1: % aa.dem
2: % AA vers. 5.3, LaTeX class for Astronomy & Astrophysics
3: % demonstration file
4: % (c) Springer-Verlag HD
5: % revised by EDP Sciences
6: %-----------------------------------------------------------------------
7: %
8: %\documentclass[referee]{aa} % for a referee version
9: %
10: %\documentclass[onecolumn]{aa}
11: \documentclass[twocolumn]{aa}
12: \usepackage{graphicx}
13: \usepackage{longtable,lscape}
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
15: \usepackage{txfonts}
16: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
17: %
18:
19: \begin{document}
20:
21: \title{Spectroscopic and Photometric evidence of two stellar
22: populations in the Galactic Globular Cluster NGC\ 6121 (M4)
23: \thanks{Based on data collected at the European Southern
24: Observatory with the VLT-UT2, Paranal, Chile.} }
25:
26: %\subtitle{Chemical abundances of RGB stars from UVES spectra. }
27:
28: \author{A.\ F.\ Marino
29: \inst{1},
30: S.\ Villanova
31: \inst{2},
32: G.\ Piotto
33: \inst{1},
34: A.\ P.\ Milone
35: \inst{1},
36: Y.\ Momany
37: \inst{3},
38: L.\ R.\ Bedin
39: \inst{4}
40: \and
41: A.\ M.\ Medling
42: \inst{5}
43: }
44:
45: %\offprints{XXX.\ XXX.}
46:
47: \institute{Dipartimento di Astronomia, Universit\`a di Padova,
48: vicolo dell'Osservatorio 2, Padova, I-35122, Italy, EU\\
49: \email{anna.marino-giampaolo.piotto-antonino.milone@unipd.it}
50: \and
51: Departamento de Astronomia, Universidad de Concepcion, Casilla
52: 160-C, Concepcion, Chile\\
53: \email{svillanova@astro-udec.cl}
54: \and
55: Osservatorio Astronomico di Padova, Vicolo dell'Osservatorio 5,
56: 35122 Padova, Italy\\
57: \email{yazan.almomany@oapd.inaf.it}
58: \and
59: Space Telescope Science Institute, Baltimore, MD 21218, USA\\
60: \email{bedin@stsci.edu}
61: \and
62: Department of Astronomy and Astrophysics, University of
63: California, Santa Cruz, CA, 95064\\
64: \email{amedling@ucolick.org}
65: }
66:
67: \date{Received Xxxxx xx, xxxx; accepted Xxxx xx, xxxx}
68:
69: %__________________________________________________________________
70: %
71:
72: % \abstract{}{}{}{}{}
73: % 5 {} token are mandatory
74:
75: \abstract
76: % context heading (optional)
77: % {} leave it empty if necessary
78: {}
79: % aims heading (mandatory)
80: {We present abundance analysis based on high resolution spectra
81: of 105 isolated red giant branch (RGB) stars in the Galactic
82: Globular Cluster NGC~6121 (M4). Our aim is to study its star population
83: in the context of the multi-population phenomenon recently discovered
84: to affect some Globular Clusters.}
85: % methods heading (mandatory)
86: {The data have been collected with FLAMES+UVES, the multi-fiber
87: high resolution facility at the ESO/VLT@UT2 telescope. Analysis was
88: performed under LTE approximation for the following elements: O, Na,
89: Mg, Al, Si, Ca, Ti, Cr, Fe, Ni, Ba, and NLTE corrections were applied
90: to those (Na, Mg) strongly affected by departure from LTE.
91: Spectroscopic data were coupled with high-precision wide-field UBVI$_{\rm C}$
92: photometry from WFI@2.2m telescope and infrared JHK photometry from 2MASS.}
93: % results heading (mandatory)
94: {We derived an average $\rm {[Fe/H]}=-1.07\pm0.01$ (internal error), and
95: an $\alpha$ enhancement of $\rm {[\alpha/Fe]}=+0.39\pm0.05$ dex (internal error).
96: We confirm the presence of an extended Na-O anticorrelation, and find
97: two distinct groups of stars with significantly different Na and O content.
98: We find no evidence of a Mg-Al anticorrelation.
99: By coupling our results with previous studies on the CN band strength,
100: we find that the CN strong stars have higher Na and Al content and are
101: more O depleted than the CN weak ones.
102: The two groups of Na-rich, CN-strong and Na-poor, CN-weak stars populate two
103: different regions along the RGB.
104: The Na-rich group defines a narrow sequence on the red side of the
105: RGB, while the Na-poor sample populate a bluer, more spread portion of the RGB.
106: In the $\rm {U}$ vs.\ $\rm {U-B}$ color magnitude diagram the RGB
107: spread is present from the base of the RGB to the RGB-tip.
108: Apparently, both spectroscopic and photometric results imply the presence
109: of two stellar populations in M4. We briefly discuss the possible
110: origin of these populations.}
111: % conclusions heading (optional), leave it empty if necessary
112: {}
113:
114: \keywords{Spectroscopy ---
115: globular clusters: individual: NGC~6121}
116:
117:
118: \titlerunning{Spectroscopic and Photometric study of the Globular Cluster M4.}
119: \authorrunning{Marino et al.}
120:
121: \maketitle
122: %
123: %________________________________________________________________
124:
125:
126: \section{Introduction}
127:
128: Observational evidence for variations in the chemical composition of
129: light elements in Globular Cluster (GC) stars were known since Cohen (1978),
130: who noted a scatter in Na among stars in M3 and M13.
131: During the last few decades, high resolution spectroscopic
132: studies have definitely confirmed that a GC stellar population is not
133: chemically homogeneous. Even if GC stars are generally homogeneous in
134: their Fe-peak element content, they show large star-to-star abundance
135: variations in the light elements such as C, N, O, Na, Mg, Al,
136: and others (see Gratton et al.\ 2004 for a review).
137:
138: During the last twenty years, it has become clear that in red giant stars the
139: abundances of some of these elements follow a well defined pattern. In
140: particular, there are clear anticorrelations between the Na and O
141: content, and between Mg and Al. Variations in the molecular CH, CN
142: and NH band strengths, due to a spread in the abundances of carbon and
143: nitrogen have been observed, as well as anticorrelations between CH
144: and CN strengths, and, in some cases, a clear bimodality in the CN
145: content.
146:
147: Despite the spectroscopic observational evidence collected in more
148: than thirty years, the pattern in the light elements is not yet well
149: understood. Two scenarios have been proposed to explain
150: this observed chemical heterogeneity: the
151: evolutionary scenario and the primordial one, both apparently
152: supported by observations.
153:
154: The observed decline of C content (e.g. in M13, as shown by Smith \&
155: Briley 2006) and the decreasing ratio $\phantom{}^{12}\rm
156: {C}/\phantom{}^{13}\rm {C}$ (observed in M4 and NGC~6528 by Shetrone 2003)
157: along the RGB phase, support the evolutionary
158: scenario. According to this theory, the origin of the observed star-to-star
159: scatter in some elements is due to the mixing of CNO-cycle-processed
160: material transported, in a way not well understood yet, to the stellar
161: surface. In this way the observed anticorrelations would be present in
162: the evolved stages of the life of stars, after the RGB bump.
163:
164: At odds with this scenario, in the last few years, spectroscopic
165: studies have revealed light element abundance variations in
166: unevolved main sequence stars and less-evolved RGB stars, fainter
167: than the RGB bump. The Na-O anticorrelation was found at the level of
168: the main sequence turn-off (TO) and sub giant branch (SGB) in M13
169: (Cohen \& Mel\'endez 2005), NGC~6397 and NGC~6752 (Carretta et al.\ 2005;
170: Gratton et al.\ 2001), NGC~6838 (Ram\'irez \& Cohen 2002) and 47~Tuc
171: (Carretta et al.\ 2004). In 47~Tuc, a bimodal distribution in the CN
172: strengths, similar to that found among RGB stars (Norris \& Freeman 1979),
173: was found also in the main sequence (MS) by Cannon et al.\ (1998).
174: Moreover, Grundahl et al.\ (2002) have shown that in
175: NGC~6752 the observed scatter in the Str$\rm {\ddot {o}}$mgren index
176: $c_{1}$ is due to the abundance variations in NH bands in stars both
177: brighter and fainter than the RGB bump. This result is consistent with a
178: primordial scenario since theory does not predict significant mixing
179: below the luminosity of the first dredge-up, observationally
180: corresponding to the magnitude of the RGB bump. These observations
181: suggest that the light element variations should be primordial,
182: i.e. they are derived from the chemical composition of the primordial site
183: where the GC stars have been generated, or alternatively, that a second
184: generation of stars has been formed from a medium enriched in some
185: elements (e.g., the self enrichment model by Ventura et al.\ 2001).
186: A primordial scenario in which such GCs have experienced multiple
187: episodes of star formation challenges the paradigm that GCs host a
188: single stellar population, i.e. that stars of a given cluster are
189: coeval and chemically homogeneous.
190:
191: Very recently, a spectacular, and somehow unexpected confirmation
192: that, at least in some GCs, the origin of the chemical anomalies must
193: be primordial came from high precision photometry from Hubble Space
194: Telescope observations. The first object challenging the paradigm of
195: GCs hosting simple stellar populations was $\omega$~Cen. As shown by
196: Bedin et al.\ (2004), the MS of $\omega$~Cen is split into two
197: sequences. But the most exciting discovery came from the
198: spectroscopic investigation by Piotto et al.\ (2005), who found that
199: the bluest MS is more metal rich than the redder one. The only way to
200: account for the spectroscopic and photometric properties of the two
201: main sequences is to assume that the bluest sequence is also strongly
202: He enhanced, to an astonishingly high Y=0.38. More recently,
203: Villanova et al.\ (2007) showed that the two main sequences split in at
204: least four sub giant branches (SGB) which must be connected in some way
205: to the multiplicity of RGBs identified by Lee et al.\ (1999) and
206: Pancino et al.\ (2000). The metal content of the different SGBs
207: measured by Villanova et al.\ (2007) also implies a large age
208: difference among $\omega$~Cen stellar populations, larger than
209: 1 Gyr. The exact age dispersion depends on the assumed abundances (including
210: the He content) of the different stellar populations, and it is still
211: controversial (Sollima et al.\ 2007). Villanova et al.\ (2007) demonstrated
212: that there is also a third MS, running on the red side of the two main MSs,
213: and likely connected with the anomalous RGB-a of Pancino et al.\ (2000).
214:
215: Omega Centauri was well-known since the seventies because of its peculiar
216: metallicity distribution. It is the only GC showing iron-peak element
217: dispersion (Freeman \& Rodgers 1975 and, more recently, Norris et al.\ 1996,
218: Suntzeff \& Kraft 1996). In a sense, the findings by Pancino et
219: al.\ (2000), Bedin et al.\ (2004), Piotto et al.\ (2005), Villanova et
220: al.\ (2007) and Sollima et al.\ (2007) could simply be considered
221: additional evidence that $\omega$~Cen is so peculiar that it might
222: not be a GC. Perhaps, as suggested by many authors
223: (Freeman 1993, Hughes \& Wallerstein 2000), it might simply be the nucleus
224: of a much larger system, likely disrupted by the tidal field of our Galaxy.\\
225: In this sense the most recent discovery by Piotto et al.\ (2007) that the MS
226: of NGC~2808 is split into three, distinct sequences came as a
227: sort of surprise, shaking at its foundation our understanding of GC
228: stellar populations. NGC~2808 has been always considered a GC, with
229: many peculiar properties regarding its metal content and its
230: color-magnitude diagram, but a genuine, massive GC. Still, it hosts multiple
231: stellar populations. Moreover, also in this case, in view of
232: the negligible dispersion in iron-peak
233: elements in NGC~2808, the only way so far available to reproduce the
234: three MSs is to assume that there are three populations, characterized
235: by three different helium contents, up to an (again) astonishingly high
236: Y=0.40. Interestingly enough, D'Antona et al.\ (2005) already made the
237: hypothesis of three groups of stars, with three different helium
238: abundances in order to explain the peculiar, multi-modal Horizontal Branch
239: (HB) of NGC~2808 (Sosin et al.\ 1997, Bedin et al.\ 2000).
240: Piotto et al.\ (2007) simply found them in the form of a MS split.
241: Piotto et al.\ (2007) also noticed that the different stellar
242: populations in NGC~2808 are consistent with the spectroscopic
243: observations by Carretta et al.\ (2006), who identified three groups of
244: stars with different Oxygen abundances. The fraction of stars in the
245: three abundance groups is in rough agreement with the fraction of stars
246: in the three MSs.
247:
248: NGC~2808 and $\omega$~Cen are, at the moment, the most extreme examples
249: of a rather complex observational scenario. In fact, evidence of
250: multiple populations has been found in other GCs, like NGC~1851
251: (Milone et al.\ 2008), NGC~6388 (Siegel et al.\ 2007, Piotto et al.\
252: 2008), and M54 (Piotto et al., in preparation) in the form
253: of a split in the SGB. In NGC~1851 the presence of a group of RGB
254: stars with enhanced Sr and Ba and strong CN bands, among the majority
255: of CN-normal RGB stars (Yong \& Grundahl 2008), and the presence of a bimodal
256: HB, agrees with the hypothesis of two stellar generations inferred by
257: the observed SGB split.
258:
259: The extremely peculiar HB of NGC~6388 (Rich et al.\ 1997), with the
260: presence of extremely hot HB stars (Busso et al.\ 2007) was already
261: interpreted in terms of multiple, helium enhanced, population by
262: Sweigart \& Catelan (1998), and in the more detailed analysis by
263: Caloi \& D'Antona (2007).
264:
265: As for M54, it has been recognized as the nucleus of the Sagittarius
266: dwarf galaxy (Da Costa \& Armandroff 1995; Bassino \& Muzzio 1995;
267: Layden \& Sarajedini 2000), and it could simply represent what
268: $\omega$~Cen was long time ago.
269:
270: It is worth noting that the clusters showing multiplicities in their CMDs are among the most
271: massive GCs of our Galaxy ($\rm {M > 10^6 M_\odot}$) and all of them have
272: peculiar HBs, as well as peculiar abundances, including Na-O
273: anticorrelations.
274: However, we must also note that much less massive GCs, with no
275: evidence of multiple populations identified so far, show very large
276: star-to-star abundance variations. It is noteworthy to recognize here
277: that the Na-O anticorrelation has been found in about 20 GCs (see
278: Carretta et al.\ 2006 for the most updated list).
279:
280: In summary, it is clear that GCs are not as simple systems as thought
281: in the past. Up to now, we lack a complete explanation for the mechanisms
282: necessary to understand the observational scenario. A
283: systematic study of the chemical abundances of many stars in GCs is
284: needed in order to better understand the star formation history of
285: these objects.
286:
287: In this work we present a study on the chemical abundances of the GC
288: NGC~6121 (M4) from high resolution spectra of its RGB
289: stars.
290:
291: As far as we know, M4 shows no evidence of multiple stellar
292: populations in its CMD, and its mass ($\rm {log {\frac
293: {M}{M\odot}}=4.8}$, Mandushev et al.\ 1991) is much smaller than the
294: mass of the clusters with the photometric peculiarities discussed above.
295:
296: Chemical abundances from high resolution spectra of M4 RGB stars
297: have been already measured by different groups of investigators:
298: Gratton, Quarta \& Ortolani 1986 (hereafter GQO86), Brown \&
299: Wallerstein 1992 (BW92), Drake et al.\ 1992 (D92),
300: Ivans et al.\ 1999 (I99), and Smith et al.\ (2005).
301: These authors have found a range of [Fe/H] between $-$1.3 dex
302: (BW92) and $-$1.05 dex (D92). For further details, see I99 who have summarized
303: the chemical abundances found in previous studies.
304: A study by Norris 1981 (hereafter N81) of 45 RGB stars showed a CN
305: bimodality in M4, e. g., stars of very similar magnitudes and colors
306: have a bimodal distribution of CN band strengths.
307: By analyzing 4 RGB stars, two selected from the CN-weak group and two from the
308: CN-strong one, D92 (and then Drake et al.\ 1994)
309: found differences in the Na and Al content. I99 found that O is
310: anticorrelated with N, whereas Na and Al abundances are larger in
311: CN-strong stars. Looking at the evolutionary states of these stars, both I99 and Smith
312: \& Briley 2005 (SB05) did not find any strong correlation between CN band
313: strength and the position on the CMD. More recently, Smith et al.\ (2005),
314: studying the fluorine abundance in seven RGB stars in M4, found a large
315: variation in $\phantom{}^{19}$F, which is anticorrelated with the Na and Al
316: abundances. In these previous studies, both the evolutionary scenario and
317: a primordial one have been taken into account in order to explain the
318: light element variations and the CN bimodality in M4.
319:
320: In this work, we analyze high resolution spectra in order to study
321: chemical abundances for a large sample of M4 RGB stars and compare our results
322: with those of previous studies.
323: In Section 2 we provide an overview of the observations and target sample.
324: The membership criterion used to separate the probable cluster stars is
325: described in Section 3, and the procedure to derive the chemical
326: abundances is in Section 4. We present our results in Section 5, and
327: discuss them in Section 6 and Section 7. Section 8 summarizes the results of this work.
328:
329: \begin{figure}[hpbt]
330: \centering
331: \includegraphics*[width=9.5cm]{Fig1.ps}
332: \caption{Distribution of the UVES target
333: stars on the $\rm {B}$ vs.\ $\rm {(B-I)}$ CMD corrected for
334: differential reddening.}
335: \label{Fig1}
336: \end{figure}
337:
338: \section{Observations and Data Reduction}
339:
340: Our dataset consists of UVES spectra collected in
341: July-September 2006, within a project devoted to the
342: detection of spectroscopic binaries in the GC M4, and to the
343: measurement of the geometric
344: distance of this cluster (Programs 072.D-0742 and 077.D-0182). Data are
345: based on single 1200-1800 sec exposures obtained with FLAMES/VLT@UT2
346: (Pasquini et al.\ 2002) under photometric conditions and a typical seeing of
347: 0.8-1.2 arcsec.
348: The 8 fibers feeding the UVES spectrograph were centered on 115 isolated stars
349: (no neighbours within a radius of 1.2 arcsec brighter than $\rm {V+2.5}$,
350: where $\rm {V}$ is the magnitude of the target star)
351: from $\sim1$ mag below the HB to the tip of the RGB of M4, in the magnitude
352: range $10.5<{\rm V}<14.0$.
353:
354: The UVES spectrograph was used in
355: the RED 580 setting. The spectra have a spectral coverage of
356: $\sim$2000 \AA \ with the central wavelength at 5800 \AA. The typical
357: signal to noise ratio is ${\rm S/N\sim 100-120}$.
358:
359: Data were reduced using UVES pipelines (Ballester et al.\ 2000),
360: including bias subtraction, flat-field correction, wavelength
361: calibration, sky subtraction and spectral rectification. Stars
362: were carefully selected from high quality photometric $\rm {UBVI_{C}}$
363: observations by Momany et al.\ (2003)
364: obtained with the Wide Field Imager (WFI) camera at the ESO/MPI 2.2m
365: telescope (total field of view $34\times33~{\rm arcmin}^{2}$) coupled
366: with infrared JHK 2MASS photometry (Skrutskie et al.\ 2006). Given
367: the spatially variable interstellar reddening across the cluster
368: (Cudworth \& Rees 1990; Drake et al.\ 1994; Kemp et al.\ 1993; Lyons
369: et al.\ 1995; and I99) we have corrected our CMDs for this effect (as
370: done in Sarajedini et al.\ 2007).
371: Fig.~\ref{Fig1} shows the position of the target stars on the
372: corrected B vs (B-I) CMD.
373:
374: \section{Radial velocities and membership}
375:
376: In the present work, radial velocities were used as the membership criterion
377: since the cluster stars all have similar motion with respect to the observer.
378: The radial velocities (${\rm v_{r}}$) of the
379: stars were measured using the IRAF FXCOR task, which cross-correlates
380: the object spectrum with a template. As a template, we used a
381: synthetic spectrum obtained through the spectral synthesis code
382: SPECTRUM (see {\sf http://www.phys.appstate.edu/spectrum/spectrum.html} for
383: more details), using a Kurucz model atmosphere with roughly the mean
384: atmospheric parameters of our stars $\rm {T_{eff}=4500}$ K, $\rm {log(g)=2.0}$,
385: $\rm {v_{t}=1.4}$ km/s, $\rm {[Fe/H]=-1.10}$. At the end, each radial velocity was
386: corrected to the heliocentric system. We calculated a first approximation
387: mean velocity and the r.m.s ($\sigma$) of the velocity distribution.
388: Stars showing $\rm {v_{r}}$ out of more than
389: $3\sigma$ from the mean value were considered probable field objects and
390: rejected, leaving us with 105 UVES spectra of probable members.
391: After this procedure, we obtained a new mean radial velocity of
392: $70.6\pm1$ km/s from all the selected spectra,
393: which agrees well with the values in the literature (Harris 1996,
394: Peterson et al.\ 1995).
395:
396: \section{Abundance analysis}
397:
398: The chemical abundances for all elements, with the exception of Oxygen,
399: were obtained from the equivalent widths (EWs) of the spectral lines.
400: An accurate measurement of EWs first requires a good determination
401: of the continuum level. Our relatively metal-poor stars, combined with our
402: high S/N spectra, allowed us to proceed in the following way. First,
403: for each line, we selected a region of 20 \AA \ centered on the line itself
404: (this value is a good compromise between having enough points, i. e. a good statistic, and
405: avoiding a too large region where the spectrum can be not flat).
406: Then we built the histogram of the distribution of the flux where the peak is a
407: rough estimation of the continuum. We refined this determination by fitting a
408: parabolic curve to the peak and using the vertex as our continuum estimation.
409: Finally, the continuum determination was revised by eye and corrected by hand
410: if a clear discrepancy with the spectrum was found.
411: Then, using the continuum value previously obtained, we fit a gaussian curve
412: to each spectral line and obtained the EW from integration.
413: We rejected lines if affected by bad continuum determination, by non-gaussian
414: shape, if their central wavelength did not agree with that expected from
415: our linelist, or if the lines were too broad or too narrow with respect to the
416: mean FWHM.
417: We verified that the gaussian shape was a good approximation for our spectral
418: lines, so no lorenzian correction was applied.
419: The typical error for these measurements is 3.6 m\AA,
420: as obtained from the comparison of the EWs measured on stars
421: having similar atmospheric parameters (see Sect. 5.1 for further details).
422:
423: \subsection{Initial atmospheric parameters}
424:
425: Initial estimates of the atmospheric parameters were derived from WFI
426: $\rm {BVI}$ photometry. First-guess effective temperatures ($\rm
427: {T_{eff}}$) for each star were derived from the $\rm {T_{eff}}$-color relations
428: (Alonso et al.\ 1999).
429: $\rm {(B-V)}$ and $\rm {(V-I)}$ colors were de-reddened using a reddening
430: of ${\rm E(B-V)=0.36}$ (Harris 1996). Surface gravities
431: ($\rm {log(g)}$) were obtained from the canonical equation:
432:
433: \begin{center}
434: ${\rm log(g/g_{\odot}) = log(M/M_{\odot}) + 4\cdot
435: log(T_{eff}/T_{\odot}) - log(L/L_{\odot}) }$
436: \end{center}
437:
438: \noindent
439: where the mass ${\rm M/M_{\odot}}$ was derived from the spectral
440: type (derived from T$_{\rm eff}$) and the luminosity class of stars (in
441: this case we have a III luminosity class) through the grid of
442: Straizys \& Kuriliene (1981).
443: The luminosity ${\rm L/L_{\odot}}$ was obtained from the absolute
444: magnitude ${\rm M_V}$, assuming an apparent distance modulus
445: of $\rm {(m-M)_{V}=12.83}$ (Harris 1996).
446: The bolometric correction ($\rm {BC}$) was derived by adopting the relation
447: $\rm {BC-T_{eff}}$ from Alonso et al.\ (1999). \\
448: Finally, microturbolence velocity ($\rm {v_{t}}$) was obtained from the
449: relation (Gratton et al. 1996):
450:
451: \begin{center}
452: ${\rm v_{t}\ (km/s) = 2.22-0.322\cdot log(g)}$
453: \end{center}
454:
455: \noindent
456: These atmospheric parameters are only initial guesses and are
457: adjusted as explained in the following Section.
458:
459: \subsection{Chemical abundances}
460:
461: The Local Thermodynamic Equilibrium (LTE) program MOOG (freely
462: distributed by C. Sneden, University of Texas at Austin) was used to
463: determine the metal abundances.
464:
465: The linelists for the chemical analysis were obtained from the VALD
466: database (Kupka et al.\ 1999) and calibrated using the Solar-inverse technique.
467: For this purpose we used the high resolution, high S/N Solar spectrum
468: obtained at NOAO ($National~Optical~Astronomy~Observatory$, Kurucz et
469: al.\ 1984).
470: We used the model atmosphere interpolated from the Kurucz (1992) grid
471: using the canonical atmospheric parameters for the Sun: $\rm {T_{eff}=5777}$ K,
472: $\rm {log(g)}=4.44$, $\rm {v_{t}=0.80}$ km/s and $\rm {[Fe/H]=0.00}$.
473:
474: The EWs for the reference Solar spectrum were obtained in the same
475: way as the observed spectra, with the exception of the strongest lines, where
476: a Voigt profile integration was used.
477: In the calibration procedure, we adjusted the value of the line
478: strength $\rm {log(gf)}$ of each
479: spectral line in order to report the abundances obtained from all the
480: lines of the same element to the mean value.
481: The chemical abundances obtained for the Sun
482: are reported in Tab.~\ref{Tab1}.
483: The derived Na and Mg abundances were corrected for the effects of
484: departures from the LTE assumption, using the prescriptions by Gratton
485: et al.\ (1999).
486:
487: \begin{table} [!hbp]
488: \caption{Measured Solar abundances ($\rm
489: {log\epsilon(X)=log(N_{X}/N_{H})+12)}$.}
490: \label{Tab1}
491: \centering
492: \begin{tabular}{lcr}
493: \hline
494: \hline
495: Element & UVES &lines\\
496: \hline
497: $\rm {OI}$ & 8.83 & 1 \\
498: $\rm {NaI_{NLTE}}$ & 6.32 & 4 \\
499: $\rm {MgI_{NLTE}}$ & 7.55 & 3 \\
500: $\rm {AlI}$ & 6.43 & 2 \\
501: $\rm {SiI}$ & 7.61 & 12 \\
502: $\rm {CaI}$ & 6.39 & 16 \\
503: $\rm {TiI}$ & 4.94 & 33 \\
504: $\rm {TiII}$ & 4.96 & 12 \\
505: $\rm {CrI}$ & 5.67 & 32 \\
506: $\rm {FeI}$ & 7.48 & 145\\
507: $\rm {FeII}$ & 7.51 & 14 \\
508: $\rm {NiI}$ & 6.26 & 47 \\
509: $\rm {BaII}$ & 2.45 & 2 \\
510: \hline
511: \end{tabular}
512: \end{table}
513: \noindent
514:
515: With the calibrated linelist, we can obtain refined atmospheric parameters
516: and abundances for our targets.
517: Firstly model atmospheres were interpolated from the grid of Kurucz models by
518: using the values of $\rm {T_{eff}}$, $\rm {log(g)}$, and $\rm {v_{t}}$
519: determined as explained in the previous section. Then, during the abundance analysis,
520: $\rm {T_{eff}}$, $\rm {v_{t}}$ and $\rm {log(g)}$ were adjusted in
521: order to remove trends in Excitation Potential (E.P.) and
522: equivalent widths vs.\ abundance respectively, and to satisfy
523: the ionization equilibrium. $\rm {T_{eff}}$ was
524: optimized by removing any trend in the relation between abundances
525: obtained from the FeI lines with respect to the E.P.
526: The optimization of $\rm {log(g)}$ was done in order to satisfy the
527: ionization equilibrium of species ionized differently: we have used
528: FeI and FeII lines for this purpose. We changed the value of log(g)
529: until the following relation was satisfied:
530:
531: \begin{center}
532: ${\rm log\epsilon(FeII)_{\odot}-log\epsilon(FeI)_{\odot}=log\epsilon(FeII)_{\ast}-log\epsilon(FeI)_{\ast}}$
533: \end{center}
534:
535: \noindent
536: We optimized $\rm {v_{t}}$ by removing any trend in the relation
537: abundance vs.\ EWs of the spectral lines. This iterative procedure
538: allowed us to optimize the values of atmospheric parameters on the
539: basis of spectral data, independent of colors. This is an
540: important advantage for those clusters, as M4, that are projected in
541: regions characterized by a differential reddening.\\
542: The adopted atmospheric parameters, together with the coordinates, the
543: U,B,V,I$_{\rm C}$, and the 2MASS J,H,K magnitudes, are listed in
544: Tab.~\ref{Tab8}. All the reported magnitudes are not corrected
545: for differential reddening.\\
546: Having $\rm {T_{eff}}$ determinations independent of
547: colors, we can verify whether the $\rm {T_{eff}}$-scale is affected
548: by systematic errors. For this purpose we used the $\rm {T_{eff}}$
549: and $\rm {[Fe/H]}$ of our stars (see Sec. 6.1) to obtain intrinsic
550: $\rm {B-V}$ colors
551: from Alonso's relations. These colors were compared with
552: WFI photometry corrected for differential reddening.
553: We obtained a reddening of:
554:
555: \begin{center}
556: $\rm E(B-V)=+0.34\pm0.02$
557: \end{center}
558:
559: \noindent
560: This value agrees well with the $\rm {E(B-V)=0.36}$ of Harris (1996)
561: and with I99 who find $\rm {E(B-V)=0.33\pm0.01}$. Therefore, we
562: conclude that our $\rm {T_{eff}}$-scale is not affected by strong
563: systematic errors. \\
564: Finally we present a $\rm {v_{t}-log(g)}$ relation obtained from our data:
565:
566: \begin{center}
567: $\rm {v_t=-(0.254\pm0.016)\cdot log(g) +(1.930\pm0.035)}$
568: \end{center}
569:
570: \noindent
571: It gives a mean microturbolence velocity 0.15 km/s lower
572: than that given by Gratton et al. (1996), but this is not a surprise
573: because also in other papers (Preston et al. 2006) Gratton et
574: al. (1996) was found to overestimate the microturbolence, especial in
575: our $\rm {T_{eff}}$ regime. We underline that our formula is valid
576: for objects in the same $\rm {T_{eff}}$-$\rm {log(g)}$ regime as
577: out targets, i.e. cold giant stars.
578:
579: \subsection{The O synthesis}
580:
581: Instead of using the EW method, we measured the O content of our stars
582: by spectral synthesis. This method is necessary because of the blending of the
583: target O line at 6300 \AA\ with other spectral lines (mainly the Ni transition
584: at 6300.34 \AA).
585: For this purpose, we used a linelist from the VALD archive
586: calibrated on the NOAO Solar spectrum as done before for the other elements,
587: but in this case we changed the log(gf) parameter of our spectral
588: lines in order to obtain a good match between our synthetic spectrum
589: and the observed one.
590:
591: With the calibrated linelist, it is possible to establish the O
592: content of our targets. To this aim, we used the standard MOOG running
593: option $synth$ that computes a set of trial synthetic spectra and
594: matches these to the observed spectrum. The synthetic spectra were
595: obtained by using the atmospheric parameters derived in
596: the previous section. In this way, the Oxygen abundances were deduced
597: by minimization of the observed-computed spectrum
598: difference. An example of a spectral synthesis is plotted in
599: Fig.~\ref{Fig2} where the spectrum of the observed star \#34240 (thick
600: line) is compared with synthetic spectra (thin lines) computed for O
601: abundances $\rm {[O/Fe]}$= +0.30, +0.47, and +0.60 dex. In the plotted
602: spectral range, the observed spectrum is contaminated by two telluric
603: features (not present in the models), but none affecting the O line.
604:
605: The O line is faint, and we could measure the O content only for a
606: sub-sample of 93 stars. In the remaining 12 spectra, the bad quality of the
607: Oxygen line prevented us from measuring accurate O abundance.
608:
609: \begin{figure}
610: \centering
611: \includegraphics[bb=18 144 592 475,width=9.6cm]{Fig2.ps}
612: \caption{Spectrum of the star \#34240, compared with synthetic spectra
613: in the region 6297-6302 \AA, which includes the Oxygen line at
614: 6300.31 \AA. Synthetic spectra were computed for O abundances
615: [O/Fe] = +0.30, +0.47, and +0.60 dex. Thick line is the
616: observed spectrum, thin lines are the synthetic ones.}
617: \label{Fig2}
618: \end{figure}
619:
620: \begin{figure}
621: \centering
622: \includegraphics[bb=18 144 592 518,width=9cm]{Fig3.ps}
623: \caption{Box plot of the M4 giant star element abundances from UVES
624: data. For each box the central horizontal line is the median
625: of the data, while the upper and lower lines indicate
626: the higher and lower 1$\sigma$ value respectively.
627: The points represent the individual measurements.}
628: \label{Fig3}
629: \end{figure}
630:
631: \section{Results}
632:
633: The wide spectral range of the UVES data allowed us to derive the
634: chemical abundances of several elements. The mean abundances
635: for each element are listed in Tab.~\ref{Tab2} together
636: with the error of the mean, the rms scatter
637: and the number of measured stars ($\rm {N_{stars}}$). The rms scatter
638: (hereafter $\sigma_{\rm {obs}}$) is assumed to be the $\rm {68.27th}$ percentile
639: of the distribution of the measures of the single stars and the error
640: of the mean is the rms divided by $\sqrt{\rm {N_{stars}-1}}$.
641: In the last column the abundances derived by I99 were listed as a comparison.
642:
643: The derived Na and Mg abundances were corrected for the effects of
644: departures from the LTE assumption (NLTE correction) using the prescriptions by Gratton
645: et al.\ (1999), as done for the Sun. In the paper all the used Na and Mg abundances are
646: NLTE corrected, also were not explicitely indicated.
647: The mean NLTE corrections obtained for [Na/Fe] and [Mg/Fe] are $-0.02$ dex and
648: $+0.07$ dex, respectively.
649: For the Fe abundance, we do not distinguish between the results obtained
650: from I and II ionization stages, because their values are necessarily the same
651: in the chosen method.
652:
653: Chemical abundances for the single stars are listed in Tab.~\ref{Tab9}.
654: A plot of our measured abundances is shown in Fig.~\ref{Fig3},
655: where, for each box, the central horizontal line is the median values for the
656: element, and the upper and lower lines indicate the 1$\sigma$ values higher and lower than the
657: median values, respectively. The points represent individual measurements.\\
658:
659: \begin{table} [!htpq]
660: \caption{The average abundances of M4 stars The results by I99 are
661: shown for comparison in Column 5.}
662: \label{Tab2}
663: \centering
664: \begin{tabular}{ l c c c c }
665: \hline \hline
666: \multicolumn{4}{c}{This work} & I99 \\
667: \hline
668: & &$\sigma_{\rm {obs}}$ & $\rm {N_{stars}}$ &\\
669: \hline
670: ${\rm [OI/Fe]}$ & $+0.39\pm0.01$ & 0.09 & 93 & $+0.25\pm0.03$ \\
671: ${\rm [NaI/Fe]_{NLTE}}$ & $+0.27\pm0.02$ & 0.17 & 105 & $+0.22\pm0.05$ \\
672: ${\rm [MgI/Fe]_{NLTE}}$ & $+0.50\pm0.01$ & 0.06 & 105 & $+0.44\pm0.02$ \\
673: ${\rm [AlI/Fe]}$ & $+0.54\pm0.01$ & 0.11 & 87 & $+0.64\pm0.03$ \\
674: ${\rm [SiI/Fe]}$ & $+0.48\pm0.01$ & 0.05 & 105 & $+0.55\pm0.02$ \\
675: ${\rm [CaI/Fe]}$ & $+0.28\pm0.01$ & 0.04 & 105 & $+0.26\pm0.02$ \\
676: ${\rm [TiI/Fe]}$ & $+0.29\pm0.01$ & 0.05 & 105 & $+0.30\pm0.01$ \\
677: ${\rm [TiII/Fe]}$ & $+0.35\pm0.01$ & 0.06 & 105 & $+0.30\pm0.01$ \\
678: ${\rm [CrI/Fe]}$ & $-0.04\pm0.01$ & 0.05 & 105 & -- \\
679: ${\rm [FeI/H]}$ & $-1.07\pm0.01$ & 0.05 & 105 & $-1.18\pm0.00$ \\
680: ${\rm [NiI/Fe]}$ & $+0.02\pm0.01$ & 0.03 & 105 & $+0.05\pm0.01$ \\
681: ${\rm [BaII/Fe]}$ & $+0.41\pm0.01$ & 0.09 & 103 & $+0.60\pm0.02$ \\
682: \hline
683: \end{tabular}
684: \end{table}
685:
686: \subsection{Internal errors associated to the chemical abundances}
687:
688: \noindent The measured abundances of of every element vary from
689: star to star as a consequence of both measurement errors and
690: intrinsic star to star abundance variations.
691: In this section, our final goal is to search for evidence of intrinsic
692: abundance dispersion in each element by comparing the observed
693: dispersion $\sigma_{\rm {obs}}$ and that produced by internal errors
694: ($\sigma_{\rm {tot}}$). Clearly, this requires an accurate analysis of
695: all the internal sources of measurement errors.
696: We remark here that we are interested in star-to-star intrinsic
697: abundance variation, i.e. we want to measure the internal intrinsic
698: abundance spread of our sample of stars. For this reason, we
699: are not interested in external sources of error which are systematic
700: and do not affect relative abundances.\\
701: It must be noted that two sources of errors mainly contribute
702: to $\sigma_{\rm {tot}}$. They are:
703: \begin{itemize}
704: \item the errors $\sigma_{\rm {EW}}$ due to the uncertainties in the
705: EWs measures;
706: \item the uncertainty $\sigma_{\rm {atm}}$ introduced by errors in the
707: atmospheric parameters adopted to compute the chemical abundances.
708: \end{itemize}
709:
710: \noindent In order to derive an estimate of $\sigma_{EW}$, we consider that
711: EWs of spectral lines in stars with the same atmospherical
712: parameters and abundances are expected to be equal; any difference in the
713: observed EWs can be attributed to measurements errors. Hence, in order
714: to estimate $\sigma_{\rm {EW}}$, we applied the following procedure:
715: \begin{itemize}
716: \item{Select two stars (\#33683 and \#33946) from our sample
717: characterized by exactly the same $\rm {T_{eff}}$, and $\rm {log(g)}$, $\rm {v_{t}}$,
718: [Fe/H] within a range of 0.05;}
719: \item {Compare the EWs of the iron spectral lines in these two stars
720: and calculate the standard deviation of the distribution of the
721: differences between the EWs divided by $\sqrt{2}$. We found a value
722: of 3.6 m\AA \ that
723: we assume to represent our estimate for the errors in the EWs.
724: Iron lines were selected because Fe abundance has always been found to be
725: the same for stars in GCs, with the exception of $\omega$~Cen. This
726: means that any differences in the EW for the same couple of lines is
727: due only to measurement errors;}
728: \item {Calculate the corresponding error in abundance measurements
729: ($\sigma_{\rm {EW}}$) by using a star (\#21728) at intermediate
730: temperature, assumed to be representative of the entire sample.
731: To this aim, we selected two spectral lines for each element
732: in order to cover the whole E.P. range,
733: changed the EWs by 3.6 m\AA \ and calculated the corresponding mean
734: chemical abundance variation. This number divided by $\sqrt {\rm
735: {N_{lines}-1}}$ is our $\sigma_{\rm {EW}}$. Results are listed in
736: Tab.~\ref{Tab3}}.
737: \end{itemize}
738:
739: \noindent A more elaborate analysis is required to determine
740: $\sigma_{\rm {atm}}$.
741: We followed two different approaches.
742:
743: In order to better understand the method, it is important
744: to summarize the role of the atmospheric parameters in the determination
745: of chemical abundances.
746: As fully described in sections 4.1, the best estimate of chemical
747: abundances, $\rm {T_{eff}}$, $\rm {v_{t}}$ and $\rm {log(g)}$ obtained
748: from the spectrum of a single star are those that satisfy at the same
749: time three conditions:
750: \begin{itemize}
751: \item {removing any trend from the straight line that best fits the
752: abundances vs.\ E.P.;}
753: \item {removing any trend from the straight line that best fits the
754: abundances as a function of equivalent widths;}
755: \item {satisfy the ionization equilibrium through the condition $\rm
756: {[FeI/H]=[FeII/H]}$.}
757: \end{itemize}
758: Therefore, any error in the slope of the best fitting lines
759: and in FeI/FeII determinations produces an error on the measured abundance.
760:
761: In order to derive the error in temperature we applied the
762: following procedure.
763: First we calculated, for each star, the errors associated with the slopes of
764: the best least squares fit in the relations between
765: abundance vs.\ E.P.
766: The average of the errors corresponds to the typical error on the slope.
767: Then, we selected three stars representative of the entire sample
768: (\#29545, \#21728, and \#34006) with high, intermediate, and low
769: $\rm {T_{eff}}$, respectively.
770: For each of them, we fixed the other parameters and varied the temperature
771: until the slope of the line that best fits the relation between
772: abundances and E.P. became equal to the respective
773: mean error. This difference in temperature represents an estimate of
774: the error in temperature itself. The value we found is $\rm {\Delta
775: T_{\rm eff} = 40}$ K.
776:
777: \begin{figure*}[!htpb]
778: \centering
779: \includegraphics[bb=170 385 450 590,width=6.5cm]{Fig4.ps}
780: \caption{$\rm {I}$ magnitude (corrected for differential reddening) as
781: a function of the atmospheric parameters for our target stars. For
782: $\rm {T_{eff}}$ the data were fitted with a parabolic curve, for
783: log(g) and $\rm {v_{t}}$ with a straight line.}
784: \label{Fig4}
785: \end{figure*}
786:
787: The same procedure was applied for $\rm {v_{t}}$, but using
788: the relation between abundance and EWs. We obtained a
789: mean error of $\Delta\rm {v_{t}}=0.06$ km/s.
790:
791: Since $\rm {log(g)}$ has been obtained by imposing the condition $\rm
792: {[FeI/H]=[FeII/H]}$, and the measures of $\rm {[FeI/H]}$ and $\rm
793: {[FeII/H]}$ have averaged uncertainties of
794: $\overline {\sigma_{\rm {star}}[\rm {FeII/H}]}$ and $\overline {\sigma_{\rm {star}}[\rm
795: {FeI/H}]}$ (where $\sigma_{\rm {star}}[\rm {Fe/H}]$ is the dispersion of
796: the iron abundances derived by the various spectral lines in
797: each spectrum and given by MOOG, divided by $\sqrt {\rm {N_{lines}-1}}$), in order to
798: associate an error to the measures of gravity we have varied the
799: gravity of the three representative stars such that the relation:
800: \begin{center}
801: $\rm {[FeI/H] - \overline{\sigma_{star}[FeI/H]} = [FeII/H] + \overline{\sigma_{star}[FeII/H]}}$
802: \end{center}
803: was satisfied. The obtained mean error is $\Delta [\rm {log(g)}]$=0.12.
804:
805: Once the internal errors associated to the atmospheric parameters were
806: calculated, we re-derived the abundances of the three reference stars by
807: using the following combination of atmospheric parameters:
808: \begin{itemize}
809: \item ($\rm {T_{eff}} \pm \Delta (\rm {T_{eff}})$, $\rm {log(g)}$, $\rm {v_{t}}$)
810: \item ($\rm {T_{eff}} $, $\rm {log(g)} \pm \Delta (\rm {log(g)})$, $\rm {v_{t}}$)
811: \item ($\rm {T_{eff}} $, $\rm {log(g)}$, $\rm {v_{t}} \pm \Delta (\rm {v_{t}})$)
812: \end{itemize}
813: where $\rm {T_{eff}}$, $\rm {log(g)}$, $\rm {v_{t}}$ are the measures
814: determined in Section 4.2.
815:
816: The resulting errors in the chemical abundances due to uncertainties in
817: each atmospheric parameter are listed in Tab.~\ref{Tab3} (columns 2, 3
818: and 4). The values of $\sigma_{\rm {atm}}$ (given by the squared sum of
819: the uncertainties introduced by each single parameter) listed in
820: column 5 are our final estimates of the error introduced by the
821: uncertainties of all the atmospheric parameters on the chemical
822: abundance measurements.
823:
824: We also used a second approach to estimate $\sigma_{\rm {atm}}$ as
825: confirmation of the first method. In order
826: to verify our derived uncertainties related to the temperature,
827: gravity and microturbolence, we considered that RGB stars with the
828: same $\rm {I}$ magnitude must have the same atmospheric parameters.
829: Hence, we started by plotting the magnitude $\rm {I}$ as a function of
830: $\rm {T_{eff}}$, $\rm {log(g)}$ and $\rm {v_{t}}$ as shown in
831: Fig.~\ref{Fig4}. We see in this figure that the typical internal error in our
832: $\rm {I}$ photometry ($\sim0.01$ mag) translates into an error in temperature
833: of $\sim3$ K, in $\rm {log(g)}$ of $\sim0.005$, and in $\rm {v_{t}}$ of $\sim0.002$ km/s, each of which is
834: absolutely negligible. The data were fitted by a parabolic curve in the case of $\rm {T_{eff}}$,
835: and by straight lines in the cases of $\rm {log(g)}$ and $\rm
836: {v_{t}}$. We determined the differences between the $\rm {T_{eff}}$,
837: log(g) and $\rm {v_{t}}$ of each star and the corresponding value on the fitting
838: curve. Assuming the 68.27th percentile of the absolute values of these
839: differences as an estimate of the dispersion of the points
840: around the fitting curve, all the stars with a distance from the curve
841: larger than $3\sigma$ were rejected (the open triangles in
842: Fig.~\ref{Fig4}). The four probable AGB stars (filled triangles) that
843: are evident on the $\rm {B-(B-I)}$ CMD of Fig.~\ref{Fig1} were also rejected.
844: With the remaining stars, the curve was refitted and the differences
845: between the atmospheric parameters of the stars and the fit were
846: redetermined. The 68.27th percentile of their absolute
847: values are our estimate of the uncertainties on the atmospheric
848: parameters.
849:
850: By using this second method we obtained the uncertainties $\Delta {\rm
851: {T_{eff}}}=$37 K, $\Delta \rm {log(g)}=$0.12 and $\Delta \rm {v_{t}}=$0.06
852: km/s, consistent with the ones determined with the first method.
853:
854: \begin{table*}[!htpq]
855: \caption{Sensitivity of derived UVES abundances to the atmospheric parameters
856: and EWs. We reported the total error due to the atmospheric parameters
857: ($\sigma_{\rm atm}$), due to the error in EW measurement
858: ($\sigma_{\rm EW}$), the squared sum of the two ($\sigma_{\rm tot}$), and the
859: observed dispersion ($\sigma_{\rm obs}$) for each element. The FeII
860: observed dispersion is not reported because it is necessarily the
861: same as FeI due to the method used in this paper to obtain
862: atmospheric parameters.}
863: \centering
864: \label{Tab3}
865: \begin{tabular}{ l r r r c c c c c}
866: \hline\hline
867: \ & $\rm {\Delta T_{eff}} (K)$ & $\Delta$log(g) & $\Delta \rm {v_{t}}$ (km/s) & \
868: $\sigma_{\rm {atm}}$ & $\sigma_{\rm {EW}}$ & $\sigma_{\rm {tot}}$ & $\sigma_{\rm {obs}}$\\
869: \hline
870: &$+40$ &$+0.12$ &$+0.06$& & & & \\ \hline
871: ${\rm [OI/Fe]}$ &-- &-- &-- & -- & -- & 0.04 &0.09 \\
872: ${\rm [NaI/Fe]}$ &$+0.00$ &$-0.01$ &$+0.02$& 0.02 & 0.04 & 0.04 &0.17 \\
873: ${\rm [MgI/Fe]}$ &$-0.01$ &$-0.01$ &$+0.01$& 0.02 & 0.06 & 0.06 &0.06 \\
874: ${\rm [AlI/Fe]}$ &$-0.01$ &$-0.02$ &$+0.02$& 0.03 & 0.06 & 0.07 &0.11 \\
875: ${\rm [SiI/Fe]}$ &$-0.04$ &$+0.03$ &$+0.02$& 0.05 & 0.03 & 0.06 &0.05 \\
876: ${\rm [CaI/Fe]}$ &$+0.01$ &$-0.02$ &$+0.00$& 0.02 & 0.02 & 0.03 &0.04 \\
877: ${\rm [TiI/Fe]}$ &$+0.04$ &$-0.02$ &$+0.00$& 0.04 & 0.02 & 0.04 &0.05 \\
878: ${\rm [TiII/Fe]}$ &$+0.03$ &$-0.01$ &$-0.01$& 0.03 & 0.05 & 0.06 &0.06 \\
879: ${\rm [CrI/Fe]}$ &$+0.02$ &$-0.01$ &$+0.00$& 0.02 & 0.06 & 0.06 &0.05 \\
880: ${\rm [FeI/H]}$ &$+0.04$ &$+0.00$ &$-0.03$& 0.05 & 0.01 & 0.05 &0.05 \\
881: ${\rm [FeII/H]}$ &$-0.03$ &$+0.06$ &$-0.01$& 0.07 & 0.04 & 0.08 &-- \\
882: ${\rm [NiI/Fe]}$ &$-0.01$ &$+0.02$ &$+0.01$& 0.02 & 0.02 & 0.03 &0.03 \\
883: ${\rm [BaII/Fe]}$ &$+0.05$ &$-0.04$ &$-0.03$& 0.07 & 0.05 & 0.09 &0.09 \\
884: \hline
885: \end{tabular}
886: \end{table*}
887:
888: Our best estimate of the total error associated to the abundance
889: measures is calculated as
890: \begin{center}
891: $\rm {\sigma_{tot}=\sqrt{\sigma_{EW}^{2}+\sigma_{atm}^{2}}}$
892: \end{center}
893: listed in the column 7 of Tab.~\ref{Tab3}.
894: For O $\rm {\sigma_{tot}}$ was obtained in a different way since its measure
895: was not based on the EW method (see Section 6.4). In all the plots for the
896: error bars associated with the measure of abundance we adopted $\sigma_{\rm tot}$.
897:
898: In addition, systematic errors on abundances can be introduced by
899: systematic errors in the scales of
900: $\rm {T_{eff}}$, log(g), $\rm {v_{t}}$, and by deviations of the real
901: stellar atmosphere from the model (i.e. deviations from LTE).
902: A study of the systematic effects goes beyond the purposes of
903: this study, where we are more interested in internal star to star variation in
904: chemical abundances.
905:
906: Comparing $\sigma_{\rm {tot}}$ with the observed
907: dispersion $\sigma_{\rm {obs}}$ (column 8) at least for the abundance
908: ratios ${\rm [Na/Fe]}$, ${\rm [O/Fe]}$ and $\rm {[Al/Fe]}$, we found
909: $\sigma_{\rm {obs}}$ significantly larger than $\sigma_{\rm {tot}}$,
910: suggesting the presence of a real spread in the content of these
911: elements within the cluster.
912:
913: \section{The chemical composition of M4}
914: In the following sections, we will discuss in detail the measured
915: chemical abundances. In addition, we will use the abundances of C and
916: N from the literature to extend our analysis to those elements.
917: The errors we give here are only internal. External ones can be estimated
918: by comparison with other works (see Tab.~\ref{Tab2}).
919:
920: \subsection{Iron and iron-peak elements}
921:
922: The weighted mean of the [Fe/H] found in the 105 cluster members is
923:
924: \begin{center}
925: ${\rm [Fe/H]=-1.07\pm0.01\ dex}$
926: \end{center}
927:
928: \noindent
929: The other Fe-peak elements (Ni and Cr) show roughly the Solar-scaled
930: abundance.
931:
932: The agreement between $\sigma_{\rm obs}$ and $\sigma_{\rm {tot}}$ (see
933: Tab.~\ref{Tab3}), allows us to conclude that there
934: is no evidence for an internal dispersion in the iron-peak
935: element content, suggesting that this cluster is homogeneous in metallicity
936: down to the $\sigma_{\rm {obs}}$ level.
937:
938: \subsection{$\alpha$ elements}
939:
940: \noindent The chemical abundances for the $\alpha$ elements
941: O, Mg, Si, Ca, and Ti are listed in Tab.~\ref{Tab2}. All
942: the $\alpha$ elements are overabundant. Ca and Ti are
943: enhanced by $\sim$0.3 dex, while Si and Mg by $\sim$0.5 dex.
944: The results for $\rm {[O/Fe]}$ will be discussed in detail in Section 6.4.
945:
946: Previous investigators (GQO86, BW92 and I99) have also found significant
947: overabundances for both Si and Mg. In particular, for $\rm {[Si/Fe]}$ they found
948: values higher than 0.5 dex, slightly higher than our results, while our
949: $\rm {[Mg/Fe]} = 0.50$ lies in the middle of
950: the literature results that range between 0.37 (BW92) and 0.68 dex (GQO86).
951: As noted by I99, the abundances of these two elements in M4 are higher if
952: compared to the ones found in other globular clusters showing a similar
953: metallicity, i. e. M5 (Ivans et al.\ 2000). I99 suggest that the higher
954: abundances in M4 should be primordial, e.g. due to the chemical composition
955: of the primordial site from which the cluster formed.
956:
957: For $\rm {[Ca/Fe]}$ we obtained a scatter of 0.04 dex,
958: differing from I99, who found $\sigma$=0.11 dex (see discussion in
959: Section $6.6$).
960: TiI and TiII show the same abundance within $3\sigma$ (see Tab.~\ref{Tab2}),
961: with a mean of the two different ionized species of
962: $\rm {[Ti/Fe]}=+0.32\pm0.04$.
963:
964: The average of all the $\alpha$ elements, gives an $\alpha$
965: enhancement for M4 of:
966:
967: \begin{center}
968: ${\rm [\alpha/Fe]=+0.39\pm0.05 \ dex}$
969: \end{center}
970:
971: \noindent
972: In this case the agreement between $\sigma_{\rm obs}$
973: and $\sigma_{\rm {tot}}$ again allows us to conclude that, down to the
974: $\sigma_{\rm obs}$ level, there
975: is no evidence of internal
976: dispersion for these elements, with the exception of O, which we will discuss in Section 6.4.
977:
978: \subsection{Barium}
979:
980: As in I99 and BW92, we have found a strong overabundance of Ba. We
981: have $\rm {[Ba/Fe]}=0.41\pm 0.01$ which is about 0.2 dex smaller than
982: in I99. In any case, our results confirm that Ba is significantly
983: overabundant in M4, at variance with what found by GQO86, possibly
984: solving this important issue. As deeply discussed by I99, this high
985: Barium content cannot be accounted for mixing processes, but must be a
986: signature of the presence of s-process elements in the primordial
987: material from which M4 stars formed. The excess of s-process elements
988: provides some support to the idea
989: that the formation of the stars we now observe in
990: M4 happened after intermediate-mass AGB stars have polluted the
991: environment, or that it lasted for long enough to allow intermediate-mass AGB
992: stars to strongly pollute the lower mass forming stars. We
993: do not find any significant dispersion in Ba content for the bulk of
994: our target stars, though there are a few outliers worth further discussion (see below).
995:
996: \subsection{Na-O anticorrelation}
997:
998: Variations in light element abundances is common among GCs,
999: and is also present in M4.
1000: In particular, Sodium and Oxygen have
1001: very large dispersions: we obtained $\sigma_{\rm {\rm [Na/Fe]}}$=0.17
1002: and $\sigma_{\rm [O/Fe]}$=0.09.
1003: A large star to star scatter in Sodium and Oxygen abundance in M4 was found
1004: also in other previous studies (I99, D92).
1005:
1006: \begin{figure}[!hpb]
1007: \centering
1008: \includegraphics[width=9cm]{Fig5.ps}
1009: \caption{${\rm [Na/Fe]}$ vs.\ $\rm {[O/Fe]}$ abundance ratios. The error
1010: bars represent the typical errors $\sigma_{\rm tot}$ from Tab.~\ref{Tab3}.}
1011: \label{Fig5}
1012: \end{figure}
1013:
1014: \begin{figure}[!hpb]
1015: \centering
1016: \includegraphics[width=8.5cm]{Fig6.ps}
1017: \caption{Histogram of the distribution of the ${\rm [Na/Fe]}$
1018: abundances. The error bars represent the Poisson errors.}
1019: \label{Fig6}
1020: \end{figure}
1021:
1022: Sodium and Oxygen show the typical anticorrelation found in many other
1023: GCs (Gratton et al.\ 2004), as shown in Fig.~\ref{Fig5}, where the ${\rm
1024: [Na/Fe]}$ values are plotted as a function of ${\rm [O/Fe]}$. The
1025: open blue circles represent the 12 stars for which it was not possible to obtain
1026: a good estimate of the Oxygen abundance.
1027:
1028: Since the Oxygen abundance was derived from just one spectral line, we adopted
1029: the following the following procedure to derive a raw estimate of the
1030: typical error $\sigma_{\rm {tot}}$ associated with each measure: we selected the group of
1031: stars with ${\rm [Na/Fe]}$$<$0.20, assuming them to be homogeneous in
1032: their O content. Under this hypothesis, the dispersion in O can be
1033: assumed to be due to the random error associated with the measured O
1034: abundance. We obtained $\sigma_{\rm [O/Fe]}=0.04$ dex, which
1035: corresponds to the error bar size in Fig.~\ref{Fig5} and to the
1036: $\sigma_{\rm tot}$ in Tab.~\ref{Tab3}.
1037:
1038: The only previous study showing the Na-O anticorrelation in M4 was that by
1039: I99. They analysed ${\rm [Na/Fe]}$ and ${\rm [O/Fe]}$ abundances
1040: for 24 giant stars from high resolution spectra taken at the Lick and
1041: McDonald Observatories. With respect to their work, our dispersion for
1042: [O/Fe] is quite a bit smaller (0.09 dex against their value of 0.14 dex),
1043: and our average value is larger (see Tab.~\ref{Tab2}).
1044: We have seven stars in common with their high resolution sample, and another
1045: seven with their medium resolution sample for which they derived only
1046: the Oxygen abundance. The
1047: comparison between their results and the present ones are listed in
1048: Tab.~\ref{Tab4}, where the last two columns give
1049: the differences between I99 and our results. We can note that our
1050: abundances are systematically higher than those of I99. The effect is
1051: higher for the ${\rm [O/Fe]}$, but is present for Sodium too, and could
1052: reflect the difference of $\sim$ 0.1 dex we found for the iron with
1053: respect to I99.
1054: This systematic difference does not affect the
1055: shape of the Na-O anticorrelation, but it does emphasize the
1056: overabundance of O in M4 stars with respect to other clusters with
1057: similar metallicity, as already noticed by I99: M4 stars formed in an
1058: environment particularly rich in O and possibly in Na.
1059:
1060: The most interesting result of our investigation is that,
1061: thanks to the large number of stars in our sample, we can show that
1062: the ${\rm [Na/Fe]}$ distribution is bimodal (Fig.~\ref{Fig6}).
1063: Setting an arbitrary separation between the two
1064: peaks in Fig.~\ref{Fig6} at $\sim$0.2 dex, we obtain, for the two
1065: groups of stars with higher and lower Na content, a mean ${\rm
1066: [Na/Fe]}=0.38\pm0.01$ and ${\rm [Na/Fe]}=0.07\pm0.01$,
1067: respectively.
1068:
1069: \begin{figure}[!hpb]
1070: \centering
1071: \includegraphics[width=8.7cm]{Fig7.ps}
1072: \caption{Left: Distribution of stars along the Na-O anticorrelation
1073: represented by the $\rm {[O/Na]}$ ratios.
1074: Right: Distribution of the projected position P of stars
1075: on the parametric curve plotted in the inset pannel. The
1076: coordinates of the edges of the curve are indicated.}
1077: \label{Fig7}
1078: \end{figure}
1079:
1080: \begin{table*}[!htpq]
1081: \caption{Comparison between our ${\rm [Na/Fe]}$ and $\rm {[O/Fe]}$ and
1082: those of I99 for the stars in common. The Oxygen values were
1083: compared both with
1084: the high resolution sample of I99 ($\rm {[O/Fe]_{I99}^{(a)}}$) and
1085: with the medium resolution sample ($\rm {[O/Fe]_{I99}^{(b)}}$). The
1086: last three columns give the differences $\rm {[el/Fe]_{this~work}-[el/Fe]_{I99}}$.}
1087: \centering
1088: \label{Tab4}
1089: \begin{tabular}{ c c c c c c c c c }
1090: \hline
1091: \hline
1092: ID &${\rm [OI/Fe]_{this~work}}$&$\rm {[OI/Fe]_{I99}^{(a)}}$&$\rm {[OI/Fe]_{\rm I99}^{(b)}}$&$\rm {[NaI/Fe]_{this~work}}$&$\rm {[NaI/Fe]_{I99}^{(a)}}$&$\Delta({\rm O^{(a)}})$&$\Delta({\rm O^{(b)}})$&$\Delta({\rm Na})$\\
1093: \hline
1094: L1411 &$+0.28$ &$+0.20$ &$+0.07$&$+0.51$ &$+0.43$&$+0.08$ &$+0.21$ &$+0.08$ \\
1095: L1501 &$+0.34$ &$+0.10$ &$ -- $&$+0.51$ &$+0.42$&$+0.24$ &$--$ &$+0.09$ \\
1096: L1514 &$+0.50$ &$+0.41$ &$+0.41$&$+0.17$ &$+0.01$&$+0.09$ &$+0.09$ &$+0.16$ \\
1097: L2519 &$+0.48$ &$+0.37$ &$+0.37$&$+0.02$ &$-0.19$&$+0.11$ &$+0.11$ &$+0.21$ \\
1098: L2617 &$+0.25$ &$+0.01$ &$+0.07$&$+0.44$ &$+0.50$&$+0.24$ &$+0.18$ &$-0.06$ \\
1099: L3612 &$+0.25$ &$+0.10$ &$+0.12$&$+0.46$ &$+0.47$&$+0.15$ &$+0.13$ &$-0.01$ \\
1100: L3624 &$+0.48$ &$+0.29$ &$+0.27$&$+0.18$ &$+0.10$&$+0.19$ &$+0.21$ &$+0.08$ \\
1101: L1403 &$+0.31$ &$--$ &$+0.22$&$+0.53$ &$--$ &$--$ &$+0.09$ &$--$ \\
1102: L2410 &$+0.39$ &$--$ &$+0.32$&$+0.11$ &$--$ &$--$ &$+0.07$ &$--$ \\
1103: L2608 &$+0.43$ &$--$ &$+0.37$&$+0.24$ &$--$ &$--$ &$+0.06$ &$--$ \\
1104: L1617 &$+0.24$ &$--$ &$+0.17$&$+0.46$ &$--$ &$--$ &$+0.07$ &$--$ \\
1105: L4415 &$+0.44$ &$--$ &$+0.37$&$+0.37$ &$--$ &$--$ &$+0.07$ &$--$ \\
1106: L4421 &$+0.28$ &$--$ &$+0.07$&$+0.44$ &$--$ &$--$ &$+0.21$ &$--$ \\
1107: L4413 &$+0.22$ &$--$ &$+0.27$&$+0.48$ &$--$ &$--$ &$-0.05$ &$--$ \\
1108: \hline
1109: mean &$+0.35$ &$+0.21$ &$+0.24$&$+0.35$ &$+0.25$& $+0.16$& $+0.11$&$+0.08$\\
1110: %$\pm$ &$~~0.03$ &$~~0.06$ &$~~0.04$&$~~0.05$&$~~0.11$&$~~0.03$&$~~0.02$&$~~0.04$\\
1111: \hline
1112: \end{tabular}
1113: \end{table*}
1114:
1115: In correspondence with the two peaks in the ${\rm [Na/Fe]}$
1116: distribution, there are also two $\rm {[O/Fe]}$ groups
1117: (Fig.~\ref{Fig5}): the first group is centered at $\rm {[O/Fe]\sim0.30}$,
1118: while the second is at $\rm {[O/Fe]\sim0.47}$, with a few stars
1119: having intermediate Oxygen abundance.\\
1120:
1121: We have followed two methods to trace the distribution of the stars on the Na-O
1122: plane.
1123: The first method uses the ratio $\rm {[O/Na]}$ between the O and the Na
1124: abundances (Carretta et al.\ 2006) which appears to be the
1125: best indicator to trace the star distribution along the Na-O
1126: anticorrelation, because this ratio continues to vary even at the
1127: extreme values along the distribution. The distribution of
1128: the [O/Na] is plotted in Fig.~\ref{Fig7} (left panel).
1129: We can identify two bulks of RGB stars in this plot: one
1130: at $\rm {[O/Na]\sim0.40}$ and the second one at $\rm {[O/Na]\sim-0.10}$.
1131: There might be another group of stars between
1132: [O/Na]$\sim$0 and [O/Na]$\sim$0.25, but we cannot provide conclusive
1133: evidence.
1134:
1135: Note that the two groups of stars in the Na-O anticorrelation have
1136: their corresponding peaks at $\rm{[O/Fe]\sim0.47}$ and $\rm {[O/Fe]\sim0.30}$,
1137: so both groups are O-enhanced.\\
1138: There are no differences in $\rm {[Fe/H]}$ content between these
1139: two groups of stars: both of them have $\rm {[Fe/H]=-1.07}$ dex.
1140:
1141: We have also analysed the distribution of stars along the Na-O
1142: anticorrelation with an alternative procedure whose steps are
1143: briefly described below:
1144:
1145: \begin{itemize}
1146:
1147: \item{A parametric curve following the observed Na-O distribution as shown
1148: in the inset of the right panel of Fig.~\ref{Fig7} has been traced;}
1149:
1150: \item{Each observed point in the Na-O distribution has been projected
1151: on the parametric curve;}
1152:
1153: \item For each projected point, the distance (P) from the origin of
1154: the parametric curve (indicated with 0 in the inset of the
1155: right panel of Fig.~\ref{Fig7}) has been calculated;
1156:
1157: \item{A histogram of the P distance distribution has been constructed.}
1158:
1159: \end{itemize}
1160: The histogram is shown in the right panel of Fig.~\ref{Fig7}.
1161: In this case also, two peaks are evident: the
1162: first one at $\rm {P\sim0.28}$, and the second one at $\rm {P\sim0.58}$.
1163:
1164: In conclusion, we obtained the Na-O anticorrelation for a large sample
1165: (93 RGB stars) in the globular cluster M4.
1166: The distribution of the objects on the Na-O anticorrelation
1167: is clearly bimodal.
1168:
1169: \subsection{Mg-Al anticorrelation}
1170:
1171: Aluminium abundances have been determined from the EWs of the AlI lines at
1172: 6696 \AA \ and 6698 \AA, and Mg abundances from the EWs of the MgI
1173: doublet at 6319 \AA, 6318 \AA , and of the line at
1174: 5711\AA. Figure~\ref{Fig8} shows the [Mg/Fe] as a function of
1175: [Al/Fe] and [O/Fe] (upper panels) and [Mg/Fe] and [Al/Fe] as a
1176: function of [Na/Fe] (lower panels).
1177:
1178: In each panel the red line represents the best least squares fit to the data.
1179: The value $\rm {a}$ is the slope of the best fit straight line $\rm {y=ax+b}$.
1180: There is no clear Mg-Al anticorrelation (upper left panel of
1181: Fig.~\ref{Fig8}), although Al is more spread out than Mg ($\rm
1182: {\sigma_{[Al/Fe]}=0.11}$ vs.\ $\rm {\sigma_{[Mg/Fe]}=0.06}$). Both Al and
1183: Mg are overabundant, with average value of $\rm {[Al/Fe]=0.54\pm0.01}$
1184: and $\rm {[Mg/Fe]=0.50\pm0.01}$ (internal errors).
1185:
1186: The right-bottom panel of Fig.~\ref{Fig8} shows that there is a
1187: correlation between $[\rm {Al/Fe]}$ and $\rm {[Na/Fe]}$ (higher Al
1188: abundances for higher Na content), though it is less pronounced than
1189: the Na-O anticorrelation. Aluminium spans over a range of $\sim0.4$ dex,
1190: while Sodium covers a range of $\sim0.6$ dex. The spread in Al content of M4
1191: is smaller than in other clusters of similar metallicity, like M5
1192: (Ivans et al.\ 1999) and M13 (Johnson et al.\ 2005).
1193:
1194: \noindent Figure~\ref{Fig9} shows that there might also be a
1195: correlation between the ${\rm [Al/Fe]}$ and ${\rm [Ba/Fe]}$.
1196:
1197: Mg does not clearly correlate neither with Na (left-bottom panel of
1198: Fig.~\ref{Fig8}) nor O (upper-right panel of Fig.~\ref{Fig8}).
1199:
1200: The absence of a Mg-Al anticorrelation is somehow surprising, in view of the
1201: presence of a strong double peaked
1202: Na-O anticorrelation, and of a correlation between Al and
1203: Na and possibly Al and Ba. All these correlations seems to indicate the
1204: presence of material that has gone through the s-process phenomenon.
1205: Assuming that the Na enhancement comes from the proton capture
1206: process at the expense of Ne, we can also expect that a similar process is
1207: the basis of the Al enhancement at the expense of Mg. Still, we do
1208: not find any significant dispersion in the Mg content. This problem has
1209: already been noted by I99, who suggested that the required Mg destruction
1210: is too low to be measured (see their Section 4.2.2 for further details).
1211:
1212: Here we only add that, despite our large sample,
1213: neither in the Al nor in the Mg (and Ba) distribution
1214: can we see any evidence of the two peaks so clearly visible as in the
1215: Na and O distribution.
1216: But, in spite of the lack of a clear Mg-Al anticorreletion, we found that Na
1217: poor stars have on average higher Mg and lower Al. The difference
1218: between the median Mg and Al abundances for the Na-poor and
1219: Na-rich samples are: $\rm {[Mg/Fe]_{Na-rich}-[Mg/Fe]_{Na-poor}=-0.04\pm0.01}$ and
1220: $\rm {[Al/Fe]_{Na-rich}-[Al/Fe]_{Na-poor}=+0.08\pm0.02}$
1221: This should be consistent with the scenario proposed by I99 (see their
1222: Sec. 4.2.2 for more details) who predicted that a drop of only 0.05
1223: dex in Mg is needed to account for the observed increase in the abundance of Al.
1224: (Obviously this requires the further hypothesis that all the
1225: Na enhanced stars are also Al enhanced.)
1226:
1227: \begin{figure}[!hpb]
1228: \centering
1229: \includegraphics[width=9cm]{Fig8.ps}
1230: \caption{Bottom panels: $\rm {[Al/Fe]}$ and $\rm {[Mg/Fe]}$ abundance ratios
1231: are plotted as a function of ${\rm [Na/Fe]}$. Upper panels:
1232: $\rm {[Mg/Fe]}$ ratio as a function of $\rm {[O/Fe]}$ and
1233: $\rm {[Al/Fe]}$. In each panel the red line is the best
1234: least squares fit and $\rm {a}$ is the slope of this
1235: line. The error bars represent the typical errors
1236: $\sigma_{\rm tot}$ from Tab.~\ref{Tab3}.}
1237: \label{Fig8}
1238: \end{figure}
1239:
1240: \begin{figure}[!hpbt]
1241: \centering
1242: \includegraphics*[width=8.cm]{Fig9.ps}
1243: \caption{$\rm {[Al/Fe]}$ vs.\ $\rm {[Ba/Fe]}$ abundance ratios. The
1244: error bars represent the typical errors $\sigma_{\rm tot}$ from
1245: Tab.~\ref{Tab3}.}
1246: \label{Fig9}
1247: \end{figure}
1248:
1249:
1250: \subsection{CN bimodality}
1251:
1252: In this section we discuss the correlation of the
1253: $\lambda$3883 CN absorption band strength with Na, O, and Al abundances.
1254: Previous studies revealed a trend of Na and Al with the CN band
1255: strength. By studying 4 RGB stars from high
1256: resolution spectra,
1257: D92 found higher Sodium abundances in the two CN-strong stars with
1258: respect to the two CN-weak ones (Suntzeff \& Smith 1991).
1259: I99 found that the Oxygen abundance is anticorrelated
1260: with Nitrogen, whereas both Na and Al are more abundant in CN-strong
1261: giants than in CN-weak ones.
1262:
1263: In our study, two peaks in the Na distribution are evident
1264: (see Fig.~\ref{Fig6}), while there is no equally strong
1265: pattern in the Al distribution (right bottom panel of
1266: Fig.~\ref{Fig8}).
1267: The CN band strengths of red giants in M4 have been measured by
1268: N81, Suntzeff \& Smith (1991), and I99.
1269: Smith \& Briley 2005 (SB05) have homogenized all the available data for
1270: CN band strengths in terms of the S(3839) index, e. g. the
1271: ratio of the flux intensities of the Cyanogen band near 3839 \AA \ and the
1272: nearby continuum.
1273: In the following, we will use the S(3839) index calculated by SB05.
1274:
1275: In Tab.~\ref{Tab5} the CN-index S(3839), the $\rm {[C/Fe]}$ and the
1276: $\rm {[N/Fe]}$ values (taken from SB05 and I99, respectively), and
1277: the identification as CN-strong (S), -weak (W), and -intermediate (I),
1278: are listed for our target stars with available CN data (34 stars in total).
1279: The IDs come from Lee (1977).
1280:
1281: For these targets, Table \ref{Tab6} lists the mean values of the
1282: chemical abundances for CN-S and CN-W(+I) stars. The Na, O, Mg, Al and
1283: Ca abundances are from this work, the $\rm {[C/Fe]}$ values are given
1284: by SB05, while the $\rm {[N/Fe]}$ are the mean values taken from I99 for all
1285: their CN-S and CN-W stars, including stars not in common with our
1286: sample (there are only five stars in our sample with nitrogen content
1287: available). In the last column the differences $\Delta(\rm {S-W})$
1288: between the mean values for CN-S and CN-W(+I) stars are listed.
1289:
1290: \begin{table*}[!htpq]
1291: \caption{CN band strengths and Carbon and Nitrogen
1292: abundances for our target stars with C, N and CN measurements
1293: in the literature. We use the
1294: star identifications from Lee (1977). In the last column, S
1295: refers to CN-strong, W CN-weak and I CN-intermediate stars. The C
1296: and N abundances are from SB05 and I99, respectively.}
1297: \centering
1298: \label{Tab5}
1299: \begin{tabular}{ c l c c c c c l c c c }
1300: \hline\hline
1301: ID & S(3839) & CN &${\rm [CI/Fe]}$&${\rm [NI/Fe]}$& &ID & S(3839)& CN&${\rm [CI/Fe]}$&${\rm [NI/Fe]}$\\
1302: \hline
1303: L1411& $0.54 $ & S &$... $&$... $ & &L2626 & $0.66 $ & S&$... $ &$... $\\
1304: L1403& $0.655$ & S &$-0.67$&$... $ & &L3705 & $0.20 $ & W&$... $ &$... $\\
1305: L1501& $0.61 $ & S &$... $&$... $ & &L3706 & $0.16 $ & W&$... $ &$... $\\
1306: L2422& $0.69 $ & S &$-0.83$&$... $ & &L2519 & $0.39 $ & I&$-0.50$ &$+0.22$\\
1307: L4404& $0.505$ & S &$-0.35$&$... $ & &L2623 & $0.245$ & W&$-0.70$ &$... $\\
1308: L1514& $0.305$ & W &$-1.12$&$+0.99$ & &L3721 & $0.58 $ & S&$... $ &$ ... $\\
1309: L4415& $0.53 $ & S &$-0.67$&$... $ & &L2617 & $0.63 $ & S&$-0.77$ &$+1.05$\\
1310: L4416& $0.62 $ & S &$-0.89$&$... $ & &L2711 & $0.65 $ & S&$... $ &$... $\\
1311: L4509& $0.62 $ & S &$-0.44$&$... $ & &L3730 & $0.24 $ & W&$... $ &$... $\\
1312: L1512& $0.61 $ & S &$... $&$... $ & &L3612 & $0.65 $ & S&$-0.73$ &$+1.02$\\
1313: L1617& $0.63 $ & S &$-0.93$&$... $ & &L3621 & $0.23 $ & W&$... $ &$... $\\
1314: L1619& $0.62 $ & S &$... $&$... $ & &L2608 & $0.57 $ & S&$-0.52$ &$... $\\
1315: L1608& $0.60 $ & S &$-0.63$&$... $ & &L3624 & $0.40 $ & I&$-0.67$ &$+0.25$\\
1316: L1605& $0.74 $ & S &$... $&$... $ & &L2410 & $0.23 $ & W&$-0.59$ &$... $\\
1317: L4630& $0.505$ & S &$-0.62$&$... $ & &L4413 & $0.655$ & S&$-0.32$ &$... $\\
1318: L3701& $0.39 $ & I &$... $&$... $ & &L4421 & $0.655$ & S&$-0.83$ &$... $\\
1319: L3404& $0.12 $ & W &$-0.40$&$... $ & &L4508 & $0.14$ & W&$ ... $ &$... $\\
1320: \hline
1321: \end{tabular}
1322: \end{table*}
1323:
1324: \begin{figure*}
1325: \centering
1326: \includegraphics[width=5cm]{Fig10a.ps}
1327: \includegraphics[width=5cm]{Fig10b.ps}
1328: \includegraphics[width=5cm]{Fig10c.ps}
1329: \includegraphics[width=5cm]{Fig10d.ps}
1330: \includegraphics[width=5cm]{Fig10e.ps}
1331: \caption{The abundance ratios ${\rm [Na/Fe]}$, $\rm {[O/Fe]}$, $\rm
1332: {[Al/Fe]}$, $\rm {[Mg/Fe]}$ and $\rm {[Ca/Fe]}$ are plotted as a
1333: function of the CN index S(3839). The filled circles represent the
1334: CN-S stars, the open circles the CN-W and the open squares the CN-I ones.
1335: In each panel, $\rm {a}$ is the slope of the straight line of the best
1336: least square fit. The error bars represent the typical errors
1337: $\sigma_{\rm tot}$ from Tab.~\ref{Tab3}.}
1338: \label{Fig10}
1339: \end{figure*}
1340:
1341: Fig.~\ref{Fig10} shows the ${\rm [Na/Fe]}$, $\rm {[Al/Fe]}$, $\rm
1342: {[O/Fe]}$, $\rm {[Mg/Fe]}$ and $\rm {[Ca/Fe]}$ values
1343: obtained in this work as a function of S(3839). In this
1344: figure CN-S objects are represented by filled circles, CN-W by open
1345: circles, and CN-I by open squares. The dotted lines represent the mean
1346: values of the abundances for the CN-S and CN-W(+I) stars; the CN-I stars
1347: were included in the CN-W group because of their similar behaviour as is
1348: clear from Fig.~\ref{Fig10}. The red
1349: lines represent the best least squares fit of the data, and $\rm {a}$
1350: is the slope of the line.
1351: We note that the CN-S objects clearly show significantly enhanced
1352: Na abundances, while the CN-W(+I) stars have a lower Na content.
1353: The three CN-I stars show low ${\rm [Na/Fe]}$ values typical of CN-weak stars.
1354: There is also a systematic difference in Al between CN-weak and
1355: CN-strong stars, with CN-strong stars richer in Al content.
1356:
1357: \begin{figure*}[!hpbt]
1358: \centering
1359: \includegraphics[width=11.0cm]{Fig11.ps}
1360: \caption{$\rm {U}$ vs.\ $\rm {(U-B)}$ CMD from WFI photometry. The distribution of
1361: the Na content for the UVES stars is shown in the inset. The stars
1362: belonging to the two different Na groups are represented in two
1363: different colors: the red triangles represent the stars with $\rm
1364: {[Na/Fe]}>0.2$ dex, and the blue circles the stars with $\rm
1365: {[Na/Fe]}\leq0.2$ dex.}
1366: \label{Fig11}
1367: \end{figure*}
1368:
1369: The CN-W(+I) stars have a higher O content than the CN-S as shown in the
1370: upper middle panel of Fig.~\ref{Fig10}.
1371: Fig.~\ref{Fig10} shows some difference in $\rm {[Mg/Fe]}$
1372: between the CN-W(+I) and the CN-S stars, with CN-W(+I) stars having
1373: slightly higher Mg abundances than the CN-S ones, but this difference is not
1374: statistically significant (see Table~\ref{Tab6}).
1375:
1376: I99 and D92 also found that the scatter in Ca abundance
1377: correlates with the CN strength. I99 found a difference in Ca content
1378: between CN-strong and CN-weak stars by $0.08\pm0.11$ dex.
1379: To compare, we have also analysed the Ca abundance as a function of the CN
1380: strength.
1381: In the bottom-right panel of Fig.~\ref{Fig10}, $\rm {[Ca/Fe]}$ is plotted as a
1382: function of S(3839); we can see that the difference between the two
1383: groups is 0.02 dex, smaller than what was found by I99
1384: and not statistically significant (see Table~\ref{Tab6}).
1385:
1386: \begin{table*}[!htpq]
1387: \caption{Mean abundance ratios and relative rms ($\sigma$) for the
1388: CN-S and CN-W(+I) groups. In the last column the difference between
1389: CN-S and CN-W(+I) stars are listed.}
1390: \centering
1391: \label{Tab6}
1392: \begin{tabular}{lccccccc}
1393: \hline\hline
1394: [el/Fe] & CN-S &$\sigma$& &CN-W(+I) &$\sigma$&&$\Delta({\rm S-W(+I)})$\\
1395: \hline
1396: ${\rm [CI/Fe]}$ &$-0.66\pm 0.05$ &0.19& &$-0.66 \pm 0.11$ &0.25&& $+0.00$ \\
1397: ${\rm [NI/Fe]}$ &$+1.08\pm 0.10$ &0.22& &$+0.42 \pm 0.19$ &0.42&& $+0.66$ \\
1398: ${\rm [OI/Fe]}$ &$+0.33\pm 0.01$ &0.07& &$+0.45 \pm 0.01$ &0.04&& $-0.12$ \\
1399: ${\rm [NaI/Fe]}$ &$+0.41\pm 0.02$ &0.07& &$+0.11 \pm 0.02$ &0.06&& $+0.30$ \\
1400: ${\rm [AlI/Fe]}$ &$+0.59\pm 0.01$ &0.06& &$+0.51 \pm 0.03$ &0.09&& $+0.08$ \\
1401: ${\rm [MgI/Fe]}$ &$+0.52\pm 0.01$ &0.04& &$+0.54 \pm 0.02$ &0.06&& $-0.02$ \\
1402: ${\rm [CaI/Fe]}$ &$+0.26\pm 0.01$ &0.04& &$+0.24 \pm 0.01$ &0.03&& $+0.02$ \\
1403: \hline
1404: \end{tabular}
1405: \end{table*}
1406:
1407: \section{Search for evolutionary effects}
1408:
1409: \subsection{Abundances along the RGB}
1410:
1411: According to the results discussed in the previous section, we can
1412: define two groups of stars: the Na-rich stars, i. e. those with $\rm
1413: {[Na/Fe]}\geq0.2$ and low Oxygen content which must be associated
1414: to the CN-S stars (Fig.~\ref{Fig10}), and the Na-poor stars with
1415: higher Oxygen, i. e. those with $\rm {[Na/Fe]<0.2}$, which correspond
1416: to the CN-W group. It is instructive to look at the position of the
1417: stars belonging to these two Na groups on the $\rm {U}$ vs.\ $\rm
1418: {(U-B)}$ CMD and on the color-color $\rm {(U-B)}$ vs.\ $\rm {(B-K)}$
1419: diagram (Fig.~\ref{Fig11} and Fig.~\ref{Fig12}, respectively). The
1420: $\rm {U}$ and $\rm {B}$ magnitudes come from our WFI photometry, the
1421: $\rm {K}$ magnitude from the 2MASS catalogue.
1422: The Na-rich stars are represented by red triangles, while the Na-poor
1423: stars by blue circles. The CMD has been corrected for differential reddening
1424: (following the same procedure used in Sarajedini et al.\ 2007) and
1425: proper motions.
1426: It is clear from Fig.~\ref{Fig11} that our sample of stars mostly
1427: includes RGB stars with only three or four probable AGB stars.
1428:
1429: Interestingly enough, the two groups of Na-rich and Na-poor stars form
1430: two distinct branches on the RGB. Na-rich stars define a narrow
1431: sequence on the red side of the RGB, while the Na-poor sample
1432: populates the blue, more spread out portion of the RGB. Even more
1433: interestingly, the anomalous broadening of the RGB is visible down to
1434: the base of the RGB, at $\rm {U} \sim 17.5$, indicating that the two
1435: abundance groups are present all over the RGB, even well below the RGB
1436: bump, where no deep mixing is expected. This evidence further
1437: strengthens the idea that the bimodal Na, O, CN distribution must have
1438: been present in the material from which the stars we presently observe
1439: in M4 originated.
1440:
1441: Figure~\ref{Fig11} unequivocally shows that the Na (CN) dichotomy
1442: is associated to a dichotomy in the color of the RGB stars.
1443: In order to quantify this split, we calculated for the Na-poor group of stars
1444: the mean difference in $\rm {(U-B)}$ color with respect to a fiducial
1445: line representative of the RGB Na-rich stars (see Fig.~\ref{Fig13}).
1446: The fiducial line has been obtained as follows:
1447:
1448: \begin{itemize}
1449:
1450: \item{the CMD has been divided into bins of 0.30 mag in $\rm {U}$, and a median
1451: color $\rm {(U-B)}$ and $\rm {U}$ magnitude have been computed for the
1452: Na-rich ($\rm {[Na/Fe]}\geq0.2$) stars in each bin. The median points have been fitted
1453: with a spline function which represent a first, raw fiducial line;}
1454:
1455: \item{for each Na-rich star, the difference in color with respect to the
1456: fiducial was calculated, and the 68.27th percentile of the absolute
1457: values of the color differences, was taken as an estimate of the
1458: color dispersion ($\sigma$). All stars with a color distance from the fiducial
1459: larger than $3\sigma$ were rejected;}
1460:
1461: \item{the median colors and magnitudes and the $\sigma$ of each bin
1462: were redetermined by using the remaining stars.}
1463:
1464: \end{itemize}
1465:
1466: The leftmost panel of Fig.~\ref{Fig13} shows the original CMD, while
1467: the rightmost one shows the CMD after subtracting from each Na-poor star
1468: the fiducial line color appropriate for its ${\rm U}$ magnitude. The
1469: differences are indicated as $\rm {\Delta(U-B)}$.
1470: We calculated the average $\rm {\Delta(U-B)}$ of Na-poor
1471: stars with $\rm {U}$$\geq$14.8. This cut in magnitude has been imposed
1472: by the poor statistics and the
1473: presence of the probable AGB stars at brighter magnitudes. The
1474: mean $\rm {\Delta(U-B)}$ value is $-0.17\pm0.02$ (dotted line in the
1475: left most panel of Fig.~\ref{Fig13}).
1476:
1477: The significant difference between the mean colors of the 2 groups of
1478: stars is a further evidence of the presence of two different
1479: stellar populations in the RGB of M4, to be associated
1480: with the different content of Na, and, because of the discussed
1481: correlations to different O, N, and C content.
1482:
1483: In order to understand the origin of the photometric dichotomy, using
1484: SPECTRUM, we simulated two synthetic spectra, one representative of
1485: the Na-rich, CN-strong stars and one for the Na-poor, CN-weak ones.
1486: The two synthetic spectra were computed using as atmospheric parameters the mean
1487: values measured for the sample of our stars for which there are literature
1488: data on the CN band strengths (see Tab.~\ref{Tab5}), and assuming for the C, N,
1489: O, Na abundances the average values calculated for the CN-W(+I) and
1490: the CN-S groups and listed in Tab.~\ref{Tab6}. We then multiplied the two
1491: spectra by the efficency curve of our $\rm {U}$ and $\rm {B}$ photometric bands
1492: (Fig.~\ref{Fig14}, lower panels), and, finally, we calculated the
1493: difference between the resulting fluxes.
1494: The upper panels of Fig.~\ref{Fig14} show the differences between the
1495: two simulated spectra as a function of the wavelength for the $\rm
1496: {U}$ (left panel) and $\rm {B}$ (right panel) panel.
1497: It is clear that the strength of the CN and NH bands strongly influences the
1498: $\rm {U} - {B}$ color. The NH
1499: band around 3360 \AA, and the CN bands around 3590 \AA, 3883 \AA\ and
1500: 4215 \AA \ are the main contributors to the effect.
1501:
1502: \begin{figure}[!hpbt]
1503: \centering
1504: \includegraphics[width=8.5cm]{Fig12.ps}
1505: \caption{$\rm {(U-B)}$ vs.\ $\rm {(B-K)}$ diagram: $\rm {U}$ and $\rm {B}$ come
1506: from WFI photometry, $\rm {K}$ from 2MASS catalogue. The UVES stars
1507: belonging to the two different Na groups are represented as in
1508: Fig.~\ref{Fig11}. }
1509: \label{Fig12}
1510: \end{figure}
1511:
1512: \begin{figure}[hpbt]
1513: \centering
1514: \includegraphics[width=8.5cm]{Fig13.ps}
1515: \caption{(Left panel) Distribution along the RGB of the stars with measured
1516: abundances. The Na-rich (red) and the Na-poor (blue) stars are represented as in
1517: Fig.~\ref{Fig11} and Fig.~\ref{Fig12}. In the middle panel the $\rm
1518: N_{red}/N_{blue}$ ratios for the 4 selected magnitude bins are shown
1519: with their relative Poisson errors. The rightmost panel shows the color
1520: difference between each analyzed star and a reference fiducial line (dotted
1521: line).}
1522: \label{Fig13}
1523: \end{figure}
1524:
1525: The differences in magnitude between the CN-S and the CN-W(+I)
1526: simulated spectra in the two bands are: $\rm
1527: {\Delta(U)_{Strong-Weak}=+0.06}$ and $\rm
1528: {{\Delta(B)_{Strong-Weak}=+0.02}}$. Consequently, the expected color
1529: difference between the two groups of stars is $\rm
1530: {{\Delta(U-B)_{Strong-Weak}=+0.04}}$.
1531: This value goes in the same direction, but it is smaller
1532: than the observed one ($\rm {\Delta(U-B)_{Strong-Weak}=+0.17\pm0.02}$).
1533: However, we note that our procedure uses simulated
1534: spectra with average abundances, and therefore should
1535: be considered a rough simulation.
1536: We cannot exclude the possibility that other effects (perhaps related to the
1537: structure and evolution of the stars with different chemical content, or
1538: effects on the stellar atmosphere associated with the complex distribution of
1539: the chemical abundances in addition to the CN band), might further contribute
1540: to the photometric dichotomy in the RGB.
1541: Surely, our simulations show that the CN-bimodality affects the
1542: $\rm {(U - B)}$ color and can be at least partly responsible for the
1543: observed spread in the $\rm U$ vs.\ $\rm {(U - B)}$ CMD.
1544:
1545: Finally, we investigated whether the chemical and photometric dichotomy
1546: is related to the evolutionary status along the RGB.
1547: To this end, we have calculated the fractions of stars in the 2
1548: Na-groups at different magnitudes along the RGB.
1549: As shown in the middle panel of Fig.~\ref{Fig13},
1550: we divided the RGB in 4 magnitude bins
1551: containing the same number of stars with measured metal content,
1552: and calculated the ratio between the number of Na-rich ($\rm
1553: {N_{red}}$) and the number of Na-poor ($\rm {N_{blue}}$) stars in each
1554: bin. The $\rm {N_{red}/N_{blue}}$ ratios with their associated Poisson errors
1555: as a function of the magnitude are plotted the middle panel of Fig.~\ref{Fig13}.
1556: The $\rm {N_{red}/N_{blue}}$ values in the different magnitude
1557: bins are the same within the errors.
1558: A similar result is obtained if we divide the RGB in two bins only,
1559: but containing a number of
1560: stars twice as large as in the previous experiment.
1561: In this case, we have $\rm {N_{red}/N_{blue}} = 1.52\pm0.44$ for $\rm
1562: {14.0<U<14.9}$, $\rm {N_{red}/N_{blue}} = 1.82\pm0.54$ for $\rm
1563: {14.9<U<15.8}$.
1564:
1565: Figure~\ref{Fig15} shows the dependence of Na, Al, and S(3839) index as
1566: a function of the magnitude $\rm {U}$ (upper panel) and $\rm {B}$
1567: (lower panel). Again, we do not see any trend of the
1568: abundances with the position of the stars along the RGB.
1569:
1570: \begin{figure*}[!hpbt]
1571: \centering
1572: \includegraphics[bb=0 220 550 600,width=8.6cm]{Fig14a.ps}
1573: \includegraphics[bb=0 220 550 600,width=8.6cm]{Fig14b.ps}
1574: \caption{The differences (expressed in $\rm
1575: {10^{4}\ erg/cm^2/sec/Angstrom}$) between the CN-S and the CN-W(+I)
1576: simulated spectra in the $\rm {U}$ and $\rm {B}$ band are represented in the
1577: upper panels. The normalized response of the two filters is also
1578: shown in the lower panels.}
1579: \label{Fig14}
1580: \end{figure*}
1581:
1582: \begin{figure*}[!hpbt]
1583: \centering
1584: \includegraphics[bb=0 250 550 620,width=13.5cm]{Fig15.ps}
1585: \caption{Dependence of Na, Al, and S(3839) index as
1586: a function of the magnitude $\rm {U}$ (upper panels) and $\rm {B}$
1587: (lower panels).}
1588: \label{Fig15}
1589: \end{figure*}
1590:
1591:
1592: In conclusion, there is no evidence for a dependence of the Na content
1593: and (because of the discussed correlations) of the O, Al abundances
1594: or of the CN strength with the evolutionary status of the stars.
1595:
1596: \subsection{The RGB progeny on the HB}
1597:
1598: In the previous sections, we have identified two groups of stars
1599: with distinct Na, O, CN content, and which populate two
1600: distinct regions in the $\rm {U}$ vs.\ $\rm {U-B}$ CMD.
1601:
1602: We note that Yong \& Grundhal (2008), by studying 8 bright red giants
1603: in the GC NGC~1851 found a large star to star abundance variation in
1604: the Na, O, and Al content, and suggested that these abundance
1605: anomalies could be associated with the presence of the two stellar
1606: populations identified by Milone et al. (2008), a situation apparently
1607: very similar to what we have found in M4. Cassisi et al. (2008)
1608: suggest that extreme CNO variations could be the basis of the SGB
1609: split found in NGC~1851. As discussed in Milone et al. (2008) there is
1610: some evidence of a split of the RGB in NGC~1851 similar to what we
1611: have found in this paper for M4.
1612: We have investigated all available HST data of M4, including
1613: CMDs from very high precision photometry based on ACS/HST images; we could
1614: find no evidence for a SGB split (as in NGC~1851)
1615: or evidence for a MS split (as in NGC~2808 or $\omega$~Cen).
1616:
1617: It is also useful to investigate where the progeny of the two RGB
1618: populations is along the HB.
1619: Milone et al.\ (2008) suggest that the bimodal HB of NGC~1851 can be
1620: interpreted in terms of the presence of two distinct stellar
1621: populations in this cluster,
1622: and tentatively link the bright SGB to the red HB (RHB) and the fainter
1623: SGB to the blue HB (BHB). M4 also has a bimodal
1624: HB, well populated on both the red and blue side of the RR Lyrae gap.
1625: Could the HB morphology of M4 be related to the CN bimodality? N81,
1626: studying the CN band strengths on a sample of 45 giant stars in M4,
1627: suggested the possibility of a relation between the HB morphology and the
1628: CN-bimodality.
1629:
1630: To test the possible relation between the chemical and consequent RGB
1631: bimodality of M4 and the morphology of its HB, we used the WFI@2.2m
1632: instrumental photometry by Anderson et al.\ 2006 (Fig.~\ref{Fig16})
1633: corrected for differential reddening, and selected the cluster members using the
1634: measured proper motions.
1635: We remark here the fact that this photometry is not the one we use in the
1636: whole paper. The homogenization between the two photometries is
1637: a difficult and long process (mainly because the fact that the
1638: $\rm {B}$ filters for the two datasets are different) beyond the scope of this work.
1639:
1640: In our test, we used the $\rm {V}$ vs.\ $\rm {(B-V)}$ CMD, as in this
1641: photometric plane the two components of the HB are more clearly
1642: distinguishable.
1643:
1644: In Fig.~\ref{Fig16}, different symbols show the red and blue HB stars
1645: that we selected. The
1646: ratio between the number ({$\rm N_{BHB}$}) of the HB stars bluer than
1647: the RR-Lyrae instability strip (BHB), and the total HB stars ({$\rm
1648: N_{TOT}$) is:
1649:
1650: \begin{center}
1651: ${\rm N_{BHB}/N_{TOT} = 0.56 \pm 0.10}$
1652: \end{center}
1653:
1654: \noindent
1655: while the ratio between the Na rich stars (${\rm N_{NaR}}$) and the
1656: total number of stars in our spectroscopic sample (${\rm N_{S}}$) is:
1657:
1658: \begin{center}
1659: ${\rm N_{NaR}/N_{S} = 0.64 \pm 0.10}$
1660: \end{center}
1661:
1662: where the associated uncertainties are the Poisson errors.
1663: Within the statistical uncertainties, accounting for the different
1664: evolutionary times along the HB,
1665: we can tentatively associate the Na-rich stars with the BHB, and the
1666: Na-poor stars with the RHB.
1667: A direct measurement of the metal content of the HB stars is needed in
1668: order to support this suggestion..
1669:
1670: \begin{figure}[hpbt]
1671: \centering
1672: \includegraphics[width=9cm]{Fig16.ps}
1673: \caption{In the top panels, the proper motions diagrams show the
1674: separation of the probable cluster stars (left panel) from the field
1675: (right panel). In the bottom panels, the cluster and field CMD ($\rm
1676: {V}$ vs.\ $\rm {B-V}$) are shown for the stars within the red circle in
1677: their respective upper panels. In the M4 CMD, the stars on the red and
1678: on the blue side of the RR Lyrae instability strip are represented
1679: with red and blue symbols respectively. For this figure, the photometry, which
1680: is not photometrically calibrated, is taken from Anderson et al.\ (2006).}
1681: \label{Fig16}
1682: \end{figure}
1683:
1684: \begin{table*}[!htpq]
1685: \caption{Mean abundance ratios and relative rms ($\sigma$) for the
1686: Na-rich and Na-poor groups. In the last column, the difference
1687: between Na-rich and Na-poor stars are listed.}
1688: \centering
1689: \label{Tab7}
1690: \begin{tabular}{lccccccc}
1691: \hline\hline
1692: [el/Fe] & Na-rich &$\sigma$& &Na-poor &$\sigma$&&$\Delta({\rm Na_{rich-poor}})$\\
1693: \hline
1694: ${\rm [OI/Fe]}$ &$+0.34\pm 0.01$ &0.08 & &$+0.47\pm 0.01$ &0.04&&$-0.13\pm0.01$ \\
1695: ${\rm [NaI/Fe]}$ &$+0.38\pm 0.01$ &0.08 & &$+0.07\pm 0.01$ &0.06&&$+0.31\pm0.01$ \\
1696: ${\rm [MgI/Fe]}$ &$+0.48\pm 0.01$ &0.06 & &$+0.52\pm 0.01$ &0.05&&$-0.04\pm0.01$ \\
1697: ${\rm [AlI/Fe]}$ &$+0.57\pm 0.01$ &0.08 & &$+0.49\pm 0.02$ &0.13&&$+0.08\pm0.02$ \\
1698: ${\rm [SiI/Fe]}$ &$+0.49\pm 0.01$ &0.04 & &$+0.46\pm 0.01$ &0.06&&$+0.03\pm0.01$ \\
1699: ${\rm [CaI/Fe]}$ &$+0.28\pm 0.01$ &0.04 & &$+0.28\pm 0.01$ &0.04&&$+0.00\pm0.01$ \\
1700: ${\rm [TiI/Fe]}$ &$+0.29\pm 0.01$ &0.03 & &$+0.29\pm 0.01$ &0.05&&$+0.00\pm0.01$ \\
1701: ${\rm [TiII/Fe]}$ &$+0.35\pm 0.01$ &0.07 & &$+0.34\pm 0.01$ &0.05&&$+0.01\pm0.01$ \\
1702: ${\rm [CrI/Fe]}$ &$-0.04\pm 0.01$ &0.05 & &$-0.03\pm 0.01$ &0.05&&$-0.01\pm0.01$ \\
1703: ${\rm [FeI/H]}$ &$-1.07\pm 0.01$ &0.05 & &$-1.07\pm 0.01$ &0.05&&$+0.00\pm0.01$ \\
1704: ${\rm [NiI/Fe]}$ &$+0.02\pm 0.01$ &0.03 & &$+0.01\pm 0.01$ &0.03&&$+0.01\pm0.01$ \\
1705: ${\rm [BaII/Fe]}$ &$+0.42\pm 0.01$ &0.07 & &$+0.40\pm 0.02$ &0.13&&$+0.02\pm0.02$ \\
1706: \hline
1707: \end{tabular}
1708: \end{table*}
1709:
1710: \section{Conclusions}
1711:
1712: We have presented high resolution spectroscopic analysis of 105 RGB
1713: stars in the GC M4 from UVES data.
1714:
1715: We have found that M4 has an iron content $\rm {[Fe/H]=-1.07\pm0.01}$
1716: (the associated error here is the internal error only),
1717: and an $\alpha$ element overabundance $\rm
1718: {[\alpha/Fe]=+0.39\pm0.05}$.
1719: Si and Mg are more overabundant than the other $\alpha$
1720: elements, suggesting a primordial overabundance for these elements.
1721: Moreover, we also find a slight overabundance of Al.
1722: The ${\rm [Na/Fe]}$ versus [O/Fe] ratios follow the well-known Na-O
1723: anticorrelation, signature of proton capture reactions at high
1724: temperature. No Mg-Al anticorrelation was found.
1725:
1726: We find a strong dichotomy in Na abundance, and show that it must be
1727: associated to a CN bimodality. Tab.~\ref{Tab7} synthesizes the
1728: spectroscopic and photometric differences we found for the two groups
1729: of stars. Comparing our results to those in the literature on
1730: the CN band strength, the CN-strong stars appear to have higher
1731: content in Na and Al, and lower O than the CN-weak ones. Apparently,
1732: M4 hosts two different stellar populations. This fact is evident from
1733: a chemical abundance distribution, and is also confirmed by
1734: photometry. In fact, an inspection of the $\rm {U}$ vs.\ $\rm {(U-B)}$
1735: CMD reveals a broadened RGB, and, as shown in our investigation, the
1736: two Na groups of stars occupy different positions (have different
1737: colors) along the RGB. Our sample of stars is composed of objects with
1738: $\rm {U}$ brighter than $\sim$16, but, as shown in Fig.~\ref{Fig11},
1739: the RGB appears to be similarly spread from the level of the SGB to the RGB
1740: tip. Since our photometry has been corrected for differential
1741: reddening, we conclude that the observed RGB spread at fainter magnitudes must
1742: also be due to the metal content dichotomy.
1743:
1744: We did not find any evident dependence of the chemical abundance
1745: distribution on the evolutionary status (along the RGB) of the target
1746: stars, from below the RGB bump to the RGB tip. This is an additional
1747: indication that the abundance spread must be primordial.
1748:
1749: The two groups of stars we have identified both spectroscopically
1750: and photometrically seem to be due to the presence of two distinct
1751: populations of stars.
1752: The abundance anomalies are very likely due to primordial variations in the
1753: chemical content of the material from which M4 stars formed, and not to
1754: different evolutionary paths of the present stellar population of M4.
1755: This is somehow surprising, because of the relatively small mass of M4 -- it
1756: is an order of magnitude smaller than the mass of $\omega$~Cen,
1757: NGC~2808, NGC~1851, NGC~6388, the other clusters in which a multiple
1758: stellar generation has been confirmed.
1759: Where did the gas which polluted the material for the CN-Na rich
1760: stars come from? Has it been ejected from a first generation of
1761: stars? How could it stay within the shallow gravitational potential of M4?
1762:
1763: The idea that M4 hosts two generations of stars makes the multipopulation
1764: phenomenon in GCs even more puzzling than originally thought. It
1765: becomes harder and harder to accept the idea that the phenomenon can
1766: be totally internal to the cluster, unless this object is what remains
1767: of a much larger system (a larger GC or the nucleus of a dwarf galaxy?).
1768: Surely, Because its orbit involves frequent passages at high inclination
1769: through the Galactic disk, always at distances from the Galactic center
1770: smaller than 5 kpc (see Dinescu et al.\ 1999), M4 must have been
1771: strongly affected by tidal shocks, and therefore it might have been
1772: much more massive in the past.
1773:
1774: %\begin{acknowledgements}
1775: %
1776: %\end{acknowledgements}
1777:
1778: \begin{thebibliography}{}
1779: \bibitem[Alonso et al. 1999]{Al99} Alonso, A., Arribas, S., \&
1780: Martinez-Roger, C. 1999, A\&A, 140, 261
1781: \bibitem[Anderson et al. 2006]{An06} Anderson, J., Bedin, L.~R., Piotto, G., Yadav, R.~S., \& Bellini, A.\ 2006, \aap, 454, 1029
1782: \bibitem[Ballester et al. 2000]{Ba00} Ballester, P., Modigliani, A.,
1783: Boitquin, O., et al. 2000, ESO Messenger, 101, 31
1784: \bibitem[Bassino \& Muzzio 1995]{BM95} Bassino, L.~P., \&
1785: Muzzio, J.~C.\ 1995, The Observatory, 115, 256
1786: \bibitem[Bedin et al.(2000)]{2000A&A...363..159B} Bedin, L.~R.,
1787: Piotto, G., Zoccali, M., Stetson, P.~B., Saviane, I., Cassisi, S.,
1788: \& Bono, G.\ 2000, \aap, 363, 159
1789: \bibitem[Bedin et al. 2004]{Be04} Bedin, L.\ R., Piotto, G., Anderson, J.,
1790: Cassisi, S., King, I.\ R., Momany, Y., \& Carraro, G. 2004, ApJ,
1791: 605, 125
1792: \bibitem[Brown \& Wallerstein 1992]{BW92} Brown, J.~A., \&
1793: Wallerstein, G.\ 1992, \aj, 104, 1818
1794: \bibitem[Busso et al. 2007]{Bu07} Busso, G., et al.\ 2007,
1795: \aap, 474, 105
1796: \bibitem[Caloi \& D'Antona 2007]{CD07} Caloi, V., \& D'Antona,
1797: F.\ 2007, \aap, 463, 949
1798: \bibitem[Cannon et al. 1998]{Ca98} Cannon, R.~D., Croke,
1799: B.~F.~W., Bell, R.~A., Hesser, J.~E., \& Stathakis, R.~A.\ 1998, \mnras, 298, 601
1800: \bibitem[Carretta et al. 2004]{Ca04} Carretta, E., Gratton,
1801: R.~G., Bragaglia, A., Bonifacio, P., \& Pasquini, L.\ 2004, \aap,
1802: \bibitem[Carretta et al. 2005]{Ca05} Carretta, E., Gratton,
1803: R.~G., Lucatello, S., Bragaglia, A., \& Bonifacio, P.\ 2005, \aap, 433, 597
1804: \bibitem[Carretta et al. 2006]{Ca06} Carretta, E., Bragaglia, A.,
1805: Gratton, R.~G., Leone, F., Recio-Blanco, A., \& Lucatello, S.\ 2006,
1806: \aap, 450, 523
1807: \bibitem[Cassisi et al. 2008]{Ca08} Cassisi, S., Salaris,
1808: M., Pietrinferni, A., Piotto, G., Milone, A.~P., Bedin, L.~R.,
1809: \& Anderson, J.\ 2008, \apjl, 672, L115
1810: \bibitem[Cohen 1978]{Co78} Cohen, J.~G.\ 1978, \apj, 223,
1811: 487
1812: \bibitem[Cohen \& Mel{\'e}ndez 2005]{CM05} Cohen, J.~G., \&
1813: Mel{\'e}ndez, J.\ 2005, \aj, 129, 303
1814: \bibitem[Cudworth \& Rees 1990]{CR90} Cudworth, K.~M., \& Rees, R.\
1815: 1990, \aj, 99, 1491
1816: \bibitem[Da Costa \& Armandroff 1995]{DA95} Da Costa, G.~S., \& Armandroff, T.~E.\ 1995, \aj, 109, 2533
1817: \bibitem[D'Antona et al. 2005]{Da05} D'Antona, F., Bellazzini, M.,
1818: Caloi, V., Pecci, F.~F., Galleti, S., \& Rood, R.~T.\ 2005, \apj,
1819: 631, 868
1820: \bibitem[Dinescu et al. 1999]{Di99} Dinescu, D.~I., Girard, T.~M., \& van
1821: Altena, W.F. 1999, AJ, 117, 1792
1822: \bibitem[Drake et al. 1992]{Dr92} Drake, J.~J., Smith, V.~V., \&
1823: Suntzeff, N.~B.\ 1992, \apjl, 395, L95
1824: \bibitem[Drake et al. 1994]{Dr94} Drake, J.~J., Smith, V.~V., \&
1825: Suntzeff, N.~B.\ 1994, \apj, 430, 610
1826: \bibitem[Freeman \& Rodgers 1975]{FR75} Freeman, K.~C., \&
1827: Rodgers, A.~W.\ 1975, \apjl, 201, L71
1828: \bibitem[Freeman 1993]{Fr93} Freeman, K.~C.\ 1993, The
1829: Globular Cluster-Galaxy Connection, 48, 608
1830: \bibitem[Gratton, Quarta, \& Ortolani 1986]{Gr86} Gratton, R.~G., Quarta, M.~L.,
1831: \& Ortolani, S.\ 1986, \aap, 169, 208
1832: \bibitem[Gratton et al.(1996)]{1996A&A...314..191G} Gratton, R.~G.,
1833: Carretta, E., \& Castelli, F.\ 1996, \aap, 314, 191
1834: \bibitem[Gratton et al. 1999]{Gr99} Gratton, R.~G.,
1835: Carretta, E., Eriksson, K., \& Gustafsson, B.\ 1999, \aap, 350, 955
1836: \bibitem[Gratton et al. 2001]{Gr01} Gratton, R.~G., et al.\
1837: 2001, \aap, 369, 87
1838: \bibitem[Gratton et al. 2004]{Gr04} Gratton, R., Sneden,
1839: C., \& Carretta, E.\ 2004, \araa, 42, 385
1840: \bibitem[Grundahl et al. 2002]{Gr02} Grundahl, F., Briley, M.,
1841: Nissen, P.~E., \& Feltzing, S.\ 2002, \aap, 385, L14
1842: \bibitem[Harris 1996]{Ha96} Harris, W.~E.\ 1996, VizieR Online Data
1843: Catalog, 7195, 0
1844: \bibitem[Hughes \& Wallerstein 2000]{HW00} Hughes, J., \&
1845: Wallerstein, G.\ 2000, \aj, 119, 1225
1846: \bibitem[Ivans et al. 1999]{Iv99} Ivans, I.~I., Sneden, C., Kraft,
1847: R.~P., Suntzeff, N.~B., Smith, V.~V., Langer, G.~E., \& Fulbright,
1848: J.~P.\ 1999, \aj, 118, 1273
1849: \bibitem[Ivans et al. 2000]{Iv00} Ivans, I.~I., Kraft, R.~P., Sneden,
1850: C., \& Smith, G.~H.\ 2000, Bulletin of the American Astronomical
1851: Society, 32, 738
1852: \bibitem[Johnson et al. 2005]{Jo05} Johnson, C.~I., Kraft, R.~P.,
1853: Pilachowski, C.~A., Sneden, C., Ivans, I.~I., \& Benman, G.\ 2005,
1854: \pasp, 117, 1308
1855: \bibitem[Kemp et al. 1993]{Ke93} Kemp, S.~N., Bates, B., \& Lyons,
1856: M.~A.\ 1993, \aap, 278, 542
1857: \bibitem[Kurucz et al. 1984]{Ku84} Kurucz, R. L., Furenlid, I., Brault, J., \&
1858: Testerman, L. Solar Flux Atlas from 296 to 1300 nm, National Solar Observatory Atlas No. 1, June 1984
1859: \bibitem[Kurucz 1992]{Ku92} Kurucz, R.\ L. 1992, in IAU Symp. 149, The
1860: Stellar Populations of Galaxies, ed. B. Barbuy \& A. Renzini
1861: (Dordrecht: Reidel), 225
1862: \bibitem[Kupka et al. 1999]{Ku99} Kupka, F., Piskunov, N.,
1863: Ryabchikova, T.~A., Stempels, H.~C., \& Weiss, W.~W.\ 1999, \aaps, 138, 119
1864: \bibitem[Layden \& Sarajedini 2000]{LS00} Layden, A.~C., \&
1865: Sarajedini, A.\ 2000, \aj, 119, 1760
1866: \bibitem[Lee 1977]{Le77} Lee, S.-W.\ 1977, \aaps, 27, 367
1867: \bibitem[Lee et al. 1999]{Le99} Lee, Y.-W., Joo, J.-M.,
1868: Sohn, Y.-J., Rey, S.-C., Lee, H.-C., \& Walker, A.~R.\ 1999, \nat, 402, 55
1869: \bibitem[Lyons et al. 1995]{Ly95} Lyons, M.~A., Bates, B., Kemp,
1870: S.~N., \& Davies, R.~D.\ 1995, \mnras, 277, 113
1871: \bibitem[Mandushev et al. 1991]{Ma91} Mandushev, G., Staneva, A., \&
1872: Spasova, N.\ 1991, \aap, 252, 94
1873: \bibitem[Milone et al. 2008]{Mi08} Milone, A.~P., et al.\
1874: 2008, \apj, 673, 241
1875: \bibitem[Momany et al.(2003)]{2003A&A...407..303M} Momany, Y.,
1876: Cassisi, S., Piotto, G., Bedin, L.~R., Ortolani, S., Castelli, F.,
1877: \& Recio-Blanco, A.\ 2003, A\&A, 407, 303
1878: \bibitem[Norris \& Freeman 1979]{NF79} Norris, J., \& Freeman,
1879: K.~C.\ 1979, \apjl, 230, L179
1880: \bibitem[Norris 1981]{No81} Norris, J.\ 1981, \apj, 248, 177
1881: \bibitem[Norris et al. 1996]{No96} Norris, J.~E., Freeman,
1882: K.~C., \& Mighell, K.~J.\ 1996, \apj, 462, 241
1883: \bibitem[Pancino et al. 2000]{Pa00} Pancino, E., Ferraro,
1884: F.~R., Bellazzini, M., Piotto, G., \& Zoccali, M.\ 2000, \apjl, 534, L83
1885: \bibitem[Pasquini et al. 2002]{Pa02} Pasquini, L., et al.\ 2002, The
1886: Messenger, 110, 1
1887: \bibitem[Peterson et al. 1995]{Pe95} Peterson, R.~C., Rees,
1888: R.~F., \& Cudworth, K.~M.\ 1995, \apj, 443, 124
1889: \bibitem[Piotto et al. 2005]{Pi05} Piotto, G., Villanova, S., Bedin, L.\ R.,
1890: Gratton, R., Cassisi, S., Momany, Y., Recio-Blanco, A., Lucatello,
1891: S., Anderson, J., King, I.\ R., Pietrinferni, A., \& Carraro,
1892: G. 2005, ApJ, 621, 777 (P05)
1893: \bibitem[Piotto et al. 2007]{Pi07} Piotto, G., et al.\ 2007, \apjl, 661, L53
1894: \bibitem[Piotto 2008]{Pi08a} Piotto, G.\ 2008, in XXI Century Challenges for
1895: Stellar Evolution, Memorie della Societa Astronomica Italiana,
1896: vol. 79/2, eds: S. Cassisi, M. Salaris (arXiv:0801.3175)
1897: \bibitem[Piotto et al. 2008]{Pi08b} Piotto, G., et al.\ 2008, in preparation
1898: \bibitem[Preston et al.(2006)]{2006AJ....132...85P} Preston, G.~W., Sneden,
1899: C., Thompson, I.~B., Shectman, S.~A., \& Burley, G.~S.\ 2006, \aj, 132, 85
1900: \bibitem[Ram{\'{\i}}rez \& Cohen 2002]{RC02} Ram{\'{\i}}rez,
1901: S.~V., \& Cohen, J.~G.\ 2002, \aj, 123, 3277
1902: \bibitem[Rich et al. 1997]{Ri97} Rich, R.~M., Sosin, C., Djorgovski, S.~G.,
1903: Piotto, G., King, I.~R., Renzini, A., Phinney, E.~S., Dorman, B., Liebert,
1904: J., Meylan, G. 1997, ApJ, 484, 25
1905: \bibitem[Sarajedini et al. 2007]{Sa07} Sarajedini, A., et
1906: al.\ 2007, \aj, 133, 1658
1907: \bibitem[Shetrone 2003]{Sh03} Shetrone, M.~D.\ 2003, \apjl,
1908: 585, L45
1909: \bibitem[Siegel et al.(2007)]{2007ApJ...667L..57S} Siegel, M.~H., et al.\
1910: 2007, \apjl, 667, L57
1911: \bibitem[Skrutskie et al. 2006]{Sk06} Skrutskie, M.~F., Cutri, R.~M.,
1912: Stiening, R., Weinberg, M.~D., Schneider, S. et al. 2006, AJ, 131, 1163
1913: \bibitem[Smith et al. 2005]{Sm05} Smith, V.~V., Cunha, K.,
1914: Ivans, I.~I., Lattanzio, J.~C., Campbell, S., \& Hinkle, K.~H.\ 2005, \apj,
1915: 633, 392
1916: \bibitem[Smith \& Briley 2005]{SB05} Smith, G.~H., \& Briley,
1917: M.~M.\ 2005, \pasp, 117, 895
1918: \bibitem[Smith \& Briley 2006]{SB06} Smith, G.~H., \& Briley, M.~M.\
1919: 2006, \pasp, 118, 740
1920: \bibitem[Sollima et al. 2007]{So07} Sollima, A., Ferraro,
1921: F.~R., Bellazzini, M., Origlia, L., Straniero, O.,
1922: \& Pancino, E.\ 2007, \apj, 654, 915
1923: \bibitem[Sosin et al.(1997)]{1997ApJ...480L..35S} Sosin, C., et al.\ 1997,
1924: \apjl, 480, L35
1925: \bibitem[Straizys \& Kuriliene 1981]{SK81} Straizys, V., \& Kuriliene, G. 1981,
1926: Ap\&SS, 80, 353
1927: \bibitem[Suntzeff \& Smith 1991]{SS91} Suntzeff, N.~B., \& Smith,
1928: V.~V.\ 1991, \apj, 381, 160
1929: \bibitem[Suntzeff \& Kraft 1996]{SK96} Suntzeff, N.~B., \&
1930: Kraft, R.~P.\ 1996, \aj, 111, 1913
1931: \bibitem[Sweigart \& Catelan 1998]{SC98} Sweigart, A.~V., \&
1932: Catelan, M.\ 1998, \apjl, 501, L63
1933: \bibitem[Ventura et al. 2001]{Ve01} Ventura, P., D'Antona,
1934: F., Mazzitelli, I., \& Gratton, R.\ 2001, \apjl, 550, L65
1935: \bibitem[Villanova et al. 2007]{Vi07} Villanova, S., Piotto, G.,
1936: King, I.~R.; Anderson, J., Bedin, L.~R., Gratton, R.~G.,
1937: Cassisi, S., Momany, Y., Bellini, A., Cool, A.~M., et al. 2007,
1938: \apj, 663, 296
1939: \bibitem[Yong \& Grundahl 2008]{YG08} Yong, D., \&
1940: Grundahl, F.\ 2008, \apjl, 672, L29
1941: \end{thebibliography}
1942:
1943: %\end{document}
1944:
1945: \longtab{8}{
1946: \begin{longtable}{lcccccccccccc}
1947: \caption{\label{Tab8} Coordinates, atmospheric parameters, U,B,V,I$_{\rm C}$,\
1948: and 2MASS J,H,K magnitudes for the analised stars. All magnitudes are not corrected for differential reddening.}\\
1949: \hline
1950: \hline
1951: ID & RA($^0$) & DEC($^0$) & T$_{\rm eff}$ (K)& log(g) & v$_{\rm t}$ (km/s) & U & B & V & I$_{\rm C}$ & J & H & K \\
1952: \hline
1953: \endfirsthead
1954: \caption{continued.}\\
1955: \hline\hline
1956: ID & RA($^0$) & DEC($^0$) & T$_{\rm eff}$ (K)& log(g) & v$_{\rm t}$ (km/s) & U & B & V & I$_{\rm C}$ & J & H & K \\
1957: \hline
1958: \endhead
1959: \hline
1960: \endfoot
1961: 19925 & 245.81007350 & -26.60153694 & 4050 & 1.20 & 1.67 & 14.83 & 12.76 & 11.04 & 8.97 & 7.59 & 6.63 & 6.38\\
1962: 20766 & 245.80811060 & -26.55680649 & 4400 & 1.80 & 1.45 & 14.89 & 13.53 & 12.10 & 10.33 & 9.07 & 8.24 & 8.07\\
1963: 21191 & 245.81900040 & -26.53602484 & 4270 & 1.60 & 1.60 & 14.79 & 13.21 & 11.70 & 9.85 & 8.50 & 7.67 & 7.45\\
1964: 21728 & 245.81082240 & -26.51303931 & 4525 & 2.00 & 1.42 & 15.05 & 13.86 & 12.51 & 10.78 & 9.50 & 8.75 & 8.51\\
1965: 22089 & 245.82041320 & -26.49663964 & 4700 & 2.28 & 1.36 & 15.28 & 14.36 & 13.14 & 11.54 & 10.34 & 9.61 & 9.47\\
1966: 24590 & 245.91667710 & -26.65088182 & 4850 & 2.66 & 1.35 & 16.89 & 14.67 & 13.44 & 11.88 & 10.79 & 10.11 & 9.94\\
1967: 25709 & 245.92048770 & -26.62319472 & 4680 & 2.20 & 1.38 & 15.11 & 14.24 & 12.96 & 11.31 & 10.16 & 9.45 & 9.28\\
1968: 26471 & 245.88130650 & -26.60513670 & 4800 & 2.40 & 1.28 & 15.38 & 14.70 & 13.49 & 11.87 & 10.73 & 10.05 & 9.90\\
1969: 26794 & 245.90940720 & -26.59866016 & 4800 & 2.45 & 1.44 & 15.69 & 14.94 & 13.71 & 12.08 & 10.92 & 10.25 & 10.07\\
1970: 27448 & 245.95330690 & -26.58586889 & 4310 & 1.57 & 1.58 & - & 13.24 & 11.73 & 9.86 & 8.57 & 7.71 & 7.53\\
1971: 28103 & 245.84921830 & -26.57493123 & 3860 & 0.50 & 1.62 & 14.75 & 12.52 & 10.71 & 8.51 & 7.01 & 6.03 & 5.75\\
1972: 28356 & 245.97324550 & -26.57086633 & 4600 & 2.22 & 1.53 & 14.93 & 13.92 & 12.57 & 10.87 & 9.67 & 8.92 & 8.72\\
1973: 28707 & 245.93837450 & -26.56579907 & 4880 & 2.74 & 1.31 & 15.59 & 14.84 & 13.61 & 11.99 & 10.80 & 10.14 & 9.95\\
1974: 28797 & 245.97434580 & -26.56458317 & 4640 & 2.35 & 1.36 & 15.13 & 14.22 & 12.93 & 11.30 & 10.12 & 9.43 & 9.21\\
1975: 28847 & 245.90471930 & -26.56390736 & 4780 & 2.40 & 1.27 & 15.43 & 14.76 & 13.54 & 11.88 & 10.67 & 9.98 & 9.77\\
1976: 28977 & 245.94551720 & -26.56218478 & 4680 & 2.33 & 1.40 & 15.22 & 14.31 & 13.02 & 11.36 & 10.14 & 9.44 & 9.26\\
1977: 29027 & 245.84810750 & -26.56146317 & 4720 & 2.40 & 1.41 & 15.18 & 14.40 & 13.16 & 11.48 & 10.29 & 9.58 & 9.39\\
1978: 29065 & 245.94135950 & -26.56081120 & 4650 & 2.10 & 1.41 & 15.08 & 14.27 & 12.98 & 11.31 & 10.10 & 9.42 & 9.24\\
1979: 29171 & 245.89969150 & -26.55958664 & 4880 & 2.64 & 1.26 & 15.72 & 14.95 & 13.73 & 12.09 & 10.88 & 10.16 & 10.00\\
1980: 29222 & 245.84606690 & -26.55896241 & 4720 & 2.50 & 1.36 & 15.28 & 14.42 & 13.17 & 11.52 & 10.32 & 9.59 & 9.41\\
1981: 29272 & 245.86286750 & -26.55825152 & 4780 & 2.50 & 1.26 & 15.32 & 14.64 & 13.42 & 11.76 & 10.54 & 9.86 & 9.67\\
1982: 29282 & 245.85544600 & -26.55811142 & 4650 & 2.30 & 1.42 & 15.17 & 14.17 & 12.88 & 11.18 & 9.91 & 9.20 & 8.98\\
1983: 29397 & 245.87588880 & -26.55671383 & 4600 & 1.50 & 1.78 & 14.60 & 13.53 & 12.20 & 10.47 & 9.25 & 8.54 & 8.32\\
1984: 29545 & 245.92749790 & -26.55485385 & 4880 & 2.61 & 1.18 & 15.53 & 14.96 & 13.79 & 12.23 & 11.08 & 10.43 & 10.24\\
1985: 29598 & 245.88908940 & -26.55414514 & 4840 & 2.50 & 1.40 & 15.65 & 14.89 & 13.68 & 12.05 & 10.86 & 10.19 & 10.01\\
1986: 29693 & 245.88094110 & -26.55307864 & 4360 & 1.10 & 1.88 & 14.67 & 13.27 & 11.80 & 9.96 & 8.63 & 7.79 & 7.64\\
1987: 29848 & 245.89219050 & -26.55108648 & 4780 & 2.52 & 1.24 & 15.48 & 14.75 & 13.52 & 11.87 & 10.69 & 10.00 & 9.83\\
1988: 30209 & 245.89243520 & -26.54681482 & 4880 & 2.62 & 1.32 & 15.52 & 14.87 & 13.66 & 12.05 & 10.84 & 10.15 & 9.99\\
1989: 30345 & 245.94994400 & -26.54524113 & 4850 & 2.73 & 1.31 & 15.64 & 15.00 & 13.84 & 12.31 & 11.19 & 10.55 & 10.39\\
1990: 30450 & 245.86144840 & -26.54411333 & 4760 & 2.53 & 1.35 & 15.44 & 14.59 & 13.34 & 11.69 & 10.47 & 9.78 & 9.59\\
1991: 30452 & 245.88768830 & -26.54411091 & 4830 & 2.56 & 1.25 & 15.72 & 15.07 & 13.88 & 12.26 & 11.07 & 10.38 & 10.22\\
1992: 30549 & 245.88815200 & -26.54294543 & 4830 & 2.52 & 1.28 & 15.52 & 14.86 & 13.66 & 12.05 & 10.76 & 10.13 & 9.99\\
1993: 30598 & 245.89661740 & -26.54235225 & 4360 & 1.75 & 1.47 & 14.72 & 13.49 & 12.05 & 10.23 & 8.83 & 8.03 & 7.75\\
1994: 30653 & 245.88793300 & -26.54184062 & 4660 & 2.30 & 1.25 & 15.23 & 14.40 & 13.14 & 11.50 & 10.29 & 9.56 & 9.38\\
1995: 30675 & 245.90236690 & -26.54157800 & 4830 & 2.58 & 1.35 & 15.52 & 14.77 & 13.58 & 12.00 & - & - & - \\
1996: 30711 & 245.85156170 & -26.54123206 & 4560 & 2.25 & 1.46 & 15.04 & 13.86 & 12.49 & 10.73 & 9.44 & 8.66 & 8.46\\
1997: 30719 & 245.90310860 & -26.54116153 & 4810 & 2.65 & 1.24 & 15.79 & 15.13 & 13.97 & 12.43 & - & - & - \\
1998: 30751 & 245.94302430 & -26.54089574 & 4430 & 1.78 & 1.47 & 14.83 & 13.61 & 12.20 & 10.45 & 9.19 & 8.36 & 8.16\\
1999: 30924 & 245.89624320 & -26.53883396 & 4810 & 2.60 & 1.28 & 15.48 & 14.79 & 13.61 & 12.03 & 9.55 & 8.81 & 8.40\\
2000: 30933 & 245.95895830 & -26.53873652 & 4800 & 2.63 & 1.30 & 15.46 & 14.70 & 13.49 & 11.92 & 10.77 & 10.12 & 9.91\\
2001: 31015 & 245.89710880 & -26.53795346 & 4800 & 2.47 & 1.37 & 14.91 & 14.21 & 13.00 & 11.38 & 10.17 & 9.48 & 9.29\\
2002: 31306 & 245.92607560 & -26.53505913 & 4900 & 2.87 & 1.33 & 15.79 & 15.12 & 13.97 & 12.44 & 11.32 & 10.64 & 10.50\\
2003: 31376 & 245.91219110 & -26.53432365 & 4800 & 2.59 & 1.36 & 15.37 & 14.52 & 13.29 & 11.68 & 10.46 & 9.79 & 9.59\\
2004: 31532 & 245.89561610 & -26.53290071 & 4770 & 2.60 & 1.21 & 15.28 & 14.53 & 13.33 & 11.76 & 10.54 & 9.85 & 9.72\\
2005: 31665 & 245.93007130 & -26.53164468 & 4650 & 2.17 & 1.34 & 14.94 & 14.12 & 12.85 & 11.18 & 9.96 & 9.25 & 9.04\\
2006: 31803 & 245.90925770 & -26.53023616 & 4850 & 2.60 & 1.34 & 15.77 & 15.14 & 13.99 & 12.46 & 11.31 & 10.64 & 10.51\\
2007: 31845 & 245.85370350 & -26.52982260 & 4700 & 2.42 & 1.31 & 15.23 & 14.34 & 13.09 & 11.44 & 10.16 & 9.49 & 9.29\\
2008: 32055 & 245.91709840 & -26.52773869 & 4300 & 1.57 & 1.52 & 14.71 & 13.36 & 11.88 & 10.06 & 8.71 & 7.90 & 7.66\\
2009: 32121 & 245.83812920 & -26.52713080 & 4840 & 2.58 & 1.33 & 15.49 & 14.67 & 13.45 & 11.84 & 10.64 & 9.93 & 9.75\\
2010: 32151 & 245.83708630 & -26.52688433 & 4770 & 2.43 & 1.38 & 15.41 & 14.59 & 13.38 & 11.75 & 10.54 & 9.84 & 9.70\\
2011: 32317 & 245.86225000 & -26.52536111 & 4510 & 1.88 & 1.43 & 15.06 & 13.93 & 12.56 & - & 9.51 & 8.76 & 8.57\\
2012: 32347 & 245.92079170 & -26.52505556 & 4640 & 2.22 & 1.38 & - & - & - & - & 9.92 & 9.20 & 9.00\\
2013: 32583 & 245.89985980 & -26.52267062 & 4850 & 2.54 & 1.34 & 15.46 & 14.80 & 13.61 & 12.03 & 10.80 & 10.16 & 9.95\\
2014: 32627 & 245.97840770 & -26.52219757 & 4750 & 2.42 & 1.30 & 15.01 & 14.34 & 13.14 & 11.59 & 10.48 & 9.79 & 9.61\\
2015: 32700 & 245.91714040 & -26.52145861 & 4560 & 2.12 & 1.36 & 14.88 & 13.86 & 12.52 & 10.80 & 9.54 & 8.81 & 8.57\\
2016: 32724 & 245.92395710 & -26.52127838 & 4850 & 2.73 & 1.30 & 15.53 & 14.82 & 13.62 & 12.04 & 10.88 & 10.20 & 10.04\\
2017: 32782 & 245.89106390 & -26.52065353 & 4880 & 2.60 & 1.24 & 15.81 & 15.10 & 13.94 & 12.38 & - & - & - \\
2018: 32871 & 245.85272450 & -26.51984366 & 4770 & 2.48 & 1.24 & 15.65 & 14.92 & 13.72 & 12.09 & 10.88 & 10.22 & 10.01\\
2019: 32874 & 245.91452240 & -26.51985129 & 4600 & 2.04 & 1.38 & 14.82 & 13.99 & 12.72 & 11.07 & 9.86 & 9.13 & 8.93\\
2020: 32933 & 245.83915100 & -26.51926550 & 4430 & 1.42 & 1.78 & 14.54 & 13.29 & 11.87 & 10.08 & 8.81 & 8.02 & 7.79\\
2021: 32968 & 245.89176500 & -26.51884367 & 4630 & 2.17 & 1.30 & 15.06 & 14.21 & 12.96 & 11.32 & 10.07 & 9.41 & 9.15\\
2022: 32988 & 245.90955560 & -26.51865553 & 4850 & 2.63 & 1.22 & 15.43 & 14.88 & 13.75 & 12.22 & 11.08 & 10.40 & 10.22\\
2023: 33069 & 245.94851290 & -26.51783666 & 4940 & 3.05 & 1.36 & 15.45 & 14.71 & 13.53 & 11.99 & 10.86 & 10.19 & 10.02\\
2024: 33195 & 245.86426280 & -26.51654913 & 4620 & 2.38 & 1.26 & 15.30 & 14.39 & 13.12 & 11.45 & 10.23 & 9.50 & 9.36\\
2025: 33414 & 245.84207930 & -26.51444711 & 4840 & 2.51 & 1.28 & 15.51 & 14.84 & 13.67 & 12.08 & 10.90 & 10.21 & 10.09\\
2026: 33617 & 245.86252960 & -26.51233938 & 4720 & 2.35 & 1.26 & 15.38 & 14.57 & 13.36 & 11.76 & 10.59 & 9.86 & 9.68\\
2027: 33629 & 245.83816250 & -26.51221901 & 4930 & 2.80 & 1.33 & 15.77 & 15.06 & 13.90 & 12.34 & 11.17 & 10.48 & 10.36\\
2028: 33683 & 245.89914420 & -26.51176320 & 4800 & 2.57 & 1.18 & 15.29 & 14.66 & 13.48 & 11.91 & 10.78 & 10.09 & 9.89\\
2029: 33788 & 245.91070220 & -26.51060047 & 4700 & 2.37 & 1.33 & 15.15 & 14.24 & 12.99 & 11.38 & 10.16 & 9.44 & 9.27\\
2030: 33900 & 245.84591550 & -26.50933155 & 4770 & 2.48 & 1.27 & 15.48 & 14.69 & 13.51 & 11.93 & 10.67 & 10.03 & 9.92\\
2031: 33946 & 245.84147130 & -26.50887807 & 4800 & 2.62 & 1.15 & 15.63 & 14.89 & 13.71 & 12.14 & 10.95 & 10.29 & 10.15\\
2032: 34006 & 245.87135820 & -26.50826984 & 4320 & 1.67 & 1.61 & 14.92 & 13.37 & 11.87 & 9.97 & 8.59 & 7.81 & 7.51\\
2033: 34130 & 245.88104790 & -26.50687941 & 4550 & 2.08 & 1.40 & 14.98 & 13.87 & 12.54 & 10.82 & 9.58 & 8.80 & 8.60\\
2034: 34167 & 245.90419020 & -26.50638230 & 4950 & 2.60 & 1.40 & 15.63 & 15.08 & 13.95 & 12.44 & 11.31 & 10.67 & 10.52\\
2035: 34240 & 245.90241860 & -26.50557893 & 4470 & 1.95 & 1.41 & 14.86 & 13.74 & 12.37 & 10.63 & 9.35 & 8.62 & 8.43\\
2036: 34502 & 245.92046770 & -26.50260567 & 4860 & 2.70 & 1.33 & 15.49 & 14.78 & 13.62 & 12.09 & 10.96 & 10.26 & 10.15\\
2037: 34579 & 245.93359340 & -26.50170174 & 4330 & 1.59 & 1.49 & 14.67 & 13.27 & 11.83 & 10.07 & 8.77 & 7.91 & 7.71\\
2038: 34726 & 245.85009660 & -26.50013473 & 4600 & 2.24 & 1.35 & 15.17 & 14.14 & 12.86 & 11.16 & 9.89 & 9.13 & 8.99\\
2039: 35022 & 245.94949140 & -26.49673818 & 4850 & 2.51 & 1.36 & 15.50 & 14.79 & 13.62 & 12.10 & 10.96 & 10.31 & 10.11\\
2040: 35061 & 245.92398380 & -26.49632071 & 4860 & 2.67 & 1.23 & 15.68 & 15.02 & 13.85 & 12.31 & 11.15 & 10.48 & 10.34\\
2041: 35455 & 245.92832620 & -26.49099726 & 4600 & 2.10 & 1.29 & 14.89 & 13.99 & 12.71 & 11.05 & 9.83 & 9.08 & 8.92\\
2042: 35487 & 245.95940170 & -26.49051723 & 4850 & 2.67 & 1.24 & 15.33 & 14.61 & 13.45 & 11.95 & 10.82 & 10.18 & 9.99\\
2043: 35508 & 245.85414850 & -26.49026951 & 4780 & 2.48 & 1.18 & 15.66 & 14.96 & 13.80 & 12.23 & 11.05 & 10.37 & 10.18\\
2044: 35571 & 245.88484770 & -26.48951670 & 4880 & 2.79 & 1.10 & 15.66 & 15.07 & 13.93 & 12.39 & 11.25 & 10.61 & 10.42\\
2045: 35627 & 245.95411440 & -26.48867681 & 4830 & 2.37 & 1.20 & 15.48 & 14.94 & 13.80 & 12.30 & 11.14 & 10.50 & 10.34\\
2046: 35688 & 245.91086870 & -26.48777246 & 4720 & 2.25 & 1.33 & 15.32 & 14.48 & 13.27 & 11.68 & 10.46 & 9.75 & 9.57\\
2047: 35774 & 245.87296550 & -26.48657771 & 4450 & 1.92 & 1.44 & 14.88 & 13.68 & 12.32 & 10.59 & 9.32 & 8.57 & 8.35\\
2048: 36215 & 245.91658540 & -26.48028356 & 4300 & 1.59 & 1.53 & 14.74 & 13.29 & 11.80 & 9.96 & 8.61 & 7.82 & 7.57\\
2049: 36356 & 245.90812220 & -26.47815339 & 4820 & 2.66 & 1.26 & 15.52 & 14.76 & 13.57 & 12.01 & 10.82 & 10.17 & 9.97\\
2050: 36929 & 245.92953710 & -26.46873630 & 4820 & 2.55 & 1.28 & 15.50 & 14.75 & 13.59 & 12.07 & 10.93 & 10.25 & 10.07\\
2051: 36942 & 245.97083610 & -26.46848900 & 4800 & 2.66 & 1.23 & 15.43 & 14.74 & 13.56 & 12.04 & 10.90 & 10.23 & 10.08\\
2052: 37215 & 245.89085360 & -26.46381984 & 4790 & 2.50 & 1.21 & 15.61 & 14.90 & 13.72 & 12.15 & 10.94 & 10.26 & 10.11\\
2053: 38075 & 245.84038430 & -26.44634617 & 4800 & 2.54 & 1.25 & 15.61 & 14.83 & 13.65 & 12.05 & 10.82 & 10.18 & 9.99\\
2054: 38383 & 245.84437580 & -26.43947709 & 4590 & 1.87 & 1.45 & 15.02 & 14.03 & 12.74 & 11.02 & 9.74 & 8.99 & 8.79\\
2055: 38399 & 245.88487200 & -26.43904923 & 4730 & 2.35 & 1.38 & 15.23 & 14.37 & 13.14 & 11.51 & 10.29 & 9.56 & 9.39\\
2056: 38896 & 245.91189860 & -26.42854610 & 4760 & 2.53 & 1.31 & 15.26 & 14.36 & 13.14 & 11.55 & 10.33 & 9.64 & 9.46\\
2057: 42490 & 246.00378460 & -26.59808973 & 4570 & 2.08 & 1.41 & 14.95 & 13.94 & 12.56 & 10.84 & 9.71 & 8.95 & 8.76\\
2058: 42620 & 245.99313500 & -26.59417329 & 4600 & 2.05 & 1.37 & 14.84 & 13.91 & 12.59 & 10.94 & 9.85 & 9.11 & 8.97\\
2059: 43370 & 246.03196790 & -26.57365632 & 4920 & 2.80 & 1.32 & 15.63 & 15.03 & 13.85 & 12.31 & 11.29 & 10.66 & 10.50\\
2060: 44243 & 246.02382800 & -26.55029031 & 4860 & 2.80 & 1.17 & 15.51 & 14.91 & 13.75 & 12.24 & 11.23 & 10.59 & 10.42\\
2061: 44595 & 246.06621120 & -26.54096589 & 4750 & 2.40 & 1.40 & 15.02 & 14.23 & 13.00 & 11.45 & 10.42 & 9.75 & 9.58\\
2062: 44616 & 245.98944100 & -26.54022147 & 4620 & 2.20 & 1.44 & 14.99 & 13.99 & 12.65 & 10.99 & 9.82 & 9.11 & 8.89\\
2063: 45163 & 246.00502890 & -26.52572879 & 4770 & 2.40 & 1.26 & 15.27 & 14.59 & 13.40 & 11.88 & 10.81 & 10.17 & 9.98\\
2064: 45895 & 246.01787760 & -26.50775258 & 4720 & 2.25 & 1.34 & 14.93 & 14.22 & 12.99 & 11.43 & 10.38 & 9.71 & 9.52\\
2065: 5359 & 245.91226500 & -26.36929347 & 4800 & 2.44 & 1.28 & 15.60 & 14.78 & 13.58 & 12.01 & 10.87 & 10.21 & 10.05\\
2066: \end{longtable}
2067: }
2068:
2069: %\end{document}
2070:
2071: \clearpage
2072:
2073: \longtab{9}{
2074: \begin{longtable}{lcccccccccccc}
2075: \caption{\label{Tab9} Chemical abundances of the analised stars.}\\
2076: \hline
2077: \hline
2078: \scriptsize{ID} & \scriptsize{[FeI/H]} & \scriptsize{[OI/Fe]} & \scriptsize{[NaI/Fe]$_{NLTE}$} & \scriptsize{[MgI/Fe]$_{NLTE}$} & \scriptsize{[AlI/Fe]} & \scriptsize{[SiI/Fe]} & \scriptsize{[CaI/Fe]} & \scriptsize{[TiI/Fe]} & \scriptsize{[TiII/Fe]} & \scriptsize{[CrI/Fe]} & \scriptsize{[NiI/Fe]} & \scriptsize{[BaII/Fe]}\\
2079: \hline
2080: \endfirsthead
2081: \caption{continued.}\\
2082: \hline\hline
2083: \scriptsize{ID} & \scriptsize{[FeI/H]} & \scriptsize{[OI/Fe]} & \scriptsize{[NaI/Fe]$_{NLTE}$} & \scriptsize{[MgI/Fe]$_{NLTE}$} & \scriptsize{[AlI/Fe]} & \scriptsize{[SiI/Fe]} & \scriptsize{[CaI/Fe]} & \scriptsize{[TiI/Fe]} & \scriptsize{[TiII/Fe]} & \scriptsize{[CrI/Fe]} & \scriptsize{[NiI/Fe]} & \scriptsize{[BaII/Fe]}\\
2084: \hline
2085: \endhead
2086: \hline
2087: \endfoot
2088: 19925 & $-$1.02 & 0.28 & 0.51 & 0.43 & 0.68 & 0.50 & 0.19 & 0.48 & 0.29 & $-$0.04 & 0.09 & 0.64 \\
2089: 20766 & $-$1.05 & 0.31 & 0.53 & 0.49 & 0.62 & 0.49 & 0.26 & 0.31 & 0.39 & $-$0.01 & 0.04 & 0.45 \\
2090: 21191 & $-$1.06 & 0.34 & 0.51 & 0.55 & 0.59 & 0.54 & 0.24 & 0.31 & 0.32 & $-$0.00 & 0.03 & 0.53 \\
2091: 21728 & $-$1.06 & 0.31 & 0.37 & 0.52 & 0.68 & 0.47 & 0.28 & 0.32 & 0.30 & $-$0.04 & 0.03 & 0.48 \\
2092: 22089 & $-$1.06 & 0.35 & 0.50 & 0.30 & -- & 0.53 & 0.31 & 0.25 & 0.32 & $-$0.15 & 0.03 & 0.46 \\
2093: 24590 & $-$1.07 & 0.50 & 0.30 & 0.61 & -- & 0.50 & 0.27 & 0.29 & 0.31 & $-$0.09 & 0.04 & 0.33 \\
2094: 25709 & $-$1.13 & 0.44 & 0.34 & 0.54 & 0.59 & 0.53 & 0.35 & 0.26 & 0.33 & $-$0.02 & $-$0.01 & 0.35 \\
2095: 26471 & $-$1.10 & 0.41 & 0.09 & 0.59 & 0.58 & 0.45 & 0.30 & 0.23 & 0.31 & 0.00 & $-$0.01 & 0.28 \\
2096: 26794 & $-$1.17 & -- & 0.36 & 0.53 & -- & 0.50 & 0.34 & 0.34 & 0.39 & $-$0.01 & 0.03 & 0.33 \\
2097: 27448 & $-$1.12 & 0.51 & 0.11 & 0.50 & 0.42 & 0.45 & 0.29 & 0.39 & 0.36 & $-$0.02 & 0.03 & 0.35 \\
2098: 28103 & $-$1.08 & 0.50 & 0.17 & 0.41 & 0.33 & 0.55 & 0.16 & 0.48 & 0.26 & $-$0.11 & 0.01 & 0.26 \\
2099: 28356 & $-$1.14 & 0.44 & 0.37 & 0.55 & 0.57 & 0.53 & 0.28 & 0.29 & 0.37 & 0.04 & 0.05 & 0.39 \\
2100: 28707 & $-$1.03 & -- & 0.22 & 0.40 & -- & 0.43 & 0.20 & 0.27 & 0.26 & $-$0.06 & 0.02 & 0.45 \\
2101: 28797 & $-$1.12 & 0.35 & 0.44 & 0.55 & 0.57 & 0.57 & 0.26 & 0.28 & 0.44 & $-$0.12 & 0.01 & 0.53 \\
2102: 28847 & $-$1.16 & 0.52 & 0.08 & 0.52 & -- & 0.51 & 0.32 & 0.31 & 0.30 & $-$0.09 & $-$0.04 & 0.46 \\
2103: 28977 & $-$1.14 & 0.39 & 0.40 & 0.51 & 0.65 & 0.57 & 0.30 & 0.31 & 0.35 & $-$0.02 & 0.05 & 0.42 \\
2104: 29027 & $-$1.10 & 0.51 & 0.02 & 0.48 & -- & 0.44 & 0.27 & 0.27 & 0.35 & $-$0.01 & $-$0.03 & 0.34 \\
2105: 29065 & $-$1.12 & 0.45 & 0.17 & 0.61 & -- & 0.48 & 0.27 & 0.21 & 0.35 & $-$0.06 & $-$0.01 & 0.35 \\
2106: 29171 & $-$0.99 & 0.22 & 0.41 & 0.43 & -- & 0.49 & 0.32 & 0.32 & 0.31 & $-$0.04 & $-$0.04 & 0.36 \\
2107: 29222 & $-$1.04 & 0.36 & 0.24 & 0.44 & 0.44 & 0.47 & 0.25 & 0.27 & 0.58 & $-$0.06 & 0.05 & 0.39 \\
2108: 29272 & $-$1.11 & 0.53 & 0.05 & 0.47 & -- & 0.43 & 0.25 & 0.30 & 0.33 & 0.01 & 0.02 & 0.35 \\
2109: 29282 & $-$1.06 & 0.28 & 0.42 & 0.50 & 0.61 & 0.49 & 0.25 & 0.23 & 0.34 & $-$0.07 & 0.02 & 0.27 \\
2110: 29397 & $-$1.12 & 0.24 & 0.46 & 0.55 & 0.51 & 0.53 & 0.20 & 0.20 & 0.15 & $-$0.11 & $-$0.03 & 0.43 \\
2111: 29545 & $-$1.06 & 0.47 &$-$0.02 & 0.44 & 0.15 & 0.36 & 0.29 & 0.29 & 0.33 & $-$0.07 & 0.00 & 0.20 \\
2112: 29598 & $-$1.06 & 0.39 & 0.40 & 0.48 & -- & 0.50 & 0.34 & 0.26 & 0.42 & $-$0.04 & 0.02 & 0.28 \\
2113: 29693 & $-$1.19 & 0.26 & 0.50 & 0.60 & 0.64 & 0.55 & 0.14 & 0.23 & 0.28 & $-$0.17 & 0.00 & 0.36 \\
2114: 29848 & $-$1.05 & 0.54 & 0.09 & 0.54 & 0.64 & 0.48 & 0.29 & 0.32 & 0.34 & $-$0.01 & $-$0.01 & 0.55 \\
2115: 30209 & $-$0.99 & 0.42 &$-$0.05 & 0.46 & -- & 0.42 & 0.30 & 0.31 & 0.46 & 0.03 & 0.05 & 0.31 \\
2116: 30345 & $-$1.06 & 0.32 & 0.43 & 0.44 & -- & 0.53 & 0.28 & 0.28 & 0.52 & $-$0.05 & 0.07 & 0.36 \\
2117: 30450 & $-$1.00 & 0.40 & 0.40 & 0.52 & 0.51 & 0.49 & 0.26 & 0.29 & 0.32 & 0.11 & 0.05 & 0.45 \\
2118: 30452 & $-$1.00 & 0.40 & 0.06 & 0.50 & 0.53 & 0.46 & 0.27 & 0.29 & 0.36 & 0.02 & 0.00 & 0.53 \\
2119: 30549 & $-$1.09 & 0.45 & 0.13 & 0.52 & -- & 0.48 & 0.34 & 0.31 & 0.35 & 0.00 & 0.06 & 0.40 \\
2120: 30598 & $-$1.07 & 0.43 & 0.05 & 0.54 & 0.37 & 0.54 & 0.25 & 0.33 & 0.41 & 0.01 & 0.05 & 0.52 \\
2121: 30653 & $-$1.06 & 0.47 & 0.15 & 0.47 & 0.46 & 0.51 & 0.27 & 0.21 & 0.32 & $-$0.06 & $-$0.01 & 0.45 \\
2122: 30675 & $-$1.07 & 0.36 & 0.41 & 0.55 & 0.65 & 0.50 & 0.36 & 0.32 & 0.42 & $-$0.02 & 0.07 & 0.35 \\
2123: 30711 & $-$1.01 & 0.37 & 0.32 & 0.47 & 0.53 & 0.45 & 0.22 & 0.31 & 0.41 & $-$0.07 & 0.02 & 0.46 \\
2124: 30719 & $-$1.19 & 0.44 & 0.42 & 0.50 & 0.68 & 0.46 & 0.33 & 0.27 & 0.48 & $-$0.09 & 0.02 & 0.45 \\
2125: 30751 & $-$1.09 & 0.40 & 0.32 & 0.58 & 0.54 & 0.48 & 0.28 & 0.38 & 0.33 & 0.10 & 0.03 & 0.42 \\
2126: 30924 & $-$1.09 & 0.48 & 0.20 & 0.47 & 0.48 & 0.50 & 0.28 & 0.30 & 0.41 & $-$0.05 & 0.02 & 0.40 \\
2127: 30933 & $-$1.07 & 0.44 & 0.44 & 0.44 & -- & 0.58 & 0.30 & 0.28 & 0.34 & $-$0.08 & 0.02 & 0.46 \\
2128: 31015 & $-$1.07 & 0.50 & 0.00 & 0.51 & 0.36 & 0.42 & 0.33 & 0.32 & 0.40 & 0.06 & $-$0.01 & 0.32 \\
2129: 31306 & $-$1.11 & -- & 0.40 & 0.41 & 0.53 & 0.45 & 0.29 & 0.33 & 0.38 & 0.03 & 0.04 & 0.36 \\
2130: 31376 & $-$1.00 & 0.33 & 0.43 & 0.46 & 0.59 & 0.43 & 0.29 & 0.29 & 0.34 & $-$0.01 & 0.03 & 0.40 \\
2131: 31532 & $-$1.03 & 0.47 & 0.38 & 0.44 & 0.55 & 0.47 & 0.26 & 0.30 & 0.39 & $-$0.08 & 0.00 & 0.49 \\
2132: 31665 & $-$1.07 & 0.48 & 0.02 & 0.49 & 0.29 & 0.46 & 0.31 & 0.26 & 0.35 & $-$0.06 & 0.01 & 0.29 \\
2133: 31803 & $-$1.13 & 0.38 & 0.25 & 0.44 & 0.49 & 0.51 & 0.32 & 0.31 & 0.28 & $-$0.10 & $-$0.01 & 0.29 \\
2134: 31845 & $-$1.06 & 0.27 & 0.38 & 0.41 & 0.67 & 0.48 & 0.27 & 0.27 & 0.35 & $-$0.09 & 0.02 & 0.47 \\
2135: 32055 & $-$1.12 & 0.31 & 0.40 & 0.55 & 0.59 & 0.52 & 0.27 & 0.39 & 0.28 & 0.03 & 0.00 & 0.45 \\
2136: 32121 & $-$0.93 & -- & 0.35 & 0.44 & 0.47 & 0.56 & 0.29 & 0.32 & 0.37 & $-$0.08 & 0.02 & 0.38 \\
2137: 32151 & $-$1.07 & -- & 0.39 & 0.42 & -- & 0.49 & 0.29 & 0.26 & 0.34 & $-$0.04 & $-$0.01 & 0.33 \\
2138: 32317 & $-$1.07 & 0.37 & 0.37 & 0.54 & -- & 0.43 & 0.31 & 0.33 & 0.34 & 0.08 & 0.03 & 0.36 \\
2139: 32347 & $-$1.09 & 0.40 & 0.23 & 0.51 & 0.57 & 0.49 & 0.27 & 0.29 & 0.32 & $-$0.06 & 0.03 & 0.43 \\
2140: 32583 & $-$1.08 & 0.46 & 0.07 & 0.46 & 0.50 & 0.39 & 0.31 & 0.32 & 0.34 & $-$0.07 & 0.00 & 0.29 \\
2141: 32627 & $-$1.10 & 0.50 &$-$0.03 & 0.51 & 0.36 & 0.34 & 0.29 & 0.26 & 0.38 & $-$0.01 & $-$0.00 & 0.36 \\
2142: 32700 & $-$1.02 & 0.49 & 0.06 & 0.61 & 0.56 & 0.47 & 0.22 & 0.37 & 0.39 & 0.06 & 0.00 & 0.52 \\
2143: 32724 & $-$1.03 & 0.51 & 0.11 & 0.48 & 0.53 & 0.39 & 0.30 & 0.34 & 0.39 & $-$0.04 & 0.00 & 0.49 \\
2144: 32782 & $-$1.05 & 0.29 & 0.46 & 0.45 & 0.58 & 0.41 & 0.34 & 0.26 & 0.27 & $-$0.07 & $-$0.03 & 0.48 \\
2145: 32871 & $-$1.01 & 0.40 & 0.06 & 0.62 & 0.81 & 0.54 & 0.29 & 0.31 & 0.34 & $-$0.05 & 0.01 & 0.65 \\
2146: 32874 & $-$1.14 & 0.41 & 0.11 & 0.55 & 0.52 & 0.42 & 0.26 & 0.27 & 0.32 & 0.02 & $-$0.02 & 0.39 \\
2147: 32933 & $-$1.13 & 0.48 & 0.02 & 0.56 & -- & 0.54 & 0.20 & 0.29 & 0.22 & $-$0.14 & 0.04 & 0.53 \\
2148: 32968 & $-$1.13 & 0.41 & 0.24 & 0.54 & 0.50 & 0.49 & 0.30 & 0.25 & 0.42 & $-$0.03 & 0.01 & 0.48 \\
2149: 32988 & $-$1.09 & 0.51 & 0.02 & 0.44 & -- & 0.50 & 0.29 & 0.29 & 0.33 & $-$0.03 & $-$0.01 & -- \\
2150: 33069 & $-$0.92 & 0.39 & 0.23 & 0.37 & 0.59 & 0.43 & 0.20 & 0.34 & 0.42 & $-$0.06 & 0.03 & 0.42 \\
2151: 33195 & $-$1.03 & 0.48 & 0.09 & 0.60 & 0.66 & 0.56 & 0.27 & 0.32 & 0.34 & $-$0.02 & 0.04 & -- \\
2152: 33414 & $-$1.05 & 0.52 & 0.21 & 0.49 & 0.30 & 0.48 & 0.32 & 0.25 & 0.26 & $-$0.05 & 0.00 & 0.36 \\
2153: 33617 & $-$1.09 & 0.29 & 0.37 & 0.54 & 0.58 & 0.51 & 0.32 & 0.26 & 0.31 & $-$0.05 & 0.02 & 0.51 \\
2154: 33629 & $-$0.98 & -- & 0.31 & 0.39 & 0.68 & 0.55 & 0.28 & 0.33 & 0.37 & $-$0.02 & 0.05 & 0.34 \\
2155: 33683 & $-$1.05 & 0.42 & 0.00 & 0.51 & 0.51 & 0.42 & 0.27 & 0.27 & 0.39 & 0.03 & $-$0.03 & 0.56 \\
2156: 33788 & $-$1.02 & 0.30 & 0.37 & 0.47 & 0.50 & 0.43 & 0.28 & 0.28 & 0.33 & $-$0.02 & 0.00 & 0.48 \\
2157: 33900 & $-$1.06 & 0.30 & 0.47 & 0.49 & 0.55 & 0.46 & 0.32 & 0.30 & 0.39 & $-$0.05 & 0.01 & 0.42 \\
2158: 33946 & $-$1.03 & 0.23 & 0.32 & 0.42 & 0.60 & 0.45 & 0.30 & 0.29 & 0.33 & $-$0.03 & 0.03 & 0.50 \\
2159: 34006 & $-$1.06 & 0.25 & 0.44 & 0.52 & 0.67 & 0.48 & 0.26 & 0.37 & 0.27 & $-$0.08 & 0.04 & 0.42 \\
2160: 34130 & $-$1.09 & 0.33 & 0.43 & 0.53 & 0.59 & 0.51 & 0.29 & 0.26 & 0.34 & $-$0.04 & $-$0.01 & 0.47 \\
2161: 34167 & $-$1.10 & -- & 0.13 & 0.56 & 0.48 & 0.35 & 0.31 & 0.25 & 0.23 & 0.01 & $-$0.02 & 0.04 \\
2162: 34240 & $-$1.10 & 0.47 & 0.08 & 0.55 & 0.55 & 0.49 & 0.25 & 0.35 & 0.30 & 0.04 & 0.03 & 0.47 \\
2163: 34502 & $-$1.08 & 0.35 & 0.34 & 0.47 & 0.49 & 0.47 & 0.28 & 0.33 & 0.38 & $-$0.06 & 0.02 & 0.34 \\
2164: 34579 & $-$1.08 & 0.25 & 0.46 & 0.52 & 0.61 & 0.50 & 0.28 & 0.38 & 0.32 & $-$0.06 & 0.02 & 0.46 \\
2165: 34726 & $-$1.01 & 0.37 & 0.42 & 0.49 & 0.60 & 0.46 & 0.28 & 0.31 & 0.34 & $-$0.06 & 0.03 & 0.40 \\
2166: 35022 & $-$1.08 & 0.30 & 0.41 & 0.52 & 0.58 & 0.50 & 0.33 & 0.32 & 0.22 & $-$0.05 & 0.02 & 0.34 \\
2167: 35061 & $-$0.99 & 0.50 & 0.06 & 0.50 & 0.52 & 0.40 & 0.24 & 0.24 & 0.33 & $-$0.06 & $-$0.01 & 0.39 \\
2168: 35455 & $-$1.06 & 0.45 & 0.01 & 0.49 & 0.49 & 0.44 & 0.24 & 0.23 & 0.37 & $-$0.01 & $-$0.02 & 0.53 \\
2169: 35487 & $-$1.00 & 0.36 & 0.30 & 0.42 & 0.55 & 0.44 & 0.27 & 0.32 & 0.43 & $-$0.05 & $-$0.01 & 0.51 \\
2170: 35508 & $-$1.05 & 0.25 & 0.38 & 0.41 & 0.45 & 0.45 & 0.27 & 0.22 & 0.27 & $-$0.11 & $-$0.05 & 0.46 \\
2171: 35571 & $-$0.99 & 0.45 &$-$0.02 & 0.46 & 0.52 & 0.45 & 0.27 & 0.24 & 0.34 & $-$0.09 & $-$0.04 & 0.55 \\
2172: 35627 & $-$1.08 & -- & 0.04 & 0.56 & 0.51 & 0.38 & 0.33 & 0.26 & 0.38 & $-$0.09 & 0.00 & 0.29 \\
2173: 35688 & $-$1.11 & 0.21 & 0.40 & 0.53 & 0.59 & 0.47 & 0.33 & 0.23 & 0.34 & $-$0.04 & 0.02 & 0.40 \\
2174: 35774 & $-$1.10 & 0.43 & 0.24 & 0.54 & 0.51 & 0.53 & 0.26 & 0.36 & 0.31 & $-$0.03 & 0.03 & 0.43 \\
2175: 36215 & $-$1.11 & 0.48 & 0.18 & 0.52 & 0.53 & 0.53 & 0.26 & 0.38 & 0.34 & $-$0.04 & 0.02 & 0.46 \\
2176: 36356 & $-$1.05 & 0.30 & 0.26 & 0.44 & 0.45 & 0.50 & 0.28 & 0.27 & 0.43 & $-$0.02 & 0.03 & 0.51 \\
2177: 36929 & $-$1.03 & 0.34 & 0.45 & 0.41 & 0.73 & 0.50 & 0.30 & 0.28 & 0.38 & $-$0.04 & 0.01 & 0.46 \\
2178: 36942 & $-$0.98 & 0.55 & 0.04 & 0.52 & 0.37 & 0.45 & 0.24 & 0.27 & 0.37 & $-$0.16 & 0.07 & 0.45 \\
2179: 37215 & $-$1.11 & 0.45 & 0.25 & 0.50 & 0.62 & 0.48 & 0.29 & 0.27 & 0.47 & $-$0.07 & $-$0.02 & 0.38 \\
2180: 38075 & $-$1.07 & -- & 0.42 & 0.53 & 0.71 & 0.54 & 0.33 & 0.27 & 0.31 & $-$0.05 & 0.08 & 0.52 \\
2181: 38383 & $-$1.10 & 0.39 & 0.11 & 0.54 & 0.39 & 0.44 & 0.25 & 0.27 & 0.26 & $-$0.06 & $-$0.01 & 0.18 \\
2182: 38399 & $-$1.08 & 0.52 & 0.19 & 0.55 & 0.64 & 0.50 & 0.29 & 0.30 & 0.32 & $-$0.03 & 0.04 & 0.42 \\
2183: 38896 & $-$1.02 & 0.31 & 0.43 & 0.48 & 0.49 & 0.47 & 0.28 & 0.33 & 0.39 & $-$0.04 & 0.01 & 0.49 \\
2184: 42490 & $-$1.07 & 0.29 & 0.38 & 0.48 & 0.58 & 0.50 & 0.24 & 0.29 & 0.33 & $-$0.09 & 0.03 & 0.43 \\
2185: 42620 & $-$1.09 & 0.22 & 0.48 & 0.49 & 0.64 & 0.52 & 0.29 & 0.26 & 0.33 & $-$0.05 & 0.03 & 0.45 \\
2186: 43370 & $-$1.05 & -- & 0.34 & 0.40 & 0.51 & 0.51 & 0.29 & 0.28 & 0.44 & $-$0.02 & 0.02 & 0.31 \\
2187: 44243 & $-$1.04 & -- & 0.41 & 0.39 & 0.47 & 0.53 & 0.22 & 0.25 & 0.49 & $-$0.04 & 0.01 & 0.43 \\
2188: 44595 & $-$1.07 & 0.20 & 0.44 & 0.52 & 0.61 & 0.46 & 0.29 & 0.31 & 0.38 & $-$0.00 & 0.03 & 0.34 \\
2189: 44616 & $-$1.04 & 0.28 & 0.44 & 0.53 & 0.65 & 0.46 & 0.29 & 0.25 & 0.33 & 0.01 & 0.02 & 0.37 \\
2190: 45163 & $-$1.10 & -- & 0.46 & 0.49 & 0.57 & 0.54 & 0.28 & 0.21 & 0.35 & $-$0.03 & $-$0.00 & 0.43 \\
2191: 45895 & $-$1.04 & 0.43 & 0.10 & 0.57 & 0.54 & 0.46 & 0.26 & 0.23 & 0.32 & $-$0.07 & 0.00 & 0.48 \\
2192: 5359 & $-$1.03 & 0.42 & 0.13 & 0.55 & 0.53 & 0.50 & 0.31 & 0.30 & 0.39 & $-$0.03 & 0.03 & 0.24 \\
2193: \end{longtable}
2194: }
2195:
2196: \end{document}
2197: