0801.2388/ms.tex
1: \documentclass[10pt,preprint]{emulateapj}
2: 
3: \newcommand{\hdf}{HDF-N}
4: \newcommand{\hdfn}{HDF-N}
5: \newcommand{\hdfs}{HDF-S}
6: \newcommand{\cdfs}{CDF-S}
7: \newcommand{\hst}{\textit{HST}}
8: \newcommand{\spitzer}{\textit{Spitzer}}
9: \newcommand{\sirtf}{\textit{Spitzer}}
10: \newcommand{\chandra}{\textit{Chandra}}
11: \newcommand{\iso}{\textit{ISO}}
12: \newcommand{\wfu}{\hbox{$U_{300}$}}
13: \newcommand{\wfb}{\hbox{$B_{450}$}}
14: \newcommand{\wfv}{\hbox{$V_{606}$}}
15: \newcommand{\wfi}{\hbox{$I_{814}$}}
16: \newcommand{\acsb}{\hbox{$B_{435}$}}
17: \newcommand{\acsv}{\hbox{$V_{606}$}}
18: \newcommand{\acsi}{\hbox{$i_{775}$}}
19: \newcommand{\acsz}{\hbox{$z_{850}$}}
20: \newcommand{\nicj}{\hbox{$J_{110}$}}
21: \newcommand{\nich}{\hbox{$H_{160}$}}
22: \newcommand{\mstar}{\hbox{$\mathcal{M}^\ast$}}
23: \newcommand{\ks}{\hbox{$K_s$}}
24: \newcommand{\AAA}{\hbox{\AA}}
25: \newcommand{\lya}{Lyman~$\alpha$}
26: \newcommand{\lyb}{Lyman~$\beta$}
27: \newcommand{\hb}{\hbox{H$\beta$}}
28: \newcommand{\ha}{\hbox{H$\alpha$}}
29: \newcommand{\lsim}{\lesssim}
30: \newcommand{\gsim}{\gtrsim}
31: \newcommand{\lstar}{\hbox{$L^\ast$}}
32: \newcommand{\mcal}{\hbox{$\mathcal{M}$}}
33: \newcommand{\rcal}{\hbox{$\mathcal{R}$}}
34: \newcommand{\scal}{\hbox{$\mathcal{S}$}}
35: \newcommand{\ebv}{\hbox{$E(B-V)$}}
36: \newcommand{\etal}{et al.}
37: \newcommand{\eg}{e.g.}
38: \newcommand{\ie}{i.e.}
39: \newcommand{\cf}{cf.}
40: \newcommand{\Msol}{\hbox{$\mathcal{M}_\odot$}}
41: \newcommand{\Zsol}{\hbox{$Z_\odot$}}
42: \newcommand{\msol}{\hbox{$\mathcal{M}_\odot$}}
43: \newcommand{\zsol}{\hbox{$Z_\odot$}}
44: \newcommand{\Lsol}{\hbox{$L_\odot$}}
45: \newcommand{\lsol}{\hbox{$L_\odot$}}
46: \newcommand{\fion}[2]{\hbox{[\ion#1#2]}}
47: \newcommand{\infinity}{\hbox{$\infty$}}
48: \newcommand{\snr}{\hbox{$\mathrm{S/N}$}}
49: \newcommand{\uit}{\textit{UIT}}
50: \newcommand{\degree}{\hbox{$^\circ$}}
51: \newcommand{\uJy}{\hbox{$\mu$Jy}}
52: \newcommand{\ujy}{\hbox{$\mu$Jy}}
53: \newcommand{\lir}{\hbox{$L_{\mathrm{IR}}$}}
54: \newcommand{\mone}{\hbox{$[3.6\mu\mathrm{m}]$}}
55: \newcommand{\mtwo}{\hbox{$[4.5\mu\mathrm{m}]$}}
56: \newcommand{\mthree}{\hbox{$[5.8\mu\mathrm{m}]$}}
57: \newcommand{\mfour}{\hbox{$[8.0\mu\mathrm{m}]$}}
58: \newcommand{\calL}{{\cal L}}
59: \newcommand \msun{\hbox{$\hbox{M}_{\odot}$}}
60: \newcommand{\todo}[1]{{\tt #1}}
61: 
62: 
63: \begin{document}
64: 
65: \title{ON THE STELLAR POPULATIONS IN FAINT RED GALAXIES IN THE
66: \textit{HUBBLE} ULTRA DEEP FIELD}
67: 
68: \author{\sc Amelia M. Stutz, 
69: Casey Papovich\altaffilmark{2}, 
70: Daniel J. Eisenstein} 
71: \affil{Department of Astronomy and Steward Observatory, 
72: University of Arizona, 933 North Cherry Avenue, Tucson, 
73: Arizona 85721 
74: \\ astutz, papovich, eisenstein@as.arizona.edu}
75: 
76: \altaffiltext{1}{This work is based in part on observations taken with
77:   the NASA/ESA Hubble Space Telescope, which is operated by the
78:   Association of Universities for Research in Astronomy, Inc.\ (AURA)
79:   under NASA contract NAS5--26555; and on observations made with the
80:   \textit{Spitzer Space Telescope}, which is operated by the Jet
81:   Propulsion laboratory, California Institute of Technology, under
82:   NASA contract 1407. }
83: \altaffiltext{2}{\spitzer\ Fellow}
84: 
85: 
86: \setcounter{footnote}{2}
87: 
88: \begin{abstract}
89: We study the nature of faint, red-selected galaxies at $z \sim 2-3$
90: using the \textit{Hubble} Ultra Deep Field (HUDF) and \spitzer\
91: Infrared Array Camera (IRAC) photometry.  Given the magnitude limit of
92: the \hst\ data, we detect candidate galaxies to
93: $H_\mathrm{AB}<26$~mag, probing lower-luminosity (lower mass) galaxies
94: at these redshifts.  We identify 32 galaxies satisfying the
95: $(\nicj-\nich)_\mathrm{AB} > 1.0$~mag color selection, 16 of which
96: have unblended \mone\ and \mtwo\ photometry from \spitzer.  Using this
97: multiwavelength dataset we derive photometric redshifts, masses, and
98: stellar population parameters for these objects.  We find that the
99: selected objects span a diverse range of properties over a large range
100: of redshifts, $1 \lsim z \lsim 3.5$.  A substantial fraction (11/32)
101: of the $(\nicj - \nich)_\mathrm{AB} > 1$~mag population appear to be
102: lower-redshift ($z \lsim 2.5$), heavily obscured dusty galaxies or
103: edge-on spiral galaxies, while others (12/32) appear to be galaxies at
104: $2 \lsim z \lsim 3.5$ whose light at rest-frame optical wavelengths is
105: dominated by evolved stellar populations.  We argue that longer
106: wavelength data ($\gsim 1$~\micron, rest-frame) are essential for
107: interpreting the properties of the stellar populations in red-selected
108: galaxies at these redshifts.  Interestingly, by including \spitzer\
109: data many candidates for galaxies dominated by evolved stellar
110: populations are rejected, and for only a subset of the sample (6/16)
111: do the data favor this interpretation.  These objects have a surface
112: density of $\sim1$~arcmin$^{-2}$.  We place an upper limit on the
113: space density of candidate massive evolved galaxies with $2.5 < z <
114: 3.5$ and $\nich(\mathrm{AB}) \leq 26$~mag of $n = 6.6^{+2.0}_{-3.0}
115: \times 10^{-4}$ Mpc$^{-3}$ with a corresponding upper limit on the
116: stellar mass density of $\rho^\ast = 5.6^{+4.4}_{-2.8}\times 10^{7}$
117: $M_\odot$ Mpc$^{-3}$.  The $z > 2.5$ objects that are dominated by
118: evolved stellar populations have a space density at most one-third
119: that of $z\sim 0$ red, early-type galaxies.  Therefore, at least
120: two-thirds of present-day early-type galaxies assemble or evolve into
121: their current configuration at redshifts below 2.5.  We find a dearth
122: of candidates for low-mass ($\lsim 2\times 10^{10}$~$\msun)$ galaxies
123: at $1.5 < z \lesssim 3$ that are dominated by passively evolving
124: stellar populations even though the data should be sensitive to them;
125: thus, at these redshifts, galaxies whose light is dominated by evolved
126: stellar populations are restricted to only those galaxies that have
127: assembled high stellar mass.
128: 
129: \end{abstract}
130: 
131: \keywords{cosmology: observations -- galaxies: evolution -- galaxies:
132: high-redshift -- galaxies: stellar content -- infrared: galaxies}
133: 
134: \section{Introduction}
135: 
136: In the hierarchical picture of galaxy evolution, galaxies assemble
137: within dark matter haloes that merge over time
138: \citep[\eg,][]{freedman01,sp03}.  This model has been highly
139: successful at reproducing galaxy clustering properties based on
140: well-understood gravitational physics \citep[\eg,][]{weinberg04}.
141: However, it is difficult to test how galaxies assemble their stellar
142: content within the context of these models because of uncertain
143: physics coupling the dark matter and baryonic matter in these haloes.
144: 
145: Observationally, the cosmic star-formation-rate (SFR) density rises by
146: roughly an order of magnitude from $z\sim 0$ to 1 and remains roughly
147: constant from $z\sim 1$-6 \citep[and references therein]{hopkins04}.
148: At the same time, roughly half of the stellar-mass density in galaxies
149: formed between $z\sim $3 and 1, corresponding to this peak in the SFR
150: density \citep[\eg,][]{dickinson03,rudnick03,glazebrook04}.  While
151: theoretical models broadly reproduce this evolution in the global
152: galaxy population \citep[\eg,][]{somerville01,hernquist03}, they have
153: difficulty in reproducing star-formation trends in individual
154: galaxies, particularly in accounting for the formation and evolution
155: of specific SFRs of the most massive objects
156: \citep{baugh03,somerville04,delucia06,croton06}.  In particular, it is
157: difficult for models to produce galaxies (at any redshift) that are
158: mostly devoid of current star-formation and evolve only passively
159: \citep[\eg,][]{croton06,dave06}.  Quantifying the number of passively
160: evolving galaxies --- and as a function of redshift and stellar mass
161: --- is helpful for constraining the processes that form stars within
162: galaxies.
163: 
164: Searches for high-redshift galaxies that are dominated by older,
165: passively evolving stellar populations are challenging.  The light
166: from these stars peaks at rest-frame optical and near-IR wavelengths
167: ($\simeq 0.4-2$~\micron), which at $z\gsim 1$ are shifted into the
168: observed-frame near-IR.  Several surveys have used deep, near-IR
169: observations to search for galaxies up to $z\sim 1- 2$ whose light is
170: dominated by older stellar populations \citep[see \eg,][]{mccarthy04}.
171: Such objects should have very red $R-K$ colors, which span the
172: 4000\AA/Balmer--break at these redshifts; several searches have
173: focused on so-called extremely red objects (EROs), with selected
174: objects typically satisfying $(R-K)_\mathrm{Vega} \gsim 5$~mag.  As a
175: class, EROs include both passively evolving early-type galaxies up to
176: $z\sim 1$ \citep[\eg,][]{cimatti02} and dust-reddened starbursts
177: \citep[\eg,][]{smail02}.  However, with only optical and near-IR
178: photometry, it was not possible to uniquely interpret the majority of
179: these objects \citep[\eg,][]{moustakas04,wilson07}.  Recent
180: observations at longer wavelengths from \spitzer\ indicate that more
181: than half of the ERO population have strong emission in the thermal IR
182: \citep{wilson04,yanl04b}, suggesting that such objects constitute a
183: significant portion of the red-selected galaxy population.
184: 
185: Recently, observers have used deep near-IR data to study red-selected
186: galaxies at higher redshifts.  \citet{franx03} used deep VLT/ISAAC
187: $JH\ks$ observations to identify a population of ``distant red
188: galaxies'' (DRGs) with $(J_s - \ks)_\mathrm{Vega} > 2.3$~mag, which
189: should select galaxies that have a strong Balmer/4000~\AA\ break
190: between the $J_s$ and \ks\ bands at $z\sim 2$-3.5 \citep[see
191: also][]{saracco01}.  Like EROs, the DRG color selection is also
192: sensitive to high redshift ($z\gsim 1$) galaxies dominated by
193: dust-enshrouded starbursts also have red observed $J-K$ colors
194: \citep[\eg,][]{smail02}.  Indeed, subsequent study of DRGs has shown
195: that the majority are massive galaxies that are actively forming stars
196: at $z\sim 1.5$-3
197: \citep{vd03,fs04,rubin04,knudsen05,reddy05,papovich06,webb06}, and
198: only a small subset appear to be completely devoid of star formation
199: and passively evolving \citep{labbe05,kriek06}.
200: 
201: However, nearly all surveys of red-selected objects have been
202: restricted to using near-IR images from ground-based telescopes, which
203: in practice limits these searches to $K_\mathrm{AB}\lsim 24$~mag
204: \citep[\eg,][]{labbe03}.  For passively evolving stellar populations
205: at $z\sim 3$, this magnitude limit acts as a limit in \textit{stellar
206: mass} of $\gsim 10^{11}$~\msol.  In contrast, the $z\sim 2-3.5$
207: UV-selected, star-forming galaxies (so-called Lyman-break galaxies,
208: LBGs) that dominate the SFR density \citep{reddy05} have typical
209: stellar masses $\sim 10^{10}$~\msol\ (for ``\lstar'' LBGs; Papovich et
210: al.\ 2001).  The inferred ages for LBGs are generally less than 1~Gyr
211: \citep{shapley01,shapley05} --- significantly less than the age of the
212: Universe at these epochs.  Thus, it is unknown if there exists a
213: population of ``faded'' LBGs, identifiable as red, passively evolving
214: galaxies with stellar masses $\sim 10^{10}$~\msol.  Deriving
215: constraints on the density of such objects would improve our
216: understanding of the evolution of these galaxies and on how galaxies
217: assemble their stellar content at these early epochs.  However,
218: detecting lower-mass, red galaxies at high redshift requires near-IR
219: surveys with higher sensitivity than what is practically available
220: from the ground.
221: 
222: Here, we use observations in the \textit{Hubble} Ultra Deep Field
223: \citep[HUDF; Beckwith et al.\ 2006 and][]{thompson05}.  This field has
224: the deepest optical \hst/ACS images to date, which combined with
225: extremely deep, near-IR \hst/NICMOS, and Spitzer (3.6, 4.5~\micron)
226: datasets form part of the Great Observatories Origins Deep Survey
227: (GOODS).  The depth of the images in this field provides a means to
228: explore high-redshift galaxies to a relatively low limiting
229: stellar-mass.  Here, we focus on a sample of galaxies selected with
230: red $\nicj - \nich$ colors in an effort to identify candidates for
231: passively evolving galaxies at $z\gsim 2$.  Similarly,
232: \citet{brammer07} use a $J-H$ color selection to identify red galaxies
233: at $z > 2$ using the deep FIRES near-IR survey imaging.  The
234: \hst/NICMOS data for the HUDF allow us to probe objects with lower
235: stellar masses than has been previously possible.  For example,
236: \citet{dickinson00} used deep NICMOS data in the \textit{Hubble} Deep
237: Field \citep[HDF;][]{williams96} to study an unusually red ($\nicj -
238: \nich \gsim 2$~mag) source, possibly a passively evolving $z\sim 3$
239: galaxy with $\mcal > 3\times 10^{10}$~\msol.  Similarly, Chen \&
240: Marzke (2004) used the HUDF \hst\ images to identify red galaxies
241: ($\acsi-\nich > 2$) with photometric redshifts $z_\mathrm{ph} > 2.8$
242: (see also Yan et al.\ 2004).  In this work we use the HUDF data to
243: study the number density and mass density of objects at $2.5 < z <
244: 3.5$ with colors consistent with older passively evolving stellar
245: populations.  To study the HUDF $\nicj - \nich > 1$ sources, when
246: possible we include GOODS IRAC 3.6-4.5~\micron\ and MIPS 24~\micron\
247: data, which allows us to better constrain the nature of the stellar
248: populations in these objects (as we discuss below).
249: 
250: Throughout this paper we assume the following cosmological parameters:
251: $\Omega_M =$ 0.3, $\Omega_{\Lambda} =$ 0.7 and H$_0$ = 70 km s$^{-1}$
252: Mpc$^{-1}$ \citep[consistent with the latest \textit{WMAP} results,
253: e.g.,][]{sp03}.  We use \acsb\acsv\acsi\acsz\nicj\nich\ to denote
254: magnitudes derived in the the \hst\ ACS and NICMOS bandpasses F435W,
255: F606W, F775W, F850LP, F110W and F160W respectively.  We denote
256: magnitudes derived from the IRAC channel 1 and channel 2 data as
257: $[3.6\micron]$ and $[4.5\micron]$.  For reference, the effective
258: wavelengths of these bandpasses are $0.43$, 0.60, 0.77, 0.91, 1.1,
259: 1.6, 3.6 and 4.5~\micron. All magnitudes henceforth are on the AB
260: system, $m(\mathrm{AB}) = 23.9 - 2.5\log(f_\nu / 1\;\ujy)$.
261: 
262: \section{Red galaxies in the HUDF: photometry and selection}\label{section:data}
263: 
264: We combine \hst/ACS and NICMOS observations of the HUDF (Beckwith et
265: al.\ 2006\footnote{http://www.stsci.edu/hst/udf}; Thompson et al.\
266: 2005) over a $2.4 \times 2.4$ arcmin$^2$ area with \spitzer\ IRAC
267: observations from GOODS (M.~Dickinson et al., in
268: preparation)\footnote{http://www.stsci.edu/science/goods/}.  The plate
269: scale of the drizzled images is $0.03$, $0.09$ and $0.6$ arcsec
270: pix$^{-1}$ for the ACS, NICMOS, and IRAC data respectively.  The $10
271: \sigma$ AB point source detection limits for the ACS and NICMOS images
272: are $29.2$, $30.0$, $29.7$, $28.7$, $27.0$ and $27.0$~mag for the
273: \acsb\acsv\acsi\acsz\nicj\nich\ images respectively.  For the IRAC
274: bands, the $5 \sigma$ AB point source detection limits are 26.3~mag
275: ($0.11\mu$Jy) and 25.6~mag ($0.2\mu$Jy), for $[3.6\micron]$ and
276: $[4.5\micron]$ (M.~Dickinson, private communication, 2005).
277: 
278: \begin{figure}[t]
279:   \begin{center}
280:   \includegraphics[angle=0,width=0.49\textwidth]{fig1.eps}  
281:   \caption{ Color-color plot of all catalog objects with $H_{160} <
282:     26.0$.  The $J_{110} - H_{160} > 1.0$ color cut is marked.  The
283:     color evolution of three \citet{bc03} models is shown, with
284:     integer red-shift steps marked.  The curve labeled SSP (orange) is
285:     a $1.0$Gyr old simple stellar population formed in a burst with no
286:     subsequent star formation.  The CSP curve (blue) is a $100$Myr
287:     constant star-forming model attenuated using \citet{dc00} dust.
288:     The purple curve a is superposition of these two SSP and CSP
289:     models, with a ratio (SSP:CSP) of $99:1$ by mass.  Although the
290:     SSP and CSP models do not individually characterize the data well,
291:     the two-component model does.  The median error for the $J_{110} -
292:     H_{160}$ color is smaller than the symbol size; the median error
293:     for the $V_{606} - z_{850}$ color is similar to the symbol
294:     size.\label{fig1_1}}
295:   \end{center}
296: \end{figure}
297: 
298: We use ACS images reprocessed by \citet{thompson05} and matched to the
299: NICMOS plate-scale and orientation.  We then convolve the ACS images
300: by a kernel to match the point spread function (PSF) of the \nich-band
301: image.  Because the PSFs of the \nicj\ and \nich\ images are similar
302: we do not convolve the \nicj-band image.  We use standard IRAF
303: packages to perform the image convolution.  Our tests show that
304: point-source photometry on the convolved images recovers the same
305: fraction of the total flux to within $5\%$ for aperture radii greater
306: than $5$ pixels ($0.45$ arcsec).  Source detection was executed using
307: the SExtractor software \citep[v2.3.2,][]{bertin96} on a summed \nicj\
308: + \nich\ image, and photometry was performed on the individual
309: PSF-matched images.  We use the same detection and photometry
310: parameters as in \citet{thompson05}.  We restrict our analysis to
311: objects with $\nich < 26$~mag measured in the SExtractor elliptical,
312: quasi-total (``Kron''; SExtractor MAG\_AUTO) apertures.  The NICMOS
313: data detect galaxies to 27th magnitude, and adopting a $\nich <
314: 26$~mag limit ensures that we can derive robust colors for objects
315: with $\nicj - \nich = 1$.  We measure ACS and NICMOS colors for
316: objects in the PSF-matched images using isophotal (MAG\_ISO) apertures
317: derived from the \nich-image; we correct the isophotal magnitudes to
318: total magnitudes using the difference between the \nich-band MAG\_AUTO
319: and MAG\_ISO magnitudes.
320: 
321: The photometric errors from SExtractor are underestimated because they
322: do not take into account pixel-to-pixel correlations from, for
323: example, the drizzling and PSF-convolution processes.  We therefore
324: estimate photometric errors from the binned, drizzled, PSF-convolved
325: ACS images themselves.  We measure the sky noise as a function of
326: aperture size in blank regions of each image.  We then parameterize
327: the noise as a power law, $\sigma = A\times n_\mathrm{pix} ^\beta$,
328: where $\sigma$ is the noise measured in an aperture with a size of
329: $n_\mathrm{pix}$ pixels, and we fit for $\beta$ and $A$.  In order to
330: measure the noise in the convolved images we tabulate 1250 randomly
331: positioned aperture fluxes; we then bin these data and fit for the
332: width of the distribution with a Gaussian function.  We exclude bins
333: in the non-Gaussian positive tail of the distribution (these
334: correspond to apertures that receive some flux from bonafide sources
335: in the image).  This procedure is repeated for 11 aperture diameters
336: in each band: 0.54, 0.60, 0.80, 0.99, 1.15, 1.30, 1.40, 1.53, 1.65,
337: 1.80 and 2$\farcs$25.  We obtain best-fit values of $\beta = 0.74$,
338: $0.92$, $0.96$, $0.93$ and $A = (5.6, 3.5, 2.0, 2.2)\times-10^{-4}$
339: ADU sec$^{-1}$, for \acsb\acsv\acsi\acsz\ respectively.  Using these
340: fitted values, we can then derive the error, $\sigma$ in each band for
341: any isophotal aperture size, $n_\mathrm{pix}$.
342: 
343: In figure~\ref{fig1_1} we plot the $J_{110} - H_{160}$ vs. $V_{606} -
344: z_{850}$ colors for all $H_{160} < 26.0$ catalog objects.  We also
345: show the color evolution of three \citet{bc03} models, a passively
346: evolving burst (SSP) model, a constant star-forming (CSP) model, and a
347: composite SSP+CSP model, at fixed age and as a function of redshift.
348: The composite model characterizes the $\nicj - \nich > 1$ selected
349: objects well.
350: 
351: \begin{figure}[t]%%[!h]
352: %%  \begin{center}
353:     %%\epsscale{0.79}\plotone{fig2.eps}\epsscale{1.0}
354:   \includegraphics[angle=0,width=0.49\textwidth]{fig2.eps}   
355:   \caption{ Color vs.\ redshift diagram of \citet{bc03} population
356:       synthesis models showing $J_{110} - H_{160}$ color evolution for
357:       different ages, star-formation histories and dust attenuations.
358:       The solid curves show a simple stellar population (SSP) model at
359:       two different ages ($1.0$ Gyr and $0.5$ Gyr).  The dotted curves
360:       show a $100$ Myr constant star-formation model (composite
361:       stellar population: CSP) with two different dust attenuations:
362:       $E(B-V) = 0.3$ (lower dotted curve) and $E(B-V) = 0.6$ (upper
363:       dotted curve).  The dot-dashed curve (labeled SSP+CSP) shows an
364:       example of a two-component model: an old population ($1.0$ Gyr
365:       SSP) and a young population ($100$ Myr constant star-formation
366:       with $E(B-V) = 0.3$) are combined with a ratio (SSP:CSP) of
367:       $99:1$ by mass.  The dashed curve shows a maximally old SSP
368:       formed at $z = 7$.  Note that our selection criterion, $J_{110}
369:       - H_{160} > 1.0$, generally excludes objects at redshifts less
370:       than $z \sim 1.5$.\label{fig1}}
371: %%  \end{center}
372: \end{figure}
373: 
374: To target galaxies whose light is dominated by evolved stellar
375: populations, we select objects with $\nicj - \nich > 1.0$~mag and
376: $\nich \leq 26$~mag, which in principle selects objects with a
377: prominent $4000$~\AA/Balmer break at $z \gsim 1.5$.  In
378: figure~\ref{fig1} we plot several fiducial stellar population
379: synthesis models from \citet{bc03}, with solar metallicity and a
380: Salpeter IMF.  Our color criterion excludes model stellar populations
381: with recent star formation (and low-to-moderate dust attenuations),
382: and should be sensitive to passively evolving, older, simple stellar
383: populations (SSPs), or dust-enshrouded objects with younger stellar
384: populations.  We find 32 objects satisfying $\nicj - \nich > 1.0$~mag
385: and $\nich \leq 26.0$~mag in the NICMOS-selected catalog; their
386: properties are summarized in table~\ref{tab1}.
387: 
388: 
389: \begin{deluxetable*}{lcccccccccccl}
390: %%\rotate
391: \tabletypesize{\scriptsize}
392: \tablecaption{Photometric properties of the NICMOS-UDF J$_{110}$ - H$_{160} > 1.0$ sample}
393: \tablewidth{0pt}
394: \tablehead{
395: \colhead{NIC ID} 
396: & \colhead{R.A.} 
397: & \colhead{Decl.} 
398: & \colhead{$z_{tot}$} 
399: & \colhead{$H_{tot}$} 
400: & \colhead{$B - z$} 
401: & \colhead{$V - z$} 
402: & \colhead{$I - z$} 
403: & \colhead{$z - H$} 
404: & \colhead{$J - H$} 
405: & \colhead{[3.6$\micron$]} 
406: & \colhead{[4.5$\micron$]}
407: & \colhead{Alternate ID} \\
408:  \colhead{ }
409: & \colhead{[J2000]}
410: & \colhead{[J2000]}
411: & \colhead{AB}
412: & \colhead{AB}
413: & \colhead{AB}
414: & \colhead{AB}
415: & \colhead{AB}
416: & \colhead{AB}
417: & \colhead{AB}
418: & \colhead{AB}
419: & \colhead{AB}
420: & \colhead{}
421: }
422: \startdata
423: 30$^{*}$ & 03:32:39.72 & -27:46:11.3 & 23.60 & 21.73 & 1.55 & 1.28 & 0.63 & 1.87 & 1.18 & 19.96 & 19.77 & 1$^{c}$  \\
424: 66 & 03:32:39.32 & -27:46:23.6 & 25.41 & 24.00 & 1.48 & 0.64 & 0.32 & 1.41 & 1.10 & \nodata & \nodata & \nodata  \\
425: 86 & 03:32:37.13 & -27:46:25.9 & 24.99 & 23.23 & 5.41 & 2.60 & 1.03 & 1.76 & 1.02 & \nodata & \nodata & 2$^{c}$  \\
426: 88$^{*}$ & 03:32:38.01 & -27:46:26.3 & 25.68 & 23.30 & 2.27 & 1.50 & 0.71 & 2.37 & 1.43 & 21.59 & 21.44 & 3$^{c}$  \\
427: 99 & 03:32:36.17 & -27:46:27.7 & 24.58 & 23.07 & 1.47 & 0.80 & 0.45 & 1.51 & 1.05 & \nodata & \nodata & \nodata  \\
428: 111$^{*}$ & 03:32:41.75 & -27:46:28.9 & 27.12 & 25.53 & 1.22 & 0.99 & 0.24 & 1.59 & 1.21 & 24.44 & 24.48 & \nodata  \\
429: 120$^{*}$ & 03:32:41.01 & -27:46:31.5 & 26.57 & 24.50 & 1.53 & 0.96 & 0.42 & 2.07 & 1.35 & 22.42 & 22.21 & \nodata  \\
430: 215 & 03:32:34.77 & -27:46:46.1 & 26.41 & 25.08 & 1.61 & 0.91 & 0.23 & 1.33 & 1.05 & \nodata & \nodata & \nodata  \\
431: 219$^{*}$ & 03:32:43.32 & -27:46:46.9 & 26.16 & 23.76 & 1.79 & 0.76 & 0.24 & 2.39 & 1.78 & 22.21 & 22.11 & 9151$^{a}$  \\
432: 223$^{*}$ & 03:32:35.06 & -27:46:47.6 & 25.79 & 23.36 & 1.97 & 0.97 & 0.43 & 2.43 & 1.62 & 21.61 & 21.48 & 9024$^{a}$, 4$^{c}$, 8$^{d}$  \\
433: 234 & 03:32:42.96 & -27:46:49.9 & 23.43 & 21.81 & 4.29 & 2.52 & 0.90 & 1.62 & 1.03 & \nodata & \nodata & \nodata  \\
434: 245 & 03:32:39.88 & -27:46:50.0 & 26.67 & 25.71 & 0.07 & 0.20 & 0.04 & 0.97 & 1.03 & \nodata & \nodata & \nodata  \\
435: 269$^{*}$ & 03:32:38.73 & -27:46:54.1 & 25.74 & 24.63 & 1.11 & 0.45 & 0.18 & 1.10 & 1.09 & 23.28 & 23.55 & \nodata  \\
436: 281 & 03:32:41.67 & -27:46:55.4 & 26.03 & 23.72 & 2.05 & 1.54 & 0.97 & 2.31 & 1.28 & \nodata & \nodata & \nodata  \\
437: 319$^{*}$ & 03:32:34.42 & -27:46:59.7 & 24.28 & 22.31 & 3.03 & 2.21 & 0.93 & 1.97 & 1.24 & 20.40 & 20.25 & 5$^{c}$  \\
438: 478 & 03:32:42.52 & -27:47:14.3 & 26.14 & 24.75 & 1.77 & 0.52 & 0.04 & 1.39 & 1.12 & \nodata & \nodata & \nodata  \\
439: 535 & 03:32:34.62 & -27:47:20.9 & 25.00 & 23.38 & 2.38 & 1.94 & 0.84 & 1.61 & 1.08 & \nodata & \nodata & \nodata  \\
440: 625$^{*}$ & 03:32:43.46 & -27:47:27.4 & 26.63 & 24.34 & 2.63 & 0.82 & 0.18 & 2.29 & 1.77 & 23.18 & 23.26 & 6140$^{a}$  \\
441: 683 & 03:32:36.96 & -27:47:27.2 & 25.22 & 23.23 & 2.33 & 1.49 & 0.62 & 1.99 & 1.08 & \nodata & \nodata & \nodata  \\
442: 706$^{*}$ & 03:32:42.73 & -27:47:33.9 & 24.37 & 22.74 & 1.90 & 1.51 & 0.75 & 1.63 & 1.05 & 21.28 & 21.27 & 5256$^{a}$  \\
443: 902 & 03:32:38.10 & -27:47:49.8 & 24.47 & 22.17 & 3.82 & 1.80 & 0.93 & 2.30 & 1.10 & \nodata & \nodata & 3650$^{b}$  \\
444: 921$^{*}$ & 03:32:39.11 & -27:47:51.6 & 25.50 & 23.11 & 3.36 & 1.55 & 0.86 & 2.39 & 1.19 & 21.92 & 21.96 & 3574$^{b}$, 13$^{d}$  \\
445: 935 & 03:32:33.26 & -27:47:52.4 & 27.23 & 24.77 & 2.05 & 1.97 & 0.98 & 2.46 & 1.33 & \nodata & \nodata & 6$^{d}$  \\
446: 989$^{*}$ & 03:32:34.93 & -27:47:56.0 & 26.21 & 24.76 & 1.70 & 0.57 & 0.35 & 1.44 & 1.06 & 23.37 & 23.44 & \nodata  \\
447: 1078$^{*}$ & 03:32:42.87 & -27:48:09.5 & 28.06 & 24.54 & 2.02 & 1.03 & 0.62 & 3.51 & 2.49 & 22.03 & 21.78 & 12182$^{a}$, 1$^{d}$  \\
448: 1082 & 03:32:38.70 & -27:48:10.3 & 27.35 & 25.84 & 1.75 & 0.73 & 0.09 & 1.51 & 1.34 & \nodata & \nodata & \nodata  \\
449: 1172$^{*}$ & 03:32:41.74 & -27:48:24.9 & 27.75 & 24.56 & 2.02 & 2.80 & 0.83 & 3.18 & 1.39 & 22.48 & 22.32 & 4$^{d}$  \\
450: 1184 & 03:32:38.75 & -27:48:27.1 & 26.43 & 24.57 & 1.84 & 0.53 & 0.17 & 1.86 & 1.30 & \nodata & \nodata & 15$^{d}$  \\
451: 1211$^{*}$ & 03:32:39.16 & -27:48:32.4 & 25.78 & 23.25 & 2.30 & 1.19 & 0.33 & 2.52 & 2.02 & 22.08 & 22.06 & 1927$^{a}$, 1446$^{b}$, 6$^{c}$, 9$^{d}$  \\
452: 1237$^{*}$ & 03:32:38.73 & -27:48:39.9 & 27.07 & 24.48 & 2.37 & 0.97 & 0.38 & 2.59 & 1.95 & 22.03 & 21.82 & 12183$^{a}$, 2$^{d}$  \\
453: 1245 & 03:32:39.58 & -27:48:42.2 & 26.21 & 25.25 & 1.49 & 0.59 & 0.15 & 0.95 & 1.07 & \nodata & \nodata & \nodata  \\
454: 1267 & 03:32:39.66 & -27:48:50.6 & 24.36 & 22.38 & 2.96 & 1.01 & 0.22 & 1.98 & 1.57 & \nodata & \nodata & 1223$^{a}$  \\
455: 
456: \enddata
457: \label{tab1}
458: \tablecomments{Objects with unblended IRAC photometry are marked with
459:   $^*$.  References for the alternate ID's listed are (a)
460:   \citet{chen04}, (b) \citet{daddi05}, (c) \citet{toft05} and (d) \citet{yan04a}.}
461: \end{deluxetable*}
462: 
463: \begin{figure}[t]
464:   %%\epsscale{0.79}\plotone{fig3.eps}\epsscale{1.0}
465:   \includegraphics[angle=0,width=0.47\textwidth]{fig3.eps}   
466:   \caption{ $\nich - \mtwo$ versus $\acsi - \nich$ colors for the 16
467:       galaxies in the $\nicj - \nich > 1$~mag sample with robust IRAC
468:       photometry.  The lines show colors of various model stellar
469:       populations formed at $z = 7$.  The dashed line shows the
470:       expected colors of a passively evolving stellar population
471:       observed at $z = 2.7$, formed in a burst as a function of age,
472:       $t$; the dot-dashed line shows the same model observed at $z =
473:       3.5$.  Tick marks denote the model ages along the dashed and
474:       dot-bashed curves as labeled.  The solid lines show colors of a
475:       stellar population with constant star formation formed at $z =
476:       7$ as a function of age and extinction with color excess $E(B-V)
477:       = 0.2$ and 0.6, as labeled.  The encircled data points indicate
478:       the 6 candidates for distant galaxies dominated by old stellar
479:       populations (see \S~4.1).  These models roughly bound the galaxy
480:       colors, and indicate that including the IRAC data we can
481:       understand the properties of the stellar populations in these
482:       galaxies.  Note that none of the objects have colors consistent
483:       with passively evolving stellar populations older than 1 Gyr.}
484:   \label{cfig1}
485: \end{figure}
486: 
487: We match $\nicj-\nich > 1$~mag objects against the \spitzer/IRAC
488: catalog from the Great Observatories Origins Deep Survey (GOODS;
489: M.~Dickinson et al. in preparation).  Source detection for this
490: catalog was performed on a summed $[3.6\micron$]+$[4.5\micron]$ image,
491: and photometry was measured in circular, 4.0\arcsec-diameter
492: apertures, which we correct to total magnitudes by adding $-0.36$ and
493: $-0.31$ respectively to the $[3.6\micron$] and $[4.5\micron]$
494: magnitudes (appropriate for GOODS, M. Dickinson et al., in prep, and
495: slightly larger than those in the IRAC data
496: handbook\footnote{http://ssc.spitzer.caltech.edu/irac/dh/}).  Because
497: the image PSF size is substantially different between the NICMOS and
498: IRAC images (FWHM $0\farcs3$ and $1\farcs6$, respectively), we do not
499: attempt to PSF match these images.
500: 
501: We matched these objects to sources in the \spitzer/IRAC catalog first
502: using a 4\arcsec-radius criterion.  We then visually inspected each
503: source and accept only those objects whose positions coincide with the
504: \hst\ source (they are well-centered) and who are isolated (unblended)
505: from other \hst\ sources within that radius.  This is a conservative
506: approach, and has the advantage of limiting our analysis to only those
507: \spitzer\ counterparts that are free from source confusion.  Of the 32
508: objects in the parent sample, 16 have unblended IRAC $[3.6\micron]$
509: and $[4.5\micron]$ counterparts.  In figures~\ref{fig2}-\ref{fig5} we
510: show the ACS, NICMOS, and IRAC images and spectral energy
511: distributions (SEDs) of these objects.
512: 
513: Including the IRAC data extends our wavelength coverage, allowing us
514: to discriminate broadly between galaxies dominated by dust-obscured
515: starbursts and galaxies with a substantial amount of stellar mass in
516: later-type stars (see, e.g., Labb\'e et al.\ 2005).  In
517: figure~\ref{cfig1}, we show a $\nich - \mtwo$ versus $\acsi-\nich$
518: plot to study the stellar populations and dust extinction properties
519: for the 16 galaxies with IRAC photometry.  We also plot fiducial
520: colors of a passively evolving stellar population and dust-enshrouded
521: starbursts (with $E(B-V) = 0.2$ and 0.6, using the extinction law from
522: Calzetti et al.\ 2000) as a function of stellar population age at
523: $z=2.7$.  These simple models bound the range of colors observed in
524: the $\nicj - \nich > 1$~mag objects, comparable to findings for red,
525: $J-\ks$-selected galaxies in other fields (Labb\'e et al.\ 2005;
526: Papovich et al.\ 2006).  Blue $\nich - \mtwo$ colors ($\lsim 2$~mag)
527: at this redshift (rest-frame $0.4-1.2$~micron) result from substantial
528: older stellar populations, which produce a significant
529: Balmer/4000~\AA\ break, causing redder $\acsi - \nich$ colors (and
530: also the red $\nicj - \nich$ color that satisfy the selection
531: criteria).  Objects at these redshifts with redder $\nich - \mtwo$
532: colors ($\gsim 2$~mag) require dust-enshrouded, ongoing star
533: formation.  The galaxies with $\nicj - \nich > 1$ and IRAC photometry
534: all have a similar range of $\acsi - \nich$ colors, making it
535: difficult to differentiate between the two types of stellar
536: populations.  However, these galaxies span a large range in $\nich -
537: \mtwo$ colors, from 1-4~mag.  Many of the objects appear to have
538: colors consistent with dust-enshrouded, young stellar populations,
539: while a few sources have colors consistent with a substantial amount
540: of their stellar mass in older stars.  Based on figure~3, there are
541: few galaxies whose colors are consistent with a pure passively
542: evolving, older stellar population.  We discuss this further in \S~4.
543: 
544: \begin{figure}[t]
545:   %%\epsscale{0.79}\plotone{fig4.eps}\epsscale{1.0}
546:     \includegraphics[angle=0,width=0.49\textwidth]{fig4.eps}   
547:   \caption{Limiting stellar mass for galaxies with a stellar
548:     population formed in a burst at $z = 7$ with observed magnitude
549:     $\nich = 26.0$ or $\ks = 23.2$.  The HUDF data are sensitive to
550:     passively evolving stellar populations with masses above these
551:     limits. \label{fig:masslimit}}
552:   \label{cfig2}
553: \end{figure}
554: 
555: We also include other multiwavelength data to interpret the nature of
556: the $\nicj - \nich > 1$~mag sample.  In particular, we use
557: \spitzer/MIPS 24~\micron\ data for this field from the catalog of
558: Papovich et al.\ (2004).  These data allow us to test for the emission
559: from dust associated with either star formation or obscured AGN in
560: these galaxies.  We also include \textit{Chandra} X--ray data from the
561: catalog of Alexander et al.\ (2003).  At the redshifts of interest for
562: our sample, the X--ray data are sensitive accretion of material onto
563: supermassive blackholes, and thus allow us to test for the presence of
564: AGN in our sample.
565: 
566: \begin{deluxetable*}{lccccccc}
567: \tabletypesize{\scriptsize}
568: \tablecaption{Best-fit parameters with HST data only}
569: \tablewidth{0pt}
570: \tablehead{
571: \colhead{NIC ID} 
572: & \colhead{z$_{phot}$} 
573: & \colhead{E(B-V)$^{a}$} 
574: & \colhead{log M$_{o}$/M$_\sun$$^{b}$} 
575: & \colhead{log M$_{y}$/M$_\sun$$^{c}$} 
576: & \colhead{f$_{old}$$^{d}$} 
577: & \colhead{$\chi_{\nu,best}^{2}$$^{e}$}
578: & \colhead{$\Delta \chi^{2}$$^{f}$}
579: }
580: \startdata
581: 30$^{*}$ & 1.4$^{+0.2}_{-0.1}$ & 0.2$\pm$0.1 &  11.2$\pm$0.1 & \phn9.0$^{+0.5}_{-0.3}$ & 0.993$^{+0.004}_{-0.010}$ & $\phn$5.2 & $\phn$0.0  \\
582: 66 & 2.6$\pm$0.2 & 0.3$^{+0.1}_{-0.0}$ &  10.6$^{+0.1}_{-0.4}$ & \phn9.6$^{+0.2}_{-0.1}$ & 0.915$^{+0.021}_{-0.207}$ & $\phn$2.5 & $\phn$0.3  \\
583: 86 & 1.1$^{+3.3}_{-0.1}$ & 1.1$^{+0.3}_{-0.8}$ &  10.3$^{+1.5}_{-0.2}$ & \phn9.8$^{+0.3}_{-0.5}$ & 0.763$^{+0.224}_{-0.136}$ & $\phn$0.0 & $\phn$0.8  \\
584: 88$^{*}$ & 2.2$^{+0.1}_{-0.2}$ & 0.6$\pm$0.1 &  10.8$^{+0.1}_{-0.2}$ &  10.0$\pm$0.4 & 0.867$^{+0.083}_{-0.231}$ & $\phn$0.7 & $\phn$0.0  \\
585: 99 & 2.1$\pm$0.1 & 0.4$\pm$0.0 &  10.7$\pm$0.1 & \phn10.$\pm$0.1 & 0.857$^{+0.018}_{-0.023}$ & $\phn$3.0 & $\phn$0.0  \\
586: 111$^{*}$ & 0.7$^{+0.1}_{-0.0}$ & 0.0$^{+0.1}_{-0.0}$ & \phn8.8$\pm$0.0 & \phn6.5$^{+0.3}_{-0.0}$ & 0.996$^{+0.000}_{-0.005}$ & 17.9 & $\phn$0.0  \\
587: 120$^{*}$ & 2.4$^{+0.3}_{-0.2}$ & 0.4$\pm$0.1 &  10.5$^{+0.2}_{-0.1}$ & \phn9.2$^{+0.3}_{-0.4}$ & 0.948$^{+0.032}_{-0.085}$ & $\phn$0.4 & $\phn$0.0  \\
588: 215 & 0.6$\pm$0.0 & 0.5$\pm$0.0 & \phn6.6$^{+0.3}_{-0.7}$ & \phn7.9$\pm$0.0 & 0.048$^{+0.051}_{-0.034}$ & 12.6 & $\phn$0.0  \\
589: 219$^{*}$ & 2.8$\pm$0.1 & 0.3$\pm$0.0 &  11.0$\pm$0.1 & \phn9.3$^{+0.0}_{-0.1}$ & 0.981$^{+0.004}_{-0.002}$ & $\phn$3.5 & $\phn$0.0  \\
590: 223$^{*}$ & 2.9$^{+0.3}_{-0.4}$ & 0.3$\pm$0.1 &  11.3$^{+0.2}_{-0.3}$ & \phn9.4$^{+0.2}_{-0.3}$ & 0.986$^{+0.009}_{-0.027}$ & $\phn$1.3 & $\phn$1.1  \\
591: 234 & 0.9$^{+0.1}_{-0.0}$ & 1.1$^{+0.2}_{-0.1}$ &  10.6$\pm$0.1 &  10.1$\pm$0.2 & 0.783$^{+0.094}_{-0.076}$ & $\phn$1.8 & $\phn$0.0  \\
592: 245 & 2.4$\pm$0.1 & 0.0$\pm$0.0 & \phn10.$\pm$0.0 & \phn8.1$\pm$0.0 & 0.989$\pm$0.001 & 14.7 & $\phn$0.0  \\
593: 269$^{*}$ & 2.7$^{+0.3}_{-0.1}$ & 0.2$^{+0.0}_{-0.1}$ &  10.4$^{+0.3}_{-0.1}$ & \phn9.2$^{+0.0}_{-0.2}$ & 0.941$^{+0.043}_{-0.020}$ & $\phn$2.3 & $\phn$0.0  \\
594: 281 & 1.9$^{+0.2}_{-0.1}$ & 0.3$^{+0.2}_{-0.1}$ &  10.7$\pm$0.1 & \phn8.7$^{+0.6}_{-0.5}$ & 0.992$^{+0.005}_{-0.029}$ & $\phn$0.6 & $\phn$0.0  \\
595: 319$^{*}$ & 1.1$\pm$0.1 & 0.9$^{+0.1}_{-0.0}$ &  10.1$^{+0.3}_{-0.5}$ &  10.2$\pm$0.2 & 0.421$^{+0.218}_{-0.200}$ & $\phn$1.9 & $\phn$0.0  \\
596: 478 & 3.1$\pm$0.2 & 0.2$^{+0.1}_{-0.0}$ &  10.7$^{+0.1}_{-0.3}$ & \phn9.2$^{+0.2}_{-0.1}$ & 0.968$^{+0.007}_{-0.086}$ & $\phn$3.1 & $\phn$2.5  \\
597: 535 & 1.1$\pm$0.1 & 0.7$^{+0.1}_{-0.5}$ &  10.0$^{+0.3}_{-0.5}$ & \phn9.3$^{+0.3}_{-1.5}$ & 0.788$^{+0.208}_{-0.330}$ & $\phn$4.4 & $\phn$0.0  \\
598: 625$^{*}$ & 3.3$\pm$0.4 & 0.2$\pm$0.1 &  11.1$^{+0.2}_{-0.3}$ & \phn8.9$\pm$0.4 & 0.993$^{+0.005}_{-0.017}$ & $\phn$3.9 & $\phn$0.0  \\
599: 683 & 1.6$^{+0.2}_{-0.1}$ & 0.6$\pm$0.1 &  10.5$^{+0.2}_{-0.3}$ & \phn9.9$^{+0.3}_{-0.5}$ & 0.800$^{+0.139}_{-0.291}$ & $\phn$0.6 & $\phn$0.0  \\
600: 706$^{*}$ & 1.2$^{+0.1}_{-0.0}$ & 0.3$^{+0.1}_{-0.2}$ &  10.6$\pm$0.1 & \phn8.7$^{+0.4}_{-0.6}$ & 0.988$^{+0.009}_{-0.022}$ & $\phn$3.5 & $\phn$0.7  \\
601: 902 & 1.9$\pm$0.1 & 0.8$\pm$0.1 &  11.1$^{+0.1}_{-0.3}$ &  10.8$^{+0.2}_{-0.4}$ & 0.682$^{+0.183}_{-0.295}$ & $\phn$2.8 & $\phn$0.0  \\
602: 921$^{*}$ & 2.0$^{+0.1}_{-0.2}$ & 0.7$\pm$0.1 &  10.8$^{+0.1}_{-0.3}$ &  10.2$^{+0.4}_{-0.5}$ & 0.811$^{+0.124}_{-0.306}$ & $\phn$1.3 & $\phn$0.0  \\
603: 935 & 1.9$\pm$0.2 & 0.5$\pm$0.4 &  10.3$^{+0.1}_{-0.3}$ & \phn8.6$\pm$1.2 & 0.978$^{+0.021}_{-0.445}$ & $\phn$0.9 & $\phn$0.7  \\
604: 989$^{*}$ & 3.1$\pm$0.3 & 0.2$\pm$0.1 &  10.7$^{+0.2}_{-0.3}$ & \phn9.1$\pm$0.3 & 0.979$^{+0.014}_{-0.049}$ & $\phn$1.6 & $\phn$0.1  \\
605: 1078$^{*}$ & 2.9$^{+0.3}_{-0.2}$ & 0.1$\pm$0.1 &  10.8$\pm$0.2 & \phn7.5$^{+0.6}_{-0.4}$ & 1.000$^{+0.000}_{-0.001}$ & $\phn$6.6 & $\phn$0.6  \\
606: 1082 & 2.7$^{+0.0}_{-0.1}$ & 0.4$\pm$0.0 & \phn9.6$^{+0.1}_{-0.2}$ & \phn9.2$\pm$0.0 & 0.707$^{+0.062}_{-0.112}$ & $\phn$6.0 & $\phn$0.0  \\
607: 1172$^{*}$ & 2.0$\pm$0.4 & 1.1$^{+0.3}_{-0.8}$ &  10.4$^{+0.2}_{-0.6}$ &  10.1$^{+0.6}_{-2.3}$ & 0.513$^{+0.485}_{-0.397}$ & $\phn$1.0 & $\phn$0.8  \\
608: 1184 & 3.2$^{+0.2}_{-0.3}$ & 0.2$\pm$0.1 &  10.9$\pm$0.2 & \phn9.0$\pm$0.3 & 0.987$^{+0.009}_{-0.029}$ & $\phn$0.4 & $\phn$0.2  \\
609: 1211$^{*}$ & 2.9$^{+0.4}_{-0.2}$ & 0.4$\pm$0.1 &  11.2$^{+0.3}_{-0.2}$ & \phn9.8$\pm$0.3 & 0.960$^{+0.028}_{-0.081}$ & $\phn$9.5 & $\phn$1.0  \\
610: 1237$^{*}$ & 3.3$^{+0.3}_{-0.4}$ & 0.1$^{+0.2}_{-0.1}$ &  11.0$\pm$0.2 & \phn8.3$^{+0.5}_{-0.4}$ & 0.998$^{+0.001}_{-0.008}$ & $\phn$2.7 & $\phn$0.0  \\
611: 1245 & 2.8$^{+0.3}_{-0.1}$ & 0.3$^{+0.0}_{-0.1}$ & \phn9.8$\pm$0.5 & \phn9.3$^{+0.1}_{-0.2}$ & 0.766$^{+0.182}_{-0.260}$ & $\phn$5.6 & $\phn$1.7  \\
612: 1267 & 3.2$^{+0.2}_{-0.3}$ & 0.4$\pm$0.1 &  11.6$^{+0.2}_{-0.3}$ &  10.6$\pm$0.3 & 0.921$^{+0.054}_{-0.180}$ & $\phn$5.1 & $\phn$0.0  \\
613: \enddata
614: \label{tab2}
615: \tablenotetext{a}{The E(B-V) extinction is applied only to the young component}
616: \tablenotetext{b}{The model fit mass in the old (SSP) component.}
617: \tablenotetext{c}{The model fit mass in the young (constant SFR) component.}
618: \tablenotetext{d}{The fraction by mass of the old component relative
619: to the total fitted mass.}
620: \tablenotetext{e}{The best-fit model $\chi_{\nu}^2$ is listed.}
621: \tablenotetext{f}{The difference between the $\chi^2$ value 
622:   for the model listed in columns $2$ through $4$ and
623:   the best $\chi^2$:  \\
624:   $\Delta \chi^2$ = $\chi_{median}^2 -$
625:   $\chi_{best}^2$ (also see Section $3.1$).}
626: \tablecomments{Columns $2$ through $4$ list the median values and
627:   the $16\%$ and $82\%$ intervals for the fitted parameters.  
628:   Objects with unblended IRAC photometry are marked with
629:   an asterisk symbol.  Objects 30, 88, 1237 and 1267 are detected in x-rays.}
630: \end{deluxetable*}
631: 
632: 
633: \section{SED fitting: mass estimates from a simple photometric
634:   redshift code}
635: 
636: \subsection{Fitting method}
637: 
638: To study the properties of our selected objects, we fit
639: stellar-population synthesis models to the photometry and derive
640: redshifts and stellar masses.  We model the galaxies as the sum of two
641: stellar components, young and old, and consider the effects of
642: reddening.  Because the star-formation history is constrained
643: generally poorly by SED fitting, and the stellar population age and
644: dust extinction have strong degeneracies (see, e.g., Papovich et al.\
645: 2001), we consider only non-negative linear superpositions of these
646: two models.  The combinations of these two models encompass the range
647: of plausible star-formation histories, and in particular, as we
648: discuss below, they allow us to derive a stringent upper limit on the
649: amount of stellar mass in old stellar populations.
650: 
651: We therefore have four model parameters: the redshift, dust
652: extinction, mass in the young stellar component, and the mass in the
653: older stellar component.  We use the \citet{bc03} population synthesis
654: code to generate the two stellar population components for our SED
655: fitting routine: a $100$ Myr constant star formation (composite
656: stellar population: CSP) model and a maximally old passively evolving
657: (simple stellar population: SSP) model formed at z = 7, both generated
658: at solar metallicity and with a Salpeter initial mass function (IMF).
659: Our maximally old model age is a function of redshift: the age at a
660: given redshift is equal to the age of the universe at that redshift
661: minus 0.75~Gyr, the age of the universe at z = 7.  We attenuate the
662: CSP model using the \citet{dc00} dust law; the old model is not
663: attenuated.  At each $z$ and $E(B-V)$, we compute model photometry by
664: convolving the \citet{bc03} spectra with the $8$ filter response
665: curves \citep[e.g.,][]{cp01}, taking in to account the HI opacity of
666: the intergalactic medium (IGM) along the line of sight \citep{pm95},
667: according the models described in \citet{fan99}.
668: 
669: We fit the object fluxes to this four-parameter model assuming that
670: the errors in the fluxes are symmetric and Gaussian.  In addition to
671: the photometric uncertainties on the magnitude of each bandpass, we
672: add a systematic uncertainty proportional to the flux density of
673: $\sigma_\mathrm{sys} / f_\nu = 0.04$, 0.08, and 0.15 for the ACS,
674: NICMOS, and IRAC photometry, respectively.  These errors represent
675: systematic uncertainties in our matched-band photometry, as well as
676: the fact that our model does not sample the full range of possible
677: parameters infinitesimally.  These systematic errors are added in
678: quadrature to either the fitted errors, for the ACS photometry, or the
679: catalog errors for the NICMOS and IRAC photometry.  We use the
680: resulting $\chi^2$ statistic to produce the likelihood $\calL \propto
681: \exp(-\chi^2/2)$ and we take the likelihood to be the probability
682: distribution for the 4 parameters given the observed fluxes.
683: 
684: \begin{deluxetable*}{lccccccc}
685: \tabletypesize{\scriptsize}
686: \tablecaption{Best-fit parameters for objects with IRAC photometry}
687: \tablewidth{0pt}
688: \tablehead{
689: \colhead{NIC ID} 
690: & \colhead{z$_{phot}$} 
691: & \colhead{E(B-V)$^{a}$} 
692: & \colhead{log M$_{o}$/M$_\sun$$^{b}$} 
693: & \colhead{log M$_{y}$/M$_\sun$$^{c}$} 
694: & \colhead{f$_{old}$$^{d}$} 
695: & \colhead{$\chi_{\nu,best}^{2}$$^{e}$}
696: & \colhead{$\Delta \chi^{2}$$^{f}$}
697: }
698: \startdata
699: 30$^{**}$ & 1.8$^{+0.2}_{-0.1}$ & 0.3$^{+0.1}_{-0.0}$ &  11.5$\pm$0.1 & \phn9.7$^{+0.5}_{-0.1}$ & 0.984$^{+0.002}_{-0.026}$ & $\phn$9.2 & $\phn$0.0  \\
700: 88$^{**}$ & 2.2$^{+0.2}_{-0.1}$ & 0.7$^{+0.0}_{-0.1}$ &  10.8$\pm$0.2 &  10.3$^{+0.1}_{-0.2}$ & 0.718$^{+0.158}_{-0.071}$ & $\phn$0.7 & $\phn$0.0  \\
701: 111 & 1.5$^{+0.1}_{-0.0}$ & 0.4$\pm$0.0 & \phn9.6$\pm$0.1 & \phn8.5$^{+0.1}_{-0.0}$ & 0.914$^{+0.010}_{-0.015}$ & 11.3 & $\phn$0.0  \\
702: 120$^{**}$ & 3.4$^{+0.1}_{-0.2}$ & 0.2$\pm$0.1 &  11.1$\pm$0.1 & \phn8.9$\pm$0.4 & 0.994$^{+0.004}_{-0.012}$ & $\phn$2.0 & $\phn$0.2  \\
703: 219$^{\bigstar}$ & 2.8$^{+0.2}_{-0.0}$ & 0.3$^{+0.0}_{-0.1}$ &  11.1$\pm$0.1 & \phn9.3$^{+0.0}_{-0.3}$ & 0.982$^{+0.011}_{-0.002}$ & $\phn$2.2 & $\phn$0.0  \\
704: 223$^{\bigstar}$ & 3.1$\pm$0.1 & 0.2$^{+0.1}_{-0.0}$ &  11.4$\pm$0.1 & \phn9.1$^{+0.3}_{-0.0}$ & 0.995$^{+0.000}_{-0.007}$ & $\phn$0.7 & $\phn$0.0  \\
705: 269$^{\bigstar}$ & 2.7$^{+0.1}_{-0.0}$ & 0.2$\pm$0.0 &  10.4$\pm$0.1 & \phn9.2$\pm$0.0 & 0.944$^{+0.007}_{-0.008}$ & $\phn$1.9 & $\phn$0.0  \\
706: 319$^{**}$ & 1.2$^{+0.0}_{-0.1}$ & 0.9$^{+0.1}_{-0.0}$ &  10.0$^{+0.2}_{-0.4}$ &  10.3$\pm$0.1 & 0.344$^{+0.129}_{-0.156}$ & $\phn$1.3 & $\phn$0.0  \\
707: 625$^{\bigstar}$ & 2.7$\pm$0.1 & 0.4$^{+0.0}_{-0.1}$ &  10.6$\pm$0.1 & \phn9.4$^{+0.1}_{-0.3}$ & 0.937$^{+0.039}_{-0.015}$ & $\phn$3.3 & $\phn$0.0  \\
708: 706 & 1.3$^{+0.2}_{-0.0}$ & 0.6$\pm$0.0 &  10.3$^{+0.2}_{-0.3}$ & \phn9.9$\pm$0.1 & 0.703$^{+0.095}_{-0.137}$ & $\phn$4.2 & $\phn$0.0  \\
709: 921 & 1.9$\pm$0.1 & 0.5$\pm$0.1 &  10.9$\pm$0.1 & \phn9.5$\pm$0.3 & 0.966$^{+0.019}_{-0.045}$ & $\phn$1.6 & $\phn$0.0  \\
710: 989$^{\bigstar}$ & 2.9$^{+0.1}_{-0.2}$ & 0.3$^{+0.0}_{-0.1}$ &  10.5$^{+0.2}_{-0.1}$ & \phn9.3$^{+0.1}_{-0.3}$ & 0.935$^{+0.042}_{-0.019}$ & $\phn$1.3 & $\phn$1.5  \\
711: 1078 & 2.9$^{+0.7}_{-0.3}$ & 1.0$^{+0.0}_{-1.0}$ &  10.6$^{+0.6}_{-0.3}$ &  10.7$^{+0.1}_{-3.3}$ & 0.386$^{+0.614}_{-0.123}$ & $\phn$5.8 & $\phn$2.4  \\
712: 1172 & 2.1$^{+0.2}_{-0.3}$ & 1.1$^{+0.3}_{-0.1}$ &  10.4$^{+0.1}_{-0.2}$ &  10.1$^{+0.1}_{-0.2}$ & 0.661$^{+0.078}_{-0.072}$ & $\phn$1.0 & $\phn$0.3  \\
713: 1211$^{\bigstar}$ & 2.7$\pm$0.1 & 0.4$^{+0.1}_{-0.0}$ &  11.0$\pm$0.1 & \phn9.9$\pm$0.2 & 0.949$^{+0.013}_{-0.073}$ & $\phn$5.0 & $\phn$0.5  \\
714: 1237$^{**}$ & 3.6$^{+0.2}_{-0.0}$ & 0.0$^{+0.1}_{-0.0}$ &  11.2$^{+0.1}_{-0.0}$ & \phn8.0$^{+0.4}_{-0.1}$ & 0.999$^{+0.000}_{-0.001}$ & $\phn$1.9 & $\phn$0.0  \\
715: \enddata
716: \label{tab3}
717: \tablenotetext{a}{The E(B-V) extinction is applied only to the young component}
718: \tablenotetext{b}{The model fit mass in the old (SSP) component.}
719: \tablenotetext{c}{The model fit mass in the young (constant SFR) component.}
720: \tablenotetext{d}{The fraction by mass of the old component relative
721: to the total fitted mass.}
722: \tablenotetext{e}{The best-fit model $\chi_{\nu}^2$ is listed.}
723: \tablenotetext{f}{The difference between the $\chi^2$ value 
724:   for the model listed in columns $2$ through $4$ and
725:   the best $\chi^2$:  \\
726:   $\Delta \chi^2$ = $\chi_{median}^2 -$
727:   $\chi_{best}^2$ (also see Section $3.1$).}
728: \tablecomments{Columns $2$ through $4$ list the median values and
729:   the $16\%$ and $82\%$ intervals for the fitted parameters.    
730:   The six galaxies with broad-band SEDs consistent with distant
731:   evolved stellar populations are marked with a star symbol.  
732:   Objects detected at $24\micron$ are indicated with
733:   a double-asterisk symbol.  Objects 30, 88, and 1237 are detected in x-rays.}
734: \end{deluxetable*}
735: 
736: We derive mean values and variances for the 4 parameters (and
737: functions thereof) by integrating over the full 4-dimensional
738: likelihood.  The integrations are done with a hybrid grid plus Monte
739: Carlo approach.  The $\chi^2$ is quadratic in two parameters (the old
740: and young masses, $M_o$ and $M_y$) and more complicated in the other
741: two (redshift and reddening).  We construct a grid in redshift and
742: reddening over the range $0<z<5$ and $0<E(B-V)<1.5$.  For each point,
743: we use linear methods to find the best fit and Gaussian covariance
744: matrix for the masses.  We generate Gaussian-distributed points given
745: this covariance region and weight points by $\exp(-\chi^2_{\rm
746: best}/2)$ for the best-fit $\chi^2$, dropping any points that have
747: negative masses.  The concatenation of all of the points is a Monte
748: Carlo of the full 4-dimensional likelihood, and any function can be
749: computed by a simple weighted sum of the function evaluated at all of
750: the points.  Similarly, we can construct quantiles by sorting the
751: points according to a given function evaluation and cumulating the
752: weights.  By using a grid in the non-quadratic directions, we ensure
753: that multiple minima are found and correctly weighted.
754:   
755: The derived physical parameters are presented in table~\ref{tab2} and
756: table~\ref{tab3}.  In these tables we define $f_\mathrm{old}$ to be
757: the ratio of the mass in the passively evolving component, $M_\mathrm
758: o$ to the total mass, $M_\mathrm o + M_\mathrm y$, where $M_\mathrm y$
759: is the mass in the star-forming component,
760: \begin{equation} 
761: f_\mathrm{old} = M_\mathrm o / (M_\mathrm o + M_\mathrm y).
762: \end{equation}
763: We characterize the probability distribution function for the various
764: parameters by the median and the $16\%$ and $84\%$ intervals
765: (equivalent to the $1\sigma$ intervals for a Gaussian distribution).
766: However, in cases with multiple minima these simple quantiles will not
767: fully represent the distribution.
768: 
769: \begin{figure*}[t]
770:   \begin{center}
771:     \scalebox{0.43}{{\includegraphics{fig5a.eps}}
772:       {\includegraphics{fig5b.eps}}}
773:     \scalebox{0.43}{{\includegraphics{fig5c.eps}}
774:       {\includegraphics{fig5d.eps}}}
775:     \caption{ Gallery of $J_{110} - H_{160} > 1.0$ and $H_{160} <
776:     26.0$ selected objects; we show imaging, broad-band SED's and the
777:     best fit model for each object.  The
778:     $B_{435}V_{606}i_{775}z_{850}J_{110}H_{160}[3.6\mu m][4.6\mu m]$
779:     postage stamps are $5\farcs4$ on a side and oriented such that
780:     North is up and East is to the left.  The
781:     $B_{435}V_{606}i_{775}z_{850}J_{110}H_{160}$ stamps are plotted at
782:     constant $f_{\nu}$ per arcsec and are displayed on a log scale.
783:     The IRAC imaging is also displayed at constant $f_{\nu}$ per
784:     arcsec on a log scale but on a different relative scale than the
785:     $B_{435}V_{606}i_{775}z_{850}J_{110}H_{160}$ imaging.  The
786:     broad-band SED's are shown in the lower panels with solid circles,
787:     $1\sigma$ error bars are indicated.  The crosses indicate the best
788:     fit model.  The dashed line indicates the fraction, by mass, of
789:     light that is attributed by the best fit to the young model;
790:     similarly, the solid line indicates the fraction attributed to the
791:     old model.\label{fig2}}
792:   \end{center}
793: \end{figure*}
794: 
795: \subsection{Fitting results}
796:   
797: We apply this SED fitting method to derive physical parameters for our
798: objects.  We present the results of the ACS and NICMOS photometry
799: fitting analysis in detail in Table~\ref{tab2} and summarize them here
800: as follows.
801: 
802: The fits to 11 objects in our sample favor models with relatively
803: lower redshifts ($z_\mathrm{phot} < 2.5)$ and lower stellar mass
804: fractions in old stellar populations ($f_\mathrm{old} < 0.9$): NIC IDs
805: $86$, $88$, $99$, $215$, $234$, $319$, $535$, $683$, $902$, $921$, and
806: $1172$.  We note that on average this set of galaxies has more dust,
807: with a mean fitted $E(B - V) \sim 0.8$, than the entire sample, which
808: has a mean fitted $E(B - V) \sim 0.4$.  The fits to 7 objects favor
809: models with lower redshifts and higher mass fractions in older stars
810: ($z_\mathrm{phot} < 2.5$ and $f_\mathrm{old} > 0.9$): $30$, $111$,
811: $120$, $245$, $281$, $706$, and $935$.  For objects 1082 and 1245, the
812: models favor a solution with high redshift ($z_\mathrm{phot} > 2.5$)
813: and a relatively low fraction of mass in old stars ($f_\mathrm{old} <
814: 0.9$).  For 12 objects the models favor solutions with high redshift
815: ($z_\mathrm{phot} > 2.5$) and a relatively high fraction of stars in
816: the old stellar population ($f_\mathrm{old} > 0.9$): 66, 219, 223,
817: 269, 478, 625, 989, 1078, 1184, 1211, 1237 and 1267.
818: 
819: We find that when we include the IRAC [3.6\micron] and [4.5\micron]
820: data in the model fitting, 12 objects (of the 16 with unblended IRAC
821: photometry) have consistent results compared to the model fits without
822: the IRAC data.  Of these 12 objects, $6$ objects are fitted with
823: high-redshift model galaxies with only small amounts of rest-frame UV
824: light associated with a young stellar population or on-going star
825: formation (for this set $z_\mathrm{phot} \geq 2.7$ and $f_\mathrm{old}
826: > 0.93$).  The fitted model parameters for these $16$ objects are
827: listed in Table~\ref{tab3}.  Here we discuss the particularities of
828: these objects and their corresponding fits.  We will not individually
829: discuss the objects without IRAC photometry.  However, we note that
830: object $1267$, is detected in the X-ray data, with a $0.5 - 2.0$ keV
831: flux of $f_x = 7.5 \times 10^{-16}$ ergs cm$^{-2}$ s$^{-1}$.
832: 
833: \begin{figure*}[t]
834:   \begin{center}
835:     \scalebox{0.43}{{\includegraphics{fig6a.eps}}
836:       {\includegraphics{fig6b.eps}}}
837:     \scalebox{0.43}{{\includegraphics{fig6c.eps}}
838:       {\includegraphics{fig6d.eps}}}
839:     \caption{Same as figure~\ref{fig2}.\label{fig3}}
840:   \end{center}
841: \end{figure*}
842: 
843: {\bf \noindent NIC ID $= 30$:} The difference in the fitted parameters
844: when we include the IRAC data is significant: $\Delta z_{phot} \sim
845: 0.4$.  As expected, the masses change as well.  However, both values
846: for $f_\mathrm{old}$ remain high.  The best-fit $\chi^2_\nu$ is very
847: large for the IRAC fit ($\chi^2_\nu = 9.2$).  This is due to the large
848: amount of IRAC flux that our models inadequately characterize (see
849: figure~\ref{fig2}); this excess is also driving the selection of a
850: higher redshift and mass.  \citet{toft05} derive a photometric
851: redshift of $z = 2.2$, roughly consistent with our IRAC fit redshift
852: $z = 1.8^{+0.2}_{-0.1}$.  However, this object is detected with MIPS
853: at $24\micron$ with $f_\nu(24\micron) = 190$~\ujy, and it has an
854: X--ray data detection of $f_X(0.5-2\mathrm{keV}) = 8.4 \times
855: 10^{-16}$ ergs cm$^{-2}$ s$^{-1}$.  Based on this data it is likely
856: that this object likely has an AGN; therefore we consider the lower
857: redshift fit more plausible.  Furthermore, it is possible that the AGN
858: contributes to the broadband photometry, and our model fits may not be
859: valid for this object.
860: 
861: {\bf \noindent NIC ID $= 88$:} The two fits, with and without the IRAC
862: data, are consistent, with $z_\mathrm{phot} \sim 2.2$ and
863: $f_\mathrm{old} \sim 0.8$.  It is detected at $24\micron$ with
864: $f_\nu(24\micron) = 70$~\ujy, although inspection of the $24\micron$
865: image shows that the object is blended (with a PSF FWHM $\simeq
866: 6$\arcsec).  This object also has a $0.5 - 2.0$ keV x-ray flux $f_X =
867: 8.8 \times 10^{-16}$ ergs cm$^{-2}$ s$^{-1}$ \citep{alex03}.  In
868: conjunction these data indicate that this object is likely host to an
869: AGN, a conclusion that is supported by the fact that this object has
870: red IRAC $[5.8\micron] - [8.0\micron]$ colors and that it has a
871: compact morphology in the HST images (see figure~\ref{fig2}).  As with
872: NIC ID 30, the presence of an AGN may imply that the models are
873: inadequate.
874: 
875: {\bf \noindent NIC ID $= 111$:} The best-fit model, derived including
876: the IRAC data, is lower redshift older stellar population:
877: $z_\mathrm{phot} = 1.5$ and $f_\mathrm{old} = 0.914$.  The model
878: derived from the ACS and NICMOS data alone have a similar
879: $f_\mathrm{old}$, but the redshift is much lower, $z = 0.7$.  This
880: discrepancy is driven by the bright IRAC photometry; our models are
881: simply not red enough at $z = 0.7$ to reproduce the mid-IR data.  This
882: indicates that the ACS and NICMOS broad-band photometry alone does not
883: adequately characterize this object.
884: 
885: {\bf \noindent NIC ID $= 120$:} The fitted photometric redshifts and
886: masses are consistent with a distant old stellar population dominating
887: the flux in the rest-frame optical and a younger stellar population
888: (or on-going star formation) dominating in the rest-frame UV, with
889: $z_\mathrm{phot} = 3.4$ and $f_\mathrm{old} = 0.994$.  The physical
890: parameters derived without the IRAC data are {\it not} consistent with
891: this interpretation, with a much smaller distance ($z = 2.4)$ and a
892: somewhat lower fraction ($f_\mathrm{old} = 0.948$).  Inspection of the
893: images reveals that the blue component dominating the rest-frame UV
894: photometry is unresolved and just off-center (see figure~\ref{fig2}).
895: Therefore, the possibility that this SED is the result of the
896: projection of more than one object cannot be ruled out.  Despite the
897: fact that $[5.8\micron]$ and $[8.0\micron]$ IRAC fluxes are in good
898: agreement with the distant old population scenario, we do not include
899: this object in our set of distant and evolved object candidates
900: because of the discrepancy between the two model fits.  Furthermore,
901: this object is detected at $24\micron$.
902: 
903: {\bf \noindent NIC ID $= 219$:} Our fitted physical parameters are
904: consistent with a high redshift galaxy dominated by an old stellar
905: population ($z_\mathrm{phot} = 2.8$ and $f_\mathrm{old} = 0.98$).  The
906: fit derived from the dataset including IRAC data is consistent with
907: this model.  The derived redshift is also consistent with the
908: photometric redshifts of \citet{chen04}.
909: 
910: \begin{figure*}[t]
911:   \begin{center}
912:     \scalebox{0.43}{{\includegraphics{fig7a.eps}}
913:       {\includegraphics{fig7b.eps}}}
914:     \scalebox{0.43}{{\includegraphics{fig7c.eps}}
915:       {\includegraphics{fig7d.eps}}}
916:     \caption{Same as figure~\ref{fig2}.\label{fig4}}
917:   \end{center}
918: \end{figure*}
919: 
920: {\bf \noindent NIC ID $= 223$:} The fitted model is a distant galaxy
921: dominated by an old stellar population, with best-fit parameters
922: $z_\mathrm{phot} = 3.1 \pm 0.1$ and $f_\mathrm{old} = 0.995$,
923: consistent with the model derived from the ACS and NICMOS data alone.
924: \citet{chen04} derive a photometric redshift of $z_\mathrm{phot} =
925: 3.11$, in good agreement with our derived redshift. \citet{yan04a}
926: derive a $z_\mathrm{phot} = 2.9$ and \citet{toft05} derive a
927: $z_\mathrm{phot} = 3.4$, however, these authors do not quote errors,
928: making it hard to compare with our model.
929: 
930: {\bf \noindent NIC ID $= 269$:} The fit parameters are consistent with
931: a distant object dominated by an old stellar population with a small
932: amount of ongoing star formation, for both fits with and without IRAC
933: data ($z_\mathrm{phot} = 2.7$ and $f_\mathrm{old} = 0.94$)
934: 
935: {\bf \noindent NIC ID $= 319$:} The most-likely model is a low
936: redshift dusty starburst: $z_\mathrm{phot} = 1.2$, $E(B-V) = 0.9$ and
937: $f_\mathrm{old} = 0.34$.  \citet{toft05} derive a $z_\mathrm{phot} =
938: 1.8$ for this object.  This object is detected at $24\micron$ with
939: $f_\nu(24\micron) = 110$~\uJy, supporting the interpretation that this
940: object is a lower redshift dust--enshrouded starburst.
941: 
942: {\bf \noindent NIC ID $= 625$:} The most-likely model is consistent
943: with a distant galaxy dominated by mass by an old stellar population
944: with a small amount of ongoing star formation ($z_\mathrm{phot} = 2.7
945: \pm 0.1$ and $f_\mathrm{old} = 0.94$).  The redshift derived without
946: the IRAC data is higher, although poorly constrained, $z_\mathrm{phot}
947: = 3.3 \pm 0.4$, and the stellar population remains dominated by an old
948: population.  \citet{chen04} derive a redshift for this object of
949: $z_\mathrm{phot} = 3.33$, consistent with our fit excluding the IRAC
950: data (see table 2).
951: 
952: {\bf \noindent NIC ID $= 706$:} The most-likely model is consistent
953: with a low-redshift dusty star forming galaxy.  \citet{chen04} reach a
954: similar conclusion for this object.
955: 
956: \begin{figure*}
957:   \begin{center}
958:     \scalebox{0.43}{{\includegraphics{fig8a.eps}}
959:       {\includegraphics{fig8b.eps}}}
960:     \scalebox{0.43}{{\includegraphics{fig8c.eps}}
961:       {\includegraphics{fig8d.eps}}}
962:     \caption{Same as figure~\ref{fig2}, arrows indicate $1\sigma$
963:     upper limits.\label{fig5}}
964:   \end{center}
965: \end{figure*}
966: 
967: {\bf \noindent NIC ID $= 921$:} The fitting results are consistent
968: with an intermediate redshift old stellar population, with
969: $z_\mathrm{phot} = 1.9 \pm 0.1$ and $f_\mathrm{old} = 0.97$.
970: \citet{daddi05} measure a consistent result with a spectroscopic
971: redshift of $z = 1.98$.  \citet{yan04a} also derive a similar
972: photometric redshift.  Our mass, $8 \times 10^{10}$ \msun\ is in good
973: agreement with the mass derived by \citet{daddi05}.
974: 
975: {\bf \noindent NIC ID $= 989$:} The resulting model fit is consistent
976: with a distant galaxy dominated by mass by an old stellar population
977: and with a small amount of ongoing star formation ($z_\mathrm{phot} =
978: 2.9$ and $f_\mathrm{old} = 0.94$).  Exclusion of the IRAC data yields
979: a similar model.
980: 
981: {\bf \noindent NIC ID $= 1078$:} The most-likely model parameters
982: derived from the HST+IRAC data are consistent with a distant,
983: $z_\mathrm{phot} = 2.9 ^{+0.7}_{-0.3}$, young and dusty stellar
984: population, with a fitted $f_\mathrm{old} = 0.4$ and $E(B - V) = 1.0$,
985: although the redshift is not particularly well constrained.  Fitting
986: the photometry while excluding the IRAC data gives a consistent
987: redshift, but a very different interpretation of this object.  In this
988: case the model is dominated by an old stellar population, with
989: $f_\mathrm{old} = 1.0$.  The relatively bright IRAC flux densities,
990: taken in conjunction with the HST photometry, are inconsistent with
991: our models of old stellar populations, which require a flattening of
992: the SED at wavelengths blueward of about $0.5\micron$ rest-frame.
993: \citet{chen04} derive a $z_\mathrm{phot} = 2.91$, in good agreement
994: with our results.  However, they find that this object is consistent
995: with a E/S$0$ SED, far more evolved than the constant star-forming
996: model we derive when we include the IRAC data.  \citet{yan04a} select
997: this object using a $f_{\nu}(3.6\micron)/f_{\nu}(z_{850}) > 20$
998: selection and derive a $z_\mathrm{phot} = 3.6$, also consistent with
999: our poorly constrained $z_\mathrm{phot}$.  \citet{yan04a} do not
1000: include this object in their analysis due to the large uncertainties
1001: in their fitted redshift.  However, we note that \citet{yan04a} report
1002: a non-detection for this object (\acsz$> 29.5$~mag; using data from
1003: Yan \& Windhorst 2004c), while we report a detection of \acsz$ =
1004: 28.1$~mag.  Upon inspection of the unconvolved ACS z-band image, we
1005: see low surface brightness flux at the location of this object and we
1006: measure a detectable flux in our PSF--convolved image.  Incidentally,
1007: although the analysis of Mobasher et al.\ (2005) supports a very high
1008: redshift interpretation for this object ($z > 6$), subsequent analyses
1009: by Dunlop et al.\ (2007, MNRAS, 376, 1054) and Chary et al.\ (2007)
1010: favor an interpretation that this source resides at substantially
1011: lower redshift $z < 4$, consistent with our interpretation.
1012: 
1013: {\bf \noindent NIC ID $= 1172$:} This objects' most-likely model
1014: parameters are consistent with an intermediate redshift old stellar
1015: population with a significant amount of UV-bright younger stars:
1016: $z_\mathrm{phot} = 2.1^{+0.2}_{0.3}$ and $f_\mathrm{old} = 0.66$.
1017: \citet{yan04a} derive a $z_\mathrm{phot} = 2.7$, somewhat larger than
1018: our value of $z_\mathrm{phot}$.
1019: 
1020: {\bf \noindent NIC ID $= 1211$:} The resulting model fit is consistent
1021: with a distant galaxy whose mass is dominated by an old stellar
1022: population.  We derive a $z_\mathrm{phot} = 2.7\pm0.1$ and an
1023: $f_\mathrm{old} = 0.95$.  \citet{chen04}, \citet{daddi05},
1024: \citet{toft05} and \citet{yan04a} derive $z_\mathrm{phot} = 3.43,
1025: 2.47, 3.24, 2.8$ respectively.  We suspect the discrepancies between
1026: the different fits result from assumptions about the star-formation
1027: histories of the models and differences in the photometry.  Our
1028: derived mass, $M_{tot} = 1.1 \times 10^{11}$ \msun, is higher than,
1029: but not inconsistent with, the mass derived by \citet{daddi05}.
1030: Although this object is selected by \citet{chen04} they do not report
1031: its mass as they classify it as a dusty starburst.  This object has an
1032: upper limit X-ray flux of $f(0.5-2\mathrm{keV}) = 3.7 \times 10^{-17}$
1033: ergs cm$^{-2}$ s$^{-1}$ \citep{alex03}, implying the possible presence
1034: of an AGN.  However, the rest-frame UV through IR photometry is
1035: consistent with with integrated stellar populations and therefore we
1036: include this object in our sample of distant, evolved galaxies.
1037: 
1038: {\bf \noindent NIC ID $= 1237$:} This objects' best-fit model is
1039: consistent with a distant, old stellar population: $z_\mathrm{phot} =
1040: 3.6$ and $f_\mathrm{old} = 0.999$.  The exclusion if the IRAC data
1041: gives a similar fit, $z_\mathrm{phot} = 3.3$ and $f_\mathrm{old} =
1042: 0.998$.  \citet{chen04} derive a much higher redshift for this object,
1043: $z_\mathrm{phot} = 4.25$.  \citet{yan04a} also select this object but
1044: do not consider it due to the fact that their $z_\mathrm{phot} = 3.4$
1045: is very uncertain.  We note that our photometry for this object is
1046: likely biased because of the proximity of a very bright spiral galaxy,
1047: just visible in the lower left corner of fig. 8.  It is likely that
1048: the image convolution causes low-surface-brightness emission for this
1049: objects to enter our aperture.  Furthermore, our photometry could be
1050: biased by the presence of nearby, faint sources in the ACS data that
1051: might be blended with this source at the NICMOS H-band resolution.
1052: This object also has a marginal $24\micron$ detection due to the fact
1053: that it is blended, with $f_{\nu}(24\micron) = 76$~\ujy, which may
1054: call in to question the validity of the best-fit model.  Because of
1055: these reasons we reject this object as a candidate for a distant,
1056: evolved galaxy.
1057: 
1058: We note that although our photometry is generally in good agreement
1059: with other published values which we are aware of, a few of our
1060: objects have significant differences (notably NIC ID 1078 and 1237).
1061: In the case of NIC ID 1078, we suspect our photometry is appropriate
1062: for our NICMOS-selected catalog.  For NIC 1237, scattered light from a
1063: nearby, large galaxy likely affects our photometry measurements.
1064: Regardless, neither of these objects enters our final sample for
1065: galaxies whose mass is dominated by older stellar populations, so
1066: these do not affect our final conclusions.
1067:  
1068: We derive masses assuming a Salpeter IMF, where the \citet{bc03}
1069: models use upper and lower mass cutoffs of $100$ $\msun$ and $0.1$
1070: $\msun$.  For an SSP \citet{bc03} model, assuming a Chabrier IMF would
1071: decrease our masses by a factor of $1.8$, relative a Salpeter IMF and
1072: for the same assumed upper and lower mass cutoffs listed above.  For a
1073: more detailed discussion of how masses depend on the IMF parameters
1074: such as the slope and the mass cutoffs, see \citet{cp01}.  We note
1075: that \citet{chen04} identify red objects in the HUDF and derive masses
1076: for six early-type candidates, 4 of which are systematically lower
1077: than the masses derived here.  Our overlap objects are $219$, $223$,
1078: $625$, $1078$, and $1237$.  The total H-band magnitudes also differ
1079: for these objects; as expected, there is a rough trend between
1080: increasing flux discrepancy and increasing mass discrepancies.  Object
1081: $223$ has the largest difference in mass, a factor of $\sim 40$.
1082: Objects $219$, $625$ and $1078$ have masses that differ by factors of
1083: 18.0, 2.1 and 1.4, respectively, from the \citet{chen04} masses,
1084: although the two estimates for objects 625 and 1078 are entirely
1085: consistent.  Object 1237 is the only for which our mass is lower than
1086: the \citet{chen04} mass, by a factor of $\sim1.2$, but as stated
1087: previously, this likely results from the higher redshift derived by
1088: Chen \& Marzke.
1089: 
1090: \begin{deluxetable*}{lccccccccc}
1091: \tabletypesize{\scriptsize}
1092: \tablecaption{Comparison between Table 3 and the $\tau$-model fitting}
1093: \tablewidth{0pt}
1094: \tablehead{
1095: \colhead{NIC ID} 
1096: & \colhead{z$_{phot}$} 
1097: & \colhead{z$_{phot}^{\tau}$} 
1098: & \colhead{log M$_{TOT}$/M$_\sun$} 
1099: & \colhead{log M$_{TOT}$/M$_\sun$$^{\tau}$} 
1100: & \colhead{f$_{old}$}
1101: & \colhead{Age (Gyr)$^{\tau}$} 
1102: & \colhead{$\tau$ (Gyr)} 
1103: & \colhead{$\chi_{\nu,best}^{2}$} 
1104: & \colhead{$\chi_{\nu,best}^{2}$$^{\tau}$} 
1105: }
1106: \startdata
1107: 30 & 1.8 & 1.5$\pm$0.1 & 11.5 &  11.2$\pm$0.1 & 0.984 & 1.47$\pm$0.36 & 1.19$^{+0.71}_{-0.32}$ & \phn9.2 & \phn8.4  \\
1108: 88 & 2.2 & 2.0$^{+0.2}_{-0.5}$ & 10.9 &  10.8$^{+0.1}_{-0.3}$ & 0.718 & 0.08$^{+0.24}_{-0.02}$ & 0.01$^{+0.74}_{-0.01}$ & \phn0.7 & \phn1.3  \\
1109: 111 & 1.5 & 0.8$\pm$0.1 & \phn9.6 & \phn7.8$^{+0.0}_{-0.2}$ & 0.914 & 0.14$\pm$0.11 & 0.13$\pm$0.10 & 11.3 & \phn4.5  \\
1110: 120 & 3.4 & 2.3$\pm$0.2 & 11.1 &  10.5$^{+0.3}_{-0.4}$ & 0.994 & 0.39$^{+0.92}_{-0.36}$ & 0.91$^{+9.92}_{-0.90}$ & \phn2.0 & \phn1.2  \\
1111: 219$^{\bigstar}$ & 2.8 & 2.5$\pm$0.1 & 11.1 &  10.9$\pm$0.1 & 0.982 & 0.80$^{+0.14}_{-0.29}$ & 0.36$\pm$0.16 & \phn2.2 & \phn2.0  \\
1112: 223$^{\bigstar}$ & 3.1 & 2.3$^{+0.7}_{-0.1}$ & 11.4 &  11.1$\pm$0.1 & 0.995 & 0.86$^{+0.44}_{-0.19}$ & 0.81$^{+3.09}_{-0.63}$ & \phn0.7 & \phn1.0  \\
1113: 269$^{\bigstar}$ & 2.7 & 2.6$\pm$0.1 & 10.4 &  10.1$\pm$0.1 & 0.944 & 0.75$^{+0.19}_{-0.22}$ & 1.68$^{+11.0}_{-1.13}$ & \phn1.9 & \phn2.2  \\
1114: 319 & 1.2 & 1.0$^{+0.0}_{-0.1}$ & 10.5 &  10.4$^{+0.1}_{-0.2}$ & 0.344 & 0.08$^{+0.09}_{-0.06}$ & 0.25$^{+3.84}_{-0.25}$ & \phn1.3 & \phn0.9  \\
1115: 625$^{\bigstar}$ & 2.7 & 2.7$\pm$0.1 & 10.6 &  10.3$\pm$0.1 & 0.937 & 0.19$^{+0.23}_{-0.08}$ & 0.04$^{+0.06}_{-0.03}$ & \phn3.3 & \phn1.1  \\
1116: 706 & 1.3 & 1.2$\pm$0.1 & 10.4 &  10.2$\pm$0.3 & 0.703 & 0.13$^{+0.60}_{-0.09}$ & 0.71$^{+7.17}_{-0.69}$ & \phn4.2 & \phn2.8  \\
1117: 921 & 1.9 & 1.9$^{+0.2}_{-1.4}$ & 10.9 &  10.6$^{+0.1}_{-1.8}$ & 0.966 & 0.26$^{+0.22}_{-0.23}$ & 0.01$^{+0.08}_{-0.01}$ & \phn1.6 & \phn1.6  \\
1118: 989$^{\bigstar}$ & 2.9 & 2.5$\pm$0.2 & 10.5 &  10.1$^{+0.1}_{-0.3}$ & 0.935 & 0.47$^{+0.34}_{-0.37}$ & 0.68$^{+4.35}_{-0.61}$ & \phn1.3 & \phn2.2  \\
1119: 1078 & 2.9 & 2.7$^{+0.2}_{-0.1}$ & 11.0 &  11.2$\pm$0.1 & 0.386 & 0.53$^{+0.75}_{-0.40}$ & 0.12$^{+0.18}_{-0.11}$ & \phn5.8 & \phn1.2  \\
1120: 1172 & 2.1 & 1.4$^{+0.3}_{-0.5}$ & 10.6 &  10.4$^{+0.2}_{-0.4}$ & 0.661 & 0.52$^{+0.56}_{-0.41}$ & 0.01$^{+0.11}_{-0.01}$ & \phn1.0 & \phn0.5  \\
1121: 1211$^{\bigstar}$ & 2.7 & 2.8$\pm$0.1 & 11.1 &  10.9$\pm$0.1 & 0.949 & 0.34$\pm$0.11 & 0.09$\pm$0.01 & \phn5.0 & \phn1.7  \\
1122: 1237 & 3.6 & 2.7$^{+0.2}_{-0.1}$ & 11.2 &  11.2$^{+0.1}_{-0.2}$ & 0.999 & 1.03$^{+0.66}_{-0.41}$ & 0.72$^{+0.54}_{-0.45}$ & \phn1.9 & \phn0.5  \\
1123: \enddata
1124: 
1125: \tablecomments{The columns with $\tau$-model parameters are indicated
1126:   with a $\tau$ symbol.  The six galaxies with broad-band SEDs
1127:   consistent with distant evolved stellar populations are marked with
1128:   a star symbol.}
1129: 
1130: \end{deluxetable*}
1131: 
1132: \subsection{Other models}
1133: 
1134: To investigate possible systematic dependences due to our modeling, we
1135: explore the effects of changing our model assumptions.
1136: 
1137: We first consider a single component fit using the same young model
1138: described above, a constant star forming model attenuated with the
1139: \citet{dc00} law.  This model does not characterize our objects well,
1140: yielding very large $\chi^2$ values compared to the two-component base
1141: model, with an average increase in $\chi^2 \sim 55$.
1142: 
1143: We also test a more complex dust model, the \citet{witt00} SMC shell
1144: geometry dust model.  We try both a single component and a
1145: two-component fit with this dust model.  The single component fit, a
1146: $100$~Myr old constant star forming model with dust as a free
1147: parameter, does not characterize the data well; the best-fit $\chi^2$
1148: values are very large compared to our base model and to the single
1149: component fit using \citet{dc00} dust discussed above.  The two
1150: component fit, the second component being a maximally old SSP model,
1151: results in an average $\chi^2$ difference of $\langle
1152: \chi^2_\mathrm{best,\ base\ model} - \chi^2_\mathrm{best} \rangle =
1153: -1.5$ over all 16 objects, somewhat worse than our base model.  This
1154: dust model gives lower $\chi^2_\mathrm{best}$ values for 9
1155: objects. Object $921$ has the largest $\Delta \chi^2$ of 3.3.  This
1156: dust model assigns systematically higher $E(B - V)$ values, in
1157: particular to objects 30, 625, 921 and 1211, with best-fit $E(B - V)
1158: =$ 1.4, 2.5, 5.3 and 2.8 respectively.  For completeness we note that
1159: the dust grid over which our models span is the grid provided by the
1160: \citet{witt00} look-up tables for this dust law.  It is not the same
1161: grid as that in our base model and it is not a regularly varying grid,
1162: having values of $E(B - V)$ ranging from $0.0$ to $12.3$, in steps of
1163: $\Delta E(B - V) = 0.0875$ at low values and in steps of $\Delta E(B -
1164: V) = 1.75$ at high values.  This difference may hamper our ability to
1165: make a fair comparison with our base model.  However, we do not think
1166: this is the case as none of our original model fits have large values
1167: of $E(B - V)$, all objects having model fit values of $E(B - V) \leq
1168: 1.1$.  We note that the best-fit masses differ by less than a factor
1169: of $3.6$ between this model and our base model, excluding object 319,
1170: whose mass increases by about an order of magnitude relative to our
1171: base model.  Similarly, the fractions change by small amounts:
1172: $\langle | \Delta f_\mathrm{old} | \rangle = 0.08$ over the set of 16
1173: objects.  The only objects with a significant change in redshift are
1174: $88$, 319 and 1172 with a $\Delta z_\mathrm{phot} = 0.9$, 1.1 and 1.1
1175: respectively. None of these objects are included in our final sample
1176: of galaxies whose mass is dominated by older stellar populations.
1177: 
1178: We next consider the possibility that our objects have metallicities
1179: lower than solar.  Again, we perform a similar analysis as outlined
1180: above, with both the young and old models at the same ages as for our
1181: base model and using the \citet{dc00} dust law, but with both
1182: components at a metallicity of $0.2Z_{\odot}$.  For this model, the
1183: average difference in $\chi^2$ is $\langle \chi^2_\mathrm{best,\ base\
1184: model} - \chi^2_\mathrm{best} \rangle = -2.1$, i.e., our base model
1185: does a better job at fitting our object SEDs.  This set of
1186: low-metallicity models yields a lower $\chi^2_\mathrm{best}$ for four
1187: of our objects.  For the entire set of 16 objects the change in
1188: redshift is $\langle | \Delta z_\mathrm{phot} | \rangle= 0.3$, with
1189: object $30$ having by far the largest difference: $z_\mathrm{phot,\
1190: base\ model} - z_\mathrm{phot} = -1.3$.  The best-fit masses change
1191: little, with a variation smaller than a factor of $\sim 3$ for all
1192: objects.  The stellar populations remain consistent for all 16 objects
1193: except 1237, which is fitted as a young starburst with $f_\mathrm{old}
1194: = 0.07$, as opposed to our base model fit with $f_\mathrm{old,\ base\
1195: model} = 0.999$.  This result lends credence to the interpretation
1196: supported by the 24$\micron$ and X-ray data that this object may host
1197: an obscured AGN.  We note that altering the metallicity of the base
1198: model does not significantly change the fitting results for the 6
1199: objects we select as candidates for distant, evolved galaxy candidates
1200: and therefore will not affect our final conclusions.
1201: 
1202: We also consider the effects of a dusty old population; we fit our
1203: data using the two-component base model outlined above and we use
1204: \citet{dc00} dust to attenuate the old component with $E(B - V) =
1205: 0.05$ (or an $A_\mathrm{V} = 0.2$).  For this model we calculate an
1206: average $\chi^2$ change of $\langle \chi^2_\mathrm{best,\ base\ model}
1207: - \chi^2_\mathrm{best} \rangle = 4.1$.  We derive
1208: $\chi^2_\mathrm{best}$ that are lower than those derived with our base
1209: models for $11$ objects (NIC ID = $30$, $88$, $111$, $120$, $269$,
1210: $625$, $706$, $986$, $1078$, $1172$, and $1211$).  For this sub-set,
1211: objects $111$, $625$ $1078$, and $1211$ have the largest differences,
1212: with $\Delta \chi^2 = 32.9$, $9.0$, $11.7$, and $6.2$ respectively.
1213: The other $7$ objects all have $\Delta \chi^2 < 3.3$.  The photometric
1214: redshifts change little, with a $\langle | \Delta z_\mathrm{phot} |
1215: \rangle = 0.15$ for all objects, except $111$, which has a $\Delta
1216: z_\mathrm{phot} = 1.9$.  The mass fractions in old and young stars
1217: change little, with $\langle | \Delta f_\mathrm{old} | \rangle = 0.03$
1218: for all objects except 219 and 1237, both of which have a decrease in
1219: $f_\mathrm{old}$ of $f_\mathrm{old,\ base\ model} - f_\mathrm{old} =
1220: 0.9$.  For the 6 objects we consider distant, evolved galaxy
1221: candidates, the derived masses change by less than $50\%$.
1222: 
1223: Finally, we consider single-component $\tau$-models with varying ages
1224: similar to other studies in the literature.  We summarize the fitting
1225: results here and compare, in more detail, the best-fit $\tau$-model
1226: parameters with our base model results in Table~4.  We generate models
1227: over a grid in ages and $\tau$ values, were $\tau$ is the exponential
1228: timescale for decay in star-formation.  Our age grid values are 0.01,
1229: 0.02, 0.04, 0.06, 0.08, 0.10, 0.20, 0.50, 0.70, 1.00, and 2.00 Gyr,
1230: and our $\tau$ grid spans 0.001~Gyr to 20.0~Gyr in quasi-logarithmic
1231: steps.  We use the same redshift and dust grid as that of our base
1232: model and we assume the same dust attenuation law, the \citet{dc00}
1233: law.  We fit the observed SEDs of our selected objects using this
1234: 4-dimensional grid of models.  With these models we obtain lower
1235: $\chi^2_\mathrm{best}$ values, relative to our base model, for 11
1236: objects (NIC ID = 30, 111, 120, 219, 319, 625, 706, 1078, 1172, 1211,
1237: and 1237), with objects 111, 1078, and 1211 having the largest
1238: differences of $\Delta \chi^2 = $ 27, 18, and 13, respectively.  For
1239: all 16 objects we measure a change in $\chi^2$ of $\langle
1240: \chi^2_\mathrm{best,\ base\ model} - \chi^2_\mathrm{best} \rangle =
1241: 5.0$; if we exclude objects 111, 1078, and 1211, the change in
1242: $\chi^2$ becomes $\langle \chi^2_\mathrm{best,\ base\ model} -
1243: \chi^2_\mathrm{best} \rangle = 1.7$.  The masses of the 11 objects
1244: with lower $\chi^2_\mathrm{best,\tau}$ change by less than a factor 2
1245: with respect to the base model fit, the only exceptions being objects
1246: 111 and 120, which have factors of $\sim$60 and $\sim$3 decrease in
1247: mass respectively.  The fitted redshifts do not change significantly,
1248: $\langle z_\mathrm{phot,base\ model} - z_\mathrm{phot} \rangle = 0.3$,
1249: with objects 111, 120, and 1172 having the largest change of $\Delta
1250: z_\mathrm{phot} \sim $ 0.7, 1.0, and 0.7 respectively.  For the entire
1251: sample of 16 it is interesting to note that 14/16 objects have base
1252: model masses that are higher than those derived with the
1253: $\tau$-models.  We consider this to be a consequence of the fact that
1254: our base model is better suited to measure an upper limit on the
1255: masses of our objects; this is in part because the maximally old
1256: component of this model has a limiting mass to light ratio.
1257: Furthermore, we note that for the 6 objects with distant evolved fits,
1258: we derive masses which are between a factor of 1.3 and 2.5 greater
1259: than those derived from the $\tau$-model fits.  Therefore we conclude
1260: that our adopted two-component base model is better suited to place
1261: upper limits on the old stellar population census in our objects than
1262: the $\tau$-model.  We note that the two objects with higher
1263: $\tau$-model masses are 1078 and 1237, which are greater by a factor
1264: of 1.6 and 1.1, respectively, still consistent within the errors with
1265: our base model.
1266: 
1267: We conclude that a two-component fit is better able to characterize
1268: the data than a one-component fit and that, although some objects are
1269: better fit by other models, in general our base model characterizes
1270: our objects well.  We show that the changes in our fitted redshifts,
1271: total masses and fractions ($f_\mathrm{old}$) are not strongly
1272: dependent on the assumed metallicity and dust properties of our
1273: models.
1274: 
1275: \section{Discussion and Summary}\label{section:discussion}
1276: 
1277: We have investigated the physical properties of $32$ galaxies selected
1278: with $(\nicj - \nich) > 1.0$ colors in the HUDF using deep, broadband
1279: imaging from \hst\ and \spitzer.  Exactly half of these objects
1280: (16/32) have unblended IRAC $[3.6\micron]$ and $[4.5\micron]$
1281: photometry.  Here, we focus on the subset of galaxies with IRAC
1282: photometry because the data span a longer wavelength baseline,
1283: allowing us to separate more cleanly the objects into those whose
1284: colors are consistent with dust--enshrouded starbursts and those
1285: dominated by old stars.  Thus, where we derive number densities and
1286: mass densities, we include a factor of 2 correction for incompleteness
1287: due to blending in the IRAC image.  Furthermore, including both the
1288: \spitzer/IRAC data allows us to better constrain the properties of the
1289: galaxies' stellar populations (see further discussion in, e.g.,
1290: Labb\'e et al.\ 2005; Papovich et al.\ 2006).
1291: 
1292: \subsection{The Properties of Galaxies with Evolved Stellar
1293: Populations at $z\gsim 2.5$}
1294: 
1295: Based on the full SED fitting, we split the subsample of objects with
1296: IRAC photometry into two broad categories: those objects whose
1297: broad-band SEDs are well represented by our two component model with a
1298: substantial component of evolved stars combined with a young
1299: star-forming component, and those objects whose interpretation changes
1300: substantially by including the IRAC photometry compared to the model
1301: fits to only the ACS and NICMOS data.  For 12/16 (75\%) of the
1302: objects, including the IRAC data give roughly consistent best-fit
1303: models as to the ACS and NICMOS data only (NIC IDs 88, 219, 223, 269,
1304: 319, 625, 706, 921, 989, 1172, 1211, and 1237).  Because we are
1305: primarily interested in placing constraints on the number density of
1306: galaxies dominated by passively evolving stellar populations, we focus
1307: here on the subset of galaxies whose model fits have $z \geq 2.5$ and
1308: $f_\mathrm{old} \geq 0.9$.  That is, the models favor a scenario where
1309: the vast majority of the galaxy's stellar mass resides in an older,
1310: passively evolving component.  There are 6 objects satisfying these
1311: conditions (NIC IDs 219, 223, 269, 625, 989, and 1211).
1312: 
1313: For three of these objects (NIC IDs 223, 625, 989), we have
1314: higher-confidence redshifts due to the presence of a Lyman ``break''
1315: between in the $\acsb -\acsv$ colors.  This occurs for galaxies with
1316: relatively blue rest-frame UV colors, and a characteristic ``break''
1317: in the colors as neutral hydrogen attenuation from the IGM absorbs the
1318: intrinsic flux shortward of 1216~\AA\ (e.g., Madau 1995).  For the ACS
1319: filters, absorption owing to IGM neutral hydrogen affects the
1320: \acsb-band at $z \gsim 3$, and the presence of this break drives the
1321: estimate of the redshifts for these galaxies ($z_\mathrm{phot} \simeq
1322: 3.5$, 3.2, and 3.1, respectively).  Interestingly, at $z\sim 3-3.5$
1323: the Balmer/4000~\AA\ is moving through the \nich-band, diminishing the
1324: $J-H$ color. Therefore the fact that these galaxies have $\nicj -
1325: \nich \geq 1$~mag implies they need a relatively large fraction of
1326: their stellar mass in older populations, and in all cases we find best
1327: fits with $f_\mathrm{old} > 0.93$.
1328: 
1329: Chen \& Marzke (2004) used a NICMOS-selected catalog in the HUDF
1330: combined with photometric redshifts to select six candidates for
1331: massive, red ($\acsi - \nich \geq 2$) early-type galaxies at $z > 3$.
1332: Interestingly, for only 2 of their candidates where we have IRAC data
1333: does our analysis reach the conclusion that old stellar populations
1334: dominate the observed colors (NIC IDs 223, 625).  We find for three of
1335: their candidates (NIC IDs 219, 1078, 1237) that when the IRAC data are
1336: included the SED fits favor solutions where most of the light
1337: originates from dusty, young stellar populations.  For the remaining
1338: candidate of Chen \& Marzke, we do not have IRAC data but our SED fit
1339: to the existing ACS + NICMOS photometry does favor a solution where
1340: the galaxy is dominated by older stellar populations.
1341: 
1342: For three of our six objects with IRAC photometry, $z\geq 2.7$, and
1343: $f_\mathrm{old} \geq 0.9$ (NIC IDs 269, 989, and 1211), Chen \& Marzke
1344: (2004) do not include these in their sample of evolved galaxies.  We
1345: derive a photometric redshift, $z=2.7$, for NIC ID 269, slightly below
1346: the $z > 2.8$ selection of Chen \& Marzke, which may be the reason
1347: they excluded it.  Chen \& Marzke interpret object NIC ID 1211 to be
1348: dominated by a dusty starburst.  However, the $\nich$ to \mone\ and
1349: \mtwo\ colors imply that while its rest-frame UV light stems from
1350: dust-obscured young stars, most of the mass ($>$88\% at 68\%
1351: confidence) resides in evolved stellar populations (see figure~8;
1352: however, we note that the upper limit on the X-ray detection of this
1353: galaxy may imply that it hosts an AGN, see \S~3.2).  NIC ID 989
1354: appears to have some ongoing star formation (with relatively blue
1355: $\acsb$ to $\acsz$ colors), but older stars appear to dominate the
1356: rest-frame optical and near-IR light (figure~7).  We suspect this
1357: object was excluded from the selection of Chen \& Marzke because of
1358: its relatively blue ACS colors.  Regardless, we find that data at
1359: wavelengths greater than rest-frame 1~\micron\ are required for
1360: interpreting the properties of the stellar populations of red
1361: galaxies.
1362: 
1363: Parenthetically, we note that there are two objects in the HUDF with
1364: $\nicj - \nich > 2$~mag (NIC IDs 1078, 1211) and one object has $\nicj
1365: - \nich = 1.95$~mag (NIC ID 1237).  All three have IRAC photometry.
1366: These objects have colors comparable to the ``unusual'' red, $\nicj -
1367: \nich$-selected object in the HDF-N reported by Dickinson et al.\
1368: (2000), which has $\nicj - \nich \simeq 2.3$~mag.  Based on our fits
1369: to the ACS, NICMOS, and IRAC photometry of the HUDF objects, we
1370: conclude that two (NIC IDs 1078, 1237) are dominated by dusty
1371: starbursts, while only one (NIC ID 1211) is dominated by a substantial
1372: population of old stars.  Dickinson et al.\ (2000) are unable to
1373: distinguish between these two possibilities (or, for that matter, the
1374: possibility that their object is a $z\approx 12$ Lyman-break-type
1375: galaxy).  This is partly because the HDF-N object is undetected in the
1376: \hst/WFPC2 data, nor did it have longer wavelength data from
1377: \spitzer/IRAC.  Two of the three objects here have \hst/ACS photometry
1378: $\gsim 1$~mag fainter than the detection limit of the WFPC2 data in
1379: the HDF-N.  Therefore, we conclude that one requires \textit{very}
1380: deep \hst\ optical imaging and \spitzer/IRAC imaging to constrain the
1381: nature of these very red, $\nicj - \nich \gsim 2$~mag objects.
1382: 
1383: None of our six candidates for galaxies with substantial population of
1384: evolved stellar stars is entirely devoid of recent star formation
1385: (i.e., all objects have $f_\mathrm{old} < 1.0$).  Indeed, in the
1386: entire HUDF there are \textit{no} candidates for galaxies consistent
1387: with a purely passively-evolving stellar population formed at much
1388: higher redshift (and in the NICMOS HDF-N, there is only one possible
1389: candidate, see discussion above; Dickinson et al.\ 2000).  Therefore,
1390: such objects are extremely rare if they exist at this redshift.  We
1391: conclude that at $z\gsim 2.5$ even galaxies with a substantial amount
1392: of evolved stellar populations still maintain some ongoing
1393: star-formation.  Papovich et al.\ (2005) argue based on galaxies'
1394: morphologies that at $z\gsim 2$, recurrent star-formation dominates
1395: the rest-frame UV and blue light, which will maintain some level of
1396: homogeneity in the galaxies' internal colors.  Our interpretation here
1397: is consistent with that scenario. However, we note that it appears
1398: that while the young stellar populations dominate the UV and blue
1399: light, in a small fraction of galaxies a substantial population of
1400: older stars dominates at rest-frame wavelengths $\gsim 1$~\micron.
1401: 
1402: \begin{figure*}[t]
1403:   \begin{center}
1404:     \scalebox{0.48}{{\includegraphics{fig9a.eps}}{\includegraphics{fig9b.eps}}}
1405:     \caption{$\nich$ magnitudes vs.\ photometric redshifts for the
1406:       subset of $(\nicj-\nich)_\mathrm{AB} > 1.0$~mag objects with
1407:       best-fit $f_\mathrm{old} > 0.9$, for objects without IRAC
1408:       photometry (left panel) and with IRAC photometry (right panel).
1409:       The dashed line shows the $\nich_\mathrm{AB} < 26$ magnitude
1410:       cut.  In the right-hand panel, cyan stars indicate those sources
1411:       among the IRAC--detected subsample with $f_\mathrm{old} > 0.9$
1412:       which have 24~\micron\ detections.  Thick--circles denote
1413:       objects with $f_\mathrm{old} > 0.9$ and $z_\mathrm{phot} > 2.5$}
1414:     \label{fig:co}
1415:   \end{center}
1416: \end{figure*}
1417: 
1418: \begin{figure*}[t]
1419:   \begin{center}
1420:     \scalebox{0.48}{{\includegraphics{fig10a.eps}}{\includegraphics{fig10b.eps}}}
1421:     \caption{Best-fit model masses and photometric redshifts for all
1422:       HUDF objects with $(\nicj-\nich)_\mathrm{AB} > 1.0$~mag (left
1423:       panel) and the subset of these with unblended IRAC photometry
1424:       (right panel).  The dashed line shows the limiting stellar mass
1425:       for galaxies with a stellar population formed in a burst at $z =
1426:       7$ with observed magnitude $\nich = 26.0$.  In both panels the
1427:       open circles indicate objects dominated by old stellar
1428:       populations ($f_\mathrm{old} > 0.9$).  In the right-hand panel,
1429:       cyan stars indicate those sources among the IRAC--detected
1430:       subsample with 24~\micron\ detections.  Thick--circles denote
1431:       those objects with $f_\mathrm{old} > 0.9$ and $z_\mathrm{phot} >
1432:       2.5$, which are the discussion of \S~4.3.}
1433:     \label{fig:co}
1434:   \end{center}
1435: \end{figure*}
1436: 
1437: \subsection{Galaxies with Evolved Stellar Populations at $z > 1.5$}
1438: 
1439: With the $\nicj - \nich > 1$~mag selection, we identify massive
1440: galaxies with substantial stellar mass and with SEDs dominated by the
1441: light from passively evolving stellar populations.  Figure~9 shows the
1442: measured $\nich$ magnitudes as a function of redshift for galaxies
1443: with $f_\mathrm{old} > 0.9$, for objects with IRAC photometry (right
1444: panel) and without IRAC photometry (left panel).  Although the data
1445: should be sensitive to such objects down to our selection limit,
1446: $H_{160} < 26$ mag, only one such object has $25 < H_{160} < 26$ mag
1447: --- the rest have $H_{160} < 25$ mag.  Figure~10 shows the derived
1448: stellar masses as a function of galaxy redshift for all galaxies in
1449: the $\nicj - \nich > 1$~mag sample.  The figure shows the stellar
1450: masses derived for the whole sample using only the ACS through NICMOS
1451: data (left panel), and it shows the stellar masses derived including
1452: the IRAC data for the IRAC--detected subsample (right panel).  Objects
1453: with more than 90\% of their stellar mass in evolved stellar
1454: populations are circled in the figure.  There exist objects whose
1455: light (and stellar mass) is dominated by passively evolving stellar
1456: populations with total stellar masses up to and exceeding
1457: $>$$10^{11}$~\msol\ over the entire redshift range.  The existence of
1458: these galaxies is not entirely unexpected, given that such objects
1459: likely dominate the stellar mass density even at $z\sim 2$ (e.g.,
1460: Rudnick et al.\ 2006; van Dokkum et al.\ 2006).
1461: 
1462: However, there is an apparent dearth of objects whose light is
1463: dominated by passively evolving stellar populations with lower stellar
1464: masses.  For example, none of the 10 IRAC--detected galaxies with
1465: $z_\mathrm{phot} > 1.6$ and with more than 90\% of their stellar
1466: masses evolved stellar populations are fainter than $\nich > 25$~mag
1467: (see figure~9), which is one magnitude above our detection limit.  (We
1468: note there is one object with $z_\mathrm{phot} \simeq 1.5$ and $H =
1469: 25.5$~mag, and a derived total stellar mass $4\times 10^{9}$~\msol,
1470: see figure~9.)  Therefore, at these high redshifts, the candidates for
1471: passively evolving galaxies are already massive.  In terms of total
1472: stellar mass, the right-hand panel of figure~10 shows that all the
1473: IRAC--detected objects with $f_\mathrm{old} > 0.9$ and $z > 1.5$ have
1474: stellar masses greater than $\sim$$2 \times 10^{10}$~\msol.  If we
1475: instead consider the results for galaxies from their stellar masses
1476: derived without the IRAC data, there is only a single candidate as a
1477: low-mass galaxy whose mass is dominated by passively evolved stellar
1478: populations (and we consider the stellar masses derived in the old
1479: component without IRAC data to be less robust).
1480: 
1481: The data should be sensitive to objects dominated by old stellar
1482: populations with stellar masses with $\sim$$3\times 10^{9}$ to
1483: $10^{10}$~\msol (see figure 4 and figure 10).  We note that this is
1484: not a bias related to the sources with IRAC detections.  The IRAC data
1485: from GOODS (5.8 \micron\ flux density of 0.11~$\mu$Jy, $5\sigma$)
1486: would detect a passively evolving stellar population formed at z=7 and
1487: M $> 10^{9.5}$~$\msun$ for all redshifts considered here.
1488: 
1489: Any plausible mass function for galaxies at $z\sim 2$ would predict a
1490: greater number of galaxies with lower stellar masses than at high
1491: stellar mass.  For example, using the Fontana et al.\ (2006) empirical
1492: galaxy mass function, we would expect roughly double the number of
1493: galaxies with $0.5 - 2.0 \times 10^{10}$~\msol\ compared to those with
1494: $>$$2 \times 10^{10}$~\msol.  This strongly contrasts with the
1495: findings here for the red galaxies dominated by passively evolving
1496: stellar populations.  Therefore, the data suggest that galaxies with
1497: more than 90\% of their stellar mass in evolved stellar populations
1498: have a minimum mass threshold of roughly several times
1499: $10^{10}$~\msol\ at $z\sim 2$.  The lack of such objects with lower
1500: masses may be related to the processes governing stellar mass
1501: build--up in galaxies at these redshifts.
1502: 
1503: 
1504: \subsection{Constraints on the Densities of Galaxies with Evolved
1505: Stellar Populations at $2.5 < z < 3.5$}
1506: 
1507: With the full multiwavelength dataset, we have separated our sample
1508: into those galaxies that are dominated by dust-enshrouded starbursts
1509: and those galaxies whose rest-frame optical and near-IR emission is
1510: consistent with being dominated by a substantial population of evolved
1511: stars (see, e.g., section 4.2, above).  Of the 16 $\nicj - \nich >
1512: 1$~mag galaxies with robust IRAC photometry, 6 have best-fit models
1513: favoring a substantial population of evolved stellar populations,
1514: which is 37.5\% of the sample.  Extending this to the full sample of
1515: 32 galaxies with $\nicj - \nich > 1$~mag, we estimate there should be
1516: 12 such objects in the area of the HUDF.  Using these numbers, we
1517: place constraints on the number and mass density for these types of
1518: objects.  If some of our 6 candidates for evolved stellar populations
1519: turned out to be dusty starbursts, our numbers would be reduced.  Due
1520: to the lower mass to light ratios of younger models, the fitted masses
1521: would also be reduced.  Therefore the densities derived here should be
1522: regarded as upper limits.
1523: 
1524: The HUDF data reach very deep limits in stellar mass for galaxies
1525: dominated by evolving stellar populations.  For a galaxy whose stellar
1526: mass formed \textit{in situ} at $z_f=7$, the $\nich = 26.0$~mag
1527: detection limit corresponds to a limit in stellar mass ranging from
1528: $6\times 10^9$~\msol\ at $z=2$ to $1\times 10^{11}$~\msol\ at $z=4$
1529: (see figures~4 and 10).  Although the $(\nicj - \nich) > 1$~mag color
1530: selection should be, in principle, sensitive to old stellar
1531: populations at $z \gsim 2$, in practice we find candidates for
1532: galaxies dominated by passively evolving stellar populations
1533: $f_\mathrm{old} > 0.94$ at $z\geq 2.7$.  Likewise, at $z > 3.5$, the
1534: Balmer/4000\AA-breaks move beyond the F160W bandpass, making our color
1535: selection less efficient.  Therefore, to set an upper limit on the
1536: number and mass densities of galaxies dominated by passively evolving
1537: stellar populations, we limit our redshift range to $2.5 < z < 3.5$.
1538: We note that requiring an IRAC detection does not limit our stellar
1539: mass selection; the limiting stellar mass for the IRAC GOODS data
1540: sensitivity limit is always less than that for the NICMOS limit at
1541: these redshifts (see above).
1542: 
1543: The volume in the area of the HUDF is $1.9 \times 10^{4}$~Mpc$^3$
1544: between $2.5 < z < 3.5$.  Restricting our analysis to only those
1545: galaxies with $2.5 < z < 3.5$ we derive a number density of
1546: 3.1$\times$$10^{-4}$~Mpc$^{-3}$, or 6.3$\times$$10^{-4}$~Mpc$^{-3}$
1547: including our 50\% incompleteness.  However, these six objects are not
1548: visible over the whole redshift range.  In the number density
1549: calculation, we compute the effective volume, $V_\mathrm{eff}$ for
1550: each galaxy from $z=2.5$ to $z_\mathrm{eff}$, where $z_\mathrm{eff}$
1551: is the lesser of $z=3.5$ and the maximum redshift to which the galaxy
1552: could be detected.  The number density derived by summing $\Sigma_i
1553: 1/V_\mathrm{eff}^{(i)}$ over all $i$ galaxies gives $3.3^{+1.0}_{-1.5}
1554: \times 10^{-4}$~Mpc$^{-3}$, and after accounting for incompleteness,
1555: $n = 6.6^{+2.0}_{-3.0} \times 10^{-4}$~Mpc$^{-3}$.
1556: 
1557: Assuming these objects will evolve to become present-day, passively
1558: evolving red galaxies, then this number density is significantly lower
1559: than what we would expect using local mass functions.  Using the
1560: $z\sim 0$ mass function parameters listed in table 4 of
1561: \citet{bell03}, we expect the number density of red, early-type
1562: galaxies to be $n(z=0) = 3.5\times 10^{-3}$~Mpc$^{-3}$ in the volume
1563: spanned between $2.5 < z < 3.5$ in the HUDF (and again taking the
1564: redshift-dependent mass limit from
1565: figure~\ref{fig:masslimit})\footnote{Note that \citep{bell03} assumed
1566: a formation redshift, $z_f = 4$, in their mass functions.  This is
1567: only marginally lower than $z_f=7$ used in this analysis.  However,
1568: the change in lookback time from $z\approx 0$ to 4 relative to that
1569: from $z\approx 0$ to 7 corresponds to a change in the mass-to-light
1570: ratio of $M/L_B = 5.8$ to 6.2.  Applying our higher formation redshift
1571: to Bell et al.\ would increase their stellar masses by $\approx 5$\%,
1572: which is negligible compared to other uncertainties.}.  This is larger
1573: than the observed number density (including incompleteness) by roughly
1574: a factor of 5.  Therefore, while we would have expected 65 objects in
1575: the HUDF (i.e., 32 after our $50\%$ completeness) we have found only
1576: 6.
1577: 
1578: This difference is substantially larger than the Poisson noise ($\pm
1579: \sqrt{6}/6$, i.e., 40\%).  It may be that these objects suffer strong
1580: field-to-field clustering (possibly expected for red galaxies, e.g.,
1581: Daddi et al.\ 2003; Quadri et al. 2007).  We estimate the number error
1582: from density variations in the volume between $2.5 < z < 3.5$ in the
1583: HUDF as $0.3\times \sigma_8$.  While we have no measurement for
1584: $\sigma_8$ for these galaxies, based on theoretical models we expect
1585: $\sigma_8 \sim 1.0$ \citep{adelberger05}, implying that the error from
1586: density variations is 30\%.  Summing these errors in quadrature brings
1587: our total error estimate to $\approx$50\% for the expected number of
1588: these galaxies in the HUDF.  Even so, the difference between the 32
1589: expected galaxies and the 6 we find is substantially larger than the
1590: combined errors from counting statistics and cosmic variance.
1591: Therefore, by number, and recalling that these candidates only define
1592: upper limits, the objects in the HUDF which are dominated by evolved
1593: stellar populations make up less than one-third of present-day red,
1594: early-type galaxies.  Thus, at least two-thirds of present-day red,
1595: early-type galaxies must assemble and/or evolve into their current
1596: configuration at $z \lsim 2.5$.
1597: 
1598: Much of the stellar mass density in galaxies $2.5 < z < 3.5$ could
1599: reside in red-selected galaxies, whose light is dominated by older,
1600: passively evolving stellar populations. For the $\nicj - \nich >
1601: 1$~mag galaxies in the HUDF, we derive an upper limit on the stellar
1602: mass density of $\log (\rho^\ast / [M_\odot \mathrm{Mpc}^{-3}] ) =
1603: 7.4\pm{0.3}$, including the effect of our redshift-evolving mass
1604: limit, where our errors are derived from a Monte--Carlo simulation of
1605: the data.  When we include our 50\% incompleteness, our total stellar
1606: mass density becomes $\log (\rho^\ast / [M_\odot \mathrm{Mpc}^{-3}] )
1607: = 7.7\pm{0.3}$.  Poisson sampling of the 6 objects gives a factor of
1608: two error and therefore is the dominant source of uncertainty in the
1609: mass density estimate.  We consider this measurement to be an upper
1610: limit on the mass density of distant evolved galaxies for two reasons:
1611: a) it is possible that some of our evolved galaxy candidates are in
1612: fact dusty starbursts, and b) because our two-component model fit
1613: incorporates an a priori upper limit on the mass--to--light ratio in
1614: the form of a maximally old SSP component (see \S 3).
1615: 
1616: \citet{rudnick03} derive a stellar mass density at $z =
1617: 2.80^{+0.40}_{-0.39}$ of log $\rho_* = 7.5^{+0.1}_{-0.1}$ and at $z =
1618: 2.01^{+0.40}_{-0.41}$ of log $\rho_* = 7.5^{+0.1}_{-0.1}$.  Because
1619: our mass density provides an upper limit on the mass limit, we
1620: conclude that our derived mass density is consistent those of Rudnick
1621: et al.  In this case, it may be that passively stellar populations
1622: dominate the stellar mass density.  This is consistent with the
1623: analysis of Dickinson et al.\ (2003) who used two--component
1624: star--formation history models to set an upper limit on the stellar
1625: mass density from passively evolving stellar populations hidden
1626: beneath the glare of younger stars in galaxies at $z < 3$.  Dickinson
1627: et al.\ find that at $z\sim 2.5-3$ an upper limit on the stellar mass
1628: density of $\log (\rho^\ast / [\msol \mathrm{Mpc}^{-3}]) =
1629: 7.89^{+0.20}_{-0.15}$.  If this reflects reality, then based on our
1630: analysis, almost all of this mass density could reside in faint,
1631: red--selected galaxies at these redshifts.  \citet{vd06} found that at
1632: $z\sim 2$ red-selected galaxies with masses greater than
1633: $10^{11}$~\msol\ dominate the mass density.  Our analysis of the HUDF
1634: galaxies suggests that at lower stellar masses red-selected galaxies
1635: still contribute significantly to the total mass budget.
1636: 
1637: Lastly, we stress that with only rest-frame UV and optical data, it is
1638: not possible to discriminate between those galaxies dominated by dusty
1639: starbursts and older stellar populations (e.g., Moustakas et al.\
1640: 2004; Smail et al.\ 2002).  Although the IRAC data does not resolve
1641: this degeneracy for all galaxies, we show that a sub-set of objects do
1642: have consistent best-fit results and we therefore use these objects in
1643: our analysis.
1644: 
1645: We wish to thank our colleagues at Steward Observatory for stimulating
1646: conversations.  In particular we would like to thank George Rieke for
1647: useful comments, Xiaohui Fan for generously providing models of the
1648: IGM attenuation, and Karl Gordon for his suggestions and help with the
1649: dust models used in this work.  AMS would also like to thank Richard
1650: Cool, Aleks Diamond-Stanic, Brandon Kelly, Juna Kollmeier, Andy
1651: Marble, John Moustakas, Jane Rigby and Yong Shi for helpful
1652: discussions.  Support for CJP was provided by NASA through the Spitzer
1653: Space Telescope Fellowship Program, through a contract issued by the
1654: Jet Propulsion Laboratory (JPL), California Institute of Technology
1655: under a contract with NASA.  DJE is also supported by an Alfred P.\
1656: Sloan Research Fellowship.
1657: 
1658: 
1659: \begin{thebibliography}{}
1660: 
1661: 
1662: \bibitem[Adelberger et al.(2005)]{adelberger05} Adelberger, K.~L., 
1663:   Steidel, C.~C., Pettini, M., Shapley, A.~E., Reddy, N.~A., \& Erb, D.~K.\ 
1664:   2005, \apj, 619, 697 
1665:   
1666: \bibitem[Alexander et al.(2003)]{alex03} Alexander, D.~M., et 
1667:   al.\ 2003, \aj, 126, 539 
1668:     
1669: \bibitem[Baugh et al.(2003)]{baugh03} Baugh, C.~M., Benson, 
1670:   A.~J., Cole, S., Frenk, C.~S., \& Lacey, C.\ 2003, The Mass of Galaxies at 
1671:   Low and High Redshift, 91
1672: 
1673: \bibitem[Beckwith et al.(2006)]{beck06} Beckwith,
1674: S.~V.~W., et 
al.\ 2006, \aj, 132, 1729 
1675: 
1676: \bibitem[Bell et al.(2003)]{bell03} Bell, E.~F., McIntosh, 
1677:   D.~H., Katz, N., \& Weinberg, M.~D.\ 2003, \apjs, 149, 289 
1678:   
1679: \bibitem[Bell et al.(2004)]{bell04} Bell, E.~F., et al.\ 2004, 
1680:   \apj, 608, 752 
1681:   
1682: \bibitem[Bertin \& Arnouts(1996)]{bertin96} Bertin, E., \& 
1683:   Arnouts, S.\ 1996, \aaps, 117, 393 
1684: 
1685: \bibitem[Brammer \& van Dokkum(2007)]{brammer07} Brammer,
1686: G.~B., 
1687: \& van Dokkum, P.~G.\ 2007, \apjl, 654, L107
1688: 
1689: \bibitem[Bruzual \& Charlot(2003)]{bc03} Bruzual, G., \& 
1690:   Charlot, S.\ 2003, \mnras, 344, 1000 
1691:   
1692: \bibitem[Calzetti et al.(2000)]{dc00} Calzetti, D., Armus, 
1693:   L., Bohlin, R.~C., Kinney, A.~L., Koornneef, J., \& Storchi-Bergmann, T.\ 
1694:   2000, \apj, 533, 682 
1695: 
1696: \bibitem[Chary et al.(2007)]{2007ApJ...665..257C} Chary, R.-R.,
1697: Teplitz, 
H.~I., Dickinson, M.~E., Koo, D.~C., Le Floc'h, E.,
1698: Marcillac, D., 
Papovich, C., \& Stern, D.\ 2007, \apj, 665, 257
1699:   
1700: \bibitem[Chen \& Marzke(2004)]{chen04} Chen, H.-W., \& Marzke, 
1701:   R.~O.\ 2004, \apj, 615, 603 
1702:   
1703: \bibitem[Cimatti et al.(2002)]{cimatti02} Cimatti, A., et al.\ 
1704:   2002, \aap, 381, L68 
1705:     
1706: \bibitem[Croton et al.(2006)]{croton06} Croton, D.~J., et al.\ 
1707:   2006, \mnras, 365, 11 
1708:   
1709: \bibitem[Daddi et al.(2005)]{daddi05} Daddi, E., et al.\ 2005, 
1710:   \apj, 626, 680 
1711:   
1712: \bibitem[Dav{\'e}(2005)]{dave06} Dav{\'e}, R., et al.\ 
1713:   2005, astro-ph\/0511532 
1714:   
1715: \bibitem[De Lucia et al.(2006)]{delucia06} De Lucia, G., 
1716:   Springel, V., White, S.~D.~M., Croton, D., \& Kauffmann, G.\ 2006, \mnras, 
1717:   366, 499 
1718:   
1719: \bibitem[Dickinson et al.(2000)]{dickinson00} Dickinson, M., et 
1720: al.\ 2000, \apj, 531, 624 
1721:     
1722: \bibitem[Dickinson et al.(2003)]{dickinson03} Dickinson, M., 
1723:   Papovich, C., Ferguson, H.~C., \& Budav{\'a}ri, T.\ 2003, \apj, 587, 25 
1724: 
1725: \bibitem[Dunlop et al.(2007)]{dunlop07} Dunlop, J.~S., Cirasuolo, M.,
1726: \& McLure, R.~J.\ 2007, \mnras, 376, 1054
1727: 
1728: \bibitem[Fan(1999)]{fan99} Fan, X.\ 1999, \aj, 117, 2528 
1729:   
1730: \bibitem[Fazio et al.(2004)]{fazio04} Fazio, G.~G., et al.\ 
1731:   2004, \apjs, 154, 10 
1732: 
1733: \bibitem[Fontana et al.(2006)]{fontana06} Fontana, A., et al.\ 
1734: 2006, \aap, 459, 745 
1735:   
1736: \bibitem[F{\"o}rster Schreiber et al.(2004)]{fs04} 
1737:   F{\"o}rster Schreiber, N.~M., et al.\ 2004, \apj, 616, 40 
1738:     
1739: \bibitem[Franx et al.(2003)]{franx03} Franx, M., et al.\ 2003, 
1740:   \apjl, 587, L79
1741:   
1742: \bibitem[Freedman et al.(2001)]{freedman01} Freedman, W.~L., et 
1743:   al.\ 2001, \apj, 553, 47 
1744:     
1745: \bibitem[Glazebrook et al.(2004)]{glazebrook04} Glazebrook, K., et 
1746:   al.\ 2004, \nat, 430, 181 
1747:   
1748: \bibitem[Hernquist \& Springel(2003)]{hernquist03} Hernquist, L., 
1749:   \& Springel, V.\ 2003, \mnras, 341, 1253 
1750:   
1751: \bibitem[Hopkins(2004)]{hopkins04} Hopkins, A.~M.\ 2004, \apj, 
1752:   615, 209
1753:     
1754: \bibitem[Knudsen et al.(2005)]{knudsen05} Knudsen, K.~K., et al.\ 
1755:   2005, \apjl, 632, L9 
1756:   
1757: \bibitem[Kriek et al.(2006)]{kriek06} Kriek, M., et al.\ 
1758:   2006, astro-ph/0603063 
1759:   
1760: \bibitem[Labb{\'e} et al.(2003)]{labbe03} Labb{\'e}, I., et 
1761:   al.\ 2003, \aj, 125, 1107 
1762:   
1763: \bibitem[Labb{\'e} et al.(2005)]{labbe05} Labb{\'e}, I., et 
1764:   al.\ 2005, \apjl, 624, L81 
1765: 
1766: \bibitem[Madau(1995)]{pm95} Madau, P.\ 1995, \apj, 441, 18 
1767:   
1768: \bibitem[McCarthy(2004)]{mccarthy04} McCarthy, P.~J.\ 2004, \araa, 
1769:   42, 477 
1770: 
1771: \bibitem[Mobasher et al.(2005)]{mobasher05} Mobasher, B., et
1772:   al.\ 
2005, \apj, 635, 832
1773: 
1774: \bibitem[Moustakas et al.(2004)]{moustakas04} Moustakas, L.~A., et 
1775:   al.\ 2004, \apjl, 600, L131 
1776:   
1777: \bibitem[Papovich et al.(2001)]{cp01} Papovich, C., 
1778:   Dickinson, M., \& Ferguson, H.~C.\ 2001, \apj, 559, 620 
1779:   
1780: \bibitem[Papovich et al.(2004)]{cp04} Papovich, C., et al.\ 
1781:   2004, \apjs, 154, 70 
1782:     
1783: \bibitem[Papovich et al.(2006)]{papovich06} Papovich, C., et al.\ 
1784:   2006, \apj, 640, 92 
1785: 
1786: \bibitem[Quadri et al.(2007)]{quadri07} Quadri, R., et al.\ 
1787: 2007, \apj, 654, 138 
1788: 
1789: \bibitem[Rubin et al.(2004)]{rubin04} Rubin, K.~H.~R., van 
1790:   Dokkum, P.~G., Coppi, P., Johnson, O., F{\"o}rster Schreiber, N.~M., Franx, 
1791:   M., \& van der Werf, P.\ 2004, \apjl, 613, L5 
1792: 
1793: \bibitem[Reddy et al.(2005)]{reddy05} Reddy, N.~A., Erb, D.~K., 
1794:   Steidel, C.~C., Shapley, A.~E., Adelberger, K.~L., \& Pettini, M.\ 2005, 
1795:   \apj, 633, 748 
1796: 
1797: \bibitem[Rudnick et al.(2003)]{rudnick03} Rudnick, G., et al.\ 
1798:   2003, \apj, 599, 847
1799: 
1800: \bibitem[Rudnick et al.(2006)]{rudnick06} Rudnick, G., et al.\ 
1801: 2006, \apj, 650, 624 
1802: 
1803: \bibitem[Saracco et al.(2001)]{saracco01} Saracco, P., Giallongo, 
1804: E., Cristiani, S., D'Odorico, S., Fontana, A., Iovino, A., Poli, F., \& 
1805: Vanzella, E.\ 2001, \aap, 375, 1 
1806:  
1807: \bibitem[Shapley et al.(2001)]{shapley01} Shapley, A.~E., 
1808:   Steidel, C.~C., Adelberger, K.~L., Dickinson, M., Giavalisco, M., \& 
1809:   Pettini, M.\ 2001, \apj, 562, 95 
1810: 
1811: \bibitem[Shapley et al.(2005)]{shapley05} Shapley, A.~E., 
1812:   Steidel, C.~C., Erb, D.~K., Reddy, N.~A., Adelberger, K.~L., Pettini, M., 
1813:   Barmby, P., \& Huang, J.\ 2005, \apj, 626, 698 
1814: 
1815: \bibitem[Smail et al.(2002)]{smail02} Smail, I., Owen, F.~N., 
1816:   Morrison, G.~E., Keel, W.~C., Ivison, R.~J., \& Ledlow, M.~J.\ 2002, \apj, 
1817:   581, 844
1818: 
1819: \bibitem[Somerville et al.(2001)]{somerville01} Somerville, R.~S., 
1820:   Primack, J.~R., \& Faber, S.~M.\ 2001, \mnras, 320, 504 
1821:   
1822: \bibitem[Somerville et al.(2004)]{somerville04} Somerville, R.~S., 
1823:   et al.\ 2004, \apjl, 600, L135 
1824: 
1825: \bibitem[Spergel et al.(2003)]{sp03} Spergel, D.~N., et al.\ 
1826:   2003, \apjs, 148, 175 
1827:   
1828: \bibitem[Thompson et al.(2005)]{thompson05} Thompson, R.~I., et 
1829:   al.\ 2005, \aj, 130, 1 
1830:   
1831: \bibitem[Toft et al.(2005)]{toft05} Toft, S., van Dokkum, P., 
1832:   Franx, M., Thompson, R.~I., Illingworth, G.~D., Bouwens, R.~J., \& Kriek, 
1833:   M.\ 2005, \apjl, 624, L9 
1834:   
1835: \bibitem[van Dokkum et al.(2003)]{vd03} van Dokkum, P.~G., 
1836:   et al.\ 2003, \apjl, 587, L83
1837: 
1838: \bibitem[van Dokkum et al.(2006)]{vd06} van Dokkum, P.~G., 
1839:   et al.\ 2006, \apjl, 638, L59 
1840:   
1841: \bibitem[Webb et al.(2006)]{webb06} Webb, T.~M.~A., et al.\ 
1842:   2006, \apjl, 636, L17 
1843:   
1844: \bibitem[Weinberg et al.(2004)]{weinberg04} Weinberg, D.~H., 
1845:   Dav{\'e}, R., Katz, N., \& Hernquist, L.\ 2004, \apj, 601, 1 
1846: 
1847: \bibitem[Williams et al.(1996)]{williams96} Williams, R.~E., et 
1848:   al.\ 1996, \aj, 112, 1335 
1849:   
1850: \bibitem[Wilson et al.(2004)]{wilson04} Wilson, G., et al.\ 
1851:   2004, \apjs, 154, 107 
1852:  
1853: \bibitem[Wilson et al.(2007)]{wilson07} Wilson, G., et al.\ 
2007,
1854: \apjl, 660, L59
1855:  
1856: \bibitem[Witt \& Gordon(2000)]{witt00} Witt, A.~N., \& Gordon, 
1857:   K.~D.\ 2000, \apj, 528, 799 
1858:   
1859: \bibitem[Yan \& Windhorst(2004)]{yan04c} Yan, H., \&
1860: Windhorst, R.~A.\ 2004, \apjl, 612, L93
1861: 
1862: \bibitem[Yan et al.(2004a)]{yan04a} Yan, H., et al.\ 2004a, 
1863:   \apj, 616, 63 
1864:   
1865: \bibitem[Yan et al.(2004b)]{yanl04b} Yan, L., et al.\ 2004b, 
1866:   \apjs, 154, 75 
1867:   
1868: \end{thebibliography}
1869: 
1870: \end{document}
1871: