Stellar populations in the Canis Major over-density
aa r X i v : . [ a s t r o - ph ] J a n Mon. Not. R. Astron. Soc. , 1– ?? (2008) Printed 1 November 2018 (MN L A TEX style file v2.2)
Stellar populations in the Canis Ma jor over-density
Giovanni Carraro ⋆ , Andr´e Moitinho , and Ruben A. V´azquez † ESO, Casilla 19001, Santiago 19, Chile SIM/IDL, Faculdade de Ciˆencias de Universidade de Lisboa, Ed. C8, Campo Grande, 1749-016, Lisboa, Portugal Facultad de Ciencias Astron´omicas y Geof´ısicas de la UNLP, IALP-CONICET, Paseo del Bosque s/n 1900, La Plata, Argentina
Accepted 1988 December 15. Received 1988 December 14; in original form 1988 October 11
ABSTRACT
We performed a photometric multicolor survey of the core of the Canis Major over-density at l ≈ o , b ≈ − . o , reaching V ∼
22 and covering 0 o . × o .
0. The main aimis to unravel the complex mixture of stellar populations toward this Galactic direction,where in the recent past important signatures of an accretion event have been claimedto be detected. While our previous investigations were based on disjointed pointingsaimed at revealing the large scale structure of the third Galactic Quadrant, we nowfocus on a complete coverage of a smaller field centered on the Canis Major over-density. A large wave-length baseline, in the
U BV RI bands, allows us to build upa suite of colour colour and colour magnitude diagrams, providing a much betterdiagnostic tool to disentangle the stellar populations of the region. In fact, the simpleuse of one colour magnitude diagram, widely employed in all the previous studiesdefending the existence of the Canis Major galaxy, does not allow one to separate theeffects of the different parameters (reddening, age, metallicity, and distance) involvedin the interpretation of data, forcing to rely on heavy modeling. In agreement withour previous studies in the same general region of the Milky Way, we recognize ayoung stellar population compatible with the expected structure and extension of theLocal (Orion) and Outer (Norma-Cygnus) spiral arms in the Third Galactic Quadrant.Moreover we interpret the conspicuous intermediate-age metal poor population asbelonging to the Galactic thick disk, distorted by the effect of strong disk warping atthis latitude, and to the Galactic halo.
Key words:
Milky Way – structure: stars.
In the last years, we have used photometric observationsof young open cluster fields to probe the spiral structurein the third Galactic Quadrant (TGQ, 180 o l o ;Carraro et al. 2005a, Moitinho et al. 2006, V´azquez et al.2008), motivated by the very poor knowledge of this por-tion of the Galaxy’s periphery. Interestingly, important lowlatitude accretion phenomena have been recently claimedto be ongoing in this part of the Galaxy, such as the Ca-nis Major over-density (CMa, Bellazzini et al. 2004), andthe Monoceros Ring (MRi, Newberg et al. 2002). Clearly,a detailed description of the structure and stellar popula-tions of the Galactic disc (thin plus thick) is mandatory todiscriminate between Galactic and extragalactic material.The TGQ is a special region of the Milky Way’s outskirts, ⋆ On leave from Dipartimento di Astronomia, Universit´a diPadova, Italy † E-mail: [email protected] (GC); [email protected](AM);[email protected] (RAV); characterized by significant absorption windows as the Pup-pis (l ∼ o ) window (Fitzgerald 1968, Moffat et al. 1979,Janes 1991, Moitinho 2001), which allows one to detect verydistant star clusters (Baume et al. 2006). Besides, and in-terestingly, young star clusters are found at low Galacticlatitudes, underlining the fact that the young Galactic diskis significantly warped in these directions (May et al. 1997,Momany et al. 2004, Moitinho et al. 2006, Momany et al.2006, L´opez-Corredoira et al. 2007).In V´azquez et al (2008), by combining optical and COobservations, we have provided a fresh and very detailedpicture of the spiral structure in the TGQ, showing thatthis region is characterized by a complicated spiral pattern.The outer (Norma-Cygnus) arm is found to be a grand de-sign spiral feature defined by young stars, whereas the re-gion closer to the Sun (d ⊙ less than 9 kpc) is dominatedby a conspicuous inter-arm structure, at l ∼ o , the Lo-cal spiral arm. In this region, Perseus is apparently definedby gas and dust, and does not appear to be traced by anevident optical young stellar population, similarly to whatcan be found in other galaxies such as M 74 (V´azquez et c (cid:13) G. Carraro et al.
Figure 1.
20 arcmin on a side field in the CMa over-density (Field1). This field is centered at RA = 07:22:51, DEC = -30:59:20.North is up, East to the left
Figure 2.
20 arcmin on a side field in the CMa over-density (Field2). This field is centered at RA = 07:20:46, DEC = -31:09:36.North is up, East to the left. al. 2008). The analysis carried out on a substantial frac-tion of the stellar fields we observed revealed a complicatedmixture of young and old populations. Although centeredon catalogued star clusters (Dias et al. 2002), a few colourmagnitude diagrams (CMD) do not reveal star clusters but,and more interestingly, show hints of a young, diffuse, anddistant stellar populations, which has become recently re-ferred to as
Blue Plume ( Bellazzini et al. 2004, Dinescu etal. 2005, Mart´ınez-Delgado et al. 2005,Carraro et al. 2005 ).Since the disk is warped and flared in these directions (Mo-
Figure 3.
20 arcmin on a side field in the CMa over-density (Field3). This field is centered at RA = 07:20:21, DEC= -31:15:43.North is up, East to the left. many et al. 2006), the lines of sight are expected to crossboth the thin and thick disk population in front of a partic-ular target, in a way that the analysis of the CMD becomesvery challenging (see for instance the analysis of the fieldtoward the star cluster Auner 1, Carraro et al. 2007).In this paper we present a photometric analysis in the
UBV RI filters of the stellar populations in 3 wide fieldpointings toward the CMa over-density. Sect. 2 describesthe observation and data reduction strategies. In Sect. 3 wediscuss various colour combination CMDs, while Sects. 4and 5 are dedicated to illustrate and analyze the TCD asa function of magnitude. Finally, Sect. 6 summarizes ourfindings.
UBV RI images of 3 overlapping fields (see Figs. 1 to3) in the Third Quadrant of the Milky Way toward theCMa over-density were obtained at the Cerro Tololo Inter-American Observatory 1.0m telescope, which is operated bythe SMARTS consortium. The telescope is equipped with anew 4k ×
4k CCD camera having a pixel scale of 0 ′′ .289/pixelwhich allows to cover a field of 20 ′ × ′ on the sky. Obser-vations were carried out on the nights of November 28 andDecember 3, 2005. The two nights were part of a 6 nightrun. In the first night we observed the fields (cid:13) , 1– ?? tellar populations in the Canis Major over-density Designation α (2000 . δ (2000 .
0) l b U B V R I Airmass Seeing[deg] [deg] secs secs secs secs secs arcsecField 1 07:22:51 -30:59:20 244.00 -07.50 20,180,1800 10,150,1500 5,60,900 5,60,900 5,60,900 1.00-1.30 0.83-1.02Field 2 07:20:46 -31:09:36 244.00 -08.00 20,180,1800 10,150,1500 5,60,900 5,60,900 5,60,900 1.00-1.30 0.90-1.12Field 33 07:20:21 -31:15:43 244.00 -08.10 20,180,1800 10,150,1500 5,60,900 5,60,900 5,60,900 1.20-1.80 1.28-2.11
Table 1.
List of pointings discussed in this paper. For each pointing, equatorial and galactic coordinates are reported together with theset of filters used, and the range of exposure time, air-mass and typical seeing. The fields are shown in Fig. 1 to 3 (U-B)(B-V)(B-V)(V-R)(V-I)
Figure 4.
Photometric solution in
UBV RI for standard stars.See Table 2 for details. σ , on the right, indicates the rms of thefit. Table 2.
Calibration coefficients. u = +3 . ± . u = +0 . ± . u = +0 . b = +2 . ± . b = − . ± . b = +0 . v = +1 . ± . v = +0 . ± . v = +0 . r = +1 . ± . r = − . ± . r = +0 . i = +2 . ± . i = +0 . ± . i = +0 . was done using the procedure developed by Philip Massey .Briefly, the procedure trims and corrects the images for bias,flat-field, and bad pixels, preparing them from photometricextraction. A series of skyflats was employed in all the filters. Three Landolt (1992) areas (TPhoenix, Rubin 149, andPG 0231+006) were observed several times each night totie instrumental magnitudes to the standard system. Allnights, except the last one, were stable and photometricwith seeing between 0.8 and 1.2 arcsec. The last night wasnon-photometric with bad seeing conditions (see Table 1). Photometry from this last night was tied to the other nightsthrough the comparison of stars in common.Since the photometric solutions were identical, all the stan-dard star measurements were used together in obtaininga single photometric solution for the entire run. This re-sulted in calibration coefficients derived using about 200standard stars. Photometric solutions have been calculatedfollowing Patat & Carraro (2001). Fig. 4 shows the run ofmagnitude differences (standard versus instrumental) for thewhole standard set. Notice that the colour baseline is suffi-ciently broad. On the right, the rms of the fit is shown foreach colour. The calibration equations read: u = U + u + u ( U − B ) + u X (1) b = B + b + b ( B − V ) + b X (2) v = V + v + v ( B − V ) + v X (3) r = R + r + v ( V − R ) + r X (4) i = I + i + i ( V − I ) + i X (5),where UBV RI are standard magnitudes, ubvri are the in-strumental magnitudes, X is the airmass, and the derivedcoefficients are presented in Table 2. We adopted the extinc-tion coefficients typical of the site (Carraro et al. 2005b). The covered areas are shown in Figs. 1 to 3. Data have beenreduced using IRAF packages CCDRED and DAOPHOT.Photometry was done employing the point spread function(PSF) fitting method (Stetson 1987). Particular care hasbeen put in defining the PSF model. A variable PSF wasadopted due to PSF variations across the CCD. In general,up to 40 bright stars have been selected for defining thePSF model. Aperture corrections were estimated from sam-ples of bright PSF stars (typically 15), and then appliedto all the stars. The corrections amounted to 0.250-0.315,0.280-0.300, 0.200-0.280, 0.190-0.270, and 0.210-0.280 magfor the UBV RI filters, respectively, over the entire run. Pho-tometric completeness was estimated following Baume et al.(2006) and was determined to be higher than 50% at V ∼
20. mag.
As discussed in Moitinho et al. (2006), up to now the analy-sis of the stellar populations in the direction of the CMaover-density (Martin et al. 2004, Mart´ınez-Delgado et al2004) has been performed using only two colours (mostly B IRAF is distributed by NOAO, which is operated by AURAunder cooperative agreement with the NSF.c (cid:13) , 1– ?? G. Carraro et al. and R). More recent analysis does not deviate from this ap-proach, and dramatically confirms the limitations and uncer-tainties of having just two colours. In Moitinho et al. (2006)and Carraro et al. (2007) we have clearly demonstrated thathaving multicolour photometry is crucial. Although beingwell known, the importance of multicolour measurements isoften overlooked. Here, as in our previous work, we stressthat the possibility of building colour-colour diagrams (ortwo-colour diagram (TCD), especially U-B vs B-V) is essen-tial when young/early-type stellar populations are present.In Fig. 5 we show the B vs B-R CMD of the center ofthe CMa over-density. Only stars with errors lower than0.10 mag (about 10,000 stars) in both filters are plotted.The diagram is in every way similar to the one presented inMart´ınez-Delgado et al. (2005), except for the magnituderange. Their CMD (their Fig. 1) is several magnitudesdeeper, while the bright stars (B blue plume appears very clearly as a sequence of blue starswhich detaches from the Main Sequence (MS) at B ∼ ∼ galaxy ,occurring 1-2 Gyrs ago (Bellazzini et al. 2005), and laterwas suggested as being the Blue Straggler population ofCMa (Bellazzini et al. 2006). In both cases, this featurewould not be populated by young stars (Carraro et al.2007), but having only two filters there is not much moreone can add. However, if a population 1-2 Gyr old werepresent, a distinctive clump of He-burning stars would beevident, which is not the case, as already also emphasized byMart´ınez-Delgado et al. (2005). Indeed, the recent analysisof de Jong et al. (2006) again highlights the difficulty ofworking with only two filters, which forces the authors torely upon heavy modeling. The lack of any spectroscopicinformation further complicates the scenario.For completeness, we show in Figs. 6 to 8 the CMDs ofthe same region in the V vs B-V, V vs V-I, and V vs U-Bplanes, respectively. The same blue plume (BP) as in the Bvs B-R CMD can be recognized in Fig. 6 and 7, while BPstars in the V vs U-B plane are mixed with the field dwarfstars. Besides the BP, Figs. 5 to 7 display a blurred, but stillconspicuous blue Turn Off (TO) at V ∼ − Following Carraro et al. (2007), we exploit the entire filterbaseline to put more stringent constraints on the propertiesof the stellar populations in the Galactic direction understudy. We employ the U filter in building the TCD in the
Figure 5.
The CMD in the B-R vs B plane of all stars havingphotometric errors smaller than 0.1 in the direction of the CMaover-density
Figure 6.
The CMD in the B-V vs V plane of all stars havingphotometric errors smaller than 0.1 in the direction of the CMaover-density. With filled triangles (red when printed in color) weindicate stars belonging to the blue plumes , as selected in the B-Rvs B CMD. (U-B) vs (B-V) plane, which is shown in Fig. 9 for all thedetected stars having photometric error lower than 0.1 mag.It is well known that the position of a star in the TCDdepends mostly on its spectral type, and does not dependon its distance. The displacement from the Zero Age MainSequences (ZAMS) is then caused by its reddening, and, to c (cid:13) , 1– ?? tellar populations in the Canis Major over-density Figure 7.
The CMD in the V-I vs V plane of all stars havingphotometric errors smaller than 0.1 in the direction of the CMaover-density. With filled triangles (red when printed in color) weindicate stars belonging to the blue plumes , as selected in the B-Rvs B CMD.
Figure 8.
The CMD in the U-B vs V plane of all stars havingphotometric errors smaller than 0.1 in the direction of the CMaover-density. With filled triangles (red when printed in color) weindicate stars belonging to the blue plumes , as selected in the B-Rvs B CMD. a minor extent, by its metallicity. This is illustrated in Fig.10, where ZAMS for dwarf stars from Girardi et al. (2000)are shown for different metallicities.The effect of interstellar absorption is to produce a dis-placement from the unreddened ZAMS (solid line) along thereddening vector represented in the bottom of Fig. 10 for anormal extinction law (solid arrow). This normal extinctionlaw - characterized by a total to selective absorption ratio R V = V V E ( B − V ) = 3.1 - is found to be valid in many regionsof the Milky Way, except for star forming regions. In par-ticular, Moitinho (2001) demonstrated that this law is validin the TGQ.The effect of metallicity is only marginally importantfor stars with spectral types earlier than A0, and becomessizable for spectral types F-G, increasing the size of the bellshaped feature introduced by the ultraviolet excess (Sandageet al. 1969, Norris et al. 1999). The larger the effect, thelower the metal content of a star. For even later spectraltypes, the trend is to have the (B-V) colour redder and the(U-B) bluer at decreasing metallicity.By inspecting Fig. 9, one can immediately recognize tworemarkable features. • The first one is the presence of a group of young stars(at B-V bluer than ∼ blue plume visible in allthe different CMDs in Figs. 5 to 7. • The other one is at the expected location of F and Gstars, namely a prominent population of metal poor stars.This population corresponds to the bulk of blue stars visiblein the CMDs of Figs. 5 to 7 in the form of a thick MS havingthe brightest TO at V ∼ − In this section we focus on the two prominent features of theTCD shown in Fig. 9 and mentioned in the previous section.To this aim, we have split the stars in different V magnitudebins, and produced the corresponding TCDs. The idea be-hind this approach is that at increasing V magnitude weare mainly picking up stars with larger reddening and dis-tance, as exhaustively illustrated in Carraro et al. (2007).The various TCDs are shown in Fig. 11.
We start by analyzing the different panels in Fig. 11 to char-acterize the young stellar population. It is straightforward torecognize how early spectral type stars composing the blueplume are mostly evident between V ∼
15 and V ∼
17, withthe peak of the distribution in the 16 V
17 panel. Herewe face a group a young stars with spectral types in therange B5-A0 reddened by E(B-V) = 0.25 ± V of these stars is in the range 0.1-0.6, and therefore we estimate them to lie at about 9.8 +1 . − . c (cid:13) , 1– ?? G. Carraro et al.
Figure 9.
TCD for all the stars having photometric errors smallerthan 0.1 mag in the direction of the CMa over-density. With filledtriangles (red when printed in color) we indicate stars belongingto the blue plumes , as selected in the B-R vs B CMD.
B0 B9A0 F0 F5 G0G5 K0 K5
Figure 10.
Location of ZAMSs in the TCD as a function ofmetallicity. The reddening vector is indicated by the arrow. Theapproximate position of the main spectral types is indicated kpc from the Sun. Having such spectral types, these starsare younger than 100 Myr or so (Carraro et al. 2005, Moit-inho et al. 2006). In all the other panels of Fig. 11 there isonly marginal evidence of the same early spectral type stel-lar population leading to the conclusion that this populationis located at any distance along the line of sight, but with aclear peak at about 10 kpc. At 10 kpc and l = 244 o , these young stars perfectlymatch the distance and position of the Galactic Outer andLocal spiral arms (Moitinho et al. 2006, V´azquez et al. 2008).Moreover, we have shown that along this line of sight theLocal (Orion arm) is a remarkable structure that stays closeto the formal Galactic plane, b = 0 o , for about 6-7 kpc, andthen starts bending, following the warping of the disk. Atthe latitude sampled in this work, the Orion arm is expectedto reach the Outer arm. So that what is seen is materiallocated all the way along the Local arm until it reaches theOuter arm, causing the appearance of a stellar over-density.This is a clear demonstration that, although remarkable,the distribution of young BP stars in CMa is that expectedfrom the warped spiral structure of the Galaxy and does notrequire postulating the presence of an accreted dwarf galaxyin CMa. The decomposition of the TCD in magnitude bins, as shownFig. 11, also allows to better understand the nature of theolder stellar population toward the CMa over-density.The conspicuous broad bell-shaped structure, likelyproduced by ultraviolet excess, is visible in all the TCDsdownward V ≈
15, and appears with increasing importanceat increasing magnitude down to the limit of our observa-tions. This morphology suggests that the majority of thesestars are F-G dwarfs spanning a variety of metallicities. It isdifficult to assign a precise metallicity range, due to photo-metric errors and different amount of reddening. However,the bulk of these dwarfs may probably span metallicitiesfrom about solar (Z=0.019) to much lower than solar (prob-ably down to Z=0.004, see Fig. 10).It is useful to compare these TCDs with the ones pre-sented by Norris et al. (1999) for metal poor stars. The samebell shape feature as in their Fig. 1c fully supports our in-terpretation of these stars being mostly dwarf metal poorstars. At such a Galactic latitude, and taking into accountthe relatively low absorption, we expect to encounter alongthe line of sight a mix of metal poor stars from the thickdisk and from the halo. Comparison of the colour of our (U-B) envelope with the one of Norris et al. (1999; U-B as blueas -0.35, see their Fig. 2), suggests the presence of stars asmetal poor as [Fe/H] ≈ -2.2 dex.In addition, the series of TCDs in Fig. 11 also revealsthe presence of stars with spectral types later than F, bothdwarfs and giants, at any magnitude bin down to V ∼ ∼ ∼ − blue plume merges with the nearby dwarfMS. The shape of this TO provides an estimate of the age c (cid:13) , 1– ?? tellar populations in the Canis Major over-density Figure 11.
TCD at different V magnitude bins. An empirical ZAMS (solid line, red when printed in color) is shown in each panel toguide the eye. of the population and its minimum distance. To this aimwe consider a mean metallicity of Z = 0.010, and choose asuitable isochrone with the purpose of guiding the eye andproviding constraints on the age and distance which matchthe shape of the TO the best. A reddening of E(V-I) = 0.18 is adopted as representative of this Galactic direction (seealso previous subsection). This value agrees with the mapsof Schlegel et al. (1998) and of Amores & L´epine (2007).In Fig. 12 we superimpose a 6 Gyr isochrone on the Vvs V-I CMD (the one where the TO is more visible), which c (cid:13) , 1–, 1–
TCD at different V magnitude bins. An empirical ZAMS (solid line, red when printed in color) is shown in each panel toguide the eye. of the population and its minimum distance. To this aimwe consider a mean metallicity of Z = 0.010, and choose asuitable isochrone with the purpose of guiding the eye andproviding constraints on the age and distance which matchthe shape of the TO the best. A reddening of E(V-I) = 0.18 is adopted as representative of this Galactic direction (seealso previous subsection). This value agrees with the mapsof Schlegel et al. (1998) and of Amores & L´epine (2007).In Fig. 12 we superimpose a 6 Gyr isochrone on the Vvs V-I CMD (the one where the TO is more visible), which c (cid:13) , 1–, 1– ?? G. Carraro et al. matches the shape of the TO for the adopted metallicityand reddening. This implies a distance modulus (m-M) V ∼ b = 0 o plane. The same isochrone is also plotted for a dis-tance modulus of ∼ ± ± We have presented a photometric analysis in the
UBV RI filters of 3 wide field pointings close to the center of theCanis Major over-density. The goal was to study the stel-lar populations in this region of the Milky Way, where aputative dwarf galaxy in the act of being cannibalized bythe Milky Way, is claimed to exist. The analysis presentedin this paper followed a different strategy from previous in-vestigations of the CMa over-density. Instead of studyingvery large fields in two filters, we have concentrated on asmaller area, but observed in several filters. This approach,frequently employed in the study of star clusters, allowedus to construct several CMDs and the classical (B-V) vs(U-B) TCD, which together constitute a very powerful toolfor detecting young stellar populations. As in our previousstudies of stellar fields in the TGQ we found evidence of adiffuse young stellar population, as expected from the pres-ence of the Local and Outer Galactic spiral arms (Carraroet al. 2005, Moitinho et al. 2006, V´azquez et al. 2008). Onceagain, no indication has been found of an ongoing accretionevent in this direction of the Galactic disk. In addition, theestimated ranges of distance, age and metallicity of the oldermetal poor population are consistent with those of thick diskstars at different distances from the Sun. These findings, to-gether with the results of previous papers by us and otherauthors (Momany et al 2004, 2007), significantly weaken theproposed scenario of a dwarf galaxy in CMa being cannibal-ized by the Milky Way. Instead, all the observational evi-dence fits our current knowledge of the Galactic disk. TheTGQ is indeed a complicated region due to the warp and theexistence of the Local Arm. Only the detailed multicolouranalysis we have been conducting in the last few years couldprovide a clear picture of the structure of the outer disk inthe TGQ.
Figure 12.
V vs V-I CMD of the CMa over-density. Twoisochrones (dashed lines, red when printed in color) have beensuperimposed to guide the eye, illustrate the position of the TO,and provide a rough estimate of the mean age of the population.
ACKNOWLEDGMENTS
This study made use of the SIMBAD and WEBDAdatabases. A.M. acknowledges support from FCT (Portu-gal) through grant PDCT/CTE-AST/57128/2004.
REFERENCES
Amores, E.B., L´epine, J.R.D., 2007, AJ 133, 1519Baume, G., Moitinho, A., V´azquez, R.A., Solivella, G., Car-raro, G., Villanova, S., 2006, MNRAS, 367, 1441Bellazzini, M., Ibata, R., Martin, N., Irwin, M.J., Lewis,G.F., 2004, MNRAS 354, 1263Bellazzini, M., Ibata, R., Martin, N., Lewis, G.F., Conn,B., Irwin, M.J., 2006, MNRAS 366, 865Bensby, T., Zenn, A.E., Oey, M.S., Feltzing, S., 2007, ApJ663, L13Carraro G., V´azquez, R.A., Moitinho, A., Baume, G.,2005a, ApJ 630, L153Carraro G., Geisler, D., Moitinho, A., Baume, G., V´azquez,R.A., 2005b, A&A, 442, 917Carraro G., Moitinho, A., Zoccali, M., V´azquez, R.A.,Baume, G., 2007, AJ, 133, 1058Dias, W.S., Alessi, B.S., Moitinho, A., L´epine, J.R.D.,2002, A&A 389, 871Dinescu, D.I., Mart´ınez-Delgado, D., Girard, T.M.,Pe¨narrubia, J., Rix, H.-W., Butler, D., van Altena, W.F.,2005, ApJ 631, L49Fitzgeralg, M.P., 1968, AJ 73, 177Girardi, L., Bressan, A., Bertelli, G., Chiosi, C., 2000,A&AS, 114, 371Janes, K.A., 1991, in Precision Photometry: Astrophysicsof the Galaxy, Proceedings of the conference held 3-4 Oc-tober, 1990 at Union College, Schenectady, NY. Edited by c (cid:13) , 1– ?? tellar populations in the Canis Major over-density A.G.D. Philip, A.R. Upgren and K.A. Janes. Schenectady,NY: Davis Press, p.233de Jong, J.T.A., Butler, D.J. Rix, H.-W., Dolphin, A.E.,Mart´ınez-Delgado, D., 2007, ApJ 662, 259Landolt, A.U., 1992, AJ 104, 340L´opez-Corredoira, M., Momany, Y., Zaggia, S., Cabrear-Lavers, A., 2007 A&A 472, L47Martin, N.F., Ibata, R.A., Bellazzini, M., Irwin, M.J.,Lewis, G.F., Dehnen, W., 2004, MNRAS 348, 12Mart´ınez-Delgado, D., Butler, D.J., Rix, H.W., Franco,V.I., Pe˜narrubia , J., Alfaro, E.J., Dinescu, D.I., 2005, ApJ633, 205May, J., Alvarez, H., Bronfman, L., 1997, A&A 327, 325Moffat, A.F.J., Jackson, P.D., Fitzgerald, M.P., 1979,A&AS 38, 197Moitinho, A., 2001, A&A 370, 436Moitinho, A., V´azquez, R.A., Carraro, G., Baume, G.,Giorgi, E.E., Lyra, W., 2006, MNRAS 368, L77Momany, Y., Zaggia, S., Bonifacio, P., Piotto, G., de An-geli, F., Bedin, L.R., Carraro, G., 2004, A&A 421, L29Momany, Y., Zaggia, S., Gilmore, G., Piotto, G., Carraro,G., Bedin, L. R., de Angeli, F., 2006, A&A 451, 515Norris, J.E., Ryan, S.G., Beers, T.C., 1999, ApJS 123, 639Patat, F., Carraro, G., 2001, MNRAS 325, 1591Sandage, A., 1969, ApJ 158, 1115Schlegel, D. J., Finkbeiner, D. P., Davis, M. 1998, ApJ,500, 525Stetson, P.,B., 1987, PASP 99, 191V´azquez, R.A., May, J., Carraro, G., Bronfman, L., Moit-inho, A., Baume, G., 2008, ApJ 672, 930 c (cid:13) , 1–, 1–