1: \documentclass[12pt,preprint]{aastex}

2: \usepackage{epsfig}

3: \shorttitle{IR Dust Bubbles}

4: \shortauthors{Watson et al.}

5: \begin{document}

6: \title{IR Dust Bubbles: Probing the Detailed Structure and Young Massive

7:   Stellar Populations of Galactic HII Regions} \author{C.

8:   Watson\altaffilmark{1}, M.S. Povich\altaffilmark{2}, E.B.

9:   Churchwell\altaffilmark{2}, B.L. Babler\altaffilmark{2}, G.

10:   Chunev\altaffilmark{1}, M.  Hoare\altaffilmark{5}, R.

11:   Indebetouw\altaffilmark{3}, M.R.  Meade\altaffilmark{2}, T.P.

12:   Robitaille\altaffilmark{6}, B.A.  Whitney\altaffilmark{4}}

13: \altaffiltext{1}{Manchester College, Dept. of Physics, 604 E. College Ave.,

14:   North Manchester, IN 46962} \altaffiltext{2}{Univ. of Wisconsin - Madison,

15:   Dept. of Astronomy, 475 N. Charter St., Madison, WI 53716}

16: \altaffiltext{3}{Department of Astronomy, University of Virginia, P.O. Box

17:   3818, Charlottesville, VA 22903-0818} \altaffiltext{4}{University of

18:   Colorado, Space Science Institute, 1540 30th St., Suite 23, Boulder, CO

19:   80303-1012} \altaffiltext{5}{School of Physics and Astronomy, University of

20:   Leeds, Leeds, West Yorkshire, LS2, 9JT} \altaffiltext{6}{5SUPA, School of

21:   Physics and Astronomy, University of St Andrews, North Haugh, KY16 9SS, St

22:   Andrews, United Kingdom}

23: \begin{abstract}

24:   We present an analysis of wind-blown, parsec-sized, mid-infrared bubbles and

25:   associated star-formation using GLIMPSE/IRAC, MIPSGAL/MIPS and MAGPIS/VLA

26:   surveys.  Three bubbles from the Churchwell et al. (2006) catalog were

27:   selected. The relative distribution of the ionized gas (based on 20 cm

28:   emission), PAH emission (based on 8 $\mu$m, 5.8 $\mu$m and lack of 4.5

29:   $\mu$m emission) and hot dust (24 $\mu$m emission) are compared. At the

30:   center of each bubble there is a region containing ionized gas and hot dust,

31:   surrounded by PAHs. We identify the likely source(s) of the stellar wind and

32:   ionizing flux producing each bubble based upon SED fitting to numerical hot

33:   stellar photosphere models.  Candidate YSOs are also identified using SED

34:   fitting, including several sites of possible triggered star formation.

35: \end{abstract}

36: \keywords{HII regions --- ISM: bubbles --- infrared: ISM --- radio continuum:

37:   ISM --- stars: formation}

38: \section{Introduction}

39: Mid-infrared (MIR) imaging surveys of the Galactic plane such as the

40: Mid-Course Space Experiment (MSX; Price et al.  2001) and the Galactic Legacy

41: Infrared Midplane Survey Extraordinaire (GLIMPSE; Churchwell et al. 2001;

42: Benjamin et al. 2003) have revealed a large number of bubbles in the disk of

43: the Galaxy.  The bubbles have bright 8 $\mu$m shells that enclose bright 24

44: $\mu$m interiors.  ISO (Infrared Space Observatory; Kessler et al.  1996) and

45: MSX were the first to reveal the existence of these objects (including the

46: association with triggered star-formation, Deharveng et al.  2005), but the

47: GLIMPSE survey (with ten times better spatial resolution and a hundred times

48: the sensitivity than MSX) has enabled the detection of many more bubbles and

49: has enabled the refinement of their morphological structure.  In fact, the

50: GLIMPSE survey has shown that bubbles are a hallmark of the diffuse emission

51: in the Galactic plane.  Small and faint bubbles are apparent only in the

52: GLIMPSE survey due to the lower resolution and sensitivity of MSX.  Churchwell

53: et al.  (2006; 2007) have cataloged almost 600 bubbles. An average of

54: $\gtrsim$5 bubbles per square degree were found within 20$^\circ$ of the

55: Galactic center.  These bubbles are identified as complete or partial rings in

56: the GLIMPSE images, which Churchwell et al.  (2006) argue are two dimensional

57: representations of three dimensional bubbles.  Churchwell et al. (2006) showed

58: that the bubbles are distributed in longitude and latitude like O and B stars,

59: $>$25\% of which are coincident with known radio HII regions, and $\sim$13\%

60: enclose known stellar clusters.  About 1/3 of the bubbles are produced by O

61: stars.  Approximately two thirds of the sample have small angular diameters,

62: which Churchwell et al.  (2006) propose to be mostly physically small bubbles

63: produced by late B type stars whose UV photon fluxes are adequate to excite a

64: shell of PAH emission but not adequate to produce a detectable radio HII

65: region.  Churchwell et al.  (2006; 2007) tabulated bubble morphological

66: properties such as angular diameter, shell thickness, and eccentricity.

67: 

68: %Here, we will examine in some detail the diffuse emission properties of dust

69: %in the IRAC and MIPS bands toward three bubbles and compare their IR and radio

70: %continuum emission distributions.  We show that apparent bubble morphology is

71: %a strong function of wavelength and examine the implications of this for the

72: %excitation and photo-destruction of polycyclic aromatic hydrocarbons (PAHs)

73: %and thermal dust emission.

74: 

75: Bubbles/HII regions identified by MIR emission present a new and powerful tool

76: to study the interaction of young hot stars with their environment.  Young

77: stars impact the ambient interstellar medium (ISM) by heating the dust,

78: ionizing the gas (in the case of O and early B stars), and heating an

79: expanding bubble.  Bubbles can be produced by stellar winds and over-pressure

80: by ionization and heating by stellar UV radiation.  Bubble sizes and

81: morphologies are determined by the density structure of the ambient ISM, the

82: luminosity of the central star(s) responsible for producing the bubble, the

83: stellar wind luminosity and the motion of the central star(s) relative to the

84: ambient ISM.  The morphologies of MIR dust bubbles/HII regions reveal

85: important information about the strength and directionalities of stellar winds

86: and the structure and physical properties of the ambient ISM into which the

87: bubbles are expanding.  Pillars, scalloping, and sharpness of the inner

88: boundaries of bubbles defined by 8$\mu$m emission provide unique insights into

89: the hydrodynamics, photo-ionization, and evaporation of gas and sublimation of

90: dust in expanding bubbles, and stellar mass loss rates during their evolution.

91: For example, based on GLIMPSE-MIR morphology of ultra-compact (UC) HII

92: regions, Hoare et al.  (2007) and references therein have found an unusually

93: high fraction of cometary shapes. They argue that the morphology of the

94: PAH-traced photo-dissociated region (PDR), combined with the radio free-free

95: emission, is suggestive of a champagne flow.

96: 

97: MIR bubbles that surround HII regions ionized by O stars and B supergiants

98: (i.e. those that have strong stellar winds) have been modeled by several

99: groups.  Analytic evolution models have been developed by Castro, McCray, and

100: Weaver (1975), Weaver et al. (1977), Whitworth et al. (1994), and Capriotti

101: and Kozminsky (2001). These models predict that wind-blown bubbles around O

102: stars and B supergiants should have the following distinct regions: a

103: hypersonic stellar-wind-evacuated central cavity surrounding the central star

104: where densities are very low; a hot, low-density, shocked stellar wind region

105: surrounding the central evacuated cavity that occupies most of the volume of

106: the HII region; a thin conduction zone where temperature sharply decreases and

107: density rapidly increases (with radius); a thin, dense shell of shocked H$^+$

108: gas surrounded by a very thin shell of non-shocked H$^+$ gas at about 10$^4$

109: K. The outer thin shell of 10$^4$ K H$^+$ gas is the region classically

110: thought to represent the photo-ionized HII region. This basic picture is

111: supported by detailed numerical evolution models calculated for a 35 M$_\odot$

112: (Freyer, Hensler, \& Yorke, 2006; hereafter FHY06), a 60 M$_\odot$ (Freyer,

113: Henseler, and Yorke, 2003), and an 85 M$_\odot$ star (Kr\"oger, Hensler, \&

114: Freyer, 2006, and Kr\"oger, Freyer, Hensler, \& Yorke, 2007).  The numerical

115: models, however, show that the sharp boundaries predicted by the analytic

116: models probably are not sharp due to turbulent mixing.  The consequence of

117: this is that clumps of cool 10$^4$ K gas are predicted to be found mixed with

118: the hot 10$^7$ K gas.  The thickness of the 10$^4$ K shell is predicted to be

119: a function both of age and luminosity of the central star; the lower the wind

120: luminosity the thicker the cool ionized shell is.  Also the outer boundaries

121: of the HII regions are predicted to be quite jagged due to instabilities

122: produced when a dense medium expands into a much lower density interstellar

123: medium (ISM).  We will compare some of the model predictions with our

124: observations in \S 5. All theoretical evolution models up to now have omitted

125: the possible role of dust in HII regions, which new MIR observations clearly

126: demonstrate is present.

127: 

128: Recent observations are providing a deeper understanding of the relative

129: distributions of thermal dust, PAHs, ionized gas, and PDRs.  Peeters et al.

130: (2005) review the ISO spectroscopic observations of compact and evolved HII

131: regions and the PDRs surrounding them. They find gas temperatures in the PDRs

132: between 100 K (NGC 2024) and 200 K - 500 K (S106IR) and densities between

133: 10$^4$ cm$^{-3}$ (W75N) and 10$^6$ cm$^{-3}$ (NGC 2024).  They find broad

134: agreement between these results and models of the illumination of PDRs by the

135: appropriate nearby hot star(s). Kassis et al. (2006), on the other hand,

136: report ground-based observations of the Orion bar that indicate PAHs are

137: present just outside the HII region. They show that [Ne II] 12.81 $\mu$m

138: emission traces the geometry of the PDR in the bar. They also conclude that

139: there is not a sharp transition between neutral and ionized PAH emission

140: within the PDR.

141: 

142: Each of the four Spitzer/IRAC (Infrared Array Camera) bandpasses are dominated

143: by different emission processes in the neighborhood of a hot star (see Draine

144: 2003 and Peeters et al. 2003).  The brightest objects in the 3.6$\mu$m band

145: are stars, but this band also has contributions from a weak, diffuse PAH

146: feature at 3.3$\mu$m, and possibly from scattered star light.  The 4.5$\mu$m

147: band samples diffuse emission in the Br$\alpha$ and Pf$\beta$ lines (from HII

148: regions), H$_2$ v=0-0, S(9), S(10), and S(11) and CO v=1-0 ro-vib lines at

149: 4.665$\mu$m (from shocked molecular gas).  The 4.5$\mu$m band contains no PAH

150: features and the brightest sources in this band are also stars.  The 5.8$\mu$m

151: band contains a strong PAH feature at 6.2$\mu$m which can dominate the diffuse

152: emission except very close to O stars where PAHs are destroyed (see \S 4.1).

153: Near hot O stars, strong contributions from thermally emitting dust plus a

154: small contribution from stochastically heated small grains probably dominate

155: the diffuse emission in the 5.8$\mu$m band.  The 8.0$\mu$m band contains two

156: very strong PAH features at 7.7 and 8.6$\mu$m which dominate the diffuse

157: emission in this band, although near hot stars it may be dominated by thermal

158: dust emission with little or no PAH emission.  Within ionized zones, Hoare

159: (1990) and Hoare et al. (1991) have shown that trapped Lyman $\alpha$ heated dust can

160: maintain T $>$ 100 K throughout the ionized region.  Because the emission

161: process that dominates in each IRAC band depends on the environment, band

162: ratios provide a powerful tool to measure the extent of each environment.

163: 

164: The main focus of this study is to use wavelength-dependent distributions of

165: MIR diffuse dust emission and high resolution radio continuum emission to

166: determine the PAH destruction radius of three bubbles and to trace the

167: relative sizes and locations of 1) the HII region, 2) the hot thermally

168: emitting dust, and 3) the location and extent of the PDR region associated

169: with each bubble.  In section 2 we introduce the different data sets used for

170: this analysis. In section 3, we discuss results of the observations toward

171: three bubbles. In section 4, we estimate the PAH destruction radius and dust

172: temperature. In section 5, we examine the stellar population within each of

173: the bubbles to identify the ionizing star(s) and any young stellar objects

174: (YSOs) associated with the bubbles. Conclusions are summarized in section 6.

175: 

176: \section{Data}

177: Data have been assembled from three imaging surveys of the galactic plane:

178: GLIMPSE, MIPSGAL, and MAGPIS. The latter two surveys were chosen based on

179: their resolution, sensitivity, wavelength coverage and overlap with the

180: GLIMPSE survey which provided the basis for the Churchwell et al. (2006)

181: bubble catalog. The GLIMPSE survey (Benjamin et al.  2003) imaged the inner

182: Galactic plane using the IRAC camera (Fazio et al. 2004) on the Spitzer Space

183: Telescope (Werner et al. 2004). The survey covered 10$^\circ$ $< |l| <$

184: 65$^\circ$ and $|b| <$ 1$^\circ$ at 3.6 $\mu$m, 4.5 $\mu$m, 5.8 $\mu$m and 8.0

185: $\mu$m with resolution from 1.5'' (3.6$\mu$m) to 1.9'' (8.0$\mu$m). Mosaicked

186: images were produced using the IPAC Montage program in the GLIMPSE pipeline

187: after image artifacts were removed. A Point Source Archive was produced of

188: all point sources detected above the 5$\sigma$ level, about 48 million

189: sources. See the GLIMPSE Data Products

190: Description\footnote{www.astro.wisc.edu/sirtf/glimpse1\_dataprod\_v2.0.pdf}

191: for further details.

192: 

193: MIPSGAL (Carey et al. 2005) is a survey of the inner Galactic plane

194: (10$^\circ$ $< |l| <$ 65$^\circ$ and $|b| <$ 1$^\circ$) at 24 $\mu$m and 70

195: $\mu$m using the MIPS (Multiband Imaging Photometer for Spitzer) instrument

196: (Rieke et al. 2004) on the Spitzer Space Telescope. MIPS has a resolution of

197: 5'' at 24 $\mu$m and 15'' at 70 $\mu$m. We only analyzed the 24 $\mu$m

198: emission because the 70 $\mu$m appears complex enough to warrant separate

199: analysis in an upcoming paper.

200: 

201: %In these cases, we used data from the MSX

202: %survey. The MSX survey (Price et al. 2001) used the SPIRIT-III instrument to

203: %cover the regions missed either by the IRAS satellite or the the COBE/DIRBE

204: %satellite, or in cases where the sensitivity of IRAS was limited by source

205: %confusion.  The MSX survey covers the galactic plane at all longitudes and

206: %$|b| <$ 5$^\circ$ along with other regions at six IR bands (8.28 $\mu$m, 4.29

207: %$\mu$, 4.35 $\mu$m, 12.13 $\mu$m, 14.65 $\mu$m, and 21.34 $mu$m) with a

208: %resolution of 18.3''.

209: 

210: The Multi-Array Galactic Plane Imaging Survey (MAGPIS; Helfand et al. 2006)

211: used the VLA in B, C and D configurations combined with the Effelsberg 100m

212: single dish data to obtain high resolution radio images (with no zero spacing

213: problems) at 1.4 GHz continuum. The survey covered 5$^\circ$ $< l <$

214: 32$^\circ$ and $|b| <$ 0.8$^\circ$ with a resolution of 6'' and a 1$\sigma$

215: sensitivity of $\sim$0.3 mJy. Full polarization was preserved.

216: 

217: \section{Results}

218: We now present the relative distribution of MIR and 20 cm emission for three

219: bubbles selected because they represent different morphological shapes and/or

220: show possible evidence for triggered star formation.  They are N10, N21 and

221: N49 in the Churchwell et al. (2006) bubble catalog.

222: 

223: \subsection{N10}

224: N10 is a bright MIR and radio continuum bubble having elliptical or slight

225: cometary shape with an opening at galactic position angle of 160$^\circ$ (see

226: Fig.  \ref{n10color}).  Its kinematic distance is 4.9$\pm$0.5 kpc (Churchwell

227: et al.  2006 and references therein).  At 1.4 GHz it has an integrated flux

228: density of 7.58 Jy (Helfand et al 2006) and we measure an average angular

229: radius to the background of 1.26' ($\sim$1.8 pc).  Using the relation of

230: (Rohlfs \& Wilson, 2006):

231: \begin{eqnarray*}

232: \frac{N_{Ly}}{s^{-1}} &=& 6.3 \times 10^{52} \left(\frac{T_e}{10^4}\right)

233: \left(\frac{\nu}{GHz}\right)^{-0.1} \left(\frac{L_\nu}{10^{20} W

234:     Hz^{-1}}\right),

235: \end{eqnarray*}

236: we calculate 1.6 $\times$10$^{49}$ ionizing photons per second are necessary

237: to maintain ionization, equivalent to a single O5V star (Martins, Schaerer \&

238: Hillier 2005, hereafter MSH05). No correction for extinction was made.

239: %In Fig. \ref{n10mono} we show a monochromatic image of N10 at 3.6 $\mu$m.

240: %At this

241: %wavelength, one sees mostly stellar emission.  A circle of diameter

242: %approximately equal to that of the 8$\mu$m bubble is centered on N10.  

243: %There are 66 stars with 4 or more detections in the GLIMPSE Archive (out of a

244: %total 2 visits $\times$ 4 bands $=$ 8 possible detections across all bands)

245: %projected within the 8$\mu$m shell of the bubble.  We will discuss the stellar

246: %population toward N10 in section 5.

247: % Five stars that lie near the

248: %peaks of the 20 cm emission and within the regions of 24$\mu$m emission are

249: %the most likely suspects for providing most of the ionizing radiation and wind

250: %power that has formed the bubble.

251: %We list in Table \ref{catalog} the

252: %magnitudes of these stars taken from the GLIMSPE Catalog.  Because of the

253: %insensitivity at NIR-MIR wavelengths to stellar spectral type we cannot

254: %determine the spectral types of these stars with any certainty, but it is

255: %certain that they are early B or hotter stars otherwise they could not provide

256: %the UV photons required to maintain ionization of the HII region.

257: 

258: The relative distributions of emission as a function of wavelength are

259: illustrated in Figs. \ref{n10lat1}-\ref{n10lat2}.  The bubble is surrounded by

260: an 8$\mu$m emission shell with angular radius (out to the Galactic background

261: level) of 1.8' (2.6 pc).  The radius to the inner face of the 8 $\mu$m shell

262: is 1.2' (1.7 pc). The 8$\mu$m emission rises very sharply on the inner face of

263: the shell and declines gradually with increasing radius.  We postulate that

264: this ring is dominated by PAH emission and the inner face of the shell defines

265: the destruction radius of PAHs (see \S 4.1).  It is bright because swept-up

266: interstellar dust densities are high here and the dust is exposed to a large

267: flux of soft UV radiation (non-H ionizing photons) that excites PAHs but is

268: not energetic enough to destroy PAHs.  The slow fall-off beyond the inner face

269: of the 8$\mu$m shell represents the PDR region of the bubble, primarily

270: delineated by PAH emission.

271: 

272: Inside the 8$\mu$m shell, 24$\mu$m, 20 cm, 8.0 $\mu$m and 5.8 $\mu$m emission

273: all peak at the same position, showing that hot dust is present inside the HII

274: region.  Note that the 24 $\mu$m emission is saturated at the center. Clearly

275: the stellar wind(s) have not yet succeeded in clearing out or destroying all

276: the dust in the bubble.  There is also some diffuse 5.8 $\mu$m and 8 $\mu$m

277: emission inside the bubble.

278: %We suggest that this is primarily

279: %due to thermal emission from the dust that produces the 24$\mu$m emission in

280: %the bubble interior and possibly some emission from small grains whose

281: %temperatures stochastically fluctuate due to absorption of UV photons.  

282: Figure \ref{n10lat2} shows that the 3.6 $\mu$m and 4.5$\mu$m emission vary

283: together because both are dominated by stellar emission.  The 5.8 $\mu$m and

284: 8.0$\mu$m band vary together also, although 5.8$\mu$m emission is generally

285: fainter than 8.0$\mu$m emission. The relative distribution of emission at

286: different wavelengths in N10 (i.e. an 8$\mu$m shell enclosing 24$\mu$m and

287: radio continuum emission) is a general property of all the bubbles for which

288: we have MAGPIS and MIPSGAL data.

289: 

290: %Source N10 is a bright and vivid example of the relative position of different

291: %components of this class of object. As is seen in the longitude slice (see

292: %Fig. XX), the source is surrounded by 8$\mu$m emission (in green). As is true

293: %for most of the this band in the GLIMPSE survey, we believe this emission to

294: %be dominated by PAH emission as opposed to thermal emission.  We measure the

295: %size of this bubble to be 2.8' by 2.7'. The edges on the minor axis (running

296: %from high-longitude and low-latitude to low-longitude and high-latitude),

297: %however, are not well-defined. 

298: %

299: %Contained in this bubble of 8 $\mu$m emission is 20 cm emission as observed by

300: %MAGPIS (shown in contours in Fig. XX). This emission originates in gas ionized

301: %by a massive star. The longitude slice clearly shows the bulk of 20 cm

302: %emission well-contained by the outer 8$\mu$m emission. The small bump of 20

303: %emission at $l$=13.21$^\circ$ in the longitude slice is probably caused by the

304: %star located at $l$=13.21$^\circ$, $b$=0.04$^\circ$. The 20 cm contours

305: %clearly peak around this source. Although some of the increase in 8$\mu$m

306: %emission is caused by this star, most of the bubble shape is not likely to be

307: %associated with this source. Also contained in the bubble is an increase in

308: %8$\mu$m emission and 24$\mu$m emission. The 24$\mu$m data from MIPSGAL quickly

309: %becomes saturated. We have examined the same region in the MSX survey, which

310: %had poorer resolution but higher saturation limits. In all three bands the

311: %source is clearly double-peaked, indicating that this bubble may dominated by

312: %multiple stars. We measured the flux contained within the bubble at 8$\mu$m

313: %and 21$\mu$m (from the MSX survey) to be 337 Jy and 2176 Jy, respectively. We

314: %measure the total emission at 8$\mu$m (including all the emission associated

315: %with the surrounding bubble) to be 814 Jy.

316: 

317: \subsection{N21}

318: N21 has a cometary morphology in 8 $\mu$m emission (see Fig \ref{n21color}),

319: but otherwise has similar relative spatial distributions with wavelength as

320: N10 with the exception that N21 is not bounded at 8$\mu$m along the lower half

321: of the bubble. N21 has a kinematic distance of 3.7$\pm$0.6 kpc (Churchwell et

322: al.  2006). At 1.4 GHz it has an integrated flux density of 7.2 Jy and an

323: angular radius of 2.6' (2.8 pc) (Helfand et al.  2006). It requires

324: 8.5$\times$10$^{48}$ ionizing photons s$^{-1}$ to maintain ionization,

325: equivalent to a single O6V star (MSH05).

326: %In Fig. \ref{n21mono} we show a 3.6 $\mu$m GLIMPSE image of N21 which, like

327: %N10, is dominated by stellar emission.  The 4.5 $\mu$m and 3.6 $\mu$m images

328: %are similar. 

329: %Within the shell there are 257 stars with 4 or more detections in

330: %the GLIMPSE Archive (see section 5 for further discussion of the stellar

331: %population toward N21).

332: 

333: The 8 $\mu$m shell has a radius to the inner face of 2.1' (2.2 pc) and a

334: radius out to the background level of of 3.0' (3.2 pc). The clearest

335: perspective on the relative emission distributions with wavelength are the

336: slices at constant longitude (see Fig.  \ref{n21lon1}-\ref{n21lon2}). Figure

337: \ref{n21lon1} shows that 8$\mu$m emission is located above the 20 cm and 24

338: $\mu$m emission. 20 cm and 24$\mu$m emission increase and decrease in near

339: unison. The 5.8 $\mu$m and 8.0 $\mu$m emission also vary together (see

340: Figures~\ref{n21lon1} \& \ref{n21lon2}). 5.8 $\mu$m and 8.0 $\mu$m are present

341: both inside the bubble and along the top boundary.  Inside the bubble, the 8

342: $\mu$m flux is likely dominated by both PAH emission from the front-side and

343: backside of the bubble and by hot dust grains inside the bubble.  We measure

344: the integrated flux density at 8$\mu$m inside the 8 $\mu$m shell (0.395$^\circ

345: < b <$ 0.41$^\circ$) to be 228 Jy and the integrated flux density at 24$\mu$m

346: in the same region to be 599 Jy. The former value may be over-estimated by

347: $\sim$30\% because of IRAC diffuse calibration errors (Cohen et al. 2007). The

348: dust temperature cannot be determined because of PAH contamination at 8

349: $\mu$m.

350:   

351: The relative distributions of hot dust, excited PAHs and the PDR region in N21

352: is similar to N10. The key difference between N10 and N21 is the absence of 8

353: $\mu$m emission along the southern half of the bubble. The detected 24 $\mu$m

354: and 20 cm emission along the southern interior, however, implies the existence

355: of ionizing photons.  The near absence of the 8 $\mu$m shell along the lower

356: half of this bubble thus implies the absence of PAHs.  This may be because the

357: ambient ISM density in this direction is low enough that the PAHs have either

358: been destroyed by direct exposure to hard stellar UV radiation and/or blown

359: out from the bubble far enough that the PAHs cannot be excited by the stellar

360: radiation due to geometric dilution.

361:   

362: %Lastly, there is an interesting extension of 24$\mu$m emission to the

363: %upper-right. Over a small part of the ring, the 24$\mu$m emission coincides

364: %with part of the 8$\mu$m emission that makes-up the brightest part of the

365: %bubble (see marked area in Fig \ref{n21_8_24}).  Upon close inspection, it is

366: %clear that in the small, filamentary structure visible in the 24 $\mu$m

367: %emission is mirrored in the 8 $\mu$m emission. The presence of 24 $\mu$m

368: %emission along the shell indicates additional physics besides PAH emission,

369: %such as thermal emission or stochastically heated grains.

370: 

371: \subsection{N49}

372: N49 is a bright MIR bubble surrounding a radio HII region that has an almost

373: spherical structure.  It has a kinematic distance of 5.7$\pm$0.6 kpc

374: (Churchwell et al.  2006). At 1.4 GHz it has an integrated flux density of 2.8

375: Jy and an angular radius out to the background of 1.5' ($\sim$2.5 pc) (Helfand

376: et al. 2006). 7.8 $\times$10$^{48}$ ionizing photons s$^{-1}$ are necessary to

377: maintain its ionization, equivalent to a single O6V star (MSH05). The radius

378: to the inner face of the 8 $\mu$m shell is 1.2' (2.0 pc) and out to the

379: background level is 1.7' (2.3 pc).

380: %In Fig.  \ref{n49mono} we show a 3.6 $\mu$m image of this bubble to

381: %highlight the stellar distribution. 

382: %There are 84 stars detected 4 or more times in the GLIMPSE Archive that lie

383: %within the projected boundaries of the 8 $\mu$m shell. The possible ionizing

384: %star(s) of the HII region and related possible YSOs are discussed in \S 5.

385: 

386: N49 has a double-shell structure, the outer traced by 8$\mu$m emission and the

387: inner traced by 24$\mu$m and 20 cm emission (see Fig \ref{n49color}). As in

388: N10 and N21, 8$\mu$m emission encloses both the 24$\mu$m and 20 cm emission.

389: The transition between the 8$\mu$m emission ring and the 24$\mu$m and 20 cm

390: emission ring can be clearly seen in the slice at constant latitude in Fig

391: \ref{n49lat1}. The 20 cm and 24$\mu$m emission are coincident, both of which

392: have a central cavity.  The 24 $\mu$m/20 cm dip appears to be the central

393: wind-evacuated cavity expected around early-O stars.

394: 

395: %With the goal of determining the dust temperature distribution within the HII

396: %region, we attempted to estimate the fraction of 8$\mu$m emission due to

397: %thermal dust emission which could then be combined with 24$\mu$m thermal

398: %emission to obtain the dust temperature distribution.  We estimated the amount

399: %of PAH emission from the front and back segments of the shell expected at

400: %8$\mu$m as a function of radius based on a spherical shell of various

401: %thicknesses.  We found that most of the emission at 8$\mu$m is due to the PAH

402: %shell, with a residual thermal component so small that it could not be

403: %reliably measured and used for a temperature determination.  This is discussed

404: %in more detail in $\S$4.2.

405: 

406: \section{Analysis}

407: We propose the following picture for the IR bubbles: ionized gas with a hot

408: dust component is surrounded by a PDR containing swept-up interstellar gas,

409: PAHs, and dust.  The ionized gas is traced by 20 cm free-free emission, the

410: hot dust within the HII region is bright at 24$\mu$m via thermal continuum

411: emission.  The IR bubbles are enclosed by a shell of 8$\mu$m emission

412: dominated by PAH emission features in IRAC bands 3.6, 5,8, and 8.0$\mu$m. The

413: inner face of the 8$\mu$m shell defines the PAH destruction radius from the

414: central ionizing star(s). In the following sections, we determine the PAH

415: destruction radii and PDR shell thicknesses based on $\frac{5.8 \mu m}{4.5 \mu

416:   m}$ and $\frac{8.0 \mu m}{4.5 \mu m}$ flux density ratios.

417: 

418: %The stellar population toward each bubble is discussed in $\S$5; in

419: %particular, we attempt to identify the ionizing star(s) and any associated

420: %YSOs toward each bubble with some discussion of their properties.

421:  

422: %and

423: %discuss the candidate ionizing stars in the context of the diffuse emission

424: %morphologies.

425: %Typically, lack of 8 $\mu$m PAH emission

426: %(e.g. only a semi-circular bubble at 8.0 $\mu$m) indicates absent ambient gas.

427: %8 $\mu$m emission within the ionized region is not likely to be caused by PAH

428: %molecules because UV radiation effectively destroys them.  Since 8 $\mu$m

429: %emission is only observed in the most luminous sources, we propose this

430: %interior 8$\mu$m originates from thermally excited dust grains.  It should be

431: %noted that stochastic heating is unlikely to contribute more than $\sim$10\%

432: %of the emission.  This detection of thermally-excited dust grains is exciting

433: %because H{\small II} regions are well-known to contain dust (citation) that

434: %have important effects on the radiative and dynamical properties.

435: 

436: \subsection{PAH Destruction}

437: 

438: Povich et al. (2007) argued that ratios of IRAC bands that contain strong PAH

439: emission features (8.0$\mu$m and 5.8$\mu$m bands) to the 4.5$\mu$m band (which

440: contains no PAH feature) can be used to determine the PAH destruction radius

441: and define the extent of PDRs around hot stars.  This technique was applied by

442: Povich et al. (2007) to derive both the PAH destruction region in M17 and the

443: extent of its PDR because the 8.0 $\mu$m, 5.8 $\mu$m and 3.6 $\mu$m IRAC bands

444: all contain PAH bands, whereas band-ratios involving the 4.5 $\mu$m PAH-free

445: band should be especially sensitive to regions containing PAHs.  They

446: supported their interpretation of these ratios by showing that the

447: 5.8$\mu$m/3.6$\mu$m ratio does not delineate the PDR boundaries. They also

448: presented IRS spectra that proved the disappearance of PAH features within the

449: M17 HII region. Povich et al. (2007) were unable to use the 8.0$\mu$m images

450: of M17 because the detector was saturated over large regions.  We have applied

451: this technique to N10, N21, and N49. The quantitative ratios are different

452: from those toward M17 because M17 is a much more luminous region but the

453: principle is the same. Since N10, N21 and N49 do not saturate the 8.0$\mu$m

454: detector, we are able to use this band in our analysis as well.

455: 

456: Figs. \ref{n10pahc} - \ref{n49pah} show false color images of the 5.8

457: $\mu$m/4.5 $\mu$m and 8.0 $\mu$m/4.5 $\mu$m band ratios, with accompanying

458: longitude or latitude cuts (averaged over 20 pixels) for all three bubbles. We

459: contour the false-color images to indicate the average values that define the

460: PDR regions.  For each source, the 5.8 $\mu$m/4.5 $\mu$m ratio is 6-7 in the

461: bubble interiors, while in the inner edge of the 8$\mu$m shells (brightest

462: part of the PDR region) it is 9-10. The 8.0 $\mu$m/4.5 $\mu$m ratio has a

463: slightly larger contrast between the shell and the interior, typically 30-35

464: versus 23-25, respectively. Using the 5.8 $\mu$m/4.5 $\mu$m ratio to determine

465: the PAH-free region, the bubble interiors have radii of 1.2', 2.2', and 1.5'

466: for N10, N21 and N49, respectively.  These are about the same as the radii of

467: the HII regions associated with the bubbles (1.26', 2.6', and 1.5',

468: respectively).  The typical radial thickness of the PDR regions are 0.6',

469: 0.6', and 0.3', respectively. Roger \& Dewdney (1992) use computer models to

470: show how the ratio of the PDR outer radius to the HII radius is related to the

471: ambient density the bubble is expanding into. Based on the ratios 1.4, 1.1 and

472: 1.2 for N10, N21 and N49, we conclude that ambient density is $\sim$10$^3$

473: cm$^{-3}$. These estimates should be taken with caution, however, since they

474: are not consistent with the estimated T$_{eff}$ of the ionizing star that we

475: determine below (see \S 5).  Lastly, there are also some very conspicuous low

476: ratios (dark spots) in the false color images, especially in the

477: 8$\mu$m/4.5$\mu$m images.  These are locations of bright stars, which are

478: bright at 4.5$\mu$m and much fainter at 5.8$\mu$m and even fainter at

479: 8.0$\mu$m.

480: %Interestingly,

481: %YSO candidate 1 (see $\S$5) on the bottom periphery of N49 does not stand out

482: %in these false color images because the star(s) and the immediate region

483: %around it are bright at all MIR bands.

484: 

485: %In Fig. \ref{n10pahc}, we show a grey-scale image of the 5.8$\mu$m/4.5$\mu$m

486: %ratio of N10.  The contour represents where the ratio equals 9. We have

487: %averaged the ratio over 20 pixels along lines of constant longitude (Fig.

488: %\ref{n10pahlat}) and constant latitude (Fix. \ref{n10pahlon}). The ratio peaks

489: %with values between 9 and 10 in the regions corresponding to the 8$\mu$m ring.

490: %Interior and exterior to the 8$\mu$m ring. The ratio has a low value between 6

491: %and 7 primarily isolated to the center of the bubble.  The diameter of

492: %PAH-free interior along b=0.40$^\circ$ is 2.4'. The diameter along a line of

493: %constant latitude is not measurable because the 8$\mu$m emission is absent

494: %toward the south.

495: %

496: %In Fig. \ref{n10pah42}, we show a grey-scale image of the 8.0$\mu$/4.5$\mu$m

497: %ratio of N10.  The contour marks where this ratio equals 25. The bubble

498: %interior has a lower average ratio than the ring or bubble exterior, but the

499: %contrast between the bubble interior and PAH-dominated ring is not as strong.

500: %This difference in behavior between the 5.8$\mu$m/4.5$\mu$m ratio and

501: %8.0$\mu$m/4.5$\mu$m ratio is may be caused by the thermal dust emission which

502: %is stronger at the longer wavelength bands or by the relative strength of the

503: %PAH emission lines in the 5.8 $\mu$m and 8.0 $\mu$m IRAC bands.

504: %

505: %In Figs. \ref{n21pahc} - \ref{n21pahlat}, we show the same grey-scale, contour

506: %maps, cuts along lines constant latitude and longitude for the

507: %5.8$\mu$m/4.5$\mu$m ratio toward N21.  The ratio has a high value over the

508: %ring of 8$\mu$m emission between 8 and 10.  In the interior of the bubble, the

509: %ratio has an average value of 6. The diameter of the PAH-free interior is

510: %difficult to determine because the 8$\mu$m ring is not a full semi-circle. The

511: %distance between the most distant positions of the PAH ring (as measured by

512: %the 5.8$\mu$m/4.5$\mu$m ratio) is 4.5'. In Fig. X, we show the grey-scale and

513: %contour image of the 8.0$\mu$m/4.5$\mu$m ratio toward N21. Similar to N10, a

514: %contour (ratio=23) does clearly delineate the bubble's interior but the

515: %contrast is not as strong as the 5.8 $\mu$m/4.5 $\mu$m ratio. There is

516: %significant variation in the 8.0$\mu$m/4.5$\mu$m ratio in the bubble interior

517: %that is not present in the 5.8$\mu$m/4.5$\mu$m ratio.

518: %

519: %In Figs. \ref{n49pahc} - \ref{n49pahlon}, we show the same grey-scale, contour

520: %maps, cuts along lines constant latitude and longitude for the

521: %5.8$\mu$m/4.5$\mu$m ratio toward N49.  The contour represents where the ratio

522: %equals 8.5. Again, the interior of the bubble has a low value of the ratio

523: %about 6 to 7. Along the 8$\mu$m ring, the ratio is between 9 and 10. The best

524: %measure of the PAH-free interior's size is along a line of constant longitude.

525: %The diameter is 3'. The 8.0$\mu$m/4.5$\mu$m (Fig. \ref{n49pah42}) ratio show

526: %the same larger variation as the sources N10 and N21.

527: 

528: \subsection{Dust Temperature}

529: The hot dust in HII regions competes with the gas for stellar UV photons. It

530: also plays an important role in the processing of radiation and cooling of HII

531: regions.  It is therefore, of high interest to measure the temperature

532: distribution of the dust as a function of distance from the source(s) of

533: ionization. To do this, it is necessary to estimate what fraction of the

534: 8$\mu$m emission interior to the inner face of the 8$\mu$m shell is due to

535: thermal continuum emission as opposed to PAH emission from the front and back

536: sides of the shell.  We assume that the PAH emission is confined to a

537: spherical shell of inner radius r and thickness dr, that the PAH emission is

538: proportional to the line-of-sight depth through the shell (front and back) and

539: that emission is homogeneous.  With these assumptions, the path length S

540: through the shell at each impact parameter y (which, for objects far from the

541: observer, is directly proportional to the angular separation) is:

542: 

543: \begin{eqnarray}

544: %S &=& 2 \times \sqrt{2r^2+2 r dr + dr^2-(2r^2+2r dr)\cos

545: %  \left[\sin^{-1}(\frac{y}{r})-\sin^{-1}\left(\frac{y}{r+dr}\right)\right]}

546: S &=& 2\left(\sqrt{r^2+2r dr +dr^2-y^2}-\sqrt{r^2-y^2}\right)

547: \end{eqnarray}

548: 

549: Figure \ref{n49dust} shows the azimuthally averaged observed radial 8$\mu$m

550: brightness of N49 with the profile predicted by eq.(1) normalized to the

551: observed flux density at the center of the bubble and superimposed on the

552: observations.  Surprisingly, the profile predicted by the assumed simple

553: geometry fits the observations very well.  A check on this assumption is the

554: fact that there is no apparent dip in brightness at 8$\mu$m at the center of

555: N49 but is very apparent at 24$\mu$m and 20 cm.  Taking the good fit of the

556: predicted profile with the observations at face value implies that almost all

557: the 8$\mu$m emission interior to the shell is due to PAH emission from the

558: front and back faces of the shell.  Only from about 45'' to 60'' is there

559: possibly a measurable excess that might be claimed to be thermal continuum

560: emission.  If this is the thermal component at 8$\mu$m, it is too small to

561: measure accurately enough to determine the dust temperature.  Using the noise

562: level of the 8$\mu$m emission at the center of the bubble along with the

563: measured 24$\mu$m emission, we can only place an upper-limit ($<$ 150 K) on

564: the interior dust temperature.

