1: \documentclass[12pt,preprint]{aastex}
2:
3: \shorttitle{Abundances in M10}
4: \shortauthors{Haynes et al.}
5:
6: \newcommand\iso[2]{$^{\rm #1}$#2}
7:
8: \begin{document}
9:
10: \title{Chemical Analysis of Five Red Giants in the Globular Cluster M10 (NGC
11: 6254)}
12:
13: \author{
14: Sharina Haynes\altaffilmark{1},
15: Geoffrey Burks\altaffilmark{1},
16: Christian I. Johnson\altaffilmark{2}, and
17: Catherine A. Pilachowski\altaffilmark{2}
18: }
19:
20: \altaffiltext{1}{Center of Excellence in Information Systems, Tennessee
21: State University, 3500 John Merritt Blvd, Box 9501, Holland Hall Room 311,
22: Nashville, Tennessee 37209, USA; burks@coe.tsuniv.edu}
23:
24: \altaffiltext{2}{Department of Astronomy, Indiana University,
25: Swain West 319, 727 East Third Street, Bloomington, IN 47405--7105, USA;
26: cijohnson@astro.indiana.edu; catyp@astro.indiana.edu}
27:
28: \begin{abstract}
29:
30: We have determined Al, $\alpha$, Fe--peak, and neutron capture elemental
31: abundances for five red giant branch (RGB) stars in the Galactic globular
32: cluster M10. Abundances were determined using equivalent width analyses
33: of moderate resolution (R$\sim$15,000) spectra obtained with the Hydra
34: multifiber positioner and bench spectrograph on the WIYN telescope. The data
35: sample the upper RGB from the luminosity level near the horizontal branch
36: to about 0.5 mag below the RGB tip. We find in agreement with previous studies
37: that M10 is moderately metal--poor with [Fe/H]=--1.45 ($\sigma$=0.04). All
38: stars appear enhanced in Al with $\langle$[Al/Fe]$\rangle$=+0.33
39: ($\sigma$=0.19), but no stars have [Al/Fe]$\ga$+0.55. We find the $\alpha$
40: elements to be enhanced by +0.20 to +0.40 dex and the Fe--peak elements to have
41: [el/Fe]$\sim$0, which are consistent with predictions from type II SNe ejecta.
42: Additionally, the cluster appears to be r--process rich with
43: $\langle$[Eu/La]$\rangle$=+0.41.
44:
45: \end{abstract}
46:
47: \keywords{stars: abundances, globular clusters: general, globular clusters:
48: individual (M10, NGC 6254). Galaxy: halo, stars: Population II}
49:
50: \section{INTRODUCTION}
51:
52: Although few chemical analysis studies of M10 exist, the general consensus is
53: that this cluster exhibits all of the classical characteristics observed
54: in other Galactic globular clusters. With a metallicity of
55: [Fe/H]$\approx$--1.5 (Kraft et al. 1995), M10 lies near the median metallicity
56: distribution for halo globular clusters (Laird et al. 1988). Small sample
57: (N$\la$15) analyses of red giant branch (RGB) stars in this cluster have
58: revealed it to have [$\alpha$/Fe]$\sim$+0.30 and [el/Fe]$\sim$0 for Fe--peak
59: elements (Kraft et al. 1995; Mishenina et al. 2003). These values are
60: consistent with the current generation of M10 stars having been polluted by
61: the ejecta of type II supernovae (SNe) without significant contributions from
62: type Ia SNe.
63:
64: While it has long been known that nearly all globular cluster giants show
65: star--to--star variations of the light elements (A$\la$27), the source of many
66: of these anomalies has yet to be determined. Numerous observations of
67: globular cluster stars from the main sequence to above the RGB luminosity bump
68: have revealed declining [C/Fe] and increasing [N/Fe] ratios as a function of
69: increasing luminosity (e.g., see reviews by Kraft et al. 1994; Gratton et al.
70: 2004; Carretta 2008). These observations show clear evidence of CN--cycle
71: products being brought to the surface and are a confirmation of first
72: dredge--up predictions (Iben 1964). Smith \& Fulbright (1997) and Smith et al.
73: (2005) have verified this trend in M10 as well as a CN band anticorrelation
74: with [O/Fe] for stars at various RGB luminosities. However, the large spread
75: in [N/Fe] of about 1.0 dex found by Smith et al. (2005) in M10 stars may be
76: evidence for primordial variations superimposed on in situ mixing.
77:
78: The C and N abundance anomalies are known to exist in both globular cluster
79: and field giants, but that likeness does not extend to the well documented
80: O/Na, Mg/Al, and O/F anticorrelations and Na/Al correlation seen solely in
81: globular cluster stars (e.g., Gratton et al. 2004). These abundance
82: relationships are clear signs of proton--capture nucleosynthesis, but where
83: these processes are operating is still a mystery. Kraft et al. (1995) examined
84: the O/Na anticorrelations of 15 bright giants in M10 along with M3 and M13,
85: which are all globular clusters of similar metallicity ([Fe/H]$\approx$--1.5),
86: because M10 and M13 have extremely blue horizontal branches (HB) but M3 has a
87: uniform distribution of blue HB, RR Lyrae, and red HB stars. The study showed
88: that M10 appears to be an intermediate case in terms of O depletion and Na
89: enhancement in that the average [O/Fe] is lower in M10 than in M3, but no M10
90: giants were super O--poor (i.e., [O/Fe]$<$--0.6), suggesting the process
91: driving O depletion does not itself determine HB morphology.
92:
93: In this paper we have examined five additional RGB stars in M10 that are
94: located above the luminosity of the horizontal branch but below the RGB tip.
95: We have derived Al, $\alpha$, Fe--peak, and heavy element abundances to
96: examine how M10 fits into context with other globular clusters of similar
97: metallicity and HB morphology.
98:
99: \section{OBSERVATIONS AND REDUCTIONS}
100:
101: The observation of cluster giants were obtained using the Hydra multifiber
102: positioner and bench spectrograph on the 3.5 meter WIYN telescope at Kitt
103: Peak National Observatory in May, 2000. The observations consisted of three,
104: 3,000 second exposures with the 200 $\mu$m red fiber bundle. The 316 line
105: mm$^{\rm -1}$ echelle grating and red camera provided a resolution of R
106: ($\lambda$/$\Delta$$\lambda$)$\sim$15,000 at 6650~\AA, with wavelength
107: coverage extending from approximately 6460--6860~\AA.
108:
109: Target stars and photometry were taken from the photometric survey by Arp
110: (1955) and astrometry was taken from the USNO Image and Catalogue
111: Archive.\footnote{http://www.nofs.navy.mil/data/fchpix/}
112: Sample selection focused on observing stars with V magnitudes brighter than
113: the HB and extending up to the RGB tip. However, the final sample only
114: includes stars with magnitudes up to about 0.5 mag below the RGB tip. The
115: Hydra configuration allowed for fiber placement on 22 objects, but only 7 of
116: those 22 had sufficient signal--to--noise (S/N) for reliable abundance
117: determinations. Two of the remaining program stars (IV--44 \& IV--87) were
118: found to have wavelength shifts and H$\alpha$ profiles inconsistent with being
119: both cluster members and low surface gravity RGB stars. Comparison with the
120: proper motion study by Chen et al. (2000) reveals that both of these stars
121: have proper motions inconsistent with other cluster members.
122:
123: The IRAF\footnote{IRAF is distributed by the National Optical Astronomy
124: Observatory, which is operated by AURA, Inc., under cooperative agreement with
125: the National Science Foundation.} task {\it ccdproc} was used to trim the bias
126: overscan region and apply the bias level correction. The IRAF routine {\it
127: dohydra} was employed to apply the flat field correction, linearize the
128: wavelength scale, correct for scattered light, remove cosmic rays, subtract
129: the sky, and extract the one--dimensional spectra. Typical S/N ratios of
130: individual spectra are 25--50 with co--added spectra having S/N ratios of about
131: 50--75. A sample spectral region for all five program stars is shown in
132: Figure \ref{f1}.
133:
134: \section{ANALYSIS}
135:
136: \subsection{Model Stellar Atmospheres}
137:
138: The significant differential reddening associated with M10 (e.g., von Braun et
139: al. 2002) makes effective temperature (T$_{\rm eff}$) and surface gravity
140: (log g) estimates based on color and photometric indices difficult.
141: Therefore, we employed an iterative method to obtain T$_{\rm eff}$ by
142: removing Fe I abundance trends as a function of excitation potential and
143: microturbulence (V$_{\rm t}$) by removing Fe I abundance trends as a function
144: of reduced width [log(EW/$\lambda$)]. Surface gravity was obtained by
145: enforcing ionization equilibrium between Fe I and Fe II because ionized Fe
146: lines in these cool giants are more sensitive to changes in log g than neutral
147: lines (e.g., Johnson \& Pilachowksi 2006; their Table 3). Despite the fact
148: that only one Fe II line (6516~\AA) was available for analysis, we used the
149: ionization equilibrium method because of the potentially large and variable
150: uncertainties in determining bolometric absolute magnitudes.
151:
152: All initial models assumed a metallicity of [Fe/H] = --1.50, which is
153: consistent with previous spectroscopic [Fe/H] estimates (e.g., Kraft et al.
154: 1995; Mishenina et al. 2003). The model stellar atmospheres (without
155: convective overshoot) were created by interpolating in the ATLAS
156: grid\footnote{Kurucz model atmosphere grids can be downloaded from
157: http://cfaku5.cfa.harvard.edu/grids.html.} (Castelli et al. 1997). The
158: temperature range of our observations covers 4450 $\le$ T$_{\rm eff}$ $\le$
159: 4750 corresponding to surface gravity values of about 1.20 $\le$ log g $\le$
160: 1.85. Our adopted values of T$_{\rm eff}$ and log g are in reasonable
161: agreement with position on the color--magnitude diagram as estimated from
162: V and B--V photometry. A summary of our adopted model atmosphere parameters
163: and associated photometry is provided in Table 1.
164:
165: \subsection{Equivalent Width Analyses}
166:
167: All abundances were determined by measuring equivalent widths using the {\it
168: splot} package in IRAF. Given the moderate resolution and S/N of our spectra,
169: we restricted measurements to isolated lines that did not suffer significant
170: blending problems and which had equivalent widths $\ga$10 m\AA. Suitable
171: lines were chosen via comparison with a high S/N, high resolution Arcturus
172: spectrum\footnote{The Arcturus Atlas can be downloaded from the NOAO Digital
173: Library at http://www.noao.edu/dpp/library.html.}, which also served as a
174: reference aiding continuum placement. The final linelist including all
175: measured equivalent widths is given in Table 2, with atomic parameters taken
176: from Johnson \& Pilachowski (2006).
177:
178: While most abundances were calculated using the {\it abfind} driver in the 2002
179: version of the LTE line analysis code MOOG (Sneden 1973), the elements Sc and
180: Eu required a modified approach. These spectral features may be sensitive
181: to hyper--fine splitting resulting from spin--orbit coupling and Eu has the
182: added complication of having two naturally occurring stable isotopes
183: (\iso{151}{Eu} and \iso{153}{Eu}). Both of these effects can cause line
184: broadening that will force single--line equivalent width measurements to
185: overestimate the abundances. Therefore, we used the {\it blends} driver in
186: MOOG with linelists including hyperfine and/or isotopic data from Prochaska \&
187: McWilliam (2000) for Sc and C. Sneden (private communication, 2006) for Eu.
188: While the 6774~\AA\ La II line may also be sensitive to hyper--fine splitting,
189: no known linelist exists in the literature for this transition. However, the
190: typically small equivalent widths of this line ($\la$30 m\AA) suggest
191: additional broadening will not affect La abundances too severely.
192:
193: \section{RESULTS AND DISCUSSION}
194:
195: \subsection{Al Abundances}
196:
197: We have determined at least upper limits of [Al/Fe] for five giants with the
198: cluster having $\langle$[Al/Fe]$\rangle$=+0.33 ($\sigma$=0.19) and a full
199: range of 0.50 dex. Both the star--to--star dispersion and average
200: [Al/Fe] ratios are in agreement with observations of other Galactic globular
201: clusters of similar metallicity (e.g., Kraft et al. 1998; Sneden et al. 2004;
202: Cohen \& Mel{\'e}ndez 2005; Johnson et al. 2005; Yong et al. 2005); however,
203: the highest [Al/Fe] ratio found in our sample is about a factor of three
204: smaller than the $>$+1.0 dex ratios observed in M3 and M13 (Pilachowski et al.
205: 1996; Sneden et al. 2004; Johnson et al. 2005; Cohen \& Mel{\'e}ndez 2005),
206: which possess similar metallicity and, in the case of M13, a similar HB
207: morphology. This may be due to our small sample size coupled with
208: observations of stars well below the RGB tip, where additional Al enhancement
209: due to extra in situ mixing may be operating (e.g., Denissenkov \& VandenBerg
210: 2003). Kraft et al. (1995) found M10 to be an intermediate case between M3
211: and M13 with regard to the amount of O depletion and Na enhancement and
212: therefore given the likely Na--Al correlation present in this cluster one
213: would not expect [Al/Fe] values much greater than about +0.80 dex. A complete
214: list of our determined abundances for Al and all other elements is provided in
215: Table 3.
216:
217: It has been shown that [Fe/H] determinations based on Fe I lines in metal--poor
218: stars suffer from larger LTE departure effects than their metal--rich
219: counterparts because of overionization due to decreased UV line blocking
220: (e.g., see review by Asplund 2005). Correcting for this effect would drive
221: the [Fe/H] abundance up, perhaps by as much as $\sim$+0.30 dex at [Fe/H]=--3
222: (Th{\'e}venin \& Idiart 1999, but see also Gratton et al. 1999; Kraft \& Ivans
223: 2003), and thus decrease the derived [Al/Fe] ratio found here. While a few
224: NLTE studies for Al exist (e.g., Gehren et al. 2004; Andrievsky et al. 2008)
225: finding offsets of order a few tenths of a dex, the actual Al NLTE correction
226: for stars in the metallicity and luminosity regime studied here are mostly
227: unknown. Fortunately, our sample does not vary widely in either metallicity or
228: luminosity and any NLTE corrections are likely to be very similar, suggesting
229: at least the relative star--to--star dispersion is a real effect.
230:
231: In Figure \ref{f2} we compare abundances of various elements in M10 versus
232: those in the similar cluster M12. The [Al/Fe] abundances for both clusters
233: are comparable and each displays a modest star--to--star dispersion. Given
234: that the scatter is about a factor of two larger than those observed in
235: the Fe--peak and $\alpha$ elements, it is likely that the Al distribution is
236: real and not an artifact of observational uncertainty. To see how M10 fits
237: into the context of other Galactic globular clusters, we have plotted [Al/Fe]
238: as a function of both horizontal branch ratio (HBR) and galactocentric
239: distance (R$_{\rm GC}$) for M10 and seven other clusters in Figure \ref{f3}.
240: The top panel suggests there is no significant relation between HBR and
241: either the average [Al/Fe] ratio or the star--to--star dispersion. However,
242: it should be noted that Carretta et al. (2007) do find a relationship between
243: the extent of O/Mg depletions and Na/Al enhancements and the maximum
244: temperature of stars located on the zero--age HB. The bottom panel may
245: indicate a trend of increasing cluster average [Al/Fe] with increasing
246: galactocentric distance; however, the sample size for each cluster varies
247: between less than 10 to nearly 100 stars. Consequently, M10 does not appear
248: to exhibit anomalous [Al/Fe] ratios compared to other globular clusters.
249:
250: \subsection{$\alpha$, Fe--Peak, and Heavy Elements}
251:
252: Nearly all globular clusters with [Fe/H]$<$--1 have [$\alpha$/Fe]$\sim$+0.30 to
253: +0.50, solar Fe--peak to Fe ratios, and are r--process rich (e.g., Gratton et
254: al. 2004). The star--to--star scatter present is usually $\la$0.10 dex for
255: the $\alpha$ and Fe--peak elements and $\sim$0.30--0.50 dex for the neutron
256: capture elements, which is still significantly less than the 0.50--1.00 dex
257: variations seen in light elements such as O, Na, and Al. In M10 we find
258: the expected enhancement and small star--to--star dispersion of the two
259: $\alpha$ elements Ca and Ti with $\langle$[Ca/Fe]$\rangle$=+0.42
260: ($\sigma$=0.12) and $\langle$[Ti/Fe]$\rangle$=+0.24 ($\sigma$=0.06). These
261: values are consistent with the results from Kraft et al. (1995) that
262: found $\langle$[Ca/Fe]$\rangle$=+0.29 ($\sigma$=0.07) and
263: $\langle$[Ti/Fe]$\rangle$=+0.21 ($\sigma$=0.12) for a set of 10 other upper
264: RGB stars in this cluster. The proxy Fe--peak elements Sc and Ni exhibit
265: near solar abundance ratios in all stars with cluster average values of
266: $\langle$[Sc/Fe]$\rangle$=+0.03 ($\sigma$=0.19) and
267: $\langle$[Ni/Fe]$\rangle$=+0.09 ($\sigma$=0.06), which are roughly consistent
268: with Kraft et al. (1995). These abundances patterns are mirrored in M12
269: (see Figure \ref{f2}), but with M10 showing a smaller range of [Cr/Fe] and
270: [Co/Fe] abundances. The combination of $\alpha$ enhancement and near solar
271: Fe--peak ratios is consistent with this cluster being primarily polluted by
272: the ejecta of type II SNe (e.g., Woosley \& Weaver 1995).
273:
274: For stars near M10's metallicity, La is produced primarily via the s--process
275: in $\sim$1--3 M$_{\odot}$ stars and Eu from the r--process in $\sim$8--10
276: M$_{\odot}$ stars (e.g., Busso et al. 1999; Truran et al. 2002). Our derived
277: La and Eu abundances are consistent with the picture of massive stars
278: producing most of the heavy elements in this cluster with
279: $\langle$[La/Fe]$\rangle$=+0.08 ($\sigma$=0.29) and
280: $\langle$[Eu/Fe]$\rangle$=+0.54 ($\sigma$=0.10). Comparing the ratio of
281: r-- to s--process elements gives [Eu/La]=+0.41 and implies M10 is slightly
282: more r--process rich than the average globular cluster. However, this
283: value is within the 1$\sigma$ range of $\langle$[Eu/Ba,La]$\rangle$=+0.23
284: ($\sigma$=0.21) found by Gratton et al. (2004) after combining data from the
285: literature on 28 globular clusters. A larger sample size of M10 stars is
286: likely to decrease the star--to--star scatter observed in our La and Eu
287: sample but will probably not change the result that the cluster is r--process
288: rich.
289:
290: \section{SUMMARY}
291:
292: We have determined abundances of the light element Al as well as several
293: $\alpha$, Fe--peak, and heavy elements in five M10 red giants using moderate
294: resolution spectroscopy (R$\sim$15,000) obtained with the Hydra multifiber
295: spectrograph on the WIYN telescope. The data sample the upper RGB with
296: luminosities ranging from above the level of the HB to about 0.5 mag below
297: the RGB tip. Model atmosphere parameters were determined via spectroscopic
298: procedures relying on abundances from equivalent width analyses.
299:
300: Our results are in agreement with previous studies that M10 is metal--poor
301: with [Fe/H]=--1.45 and has a very small metallicity spread ($\sigma$=0.04).
302: Al abundances indicate that while cluster stars maintain supersolar [Al/Fe]
303: values, there is a paucity of high--Al stars (i.e., [Al/Fe]$\ga$+1.0). This
304: result corroborates the O and Na data from Kraft et al. (1995) who found no
305: stars with [O/Fe]$<$--0.6, despite the cluster's similarity to M13 which has
306: several super O--poor/high--Al RGB stars. The modest average Al enhancement
307: of [Al/Fe]=+0.33 may be a consequence of its galactocentric distance of
308: $\sim$5 Kpc because comparison with several other similar metallicity globular
309: clusters at different R$_{\rm GC}$ shows a possible trend of increasing
310: $\langle$[Al/Fe]$\rangle$ with increasing R$_{\rm GC}$.
311:
312: We find all stars to have enhancements in [Ca/Fe] and [Ti/Fe] by about +0.20
313: to +0.40 dex and [el/Fe]$\sim$0 for Fe--peak elements. These data suggest the
314: current generation of M10 stars were heavily polluted with the by--products of
315: type II SNe without significant type Ia contributions, which would result in
316: lower [$\alpha$/Fe] ratios. The neutron capture elements also suggest
317: enrichment from massive stars because the cluster appears to be quite
318: r--process rich with $\langle$[Eu/La]$\rangle$=+0.41.
319:
320: \acknowledgements
321:
322: We are grateful to Diane Harmer for obtaining all observations used in this
323: paper. We would like to thank the NSF REU program for supporting SH via
324: AST--0453437. Support of the College of Arts and Sciences at Indiana
325: University Bloomington for CIJ is gratefully acknowledged.
326:
327: \begin{thebibliography}{}
328:
329: \bibitem[Andrievsky et al.(2008)]{2008A&A...481..481A} Andrievsky, S.~M.,
330: Spite, M., Korotin, S.~A., Spite, F., Bonifacio, P., Cayrel, R., Hill, V., \&
331: Fran{\c c}ois, P.\ 2008, \aap, 481, 481
332:
333: \bibitem[Arp(1955)]{1955AJ.....60..317A} Arp, H.~C.\ 1955, \aj, 60, 317
334:
335: \bibitem[Asplund(2005)]{2005ARA&A..43..481A} Asplund, M.\ 2005, \araa, 43, 481
336:
337: \bibitem[Busso et al.(1999)]{1999ARA&A..37..239B} Busso, M., Gallino, R., \&
338: Wasserburg, G.~J.\ 1999, \araa, 37, 239
339:
340: \bibitem[Carretta et al.(2007)]{2007ApJ...671L.125C} Carretta, E.,
341: Recio-Blanco, A., Gratton, R.~G., Piotto, G., \& Bragaglia, A.\ 2007, \apjl,
342: 671, L125
343:
344: \bibitem[Carretta(2008)]{2008MmSAI..79..508C} Carretta, E.\ 2008, Memorie
345: della Societa Astronomica Italiana, 79, 508
346:
347: \bibitem[Castelli et al.(1997)]{1997A&A...318..841C} Castelli, F.,
348: Gratton, R.~G., \& Kurucz, R.~L.\ 1997, \aap, 318, 841
349:
350: \bibitem[Chen et al.(2000)]{2000A&AS..145..223C} Chen, L., Geffert, M., Wang,
351: J.~J., Reif, K., \& Braun, J.~M.\ 2000, \aaps, 145, 223
352:
353: \bibitem[Cohen \& Mel{\'e}ndez(2005)]{2005AJ....129..303C} Cohen, J.~G., \& Mel{\'e}ndez, J.\ 2005, \aj, 129, 303
354:
355: \bibitem[Denissenkov \& VandenBerg(2003)]{2003ApJ...593..509D} Denissenkov, P.~A., \& VandenBerg, D.~A.\ 2003, \apj, 593, 509
356:
357: \bibitem[Dinescu et al.(1999)]{1999AJ....117.1792D} Dinescu, D.~I., Girard,
358: T.~M., \& van Altena, W.~F.\ 1999, \aj, 117, 1792
359:
360: \bibitem[Fulbright(2000)]{2000AJ....120.1841F} Fulbright, J.~P.\ 2000, \aj,
361: 120, 1841
362:
363: \bibitem[Gehren et al.(2004)]{2004A&A...413.1045G} Gehren, T., Liang, Y.~C.,
364: Shi, J.~R., Zhang, H.~W., \& Zhao, G.\ 2004, \aap, 413, 1045
365:
366: \bibitem[Gratton et al.(1999)]{1999A&A...350..955G} Gratton, R.~G., Carretta,
367: E., Eriksson, K., \& Gustafsson, B.\ 1999, \aap, 350, 955
368:
369: \bibitem[Gratton et al.(2004)]{2004ARA&A..42..385G} Gratton, R., Sneden, C.,
370: \& Carretta, E.\ 2004, \araa, 42, 385
371:
372: \bibitem[Iben(1964)]{1964ApJ...140.1631I} Iben, I.~J.\ 1964, \apj, 140,
373: 1631
374:
375: \bibitem[Ivans et al.(1999)]{1999AJ....118.1273I} Ivans, I.~I., Sneden, C.,
376: Kraft, R.~P., Suntzeff, N.~B., Smith, V.~V., Langer, G.~E.,
377: \& Fulbright, J.~P.\ 1999, \aj, 118, 1273
378:
379: \bibitem[Ivans et al.(2001)]{2001AJ....122.1438I} Ivans, I.~I., Kraft,
380: R.~P., Sneden, C., Smith, G.~H., Rich, R.~M.,
381: \& Shetrone, M.\ 2001, \aj, 122, 1438
382:
383: \bibitem[Johnson et al.(2005)]{2005PASP..117.1308J} Johnson, C.~I., Kraft,
384: R.~P., Pilachowski, C.~A., Sneden, C., Ivans, I.~I.,
385: \& Benman, G.\ 2005, \pasp, 117, 1308
386:
387: \bibitem[Johnson \& Pilachowski(2006)]{2006AJ....132.2346J} Johnson, C.~I., \&
388: Pilachowski, C.~A.\ 2006, \aj, 132, 2346
389:
390: \bibitem[Kraft(1994)]{1994PASP..106..553K} Kraft, R.~P.\ 1994, \pasp, 106,
391: 553
392:
393: \bibitem[Kraft et al.(1995)]{1995AJ....109.2586K} Kraft, R.~P., Sneden, C.,
394: Langer, G.~E., Shetrone, M.~D., \& Bolte, M.\ 1995, \aj, 109, 2586
395:
396: \bibitem[Kraft et al.(1998)]{1998AJ....115.1500K} Kraft, R.~P., Sneden, C.,
397: Smith, G.~H., Shetrone, M.~D., \& Fulbright, J.\ 1998, \aj, 115, 1500
398:
399: \bibitem[Kraft \& Ivans(2003)]{2003PASP..115..143K} Kraft, R.~P., \& Ivans, I.~I.\ 2003, \pasp, 115, 143
400:
401: \bibitem[Laird et al.(1988)]{1988AJ.....96.1908L} Laird, J.~B., Carney,
402: B.~W., Rupen, M.~P., \& Latham, D.~W.\ 1988, \aj, 96, 1908
403:
404: \bibitem[Mishenina et al.(2003)]{2003ARep...47..248M} Mishenina, T.~V.,
405: Panchuk, V.~E., \& Samus', N.~N.\ 2003, Astronomy Reports, 47, 248
406:
407: \bibitem[Pilachowski et al.(1996)]{1996AJ....112..545P} Pilachowski, C.~A.,
408: Sneden, C., Kraft, R.~P., \& Langer, G.~E.\ 1996, \aj, 112, 545
409:
410: \bibitem[Prochaska \& McWilliam(2000)]{2000ApJ...537L..57P} Prochaska,
411: J.~X., \& McWilliam, A.\ 2000, \apjl, 537, L57
412:
413: \bibitem[Ryan et al.(1996)]{1996ApJ...471..254R} Ryan, S.~G., Norris,
414: J.~E., \& Beers, T.~C.\ 1996, \apj, 471, 254
415:
416: \bibitem[Smith \& Fulbright(1997)]{1997PASP..109.1246S} Smith, G., \&
417: Fulbright, J.\ 1997, \pasp, 109, 1246
418:
419: \bibitem[Smith et al.(2005)]{2005AJ....129.1589S} Smith, G.~H., Briley,
420: M.~M., \& Harbeck, D.\ 2005, \aj, 129, 1589
421:
422: \bibitem[Sneden(1973)]{1973ApJ...184..839S} Sneden, C.\ 1973, \apj, 184,
423: 839
424:
425: \bibitem[Sneden et al.(1997)]{1997AJ....114.1964S} Sneden, C., Kraft,
426: R.~P., Shetrone, M.~D., Smith, G.~H., Langer, G.~E.,
427: \& Prosser, C.~F.\ 1997, \aj, 114, 1964
428:
429: \bibitem[Sneden et al.(2004)]{2004AJ....127.2162S} Sneden, C., Kraft,
430: R.~P., Guhathakurta, P., Peterson, R.~C.,
431: \& Fulbright, J.~P.\ 2004, \aj, 127, 2162
432:
433: \bibitem[Th{\'e}venin \& Idiart(1999)]{1999ApJ...521..753T} Th{\'e}venin, F.,
434: \& Idiart, T.~P.\ 1999, \apj, 521, 753
435:
436: \bibitem[Truran et al.(2002)]{2002PASP..114.1293T} Truran, J.~W., Cowan,
437: J.~J., Pilachowski, C.~A., \& Sneden, C.\ 2002, \pasp, 114, 1293
438:
439: \bibitem[von Braun et al.(2002)]{2002AJ....124.2067V} von Braun, K., Mateo,
440: M., Chiboucas, K., Athey, A., \& Hurley-Keller, D.\ 2002, \aj, 124, 2067
441:
442: \bibitem[Woosley \& Weaver(1995)]{1995ApJS..101..181W} Woosley, S.~E., \&
443: Weaver, T.~A.\ 1995, \apjs, 101, 181
444:
445: \bibitem[Yong et al.(2005)]{2005A&A...438..875Y} Yong, D., Grundahl, F.,
446: Nissen, P.~E., Jensen, H.~R., \& Lambert, D.~L.\ 2005, \aap, 438, 875
447:
448: \end{thebibliography}
449:
450: \newpage
451: \begin{figure}
452: \epsscale{.8}
453: \plotone{f1.eps}
454: \caption{Spectra from all five program stars are shown above with a pair of
455: Fe I and Al I lines indicated for reference. The spectra have been offset
456: for display purposes.}
457: \label{f1}
458: \end{figure}
459:
460: \newpage
461: \begin{figure}
462: \epsscale{1.00}
463: \plotone{f2.eps}
464: \caption{Average abundances for various elements in M10 are shown as filled
465: circles with error bars representing the 1$\sigma$ values. Similar data for
466: M12 from Johnson \& Pilachowski (2006) are shown as shaded boxes overlapping
467: the M10 results.}
468: \label{f2}
469: \end{figure}
470:
471: \newpage
472: \begin{figure}
473: \epsscale{.8}
474: \plotone{f3.eps}
475: \caption{The top panel shows [Al/Fe] versus the horizontal branch parameter,
476: where a value of +1.0 means only BHB stars, 0.0 is a mix of RHB, BHB, and RR
477: Lyrae stars, and --1.0 means only RHB stars. The various symbols indicate the
478: average [Al/Fe] abundance in a cluster with the error bars showing the
479: 1$\sigma$ values. The bottom panel shows [Al/Fe] versus the Galactocentric
480: distance for each cluster. The cluster data are from: M12 (Johnson \&
481: Pilachowski 2006), M3 and M13 (Johnson et al. 2005), NGC 7006 (Kraft et al.
482: 1998), M4 (Ivans et al. 1999), M5 (Ivans et al. 2001), and M15 (Sneden et al.
483: 1997).}
484: \label{f3}
485: \end{figure}
486:
487: \clearpage
488: \input{tab1.tex}
489: \clearpage
490: \input{tab2.tex}
491: \clearpage
492: \pagestyle{empty}
493: \input{tab3.tex}
494:
495: \end{document}
496: