arXiv:astro-ph/0701280v1 10 Jan 2007 wyfo h agtglx S 6-00wsfalling was 269-G070 arcmin ESO 7.7 20 galaxy only the 5011C into target NGC the galaxy of from the one away 1999, by out in carried (HJ) Observatory us Spring Siding at scope ulse esitmaueet foe 00k s and km 3000 sky over the in of a position measurements suggested its redshift galaxy on lit- this Based published the about picture. through details different search more a for However, central erature galaxy. low feature- characteristic brightness. the elliptical the exhibits surface dwarf with system distribution nearby light stellar less a the of strikingly Most that closely resem- appearance bling morphological its of because attention CPSA H ASLAOBSERVATORY SILLA LA THE AT SCOPES 1)? ∗ Fig. surface (see stellar galaxy the dwarf luminous larger with low but size a 5011B, physical of NGC member, density a cluster galaxy arises with Centaurus S0 neighboring question is Cen- the the immediate this distant of that galaxy the the than of of true, kind member If what a cluster. be taurus should 5011C NGC lcrncades [email protected] address: Electronic [email protected] address: Electronic AE NOSRAIN AEWT S TELE- ESO WITH MADE OBSERVATIONS ON BASED uigaCDiaigcmag ihte4-nhtele- 40-inch the with campaign imaging CCD a During h aayNC51C(S 6-08 slctdin located is 269-G068) (ESO 5011C NGC galaxy The rpittpstuigL using 2018 typeset 5, Preprint November version Draft eerhSho fAtooyadAtohsc,TeAustral The Astrophysics, and Astronomy of School Research n htNC51Bhsalmnst-egtdaeo 4 of age luminosity-weighted a has 5011B NGC that find uioiywihe g n ealct fNC51Cae0 are 5011C NGC of metallicity and age luminosity-weighted ae npooercprmtr lofvrti cnro saby a at 2 As 5011B past scenario. NGC the this for for favor believed also redshift parameters as photometric cluster on Centaurus based distant the of member siaeaselrms fNC51Ccmaal ota fdafsp dwarf of that to comparable headings: Subject 5011C NGC of mass Group. stellar a estimate trce u teto ic,a uaiedsac f45 of distance putative redsh a wrong at and since, NG a confusion attention with coordinate our listed m system of attracted unnoticed stellar known because even previously well hidden a a surveys, is remained of galaxy extensive discovery the Despite While the 5011C. report NGC available. we is Indeed members prises. group of sus ogsi oe h esit fbt G 01 n t egbrNGC neighbor of its stan redshift and velocity low radial ob 5011C a a were indeed NGC of spectra has that both velocity. the with of pecu and spectra recession redshifts their this telescope, t its cross-correlating The reject ESO at of 3.6m or mode. measurement and the confirm long-slit galaxy, at reliable To EFOSC2 dwarf first with galaxy. a the out S0 have of or to typical spiral was brightness early-type surface an low of very a ing rtclsuyo h rpriso ruso aaiscnb done be can galaxies of groups of properties the of study critical A × G 01:A VROKDDAFGLX NTECNARSAGROU A CENTAURUS THE IN GALAXY DWARF OVERLOOKED AN 5011C: NGC 0s rmnfil-fve n trce the attracted and field-of-view arcmin sq 20 1. A INTRODUCTION T E tl mltajv 05/04/06 v. emulateapj style X aais lses niiul e ru aais wr galaxies – dwarf galaxies: – group A Cen individual: clusters: galaxies: G 01,NC5011C NGC 5011B, NGC uoenSuhr bevtr,A eCroa30,Santiag 3107, Cordova de A. Observatory, Southern European v v ⊙ ⊙ 3227 = 647 = Dtd eevdxx cetdxxx) accepted xxx; Received (Dated: rf eso oebr5 2018 5, November version Draft ± ± 6k sec km 96 C 61 Australia 2611, ACT 0k sec km 50 emtJerjen Helmut ABSTRACT v Saviane Ivo − 1 a ainlUiest,M tol bevtr,Cte Ro Cotter Observatory, Stromlo Mt University, National ian , and − − 1 1 n hsi erydafglx ahrta a than rather galaxy dwarf nearby a is thus and pligpplto ytei ehius we techniques, synthesis population Applying . igei20b aahnsve l 06 n enhe- mean a velocity and & 2006) liocentric Freeman al. dis- (Jerjen, Cen- et mean Karachentsev Mpc the a 3.7 2000b; of Binggeli has and Way group Milky group the A Cen from A tance nearby Cen The the cluster. of taurus direction general the itn ihte30 ms km 3200 the with sistent ute wya 53Mc(ise ikr&Infante red- a & reported Hilker (1986) s km Lucey 3195 (Mieske, of & shift Currie Mpc 45.3 Dickens, at 2005). away further nteohrhn,teCnarscutris cluster Centaurus the hand, other the On ut rmDese 19) h aayctlglse a listed s catalog km (3293 galaxy 5011C NGC The for only (1991). redshift Dressler from sults interaction of sign optical two. obvious the no of between is distance there projected kpc 14 small only the despite However, #530). oeto h etuu lse hthsama velocity mean a has that com- cluster of Centaurus 30 Cen the the of with ponent association an implies immediately oteNrh(i.1 n a eysmlrredshift similar very a has and 1) arcmin 1.1 s km (Fig. 3294 only of North is which the pair galaxy 269-G067), to S0 physical (ESO the a member, 5011B form cluster NGC to Centaurus other seems the 5011C with NGC location, its hl omaueetwsrpre o G51B This NGC5011B. for reported was measurement no while oecnuinwsaddt h itr ytere- the by picture the to added was confusion More ≈ ± 01k s km 3041 . y n oa ealct,adta the that and metallicity, solar a and Gyr 1 p,i ol eapcla bethav- object peculiar a be would it Mpc, 3 . 9 adsa.W on htNC5011C NGC that found We star. dard ± − rdc forsuyw paethe update we study our of product 0 1 er.Ruhdsac estimates distance Rough years. 3 me fteCnarsAGroup, A Centaurus the of ember fsi h ieaue G 5011C NGC literature. the in ifts − . y n 1 and Gyr 1 srpre ytesm uhr (galaxy authors same the by reported as 1 irt,orimdaeobjective immediate our liarity, erygop a edt sur- to lead can groups nearby − h bevtoswr carried were observations The Lcy ure&Dces18) From 1986). Dickens & Currie (Lucey, esm iehvn h size the having time same he eodlglxe nteLocal the in galaxies heroidal andwt h ntuetin instrument the with tained 1 h v o G 01 glx 51,con- #531), (galaxy 5011C NGC for ⊙ ,Chile o, nyoc opeecen- complete a once only i 01 eecmue by computed were 5011B ubristu identity true its number C 5 ms km 551 = − 1 / oa.Fnlywe Finally solar. 5 rmFial(93,which (1983), Fairall from − individual: : 1 Co ´ ta.1997). (Cˆot´e al. et − P 1 aay#456) galaxy , ∗ ≈ d Weston ad, 2times 12 2 appeared somewhat strange as the central surface bright- fitted with a two-dimensional polynomial, with degree 3 ness of NGC5011B is approximately four magnitudes along the dispersion axis, and degree 2 along the spatial higher than that of NGC5011C making the former galaxy direction, which is enough to correct the curvature of the a more logical target for a redshift measurement. arc lines. The r.m.s. error of the wavelength solution is The present paper reports on our spectroscopic and 0.26 A,˚ and the calibrated spectra cover the range 3383 optical follow up of the galaxy pair to shed more light to 7505 A˚ with a dispersion of 4.23 A/pix.˚ The width of on this confusing situation around the two galaxies the arc lines at 5000 A˚ is ≈ 13.3 A,˚ so a velocity mea- NGC5011B and C. In the next section we describe our surement based on a single spectral line has a 1-sigma 3.6m observations and data reduction, and the age and error of ≈ 340kmsec−1. metallicity of the stellar populations dominating the During the night, after a new spectrum was taken, it galaxies light are determined in §3. The new recession was reduced using the calibration data prepared above: velocities for the two galaxies are presented in §4 followed it was rotated and trimmed, bias-subtracted and flat- by a photometric analysis with a rough distance estimate fielded, and the wavelength calibration was applied, for NGC5011C in §5 and a discussion in §6. which transforms the bidimensional frames from the pixel (x, y) space to the (λ, s) space, where s is the spatial di- 2. SPECTROSCOPY rection after correction for the distortion introduced by The observations were carried out with the ESO 3.6m the grism. The one-dimensional spectra were extracted telescope at La Silla, during an EFOSC2 (Buzzoni et with a simple average over the rows covered by the pro- al. 1984) technical night, with the purpose of testing all file, and then multiplied by the number of rows. The science and calibration templates, and to estimate the extraction window was defined by inspecting the spa- photometric zero points and the grism efficiencies. The tial profile of the spectra, and taking the rows were the “rotate to slit” template was tested by putting the nu- spectrum was significantly brighter than the background. clei of NGC 5011B and C into the same 1arcsec slit, and Prior to the extraction, the sky background was sub- taking 3 × 1200sec spectra. Before and after the galax- tracted from each frame: two template sky spectra were ies’ observation, spectrophotometric standards were ob- defined in two windows preceding and following the tar- served, and immediately after the end of the last spec- get spectrum, and the sky background at the position of trum, 3 × 1sec spectra of a radial velocity standard were the target was defined by a linear interpolation between also taken. Although several grisms were tested dur- these two spectra. ing the night, for this particular set of data grism #11 The spectrophotometric standard LTT 3218 was used was used, since it gives a good wavelength coverage and to define the response function of the grism: in this case its intermediate dispersion allows to obtain a good S/N the spectrum was extracted using a very wide window for faint targets in relatively short times. An observing and then compared to the spectrum published by Hamuy log is listed in Table 1. The radial velocity standard is et al. (1992). The spectra were flux-calibrated by first HD125184 (from Udry et al. 1999), and the rest of the ob- normalizing by their exposure time and then dividing by jects in the table are spectrophotometric standards from the response function. Prior to this, they were corrected Hamuy et al. (1992; 1994). for atmospheric extinction using the average curve at La All reductions were carried out with the ESO-Midas Silla. The fully reduced and calibrated spectra of the two data processing system (e.g. Warmels 1991), and in par- galaxies are shown in Fig. 2. ticular within the long context, which provides com- mands to perform the standard reduction steps of long- 3. AGES AND METALLICITIES slit spectroscopy. A series of scripts were written to facil- The spectral energy distribution (SED) of NGC 5011B itate as much as possible the batch reduction of several is obviously redder than that of NGC 5011C, but a more spectra, and they are publicly available in the EFOSC2 quantitative comparison of these two spectra is not pos- instrument pages. We first prepared the system to re- sible, since the slit was aligned along the two galaxy cen- duce the night spectra, using the afternoon calibrations: ters, and not along the parallactic angle. To overcome the frames were rotated to have the wavelength increas- this problem, we started from the fact that NGC 5011B ing along the CCD rows, a median bias frame was cre- is classified as an S0 galaxy (Lauberts 1982). Looking ated and subtracted from the rest of the frames, and a at Fig. 1 of Zaritsky et al. (1995) one can see that normalized flat-field was also generated. To do this, five the SEDs of this class of galaxies are rather uniform, flat-fields taken with slit 1′′. 0 were median combined, and so by selecting a template spectrum and comparing it then a two-dimensional polynomial was fitted to the re- to our observed one, we can estimate the degree of loss sulting image: the polynomial had degree 6 along the of blue flux induced by the differential atmospheric re- dispersion axis, and it was constant along the spatial fraction. Zaritsky et al. used the spectral atlas of Ken- direction. This artificial image was subtracted from the nicutt (1992), which includes five S0 galaxies. We chose median flat, thus removing the lamp continuum and leav- NGC 3245 since, compared to the other templates, it has ing only the pixel-to-pixel variation. The wavelength so- no emission lines, its SED is not rising in the blue, it ex- lution was established with a HeAr arc spectrum taken tends to redder wavelengths, and it has better spectral with the 1′′. 0 slit. A database of wavelength solutions for resolution. By dividing the NGC 5011B spectrum by each EFOSC2 grism is also published in the instrument the template (redshift corrected), a correction function web pages, so by using the appropriate guess session for was found, which is smoothly declining from red to blue, grism #11 it was straightforward to get the solution for reaching at 3400 A˚ 60% of the value at 7500 A.˚ In other the current data set. After searching the master arc for words, 40% of the blue flux was lost during the observa- emission lines, and after defining a two-dimensional grid tion. The trend of the correction is well fit by a straight of points distributed on the arc ridge lines, the grid was line, which was then used to compute the intrinsic SEDs 3 of the two galaxies. The corrected spectra are shown in Finally, from Fig. 1 of BC03 one can estimate that, for Fig. 2 together with the original ones, and after normal- an age of 1 Gyr, the mass-to-light ratio is M/LV ≈ 0.3, 6 izing their fluxes to unity at 6000 A.˚ The original spectra so the mass of the galaxy is M ∼ 8 × 10 M⊙. This value of NGC 5011B and NGC 5011C have fluxes reaching a of the mass is only valid in the single age approximation, maximum of 1.5 × 10−15 erg cm−2 sec−1 A˚−1 in the R but it’s very likely that the SFH of the galaxy is more ex- band and 0.06 × 10−15 erg cm−2 sec−1 A˚−1 in the B tended, so realistically that mass represents only a lower band, respectively. In the R band, NGC 5011C is four limit. But even with this caveat, the stellar mass is in magnitudes fainter than NGC 5011B. the regime of that of dwarf galaxies. A first check of the correctness of our SEDs can be done by using them to perform synthetic broad-band 4. RECESSION VELOCITIES OF NGC 5011B AND C photometry in order to compute colors, and then compar- A brief inspection of the prominent Balmer lines in the ing the result to that obtained with our direct imaging. spectra (see Fig. 2) quickly reveals that the two galax- To compute synthetic colors, a stellar A0I template was ies must have quite different redshifts in contrast to what taken from Jacoby et al. (1984) library, and instrumen- has been reported in the literature. An initial estimate of tal magnitudes and color were computed by integrating the recession velocities can be obtained by measuring the it within the Johnson B and V filter passbands from central wavelength of one of these lines. For example, the Bessell (2005). A zero-point correction was then com- center of Hβ is at 4915A˚ in the NGC 5011B spectrum but puted, to calibrate the color to the standard Johnson at 4873A˚ in the NGC 5011C spectrum (see also Fig. 3). system; the color of an A0I star (B − V = −0.01) was This translates into a velocity of 3290 ± 340kmsec−1 taken from Allen’s Astrophysical Quantities (Cox 1999). for NGC 5011B while that of NGC 5011C would be By repeating this procedure with the two galactic SEDs 750 ± 340kmsec−1, where the error is the minimum un- as input, we found B − V = 0.95 and B − V = 0.47 for certainty estimated above at ∼ 5000A.˚ To compute more NGC 5011B and NGC 5011C, respectively. The spec- precise velocities, we cross-correlated our spectra with trum of NGC 5011C has been extracted in the central that of the radial velocity standard HD125184, thus ob- ′′ 5. 3, and in Fig. 4 one can see that (B − V ) ≈ 0.4 ± 0.08 taining the relative velocity of each galaxy with respect at the center of the galaxy (see below). The synthetic to that of the standard star. If we call this relative ve- color is thus in good agreement with that obtained from locity ∆vgal, then the true heliocentric velocity of the broad-band surface photometry, which means that our galaxy is correction indeed yields a reliable SED. gal gal std gal std After this consistency check, we can interpret the two v⊙ = ∆v + v⊙ + δv⊙ − δv⊙ SEDs using population synthesis techniques. For that where δv⊙ is the heliocentric correction listed in Ta- purpose we used GALAXEV from Bruzual & Charlot ble 1, and the heliocentric velocity of the standard star (2003; BC03), to estimate the luminosity-weighted age std −1 is v⊙ = −12.40 ± 0.05kmsec . From the same table, and metallicity. A grid of model SEDs were generated, gal std −1 based on the Padova 1994 tracks (see BC03 for the ref- one can see that δv⊙ − δv⊙ = −3.7kmsec , where erences) with an input Salpeter stellar mass function. the individual heliocentric corrections have been com- Computations were done for ages ranging from 0.1 Gyr puted using iraf’s task rvcor. The cross-correlation to 20 Gyr, and for five of the available metallicities of the spectra has been done using iraf’s task fxcor. Z =0.0001, 0.0004, 0.004, 0.008, and 0.02. The age was To improve the signal, the spectra were first trimmed to increased in steps of 0.1 Gyr up to 1 Gyr, and in steps 3900 − 6700A˚ and then continuum-normalized, as shown of 1 Gyr for older ages, and the spectra were degraded in Fig. 3. Additionally, in the case of NGC 5011C the to our instrumental resolution with software provided by region containing Hγ and the G band was masked out BC03. The best-fit was searched by minimizing the dif- before cross-correlating the two spectra: the two fea- ference between model and observed SEDs. In this way tures are located close to each other at our spectral res- we found that the best solution is obtained for an age of olution and as Hγ is basically absent in the RV stan- 0.9 ± 0.1 Gyr and Z = 0.004 for NGC 5011C, and for dard spectrum, the algorithm would overestimate the an age of 4 ± 1 Gyr and Z = 0.02 for NGC 5011B. In velocity as it attempts to correlate Hγ in the galaxy the BC03 scale the metallicity of the Sun is Z =0.02, so spectrum with the G band in the standard star spec- the two metallicities are 1/5 solar and solar, respectively. trum. Masking the two spectra accordingly removed The good match between the theoretical and observed the confusion leading to a more realistic velocity esti- mate with a reduced measurement error. At the end of SEDs can be checked in Fig. 2. Note that these results NGC5011B −1 only apply to the central regions of the two galaxies; for this analysis we get ∆v = 3243 ± 50kmsec −1 example the NGC 5011C spectrum was extracted in the and ∆vNGC5011C = 663 ± 96kmsec , so the heliocen- ′′ −1 central 5. 3, which correspond to 95 pc at the estimated tric velocities are v⊙ = 3227 ± 50kmsec and v⊙ = distance of the galaxy (see below). The age and metallic- 647 ± 96kmsec−1 for NGC 5011B and NGC 5011C, re- ity thus represent the central stellar population only, and spectively. since the rest of the galaxy has redder colors presumably it hosts older populations, as is evident from Fig. 4. In- 5. SURFACE PHOTOMETRY deed a kind of tiny, whiter, bulge-like core can be seen To gain insight into NGC5011C’s colors and photomet- in Fig. 1 as well. ric parameters, shallow broad-band B, V, and R frames The total V magnitude computed in §?? allows us to of 90s, 60s, and 30s integration time, respectively, were estimate the mass of NGC 5011C as well. At a fiducial acquired with the EFOSC2 instrument during the same distance of 3.7 Mpc it converts into MV ∼−13.9, and as- observing night (see Table 1). A montage image is shown ⊙ suming MV =4.7, the total luminosity can be computed. in Fig. 1. The two galaxies have similar angular size, 4 but they are of entirely different morphological type. ies in the Virgo cluster (Binggeli & Cameron 1991) pre- NGC 5011B is a classical lenticular (S0) galaxy seen al- dicts an apparent B magnitude of 15.5 ± 1.5mag for most edge-on, while NGC 5011C appears as a low surface a galaxy like NGC5011C with an extinction-corrected 0 brightness, early-type system with a uniform color and a mean effective B surface brightness of hµieff = 22.94 mag hint of a tiny nucleus or central star concentration. arcsec−2 at the Virgo distance. Employing the latest Continuous MeteoMonitor measurements of the atmo- cluster distance modulus of (m − M)Virgo = 31.05mag spheric conditions at La Silla observatory showed that (Jerjen, Binggeli & Barazza 2004, Mei et al. 2006) this the extinction RMS was less than 0.02 magnitudes indi- translates into an absolute magnitude of MB = −15.55± cating that the sky transparency was good. From 9–11 1.5 mag. Combining that number with the apparent B measurements spread in time the photometric zero points magnitude in Table 2 we infer a distance modulus of for the three passbands were calculated: ZP(B)= 26.06± (m − M) = mB − AB − MB = 29.85 ± 1.5mag or a 0.07, ZP(V)= 26.12 ± 0.03, and ZP(R)= 26.19 ± 0.04. distance range of 4.7–18.6 Mpc. The pre-reduction of the raw CCD frames was carried (B) As an alternative, we can use the S´ersic parame- out within MIDAS in the usual manner. After the bias ter n that controls the overall shape of a galaxy to esti- subtraction, pixels that do not respond linearly to flux mate the distance interval. Equation (6) from Binggeli were replaced by a median of the surrounding area, using & Jerjen (1998) that describes the log(n) − mB rela- the latest EFOSC2 bad pixel mask, and then each sci- tion for early-type dwarfs in the Virgo cluster predicts ence frame was flattened using high signal-to-noise sky an apparent B magnitude of mB = 16.29 ± 0.92 mag for flats. All reduced frames were flat to less than 0.08%. NGC5011C at the Virgo distance. Using the same clus- The sky level was determined from star-free regions well ter distance modulus as before yields an absolute magni- away from the galaxies. Then foreground stars and back- tude of −14.76 ± 0.92 mag, a galaxy distance modulus of ground galaxies were removed from the images to allow 29.06 ± 0.92 mag, and a distance range of 4.3–9.9 Mpc. uncontaminated photometric measurements using proce- The main conclusion we can draw from these estimates dures written within the IRAF package. Stars in the based on photometric properties is that the derived dis- vicinity of NGC5011C were replaced by small patches of tance intervals are consistent with NGC5011C being a empty sky. The area underneath the neighbor galaxy member of the CenA Group at 3.7Mpc (Jerjen, Free- NGC5011B was restored using a rectangular sub-image man & Binggeli 2000; Karachentsev 2005) but incom- from the opposite side of the galaxy. The center of patible with the Centaurus cluster distance of 45.3 Mpc NGC5011C was given by the intensity-weighted centroid. (Mieske, Hilker & Infante 2005). The galaxy flux was integrated using circular apertures 6. DISCUSSION AND CONCLUSIONS on the clean BV R images to produce growth curves as a function of the geometric radius r = (ab)1/2 (where a The deep spectroscopy and shallow imaging with the and b are the major and minor axis). The total magni- 3.6m ESO telescope has confirmed a new member of the nearby Cen A Group whose true identity remained hid- tude mT was determined by extrapolation to the asymp- den because of coordinate confusion and wrong redshifts totic intensity, and the effective radius reff was defined in the literature since 23 years. The galaxy’s morphol- as the value at which the intensity reached half of mT . ogy provided the deciding clue. NGC 5011C is only Within reff the mean surface brightness hµieff was also computed. These photometric parameters are listed in 2.25 degrees to the West of the dominant group galaxy Table 2. Surface brightness profiles in the three bands NGC 5128 (Centaurus A). Adopting a mean group dis- tance of 3.7 Mpc this angular separation translates into and extinction-corrected colour gradients (B–R)0, (V– R) , and (B–V) are shown in Figure 4. Thereby we 150 kpc. The two galaxies have also comparable veloc- 0 0 −1 used the extinction values AB =0.503, AV =0.386, and ities with v⊙ = 647 ± 96kmsec for NGC 5011C and −1 AR = 0.311 from Schlegel, Finkbeiner & Davis (1998). v⊙ = 547 ± 5kmsec for NGC 5128. The sky distri- Additionally, S´ersic parameters were measured by fitting bution of all known group members in the vicinity of a S´ersic model (S´ersic 1968): NGC 5128 is shown in Fig. 5. Heliocentric velocities and distances taken from the literature (Jerjen, Binggeli µ(r)= µ +1.086(r/r )n 0 0 & Freeman 2000a; Karachentsev 2005; Karachentsev et to the observed profiles. These model parameters are also al. 2006) are given in parentheses. listed in Table 2 and the corresponding model curves su- NGC 5011C is well within the velocity distribution of perimposed onto the surface brightness profiles in Fig. 4. the Cen A group, and it is definitely not compatible with The quoted uncertainties include formal errors owing to the velocity distribution of the Centaurus cluster. The the sky background removal, photometric uncertainty Cen 30 component of the cluster has a velocity disper- −1 and the fit to the surface brightness profile. The total sion of σv = 586 km sec , so NGC 5011C’s velocity magnitudes are in good agreement with the values pub- is 4 σ lower than the cluster’s average. Although very lished in the ESO-Uppsala catalogue (Lauberts & Valen- high peculiar velocities can be observed in galaxy clus- tijn 1989). ters, this is not the case for the Centaurus cluster, and all Although the current imaging data is not deep enough its members have velocities greater than 1500 km sec−1 to measure an accurate distance for NGC5011C by means (see Fig. 3 in Lucey et al. 1986). of the surface brightness fluctuations method (e.g. Jerjen At 3.7 Mpc, NGC 5011C has an absolute blue magni- et al. 2001) we can make a guess at the distance interval tude of MB = −13.54 mag that makes it the 8th bright- in which NGC5011C is likely to be found with the help of est member of the group comparable in luminosity to the two independent quantities hµieff and n if we accept ESO 269-066 and ESO 384-016. In terms of photometric the galaxy to be of early-type morphology. properties such as the overall (B–R)0 color and the S´ersic (A) The hµieff −mB relation for early-type dwarf galax- parameters the galaxy resembles closely the intermediate 5 type (dS0/Im) dwarf ESO 384-016 (see Table 2 of Jerjen, sence of H I in NGC 5011C. At the tentative distance of Binggeli & Freeman 2000a). 3.7 Mpc a non-detection corresponds to an upper limit of 7 NGC 5011C has a bluer global appearance (hB–Ri0 ≈ MHI ∼ 1.4 × 10 M⊙ which is still a factor of 2.5 higher 6 1.1 mag) than NGC 5011B but unlike for instance the than the Hi mass of (6.0 ± 0.5) × 10 M⊙ detected in blue LSB galaxy UGC 5889 (Vallenari et al. 2005) it ESO 384-016 (Beaulieu et al. 2006). For a most recent shows no sign of on-going star formation neither in the list of H I detections and upper limits in Cen A group 2D image (e.g. young star clusters, H II regions) nor in dwarf galaxies see Bouchard et al. (2006). our 1D-spectrum (emission lines). Our estimated age To obtain a better Hi mass estimate and to measure of 0.9 Gyr for the central stellar population, and the an accurate distance from surface brightness fluctuations presumably older disk, is compatible with this scenario, or the tip magnitude of the red giant branch will be part since much younger stars would be needed to ionize any of a follow-up study of NGC 5011C. existing gas. NGC 5011C is neither a compact blue dwarf nor does it exhibit an irregular stellar distribution or spiral arms. We acknowledge helpful discussions with Dominique Optically, the galaxy has smooth, symmetric isophotes Naef and Linda Schmidtobreick, and we thank the anony- suggesting a classification as dwarf elliptical or interme- mous referee for comments that helped to improve the diate (dE/Im) type galaxy. results of this work. We made use of the HIPASS public A look at the 21cm spectrum from the Hi Parkes data release. The Parkes telescope is part of the Aus- All-Sky Survey (Barnes et al. 2001) observed in the tralia Telescope which is funded by the Commonwealth direction of NGC 5011C gives the impression of some of Australia for operation as a National Facility man- emission at the optical velocity (see Fig. 6). How- aged by CSIRO. This paper has been completed thanks ever, the periodic baseline structure in the velocity range to a stay of I.S. at Mt Stromlo Observatory, supported 150kms−1 < cz < 1500kms−1 makes it impossible to both by ESO (through a Director General grant) and the draw a final conclusion on the possible presence or ab- Australia National Unversity (grant ARC DP0343156).

REFERENCES Barnes, D.G., Staveley-Smith, L., de Blok, W.J.G., et al. 2001, Jerjen, H., Rekola, R., Takalo, L., Coleman, M., Valtonen, M. 2001, MNRAS, 322, 486 A&A, 380,90 Beaulieu, S.F., Freeman, K.C., Carignan, C., Lockman, F.J., Karachentsev, I.D 2005, AJ, 129, 178 Jerjen, H. 2006, AJ, 131, 325 Karachentsev, I.D., Tully, R.B., Dolphin, A., Sharina, M., Bessell, M. S. 2005, ARA&A, 43, 293 Makarova, L., Makarov, D., Kashibadze, O.G., Karachentseva, Binggeli, B. & Cameron, L.M. 1991, A&A, 252, 27 V., Sakai, S., Shaya, E.J., Rizzi, L. 2006, AJ submitted Binggeli, B. & Jerjen, H. 1998, A&A, 333, 17 (astro/ph0603091) Bouchard, A., Jerjen, H., Da Costa, G. S., Ott, J. 2006, AJ, in Kennicutt, R. C., Jr. 1992, ApJ, 388, 310 press (astro-ph/0611423). Lauberts, A., & Valentijn, E. A. 1989, The Surface Photometry Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 Catalogue of the ESO-Uppsala Galaxies, ESO Buzzoni, B., Delabre, B., Dekker, H., Dodorico, S., Enard, D., Lauberts, A. 1982, ESO/Uppsala survey of the ESO(B) atlas, ESO Focardi, P., Gustafsson, B., Nees, W., Paureau, J., Reiss, R. Lucey, J.R., Currie, M.J., Dickens, R.J. 1986 MNRAS, 221, 453 1984, Msngr, 38, 9 Mei, S., Blankeslee, J.P. Cˆot´e, P., Tonry, J.L., West, M.J., Cˆot´eS., Freeman, K.C., Carignan, C., Quinn, P.J. 1997, AJ, 114, Ferrarese, L., Jord´an, A., Peng, E.W., Anthony, A., Merritt, D. 1313 ApJ, in preparation Cox, A.N. (ed), 1999, Allen’s Astrophysical Quantities, Fourth Mieske, S. Hilker, M., Infante, L. 2005, A&A, 438, 103 Edition, Los Alamos, NM Udry, S., Mayor, M., Queloz, D. 1999, In: Precise Stellar Radial Dickens, R.J., Currie, M.J., Lucey, J.R. 1986, MNRAS, 220, 679 Velocities, ASP Conference Series 185, IAU Colloquium 170. Eds. Dressler, A. 1991, ApJS, 75, 241 J. B. Hearnshaw and C. D. Scarfe, p. 367 Fairall, A.P. 1983, MNRAS, 203, 47 Schlegel, D.J., Finkbeiner, D.P. & Davis, M. 1998, ApJ, 500, 525 Hamuy, M., Walker, A.R., Suntzeff, N.B., Gigoux, P., Heathcote, S´ersic, J.-L. 1968, Atlas de galaxias australes, Observatorio S.R., Phillips, M.M. 1992, PASP, 104, 533 Astronomico, Cordoba Hamuy, M., Suntzeff, N.B., Heathcote, S.R., Walker, A.R., Gigoux, Vallenari, A, Schmidtobreick, L. & Bomans, D.J. 2005, A&A, 435, P., Phillips, M.M. 1994, PASP, 106, 566 821 Jacoby, G. H., Hunter, D. A., & Christian, C. A. 1984, ApJS, 56, Warmels, R.H. 1991, in Astronomical Data Analysis Software and 257 Systems I, PASP Conf. Series, Vol. 25, p. 115 Jerjen, H., Bingeli, B., Barazza, F. 2004, AJ, 127, 771 Zaritsky, D., Zabludoff, A. I., & Willick, J. A. 1995, AJ, 110, 1602, Jerjen, H., Binggeli, B., Freeman, K.C. 2000a, AJ, 119, 593 Jerjen, H., Freeman, K.C., Binggeli, B. 2000b, AJ, 119, 166 6

Fig. 1.— EFOSC2 true-color image of NGC 5011B (top red galaxy) and NGC 5011C (bottom blue galaxy). The composite was created from BVR frames and the field of view is ∼ 5′ × 5′. North is up, East to the left. 7

Fig. 2.— The flux-calibrated, and normalized, spectra of NGC 5011B and NGC 5011C are shown in the top and bottom panels, respectively. The dotted curves show the original spectra, while the solid curves show the spectra after correcting for the missing blue flux, and with their flux normalized in the [6000 : 6050]A˚ region. The spectrum of NGC 5011C has been smoothed with a window of three pixels. The vertical dotted lines show the location of the first five Balmer lines (Hα, β, γ, δ, ǫ) red-shifted with the individual velocities of the galaxies. The thick segment shows the (red-shifted) location of the G band. NGC 5011C has evident Balmer lines and only a hint of G band, while the G band and the H and K Ca lines at ∼ 3950 A˚ are clearly present in the spectrum of NGC 5011B. Moreover, only weak Hα and Hβ can be seen. The blue curves are model single stellar populations, computed for an age of 4 Gyr and 0.9 Gyr, and a metallicity Z = 0.02 and Z = 0.004 for NGC 5011B and NGC 5011C, respectively (see text for details). 8

Fig. 3.— The radial velocities were computed by cross-correlating the upper two and lower two spectra shown in this figure. The upper two continuum-normalized spectra are those of NGC 5011B and the RV standard; the lower two spectra are those of NGC 5011C and the same standard, after masking the Hγ and G band region. The vertical solid segments identify Balmer lines, while the dotted ones identify features typical of G-type stars. In particular the heavy segment marks the position of the G band. For the two galaxies, the wavelengths of the lines have been red-shifted using the individual recession velocities computed in Sec. 4. 9

Fig. 4.— (Top panel): BVR surface brightness profiles for NGC5011C as a function of geometric radius r = (ab)1/2, where a and b are the major and minor axis. The best-fitting analytical Sersic profiles are superimposed as black lines. (Bottom panel): extinction-corrected color gradients smoothed with a third-order polynomial. 10

-39o

-40o HIPASS1337-39 (492) Cen 6 (617)

-41o

ESO324-024 (516/3.73) o -42 ESO325-G011 (545/3.4) CenA-dE1 SGC1319.1-4216

o -43 NGC5237 (361) NGC5128 (547/3.66) Decl. (J2000) NGC5011C (647) -44o

-45o AM 1339-445 (3.75) ESO269-066 (784/3.54) AM1318-444 (741) AM1331-451 (828) -46o AM 1343-452 (3.97) 100 kpc

13h40m 13h30m 13h20m 13h10m R.A. (J2000)

Fig. 5.— The galaxy distribution around NGC 5128 (triangle). Early-type galaxies are indicated as filled circles and late-type galaxies as stars. Heliocentric velocities (in km sec−1) and distances (in Mpc) are given in parenthesis if available from the literature. 11

0.04

0.03

) 0.02 -1

0.01

0.00

-0.01

Flux Density (Jy beam -0.02

-0.03

-0.04

0 1000 2000 3000 -1 Velocity, cz (km s )

Fig. 6.— The 21 cm spectrum in the direction of NGC 5011C obtained from the HIPASS public data release - v1.2 May 13, 2000 (south). The vertical line and the grey band indicate the velocity and the 1σ velocity uncertainty as derived from our redshift measurement. Superimposed is a 5-passes boxcar smoothed version of the spectrum (thick line) that suggests a possible signal from Hi emission at the optical position. However the periodic baseline structure in the velocity range 150 km s−1

TABLE 1 Journal of the observations.

−1 UT start Target R.A.(J2000) DEC(J2000) texp(sec) slit Filter/Grism Airmass δv⊙(km sec ) 2006-03-02T06:59:26 NGC5011B/C 13 13 11.95 −43 15 56.0 30 – R 1.033 – 2006-03-02T07:00:32 NGC5011B/C 13 13 11.95 −43 15 56.0 60 – V 1.032 – 2006-03-02T07:02:11 NGC5011B/C 13 13 11.95 −43 15 56.0 90 – B 1.032 – 2006-03-02T00:06:12 LTT2415 05 56 24.30 −27 51 28.8 40 5′′ #11 1.001 −17.85 2006-03-02T00:26:40 LTT3218 08 41 34.10 −32 57 00.1 40 5′′ #11 1.153 −4.80 2006-03-02T01:18:05 LTT3218 08 41 34.10 −32 57 00.1 40 5′′ #11 1.058 −4.88 2006-03-02T03:28:29 LTT3218 08 41 34.10 −32 57 00.1 40 5′′ #11 1.015 −5.10 2006-03-02T04:58:15 LTT4816 12 38 50.94 −49 47 58.8 120 5′′ #11 1.137 17.92 2006-03-02T05:16:36 NGC5011B/C 13 13 12.13 −43 14 47.3 1200 1′′ #11 1.130 20.73 2006-03-02T05:36:36 NGC5011B/C 13 13 11.88 −43 15 55.7 1200 1′′ #11 1.097 20.70 2006-03-02T05:56:36 NGC5011B/C 13 13 12.00 −43 15 21.5 1200 1′′ #11 1.072 20.68 2006-03-02T06:28:37 HD125184 14 18 00.73 −07 32 32.6 1 1′′ #11 1.210 24.42 2006-03-02T06:29:37 HD125184 14 18 00.73 −07 32 32.6 1 1′′ #11 1.208 24.42 2006-03-02T06:30:37 HD125184 14 18 00.73 −07 32 32.6 1 1′′ #11 1.206 24.43 2006-03-02T09:07:31 LTT6248 15 39 00.02 −28 35 33.1 40 5′′ #11 1.009 29.15 Note. — For the target “NGC5011B/C”, the coordinates of NGC 5011B, of NGC 5011C, and of a point in-between the two galaxies, are listed in the three rows. Likewise the heliocentric velocity correction has been computed for the three positions and UTs, in order to show that an average correction can be safely adopted. 13

TABLE 2 Photometric Parameters B V R mT (mag) 14.80 ± 0.08 13.95 ± 0.07 13.50 ± 0.06 reff (arcsec) 21.3 ± 0.7 23.3 ± 0.5 22.3 ± 0.2 −2 hµieff (mag as ) 23.44 ± 0.08 22.77 ± 0.07 22.23 ± 0.06 −2 µ0 (mag as ) 21.71 ± 0.21 21.15 ± 0.14 20.74 ± 0.16 r0 (arcsec) 6.7 ± 1.6 8.3 ± 1.2 9.2 ± 1.3 n (S´ersic) 0.73 ± 0.07 0.76 ± 0.05 0.81 ± 0.05