565: 

566: The turndown in 8$\mu$m brightness beginning at $\sim$ 80'' outward indicates

567: the decline in brightness of the PDR.  It declines at larger radii because of

568: dilution of stellar photons that can excite the PAH features.  We have not

569: done a similar analysis for N21 because of its lack of symmetry.

570: 

571: In N10, on the other hand, there appears to be 8 $\mu$m and 5.8 $\mu$m

572: emission coincident with the central 20 cm peak that is significantly stronger

573: than the shell emission (see Figure \ref{n10lat2}). This cannot be front or

574: back-face emission, as in N49, but rather it is likely thermal emission from

575: hot dust.  Unfortunately, the 24 $\mu$m emission is saturated.  Using only the

576: IRAC bands, we can only place an upper-limit ($<$250 K) on the interior dust

577: temperature.

578: 

579: \subsection{Morphology Comparison}

580: It is instructive to compare the MIR morphology of these three bubbles since

581: they are each unique in interesting ways. The most important difference is

582: that in N10 and N21, 24 $\mu$m and 20 cm emission is centrally peaked, whereas

583: in N49 there is a cavity. That is, the driving star in N49 appears to have

584: evacuated its immediate surroundings of hot dust (as traced by 24 $\mu$m

585: emission) and gas (as traced by 20 cm emission). This difference in hot dust

586: structure could be caused by either N49 containing an earlier ionizing star

587: (O5V vs. O6.5V, see \S 5) or by a difference in age (N49 being more evolved

588: than N10). A second difference is the detection of 8 $\mu$m emission toward

589: the center of N10 and N21 but its absence toward the center of N49. This

590: difference implies that the dust at the center of N10's and N21 is hotter than

591: in N49.

592: 

593: Since even cooler O stars destroy PAHs in their neighborhood (\S 4.1), the 8

594: $\mu$m emission at the centers of N10 and N21 is likely dominated by thermal

595: emission from hot dust.  Presumably, small dust grains that give rise to

596: stochastic emission are destroyed near O stars.  Assuming that the 24 and 8

597: $\mu$m emission at the center of N21 are due to thermal emission from large

598: dust grains, we find T$_{dust}$ $\sim$ 200 K.  A similar estimate cannot be

599: made for N10 because the 24 µm emission is saturated; however, using 5.8 and

600: 3.5 $\mu$m emission, which probably samples hotter dust still closer to the

601: central star than in N21, we find T$_{dust}$ $\sim$380 K for N10.  The dust

602: temperatures in N21 and N10 are higher than in N49 even though N49 is ionized

603: by a hotter star because the dust is much closer to the central star(s) in N10

604: and N21 than in N49.

605: 

606: \section{Associated YSOs and Ionizing Stars}

607: 

608: In this section, we examine the stellar populations observed at NIR-MIR

609: wavelengths toward each bubble. We are interested in the YSOs associated with

610: each bubble, especially those appearing on the bubble rims that may have been

611: triggered by the expanding bubble, as well as the ionizing stars responsible

612: for producing each bubble.

613: 

614: To identify YSO candidates, we fit GLIMPSE Archive sources combined with 24

615: $\mu$m photometry from MIPSGAL images with spectral energy distributions

616: (SEDs) from a large, pre-computed grid of YSO models (Robitaille et al.\ 

617: 2006).  The grid consists of 20,000 2-D Monte Carlo radiation transfer models

618: (Whitney et al.\ 2003a,b; Whitney et al.\ 2004) spanning a complete range of

619: stellar mass and evolutionary stage and output at 10 viewing angles (a total

620: of 200,000 SEDs).  The model fitting tool uses a fast $\chi^2$-minimization

621: fitting algorithm (Robitaille et al.\ 2007) and includes a grid of Kurucz

622: (1993) stellar photosphere SEDs. We can robustly distinguish between YSOs and

623: extincted photospheres of main-sequence and giant stars because YSOs require a

624: thermal emission component from circumstellar dust to reproduce the shapes of

625: their mid-IR excesses.  The concept of fitting SEDs from a large grid of

626: models was tested by fitting the SEDs of several known YSOs in Taurus and

627: deriving physical properties in agreement with previous determinations based

628: on other methods (Robitaille et al. 2007). This model-based approach does a

629: reliable job of identifying YSO candidates, and additionally provides

630: information on the physical properties of the YSOs.

631: 

632: In Table \ref{YSOs} we present the YSO candidates listed in order of ascending

633: Galactic longitude for each bubble. For each YSO candidate, the set of

634: well-fit models (numbering $N_{\rm fits}$) was selected on the basis of

635: $\chi^2$ according to

636: \begin{equation}\label{good}

637:   \chi^2-\chi^2_{\rm min} \le 2N_{\rm data},

638: \end{equation}

639: where $\chi_{\rm min}^2$ is the goodness-of-fit parameter of the best-fit

640: model and $4\le N_{\rm data}\le 8$ is the number of flux datapoints used for

641: the fit. From the set of well-fit models we construct the cumulative

642: probability distributions of the YSO parameters. These include mass $M_\star$

643: of the central star, total luminosity (stellar plus disk accretion) $L_{\rm

644:   TOT}$ of the YSO, and envelope accretion rate $\dot{M}_{\rm env}$. In Table

645: \ref{YSOs} we report the ``best'' values of these YSO parameters, defined as

646: the value of each parameter for which the slope of the cumulative probability

647: distribution is maximized. We include the minimum and maximum parameter values

648: representing 95\% confidence intervals.  Table \ref{YSOs} includes the most

649: probable evolutionary stage of the YSO, a classification of the models

650: introduced by Robitaille et al.\ (2006) that parallels the observational

651: ``Class'' taxonomy (Lada 1987).  A Stage I YSO is heavily embedded in its

652: infalling envelope, while a Stage II YSO is a more evolved disk-dominated

653: object. We do not detect any candidate Stage III objects, young stars with

654: remnant dust disks. This may be a selection effect, however, because the SEDs

655: of such sources tend to most closely resemble stellar photospheres, and we

656: have been conservative in our selection of YSO candidates.  The comments

657: column of Table \ref{YSOs} gives the apparent location of the YSO: (Rim) YSO

658: on the projected rim of the bubble and hence a possible example of triggered

659: star formation; (IRDC) YSO within an infrared dark cloud; (Bub) YSO inside the

660: bubble in projection; and (PDR) YSO within the bright diffuse PAH emission

661: from the PDR of the bubble.  We also include a [4.5] flag for two YSO

662: candidates that appear to be associated with bright, extended emission at 4.5

663: $\mu$m.  Excess emission in IRAC [4.5] probably indicates shocked molecular

664: (H$_2$ or CO) outflow or jet, a signpost of the early stages of massive star

665: formation (Smith et al. 2006; Davis et al.2007; Shepherd et al. 2007).

666: 

667: Identifying each bubble's ionizing stars is complicated by the fact that such

668: stars, while characteristically luminous, have reddened Rayleigh-Jeans

669: spectral slopes at NIR--MIR wavelengths. While the stellar photosphere models

670: included in the fitting tool are primarily intended to facilitate the

671: separation of YSO candidates from field stars, it is possible to use the

672: results of the photosphere fits to identify candidate ionizing stars lying

673: within the bubbles. Robitaille et al.\ (2007) have incorporated into the

674: fitting tool an interstellar extinction model using the MIR extinction

675: properties derived by Indebetouw et al. (2005).  To identify candidate hot

676: stars in the bubble, we forced the minimum extinction used by the fitting tool

677: to be 8--10 mag and selected sources for which the best-fit Kurucz model had a

678: photospheric temperature $T_{\rm eff}>20,000$ K. This revealed ${\sim}10$

679: candidate ionizing stars within the boundaries of each bubble. We then fit

680: these sources a second time, allowing the extinction to range from 0--30 mag.

681: The resulting set of well-fit models, selected according to Equation

682: \ref{good}, produces a tight curve of $T_{\rm eff}$ versus stellar radius $R$

683: when scaled to the distance of the bubble. This curve intersects with the

684: theoretical O-star $T_{\rm eff}$--$R$ relations of MSH05 (see Fig.

685: ~\ref{fitter}).  This selects a particular group of models, providing

686: estimates of both the spectral types and the extinction toward each of the

687: candidate ionizing stars in each bubble.  The results of this analysis are

688: presented in Table \ref{ionizers}. The candidate stars are listed in

689: decreasing order of their likely importance in ionizing the \ion{H}{2} regions

690: of the bubbles. The fifth column indicates our best estimate of the ionizing

691: star based on location, spectral type and radio continuum emission. It is

692: important to note that this method of assigning spectral type is not

693: definitive. Any one of the candidate ionizing stars could be a less reddened,

694: foreground, cooler main sequence star or red giant.  Geometrical arguments,

695: however, strongly support the identification of the best candidates as real O

696: stars responsible for producing the bubbles.

697: 

698: \subsection{N10}

699: Out of 687 GLIMPSE Archive sources analyzed within a 3.6' (5.2 pc) radius from

700: the center of N10, 15 were fit with high confidence by YSO models.  These YSO

701: candidates surround N10 (yellow circles in Figure \ref{N10}), and the

702: configuration is highly suggestive of triggered massive star formation.  The

703: bubble is bordered on 2 sides by infrared dark clouds (IRDCs).  4 candidate

704: highly-embedded Stage I massive YSOs appear to be located on the bubble rims

705: (sources N10-7, 5.8--16.7 M$_{\sun}$; N10-9, 8.3--13.5 M$_{\sun}$; N10-11,

706: 8.4--12.7 M$_{\sun}$; and N10-12, 8.9--17.1 M$_{\sun}$, see Table \ref{YSOs}).

707: Two of these YSOs are also in an IRDC and the other two YSOs appear close to

708: an IRDC.

709: 

710: Four stars have been identified as possible ionizing stars located in

711: projection inside the bubble.  It should be emphasized that these stars are

712: well-fit with stellar photospheres and are not YSOs. The spectral energy

713: distributions of these stars suggest spectral types ranging from O7.5V to O6V

714: (see Table 2 and cyan circles in Figure 19).  Although the spectral types are

715: rather uncertain, if we take them at face value and use the models of MSH05,

716: their combined UV photon flux is $\sim$2.2x10$^{49}$ s$^{-1}$, more than

717: adequate to maintain ionization of the HII region (1.6$\times$10$^{49}$

718: s$^{-1}$, based on 20 cm emission). Some UV, however, is required to heat the

719: dust at the center of N10. IN10-1 appears to lie at the center of the bubble

720: and near the center of the radio and saturated diffuse 24 $\mu$m emission

721: filling the bubble, while IN10-2 appears to lie in a sub-cavity near a second

722: peak of bright diffuse MIR emission that may be hiding a cluster of later-type

723: OB stars or YSOs. Preliminary optical spectroscopy obtained using the Wyoming

724: Infrared Observatory (WIRO) indicates that IN10-3 is an O8-B0 V star

725: (Kobulnicky, priv. comm.). Thus, the above method appears to be a robust

726: method of identifying ionizing star candidates. The discrepancy between the

727: spectral identification and the analysis presented above may be caused by the

728: uncertain kinamtic distance used for the bubble.  The presence of the IRDCs on

729: 2 sides of N10 suggest that this bubble may be density-bounded in those

730: directions, allowing the bubble to expand asymmetrically, with the center of

731: influence from these 2 stars offset from the geometric center of the bubble.

732: 

733: %Two stars may be working together to produce this bubble.  IN10-1 (O7.5 V) and

734: %IN10-2 (O6.5 V) in Table \ref{ionizers} (cyan circles in Figure \ref{N10})

735: %together appear to produce only 50\% of the ionizing photons necessary to

736: %maintain the \ion{H}{2} region. If we consider all 4 candidate ionizing stars

737: %of N10 in Table 2, together they could produce 91\% of the UV photon flux

738: %required forionization. Our spectral type determinations are uncertain,

739: %however, and geometrical arguments strongly support the identification of

740: %IN10-1 and IN10-2 as the primary ionizers of this bubble.  

741: 

742: \subsection{N21}

743: N21 is associated with a larger HII region complex. While 21 YSO candidates

744: (yellow circles in Figure \ref{N21}) were selected from the 2333 GLIMPSE

745: Archive sources within 6' (6.5 pc, white circle) from the center of the

746: bubble, more YSO candidates probably lie beyond our search radius. The

747: greatest concentration of YSOs appears in the IRDC located midway between N21

748: and a neighboring bubble in the upper-left corner of Figure \ref{N21} (N22 in

749: the Churchwell et al.  2006 catalog), but it is unclear whether this IRDC is

750: physically associated with either of the bubbles.  Perhaps the most

751: interesting single point source in this image is the very bright greenish-red

752: source that appears to lie inside the bubble.  This source has an extremely

753: positive spectral index in the NIR--MIR (it is undetected in 2MASS $J$ and $H$

754: and saturated in all 4 GLIMPSE bands) and is bright enough at longer MIR

755: wavelengths to be detected in all 4 MSX bands and to saturate the MIPS 24

756: $\mu$m band. Although this spectral index is suggestive of a massive Class I

757: YSO, this source is undetected by MIPS at 70 $\mu$m, while a YSO SED should

758: peak near 70 $\mu$m. We can estimate ranges of extinction and distance

759: consistent with the source being an AGB star by assuming a typical K-magnitude

760: of -7.56 (Sohn et al. 2006). Assuming A$_V <$ 40 mag (larger than any source

761: detected in the field), the distance is $>$ 10 kpc, indicating a background

762: AGB star.

763: 

764: %Although AGB stars can possess powerful winds, it is unlikely that this AGB

765: %star is responsible for producing the bubble. The UV emission from a cool AGB

766: %photosphere would be insufficient to excite the PAH emission on the bubble

767: %rim.  The most massive stars to reach the AGB are 5--6 M$_\sun$ with

768: %main-sequence lifetimes of ${\ga}70$ Myr (Bressan et al.\ 1993), while the

769: %environment of N21 and its \ion{H}{2} region, characterized by numerous IRDCs

770: %and YSOs, appears to be not older than a few Myr. The AGB star is offset from

771: %the bright diffuse 24 $\mu$m and radio emission inside the bubble.  This

772: %source is therefore most likely an unassociated, highly-extincted background

773: %AGB star that appears to lie inside the bubble in projection.

774: 

775: The best candidate star for producing N21 is IN21-1, which lies at the center

776: of the brightest radio and diffuse 24 $\mu$m emission. The model fits to this

777: star, when scaled to the 3.7 kpc kinematic distance of the \ion{H}{2} region,

778: produce a stellar radius too large to be a main-sequence star (see

779: Fig~\ref{fitter}). Instead, the models lie exactly along the $T_{\rm

780:   eff}$--$R$ curve derived by MSH05 for O supergiants, a highly suggestive

781: correspondence. Because the 2 curves overlap the spectral type is degenerate.

782: Preliminary optical spectroscopy obtained using WIRO indicates IN21-1 is a

783: late-O supergiant (Kobulnicky, priv. comm.), again confirming the method of

784: identifying ionizing star candidates. The ionizing photon flux required to

785: maintain the \ion{H}{2} region suggests an early BI.

786: %An O9I star could

787: %supply $\sim$75\% of the required UV photon flux, suggesting either a slightly

788: %earlier spectral type or that other candidate stars in Table 2 are

789: %contributing to ionization.

790: 

791: \subsection{N49}

792: Within a radius of 3.6' (6 pc) from the center of N49, 722 GLIMPSE Archive

793: sources were analyzed and 7 were fit with high confidence as YSOs (yellow

794: circles in Figure \ref{N49}). N49-1 (see Table \ref{YSOs} and Figure

795: \ref{N49YSOs}) was fit with Stage I models ranging in mass from 14 to 29

796: M$_{\sun}$, making it potentially the most massive YSO in our sample.  N49-3

797: sports a spectacular example of a bipolar outflow seen in 4.5 $\mu$m emission.

798: Both YSO candidates appear to lie in an IRDC just beyond the rim of the bubble

799: and adjacent to a bright knot of diffuse 8.0 $\mu$m and 24 $\mu$m emission that

800: probably hides additional YSOs undetected by GLIMPSE because they are masked

801: by the diffuse MIR emission.  Other potential YSOs in the IRDC may not have

802: been detected by GLIMPSE due to confusion and high extinction.  Hence N49-1

803: and N49-3 may be massive members of a YSO cluster that has been triggered by

804: the expansion of the bubble.

805: 

806: N49 contains a star, IN49-1, at the center of the wind-evacuated cavity

807: (Figure \ref{N49}). Given the strong circular symmetry observed in all bands,

808: this star is very likely the ionizing star responsible for producing the

809: bubble. IN49-1 is fit both by an O8 III and an O5 V star (see Fig

810: \ref{fitter}). The UV photon flux implied by the radio emission (10$^{48.89}$

811: s$^{-1}$) is in closer agreement with an O8 III star (10$^{48.88}$ s$^{-1}$)

812: than an O5 V star (10$^{49.22}$ s$^{-1}$). However, the age implied by the

813: presence of triggered star-formation is closer to that of an O5 V star

814: ($\sim$10$^5$ yrs) than that of an O8 III star (10$^{6}$ yrs), so we favor the

815: O5 V classification. Presumably the excess UV is responsible for heating the

816: dust that produces the 24 $\mu$m emission inside the HII region. The model

817: SEDs fit to the 2 brightest sources (N49-1 and N49-3) were based upon the 4

818: IRAC bands plus a 24 $\mu$m lower limit. The 24 $\mu$m lower limit effectively

819: places lower limits on the YSO masses and luminosities, but the upper limits

820: are difficult to constrain based upon the available photometry.

821:  

822: %Since there are several well developed YSOs on the periphery of N49 that we

823: %suspect have been triggered by the bubble's expansion and compression of

824: %ambient ISM, we know that the dynamical age of the bubble has to be at least

825: %several times 10$^5$ years.  This being the case it is of interest to see what

826: %conditions this places on the density of the ambient ISM and to see if this is

827: %consistent with the numerical models of FHY06.  

828: %

829: %Using the analytic relations of Weaver et al. (1977) we can estimate the

830: %dynamical age of the bubble.  We use the stellar parameters of MSH05 to

831: %estimate the mass loss rate and wind luminosities of the ionizing stars using

832: %the prescription of Vink, deKoter, and Lamers (2001).  Having the wind

833: %luminosity and measured bubble radius, we can estimate the age of the bubble

834: %for an assumed ambient density.  Since several Stage I massive YSOs have been

835: %identified along the periphery of the bubble, we know that the bubble has to

836: %be at least several $\times$ 10$^5$ years to have triggered the YSOs because

837: %this is the approximate time scale for massive YSO formation.  we find that an

838: %ISM density of at least 10$^5$ cm$^{-3}$ is required to achive an age of

839: %5.5x10$^5$ yr for an O5V star, which seems consistent with our understanding

840: %of the environments in which O stars form.  Ages greater than $\sim$10$^6$ yrs

841: %begin to require unrealistically high ISM densities.  Freyer et al. (2006),

842: %however, have shown using numerical simulations that the energy transfer

843: %efficiency may be lower than assumed in analytical models by an order of

844: %magnitude. If this result holds, than the age here would be underestimated and

845: %the necessary ambient ISM density would be significantly less. Furthermore, if

846: %the HII region's age is significantly greater, the star-formation may not be

847: %triggered.

848: %

849: %A second approach is to apply the models of FHY06. These models show a HII

850: %region to have already achieved a radius of $\sim$11 pc at an age of

851: %4$\times$10$^5$ yr for an ambient density of 20 cm$^{-3}$.  In analytic models

852: %radius is proportional to density n$^{-0.2}$, which would change the density

853: %from 20 to 10$^5$ cm$^{-3}$ and decrease the radius at 4$\times$10$^5$ years

854: %from $\sim$11 pc to $\sim$2 pc.  This result is consistent with a radius of

855: %2.5 pc at an age $\ge$6x10$^5$ yr for an ambient density of 10$^5$ cm$^{-3}$.

856: %However, some physical processes like instability, turbulent mixing and energy

857: %loss may not scale in such a simple manner, so our consistency should be taken

858: %cautiously.  Garcia-Segura \& Franco (1996) present numerical simulations of

859: %HII regions expanding into dense ambient environments and find consistent

860: %results.  We conclude that the analytic and numerical models of bubble

861: %evolution are in reasonable agreement with observations, although numerical

862: %models that explore higher density regimes are necessary to confirm this

863: %result.

864: 

865: The central cavity in N49 seems to indicate that this bubble is stellar

866: wind-dominated and therefore should be describable by the analytic relations

867: of Weaver et al. (1977):

868: 

869: \begin{eqnarray*}

870: R(t) &\propto &n_o^\frac{-1}{5} L_w^\frac{1}{5} t^\frac{3}{5}

871: \end{eqnarray*}

872: where R(t) is the radius, n$_o$ is the initial ambient density, L$_w$ is the

873: stellar wind luminosity, and t is the dynamical age of the bubble.  From this,

874: one can estimate the age of a bubble as a function of no given the wind

875: luminosity and a measured radius.  We use the stellar parameters of MSH05 to

876: estimate the mass loss rate (1.5 x 10$^{-6}$ M$_\odot$ yr$^{-1}$) and wind

877: luminosity of the ionizing star ($\sim$4 x 10$^{36}$ erg s$^{-1}$) using the

878: prescription of Vink, de Koter, and Lamers (2001).  Having the wind luminosity

879: and measured bubble radius, we can estimate the age of the bubble for an

880: assumed ambient density.  Several Class I or younger YSOs have been identified

881: along the periphery of the bubble.  Assuming that the YSOs were triggered by

882: the expansion of the bubble would require the bubble to be at least 10$^{5}$

883: yr old because this is the time scale for massive YSO formation.

884: %We estimate the

885: %swept-up mass to be $\sim$3 $\times$ 10$^{4-5}$ M$_\odot$ (based on NH3

886: %imaging and assuming that the PDR shell represents the volume of the swept-up

887: %shell, Cyganowski, priv. comm.), which implies an initial ambient density of 2

888: %$\times$ 10$^{3-4}$ cm$^{-3}$.  

889: Using the estimated initial ambient density of $\ge$10$^3$ cm$^{-3}$, the

890: Weaver et al.  (1977) relations implies a dynamical age $\ge$10$^{5}$ yr, a

891: very young bubble but easily old enough to have spawned a second generation of

892: star formation.  Freyer et al. (2006), however, have shown using numerical

893: simulations that the energy transfer efficiency may be lower than assumed in

894: analytic models.  If this is the case, then the age estimates based on the

895: analytic relations will be under-estimated, giving even more time for

896: triggered star formation to occur.

897: 

898: 

899: \section{Conclusions}

900: 

901: Based on MIR and 20 cm observations of three bubbles, we conclude the

902: following:\\

903: $\bullet$ At the center of each bubble there is a region containing ionized

904: gas and hot dust delineated by radio free-free emission and 24 $\mu$m

905: emission, respectively.\\

906: $\bullet$ At the center of N10 and N21 (but not N49), the hot dust is also

907: traced by 5.8 and 8.0 $\mu$m emission.\\

908: %$\bullet$ The dust is, at most, 150 K.\\

909: $\bullet$ Based on the decreased ratio of 8.0 $\mu$m/4.5 $\mu$m, 5.8

910: $\mu$m/4.5 $\mu$m and the lack of a similar decrease in 8.0 $\mu$m/3.6 $\mu$m,

911: we conclude that inside the 8 $\mu$m shell PAHs are destroyed by hard, direct

912: stellar UV radiation.\\

913: $\bullet$ Based on the increased 8.0 $\mu$m/4.5 $\mu$m, 5.8 $\mu$m/4.5 $\mu$m,

914: the bright 8.0 $\mu$m ring emission is dominated by PAH emission that defines

915: the PDR region around each bubble.\\

916: $\bullet$ We have identified YSO candidates and probable ionizing sources for

917: each bubble. This was accomplished by employing the SED model fitter developed

918: by Robitaille et al. (2007) to fit model SEDs of YSOs and hot stellar

919: photospheres (Kurucz 1993) to the observed fluxes.\\

920: $\bullet$ Based on morphology and environment, several of the identified YSO

921: candidates in N10 and N49 appear to be triggered by expansion of the

922: bubbles.\\

923: $\bullet$ The wind-blown cavity at the center of N49 appears to be produced by

924: a central 05 V star, the hottest ionizing star observed here. This bubble

925: appears to be dominated by the wind from the O5 star and have a dynamical age

926: of $\ge$10$^5$ yrs.\\

927: %  $\bullet$ When the ambient density is scaled to 10$^{5}$

928: %cm$^{-3}$, the numerical models of FHY06 are consistent with the measured

929: %radii of N49 and with the age (10$^5$ yrs) necessary for YSOs to be

930: %triggered by the bubble expansion.\\

931: 

932: \acknowledgements 

933: We would like to acknowledge Chip Kolbunicky for obtaining

934: optical spectra of the candidate ionizing stars. An anonymous referee made

935: many comments which improved the paper. E.B.C would like to acknowledge

936: support through NASA contract \# 1275394.

937: 

938: \begin{thebibliography}{}

939: %\bibitem[Allen et al.(2004)]{2004ApJS..154..363A} Allen, L.~E., et al.\ 

940: %2004, \apjs, 154, 363 

941: %\bibitem[Arthur \& Hoare(2006)]{2006ApJS..165..283A} Arthur, S.~J., \& 

942: %Hoare, M.~G.\ 2006, \apjs, 165, 283 

943: \bibitem[Benjamin et al.(2003)]{2003PASP..115..953B} Benjamin, R.~A., et 

944: al.\ 2003, \pasp, 115, 953 

945: %\bibitem[Bressan et al.(1993)]{AB93} Bressan, A. Fagotto, F., Bertelli,

946: %  G., \& Chiosi, C.  1993, \aaps, 100, 647

947: 

948: \bibitem[Carey et al.(2005)]{2005AAS...207.6333C} Carey, S.~J., et al.\ 

949: 2005, Bulletin of the American Astronomical Society, 37, 1252 

950: 

951: \bibitem[aa]{aa}Castor, J., McCray, R., \& Weaver, R. 1975, ApJ, 200, L107

952: 

953: \bibitem[baa]{baa} Capriotti, E.R. \& Kozminski, J.F. 2001, PASP, 113, 677

954: 

955: \bibitem[Churchwell \& GLIMPSE Team(2001)]{2001AAS...198.2504C} Churchwell, 

956: E., \& GLIMPSE Team 2001, Bulletin of the American Astronomical Society, 

957: 33, 821 

958: 

959: \bibitem[Churchwell et al.(2006)]{2006ApJ...649..759C} Churchwell, E., et 

960: al.\ 2006, \apj, 649, 759 

961: 

962: \bibitem[Churchwell et al.(2007)]{2006ApJ...649..759C} Churchwell, E., et 

963: al.\ 2007, \apj, submitted

964: 

965: \bibitem[Cohen et al.(2007)]{2007MNRAS.374..979C} Cohen, M., et al.\ 2007, 

966: \mnras, 374, 979

967: 

968: %\bibitem[aa]{aa}Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F.,

969: 

970: \bibitem[Davis et al.(2007)]{CD07} Davis, C. J., Kumar, M. S. N.,

971:   Sandell, G., Froebrich, D., Smith, M. D., \& Currie, M. J. 2007,

972:   \mnras, 374, 29  

973: %Perley, R. A., Taylor, G. B., \& Broderick, J. J. 1998, AJ, 115, 1693

974: 

975: \bibitem[Deharveng et al.(2005)]{2005A&A...433..565D} Deharveng, L., 

976: Zavagno, A., \& Caplan, J.\ 2005, \aap, 433, 565 

977: 

978: \bibitem[Draine(2003)]{2003ARA&A..41..241D} Draine, B.~T.\ 2003, \araa, 41, 

979: 241 

980: 

981: \bibitem[asdf]{asdf}Fazio, G.G., et al. 2004 ApJS, 154, 87

982: 

983: \bibitem[Freyer et al.(2003)]{2003ApJ...594..888F} Freyer, T., Hensler, G., 

984: \& Yorke, H.~W.\ 2003, \apj, 594, 888 

985: 

986: \bibitem[aa]{aa}Freyer, T., Hensler, G., Yorke, H.W. 2006, ApJ, 628, 262 [FHY06]

987: 

988: \bibitem[Garcia-Segura \& Franco(1996)]{1996ApJ...469..171G} Garcia-Segura, 

989: G., \& Franco, J.\ 1996, \apj, 469, 171 

990: 

991: \bibitem[Hoare(1990)]{1990MNRAS.244..193H} Hoare, M.~G.\ 1990, \mnras, 244, 

992: 193 

993: 

994: \bibitem[Hoare et al.(1991)]{1991MNRAS.251..584H} Hoare, M.~G., Roche, 

995: P.~F., \& Glencross, W.~M.\ 1991, \mnras, 251, 584 

996: \bibitem[Hoare et al.(2007)]{2007prpl.conf..181H} Hoare, M.~G., Kurtz, 

997: S.~E., Lizano, S., Keto, E., \& Hofner, P.\ 2007, Protostars and Planets V, 

998: 181 

999: 

1000: \bibitem[Helfand et al.(2006)]{2006AJ....131.2525H} Helfand, D.~J., Becker, 

1001: R.~H., White, R.~L., Fallon, A., \& Tuttle, S.\ 2006, \aj, 131, 2525 

1002: \bibitem[Indebetouw et al.(2005)]{IM05} Indebetouw, R. et al.\ 2005,

1003:   \apj, 619, 931

1004: \bibitem[Kassis et al.(2006)]{2006ApJ...637..823K} Kassis, M., Adams, 

1005: J.~D., Campbell, M.~F., Deutsch, L.~K., Hora, J.~L., Jackson, J.~M., \& 

1006: Tollestrup, E.~V.\ 2006, \apj, 637, 823 

1007: 

1008: \bibitem[Kessler et al.(1996)]{1996A&A...315L..27K} Kessler, M.~F., et al.\ 

1009: 1996, \aap, 315, L27 

1010: 

1011: \bibitem[aa]{aa}Kr\"oger, D., Hensler, G., Freyer, T., 2006 A\&A, 450, L5

1012: 

1013: \bibitem[aa]{aa}Kr\"oger, D., Freyer, T., Hensler, G., Yorke, H.W. 2007 A\&A, submitted

1014: 

1015: \bibitem[Kurucz(1993)]{1993yCat.6039....0K} Kurucz, R.~L.\ 1993, VizieR 

1016: Online Data Catalog, 6039, 0 

1017: \bibitem[Lada(1987)]{CL87} Lada, C. J.  1987, in IAU Symp. 115: Star

1018:   Forming Regions, eds. M. Peimbert \& J. Jugaku, 1

1019: 

1020: \bibitem[aa]{aa}Martins, F., Schaerer, D., Hillier, D.J. 2005, A\&A, 436, 1049

1021: 

1022: \bibitem[Peeters et al.(2003)]{2003asdu.confE..42P} Peeters, E., Tielens, 

1023: A.~G.~G.~M., Allamandola, L.~J., Bauschlicher, C.~W., Boogert, A.~C.~A., 

1024: Hayward, T.~L., Hudgins, D.~M., \& Sandford, S.~A.\ 2003, Astrophysics of 

1025: Dust p.42

1026: 

1027: \bibitem[Peeters et al.(2005)]{2005SSRv..119..273P} Peeters, E., 

1028: Mart{\'{\i}}n-Hern{\'a}ndez, N.~L., Rodr{\'{\i}}guez-Fern{\'a}ndez, N.~J., 

1029: \& Tielens, X.\ 2005, Space Science Reviews, 119, 273 

1030: \bibitem[Povich et al.(2007)]{2007ApJ...660..346P} Povich, M.~S., et al.\ 

1031: 2007, \apj, 660, 346 

1032: 

1033: \bibitem[aa]{aa} Price, S.D., Egan, M.P., Carey, S.J., Mizuno, D., Kuchar, T.

1034:   2001, A\&A, 121, 2819

1035: 

1036: \bibitem[Rieke et al.(2004)]{2004ApJS..154...25R} Rieke, G.~H., et al.\ 

1037: 2004, \apjs, 154, 25 

1038: 

1039: \bibitem[Robitaille et al.(2006)]{2006ApJS..167..256R} Robitaille, T.~P., 

1040: Whitney, B.~A., Indebetouw, R., Wood, K., \& Denzmore, P.\ 2006, \apjs, 

1041: 167, 256 

1042: 

1043: \bibitem[Robitaille et al.(2007)]{2007ApJS..169..328R} Robitaille, T.~P., 

1044: Whitney, B.~A., Indebetouw, R., \& Wood, K.\ 2007, \apjs, 169, 328 

1045: \bibitem[Smith et al.(2006)]{HS06} Smith, H. A., Hora, J. L., Marengo, M., \& Pipher, J. L. 2006 \apj, 645, 1264 

1046: \bibitem[Shepherd et al.(2007)]{SP07} Shepherd, D. S. et al.\ 2007,

1047:   \apj, submitted

1048: 

1049: \bibitem[Sohn et al.(2006)]{2006A&A...445...69S} Sohn, Y.-J., Kang, A., 

1050: Rhee, J., Shin, M., Chun, M.-S., \& Kim, H.-I.\ 2006, \aap, 445, 69 

1051: 

1052: \bibitem[aa]{aa}Weaver, R., McCray, R., Castor, J., Shapiro, P., Moore, R. 1977, ApJ, 218, 377

1053: 

1054: \bibitem[Werner et al.(2004)]{2004ApJS..154....1W} Werner, M.~W., et al.\ 

1055: 2004, \apjs, 154, 1 

1056: 

1057: \bibitem[Whitney et al.(2003a)]{WI} Whitney, B. A., Wood, K.,

1058:   Bjorkman, J. E., \& Wolff, M. J. 2003, \apj, 591, 1049

1059: 

1060: \bibitem[Whitney et al.(2003b)]{2003ApJ...598.1079W} Whitney, B.~A., Wood, 

1061: K., Bjorkman, J.~E., \& Cohen, M.\ 2003, \apj, 598, 1079 

1062: 

1063: \bibitem[Whitney et al.(2004)]{2004ApJ...617.1177W} Whitney, B.~A., 

1064: Indebetouw, R., Bjorkman, J.~E., \& Wood, K.\ 2004, \apj, 617, 1177 

1065: 

1066: \bibitem[aa]{aa}Whitworth, A.P., Bhattal, A.S., Chapman, S.J., Disney, J.J. \& Turner,

1067: J.A. 1994, MNRAS, 268, 291

1068: 

1069: \bibitem[aa]{aa}Vink, J.S., deKoter, A., Lamers, H.J.G.L. 2001, A\&A, 369, 574

1070: \end{thebibliography}

1071: 

1072: \begin{figure}

1073: \caption{N10, 24 $\mu$m (red), 8 $\mu$m (green) and 4.5 $\mu$m (blue). 20 cm

1074:   (contours) in bottom panel. Note that the 24 $\mu$m emission is saturated at

1075:   the center of the image. The white dashed line on the top figure indicates

1076:   the location of the cross-cut in figures 2-4.}

1077: \label{n10color}

1078: %\plotone{ps/N10colornocontour.ps}

1079: \plotone{f1a.eps}

1080: \plotone{f1b.ps}

1081: \end{figure}

1082: 

1083: %\begin{figure}

1084: %\caption{N10: 3.6 $\mu$m, the circle indicates the 8 $\mu$m shell size.}

1085: %\label{n10mono}

1086: %\plotone{ps/N10mono.ps}

1087: %\end{figure}

1088: 

1089: %\begin{figure}

1090: %\caption{N10: Slice at constant longitude. 20 cm (solid, magnified 100x), 24 $\mu$m (dotted) and 8 $\mu$m (dashed)}

1091: %\label{n10long1}

1092: %\psfig{file=ps/N10sliceb.ps,width=3in}

1093: %\end{figure}

1094: 

1095: %\begin{figure}

1096: %\caption{N10: Slice at constant longitude. 3.6 $\mu$m (solid), 4.5 $\mu$m (dotted) and

1097: %  5.8 $\mu$m (dashed)}

1098: %\label{n10long2}

1099: %\psfig{file=ps/N10sliceb2.ps,width=3in}

1100: %\end{figure}

1101: 

1102: \begin{figure}

1103: %\caption{N10: Slice at latitude b=0.04$^\circ$. 20 cm (solid, magnified

1104: %  100x), 24 $\mu$m (dotted) and 8 $\mu$m (dashed, magnified 5x). The location

1105: %  of the 8 $\mu$m shell and central 24 $\mu$m hot dust emission are indicated.

1106: %  Note that the 24 $\mu$m emission is saturated in the center of the slice,

1107: %  resulting in the strong dip and missing data between longitudes

1108: %  13.18$^\circ$ and 13.20$\circ$.}

1109: \caption{N10: Slice at latitude b=0.04$^\circ$. 20 cm (solid, magnified

1110:   10$^6$), 24 $\mu$m (dotted) and 8 $\mu$m (dashed, magnified 5x). The location

1111:   of the 8 $\mu$m shell and central 24 $\mu$m hot dust emission are indicated.

1112:   Note that the 24 $\mu$m emission is saturated in the center of the slice,

1113:   resulting in the strong dip and missing data between longitudes

1114:   13.18$^\circ$ and 13.20$\circ$.}

1115: \label{n10lat1}

1116: %\plotone{ps/N10slicel.ps}

1117: \plotone{f2.eps}

1118: \end{figure}

1119: 

1120: \clearpage

1121: \begin{figure}

1122: \caption{N10: Slice at latitude b=0.04$^\circ$. 3.6 $\mu$m (solid), 4.5

1123:   $\mu$m (dotted) and 5.8 $\mu$m (dashed). The spikes in 3.6 $\mu$m and 4.5

1124:   $\mu$m emission indicate stars.}

1125: \label{n10lat2}

1126: %\plotone{ps/N10slicel2.ps}

1127: \plotone{f3.eps}

1128: \end{figure}

1129: 

1130: %\begin{figure}

1131: %\caption{N10: Slice at latitude b=0.04$^\circ$. 5.8 $\mu$m (solid) and 8.0

1132: %  $\mu$m (dashed). The location of the 8 $\mu$m shell is marked.}

1133: %\label{n10lat3}

1134: %%\plotone{ps/N10slicel34.ps}

1135: %\plotone{f4.eps}

1136: %\end{figure}

1137: 

1138: %\begin{figure}

1139: %\caption{N10: Dust temperature map measured from 21 $\mu$m (MSX) / 8 $\mu$m}

1140: %\label{n10temp}

1141: %\psfig{file=ps/N10temp.ps,width=3in}

1142: %\end{figure}

1143: \begin{figure}

1144: \caption{N21, 24 $\mu$m (red), 8 $\mu$m (green), 4.5 $\mu$m (blue) and 20 cm

1145:   (contours) in the bottom panel.The white dashed line on the top figure

1146:   indicates the location of the cross-cut in figures 6 \& 7.}

1147: \label{n21color}

1148: %\plotone{ps/N21colornocontour.ps}

1149: \plotone{f4a.eps}

1150: \plotone{f4b.ps}

1151: \end{figure}

1152: 

1153: %\begin{figure}

1154: %\caption{N21: 3.6 $\mu$m}

1155: %\label{n21mono}

1156: %\plotone{ps/n21mono.ps}

1157: %\end{figure}

1158: 

1159: \begin{figure}

1160: \caption{N21: Slice at longitude l=18.19$^\circ$. 20 cm (solid, magnified

1161:   10$^6$x), 24 $\mu$m (dotted) and 8 $\mu$m (dashed, magnified 5x)}

1162: \label{n21lon1}

1163: \plotone{f5.eps}

1164: \end{figure}

1165: 

1166: \begin{figure}

1167: \caption{N21: Slice at longitude l=18.19$^\circ$. 3.6 $\mu$m (solid), 4.5

1168:   $\mu$m (dotted) and 5.8 $\mu$m (dashed). The spikes in 3.6 $\mu$m and 4.5

1169:   $\mu$m emission indicate stars.}

1170: \label{n21lon2}

1171: \plotone{f6.eps}

1172: \end{figure}

1173: 

1174: %\begin{figure}

1175: %\caption{N21: A comparison of 8 $\mu$m (green) and 24 $\mu$m (red) emission

1176: %  along the shell. Note that the small-scale structure in 8 $\mu$m and 24

1177: %  $\mu$m is similar. Since there is not significant PAH emission in the 24

1178: %  $\mu$m band, there appears to be significant thermal emission from the

1179: %  shell.}

1180: %\label{n21_8_24}

1181: %\plotone{ps/f8a.eps}

1182: %\plotone{ps/f8b.eps}

1183: %\end{figure}

1184: 

1185: %\begin{figure}

1186: %\caption{N21: Latitude slice at constant longitude (as marked in Fig. \ref{n10pah} of PAH destruction as measured from 5.8 $\mu$m (MSX) / 4.5 $\mu$m}

1187: %\label{n21pahlat}

1188: 

1189: %\end{figure}

1190: %\begin{figure}

1191: %\caption{N21: Latitude cut (as marked in Fig. \ref{n10pah} of PAH destruction as measured from 5.8 $\mu$m (MSX) / 4.5 $\mu$m}

1192: %\label{n21pahlon}

1193: %\psfig{file=ps/n21pahlon.ps,width=3in}

1194: %\end{figure}

1195: 

1196: %\begin{figure}

1197: %\caption{N21: PAH destruction as measured from 8.0 $\mu$m (MSX) / 4.5 $\mu$m}

1198: %\label{n21pah42}

1199: 

1200: %\end{figure}

1201: %\begin{figure}

1202: %\caption{N21: Latitude slice at constant longitude (as marked in Fig. \ref{n21pah42} of PAH destruction as measured from 8.0 $\mu$m (MSX) / 4.5 $\mu$m}

1203: %\label{n21pah42lat}

1204: %

1205: %\end{figure}

1206: %\begin{figure}

1207: %\caption{N21: Latitude cut (as marked in Fig. \ref{n21pah42} of PAH destruction as measured from 8.0 $\mu$m (MSX) / 4.5 $\mu$m}

1208: %\label{n21pah42lon}

1209: %\psfig{file=ps/n21pah42lon.ps,width=3in}

1210: %\end{figure}

1211: 

1212: 

1213: \clearpage

1214: %\begin{figure}

1215: %\caption{N49: 3.6 $\mu$m}

1216: %\label{n49mono}

1217: %\plotone{ps/n49mono.ps}

1218: %\end{figure}

1219: 

1220: \begin{figure}

1221: \caption{N49, 24 $\mu$m (red), 8 $\mu$m (green), 4.5 $\mu$m (blue) and 20 cm

1222:   (contours) in bottom panel. The white dashed line on the top figure

1223:   indicates the location of the cross-cut in figures 10 \& 11.}

1224: \label{n49color}

1225: %\plotone{ps/N49colornocontour.ps}

1226: \plotone{f7a.eps}

1227: \plotone{f7b.ps}

1228: \end{figure}

1229: 

1230: %\begin{figure}

1231: %\caption{N49: Slice at constant longitude. 20 cm (solid, magnified 100x), 24 $\mu$m (dotted) and 8 $\mu$m (dashed)}

1232: %\label{n49lon1}

1233: %\psfig{file=ps/N049sliceb.ps,width=3in}

1234: %\end{figure}

1235: %

1236: %\begin{figure}

1237: %\caption{N49: Slice at constant longitude. 3.6 $\mu$m (solid), 4.5 $\mu$m (dotted) and

1238: %  5.8 $\mu$m (dashed)}

1239: %\label{n49lon2}

1240: %\psfig{file=ps/N049sliceb2.ps,width=3in}

1241: %\end{figure}

1242: 

1243: \begin{figure}

1244: \caption{N49: Slice at latitude b=-0.23$^\circ$. 20 cm (solid, magnified

1245:   10$^6$x), 24 $\mu$m (dotted) and 8 $\mu$m (dashed, magnified 5x). Note that

1246:   there is no central peak at 24 $\mu$m as there is in N10 and N21.}

1247: \label{n49lat1}

1248: \plotone{f8.ps}

1249: \end{figure}

1250: 

1251: %\begin{figure}

1252: %\caption{N49: Slice at latitude b=-0.23$^\circ$. 3.6 $\mu$m (solid), 4.5

1253: %  $\mu$m (dotted) and 5.8 $\mu$m (dashed). The spikes in 3.6 $\mu$m and 4.5

1254: %  $\mu$m emission are caused by stars.}

1255: %\label{n49lat2}

1256: %\plotone{f10.ps}

1257: %\end{figure}

1258: 

1259: %\begin{figure}

1260: %\caption{N49: Dust temperature map measured from 24 $\mu$m (MIPSGAL) / 8 $\mu$m}

1261: %\label{n49temp}

1262: %\psfig{file=ps/N49temp.ps,width=3in}

1263: %\end{figure}

1264: \begin{figure}

1265: \caption{N10: PAH destruction as measured from 5.8 $\mu$m / 4.5 $\mu$m

1266:   (upper-left) and 8.0 $\mu$m / 4.5 $\mu$m (lower-left). The color scale

1267:   ranges from 5 to 12 (upper-left, blue-to-red) and 24 to 44 (lower-left,

1268:   blue-to-red). The contours represent ratios of 8.5 (upper-left) and 25

1269:   (lower-left). Dashed white lines at left indicate position of slices show at

1270:   right.}

1271: \label{n10pahc}

1272: \epsscale{1.2}

1273: \plottwo{f9a.ps}{f9b.ps}

1274: \plottwo{f9c.ps}{f9d.ps}

1275: \end{figure}

1276: 

1277: \begin{figure}

1278: \caption{N21: PAH destruction as measured from 5.8 $\mu$m / 4.5 $\mu$m

1279:   (upper-left) and 8.0 $\mu$m / 4.5 $\mu$m (lower-left). The color scale

1280:   ranges from 4 to 12 (upper-left, blue-to-red) and 15 to 40 (lower-left,

1281:   blue-to-red). The contours represent ratios of 7.5 (upper-left) and 25

1282:   (lower-left). Dashed white lines at left indicate position of slices show at

1283:   right.}

1284: \label{n21pah}

1285: \plottwo{f10a.ps}{f10b.ps}

1286: \plottwo{f10c.ps}{f10d.ps}

1287: \end{figure}

1288: 

1289: \begin{figure}

1290: \caption{N49: PAH destruction as measured from 5.8 $\mu$m / 4.5 $\mu$m

1291:   (upper-left) and 8.0 $\mu$m / 4.5 $\mu$m (lower-left). The color scale

1292:   ranges from 7 to 13 (upper-left, blue-to-red) and 15 to 42 (lower-left,

1293:   blue-to-red). The contours represent ratios of 8.5 (upper-left) and 28

1294:   (lower-left). Longitude slices at constant latitude (upper-right and

1295:   lower-right). Dashed white lines at left indicate position of slices show at

1296:   right.}

1297: \label{n49pah}

1298: \plottwo{f11a.ps}{f11b.ps}

1299: \plottwo{f11c.ps}{f11d.ps}

1300: \end{figure}

1301: %\begin{figure}

1302: %\caption{N49: Longitude cut (as marked in Fig. \ref{n10pah} of PAH destruction as measured from 5.8 $\mu$m (MSX) / 4.5 $\mu$m}

1303: %\label{n49pahlat}

1304: %\psfig{file=ps/n49pahlat.ps,width=3in}

1305: %\end{figure}

1306: %\begin{figure}

1307: %\caption{N49: Longitude slice at constant latitude (as marked in Fig. \ref{n10pah} of PAH destruction as measured from 5.8 $\mu$m (MSX) / 4.5 $\mu$m}

1308: %\label{n49pahlon}

1309: %

1310: %\end{figure}

1311: %

1312: %\begin{figure}

1313: %\caption{N49: PAH destruction as measured from 8.0 $\mu$m (MSX) / 4.5 $\mu$m}

1314: %\label{n49pah42}

1315: %\end{figure}

1316: %%\begin{figure}

1317: %%\caption{N49: Longitude cut (as marked in Fig. \ref{n10pah} of PAH destruction as measured from 8.0 $\mu$m (MSX) / 4.5 $\mu$m}

1318: %%\label{n49pah42lat}

1319: %%\psfig{file=ps/n49pah42lat.ps,width=3in}

1320: %%\end{figure}

1321: %\begin{figure}

1322: %\caption{N49: Longitude slice at constant latitude (as marked in Fig. \ref{n10pah} of PAH destruction as measured from 8.0 $\mu$m (MSX) / 4.5 $\mu$m}

1323: %\label{n49pah42lon}

1324: %\end{figure}

1325: 

1326: \begin{figure}

1327: \caption{N49: Observed, azimuthally averaged, radial profile of 8 $\mu$m

1328:   emission (+) compared with a model of shell emission (solid line) normalized

1329:   to the observed brightness at the center of N49.}

1330: \label{n49dust}

1331: \epsscale{1}

1332: \plotone{f12.ps}

1333: \end{figure}

1334: 

1335: \begin{figure}

1336:   \plotone{f13.eps}

1337:   \caption{Examples of spectral type determinations for candidate ionizing  

1338:     stars. The set of well-fit Kurucz (1993) stellar atmosphere models to the

1339:     NIR--MIR SEDs of the best candidate ionizing stars in N49 and N21 have

1340:     been plotted in $T_{\rm eff}$--$R$ space, where $R$ is calculated by

1341:     scaling the model fits to the distance of their respective bubbles. The

1342:     O-star $T_{\rm eff}$--$R$ relations of Martins et al. (2005) are

1343:     overplotted as heavy curves for dwarfs, giants, and supergiants. The

1344:     curves extend from types O3 to O9.5 for each luminosity class. From the

1345:     intersections of the curves with the loci of model fits, a spectral type

1346:     of O5 V or late O III is consistent with the observed fluxes of IN49-1.

1347:     The spectral type of IN21-1 is degenerate among O supergiants. These

1348:     degeneracies in spectral type can be lifted by considering the ionizing

1349:     flux required to produce the \ion{H}{2} regions in which the stars are

1350:     located. \label{fitter}}

1351: \end{figure}

1352: 

1353: 

1354: 

1355: \begin{figure}

1356:   \plotone{f14.eps}

1357:   \caption{N10 image in 3.6 $\mu$m (blue), 4.5 $\mu$m (green) and 8.0 $\mu$m

1358:     (red). Positions of YSO candidates in Table \ref{YSOs} and candidate

1359:     ionizing stars in Table \ref{ionizers} are marked with yellow and cyan

1360:     circles, respectively. The large white circle shows the area in which the

1361:     GLIMPSE Point Source Archive was searched for YSO candidates.\label{N10}}

1362: \end{figure}

1363: 

1364: %\begin{figure}

1365: %  \plottwo{ps/f20a.eps}{ps/f20b.eps}

1366: %  \caption{Model fits to the SEDs of 2 candidate massive YSOs

1367: %    appearing on the rim of bubble N10. Measured broadband fluxes and lower

1368: %    limits are plotted as heavy dots with error bars and triangles,

1369: %    respectively. The set of well-fit model SEDs are overplotted as curves,

1370: %    with the heavy curve showing the best fit. {\it Left panel:} While the

1371: %    best-fit model for YSO N10-9 gives a low-mass star, the majority of good

1372: %    fits indicate a massive star.  The 24 $\mu$m flux of this object is very

1373: %    uncertain due to the steep, bright diffuse background gradient on the rim

1374: %    of the bubble. {\it Right panel:} YSO N10-12 is very bright in GLIMPSE and

1375: %    saturated at 24 $\mu$m.\label{N10YSOs}}

1376: %\end{figure}

1377: 

1378: \clearpage

1379: \begin{figure}

1380:   \plotone{f15.eps}

1381:   \caption{Same as Fig.\ \ref{N10}, but showing bubble N21. The

1382:     MIR-bright asymptotic giant branch star is labeled. Although this star

1383:     appears to lie inside the bubble, it is most likely a background AGB

1384:     star.\label{N21}}

1385: \end{figure}

1386: 

1387: \clearpage

1388: \begin{figure}

1389:   \plotone{f16.eps}

1390:   \caption{Same as Fig.\ \ref{N10}, but showing bubble N49.\label{N49}}

1391: \end{figure}

1392: 

1393: \clearpage

1394: \begin{figure}

1395:   \plottwo{f17a.eps}{f17b.eps}

1396:   \caption{Model fits to the SEDs of 2 candidate massive YSOs

1397:     appearing on the rim of bubble N49. Measured broadband fluxes and lower

1398:     limits are plotted as heavy dots with error bars and triangles,

1399:     respectively. The set of well-fit model SEDs are overplotted as curves,

1400:     with the heavy curve showing the best fit. {\it Left panel:} YSO N49-1 is

1401:     extremely red in the GLIMPSE bands. {\it Right panel:} YSO N49-3 is

1402:     associated with the brightest extended 4.5 $\mu$m emission in our sample,

1403:     indicative of a powerful molecular outflow. Because the 4.5 $\mu$m emission is

1404:     extended, and the YSO models do not incorporate molecular line emission,

1405:     this band was not used in fitting the SED of this object.  These sources

1406:     are very close together, and their 24 $\mu$m emission is borderline

1407:     confused and saturated, so lower limits were employed for the

1408:     fitting.\label{N49YSOs}}

1409: \end{figure}

1410: 

1411: \begin{deluxetable}{lcrrrrcccccccccrr}

1412: \tabletypesize{\tiny}

1413: \rotate

1414: \tablecaption{Model Parameters for YSO Candidates Associated with the

1415:   Bubbles \label{YSOs}}

1416: 

1417: \tablewidth{0pt}

1418: %%\tablecolumns{20}

1419: \tablehead{

1420:   \multicolumn{3}{c}{ } & \multicolumn{3}{c}{$M_{\star}$ (M$_{\sun}$)} & \multicolumn{3}{c}{$L_{\rm TOT}$ (L$_{\sun})$} & \multicolumn{3}{c}{$\dot{M}_{\rm env}$ (M$_{\sun}$ yr$^{-1}$)} & \multicolumn{2}{c}{ } \\ 

1421:   \colhead{ID} & \colhead{Name (G$l+b$)} & \colhead{$N_{\rm fits}$} & \colhead{best} & \colhead{min} & \colhead{max} & \colhead{best} & \colhead{min} & \colhead{max}  & \colhead{best} & \colhead{min} & \colhead{max} & Stage & Comments\tablenotemark{a} \\

1422: }

1423: \startdata

1424: N10-1 &  G013.1536+00.0040 &   86 &    .7 &    2.9 &    7.3 &    315 &     76 &    532 &   2.3E-04 &  4.1E-05 &  9.4E-04 & I &  \\

1425: N10-2 &  G013.1667+00.0074 &   12 &    .1 &    1.0 &    5.5 &     26 &     15 &    682 &   1.5E-04 &  0.0E+00 &  4.1E-04 & I &  \\

1426: N10-3 &  G013.1670+00.0794 &    7 &    .6 &    0.9 &    3.0 &     43 &     33 &    128 &   3.3E-04 &  2.1E-04 &  4.1E-04 & I & IRDC \\

1427: N10-4 &  G013.1686+00.0798 &    3 &   1.7 &    3.7 &    3.7 &    107 &    107 &    107 &   2.1E-04 &  2.1E-04 &  2.1E-04 & I & IRDC \\

1428: N10-5 &  G013.1711-00.0055 &  351 &   3.4 &    0.3 &    7.1 &     35 &      9 &    766 &   0.0E+00 &  0.0E+00 &  3.8E-04 & II &  \\

1429: N10-6 &  G013.1725+00.0333 &  597 &    .1 &    0.7 &    7.5 &    152 &     12 &    600 &   2.2E-05 &  1.6E-06 &  6.0E-04 & I &  Bub \\

1430: N10-7 &  G013.1758+00.0604 &   57 &    .4 &    5.8 &   16.7 &   6479 &    273 &   8570 &   8.9E-04 &  2.2E-04 &  1.6E-03 & I & Rim, IRDC \\

1431: N10-8 &  G013.1814+00.0300 & 2395 &   3.6 &    0.9 &   11.1 &    246 &     24 &   3225 &   0.0E+00 &  0.0E+00 &  1.4E-03 & I & Bub \\

1432: N10-9 &  G013.1818+00.0610 &  197 &    .3 &    0.4 &   13.5 &    2820 &    142 &  13880 &   1.0E-03 &  3.7E-06 &  2.1E-03 & I &  Rim, IRDC, [4.5] \\

1433: N10-10 &  G013.1931+00.0885 &   84 &   5.1 &    1.9 &    6.2 &    440 &     62 &    440 &   2.2E-03 &  2.2E-05 &  2.2E-03 & I & PDR \\

1434: N10-11 &  G013.2124+00.0401 &   42 &   0.7 &    8.4 &   12.7 &   3858 &   2170 &   6900 &   9.3E-04 &  1.5E-04 &  1.0E-03 & I & Rim \\

1435: N10-12 &  G013.2127+00.0476 &  244 &   8.9 &    8.9 &   17.1 &   5505 &   3342 &  29590 &   2.2E-04 &  6.4E-05 &  5.8E-03 & I & Rim \\

1436: N10-13 &  G013.2142-00.0101 &   27 &   3.6 &    2.5 &    4.2 &    206 &     44 &    258 &   0.0E+00 &  0.0E+00 &  0.0E+00 & II &  \\

1437: N10-14 &  G013.2172+00.0582 & 2400 &   4.2 &    1.4 &    8.7 &    144 &     33 &   1349 &   0.0E+00 &  0.0E+00 &  4.6E-04 & I & IRDC \\

1438: N10-15 &  G013.2227-00.0093 &   13 &   0.5 &    0.2 &    0.5 &     16 &     15 &     20 &   2.9E-06 &  2.3E-06 &  3.9E-06 & I & IRDC \\

1439: \tableline

1440: N21-1 &  G018.0981-00.3692 & 1302 &   .8 &    0.4 &    4.6 &     27 &      6 &     87 &   1.3E-05 &  1.2E-06 &  2.3E-04 & I & IRDC \\

1441: N21-2 &  G018.1192-00.3359 & 2212 &   .3 &    3.2 &    6.9 &    314 &     78 &   1563 &   0.0E+00 &  0.0E+00 &  4.3E-05 & II & PDR \\

1442: N21-3 &  G018.1227-00.3524 &   49 &   .4 &    4.6 &    6.8 &   1151 &    383 &   1543 &   0.0E+00 &  0.0E+00 &  0.0E+00 & II & PDR \\

1443: N21-4 &  G018.1403-00.3721 & 1581 &   .5 &    1.5 &    8.6 &    102 &     30 &   1216 &   0.0E+00 &  0.0E+00 &  5.4E-04 & I & PDR \\

1444: N21-5 &  G018.1412-00.3763 & 1166 &   .9 &    2.6 &    5.8 &    163 &     76 &    719 &   0.0E+00 &  0.0E+00 &  1.4E-04 & II & PDR \\

1445: N21-6 &  G018.1611-00.3480 &  354 &   .1 &    2.6 &    4.2 &     80 &     42 &    141 &   0.0E+00 &  0.0E+00 &  1.3E-07 & II & PDR \\

1446: N21-7 &  G018.1662-00.4833 &  799 &   .6 &    1.6 &    4.0 &     44 &     16 &    197 &   0.0E+00 &  0.0E+00 &  1.1E-08 & II &  \\

1447: N21-8 &  G018.1677-00.3034 &   60 &   .5 &    1.2 &    8.0 &    474 &     35 &    903 &   5.0E-05 &  1.0E-05 &  2.9E-04 & I &  PDR \\

1448: N21-9 &  G018.1808-00.4686 & 3546 &   .8 &    0.5 &    4.7 &     58 &      8 &    140 &   0.0E+00 &  0.0E+00 &  2.3E-04 & I &  \\

1449: N21-10 &  G018.1992-00.3520 &  276 &  4.2 &    3.2 &    5.2 &    258 &    144 &    583 &   0.0E+00 &  0.0E+00 &  0.0E+00 & II &  PDR \\

1450: N21-11 &  G018.2044-00.4412 & 1015 &  2.9 &    0.9 &    5.1 &     59 &     26 &    143 &   0.0E+00 &  0.0E+00 &  3.4E-04 & I &  \\

1451: N21-12 &  G018.2125-00.4857 &   25 &  1.0 &    0.5 &    2.0 &     11 &      7 &     37 &   2.4E-05 &  6.8E-06 &  4.7E-04 & I &  \\

1452: N21-13 &  G018.2157-00.3419 &   33 &  8.6 &    1.2 &    8.6 &   1443 &     62 &   1735 &   1.3E-03 &  1.2E-05 &  1.3E-03 & I &  IRDC \\

1453: N21-14 &  G018.2171-00.3426 &   83 &  0.5 &    0.2 &    6.4 &     16 &     16 &    379 &   2.9E-06 &  2.3E-06 &  1.3E-03 & I &  IRDC \\

1454: N21-15 &  G018.2249-00.3352 &  371 &  6.1 &    1.1 &    6.1 &    358 &     19 &    358 &   6.4E-05 &  0.0E+00 &  5.8E-04 & I &  IRDC \\

1455: N21-16 &  G018.2262-00.3348 &   63 &  1.3 &    0.4 &    2.4 &     44 &     34 &    134 &   6.7E-06 &  4.3E-06 &  3.7E-05 & I &  IRDC \\

1456: N21-17 &  G018.2277-00.3303 & 3136 &  1.5 &    0.7 &    5.2 &     69 &     12 &    154 &   0.0E+00 &  0.0E+00 &  3.0E-04 & I &  IRDC \\

1457: N21-18 &  G018.2311-00.3150 & 5360 &  1.7 &    0.7 &    5.7 &     18 &     10 &    546 &   0.0E+00 &  0.0E+00 &  2.9E-04 & I &  PDR \\

1458: N21-19 &  G018.2351-00.3532 &   83 &  5.1 &    0.4 &    6.0 &    182 &     26 &    480 &   9.4E-04 &  0.0E+00 &  9.4E-04 & I &  IRDC \\

1459: N21-20 &  G018.2466-00.4717 &   20 &  2.9 &    1.2 &    4.1 &    139 &     56 &    235 &   1.3E-03 &  0.0E+00 &  1.3E-03 & I &  \\

1460: N21-21 &  G018.2470-00.4728 &   19 &  5.5 &    2.0 &    6.0 &    161 &     46 &    194 &   6.3E-04 &  1.5E-04 &  9.9E-04 & I &  \\

1461: \tableline

1462: N49-1 &  G028.8299-00.2532 &   47 &    9.0 &   14.0 &   29.0 &  30040 &   5251 & 102400 &   8.9E-04 &  4.1E-04 &  6.7E-03 & I &  Rim, IRDC \\

1463: N49-2 &  G028.8318-00.2808 & 1461 &   34.0 &    1.6 &    4.7 &     60 &     19 &    205 &   0.0E+00 &  0.0E+00 &  1.7E-05 & II &  \\

1464: N49-3 &  G028.8326-00.2531 &    4 &    0.5 &    9.3 &   12.7 &   2942 &   2288 &   3657 &   3.4E-04 &  1.2E-04 &  3.4E-04 & I &  Rim, IRDC,[4.5] \\

1465: N49-4 &  G028.8352-00.2354 & 1578 &   33.7 &    2.8 &    7.9 &    207 &    101 &   1050 &   0.0E+00 &  0.0E+00 &  7.0E-04 & I &  Bub \\

1466: N49-5 &  G028.8547-00.2192 &   72 &    1.3 &    0.3 &    8.1 &     44 &     34 &    924 &   6.7E-06 &  4.0E-06 &  1.0E-03 & I &  PDR \\

1467: N49-6 &  G028.8573-00.2184 &  276 &    0.5 &    0.2 &    6.9 &     16 &     15 &    379 &   2.9E-06 &  2.3E-06 &  1.2E-03 & I &  PDR \\

1468: N49-7 &  G028.8619-00.2072 &  812 &    2.3 &    0.7 &    3.9 &     18 &      7 &     35 &   1.2E-06 &  8.6E-08 &  9.3E-05 & I &  \\

1469: \enddata

1470: \tablenotetext{a}{The flags in the comments column are as follows:

1471:   Rim=source on the rim of the bubble; IRDC=source within an infrared

1472:   dark cloud; PDR=source within bright diffuse PAH background emission

1473:   in the photodissociation region; and [4.5]=source exhibits extended

1474:   excess emission at 4.5 $\mu$m.}

1475: \end{deluxetable}

1476: 

1477: \clearpage

1478: \begin{deluxetable}{lcccc}

1479:   \tablecaption{Candidate Ionizing Stars\label{ionizers}}

1480:   \tablewidth{0pt}

1481:   \tablehead{\colhead{ID} & \colhead{Name (G$l+b$)} &

1482:     \colhead{Spectral Type} & \colhead{$A_V$} & \colhead{Best?} \\

1483:     }

1484:     \startdata

1485:     IN10-1 & 13.1887+00.0421 & O7.5 V & 7 & \checkmark \\

1486:     IN10-2 & 13.1942+00.0521 & O6.5 V & 7 &  \\

1487:     IN10-3 & 13.1786+00.0331 & O6 V & 5 & \checkmark \\

1488:     IN10-4 & 13.1777+00.0346 & O7 V & 8 & \\

1489:     \tableline

1490:     IN21-1 & 18.1893-00.4041 & early B I\tablenotemark{a} & 6 & \checkmark \\

1491:     IN21-2 & 18.1742-00.3918 & O6 V & 9 & \\

1492:     IN21-3 & 18.1798-00.4275 & O8 V & 7.5 & \\

1493:     IN21-4 & 18.1977-00.3886 & O8.5 V & 13 & \\

1494:     IN21-5 & 19.1928-00.4147 & early B V& 8 & \\

1495:     \tableline

1496:     IN49-1 & 28.8263-00.2287 & O5 V & 10.5 & \checkmark \\

1497:     IN49-2 & 28.8142-00.2241 & O5.5 V & 7.5 & \\

1498:     IN49-3 & 28.8174-00.2464 & O7 V & 7.5 & \\

1499:     IN49-4 & 28.8119-00.2383 & O9 & 10 & \\

1500:     IN49-5 & 28.8098-00.2270 & B0 & 6 & \\

1501:     \enddata

1502:     \tablenotetext{a}{For this star, the observed $T_{\rm eff}$--$R$ relation

1503:       at the 3.7-kpc kinematic distance of N21 overlaps with the MSH05 curve

1504:       for O supergiants, so the spectral type is degenerate. We have assigned

1505:       a spectral type based upon the Lyman continuum photon flux required to

1506:       ionize the \ion{H}{2} region in N21.}

1507: \end{deluxetable}

1508: 

1509: 

1510: \end{document}

1511: 

1512: