1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{natbib}
3: %\usepackage{emulateapj5}
4: \newcommand{\etal}{et al.}
5: \shorttitle{Globular Cluster Formation in M82}
6: \shortauthors{Lipscy \& Plavchan}
7:
8: %\received{2003 September 25}
9: \begin{document}
10: \title{Globular Cluster Formation in M82}
11: \author{S.~J. Lipscy and P. Plavchan}
12: \affil{Department of Physics and Astronomy, University of California,\\
13: Box 951547, Knudsen Hall, Los Angeles, CA 90095-1562; \\
14: lipscy@astro.ucla.edu; plavchan@astro.ucla.edu}
15:
16:
17: \begin{abstract}
18: We present high resolution mid-infrared (mid-IR; 11.7 and 17.65
19: $\micron$) maps of the central 400 pc region of the starburst galaxy
20: M82. Seven star forming clusters are identified which together
21: provide $\sim$ 15\% of the total mid-IR luminosity of the galaxy.
22: Combining the mid-IR data with thermal radio measurements and near-
23: and mid-IR line emission, we find that these young stellar clusters
24: have inferred masses and sizes comparable to globular clusters. At
25: least 20\% of the star formation in M82 is found to occur in
26: super-star clusters.
27: \end{abstract}
28:
29: \keywords{galaxies: individual(M82) --- galaxies: starburst --- galaxies: star
30: clusters --- infrared: galaxies}
31:
32:
33: \section{Introduction}
34:
35: M82 (NGC 3034) is often considered the archetypical starburst galaxy
36: since it has a derived star formation rate ($\sim$10 M$_{\odot}$
37: yr$^{-1}$; \citealt{om78}) that would deplete the observed molecular
38: gas in $<$ 10$^8$ yrs (i.e. short on a Hubble timescale;
39: \citealt{lord96}). The starburst phenomena traces recent star
40: formation since it is the massive stars ($>$8 M$_{\odot}$), which have
41: short lives, that dominate the energetic output of the host galaxy.
42: The nuclear starburst of M82 dominates the infrared (IR) luminosity of
43: the galaxy - essentially all of the galaxy's L$_{IR} \sim$ 3.6
44: $\times$ 10$^{10}$ L$_{\odot}$\footnote{\citet{telesco80} published
45: L$_{IR}$ = L(1-300$\micron$) = 3.0 $\times$ 10$^{10}$ L$_{\odot}$
46: assuming a distance to M82 of 3.3 Mpc. We have updated L$_{IR}$ for
47: the distance of 3.6 Mpc \citep{freedman94, sakai99} used in this
48: paper.} comes from the central kpc \citep{telesco80}.
49:
50: Optical and near-IR imaging of M82 has revealed numerous
51: super-star clusters in its active star-forming nucleus. Ground-based
52: optical imaging of the central region detected eight young knots
53: distributed throughout the region \citep{om78}. Further study of
54: these knots broke them into smaller star clusters (half light
55: diameters $\sim$3-4 pc) and suggested cluster dynamical masses
56: 10$^4$-10$^6$ M$_{\odot}$ \citep{smith01, degrijs01, mccrady03}.
57:
58: M82's current starburst is thought to have been triggered by its
59: interaction with M81 $\sim$10$^8$ yrs ago
60: \citep{cottrell77,achtermannlacy95}. It has been argued that the
61: conditions resulting from interactions and mergers of galaxies are
62: favorable for globular cluster formation (eg. \citealt{taniguchi99}
63: and ref. therein). UV studies of global properties of starbursting
64: galaxies have shown that as much as 20\% of the light is produced in
65: the luminous knots \citep{meurer95, zepf99}, suggesting a high
66: efficiency of cluster formation in starbursts. Understanding the star
67: formation occurring in M82's nuclear region can provide insight into
68: both the general process of star formation in starburst environments
69: and also the process and efficiency of forming super-star
70: clusters.
71:
72: The nearly edge-on geometry of M82 combined with heavy optical
73: extinction (A$_V \sim$ 5-25 mag; \citealt{lester90, telesco91,
74: larkin94, satyapal95}) has made the galaxy's central 400 pc difficult
75: to study. Even in the near-IR where A$_{2.2 \micron} \sim$ 1/10A$_V$
76: \citep{rieke85}, it is difficult to directly measure the deepest star
77: forming regions. The mid-IR region of the spectrum is essential
78: for probing deep into the central regions of M82 and revealing the
79: details of the intense star formation occurring in the central
80: regions. Previously, the highest resolution mid-IR map of the central
81: region of M82 was a 12.4 $\micron$ map by \citet{telescogezari92},
82: which had a resolution $\sim$1.1$\arcsec$. Mid-IR maps of M82 with
83: lower resolution have also been published by \citet{rieke80},
84: \citet{telesco89}, \citet{deitz89}, \citet{telesco91} and
85: \citet{forsterschreiber03}. In this paper, we improve upon previous
86: observations by presenting higher resolution mid-IR maps of M82. We
87: discuss evidence suggesting the sources in the maps are young
88: counterparts to globular clusters and estimate the efficiency of
89: super-star cluster formation in M82.
90:
91: \section{Observations}
92:
93: On 2003 April 23, we imaged M82 at 3.5 (L-band), 11.7, and 17.65
94: $\micron$ with the Long Wavelength Spectrograph (LWS;
95: \citealt{jones93}), a facility instrument at the W.M. Keck
96: Observatory. LWS uses a 128 $\times$ 128 pixel Boeing Si:As detector
97: and has a plate scale of 0$\farcs$08 pixel$^{-1}$, resulting in a
98: 10$\farcs$2 $\times$ 10$\farcs$2 field of view. We used the
99: ``chop-nod'' mode of observing, with a chop throw of 15$\arcsec$
100: north. The bad pixels in the images have been smoothed over, and a
101: mask has been applied to remove the portion of the chip not
102: illuminated by the source. The seeing varied during the course of the
103: observations so the resolution in individual frames ranged from
104: 0.4-1.0$\arcsec$ at 11.7 $\micron$ and from 0.5-0.7$\arcsec$ at 17.65
105: $\micron$. At each of seven pointings across M82's nuclear region, we
106: imaged at all three wavelengths before moving to the next pointing.
107: Images from the seven pointings were mosaiced by centroiding on the
108: bright sources in each frame. Assigning coordinates to the field was
109: accomplished by aligning 2MASS sources with the centroids of the two
110: bright sources in the L-band frames. We estimate that our positions
111: are accurate to $\sim$0$\farcs$5. The star $\mu$ UMa was used as the
112: primary standard for flux density and point-spread function
113: calibration, and $\alpha$ Her and $\eta$ Sgr were used to estimate a
114: calibration error of $\sim$ 20\%.
115:
116: \section{Characterization of mid-IR sources}
117:
118: Our mid-IR maps, presented as Figures 1 and 2, contain several bright,
119: resolved sources as well as diffuse emission connecting the brighter
120: sources. The mid-IR sources, labeled A-G from west to east, are
121: denoted by black circles (with radii 0$\farcs$5, corresponding to the
122: positional error). For reference, the 2 $\micron$ peak is marked in
123: the Figures as a yellow cross \citep{deitz86} and the dynamical center
124: measured by \citealt{lester90} is marked with a yellow circle. The
125: sharp edges visible in the maps (eg. southeast of source B in the
126: 11.7 $\micron$ map) are artifacts that resulted from the mosaicing of
127: images with differing thermal backgrounds and do not affect the results
128: of this paper.
129:
130: The earlier, lower resolution 12.4 $\micron$ map by
131: \citet{telescogezari92} contained two bright regions, one to either
132: side of the galaxy's center. The overall structure in our higher
133: resolution maps is comparable, though we identify individual resolved
134: sources within the Telesco \& Gezari sources. To compare their flux
135: densities with those previously published, Telesco \& Gezari reported
136: a flux of 6.5 $\pm$ 0.7 Jy at 12.4 $\micron$ for the 4$\arcsec \times$
137: 4$\arcsec$ region containing our sources C and D. We measure a flux
138: for the same region of 5.3 $\pm$ 1.3 Jy at 11.7 $\micron$, in
139: agreement with the previous observation.
140:
141: The flux densities and half intensity major and minor axes measured
142: from the maps for the mid-IR sources are listed in Table 1. For
143: sources covered by more than one image, photometry was performed on
144: each image individually and the average of the measurements is
145: presented here. The standard deviation of the measurements due to
146: background variations was $\leq$ 15\% and, combined quadratically with
147: the calibration error, results in an overall flux density error of
148: $\leq$ 25\% in our measurements.
149:
150: Using our 11.7 $\micron$ and 17.65 $\micron$ flux densities, we
151: estimate the color temperature of the mid-IR emitting dust to be in
152: the range 150-270 K, assuming the dust particles radiate as
153: blackbodies. The mid-IR luminosities (L$_{MIR} \equiv$ L$_{12-18
154: \micron}$) of the sources, estimated by fitting the measured flux
155: densities to a blackbody with the color temperatures for each source,
156: are all between 0.2-6 $\times$ 10$^{8}$ L$_{\odot}$ and sum to 2.4
157: $\times$ 10$^{9}$ L$_{\odot}$. L$_{MIR}$ for the entire galaxy
158: estimated using the same procedure with the uncorrected IRAS 12 and 25
159: $\micron$ flux densities of 53 Jy and 274 Jy, respectively, gives 1.7
160: $\times$ 10$^{10}$ L$_{\odot}$, thus the seven mid-IR sources
161: contribute $\sim$15\% of M82's total L$_{MIR}$.
162:
163: Using radio images at five frequencies, \citet{allen99} created
164: spectrally decomposed images of thermal (free-free) and non-thermal
165: (synchrotron) emission. We find that not only do all the centers of
166: the thermal H II regions (marked in Figures 1 \& 2 with red squares)
167: match the centers of the mid-IR sources reasonably well, the diffuse
168: structure in the mid-IR maps follows closely the structure in the
169: thermal free-free map. \citet{golla96} also noted the correspondence
170: of the diffuse 1.5 \& 22 GHz emission with the \citet{telescogezari92}
171: 12.4 $\micron$ emission. This supports the hypothesis that the mid-IR
172: sources are heavily obscured H II regions.
173:
174: The [Ne II] (12.8 $\micron$) map published by
175: \citet{achtermannlacy95}, with a resolution of $\sim$1$\arcsec$, also
176: correlates well with our mid-IR maps (see Figure 3). It should be
177: noted that while the peak of the [Ne II] line lies within our 11.7
178: $\micron$ filter bandpass ($\Delta\lambda$ = 2.4 \micron), based on
179: the 5-16.5 $\micron$ spectra presented in \citet{forsterschreiber03},
180: the [Ne II] line contributes $\leq$15\% of the flux in our bandpass.
181: We interpret the correlation between the mid-IR and the [Ne II]
182: emission as confirmation that the source of the mid-IR emission is
183: dust heated by UV from young stars which ionize Ne I. Further
184: evidence to support this hypothesis comes from the Br$\gamma$ map from
185: \citet{larkin94}. Though not covering the entire mid-IR field, the
186: Br$\gamma$ emission observed is also well correlated with the mid-IR
187: emission. Both emission line maps agree with the mid-IR map in the
188: apparent lack of emission toward the dynamical center of M82.
189:
190: Non-thermal radio sources from \citet{mcdonald02} and \citet{allen99}
191: are shown in Figures 1 \& 2 as magenta crosses; these are assumed
192: to be supernova remnants since they have inverted radio spectra. We
193: find no correlation between the radio supernova remnants and the
194: mid-IR sources, but notice that most of the supernova remnants follow
195: the outer edge of the mid-IR emission at a flux levels $\lesssim$0.3
196: Jy arcsec$^{-2}$ at 11.7 $\micron$ and $\lesssim$0.7 Jy arcsec$^{-2}$
197: at 17.65 $\micron$.
198:
199: Following \citet{beck01}, we assume a reference OB star luminosity of
200: 2.5 $\times$ 10$^5$ L$_{\odot}$ for an O7 star \citep{vacca96} to
201: calculate the OB star content of each mid-IR source from its
202: L$_{MIR}$. These values are listed in Table 1. For the brightest
203: regions (B, C, and G) we find that $\sim$2000-2500 O7 stars are
204: required to generate the mid-IR luminosity.
205: \citet{fosterschreiber01} derived comparable numbers of OB stars for
206: regions near sources B and C (3200 and 4400 O7 stars, respectively)
207: using He I/Br$\gamma$ line ratios to estimate the Lyman continuum flux
208: from their regions.
209:
210: Since we find that the [Ne II] emission correlates spatially with the
211: mid-IR emission, we use [Ne II] channel maps from
212: \citet{achtermannlacy95}, which have a resolution of 16 km s$^{-1}$,
213: to estimate velocity dispersions ($\sigma_{vel}$) in the ionized gas
214: of $\sim$15-30 km s$^{-1}$ at the positions of the mid-IR sources.
215: Estimates of the velocity dispersions from the Br$\gamma$
216: \citep{larkin94} and $^{13}$CO \citep{neininger98} channel maps, each
217: with lower velocity resolution, give similar results. A caveat to
218: using ionized emission to estimate a velocity dispersion is that the
219: ionized gas may well be accelerated by shocks, and therefore its
220: velocity should be considered an upper limit to the stellar velocity.
221:
222: From the range of estimated velocity dispersions and assuming
223: virialized systems, we use $M = \eta \sigma_{vel}^2 r_{h}/G$, to
224: calculate a range for the total mass in each these systems. In this
225: formulation, G is the gravitational constant, $r_{h}$ is the projected
226: half intensity radius defined to be the geometric mean of the
227: semi-major and semi-minor axes, and $\eta$ = 10 \citep{smith01}. The
228: mass ranges for each mid-IR source are listed in Table 1 and should be
229: considered upper limits since H II regions are generally found to be
230: freely expanding systems. For the largest sources, the mass range we
231: find is 6-25 $\times$ 10$^6$ M$_{\odot}$. In every case, the range
232: contains the system mass found by extrapolating from the number of O7
233: stars using a Salpeter mass function. Comparing the sizes and masses
234: of the star clusters forming in M82 to three of the largest Galactic
235: globular clusters, which have radii of 5-10 pc and masses 1-2 $\times$
236: 10$^6$ M$_{\odot}$ (see Table 1), we conclude that the star clusters
237: forming in the nuclear region of M82 are young analogs to globular
238: clusters.
239:
240: The existence of present day globular cluster-sized knots of star
241: formation is not unique to M82. Indeed, super-star clusters
242: containing quantities of OB stars similar to those we find in M82 have
243: been observed in a number of galaxies (e.g. NCG 5253,
244: \citealt{turner03}; He 2-10, \citealt{johnson03}; NGC 4038/9,
245: \citealt{mengel02}). Additionally, two star forming regions in the
246: Milky Way, the Arches cluster near the Galactic center and the Cygnus
247: OB2 association, are estimated to each weigh in at 6 $\times$ 10$^4$
248: M$_{\odot}$ \citep{serabyn98, knodlseder00}, comparable to a small
249: globular cluster \citep{mandushev91,pryor93}. However, the Arches
250: cluster and Cygnus OB2 each contain around 100 O stars, which
251: contribute together only $\sim$3\% of the total Galactic O stars
252: \citep{terzian74}, and yet far outnumber all other Galactic star
253: forming regions in their O star content.
254:
255: In M82, the seven mid-IR sources in the nuclear region together
256: contribute $\sim$ 15\% of the total L$_{MIR}$ of the galaxy. The [Ne
257: II] map suggests that there may be several H II regions outside our
258: mid-IR field which may contribute up to an additional $\sim$5\% to
259: the total L$_{IR}$. Assuming all the mid-IR luminosity in M82 comes
260: from star formation \citep{telesco88}, it follows that $\gtrsim$20\%
261: of M82's star formation is in the form of super-star clusters - in
262: contrast to the mere 3\% in the Milky Way. This may be an important
263: feature of starbursts in general; not only do they provide an
264: environment suitable for forming globular clusters, but the super-star
265: cluster formation efficiency in starbursts is $\gtrsim$20\%.
266:
267: \section{Conclusions}
268:
269: This paper presents mid-IR (11.7 and 17.65 $\micron$) maps with
270: $\sim$0.5$\arcsec$ resolution of the central 400 pc of the
271: starbursting galaxy M82. We find 7 resolved sources in this region of
272: M82 with luminosities summing to 15\% of the total IRAS flux of the
273: entire galaxy. The mid-IR maps exhibit features comparable to those
274: found in maps of [Ne II] emission, Br$\gamma$ emission, and thermal
275: free-free emisson. We present evidence implying that the mid-IR
276: sources are giant H II regions in which globular cluster-sized star
277: clusters are forming. Our data imply that $\gtrsim$20\% of the star
278: formation in M82 is occurring in super-star clusters.
279:
280: \acknowledgments This work has been supported by funding from NASA.
281: The authors wish to thank Mike Jura, Jean Turner, and James Larkin for
282: helpful comments. Data presented herein were obtained at the
283: W.M. Keck Observatory, which is operated as a scientific partnership
284: among the California Institute of Technology, the University of
285: California and NASA. The Observatory was made possible by the generous
286: financial support of the W.M. Keck Foundation. The authors wish to
287: recognize and acknowledge the very significant cultural role that the
288: summit of Mauna Kea has always had within the indigenous Hawaiian
289: community. We are most fortunate to have the opportunity to conduct
290: observations from this mountain. This publication makes use of data
291: products from the Two Micron All Sky Survey, which is a joint project
292: of the University of Massachusetts and the Infrared Processing and
293: Analysis Center/California Institute of Technology, funded by the NASA
294: and the National Science Foundation.
295:
296: \begin{thebibliography}{}
297: \bibitem[Achtermann \& Lacy(1995)]{achtermannlacy95} Achtermann,
298: J.~M.~\& Lacy, J.~H.\ 1995, \apj, 439, 163
299: \bibitem[Allen(1999)]{allen99} Allen, M.~L.\ 1999,
300: Ph.D.~Thesis
301: \bibitem[Beck, Turner, \& Gorjian(2001)]{beck01} Beck, S.~C.,
302: Turner, J.~L., \& Gorjian, V.\ 2001, \aj, 122, 1365
303: \bibitem[Cottrell(1977)]{cottrell77} Cottrell, G.~A.\ 1977,
304: \mnras, 178, 577
305: \bibitem[de Grijs, O'Connell, \& Gallagher(2001)]{degrijs01} de
306: Grijs, R., O'Connell, R.~W., \& Gallagher, J.~S.\ 2001, \aj, 121, 768
307: \bibitem[Dietz et al.(1989)]{deitz89} Dietz, R.~D., Gehrz,
308: R.~D., Jones, T.~J., Grasdalen, G.~L., Smith, J., Gullixson, C., \&
309: Hackwell, J.~A.\ 1989, \aj, 98, 1260
310: \bibitem[Dietz et al.(1986)]{deitz86} Dietz, R.~D., Smith, J.,
311: Hackwell, J.~A., Gehrz, R.~D., \& Grasdalen, G.~L.\ 1986, \aj, 91, 758
312: \bibitem[Freedman et al.(1994)]{freedman94} Freedman, W.~L.~et
313: al.\ 1994, \apj, 427, 628
314: \bibitem[F{\" o}rster-Schreiber et al.(2001)]{fosterschreiber01}
315: F{\" o}rster-Schreiber, N.~M., Genzel, R., Lutz, D., Kunze, D., \& Sternberg,
316: A.\ 2001, \apj, 552, 544
317: \bibitem[F{\" o}rster-Schreiber et al.(2003)]{forsterschreiber03}
318: F{\" o}rster-Schreiber, N.~M., Sauvage, M., Charmandaris, V., Laurent, O.,
319: Gallais, P., Mirabel, I.~F., \& Vigroux, L.\ 2003, \aap, 399, 833
320: \bibitem[Golla, Allen, \& Kronberg(1996)]{golla96} Golla, G.,
321: Allen, M.~L., \& Kronberg, P.~P.\ 1996, \apj, 473, 244
322: \bibitem[Johnson \& Kobulnicky(2003)]{johnson03} Johnson, K.~E., \&
323: Kobulnicky, H.~A.\ 2003, \apj, in press (astro-ph/0308303)
324: \bibitem[Jones \& Puetter(1993)]{jones93} Jones, B.~\& Puetter,
325: R.~C.\ 1993, \procspie, 1946, 610
326: \bibitem[Kn{\" o}dlseder(2000)]{knodlseder00} Kn{\" o}dlseder, J.\
327: 2000, \aap, 360, 539
328: \bibitem[Larkin et al.(1994)]{larkin94} Larkin, J.~E., Graham,
329: J.~R., Matthews, K., Soifer, B.~T., Beckwith, S., Herbst, T.~M., \&
330: Quillen, A.~C.\ 1994, \apj, 420, 159
331: \bibitem[Lester et al.(1990)]{lester90} Lester,
332: D.~F., Carr, J.~S., Joy, \& M.Gaffney, N.\ 1990, \apj, 352, 544
333: \bibitem[Lord et al.(1996)]{lord96} Lord, S.~D., Hollenbach,
334: D.~J., Haas, M.~R., Rubin, R.~H., Colgan, S.~W.~J., \& Erickson, E.~F.\
335: 1996, \apj, 465, 703
336: \bibitem[Mandushev, Staneva, \& Spasova(1991)]{mandushev91}
337: Mandushev, G., Staneva, A., \& Spasova, N.\ 1991, \aap, 252, 94
338: \bibitem[McCrady, Gilbert, \& Graham(2003)]{mccrady03} McCrady, N., Gilbert,
339: A.~M. \& Graham, J.~R.\ 2003, \apj, in press (astro-ph/0306373)
340: \bibitem[McDonald et al.(2002)]{mcdonald02} McDonald, A.~R.,
341: Muxlow, T.~W.~B., Wills, K.~A., Pedlar, A., \& Beswick, R.~J.\ 2002,
342: \mnras, 334, 912
343: \bibitem[Mengel et al.(2002)]{mengel02} Mengel, S., Lehnert, M.~D.,
344: Thatte, N., \& Genzel, R.\ 2002, \aap, 383, 137
345: \bibitem[Meurer et al.(1995)]{meurer95} Meurer, G.~R., Heckman,
346: T.~M., Leitherer, C., Kinney, A., Robert, C., \& Garnett, D.~R.\ 1995, \aj,
347: 110, 2665
348: \bibitem[Neininger et al.(1998)]{neininger98} Neininger, N.,
349: Guelin, M., Klein, U., Garcia-Burillo, S., \& Wielebinski, R.\ 1998, \aap,
350: 339, 737
351: \bibitem[O'Connell \& Mangano(1978)]{om78} O'Connell, R.~W.~\&
352: Mangano, J.~J.\ 1978, \apj, 221, 62
353: \bibitem[Pryor \& Meylan(1993)]{pryor93} Pryor, C.~\& Meylan,
354: G.\ 1993, ASP Conf.~Ser.~ 50: Structure and Dynamics of Globular Clusters,
355: 357
356: \bibitem[Rieke \& Lebofsky(1985)]{rieke85} Rieke, G.~H.~\&
357: Lebofsky, M.~J.\ 1985, \apj, 288, 618
358: \bibitem[Rieke et al.(1980)]{rieke80} Rieke, G.~H., Lebofsky,
359: M.~J., Thompson, R.~I., Low, F.~J., \& Tokunaga, A.~T.\ 1980, \apj, 238, 24
360: \bibitem[Sakai \& Madore(1999)]{sakai99} Sakai, S.~\& Madore,
361: B.~F.\ 1999, \apj, 526, 599
362: \bibitem[Satyapal et al.(1995)]{satyapal95} Satyapal, S.~et al.\
363: 1995, \apj, 448, 611
364: \bibitem[Serabyn, Shupe, \& Figer(1998)]{serabyn98} Serabyn, E.,
365: Shupe, D., \& Figer, D.~F.\ 1998, \nat, 394, 448
366: \bibitem[Smith \& Gallagher(2001)]{smith01} Smith, L.~J.~\&
367: Gallagher, J.~S.\ 2001, \mnras, 326, 1027
368: \bibitem[Spitzer(1987)]{spitzer87} Spitzer, L.\ 1987, Princeton,
369: NJ, Princeton University Press, 1987, 191 p.,
370: \bibitem[Taniguchi, Trentham, \& Ikeuchi(1999)]{taniguchi99}
371: Taniguchi, Y., Trentham, N., \& Ikeuchi, S.\ 1999, \apjl, 526, L13
372: \bibitem[Telesco(1988)]{telesco88} Telesco, C.~M.\ 1988, \araa,
373: 26, 343
374: \bibitem[Telesco et al.(1991)]{telesco91} Telesco, C.~M., Campins,
375: H., Joy, M., Dietz, K., \& Decher, R.\ 1991, \apj, 369, 135
376: \bibitem[Telesco, Decher, \& Joy(1989)]{telesco89} Telesco,
377: C.~M., Decher, R., \& Joy, M.\ 1989, \apjl, 343, L13
378: \bibitem[Telesco \& Gezari(1992)]{telescogezari92} Telesco, C.~M.~\&
379: Gezari, D.~Y.\ 1992, \apj, 395, 461
380: \bibitem[Telesco \& Harper(1980)]{telesco80} Telesco, C.~M.~\&
381: Harper, D.~A.\ 1980, \apj, 235, 392
382: \bibitem[Terzian(1974)]{terzian74} Terzian, Y.\ 1974, \apj, 193,
383: 93
384: \bibitem[Turner et al.(2003)]{turner03} Turner, J.~L., Beck,
385: S.~C., Crosthwaite, L.~P., Larkin, J.~E., McLean, I.~S., \& Meier, D.~S.\
386: 2003, \nat, 423, 621
387: \bibitem[Vacca, Garmany, \& Shull(1996)]{vacca96} Vacca, W.~D.,
388: Garmany, C.~D., \& Shull, J.~M.\ 1996, \apj, 460, 914
389: \bibitem[van den Bergh, Morbey, \& Pazder(1991)]{vandenbergh91} van
390: den Bergh, S., Morbey, C., \& Pazder, J.\ 1991, \apj, 375, 594
391: \bibitem[Zepf et al.(1999)]{zepf99} Zepf, S.~E., Ashman,
392: K.~M., English, J., Freeman, K.~C., \& Sharples, R.~M.\ 1999, \aj, 118, 752
393:
394:
395: \end{thebibliography}
396:
397: \clearpage
398:
399: \begin{deluxetable}{ccccccccc}
400: \tablenum{1}
401: \tablecaption{Properties of Proto-Globular Clusters in
402: M82 and Galactic Globular Clusters}
403: \label{table1}
404: \tablewidth{0pt}
405: \tablehead{
406: \colhead{Source} & \colhead{F$_{\nu}$(11.7 $\micron$)} &
407: \colhead{F$_{\nu}$(17.65 $\micron$)} & \colhead{Size\tablenotemark{a}} &
408: \colhead{T$_c$} & \colhead{L$_{MIR}$} &
409: \colhead{N(O7)} & \colhead{Mass\tablenotemark{b}}\\
410: \colhead{} & \colhead{(Jy)} &
411: \colhead{(Jy)} & \colhead{(pc)} & \colhead{(K)} &
412: \colhead{(10$^8$ L$_{\odot}$)} &
413: \colhead{} &\colhead{(10$^6$ M$_{\odot}$)}
414: }
415: \startdata
416: A & 0.16 & 0.15 & 10 $\times$ 10 & 270 & 0.20 & 80 & 0.3\tablenotemark{c} \\
417: B & 3.17 & 5.39 & 30 $\times$ 21 & 195 & 5.9 & 2400 & 6-25 \\
418: C & 2.38 & 5.95 & 26 $\times$ 19 & 165 & 5.5 & 2200 & 6-22 \\
419: D & 1.23 & 3.31 & 21 $\times$ 12 & 160 & 3.3 & 1300 & 4-16 \\
420: E & 0.67 & 1.74 & 13 $\times$ 13 & 160 & 1.7 & 700 & 3-13 \\
421: F & 0.77 & 2.24 & 28 $\times$ 8 & 155 & 2.3 & 900 & 4-15 \\
422: G & 1.56 & 4.96 & 25 $\times$ 16 & 150 & 4.9 & 2000 & 5-20 \\
423: $\omega$ Cen & & & 9\tablenotemark{d} & & & & 2\tablenotemark{e} \\
424: NGC 6273 & & & 3\tablenotemark{d} & & & & 1\tablenotemark{e} \\
425: NGC 104 & & & 5\tablenotemark{d} & & & & 1\tablenotemark{e} \\
426:
427: \enddata
428: \tablenotetext{a}{Sizes are full width at half intensity major and minor
429: axes of elliptical sources, as observed, and include a $\sim$0.5$\arcsec$
430: (9 pc) PSF.}
431: \tablenotetext{b}{Upper limit to mass of systems}
432: \tablenotetext{c}{Source A is not identifiable in the [Ne II] channel
433: maps. Stellar mass extrapolated from the number of O7 stars using a
434: Salpeter initial mass function}
435: \tablenotetext{d}{Sizes are averages of values in \citet{vandenbergh91} and
436: \citet{mandushev91}}
437: \tablenotetext{e}{From \citet{mandushev91}}
438:
439: \end{deluxetable}
440:
441: \clearpage
442:
443:
444: \figcaption[f1.eps]{Central region of M82 at
445: 11.7 $\micron$ smoothed with a 0$\farcs$4 boxcar function. The
446: seven mid-IR sources are labeled with white letters and their
447: positional errors are represented by 0$\farcs$5 radius black circles.
448: The 2 $\micron$ peak \citep{deitz86} is labeled with a yellow cross
449: and M82's dynamical center as determined by \citet{lester90} is
450: marked by a yellow circle. Magenta crosses mark positions for
451: non-thermal radio sources \citep{mcdonald02, allen99} and red
452: squares mark position of H II regions \citep{allen99}.}
453:
454: \figcaption[f2.eps]{Central region of M82 at
455: 17.65 $\micron$ smoothed with a 0$\farcs$4 boxcar function. Symbols
456: are the same as in Figure 1.}
457:
458: \figcaption[f3.eps]{Smoothed 11.7 $\micron$ data overlaid with
459: contours from the \citet{achtermannlacy95} [Ne II] line map deconvolved
460: with the maximum entropy method (their Figure 5).}
461:
462: \clearpage
463:
464: \begin{figure}
465: \plotone{f1.eps}
466: \end{figure}
467:
468: \begin{figure}
469: \plotone{f2.eps}
470: \end{figure}
471:
472: \begin{figure}
473: \plotone{f3.eps}
474: \end{figure}
475:
476: \end{document}
477: