. 1991, A&A, 246, 32 Sahu, D.K., Pandey, S.K., Kembhavi, A.K. 1996, BASI, 24, 777 van den Bergh, S. 1960, ApJ, 131, 215 Sandage, A., & Binggeli, B. 1984, AJ, 89, 919 van den Bergh, S. 1994, ApJ, 428, 617 Sandage, A., Binggeli, B., & Tammann, G. 1985, van den Bergh, S. 1999, A&ARv, 9, 273 AJ, 90, 395 van den Bergh, S. 2000, “The galaxies of the Local Schlegel, D.J., Finkbeiner, D.P., & Davis, M. 1998, Group”, Cambridge University Press ApJ, 500, 525 van der Marel, R.P. 1999, AJ, 117, 744 Schombert, J.M., Pildis, R.A., Eden, J.A., Oem- van Dokkum, P. 2001, PASP, 113, 1420 ler, A.Jr. 1995, AJ, 110, 2067 Vikhlinin, A., McNamara, B.R., Forman, W., Schwarz, M.P. 1981, ApJ, 247, 77 Jones, C., Quintana, H., & Hornstrup, A. 1998, Scorza, C., & Bender, R. 1995, A&A, 293, 20 ApJ, 502, 558 Scorza, C., Bender, R., Winkelmann, C., Capac- Young, L.M. 2000, AJ 119, 188 cioli, M., Macchetto, D.F. 1998, A&AS, 131, Zabludoff, A.I., & Franx, M. 1993, AJ, 106, 1314 265 Zabludoff, A.I., & Mulchaey, J.S. 2000, ApJ, 539, Seigar, M.S., Chorney, N.E., & James, P.A. 2003, 136 MNRAS, 342, 1 Seigar, M.S., & James, P.A. 1998, MNRAS, 289, 685 Sellwood, J.A., & Carlberg, R.G. 1984, ApJ, 282, 61 Sellwood, J.A., & Sparke, L.S. 1988, MNRAS, 231, 25 Sellwood, J.A., & Kahn, F.D. 1991, MNRAS, 250, 278 Skillman, E.D., & Bender, R. 1995, RMxAC, 3, 25 Sofue, Y. 1994, ApJ, 423, 207 Sundelius, B., Thomasson, M., Valtonen, M.J., & Byrd, G.G. 1987, A&A, 174, 67

A Toniazzo, T., & Schindler, S. 2001, MNRAS, 325, This 2-column preprint was prepared with the AAS L TEX 509 macros v5.0.

10 Fig. 1.— a) HST F606W image of GMP 3292. b) Residual image of GMP 3292: following Jerjen et al. (2001) the axisymmetric component of the galaxy has been subtracted. The labels “arm” are intended only to highlight the visible portions of the arms, they do not necessarily imply that there Fig. 2.— Same as figure 1 but using GMP 3629. are 4 arms, there may well only be two (broken) The image size is 18.6′′ 20′′. × arms. c) Result of unsharp masking. The image size is 20′′ 20′′. ×

11 Fig. 4.— Mean-axis (r = √ab)F 606W surface Fig. 3.— Adaption of the B-band central disk brightness profile of GMP 3629. The dashed line surface brightnesses vs. disk scale-length diagram is a Moffat-convolved Gaussian fitted to the cen- from Graham & de Blok (2001). With the excep- tral star cluster; the best-fitting, seeing-convolved tion of GMP 3292 (star), all of the disks are in bulge and disk models are shown with the solid isolated, non-disturbed field galaxies. The solid curves. All components were simultaneously fitted lines are tracks of constant disk luminosity (slope to the filled circles; the less reliable, low surface = 5); for comparison, the dotted lines with a brightness data points (denoted by the open cir- slope of 2.5 delineate the region where most early- cles) were not used in the fitting process. The ex- type galaxies reside. No inclination correction C trapolation of the models are shown by the dotted has been applied to the surface brightness terms. lines. For comparison, a fit without a disk compo- GMP 3292’s value of µ0 has been converted from nent is presented in Graham & Guzm´an (2003). F 606W to the B-band assuming a color of 1.08. No change has been made to its scale-length.

12 Table 1 Galaxy Data

GMP mF 606W MB B/D µ0,F 606W hF 606W B R No. mag mag magarcsec−2 kpc mag−

3292 16.74 -17.28 0.3 20.7 1.0 1.5 3629 18.02 -16.00 2.6 25.6 3.1 1.4

Note.— Column 1: Godwin et al. 1983 (1983) catalog number. Column 2: uncorrected apparent F 606W -band magnitude, derived by extrapolating the fitted models to infinity. Column 3: Corresponding corrected (see text) absolute B-band magnitude. Column 4: Bulge-to-disk luminosity ratio. Column 5 & 6: uncorrected F 606W -band central disk surface brightness and scale-length. Column 7: Global B R color term from Mobasher et al. (2001) and Guzm´an (2003, priv. comm.).−

13