astro-ph0408197/ms.tex
1: %
2: % LaTeX sample file for
3: % Publications of the Astronomical Society of Australia.
4: % Version 4.0 - November 2003
5: %
6: %
7: % Other information on PASA can be found at:
8: % http://www.publish.csiro.au/journals/pasa/
9: %
10: % General instruction for PASA authors can be found at:
11: % http://www.publish.csiro.au/nid/141/aid/442.htm
12: %
13: % \label and \ref cross-referencing is encouraged for referencing
14: % figures and sections.
15: %
16: % Natbib may be used for citations.
17: %
18: \documentclass[useAMS,usenatbib,usegraphicx]{pasa}
19: 
20: \title[The Canis Major Dwarf Galaxy]{
21: The Canis Major Dwarf Galaxy}
22: 
23: \author[Lewis et al.]{
24: Geraint F. Lewis$^{1,5}$,
25: Rodrigo A. Ibata$^2$,
26: Michael J. Irwin$^3$, \\
27: Nicolas F. Martin$^2$,
28: Michele Bellazzini$^4$ \&
29: Blair Conn$^1$
30: \affil{
31: $^1$Institute of Astronomy, School of Physics, A29, 
32: University of Sydney, NSW 2006, Australia \\
33: $^2$Observatoire de Strasbourg, 11 Rue de l'Universite, 
34: F-6700 Strasbourg, France\\
35: $^3$Institute of Astronomy, Madingley Road, Cambridge, CB3 0HA, U.K.\\
36: $^4$INAF - Osservatorio Astronomico di Bologna, Via Ranzani 1, 40127, Bologna, Italy\\
37: $^5$Email: {\tt gfl@physics.usyd.edu.au}
38: }}
39: \begin{document}
40: 
41: \maketitle
42: 
43: \label{firstpage}
44: 
45: \begin{abstract}
46: Recent  observational  evidence suggests  that  the Sagittarius  dwarf
47: galaxy  represents  the only  major  ongoing  accretion  event in  the
48: Galactic  halo,   accounting  for  the  majority   of  stellar  debris
49: identified  there.   This paper  summarizes  the  recent discovery  of
50: another  potential Milky Way  accretion event,  the Canis  Major dwarf
51: galaxy.  This  dwarf satellite galaxy is  found to lie  just below the
52: Galactic  plane and  appears to  be  on an  equatorial orbit.   Unlike
53: Sagittarius, which is contributing  to the Galactic halo, the location
54: and  eventual demise  of Canis  Major  suggests that  it represents  a
55: building block of the thick disk.
56: \end{abstract}
57: 
58: \begin{keywords}
59: Galaxy: structure -- Galaxy: evolution -- galaxies: dwarf 
60: \end{keywords}
61: 
62: \section{Introduction}\label{introduction}
63: ${\rm \Lambda}$CDM  represents the current paradigm  for the formation
64: and evolution of  structure in the Universe. While  successul on large
65: scales,  focus  has recently  turned  to  its  inability to  correctly
66: predict   the  number   of   satellite  systems   in  galactic   halos
67: \citep{1999ApJ...522...82K}. This  {\it missing satellite  problem} is
68: apparent with the Milky  Way, with recent observations indicating that
69: the Milky Way has undergone  a single large accretion, the Sagittarius
70: Dwarf            galaxy,             in            the            last
71: $\sim7$Gyrs~\citep{2002MNRAS.332..921I,2003ApJ...599.1082M},   although
72: some        older       accretions        are        apparent       in
73: phase-space~\citep{1999Natur.402...53H,2003ApJ...585L.125B}. 
74: 
75: \begin{figure*}
76: \includegraphics[height=4.4in]{CanisMajorfig05.ps}
77: \caption{\label{fig1} Binned M-giant  counts about the Galactic plane.
78: The  three  panels denote  different  magnitude  (and hence  distance)
79: ranges,  with the  galactocentric distance  modulus labeled  upon each
80: panel. Also in each panel,  the counts represent the asymmetry between
81: the north and south (ie -  southern data subtracted from the north and
82: vice-versa). The main body of  Canis Major and prominent structures in
83: the    M-giant    distribution    are    noted    (Figure    5    from
84: \citet{2004MNRAS.348...12M}).}
85: \end{figure*}
86: 
87: To  compare  with  predictions  from ${\rm  \Lambda}$CDM  models,  the
88: accretion history  of the Milky Way  needs to be mapped  out.  As with
89: many studies of the Universe,  however, our view of the Galactic halo,
90: and any accreting  systems, is obscured by the disk  of the Milky Way.
91: Recent large scale  surveys of this region, however,  have uncovered a
92: curious overdensity of  stars that has been interpreted  as being part
93: of a current accretion event that  is taking place within the plane of
94: the Milky Way. This paper reviews the discovery of this object and our
95: current understanding of its properties.
96: 
97: \section{The ring around the Galaxy}\label{ring}
98: The first  indication of  an additional halo  population of  stars was
99: found by ~\citet{2002ApJ...569..245N} while examining halo stars drawn
100: from  the  Sloan  Digital  Sky  Survey.  Taken in  a  narrow  band  of
101: $\sim2.5^o$  about the  celestial equator,  and selected  to  have the
102: colours  of F-stars, these  data revealed  a prominent  overdensity of
103: halo stars  which has been  interpreted as the survey  slicing through
104: the stream  of tidal  debris from the  Sagittarius Dwarf  Galaxy.  One
105: additional prominent  overdensity of  stars was identified  within the
106: halo, towards the Galactic anti-centre, in the direction of Monoceros,
107: at a galactocentric distance  of $\sim18$kpc and width $<6$kpc.  While
108: a   Galactic   origin   could   not   be   conclusively   ruled   out,
109: \citet{2002ApJ...569..245N}  suggested  that  this too  represented  a
110: tidal stream, but of a yet unknown disrupting companion galaxy,
111: 
112: Spurred  by this  discovery, \citet{2003MNRAS.340L..21I}  searched for
113: the signature  of the  Monoceros stream of  stars in the  Isaac Newton
114: Telescope   Wide    Field   Survey   data    archive.    Analysis   of
115: colour-magnitude  diagrams, \citet{2003MNRAS.340L..21I}  confirmed the
116: identification of  the distinct stellar population in  the vicinity of
117: Monoceros. This study also  identified the Monoceros stream population
118: in a number of additional  fields, revealing this population to extend
119: $\sim100^o$ over  the sky, within $\sim30^o$ of  the Galactic equator,
120: and  \citet{2003MNRAS.340L..21I}  suggest  that  the  Monoceros  stream
121: actually rings the  Galaxy. Fitting the main sequence  of this stellar
122: population in  each field, this  study estimates that the  distance to
123: the  stream ranges  from  $\sim15$ to  $\sim20$kpc,  with an  apparent
124: scale-height  of $\sim0.75$kpc.  While  they consider  the possibility
125: that this stellar population was an accreting dwarf, they also pointed
126: out that their  data is consistent with other  hypotheses including an
127: outer spiral  arm or unknown  flare/warp generated via a  resonance in
128: the Galactic disk.
129: 
130: At   the  same  time   as  the   \citet{2003MNRAS.340L..21I}  results,
131: \citet{2003ApJ...588..824Y}  presented  a   kinematic  analysis  of  a
132: Galactic  halo  stars  drawn  from  the SDSS.   Focusing  upon  F-star
133: candidates, this  study obtained spectra in  several regions, allowing
134: the determination of their  kinemtatic properties.  Accounting for the
135: Galactic contribution,  the Monoceros population  was found to  have a
136: velocity dispersion of $\sim25-30$km/s.  While of similar order to the
137: tidal debris torn from the Sagittarius dwarf, this velocity dispersion
138: is quite distinct  from the spheroid, thick disk or  any known warp or
139: flare. From these  velocities, \citet{2003ApJ...588..824Y} deduce that
140: the orbital velocity of the  stream of stars is prograde and (assuming
141: circular orbits) is $215\pm25$km/s, [An erratum to the original result
142: of $110\pm25$kms/s was  presented in \citet{2004ApJ...605..575Y}], and
143: that    the    stars    appear    to    be    relatively    metal-poor
144: $([\frac{Fe}{H}]=-1.6)$.    \citet{2003ApJ...588..824Y}   conclude  by
145: proposing  a simple  model for  the Monoceros  stream as  a disrupting
146: dwarf, orbiting the Milky Way  at a distance of $\sim18$kpc; the dwarf
147: galaxy's stars are  undulating above and below the  plane of the Milky
148: Way by $\sim6$kpc.
149: 
150: Several   additional  programs   have  focused   upon   the  Monoceros
151: stream~\footnote{The stream of stars  has acquired several names since
152: its  discovery. For  the  sake of  consistency,  in this  paper it  is
153: referred  to solely  as the  Monoceros  Stream.}. Using  M giant  star
154: candidates  drawn   from  2MASS,  \citet{2003ApJ...594L.115R}  further
155: confirmed the existence of the  Monoceros Stream as a distinct stellar
156: population  beyond the  edge of  the  disk of  the Milky  Way. With  a
157: galactocentric distance of $18\pm2$kpc,  they find the arc of material
158: possesses an angular extent of  at $\sim170^o$, with the presence of M
159: giants  indicating  the  stellar  population of  the  Monoceros  stars
160: possesses   a    higher   metallicity   than    previously   estimated
161: $([\frac{Fe}{H}]\sim0.4\pm0.3)$.  \citet{2003ApJ...594L.119C} extended
162: this  work,   obtaining  velocities  of  2MASS   selected  stars  over
163: $\sim100^o$  of the  Monoceros  stream.  While  confirming a  velocity
164: dispersion of $\sim20$km/s, these data  also indicate the stars in the
165: Monoceros stream orbit  the Galaxy in a prograde  fashion, with little
166: eccentricity.   While  this  is  somewhat worrisome  for  the  tidally
167: disrupting dwarf  galaxy hypothesis, \citet{2003ApJ...594L.119C} point
168: out  that  such features  are  apparent  in  numerical simulations  of
169: in-plane   dwarf  accretion\citep{2003ApJ...592L..25H}.   Furthermore,
170: this study  identifies four globular  clusters that are  spatially and
171: kinematically  aligned  with  the  putative stellar  stream;  such  an
172: alignment argues against  a Galactic origin for the  stream, such as a
173: spiral  arm.   Finally,  \citet{2004ApJ...602L..21F} noted  that  five
174: globular  clusters  aligned with  the  Monoceros  stream,  as well  as
175: $\sim15$ outer, old stellar clusters that may also be part of the same
176: population;  these clusters  lie  in a  plane  which is  significantly
177: tilted  $(\sim17^o)$ to  that of  the  Milky Way.   The population  of
178: globular clusters is reminiscent of the Sagittarius Dwarf galaxy which
179: has appears  to have deposited  a similar number of  globular clusters
180: into  the halo  of the  Milky Way,  bolstering the  argument  that the
181: Monoceros stream represets a similarly disrupting dwarf galaxy.
182: 
183: The  extensive nature  of  the material  strongly  suggested that  its
184: origin lay  in a  disrupting system,  in an event  similar to  that of
185: Sagittarius, but occurring in the  plane of the Galaxy. This, however,
186: was not  the only  explanation for  its origin as  it may  represent a
187: previously  unidentified   aspect  of  Galactic   structure.   If  the
188: disrupting system  hypothesis is correct,  and its destruction  is not
189: yet  complete,  we should  expect  to  identify  some remnant  of  the
190: original  dwarf  galaxy.   However,  the expected  location,  observed
191: against the plane of the Galaxy, makes the detection of such a remnant
192: difficult.
193: 
194: \section{The Canis Major Dwarf}\label{cmdwarf}
195: As   with   \citet{2003ApJ...594L.115R},   \citet{2004MNRAS.348...12M}
196: employed 2MASS to search for a signature of Monoceros stream of stars.
197: Pushing the search to $|b|\sim5^o$, \citet{2004MNRAS.348...12M} mapped
198: the density of M-giant stars around the Galactic equator. This reveals
199: a strong  asymmetry about the  Galactic plane which is  interpreted as
200: being the Monoceros ring of  stars snaking around the outskirts of the
201: Milky Way.   Several prominent features are noted,  including a strong
202: Northern arc  and weaker Southern arc,  both of which  extend for more
203: than $\sim100^o$ on the sky (see Figure~\ref{fig1}).
204: 
205: Intriguingly,  \citet{2004MNRAS.348...12M} also  identified  a strong,
206: elliptical  overdensity of  M-stars  at $(l,b)=(240^o,-8^o)$,  aligned
207: somewhat with  the Galactic disk. The (heliocentric)  distance to this
208: stellar overdensity is $D_\odot =  7.1\pm0.1$kpc, with a major axis is
209: $\sim4.2$kpc.   With $\sim2300$  M-giant  stars within  $10^o$ of  its
210: centre, this overdensity contains a  similar number of M-giants to the
211: Sagittarius Dwarf  galaxy, a system  which we know is  currently being
212: cannibalised by the Milky Way.   Given that this implies that the mass
213: of  the  Canis  Major  overdensity  is  $\sim10^8-10^9{\rm  M_\odot}$,
214: \citet{2004MNRAS.348...12M} concluded  that it too  represents a dwarf
215: galaxy also undergoing tidal disruption, and possibly representing the
216: progenitor of the Monoceros stream of stars.
217: 
218: Unlike Sagittarius, which passes over  the poles of the Milky Way, the
219: identification of  the Canis Major  dwarf galaxy (as  this overdensity
220: will  now  be  referred  to)  represents the  first  detection  of  an
221: accretion occurring within the plane of the Galaxy.
222: 
223: \citet{2004MNRAS.348...12M}  also noted  that  possibly five  globular
224: clusters  were associated in  phase-space with  the Canis  Major dwarf
225: galaxy.  This is  a similar number to the  globular cluster population
226: currently  associated  with the  Sagittarius  dwarf galaxy,  providing
227: further  evidence  for  the   origin  of  Canis  Major.   Furthermore,
228: \citet{2004MNRAS.348...12M} noted the phase-space grouping of a number
229: of four open galactic clusters  directly associated with the main body
230: of Canis  Major\footnote{Several of these clusters were  also noted by
231: \citet{2004ApJ...602L..21F} as being part of the Monoceros stream.}.
232: 
233: In  an attempt  to understand  the observed  distribution  of M-stars,
234: \citet{2004MNRAS.348...12M}  also  undertook  a  series  of  numerical
235: simulations. Utilizing  detailed models  for the mass  distribution of
236: the Milky Way  \citep{1998MNRAS.294..429D}, these simulations involved
237: following the  dynamical dissociation  of dwarf galaxies  (modeled as
238: King  profiles) as  they  orbited  the Galaxy.   The  results of  this
239: procedure   favoured  a  dwarf   galaxy  with   an  initial   mass  of
240: $\sim5\times10^8{\rm   M_\odot}$,    with   a   orbital    period   of
241: $\sim0.4$Gyrs. While  such simulations can reasonably  account for the
242: observed distribution of M-stars not  only in the body of Canis Major,
243: but also along the extensive  arcs above and below the Galactic plane,
244: the current  data does not allow a  definitive differentiation between
245: prograde and  retrograde orbits. \citet{2004MNRAS.348...12M}, however,
246: point out that the resultant orbits of the debris in such an encounter
247: closely  mimic those  of  stars in  the  thick disk.   Given that  the
248: estimated  mass of  this single  dwarf  is roughly  $\sim10\%$ of  the
249: entire thick disk, then the thick  disk could be formed via only a few
250: such accretion events.
251: 
252: Further observational evidence for the nature of the Canis Major dwarf
253: came from the study of \citet{aa} who identified the main-sequence and
254: red giant branch  populations of Canis Major in  the background to the
255: Galactic open  clusters NGC2477, Tombaugh 1 and  Berkeley 33. Analysis
256: of   this    population   suggested   it    is   somewhat   metal-rich
257: $(-0.7<[\frac{Fe}{H}]<0.0)$, with an age of $\sim2-7$Gyrs, although an
258: apparent blue  plume of stars  is taken as  evidence of a  more recent
259: episode  of  star formation.   This  study  also  finds a  photometric
260: parallax  for the  main  body of  Canis  Major of  $\sim8.3\pm1.2$kpc,
261: larger  than the M-giant  study of  \citep{2004MNRAS.348...12M}, whose
262: smaller distance determination they  put down to poorer systematics in
263: the  M-giant photometric  parallax. Finally,  \citet{aa}  also suggest
264: that as  well as the  identified globular cluster population,  two old
265: open clusters, AM-2  and Tombaugh 2, are possibly  associated with the
266: Canis Major dwarf.
267: 
268: Utilizing  the Second  U.S. Naval  Observatory CCD  Astrograph Catalog
269: (UCAC2),  \citet{bb}  examined the  proper  motions  of  stars in  the
270: vicinity of  the Canis Major dwarf  galaxy.  Cross-identifying M-giant
271: stars drawn  from the 2MASS catalog,  the Galactic-longitudinal motion
272: of  Canis Major  was  found to  be  $\mu_l=-4.0\pm0.4$mas/yr, with  no
273: measurable motion  in Galactic latitude. At a  distance of $D=8.3$kpc,
274: this  corresponds to transverse  velocity of  $\sim238\pm28$km/s, with
275: Canis Major on a prograde orbit  about the Milky Way. Since this value
276: is  higher than that  predicted by  \citet{2004MNRAS.348...12M}, drawn
277: from   their   numerical   simulation,    it   is   clear   the   full
278: three-dimensional  velocity of Canis  Major will  be required  for the
279: detailed numerical modeling of this accretion event.
280: 
281: Recently, \citet{cc} further  examined the globular cluster population
282: associated  with  the  Canis  Major  dwarf galaxy,  finding  that  the
283: age-metallicity  relationship  for these  is  distinct  from the  main
284: globular  cluster population  of  the Milky  Way.  Furthermore,  these
285: globular  clusters are  somewhat smaller  than expected  if  they were
286: drawn from the  Galactic population. Both these lines  of evidence add
287: further  weight to  the possibility  that the  globular  clusters, and
288: Canis Major itself, were formed elsewhere, were drawn in presumably as
289: the  initial orbit  decayed by  dynamical friction,  and  represent an
290: on-going accretion event onto the Milky Way galaxy.
291: 
292: \section{Further Observations}\label{further}
293: While the tidally disrupting  dwarf galaxy is the favoured explanation
294: for  the  observed  overdensity   of  star  in  Canis  Major,  current
295: observations are  not yet  completely conclusive, and  the possibility
296: that  the  Monoceros  Ring  and  Canis  major  dwarf  galaxy  actually
297: represent some unknown aspect of Galactic structure has not been ruled
298: out.  Hence,  several observational  programs are underway  to address
299: this issue.
300: 
301: The    first    is    an    extension    of    the    earlier    study
302: by~\citet{2003MNRAS.340L..21I}, systematically mapping above and below
303: the Galactic plane with  wide-field camera observations. This has been
304: completed  in the  north, using  the  Wide-Field Camera  on the  Isaac
305: Newton Telescope; these data are  being analyzed and will be published
306: shortly (Conn et al. {\it  in preparation}). The southern survey, with
307: the Wide-Field Imager at  the Anglo-Australian Telescope, is currently
308: underway  and  will be  completed  by  mid-2004.   These will  provide
309: important probes  of the  extent of the  stellar material, as  well as
310: constraining its distance (via main sequence fitting).
311: 
312: \begin{figure}
313: \includegraphics[height=4.7in]{CanisMajorfig14.ps}
314: \caption{\label{fig2}  A numerical  simulation  of the  demise of  the
315: Canis Major dwarf  galaxy. The top panel presents  the view from above
316: the Milky  Way [with  the Sun located  at $(x,y,z)=-8,0,0\  $kpc]. The
317: lower panel presents the side-on view of the debris. The colour coding
318: in both denoted the heliocentric velocity of the stars (as given in the
319: lower key), while  the symbols represent data from  the INT survey and
320: 2MASS [Figure 14 from \citet{2004MNRAS.348...12M}].}
321: \end{figure}
322: 
323: To augment these  studies, 2dF observations of the  main body of Canis
324: Major and the extensive stream  of stars are being conducted to obtain
325: stellar kinematics,  via the calcium triplet, over  $\sim100^o$ of the
326: sky.   Coupled with  the  spatial data  obtained  with the  wide-field
327: camera surveys, these data should test the hypotheses that Canis Major
328: represents a  truly disrupting dwarf galaxy,  or (equally interesting)
329: is a  currently unknown aspect of Galactic  structure. Furthermore, if
330: the dwarf galaxy is confirmed,  then these data will provide important
331: constraints to numerical simulations of its orbit and eventual demise.
332: 
333: \section*{Acknowlegements}
334: GFL  thanks Joss  Bland-Hawthorn for  inviting him  to the  Little Bay
335: Meeting.  The  anonymous referee  is  thanked  for their  constructive
336: comments.
337: 
338: \begin{thebibliography}{}
339: %
340: \bibitem[\protect\citeauthoryear{Bellazzini     et     al.}{2004}]{aa}
341: Bellazzini  M., Ibata  R., Monaco  L., Martin  N., Irwin  M.~J., Lewis
342: G.~F., 2004, MNRAS in press, {\it astro-ph/0311119}
343: %
344: \bibitem[\protect\citeauthoryear{Brook                               et
345: al.}{2003}]{2003ApJ...585L.125B} Brook C.~B., Kawata D., Gibson B.~K.,
346: Flynn C., 2003, ApJ, 585, L125
347: %
348: \bibitem[\protect\citeauthoryear{Crane et al.}{2003}]{2003ApJ...594L.119C} 
349: Crane J.~D., Majewski S.~R., Rocha-Pinto H.~J., Frinchaboy P.~M., Skrutskie 
350: M.~F., Law D.~R., 2003, ApJ, 594, L119 
351: %
352: \bibitem[\protect\citeauthoryear{Dehnen \& 
353: Binney}{1998}]{1998MNRAS.294..429D} Dehnen W., Binney J., 1998, MNRAS, 294, 
354: 429 
355: %
356: \bibitem[\protect\citeauthoryear{Forbes et al.}{2004}]{cc} 
357: Forbes, D.~A., Strader, J., Brodie, J.P., 2004, {\it astro-ph/0403136}
358: %
359: \bibitem[\protect\citeauthoryear{Frinchaboy et 
360: al.}{2004}]{2004ApJ...602L..21F} Frinchaboy P.~M., Majewski S.~R., Crane 
361: J.~D., Reid I.~N., Rocha-Pinto H.~J., Phelps R.~L., Patterson R.~J., Mu{\~ 
362: n}oz R.~R., 2004, ApJ, 602, L21 
363: %
364: \bibitem[\protect\citeauthoryear{Helmi et al.}{2003}]{2003ApJ...592L..25H} 
365: Helmi A., Navarro J.~F., Meza A., Steinmetz M., Eke V.~R., 2003, ApJ, 592, 
366: L25 
367: %
368: \bibitem[\protect\citeauthoryear{Helmi et al.}{1999}]{1999Natur.402...53H} 
369: Helmi A., White S.~D.~M., de Zeeuw P.~T., Zhao H., 1999, Natur, 402, 53 
370: %
371: \bibitem[\protect\citeauthoryear{Ibata et al.}{2003}]{2003MNRAS.340L..21I} 
372: Ibata R.~A., Irwin M.~J., Lewis G.~F., Ferguson A.~M.~N., Tanvir N., 2003, 
373: MNRAS, 340, L21 
374: %
375: \bibitem[\protect\citeauthoryear{Ibata et al.}{2004}]{bb} 
376: Ibata R., Bellazzini, M., Irwin M., Lewis G.~F., Martin, N.~F., 
377: 2004, MNRAS (Submitted)
378: %
379: \bibitem[\protect\citeauthoryear{Ibata et al.}{2002}]{2002MNRAS.332..921I} 
380: Ibata R.~A., Lewis G.~F., Irwin M.~J., Cambr{\' e}sy L., 2002, MNRAS, 332, 
381: 921 
382: %
383: \bibitem[\protect\citeauthoryear{Ibata et al.}{2001}]{2001ApJ...551..294I} 
384: Ibata R., Lewis G.~F., Irwin M., Totten E., Quinn T., 2001, ApJ, 551, 294 
385: %
386: \bibitem[\protect\citeauthoryear{Klypin et al.}{1999}]{1999ApJ...522...82K} 
387: Klypin A., Kravtsov A.~V., Valenzuela O., Prada F., 1999, ApJ, 522, 82 
388: %
389: \bibitem[\protect\citeauthoryear{Majewski et 
390: al.}{2003}]{2003ApJ...599.1082M} Majewski S.~R., Skrutskie M.~F., Weinberg 
391: M.~D., Ostheimer J.~C., 2003, ApJ, 599, 1082 
392: %
393: \bibitem[\protect\citeauthoryear{Martin et al.}{2004}]{2004MNRAS.348...12M} 
394: Martin N.~F., Ibata R.~A., Bellazzini M., Irwin M.~J., Lewis G.~F., Dehnen 
395: W., 2004, MNRAS, 348, 12 
396: %
397: \bibitem[\protect\citeauthoryear{Newberg et al.}{2002}]{2002ApJ...569..245N} 
398: Newberg H.~J.~et al., 2002, ApJ, 569, 245 
399: %
400: \bibitem[\protect\citeauthoryear{Rocha-Pinto et 
401: al.}{2003}]{2003ApJ...594L.115R} Rocha-Pinto H.~J., Majewski S.~R., 
402: Skrutskie M.~F., Crane J.~D., 2003, ApJ, 594, L115 
403: %
404: \bibitem[\protect\citeauthoryear{Yanny et al.}{2003}]{2003ApJ...588..824Y} 
405: Yanny B.~et al., 2003, ApJ, 588, 824 
406: %
407: \bibitem[Yanny et al.(2004)]{2004ApJ...605..575Y} Yanny, B., et al.\ 2004, 
408: ApJ, 605, 575 
409: %
410: \end{thebibliography}
411: 
412: 
413: 
414:  \label{lastpage}
415: 
416: \end{document}
417: