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

2: %\usepackage{natbib}

3: %\usepackage{epsfig}

4: 

5: \begin{document}

6: 

7: \title{Clustered and Triggered Star Formation in W5: Observations with {\it Spitzer}}

8: 

9: \author{Xavier P. Koenig,\altaffilmark{1} Lori E. Allen,\altaffilmark{1} Robert A. Gutermuth,\altaffilmark{1} Joseph L. Hora,\altaffilmark{1} Christopher M. Brunt,\altaffilmark{2} James Muzerolle\altaffilmark{3}}

10: \altaffiltext{1}{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, 

11: Cambridge, MA 02138}

12: \altaffiltext{2}{The School of Physics, University of Exeter, The Queens Drive, Exeter, Devon, UK, EX4 4QL}

13: \altaffiltext{3}{Steward Observatory, 933 North Cherry Avenue, University of Arizona, Tucson, AZ 85721}

14: 

15: \begin{abstract}

16:   We present images and initial results from our extensive {\it

17:     Spitzer} Space Telescope imaging survey of the W5 H\,{\sc ii}

18:   region with the Infrared Array Camera (IRAC) and Multiband Imaging

19:   Photometer for {\it Spitzer} (MIPS). We detect dense clusters of

20:   stars, centered on the O stars: HD 18326, BD +60 586, HD 17505 and

21:   HD 17520. At 24 $\micron$, substantial extended emission is visible,

22:   presumably from heated dust grains that survive in the strongly

23:   ionizing environment of the H\,{\sc ii} region. With photometry of

24:   more than 18000 point sources, we analyze the clustering properties

25:   of objects classified as young stars by their IR spectral energy

26:   distributions (a total of 2064 sources) across the region using a

27:   minimal-spanning-tree algorithm. We find $\sim$40--70\% of infrared

28:   excess sources belong to clusters with $\geq$10 members. We find

29:   that within the evacuated cavities of the H\,{\sc ii} regions that

30:   make up W5, the ratio of Class II to Class I sources is $\sim$7

31:   times higher than for objects coincident with molecular gas as

32:   traced by $^{12}$CO emission and near-IR extinction maps. We

33:   attribute this contrast to an age difference between the two

34:   locations, and postulate that at least two distinct generations of

35:   star formation are visible across W5. Our preliminary analysis shows

36:   that triggering is a plausible mechanism to explain the multiple

37:   generations of star formation in W5, and merits further

38:   investigation.

39: \end{abstract}

40: 

41: \keywords{HII regions --- infrared: stars --- ISM: globules --- stars: early-type --- stars: formation --- stars: pre-main sequence}

42: 

43: \section{Introduction}

44: Star formation is a self-regulating process---once massive stars form

45: they immediately begin to disrupt their natal environment with their

46: stellar winds and the emission of ionizing radiation. Eventually their

47: parental molecular clouds are destroyed, halting further star

48: formation. However, it has also been argued that the energy input by

49: these massive stars can promote and induce subsequent star formation

50: in the surrounding molecular gas before it disperses, above that which

51: would be produced without external forcing. This process is given the

52: name `triggering' \citep[see][for reviews of this

53: subject]{elmegreen98,zinnecker07}. It is vital to understand the

54: balance of cloud destruction and triggered star formation if we are to

55: develop a theory that explains the morphology and evolution of star

56: forming regions, star formation efficiencies and the initial mass

57: function of stars in clusters. It also has relevance for star

58: formation on galactic scales in understanding how star formation

59: progresses with time, and indeed how spiral structure of galaxies

60: evolves with time \citep{seiden90,jungwiert94}. Feedback in star

61: formation can be clearly observed in bright-rimmed clouds, where an

62: edge-on molecular cloud is externally illuminated by nearby young

63: massive stars, creating a cross-section of the photo-evaporation

64: process as the ionization fronts they produce advance into the

65: molecular cloud.

66: 

67: Two triggering mechanisms are of interest in regard to W5. The first

68: is the creation of an ionized H\,{\sc ii} region bubble by an initial

69: generation of massive stars within their parental molecular cloud, and

70: its subsequent expansion. In this `collect and collapse' mechanism,

71: investigated analytically by \citet{whitworth94} and numerically by

72: \citet{dale07}, the expansion creates a shock front that sweeps up

73: neutral material ahead of it as it progresses outward. The gas

74: accumulated eventually exceeds a critical threshold for collapse and

75: gives rise to a second generation of star formation. Secondly,

76: inhomogeneity in the ISM often leads to small clumps of material

77: remaining exposed inside an H\,{\sc ii} region. In this environment,

78: the high pressure of the surrounding ionized gas has been suggested as

79: a mechanism to compress the clumps and form stars \citep{stutzki88}.

80: 

81: The star forming region W5 is a part of the chain of molecular clouds

82: W3/4/5 \citep{westerhout58}. Its distance has not been definitively

83: established---\citet{becker71} found a photometric distance of 2.2

84: kpc, but more recently, \citet{hillwig06} found that a distance of 1.9

85: kpc gave the most consistent results of evolutionary model fits to

86: stellar radii in W5. \citet{xu06} found a distance to the neighboring

87: region W3OH of 1.95$\pm$0.04 kpc using maser parallaxes---we adopt a

88: distance to W5 of 2 kpc as a conservative intermediate between these

89: different estimates. W5 is a relatively isolated star forming region,

90: with an apparently simple morphology. Optical imaging shows it is made

91: up of two roughly circular adjoining H\,{\sc ii} regions W5 East and

92: W5 West \citep{karr03}, containing one and four O stars

93: respectively. At least two of the four O stars in W5 West are multiple

94: systems \citep{hillwig06}. Both $^{12}$CO and 21 cm radio emission

95: \citep[see][]{normandeau96} demonstrate the same overall

96: shape---$^{12}$CO emission traces the molecular hydrogen gas in W5,

97: which appears as two broken rings of emission. Figure 1 shows a map of

98: $^{12}$CO ($\lambda$ = 2.6 mm) emission integrated over the range

99: $-$28 to $-$47 km s$^{-1}$, made using observations with the 14m FCRAO

100: telescope. The bulk of the emission is found between a velocity of

101: $-$31 and $-$47 km s$^{-1}$, and thus is likely roughly in the same

102: plane. In this image, following the naming scheme of

103: \citet{wilking84}, to the northwest of HD 17505 is the molecular cloud

104: W5NW, in between HD 18326 and BD +60 586 is W5NE, and to the east of

105: HD 18326 is W5A. Due to this relatively simple morphology, W5 presents

106: a useful test case for investigating models of triggered star

107: formation and the influence of massive star formation on its

108: surroundings.

109: 

110: \begin{figure}

111: \begin{center}

112: \includegraphics[width=4.5in]{f1.eps}

113: \caption{$^{12}$CO (1-0) greyscale image of W5 from FCRAO telescope

114:   (integrated T$_A^*$ corrected for main beam efficiency

115:   $\eta_{MB}=0.45$, ranging from 0--110 K km s$^{-1}$), oriented

116:   North-up and East-left. We label molecular clouds in black following

117:   the naming scheme of \citet{wilking84}. We label the two H\,{\sc ii}

118:   regions in white. Prominent O stars are marked with white

119:   asterisks: HD 18326 in W5 East, and in W5 West from North-East to

120:   South-West are BD +60 586, HD 237019 and HD 17505 (upper of pair), HD

121:   17520 (lower of pair).}

122: \end{center}

123: \end{figure}

124: 

125: In this paper we present initial results from our mid-IR {\it Spitzer}

126: survey of W5 with the IRAC and MIPS instruments. In $\S$ 2 we describe

127: our observations and data reduction techniques and our classification

128: and clustering analysis in $\S$ 3. In $\S$ 4 we perform a trial,

129: simplified investigation of triggered star formation models as a means

130: to explain the observed distribution of young stars in different

131: evolutionary states across the region. In $\S$ 5 we present our

132: conclusions and directions of future work.

133: 

134: \section{Observations and Data Reduction}

135: W5 was observed with the {\it Spitzer} IRAC instrument \citep{fazio04}

136: in all four bands (3.6, 4.5, 5.8 and 8.0 $\micron$). The observations

137: were broken down into three rectangular Astronomical Observing

138: Requests (AORs) covering $\sim$1.8$\times$1.6 degrees, in order to

139: observe at multiple rotation angles and help minimize artifacts

140: aligned along columns or rows of the array. In Table 1 we list the

141: dates and coordinates of each AOR. Each AOR had a coverage of 1 High

142: Dynamic Range (HDR) frame. HDR mode results in a 10.4 second and 0.4

143: second exposure being taken at each position in each map. We used

144: software tools ({\it clustergrinder}) developed by one of us (RAG) to

145: produce final image mosaics from these data in each wavelength

146: band. {\it Clustergrinder} incorporates all necessary image treatment

147: steps, for example, saturated pixel processing and distortion

148: corrections \citep[see][for a more complete description of the

149: processing performed]{gutermuth08}. {\it Clustergrinder} uses the

150: short 0.4 sec exposures only in saturated or near-saturated regions,

151: so that the combined map has an effective total integration time of

152: 3$\times$10.4 = 31.2 sec in most of the overlapping areas. We also

153: incorporated in our data processing archival data covering AFGL 4029

154: \citep[from {\it Spitzer} GTO program PID 201][]{allen05}, at the

155: eastern end of W5.

156: 

157: \input{tab1.tex}

158: 

159: The MIPS \citep{rieke04} observations were carried out on 2006

160: February 23 UT under our GO-2 program, PID 20300. Images were taken in

161: scan map mode using the medium scan speed for an average exposure time

162: of 41.9s per pixel once frames were combined. The raw data were

163: processed with pipeline version S13.2.0. We produced final mosaics

164: using the MIPS instrument team Data Analysis Tool, which calibrates

165: the data and applies a distortion correction to each individual

166: exposure before combining \citep{gordon05}. We used only the 24

167: $\micron$ band data for our analysis in this paper, since strong

168: background emission dominates at the longer wavelength (70 and 160

169: $\micron$) bands of MIPS, and lower sensitivity reduces the number of

170: detectable objects to an un-useful level for the present study.

171: 

172: In Figure 2 we present a 3-color IRAC image of W5, using the color

173: scheme: blue=3.6 $\micron$, green=4.5 $\micron$ and red=8.0

174: $\micron$. In Figure 3 we show our MIPS 24 $\micron$ mosaic. Figure 4

175: presents a composite image incorporating MIPS, with blue=4.5

176: $\micron$, green=5.8 $\micron$ and red=24 $\micron$ (MIPS).

177: 

178: \begin{figure}

179: \begin{center}

180: \includegraphics[angle=90,width=6.1in]{f2.eps}

181: \caption{W5 {\it Spitzer} IRAC bands 1, 2 and 4 color

182:   composite. Bright emission from PAH grains at 8 $\micron$ traces

183:   the boundary of the H\,{\sc ii} region giving the bright pinkish

184:   color. Dense clusters of stars surround the O stars in the interior,

185:   with smaller clusters appearing amidst the diffuse PAH emission.}

186: \end{center}

187: \end{figure}

188: 

189: \begin{figure}

190: \begin{center}

191: \includegraphics[angle=90,width=6.1in]{f3.eps}

192: \caption{{\it Spitzer} MIPS 24 $\micron$ mosaic. The O stars labelled

193:   in Fig. 1 are marked here with white diamonds. Bright dust continuum

194:   emission traces the boundary of the H\,{\sc ii} region, while

195:   extended emission from heated dust grains is also present in the

196:   interior (see $\S$ 3.2). Bright point sources are protostars.}

197: \end{center}

198: \end{figure}

199: 

200: \begin{figure}

201: \begin{center}

202: \includegraphics[angle=90,width=6.1in]{f4.eps}

203: \caption{Composite of emission in IRAC bands 2 and 4, plus MIPS 24

204:   $\micron$. Emission at 24 $\micron$---presumably from heated

205:   dust---fills the interior of the H\,{\sc ii} region cavities, see

206:   $\S$ 3.2.}

207: \end{center}

208: \end{figure}

209: 

210: We carried out point source extraction and aperture photometry of all

211: point sources on the final IRAC mosaics with PhotVis version

212: 1.10beta3. PhotVis is an IDL GUI-based photometry visualization tool

213: \citep[see][]{gutermuth04} that utilizes DAOPHOT modules ported to IDL

214: in the IDL Astronomy User's Library \citep{landsman93}. We used

215: PhotVis to visually inspect the detected sources in IRAC bands 1, 3

216: and 4, adding sources not detected automatically, but clearly visible

217: in the images with the GUI tool and rejected any structured nebulosity

218: or cosmic rays mistaken for stellar sources by the automatic detection

219: algorithm. To save time, we did not visually check the band 2

220: photometry in this manner. Instead, we took the cleaned band 1 source

221: list as the start point for finding objects in the image and extracted

222: photometry at each position. Radii of the apertures and inner and

223: outer limits of the sky annuli were 2.4$''$, 2.4$''$ and 7.2$''$

224: respectively. The resultant photometry was calibrated using

225: large-aperture in-flight measurements of standard stars, with an

226: appropriate aperture correction in each channel to correct for the

227: smaller apertures used in this study. In this paper we restrict our

228: source catalog to only those objects with magnitude error $\leq$ 0.2

229: in all four IRAC filters, a total of 18518 objects.

230: 

231: We estimate the completeness in each IRAC filter by breaking up each

232: image into a 100 by 100 pixel grid, adding artificial stars to each

233: grid cell and counting the number retrieved as a function of

234: magnitude. Averaged over the whole W5 field, our source catalog is

235: 90\% complete to a magnitude of 15.5 at 3.6 $\micron$, 15.5 at 4.5

236: $\micron$, 14.0 at 5.8 $\micron$ and 12.7 at 8 $\micron$. The

237: completeness is less in regions of bright diffuse emission: $\sim$14

238: at 3.6 and 4.5 $\micron$, $\sim$11 at 5.8 $\micron$ and $\sim$9.5 at

239: 8.0 $\micron$.

240: 

241: We conducted point source extraction and aperture photometry of point

242: sources in the 24 $\micron$ MIPS mosaic using the

243: point-spread-function fitting capability in IRAF DAOPHOT

244: \citep{stetson87}. We visually inspected the image to pick out point

245: sources not automatically detected due to bright diffuse emission

246: evident throughout the image. We match the four-band IRAC source list

247: to the MIPS catalog using a 2$''$ search radius, selecting the object

248: closest to the MIPS point spread function centroid in cases where more

249: than one IRAC object is a match. Of the 1874 MIPS objects that match

250: with entries in our IRAC 4-band list, 6 had a second object within the

251: 2$''$ search radius which possibly contributes to its 24 $\micron$

252: flux. We consider this a small effect on our analysis.

253: 

254: \section{Analysis}

255: \subsection{Source Classification}

256: In order to characterize the progress of star formation throughout W5

257: it is important to establish the evolutionary status of stars within

258: the region. The presence of massive O stars \citep{hillwig06} and

259: significant molecular material in W5 \citep{lada78}, suggest that the

260: stars associated with the region will be of relatively young age ($<$

261: 10$^7$yrs). Recently formed stars exhibit infrared excess emission in

262: their spectra above that produced by the stellar photosphere. This

263: excess emission arises from heated dust, either within the

264: circumstellar material close to the young star, perhaps falling onto

265: it as a part of protostellar collapse, or left behind after the end of

266: star formation. This material gradually disappears with time, and

267: hence the excess emission evolves as a function of stellar age. We

268: make use of this by measuring $\alpha _{IR}$ \citep[see,

269: e.g.][]{lada87,stahler05}:

270: 

271: \begin{equation}

272: \alpha _{IR} = \frac{d\log \left( \lambda F_{\lambda}\right)}{d\log \lambda}

273: \end{equation}

274: 

275: the value of which decreases with the progression of the star's

276: evolutionary state, whether through its own aging, or the effects of

277: its environment.

278: 

279: In the case of most surveys of star forming regions, including W5,

280: infrared colors serve as proxies for directly measuring $\alpha

281: _{IR}$. Schemes developed in \citet{whitney03}, \citet{allen04},

282: \citet{megeath04} and \citet{muzerolle04} and tested by (for example)

283: \citet{hartmann05} have demonstrated the power of this technique for

284: classifying young stellar objects using {\it Spitzer} IRAC and MIPS

285: photometry. The classification scheme we use is described fully in

286: \citet{gutermuth08}. We use an iterative method combining IRAC and

287: MIPS photometry with an extinction map generated from 2MASS near-IR

288: photometry \citep{skrutskie06}. The scheme classifies the stars and

289: filters out extra-galactic contamination.

290: 

291: We generate an extinction map from 2MASS $H - K_S$ colors of stars

292: detected across the entire W5 field using the method described in

293: \citet{gutermuth05} which itself is based on the NICE and NICER

294: algorithms of \citet{lada94} and \citet{lombardi01}. The map has an

295: angular resolution of $\sim$35 arcsec, and is sensitive up to $A_V

296: \sim$15. However, the map is limited by the sensitivity of the 2MASS

297: survey. As a result, $A_V$ values in the map above $\sim$3 may be

298: underestimates of the true extinguishing column through the entire

299: cloud, since objects behind are too faint to be detected. This is only

300: the case for $<$1\% of pixels in the map, so we ignore this effect for

301: our analysis. With $A_V$ values in hand for all sources, we deredden

302: their IRAC magnitudes using the IR extinction law presented in

303: \citet{flaherty07}, assuming each source is seen behind the full

304: extinguishing column in each case.

305: 

306: We first remove star forming (`PAH') galaxies and weak line AGN via a

307: series of cuts in the four band IRAC color-color diagrams after the

308: procedure developed by \citet{gutermuth08}. Because W5 is

309: $\sim$4$\times$ more distant than the nearby regions studied by

310: Gutermuth et al., this filter removes many apparent young stellar

311: objects (YSOs). In the Appendix we discuss how we characterize and

312: account for this effect in our subsequent analysis. Next we filter out

313: unresolved shocked blobs of PAH emission by cutting objects with a

314: large 4.5 $\micron$ excess, i.e. very red $[3.6]-[4.5]$ color

315: \citep[][]{smith06}. We categorize the remaining---presumably

316: stellar---sample primarily relying largely on the $[4.5]-[5.8]$ color

317: to discriminate among SED classes. In Figure 5 we show IRAC

318: color-color diagrams for sources in W5, marking the location of

319: protostars (Class I/0, red dots), stars with disks (Class II, green

320: dots) and stars exhibiting only photospheric colors (Class III, black

321: dots) as identified using our scheme.

322: 

323: We also use 2MASS $H$ and $K_S$ photometry, combined with IRAC 3.6 and

324: 4.5 $\micron$ data to classify objects lacking either an IRAC $[5.8]$

325: or $[8.0]$ detection. To make sure the 2MASS sources have reliable

326: photometry we require a magnitude error $\leqslant$0.1 in both $H$ and

327: $K_S$ bands. We first deredden the photometry in these four bands, and

328: then identify IR excess sources as those having red $[3.6]-[4.5]$ and

329: $K_S-[3.6]$ colors. The results are shown in Figure 6, left panel.

330: 

331: MIPS 24 $\micron$ photometry provides us with additional

332: classification information. Figure 6 (right panel) shows the

333: color-color diagram combining MIPS and IRAC photometry. Stellar

334: sources not classified as either Class I or Class II by their IRAC or

335: near-IR colors may still have red colors ($[5.8]-[24]$ $> $1.5) and

336: thus be candidate ``cold disk'' or ``transition disk'' objects---in

337: other words: Class II objects with significant clearing of their inner

338: disk region \citep[see for

339: example,][]{muzerolle04,lada06,cieza07,najita07,flaherty08,brown08}. These

340: are marked with yellow dots in Figures 5 and 6. MIPS photometry is

341: also used as a check on the AGN/galaxy/shocked blob filtering, and

342: Class I classification. AGN candidates with bright MIPS detection can

343: be re-classified as protostars, and Class I objects with

344: insufficiently red $[5.8]-[24]$ or $[4.5]-[24]$ colors are `demoted'

345: to Class II.

346: 

347: \begin{figure}

348: \begin{center}

349: \includegraphics*[width=3.0in]{f5a.eps}

350: \includegraphics*[width=3.0in]{f5b.eps}

351: \caption{Left panel: $[3.6]-[4.5]$ vs. $[4.5]-[5.8]$; right panel:

352:   $[4.5]-[5.8]$ vs. $[5.8]-[8.0]$ IRAC color-color diagrams used for

353:   identifying candidate protostars. Black dots---Class III,

354:   green---Class II, red---Class I, yellow---transition disk

355:   candidates. AGN and PAH galaxy candidates are not included here.}

356: \end{center}

357: \end{figure}

358: 

359: \begin{figure}

360: \begin{center}

361: \includegraphics*[width=3in]{f6a.eps}

362: \includegraphics*[width=3in]{f6b.eps}

363: \caption{Left panel: dereddened 2MASS $K_S$ vs. $K_S-[4.5]$

364:   color-magnitude diagram. Dashed red box shows location of AGN

365:   detected by IRAC Shallow survey \citep{eisenhardt04}. Right panel:

366:   IRAC and MIPS $[3.6]-[5.8]$ vs. $[8.0]-[24]$ color-color

367:   diagram. Color scheme: black dots---Class III, green---Class II,

368:   red---Class I, yellow---transition disk candidates. Suspected AGN

369:   and PAH galaxies are not included here.}

370: \end{center}

371: \end{figure}

372: 

373: Lastly, we also visually inspected the SEDs of all objects in Class I

374: and Class II (a total of 2064 objects), where available making use of

375: 2MASS $JHK_S$ and/or 24 $\micron$ photometry. In most cases only IRAC

376: 4-band data were present. We give the full list of objects: Class I,

377: Class II, Class III and transition disk candidates in Table 2. We list

378: source coordinates and photometry in near-IR $JHK_S$ (from 2MASS) and

379: {\it Spitzer} bands, and give a column denoting infrared SED source

380: class and a `flag' column for objects with unusual SEDs. Objects are

381: sorted and indexed by ascending right ascension order. Note that the

382: full machine-readable table is available in the online edition of this

383: paper. We note Class I sources with bright emission at 24 $\micron$

384: and very red IRAC$-$24 $\micron$ color ($[X]-[24] > 4.5$, where $X$ is

385: the magnitude in any of the four IRAC bands) as a subclass: `deeply

386: embedded protostars' in the table. Table 3 gives a summary of the

387: classification results.

388: 

389: \input{stub.tab2.tex}

390: 

391: \input{tab3.tex}

392: 

393: In Figure 7 we show the spatial distribution of young stars (Class I

394: and deeply embedded protostars, and Class II and transition disk

395: candidates) in W5 overlaid on the IRAC 4.5 $\micron$ image. Class I

396: and Class II sources are marked in red and green respectively,

397: transition disk candidates and embedded protostars with yellow and

398: blue dots respectively. O stars are labelled and marked with white

399: asterisks.

400: 

401: \begin{figure}

402: \begin{center}

403: \includegraphics[angle=90,width=6.1in]{f7.eps}

404: \caption{Source distribution overlaid on the MIPS 24$\micron$

405:   greyscale image. Color scheme: red---Class I, green---Class II,

406:   yellow dots---transition disk candidates. O stars marked with

407:   asterisks.}

408: \end{center}

409: \end{figure}

410: 

411: \subsection{Spatial Distribution of Young Stars in W5}

412: Figure 7 clearly shows that neither Class I nor Class II type objects

413: in W5 are distributed uniformly, but rather (for the most part) in

414: clustered, or filamentary formations. Within the larger cavity of W5W,

415: clusters of Class II YSOs can be seen centered on the O stars HD

416: 17505, HD 17520 and BD +60 586 of the open cluster OCl 364

417: \citep{alter70}. Similarly in the smaller W5E bubble, a cluster of

418: young stars is visible centered on the O star HD 18326. Smaller

419: clusters of Class I and Class II objects are also visible in the large

420: region of bright diffuse emission to the North-West of HD17505. Based

421: on the higher proportion of Class I stars relative to Class II's

422: (quantified in $\S$3.4), these are probably younger clusters, just

423: emerging from their parental molecular cloud.

424: 

425: Extending out from the central clusters of Class II objects are more

426: filamentary groupings of stars. In W5 West, a chain of young stars

427: appears to extend from the cometary bright rimmed clouds at the

428: Southern rim of the bubble \citep[object 11 in the survey

429: of][]{sugitani91} up to the clusters around HD 17505 and HD 17520, and

430: through the cluster around BD +60 586 to the bright rimmed clouds on

431: the Eastern rim of the bubble. In W5 East, a chain of young stars

432: extends East-West from each of the two peaks of AFGL4029

433: \citep{price83} through the cluster around HD 18326 to the bright PAH

434: ridge that marks the Western edge of the bubble.

435: 

436: Numerous protostars (Class I objects) are seen in projection on the

437: cloud rim at the PAH emission/H\,{\sc ii} region cavity

438: boundaries. For example, the bright rimmed clouds that make up the

439: eastern border of W5 East, the bright rimmed clouds at the

440: North-Eastern rim of W5 West, and the cometary clouds at the southern

441: rim of W5 West all contain several Class I objects. Several isolated

442: cloud remnants and small `elephant trunk' formations within the

443: H\,{\sc ii} region cavity also contain Class I or Class II

444: objects---these are listed in Table 4. We give a `type' to each object

445: to distinguish the different morphologies. These objects are

446: interesting candidates for triggered star formation on small scales,

447: see the discussion in $\S$ 4.3. Figure 8 shows several examples of

448: these objects.

449: 

450: \input{tab4.tex}

451: 

452: \begin{figure}

453: \begin{center}

454: \includegraphics*[width=5.5in]{f8a.eps}

455: \includegraphics*[width=5.5in]{f8b.eps}

456: \end{center}

457: \end{figure}

458: 

459: \begin{figure}[ht]

460: \begin{center}

461: \includegraphics*[width=3in]{f8c.eps}

462: \caption{Examples of star forming globules in W5 shown at 8

463:   $\micron$. PAH emission is seen at or near the globule

464:   heads. Objects are labelled as in Table 4.}

465: \end{center}

466: \end{figure}

467: 

468: The influence of star formation on its environment is clear to see in

469: both the IRAC (Figure 2) and MIPS images (Figures 3 and 4). Bright

470: diffuse emission in the four IRAC filters \citep{draine07} is only

471: present around the edge of the two main H\,{\sc ii} region bubbles in

472: roughly ring-like structures, and brightly lit-up over the molecular

473: clouds W5NE and W5NW. Presumably the PAHs responsible for much of the

474: emission in IRAC bands 1, 3 and 4 are destroyed in the harsh ionizing

475: environment of the H\,{\sc ii} region. At 24 $\micron$ the morphology

476: is similar in the bubbles' rims, but very different diffuse emission

477: is seen in the interior surrounding the 5 O stars (as labelled in

478: Figure 3). Seen in numerous massive star forming regions by

479: \citet{church06}, \citet{smith07} and recently \citet{harvey08}, this

480: additional component of diffuse emission is consistent with the

481: picture of \citet{crete99} whereby emission within H\,{\sc ii} regions

482: at mid-IR wavelengths is produced by heated larger dust grains (radius

483: $>$ 12 $\textrm{\AA}$) that survive being destroyed by ionizing

484: radiation. As \citet{draine07} have shown, emission in IRAC bands 1

485: and 4 is much more sensitive to the PAH mass fraction than is emission

486: in the MIPS 24$\micron$ band. Further, 24 $\micron$ band emission

487: increases rapidly when the interstellar radiation field (ISRF) rises

488: to 10$^4$--10$^5$ times the local value. Thus a region with low PAH

489: mass fraction but high ISRF can have enhanced 24 $\micron$ emission,

490: but will have low IRAC 3.6 and 8.0 $\micron$ PAH emission. The

491: morphology of the 24 $\micron$ emission suggests that these larger

492: grains in the {\it immediate} vicinity of the O stars are blown away

493: by the strong stellar winds.

494: 

495: \subsection{Spatial Variation in SED types}

496: Class I and II objects are present in varying amounts across

497: W5. Although with our present dataset we cannot tell if any one Class

498: I object is younger than a given Class II object, these classes do

499: represent different evolutionary stages in the formation of stars. In

500: W5, we find that in regions where our $^{12}$CO emission map exceeds

501: T$_A^*$ = 7.5 K km s$^{-1}$, there are on average 0.23 Class I objects

502: for every Class II. In regions with T$_A^* <$ 7.5 K km s$^{-1}$, the

503: ratio is 0.04. This suggests that the stellar clusters associated with

504: significant molecular material in W5 are systematically younger than

505: those in cleared out regions, for example the H\,{\sc ii} region

506: cavities.

507: 

508: The varying levels of completeness in each IRAC map as a function of

509: association with the bright, diffuse emission affect our result. As

510: described in $\S$ 2, the 90\% limiting magnitude decreases by

511: $\sim$1.5 mag in IRAC channels 1 and 2, by $\sim$2.5 mag in channel 3

512: and by $\sim$3.2 mag in channel 4 in going from the evacuated H\,{\sc

513:   ii} region to the brightest regions of diffuse emission. Recently,

514: \citet{chavarria08} studied the effects of reduced sensitivity in the

515: 5.8 and 8 $\micron$ {\it Spitzer} bands, combined with the obscuring

516: effect of bright, diffuse PAH emission on the detection of Class I and

517: Class II sources in IRAC surveys. They found that Class II sources are

518: on average fainter than Class I's, and are likely preferentially

519: missed in bright PAH emission regions as a result.

520: 

521: \subsection{Clustered versus Distributed Star Formation in W5}

522: We want to describe in an objective way how the stars in W5 are

523: grouped and address the following question: is there a population of

524: stars that do not belong to identifiable groups that may have formed

525: in isolation, in other words a `distributed' population? Lacking a

526: dataset that describes the masses and dynamics of the stars and gas

527: which would allow us to assign stars to clusters based on

528: gravitational association, we base cluster membership {\it only} on

529: spatial arrangement on the sky and so we do not require that these

530: clusters be gravitationally bound. We restrict ourselves to infrared

531: excess sources only to simplify the issue of whether stars are

532: associated with W5. We include the deeply embedded sources and

533: transition disk candidates along with the Class I and Class II objects

534: and exclude stars exhibiting only photospheric emission (Class III

535: sources), AGN and PAH galaxies for a total of 2064 objects. The source

536: distribution is shown in Figure 7, and reproduced in Figure 9, without

537: the background image.

538: 

539: \begin{figure}

540: \begin{center}

541: \includegraphics[angle=90,width=5.4in]{f9.eps}

542: \caption{Distribution of stellar sources as in Figure 7. Green points:

543:   Class II, red: Class I, yellow: transition disk candidates. Black

544:   asterisks mark the location of the known O stars.}

545: \end{center}

546: \end{figure}

547: 

548: Determining a criterion for cluster membership from spatial

549: distributions alone is somewhat arbitrary. In this paper we use the

550: so-called `minimal spanning tree' (MST) method to identify and

551: characterize clustering in W5 \citep[see recent work

552: by][]{cartwright04,bastian07}. \citet{gower69} describe how such a

553: tree is constructed from a list of source positions. To construct the

554: tree, we first generate the network of lines that joins together the

555: positions of all objects in our input list, such that the total length

556: of all lines joining points is minimized and there are no closed

557: loops. Each object is assigned one `branch length,' that is: the

558: projected distance to its nearest neighbor. We plot the distribution

559: of branch lengths in Figure 10.

560: 

561: \begin{figure}[ht]

562: \begin{center}

563: \includegraphics[width=4.0in]{f10.eps}

564: \caption{Minimal spanning tree length distribution, showing each

565:   object's branch length (y-axis) versus its number in the sorted

566:   list. Straight line fits (dashed lines) to the short and long length

567:   parts of the distribution intersect at 0.86 pc.}

568: \end{center}

569: \end{figure}

570: 

571: To distinguish `clustered' objects from `distributed' using the MST

572: requires 2 things: a break/cutoff length, $d_s$ and a minimum group

573: size $N$. A group is then a collection of objects linked by branches

574: shorter than $d_s$ with at least $N$ stars (we choose $N$ = 10). We

575: can estimate a value for $d_s$ from the distribution of branch lengths

576: in Figure 10. Following Gutermuth et al. (2008, in prep) we fit

577: straight lines through the long and short-length portions of the

578: distribution. Where these lines cross defines the break length for

579: clustering. For the IR excess sources in W5 this method yields $d_s$ =

580: 0.86 pc. Using this break length we find 16 groups containing at least

581: 10 stars and a clustered fraction of 70\%.

582: 

583: A single value of $d_s$ chosen in this way may not be representative

584: of the varying spatial density of stars across the region, and may

585: underestimate the amount of substructure within groups. Altering $d_s$

586: changes the number of groups and the relative fraction of clustered

587: and distributed sources. As the cutoff length is increased, an

588: increasing fraction of objects qualify as `clustered.' The number of

589: groups ($N_{grp}$) we find also increases up to a maximum before

590: falling off as groups start to merge until all groups belong to a

591: single `cluster.' These trends are shown in Figure 11. An alternate

592: value for the break length $d_s$ can be derived from the peak in the

593: plot of $N_{grp}$ versus increasing $d_s$. As argued by

594: \citet{batt91}, this method can be thought of as returning a maximum

595: of information from our source distribution. In this case we find a

596: cutoff of $d_s$ = 0.54 pc, $N_{grp}$ = 27, and a clustered fraction of

597: 44.2\%. In Figure 12 we compare the groupings produced by the two

598: methods side by side to demonstrate this.

599: 

600: Both values of $d_s$ used here are large when compared to the typical

601: separations of stars in nearby regions such as Taurus

602: \citep{hartmann05}, where a cutoff length of 0.54 pc would incorporate

603: many objects not associated with any cluster. However, Taurus is a

604: low-mass star forming region and so the characteristic scales of

605: clustering are likely to be smaller there. In recent surveys of low

606: mass star forming regions \citep{enoch07} and the high mass star

607: forming regions in Cygnus \citep{motte07}, typical star-forming dense

608: cores are $\sim$0.1 pc in size. However, \citep{motte07} also found

609: that massive dense {\it clumps} can be much larger: 0.5--0.8 pc, which

610: may be a more relevant scale for cluster and massive star formation

611: than for low density, more isolated star formation seen in Taurus.

612: 

613: \begin{figure}

614: \begin{center}

615: \includegraphics*[width=4.5in]{f11.eps}

616: \caption{Minimal spanning tree results. Solid line and left y-scale:

617:   clustered fraction of stars (belonging to all groups $\geq$10 stars)

618:   as a function of branch length cutoff. Dashed line and right

619:   y-scale: number of groups ($N_{grp}$) containing 10 or more stars

620:   identified by the MST algorithm. NOTE: coarse steps in x-axis miss

621:   finer detail in $N_{grp}$; true peak is 27 groups at $d_s$ = 0.54

622:   pc.}

623: \end{center}

624: \end{figure}

625: 

626: \begin{figure}

627: \begin{center}

628: \includegraphics*[width=5.0in]{f12a.eps}

629: \includegraphics*[width=5.0in]{f12b.eps}

630: \caption{Upper panel: the MST groups as found using the straight-line

631:   fit technique: $d_s$ = 0.86 pc, $N_{grp}$ = 16. Stars are colored

632:   according to their group for visual identification only. Lower

633:   panel: MST groups picked out at the maximum in $N_{grp}$ vs. $d_s$

634:   plot, using $d_s$ = 0.54 pc, $N_{grp}$ = 27. In both plots, black

635:   dots represent stars not associated with any cluster. Black

636:   asterisks mark the location of the known O stars.}

637: \end{center}

638: \end{figure}

639: 

640: Table 5 summarizes the clustering results from the two methods. The

641: value for the fraction of stars in clusters from both these methods is

642: lower than that reported by \citet{allen07} in {\it Spitzer} surveys

643: of Orion and Ophiuchus (74--78\%). \citet{carpenter00}, in his

644: analysis of the Orion A and B, Mon R2 and Perseus molecular clouds

645: with 2MASS found a range of clustering fractions, different for every

646: cloud, between 56--100\%. This fraction could be as much as a factor

647: two lower depending on how the age of the distributed population is

648: modeled. In W5---at a distance of 2 kpc---we are sampling much less of

649: the IMF than in these nearby regions, which lie within 1 kpc of the

650: Sun. We also do not know the disk-fraction of YSOs in the W5

651: clusters. \citet{cieza07} have shown that a significant fraction of

652: stars lose detectable circumstellar disks (in the mid-IR) on a

653: timescale $\sim$1 Myr. These objects would not be included in our

654: sample of IR-excess sources.

655: 

656: The mass completeness limit for our survey is dictated by our

657: photometric completeness and the wide span of YSO mid-IR colors. The

658: primary constraint in completeness for the IRAC 4-band sample is the 8

659: $\micron$ detection limit, since this has low sensitivity and very

660: bright background contamination from PAH feature emission. A typical

661: photosphere with $[3.6]-[8.0]\sim$0.1 at age 2 Myr, and $[8.0]$ =

662: 12.7, equates to a limiting mass $\approx$2 M$_{\sun}$ (using the

663: evolutionary models of D'Antona \& Mazzitelli 1994, and adopting $A_V$

664: = 2, a typical value for the H\,{\sc ii} region cavity: Hillwig

665: et~al. 2006). On the bright diffuse emission the 8 $\micron$ detection

666: limit for a photosphere equates to $\sim$8 M$_{\sun}$. The additional

667: 2MASS+IRAC bands 1 \& 2 sample is limited by the $K_S$ band

668: completeness. We estimate that the 90\% completeness in this band is

669: 14.2. Using the models of \citet{baraffe98}, this equates to a star of

670: 1 M$_{\sun}$ at 2--2.5 Myr, at an $A_V$ of 2 and a distance of 2

671: kpc. Both the 8 $\micron$ and $K_S$ band mass completeness limits are

672: conservative estimates, since YSOs with infrared excess emission will

673: be detected more readily. For example, the $[3.6]-[8.0]$ color peaks

674: at $\sim$1.4 for IR excess YSOs in W5. In this case, $[8.0]$ = 12.7

675: converts to a mass of $\approx$0.8 M$_{\sun}$ at 2 Myr in the

676: cavity. In summary, we estimate that for a young population of 2 Myr

677: age, we are complete to $\sim$8 M$_{\sun}$ for photospheres and

678: $\sim$4 M$_{\sun}$ for typical disk excess objects in the IRAC 4-band

679: sample seen against the bright background. In the cavity we are

680: complete to $\sim$2 M$_{\sun}$ for photospheres and $\sim$0.8

681: M$_{\sun}$ for disk excess objects. In the additional 2MASS sample we

682: are complete to a mass $\sim$1 M$_{\sun}$.

683: 

684: \input{tab5.tex}

685: 

686: %\begin{figure}

687: %\begin{center}

688: %\includegraphics[width=3.5in]{clustering_result.eps}

689: %\caption{Breakdown of clustering results by group size. The bins have height 0.33, 0.09, 0.04, 0.55.}

690: %\end{center}

691: %\end{figure}

692: 

693: \section{Discussion}

694: \subsection{Origin of the Distributed Population}

695: Our clustering analysis shows that there is a significant distributed

696: population of young stars (30--57\% of IR excess objects).  These may

697: have formed in groups or clusters and dispersed through random motions

698: or been ejected at high velocity.  Conversely, they may have formed in

699: their current locations in largely unrelated, isolated events.

700: 

701: On a $\sim$10 Myr timescale, the distributed population may be a

702: natural consequence of dynamical interactions between young stars in

703: groups throughout the original giant molecular cloud.  Small groups

704: ({\it N} $<$ 36) have short relaxation times relative to their

705: crossing times and, once the gas is removed, will not remain visible

706: as clusters for much longer than this \citep{adams01}.  However, if we

707: assume that the distributed stars are only as old as the oldest O star

708: in W5 (a reasonable assumption, given that we have identified them

709: through their IR excess emission), they will have had $<$ 5 Myr to

710: travel.

711: 

712: Extreme dynamical interactions, namely close encounters between stars

713: in multiple systems, can produce high velocity runaway stars.

714: Numerical studies show that a given triple system will eject a member

715: within about 100 crossing times \citep[$\sim$30000

716: years:][]{reipurth00}. \citet{sterzik95} calculate similar timescales

717: for close systems consisting of 5 stars, and predict velocities of

718: 3--4 km s$^{-1}$ for the ejected stars.  Few have been found;

719: \citet{goodman04} find 7 high velocity ($>$10 km s$^{-1}$) candidates

720: in the literature.

721: 

722: F{\H u}r{\'e}sz et al. (2006, 2008) measured the radial velocities of

723: young stars in NGC~2264 and the Orion Nebula Cluster, and found

724: dispersions of 3.5 and 3.1 km s$^{-1}$, respectively.  In W5, the

725: average projected distance of distributed stars from the nearest

726: cluster is 8.5 pc.  Assuming a velocity of 3 km s$^{-1}$ perpendicular

727: to the line of sight, this distance could be traversed in $\sim$3

728: Myr---well within the current upper age upper limit for W5 \citep[5

729: Myr, see:][]{karr03}. Thus it is plausible that at least some of the

730: distributed population originated in nearby clusters.

731: 

732: \subsection{Star Formation Efficiency}

733: We calculate the star formation efficiency in W5 using the equation:

734: 

735: \begin{equation}

736:   \epsilon = \frac{M_{stars}}{M_{stars}+M_{gas}}

737: \end{equation}

738: 

739: \noindent Here $M_{stars}$ represents the mass in {\it young} stars

740: only, as identified by their infrared excess, so $\epsilon$ is a

741: current efficiency, averaged over the last few million years. For the

742: entire W5 region we count 2064 young stars and assume an average mass

743: of 1 M$_{\sun}$. To find the gas mass we use the $A_V$ map of W5

744: described in $\S$ 3.1 and estimate the molecular gas column density

745: from the standard relation: $N(H_2)$ = 1.9 $\times$10$^{21} A_V$

746: protons cm$^{-2}$ \citep[see, for example:][]{bohlin78}. Assuming all

747: gas is associated with W5, we convert $N(H_2)$ to a projected

748: two-dimensional mass density at each pixel in the $A_V$ map: $\kappa$

749: = 15 $\times A_V$ M$_{\sun}$ pc$^{-2}$ \citep[as

750: in][]{lombardi01}. Each $A_V$ map pixel has a size 0.34$\times$0.34

751: pc. Over our whole survey field, we find a total gas mass of

752: 6.5$\times$10$^4$ M$_{\sun}$ and derive an efficiency of 3\%.

753: 

754: We calculate $\epsilon$ for the clusters identified by our MST

755: treatment in $\S$ 3.4. For each of the 16 clusters, we count the

756: number of stars contained, assume a stellar mass of 1 M$_{\sun}$ and

757: add up the gas in the corresponding pixels. Inside the H\,{\sc ii}

758: region cavity, where the molecular gas has been destroyed, the

759: efficiencies are high (26--39\%). In the rim of W5, where molecular

760: gas remains and embedded clusters are found, $\epsilon$ is

761: 10--17\%. These latter values are likely a lower limit, as our census

762: of young stars is incomplete in the bright extended emission

763: coincident with the embedded clusters.

764: 

765: Recent results from the c2d survey (Evans et~al. 2008 in prep.) show

766: that nearby dark clouds where low mass star formation is occurring

767: have star formation efficiencies of 2--4\% cloud-wide and 15--20\% in

768: clusters, similar to the values we have obtained for the high-mass

769: star forming region W5. Our results for $\epsilon$ in the clusters are

770: also comparable to those of \citet{chavarria08} for the S255 region.

771: 

772: %we make a 744$\times$700 grid of points

773: %across the W5 surveyed region. At each point in the grid we find the

774: %neharest 20 Class I or Class II objects and calculate their ratio. We

775: %convert this ratio map to a greyscale image in Figure 13 below. The

776: %greyscale ranges from a value of 0 (i.e. none in 20) up to 1.0 at its

777: %maximum near AFGL 4029. Overlaid on this image are contours of

778: %integrated $^{12}$CO intensity (W$_{CO}$ ranging from a minimum of 7.5

779: %K km s$^{-1}$ up to 157.1 K km s$^{-1}$)---a clear correlation between

780: %enhanced Class I numbers relative to Class II can be seen associated

781: %with higher CO emission. We believe this to be indicative of an age

782: %difference between the stellar population in the H\,{\sc ii} region

783: %cavity, and that associated with the molecular cloud. Note, however

784: %that edge effects due to low surface density of IR excess sources

785: %somewhat exaggerates the extent of regions of high Class I number at

786: %the far south-western part of W5, and in the dark region in the ??????

787: 

788: %\begin{figure}

789: %\begin{center}

790: %\includegraphics*[width=5.0in]{cl1cl2ratio.eps}

791: %\caption{Greyscale image demonstrating the variation in Class I to Class II object ratio in W5. The color scale runs from 0.0 (lightest) to 1.0 (darkest). Overlaid in black are $^{12}CO$ integrated intensity contours. Contours range from 7.5 to 157.1 K km s$^{-1}$ in uniform intervals.}

792: %\end{center}

793: %\end{figure}

794: 

795: \subsection{Triggered Star Formation in W5}

796: We consider two mechanisms of triggered star formation relevant to

797: W5. The first mechanism, called radiatively driven implosion (RDI), is

798: the compression of pre-existing density enhancements (small cloud

799: clumps or globules) inside and on the boundary of the ionized bubble

800: by the high pressure of the H\,{\sc ii} region \citep[for a review,

801: see][]{klein85}. Clumps of material visible as bright PAH emission in

802: IRAC bands 1, 3 and 4 with spatially coincident Class I or Class II

803: objects are summarized in Table 4. These include isolated clumps,

804: `elephant trunk' formations and bright rims (as shown in Figure 8). As

805: discussed in \citet{elmegreen98}, if globule squeezing does occur, its

806: timescale is expected to be short, since an isolated overdensity can

807: be compressed immediately on being engulfed by the H\,{\sc ii} region

808: or stellar wind. \citet{thompson04} investigated the cometary features

809: on the southern rim of W5 West for signs of star formation triggered

810: via the RDI mechanism (globules 4--6 in our Table 4). They found that

811: the H$\alpha$, CO-molecular and dust morphologies of the three pillars

812: are `reasonably consistent' with the model of \citet{lefloch94} for

813: radiatively driven implosion, at the early collapse phase. They used

814: near-IR photometry to detect several candidate young stellar

815: objects---these we confirm through our {\it Spitzer} photometry as

816: Class I and Class II objects. They estimate that the timescales for

817: duration of UV illumination, shock crossing times across the features,

818: and protostar/YSO ages are all $\sim$10$^5$ yrs. Since

819: \citet{thompson04} were only able to establish that approximate

820: pressure equilibrium holds in the pillar heads from their data, we

821: need a more detailed study of the gas dynamics in these pillars to

822: determine if they are currently collapsing due to external pressure

823: from the H\,{\sc ii} region.

824: 

825: The second mechanism, `collect and collapse,' is the large scale

826: expansion of an ionized bubble, powered by an over-pressurized H\,{\sc

827:   ii} region and stellar winds. This expansion can drive shock fronts

828: into the surrounding medium, sweeping up a dense ridge or shell which

829: collapses into stars when a critical density of material

830: accumulates. \citet{whitworth94} presented an analytical treatment of

831: this process. A shock front forms and gathers material until it is

832: able to fragment and collapse to form stars, at a time $t_{frag}$ and

833: radius $R_{frag}$ given by:

834: 

835: \begin{equation}

836: t_{frag} = 1.56 \textrm{ Myr }a_{0.2}^{7/11}\textrm{ }L_{49}^{-1/11}\textrm{ }n_3^{-5/11}

837: \end{equation}

838: \begin{equation}

839: R_{frag} = 5.8 \textrm{ pc }a_{0.2}^{4/11}\textrm{ }L_{49}^{1/11}\textrm{ }n_3^{-6/11}

840: \end{equation}

841: 

842: where $a_{0.2}$ is the sound speed inside the shocked layer in units

843: of 0.2 km s$^{-1}$, $L_{49}$ is the central source ionizing flux in

844: units of 10$^{49}$ photons s$^{-1}$ and $n_3$ is the initial gas

845: atomic number density in units of 10$^3$ cm$^{-3}$. 

846: 

847: In any massive star forming region, presumably some combination of

848: H\,{\sc ii} region expansion and stellar wind forces operate. Stellar

849: winds certainly play a significant role in shaping the morphology of

850: an H\,{\sc ii} region only until an age of $\sim 10^5$years. As shown

851: by \citet{mckee84} and \citet{weaver77}, weak stellar winds (wind

852: luminosity $\ll$ 1.26$\times$10$^{36}\times L_{49}\times n$, where $n$

853: is the ambient density) are likely to be confined by the H\,{\sc ii}

854: region pressure to a smaller bubble around the ionizing source, since

855: gas that is evaporated from clumps of material in the H\,{\sc ii}

856: region mixes with the hot stellar wind bubble and can radiate energy

857: away efficiently. Strong winds can overcome this confinement---they

858: create a bubble with size determined by the photoevaporation and

859: displacement of surrounding inhomogeneities in the ambient gas. This

860: then expands along with the H\,{\sc ii} region.  In estimating the

861: lifetimes of the W5 bubbles, we thus consider only the effects of the

862: H\,{\sc ii} region expansion.

863: 

864: As described in $\S$1, the gross morphology of W5 in the mid-IR is

865: defined by two large, roughly circular rings of bright PAH emission

866: that mark the smaller W5 East (W5E, radius $\sim$10 pc), and larger,

867: more irregular W5 West (W5W, overall radius $\sim$22 pc). At the

868: approximate center of W5E lies HD 18326, an O7V star surrounded by a

869: dense cluster of young stars. It seems reasonable to assume that

870: whether this is a sphere or a ring of gas, that the cluster lies near

871: the true center, and is not significantly offset either to the

872: foreground or background. Adopting stellar parameters from

873: \citet{martins05}, an O7V star will have ionizing flux $L_{49}$ =

874: 0.43. Since we cannot calculate appropriate values for $a_{0.2}$ and

875: $n_3$ from our current datasets, we adopt 1.0 for both. The H\,{\sc

876:   ii} region fragmentation time we calculate is 1.69 Myr at a radius

877: of 5.37 pc. Both \citet{whitworth94} and \citet{dale07} note that the

878: density in a real region is likely to be lower---this would make the

879: fragmentation time longer and the radius larger by a factor

880: $\sim$2. 

881: 

882: W5W presents a less clear-cut picture. It contains an isolated O8V

883: star: HD 237019, as well as three dense clusters centered on BD +60

884: 586 (O7.5V), and the multiple systems HD 17505 (two O7.5V((f)) stars,

885: an O6.5III((f)) and an O8.5V) and HD 17520 (O9V + Be). See

886: \citet{hillwig06} for a more detailed study of the O stars in W5. The

887: relative configuration of these four systems, and their relationship

888: to the surrounding diffuse gas (whether in front or behind) is not

889: known. We consider here two simple scenarios and assume that all four

890: objects and the ring of PAH emission are in the same plane. Scenario

891: 1: HD 237019 represents an initial episode of star formation, and its

892: isolation is due to its cluster having had time to disperse. In the

893: Whitworth model, an O8V star has a fragmentation time of 1.8 Myr and

894: collapse of swept up gas occurs at a radius of 5.0 pc. In this

895: picture, this event triggered the formation of the three dense

896: clusters in W5W which all lie at roughly 6.8 pc from HD 237019. These

897: clusters, and the continuing expansion of the H\,{\sc ii} region then

898: triggered the current, ongoing star formation seen in association with

899: molecular clouds W5NW and W5NE. Scenario 2: the three dense clusters

900: form together---with their present arc-like arrangement due to some

901: initial filamentary distribution of molecular material. They then

902: trigger star formation in the remaining molecular material to the

903: north, north-east and south. The O7.5V star BD +60 586 by itself has a

904: fragmentation time of 1.75 Myr and collapse occurs at a radius of 5.16

905: pc. The combined systems HD 17505 and HD 17520 \citep[with at least 5

906: O stars, see:][]{hillwig06} have a fragmentation time of 1.46 Myr and

907: collapse occurs at a radius of 6.19 pc. In this picture, HD 237019

908: would have been ejected from one of the multiple O star

909: systems. Located at $\approx$10 pc from the nearest cluster, assuming

910: a velocity entirely transverse to our line of sight of 10 km s$^{-1}$,

911: it would take $\sim$1 Myr to arrive at its current location,

912: consistent with the likely young age of the cluster as a

913: whole. \citet{moffat98} found a runaway frequency percentage of 14\%

914: among Galactic O stars and included this star as a marginal runaway

915: candidate. Although the most favored creation mechanism for runaways

916: is supernova ejection \citep{blaauw61}, ejection from compact young

917: stellar clusters or through binary interaction is also thought to be

918: responsible for some O star runaways

919: \citep{poveda67,clarke92}. \citet{moffat98} found HD 237019 to be only

920: a marginal runaway candidate, given the large uncertainty in its

921: tangential velocity: 15$\pm$17 pc Myr$^{-1}$.

922: 

923: There are several observations consistent with the collect and

924: collapse mechanism in W5. The low level of non-thermal radio emission

925: suggests there have been no supernovae during its lifetime

926: \citep{vallee79}. The main sequence lifetimes of the O stars are

927: estimated to be $<$ 5 Myr \citep{karr03}. The timescales for

928: triggering presented above are shorter than this. \citet{wilking84}

929: found several young embedded OB stars around W5 in the molecular

930: clouds, and argued that their locations near to the cloud edge,

931: together with some evidence of magnetic field alignment parallel to

932: the ionization front are suggestive of triggering via external

933: compression. Finally, \citet{nakano08} found evidence of an age

934: difference of $\sim$3 Myr between the cluster around HD 18326 in the

935: W5 East bubble and the young stars in the rim. This is longer than our

936: estimate above of $\sim$1 Myr, but given the uncertainties in gas

937: sound speed $a$ and initial cloud number density $n$, is an allowable

938: timescale in the collect and collapse model.

939: 

940: \citet{karr03} argue that the distribution of young stars in W5

941: indicates a shorter timescale, more consistent with RDI than with

942: collect and collapse. They investigated the locations of IRAS point

943: sources with colors corresponding to YSOs. Their results did suggest a

944: 2-generation model, with older stars at the centers of the H\,{\sc ii}

945: regions and ongoing star formation surrounding them, however they

946: found that the number density of young objects peaks $\sim$5 pc inside

947: the H\,{\sc ii} region (as defined by the 6.2K 1420 MHz contour),

948: which corresponds to a triggering timescale of 0.5--1 Myr. Of the 42

949: IRAS sources from their list that fall within our {\it Spitzer} image

950: we find only 16 in our source list at 24 $\micron$. The remainder may

951: be spurious (they are also not seen in our 70 $\micron$ image) and in

952: fact may be small knots of IR emission that masquerade as point

953: sources in the IRAS survey. All are associated with bright PAH

954: emission, including one coincident with globule 8 in Table 4. The peak

955: in IRAS-classified YSOs found by \citet{karr03} is also not upheld by

956: an analysis of our {\it Spitzer} sample which shows that the Class I

957: objects are found predominantly along the cloud rims or at the cloud

958: centers. This distribution, with the YSOs further from the O stars

959: argues for the longer timescale of collect and collapse.

960: 

961: Studies of a young embedded object G138.295+1.555 and the UCH\,{\sc

962:   ii} region G138.300+1.558 in AFGL4029 (W5E) with near- and mid-IR

963: imaging \citep{deharveng97,zavagno99} suggest that collect and

964: collapse is not playing a role in star formation here, since the

965: youngest object (the former) is apparently closer to the ionization

966: front than the older, latter object. However the true 3-dimensional

967: configuration of all the objects involved is not exactly known, this

968: may be due to projection effects. The clustering of Class II objects

969: immediately outside the cloud to the west of G138.295+1.555 and

970: G138.300+1.558 is certainly suggestive that the sequence of star

971: formation here is west to east, i.e. outside-in, away from the

972: ionizing star HD 18326.

973: 

974: In consideration of our simple model and the results in the

975: literature, it seems plausible that both the RDI and collect and

976: collapse mechanisms are at work in W5; RDI on the smaller scale of

977: cometary globules, and collect and collapse on the larger scale of the

978: H\,{\sc ii} region. Detailed investigations (Koenig et al. 2008, in

979: preparation) of the spatial distributions and relative ages of the

980: YSOs may further constrain the scenarios presented here.

981: 

982: \section{Conclusions and Future Work}

983: We have presented initial results from our extensive {\it Spitzer}

984: survey of W5. Shorter wavelength data from IRAC (3--8 $\micron$) and

985: longer (24 $\micron$) wavelength data from the MIPS instrument were

986: combined to maximize spectral coverage of detected sources.

987: 

988: Even before photometric analysis, dense clusters of stars are clearly

989: visible across the region, centered on the O stars HD 18326, BD +60

990: 586, HD 17505 and HD 17520, and also across the extensive PAH emission

991: that defines the outline of W5. At 24 $\micron$, substantial extended

992: emission is visible, presumably from heated dust that survives in the

993: strongly ionizing environment of the H\,{\sc ii} region.

994: 

995: We used photometry of more than 18000 point sources to analyze the

996: spatial distributions of young stars, establish their evolutionary

997: status via their infrared colors and magnitudes, and assess their

998: clustering properties across this large star forming region. The large

999: clusters that dominate the region, centered on the massive O stars,

1000: contain numerous infrared excess sources. Looking at the large scale

1001: distributions of stars at different evolutionary stages, we find that

1002: within the evacuated cavity of the H\,{\sc ii} regions that make up

1003: W5, the ratio of Class II (older) to Class I (younger) sources is

1004: $\sim$7 times higher than for objects detected coincident with the

1005: molecular clouds in the rim. We attribute this difference to an age

1006: difference between the two locations, and consequently postulate that

1007: at least two, distinct generations of star formation are visible in

1008: the region. An isolated O star in W5 West, HD 237019, may represent an

1009: initial episode of star formation in the region, preceding the

1010: formation of the large clusters in the cavity, although we cannot rule

1011: out its ejection from an O star multiple system, for example HD 17505.

1012: 

1013: The clustering results show that, considering infrared excess sources

1014: alone, (2064 objects) $\sim$45--70\% are found in clusters with

1015: $\geq$10 members. Incorporating the sources apparently misclassified

1016: as AGN and PAH galaxies (see Appendix) extends this range to

1017: $\sim$40--70\%. The remainder are in the distributed mode, many of

1018: which could have formed in nearby clusters.

1019: 

1020: We looked at the role that {\it triggered} star formation may have

1021: played in W5. We catalogued isolated globules of diffuse PAH emission,

1022: and so-called `elephant-trunk' structures that contain young

1023: protostars or stars with protostellar disks. These are examples of

1024: possible RDI-triggered star formation events. On the larger scale, we

1025: tested the analytical formulations of \citet{whitworth94} for the

1026: collect and collapse mechanism. Our simple estimates show that

1027: triggering remains a plausible mechanism to explain the multiple

1028: generations of star formation in W5 and merits further investigation.

1029: 

1030: Substantial work is ongoing to refine our understanding of star

1031: formation in W5 in the light of triggering models and the clustering

1032: of stars. We have undertaken a near infrared survey of the whole

1033: region in $J$, $H$ and $K_S$ bands to extend the stellar SEDs to

1034: shorter wavelengths and detect more young stars against the bright

1035: background emission in W5. This will permit a full analysis of the

1036: clustering of young stars in the region.  We have obtained optical

1037: spectra of several hundred stars across W5 with the aim of determining

1038: their relative ages by constructing H-R diagrams. This will allow us

1039: to better understand the history of star formation across the entire

1040: region. A much clearer picture of the progression of star formation

1041: across W5 and the role of feedback in this process should result.

1042: 

1043: \acknowledgements The authors would like to thank Paul Harvey and Phil

1044: Myers for a very helpful analysis of the manuscript and useful

1045: discussions, and Fred Adams, Tom Megeath and Neal Evans for helpful

1046: discussions on clustering analysis. This work is based (in part) on

1047: observations made with the {\it Spitzer} Space Telescope, which is

1048: operated by the Jet Propulsion Laboratory, California Institute of

1049: Technology under a contract with NASA. Support for this work was

1050: provided by NASA. This publication makes use of data products from the

1051: Two Micron All Sky Survey, which is a joint project of the University

1052: of Massachusetts and the Infrared Processing and Analysis

1053: Center/California Institute of Technology, funded by the National

1054: Aeronautics and Space Administration and the National Science

1055: Foundation. This research has made use of NASA's Astrophysics Data

1056: System. This research has made use of the SIMBAD database, operated at

1057: CDS, Strasbourg, France.

1058: 

1059: {\it Facilities:} \facility{2MASS ($JHK_S$)}, \facility{Spitzer (IRAC, MIPS)}

1060: 

1061: \section*{Appendix}

1062: \appendix

1063: \section{AGN/PAH Galaxy filtering}

1064: Following \citet{gutermuth08} PAH galaxy candidates are identified on

1065: the basis of the following criteria:

1066: 

1067: \begin{displaymath}

1068: [4.5]-[5.8] < \frac{1.05}{1.2} ([5.8]-[8.0]-1),

1069: \end{displaymath}

1070: \begin{displaymath}

1071: [4.5]-[5.8] < 1.05,

1072: \end{displaymath}

1073: \begin{displaymath}

1074: [5.8]-[8.0] > 1,

1075: \end{displaymath}

1076: \begin{displaymath}

1077: [3.6]-[5.8] < \frac{1.5}{2} ([4.5]-[8.0]-1),

1078: \end{displaymath}

1079: \begin{displaymath}

1080: [3.6]-[5.8] < 1.5,

1081: \end{displaymath}

1082: \begin{displaymath}

1083: [4.5]-[8.0] > 1.

1084: \end{displaymath}

1085: 

1086: After removing these sources, AGN are then picked out following these

1087: criteria:

1088: 

1089: \begin{displaymath}

1090: [4.5]-[5.8] > 0.5,

1091: \end{displaymath}

1092: \begin{displaymath}

1093: [4.5] > 13.5 + ([4.5]-[8.0]-2.3)/0.4,

1094: \end{displaymath}

1095: \begin{displaymath}

1096: [4.5] > 13.5.

1097: \end{displaymath}

1098: 

1099: We plot the distribution of 729 AGN and 198 PAH galaxy candidates in

1100: W5 in Figure 13. It is notable that these objects show obvious

1101: non-uniformity. The non-uniformity arises from several sources: 1)

1102: variation in the intrinsic distribution of galaxies toward (behind)

1103: W5, 2) variable extinction across the region due to the molecular

1104: clouds that make up W5 and material along the line of sight through

1105: the Galaxy in this direction, 3) variable levels of photometric

1106: completeness across the field due to bright extended nebular emission

1107: and 4) clustering coincident with the positions of the dense clusters

1108: of young stars around the O stars in the W5 H\,{\sc ii} region cavity,

1109: due to faint (low mass) stars misclassified as AGN or PAH

1110: galaxies. This last issue affects all of our analysis, thus we need to

1111: characterize---statistically---the properties of these misclassified

1112: objects and how adding them back to the original sample affects our

1113: previous results.

1114: 

1115: \begin{figure}

1116: \begin{center}

1117: \includegraphics[width=4in]{f13.eps}

1118: \caption{Distribution of extragalactic contaminants as classified by {\it Spitzer} photometry: AGN (blue points) and PAH galaxies (red).}

1119: \end{center}

1120: \end{figure}

1121: 

1122: We consider the AGN first. In the extreme case where all objects are

1123: misclassified, we return the full list of 729 AGN to the stellar list,

1124: classify the returned objects into Class I, II etc., and examine the

1125: clustering fraction. With the MST straight line fit method, we find

1126: $d_s$ = 0.88 pc, $N_{grp}$ = 21 and a clustered fraction = 68.6\%. If

1127: we use the maximum in the $N_{grp}$ distribution we find $d_s$ = 0.5

1128: pc, $N_{grp}$ = 26 and a clustered fraction = 40.4\%.

1129: 

1130: Instead of this, presumably some intermediate fraction of AGN should

1131: be returned as AGN. We must first calculate the expected number and

1132: averaged distribution of AGN in W5, using a sample of AGN extracted by

1133: \citet{stern05} from the {\it Spitzer} IRAC shallow survey of the

1134: Bo\"otes field \citep{eisenhardt04}, convolved with our extinction map

1135: (as described in $\S$ 2.1) and with completeness estimates derived for

1136: our {\it Spitzer} survey of W5 via a simple Monte Carlo simulation

1137: code. The model AGN sample is drawn from a 7.7 square degree section

1138: of the \citet{stern05} survey, identifying sources as AGN based on the

1139: same color classification scheme that we use in this paper

1140: \citep{gutermuth08}.

1141: 

1142: We use the IRAC completeness maps as described in $\S$ 3.1. The area

1143: imaged in our W5 survey (3.5 square degrees) is smaller than that

1144: covered by the Bo\"otes data. In our Monte Carlo simulation, we input

1145: a proportionately smaller, random selection of AGN from the Bo\"otes

1146: list and distribute these randomly in the field of W5. Each object is

1147: assigned an $A_V$ value based on its location in the extinction map.

1148: We apply this extinction to its IRAC photometry using the infrared

1149: extinction relation presented in \citet{flaherty07}. The corresponding

1150: location within the completeness maps gives us the completeness value

1151: (a number $\leq$1) at each of the four IRAC wavelengths given its

1152: extincted magnitude. A random number generator is then used to

1153: determine whether or not an object is detected, with the completeness

1154: estimates as a measure of the probability of detection. For detection

1155: we require that an object be detected in all four IRAC bands. This

1156: simulation predicts on average 270 AGN detected in W5.

1157: 

1158: This distribution of AGN has a certain characteristic space density

1159: distribution. At each iteration we measure the `nearest-neighbor'

1160: distance distribution \citep{casertano85}. For each object we find the

1161: projected distance to its sixth nearest neighbor in arcsec ($d_6$) and

1162: construct a histogram of distances. The final output of the code is an

1163: averaged histogram combining 1000 outcomes of the simulation. For

1164: comparison, we generate the distribution of $d_6$ for the real AGN

1165: candidates in W5 and compare the results in Figure 14 (leftmost

1166: panel). In the right-hand panels we compare the output magnitude and

1167: color distributions of the simulated AGN (averaged over 1000 outcomes)

1168: with our W5 AGN candidates.

1169: 

1170: \begin{figure}[ht]

1171: \begin{center}

1172: \includegraphics*[width=2.12in]{f14a.eps}

1173: \includegraphics*[width=2.12in]{f14b.eps}

1174: \includegraphics*[width=2.12in]{f14c.eps}

1175: \caption{Distributions of nearest neighbor distances, [4.5] magnitudes

1176:   and $[4.5]-[8.0]$ colors for objects classified as AGN in W5 (black

1177:   histograms). Plotted in gray are the simulated distributions of

1178:   simulated AGN according to our completeness estimates. We attribute

1179:   the excess of objects in the W5 sample to young stars misclassified

1180:   as AGN by our classification scheme.}

1181: \end{center}

1182: \end{figure}

1183: 

1184: These distributions can be used to generate a probability that an

1185: object in our AGN list, of given $d_6$, $[4.5]$ magnitude and

1186: $[4.5]-[8.0]$ color is a YSO. In Table 6 we present the source

1187: properties of the 729 W5 AGN candidates. The full machine-readable

1188: table is given in the online material for this paper. Each object is

1189: listed with its coordinates and photometry in 2MASS and {\it Spitzer}

1190: bands, and given a combined `P(YSO)' value, that is the probability

1191: given its local surface density, $[4.5]$ magnitude and $[4.5] - [8.0]$

1192: color.

1193: 

1194: We statistically return a sample of AGN to the YSO list based on these

1195: probabilities to test the effect on the clustered fraction. We return

1196: samples based on the density probability alone, and then the combined

1197: density, color and magnitude probability. We run our MST clustering

1198: analysis on the resultant list of YSOs plus YSO-AGN. In Table 7 below

1199: we show the averaged results of adding AGN back to the YSO sample

1200: according to the different filters and compare to the original

1201: YSO-only sample result.

1202: 

1203: \input{stub.tab6.tex}

1204: 

1205: The space density filter typically returns $\approx$460 objects to the

1206: list. The combined filter returns $\sim$230.  Although a significant

1207: fraction of objects that are returned from AGN to YSOs join the

1208: clustered population, many are added to the distributed sources which

1209: increases the fraction outside groups and clusters.

1210: 

1211: The objects returned to the stellar sample will also affect the Class

1212: I/Class II ratio. Applying the density filter returns a sample with

1213: $\sim$16\% Class I objects and the remainder Class II. If we apply the

1214: combined filter, typically 4\% are Class I, 96\% Class II. The effect

1215: on the ratio test is small: we increase the proportion of Class I

1216: objects in the H\,{\sc ii} cavity from 4\% to $\sim$7\%. The fraction

1217: associated with the molecular gas remains at 23\%, thus our earlier

1218: result and conclusion still holds. A summary of the results of these

1219: results is given in Table 7 below.

1220: 

1221: \input{tab7.tex}

1222: 

1223: For the smaller sample of PAH galaxies, statistical analysis carried

1224: out as above predicted on average 211 objects. However, we find 198

1225: candidate PAH galaxies in W5, of which certainly some fraction are

1226: stellar objects as can be seen from the non-uniformity in Figure

1227: 13. Although our model predicts an excess of PAH galaxies over that

1228: detected, we can still use the same procedure for these objects as for

1229: the AGN to generate a probability P(YSO), based on colors and space

1230: density. We present the 198 PAH galaxy candidates in Table 6 along

1231: with the AGN with the same format of coordinates and photometry. We

1232: provide a P(YSO) value for each, derived from local space density and

1233: $[3.6] - [5.8]$ and $[4.5] - [8.0]$ colors.

1234: 

1235: %\begin{figure}

1236: %\begin{center}

1237: %\includegraphics[width=5in]{pah_nnd6.eps}

1238: %\caption{Nearest neighbor distance distribution for objects classified as PAH galaxies by our pipeline.}

1239: %\end{center}

1240: %\end{figure}

1241: 

1242: %\bibliographystyle{apj} \bibliography{koenig}

1243: \begin{thebibliography}{59}

1244: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

1245: 

1246: \bibitem[{{Adams} \& {Myers}(2001)}]{adams01}

1247: {Adams}, F.~C., \& {Myers}, P.~C. 2001, \apj, 553, 744 

1248: 

1249: \bibitem[{{Allen} {et~al.}(2007){Allen}, {Megeath}, {Gutermuth}, {Myers},

1250:   {Wolk}, {Adams}, {Muzerolle}, {Young}, \& {Pipher}}]{allen07}

1251: {Allen}, L., et~al.\ 2007, in  Protostars and Planets V, ed. B.~{Reipurth}, D.~{Jewitt}, \& K.~{Keil},  361--376

1252: 

1253: \bibitem[{{Allen} {et~al.}(2004){Allen}, {Calvet}, {D'Alessio}, {Merin},

1254:   {Hartmann}, {Megeath}, {Gutermuth}, {Muzerolle}, {Pipher}, {Myers}, \&

1255:   {Fazio}}]{allen04}

1256: {Allen}, L.~E., et~al.\ 2004, \apjs, 154, 363

1257: 

1258: \bibitem[{{Allen} {et~al.}(2005){Allen}, {Hora}, {Megeath}, {Deutsch}, {Fazio},

1259:   {Chavarr\'ia}, \& {Dell}}]{allen05}

1260: {Allen}, L.~E., et~al.\ 2005, in IAU Symposium, Vol. 227,  Massive Star Birth: A Crossroads of Astrophysics, ed. R.~{Cesaroni},  M.~{Felli}, E.~{Churchwell}, \& M.~{Walmsley}, 352--357

1261: 

1262: \bibitem[{{Alter} {et~al.}(1970){Alter}, {Balazs}, {Ruprecht}, \& {Vanysek}}]{alter70}

1263: {Alter}, G., {Balazs}, B., {Ruprecht}, J., \& {Vanysek}, J. 1970, Catalogue of Star Clusters and Associations (2nd ed.; Budapest: Akademiai Kiado)

1264: 

1265: \bibitem[{{Baraffe} {et~al.}(1998){Baraffe}, {Chabrier}, {Allard}, \& {Hauschildt}}]{baraffe98}

1266: {Baraffe}, I., {Chabrier}, G., {Allard}, F., \& {Hauschildt}, P.~H. 1998, \aap, 337, 403

1267: 

1268: \bibitem[{Bastian} {et~al.}(2007){Bastian}, {Ercolano}, {Gieles}, {Rosolowsky}, {Scheepmaker}, {Gutermuth}, {Efremov}]{bastian07}

1269: {Bastian}, N., {Ercolano}, B., {Gieles}, M., {Rosolowsky}, E., {Scheepmaker}, R.~A., {Gutermuth}, R., \& {Efremov}, Y. 2007, \mnras, 379, 1302

1270: 

1271: \bibitem[{{Battinelli} (1991)}]{batt91}

1272: {Battinelli}, P. 1991, \aap, 244, 69

1273: 

1274: \bibitem[{{Becker} \& {Fenkart}(1971)}]{becker71}

1275: {Becker}, W., \& {Fenkart}, R. 1971, \aaps, 4, 241

1276: 

1277: \bibitem[{{Bessell} \& {Brett}(1988)}]{bessell88}

1278: {Bessell}, M.~S., \& {Brett}, J.~M. 1988, \pasp, 100, 1134

1279: 

1280: \bibitem[{{Blaauw}(1961)}]{blaauw61}

1281: {Blaauw}, A. 1961, \bain, 15, 265

1282: 

1283: \bibitem[{{Bohlin} {et~al.}(1978){Bohlin}, {Savage}, \& {Drake}}]{bohlin78}

1284: {Bohlin}, R.~C., {Savage}, B.~D., \& {Drake}, J.~F. 1978, \apj, 224, 132

1285: 

1286: \bibitem[{{Brown} {et~al.}(2008){Brown}, {Blake}, {Qi}, {Dullemond}, \& {Wilner}}]{brown08}

1287: {Brown}, J.~M., {Blake}, G.~A., {Qi}, C., {Dullemond}, C.~P., \& {Wilner}, D.~J. 2008, \apjl, 675, L109

1288: 

1289: \bibitem[{{Carpenter}(2000)}]{carpenter00}

1290: {Carpenter}, J.~M. 2000, \aj, 120, 3139

1291: 

1292: \bibitem[{{Cartwright} \& {Whitworth}(2004)}]{cartwright04}

1293: {Cartwright}, A., \& {Whitworth}, A.~P. 2004, \mnras, 348, 589

1294: 

1295: \bibitem[{{Casertano} \& {Hut}(1985)}]{casertano85}

1296: {Casertano}, S., \& {Hut}, P. 1985, \apj, 298, 80

1297: 

1298: \bibitem[{{Chavarr\'ia} {et~al.}(2008){Chavarr\'ia}, {Allen}, {Hora}, {Brunt}, \& {Fazio}}]{chavarria08}

1299: {Chavarr\'ia}, L., {Allen}, L., {Hora}, J.~L., {Brunt}, C., \& {Fazio}, G.~G. 2008, \apj, 682, 445

1300: 

1301: \bibitem[{{Churchwell} {et~al.}(2006){Churchwell}, {Povich}, {Allen}, {Taylor}, {Meade}, {Babler}, {Indebetouw}, {Watson}, {Whitney}, {Wolfire}, {Bania}, {Benjamin}, {Clemens}, {Cohen}, {Cyganowski}, {Jackson}, {Kobulnicky}, {Mathis}, {Mercer}, {Stolovy}, {Uzpen}, {Watson}, \& {Wolff}}]{church06}

1302: {Churchwell}, E., et~al.\ 2006, \apj, 649, 759

1303: 

1304: \bibitem[{{Cieza} {et~al.}(2007){Cieza}, {Padgett}, {Stapelfeldt}, {Augereau}, {Harvey}, {Evans}, {Mer{\'{\i}}n}, {Koerner}, {Sargent}, {van Dishoeck}, {Allen}, {Blake}, {Brooke}, {Chapman}, {Huard}, {Lai}, {Mundy}, {Myers}, {Spiesman}, \& {Wahhaj}}]{cieza07}

1305: {Cieza}, L., et~al.\ 2007, \apj, 667, 308 

1306: 

1307: \bibitem[{{Clarke} \& {Pringle}(1992)}]{clarke92}

1308: {Clarke}, C.~J., \& {Pringle}, J.~E. 1992, \mnras, 255, 423

1309: 

1310: \bibitem[{{Cr{\'e}t{\'e}} {et~al.}(1999){Cr{\'e}t{\'e}}, {Giard}, {Joblin},

1311:   {Vauglin}, {L{\'e}ger}, \& {Rouan}}]{crete99}

1312: {Cr{\'e}t{\'e}}, E., {Giard}, M., {Joblin}, C., {Vauglin}, I., {L{\'e}ger}, A.,

1313:   \& {Rouan}, D. 1999, \aap, 352, 277

1314: 

1315: \bibitem[{{D'Antona} \& {Mazzitelli}(1994)}]{dantona94}

1316: {D'Antona}, F. \& {Mazzitelli}, I. 1994, \apjs, 90, 467

1317: 

1318: \bibitem[{{Dale} {et~al.}(2007{\natexlab{a}}){Dale}, {Bonnell}, \&

1319:   {Whitworth}}]{dale07}

1320: {Dale}, J.~E., {Bonnell}, I.~A., \& {Whitworth}, A.~P. 2007{\natexlab{a}},

1321:   \mnras, 375, 1291

1322: 

1323: \bibitem[{{Deharveng} {et~al.}(1997){Deharveng}, {Zavagno}, {Cruz-Gonzalez},

1324:   {Salas}, {Caplan}, \& {Carrasco}}]{deharveng97}

1325: {Deharveng}, L., {Zavagno}, A., {Cruz-Gonzalez}, I., {Salas}, L., {Caplan}, J.,

1326:   \& {Carrasco}, L. 1997, \aap, 317, 459

1327: 

1328: \bibitem[{{Draine} \& {Li}(2007)}]{draine07}

1329: {Draine}, B.~T., \& {Li}, A. 2007, \apj, 657, 810

1330: 

1331: \bibitem[{{Eisenhardt} {et~al.}(2004){Eisenhardt}, {Stern}, {Brodwin}, {Fazio},

1332:   {Rieke}, {Rieke}, {Werner}, {Wright}, {Allen}, {Arendt}, {Ashby}, {Barmby},

1333:   {Forrest}, {Hora}, {Huang}, {Huchra}, {Pahre}, {Pipher}, {Reach}, {Smith},

1334:   {Stauffer}, {Wang}, {Willner}, {Brown}, {Dey}, {Jannuzi}, \&

1335:   {Tiede}}]{eisenhardt04}

1336: {Eisenhardt}, P.~R., et~al.\ 2004, \apjs, 154, 48

1337: 

1338: \bibitem[{{Elmegreen}(1998)}]{elmegreen98}

1339: {Elmegreen}, B.~G. 1998, in Astronomical Society of the Pacific Conference

1340:   Series, Vol. 148, Origins, ed. C.~E. {Woodward}, J.~M. {Shull}, \& H.~A.

1341:   {Thronson}, Jr., 150--+

1342: 

1343: \bibitem[{{Enoch} {et~al.}(2007){Enoch}, {Glenn}, {Evans}, {Sargent}, {Young}, \& {Huard}}]{enoch07}

1344: {Enoch}, M.~L., {Glenn}, J., {Evans}, II, N.~J., {Sargent}, A.~I., {Young}, K.~E., \& {Huard}, T.~L. 2007, \apj, 666, 982

1345: 

1346: \bibitem[{{Evans} {et~al.}(2003){Evans}, {Allen}, {Blake}, {Boogert}, {Bourke}, {Harvey}, {Kessler}, {Koerner}, {Lee}, {Mundy}, {Myers}, {Padgett}, {Pontoppidan}, {Sargent}, {Stapelfeldt}, {van Dishoeck}, {Young}, \& {Young}}]{evans03}

1347: {Evans}, N.~J., II, et~al.\ 2003, \pasp, 115, 965

1348: 

1349: \bibitem[{{Fazio} {et~al.}(2004){Fazio}, {Hora}, {Allen}, {Ashby}, {Barmby},

1350:   {Deutsch}, {Huang}, {Kleiner}, {Marengo}, {Megeath}, {Melnick}, {Pahre},

1351:   {Patten}, {Polizotti}, {Smith}, {Taylor}, {Wang}, {Willner}, {Hoffmann},

1352:   {Pipher}, {Forrest}, {McMurty}, {McCreight}, {McKelvey}, {McMurray}, {Koch},

1353:   {Moseley}, {Arendt}, {Mentzell}, {Marx}, {Losch}, {Mayman}, {Eichhorn},

1354:   {Krebs}, {Jhabvala}, {Gezari}, {Fixsen}, {Flores}, {Shakoorzadeh}, {Jungo},

1355:   {Hakun}, {Workman}, {Karpati}, {Kichak}, {Whitley}, {Mann}, {Tollestrup},

1356:   {Eisenhardt}, {Stern}, {Gorjian}, {Bhattacharya}, {Carey}, {Nelson},

1357:   {Glaccum}, {Lacy}, {Lowrance}, {Laine}, {Reach}, {Stauffer}, {Surace},

1358:   {Wilson}, {Wright}, {Hoffman}, {Domingo}, \& {Cohen}}]{fazio04}

1359: {Fazio}, G.~G., et~al.\ 2004, \apjs, 154, 10

1360: 

1361: \bibitem[{{Flaherty} {et~al.}(2007){Flaherty}, {Pipher}, {Megeath}, {Winston},

1362:   {Gutermuth}, {Muzerolle}, {Allen}, \& {Fazio}}]{flaherty07}

1363: {Flaherty}, K.~M., {Pipher}, J.~L., {Megeath}, S.~T., {Winston}, E.~M.,

1364:   {Gutermuth}, R.~A., {Muzerolle}, J., {Allen}, L.~E., \& {Fazio}, G.~G. 2007,

1365:   \apj, 663, 1069

1366: 

1367: \bibitem[{{Flaherty} \& {Muzerolle}(2008)}]{flaherty08}

1368: {Flaherty}, K.~M., \& {Muzerolle}, J. 2008, \aj, 135, 966

1369: 

1370: \bibitem[{{F{\H u}r{\'e}sz} {et~al.}(2006){F{\H u}r{\'e}sz}, {Hartmann}, {Szentgyorgyi}, {Ridge}, {Rebull}, {Stauffer}, {Latham}, {Conroy}, {Fabricant}, \& {Roll}}]{furesz06}

1371: {F{\H u}r{\'e}sz}, G. et~al. 2006, \apj, 648, 1090

1372: 

1373: \bibitem[{{F{\H u}r{\'e}sz} {et~al.}(2008){F{\H u}r{\'e}sz}, {Hartmann}, {Megeath}, {Szentgyorgyi}, \& {Hamden}, E.~T.}]{furesz08}

1374: {F{\H u}r{\'e}sz}, G., {Hartmann}, L.~W., {Megeath}, S.~T., {Szentgyorgyi}, A.~H., \& {Hamden}, E.~T. 2008, \apj, 676, 1109

1375: %\bibitem[{{Geller} \& {Beers}(1982)}]{geller82}

1376: %{Geller}, M.~J., \& {Beers}, T.~C. 1982, \pasp, 94, 421

1377: 

1378: \bibitem[{{Goodman} \& {Arce}(2004)}]{goodman04}

1379: {Goodman}, A.~A., \& {Arce}, H.~G. 2004, \apj, 608, 831

1380: 

1381: \bibitem[{{Gordon} {et~al.}(2005){Gordon}, {Rieke}, {Engelbracht}, {Muzerolle},

1382:   {Stansberry}, {Misselt}, {Morrison}, {Cadien}, {Young}, {Dole}, {Kelly},

1383:   {Alonso-Herrero}, {Egami}, {Su}, {Papovich}, {Smith}, {Hines}, {Rieke},

1384:   {Blaylock}, {P{\'e}rez-Gonz{\'a}lez}, {Le Floc'h}, {Hinz}, {Latter},

1385:   {Hesselroth}, {Frayer}, {Noriega-Crespo}, {Masci}, {Padgett}, {Smylie}, \&

1386:   {Haegel}}]{gordon05}

1387: {Gordon}, K.~D., et~al.\ 2005, \pasp, 117, 503

1388: 

1389: \bibitem[{{Gower} \& {Ross}(1969)}]{gower69}

1390: {Gower}, J.~F.~R., \& {Ross}, G.~J.~S. 1969, Appl. Stat., 18, 54

1391: 

1392: \bibitem[{{Gutermuth} {et~al.}(2004){Gutermuth}, {Megeath}, {Muzerolle},

1393:   {Allen}, {Pipher}, {Myers}, \& {Fazio}}]{gutermuth04}

1394: {Gutermuth}, R.~A., {Megeath}, S.~T., {Muzerolle}, J., {Allen}, L.~E.,

1395:   {Pipher}, J.~L., {Myers}, P.~C., \& {Fazio}, G.~G. 2004, \apjs, 154, 374

1396: 

1397: \bibitem[{{Gutermuth} {et~al.}(2005){Gutermuth}, {Megeath}, {Pipher},

1398:   {Williams}, {Allen}, {Myers}, \& {Raines}}]{gutermuth05}

1399: {Gutermuth}, R.~A., {Megeath}, S.~T., {Pipher}, J.~L., {Williams}, J.~P.,

1400:   {Allen}, L.~E., {Myers}, P.~C., \& {Raines}, S.~N. 2005, \apj, 632, 397

1401: 

1402: \bibitem[{{Gutermuth} {et~al.}(2008){Gutermuth}, {Myers}, {Megeath}, {Allen},

1403:   {Pipher}, {Muzerolle}, {Porras}, {Winston}, \& {Fazio}}]{gutermuth08}

1404: {Gutermuth}, R.~A., et~al.\ 2008,  \apj, 674, 336

1405: 

1406: \bibitem[{{Hartmann} {et~al.}(2005){Hartmann}, {Megeath}, {Allen}, {Luhman},

1407:   {Calvet}, {D'Alessio}, {Franco-Hernandez}, \& {Fazio}}]{hartmann05}

1408: {Hartmann}, L., {Megeath}, S.~T., {Allen}, L., {Luhman}, K., {Calvet}, N.,

1409:   {D'Alessio}, P., {Franco-Hernandez}, R., \& {Fazio}, G. 2005, \apj, 629, 881

1410: 

1411: \bibitem[{{Harvey} {et~al.}(2008){Harvey}, {Huard}, {J{\o}rgensen}, {Gutermuth}, {Mamajek}, {Bourke}, {Mer{\'{\i}}n}, {Cieza}, {Brooke}, {Chapman}, {Alcal{\'a}}, {Allen}, {Evans}, {Di Francesco}, \& {Kirk}}]{harvey08}

1412: {Harvey}, P.~M., et~al.\ 2008, \apj, 680, 495

1413: 

1414: \bibitem[{{Hillwig} {et~al.}(2006){Hillwig}, {Gies}, {Bagnuolo}, {Huang},

1415:   {McSwain}, \& {Wingert}}]{hillwig06}

1416: {Hillwig}, T.~C., {Gies}, D.~R., {Bagnuolo}, Jr., W.~G., {Huang}, W.,

1417:   {McSwain}, M.~V., \& {Wingert}, D.~W. 2006, \apj, 639, 1069

1418: 

1419: \bibitem[{{Jungwiert} \& {Palous}(1994)}]{jungwiert94}

1420: {Jungwiert}, B., \& {Palous}, J. 1994, \aap, 287, 55

1421: 

1422: \bibitem[{{Karr} \& {Martin}(2003)}]{karr03}

1423: {Karr}, J.~L., \& {Martin}, P.~G. 2003, \apj, 595, 900

1424: 

1425: \bibitem[{{Klein} {et~al.}(1985){Klein}, {Whitaker}, \& {Sandford}}]{klein85}

1426: {Klein}, R.~I., {Whitaker}, R.~W., \& {Sandford}, II, M.~T. 1985, in Protostars

1427:   and Planets II, ed. D.~C. {Black} \& M.~S. {Matthews}, 340--+

1428: 

1429: \bibitem[{{Lada}(1987)}]{lada87}

1430: {Lada}, C.~J. 1987, in IAU Symposium, Vol. 115, Star Forming Regions, ed.

1431:   M.~{Peimbert} \& J.~{Jugaku}, 1--17

1432: 

1433: \bibitem[{{Lada} {et~al.}(1978){Lada}, {Elmegreen}, {Cong}, \&

1434:   {Thaddeus}}]{lada78}

1435: {Lada}, C.~J., {Elmegreen}, B.~G., {Cong}, H.-I., \& {Thaddeus}, P. 1978,

1436:   \apjl, 226, L39

1437: 

1438: \bibitem[{{Lada} {et~al.}(1994){Lada}, {Lada}, {Clemens}, \& {Bally}}]{lada94}

1439: {Lada}, C.~J., {Lada}, E.~A., {Clemens}, D.~P., \& {Bally}, J. 1994, ApJ, 429, 694

1440: 

1441: \bibitem[{{Lada} {et~al.}(2006){Lada}, {Muench}, {Luhman}, {Allen}, {Hartmann}, {Megeath}, {Myers}, {Fazio}, {Wood}, {Muzerolle}, {Rieke}, {Siegler}, \& {Young}}]{lada06}

1442: {Lada}, C.~J., et~al.\ 2006, \aj, 131, 1574

1443: 

1444: \bibitem[{{Landsman}(1993)}]{landsman93}

1445: {Landsman}, W.~B. 1993, in ASP Conf. Ser. 52: Astronomical Data Analysis

1446:   Software and Systems II, ed. R.~J. {Hanisch}, R.~J.~V. {Brissenden}, \&

1447:   J.~{Barnes}, 246--+

1448: 

1449: \bibitem[{{Lefloch} \& {Lazareff}(1994)}]{lefloch94}

1450: {Lefloch}, B., \& {Lazareff}, B. 1994, \aap, 289, 559

1451: 

1452: \bibitem[{{Lombardi} \& {Alves}(2001)}]{lombardi01} 

1453: {Lombardi}, M., \&  Alves, J. 2001, \aap, 377, 1023

1454: %\bibitem[{{Lutz} {et~al.}(1998){Lutz}, {Kunze}, {Spoon}, \&

1455: %  {Thornley}}]{lutz98}

1456: %{Lutz}, D., {Kunze}, D., {Spoon}, H.~W.~W., \& {Thornley}, M.~D. 1998, \aap,

1457: %  333, L75

1458: 

1459: \bibitem[{{Martins} {et~al.}(2005){Martins}, {Schaerer}, \& {Hillier}}]{martins05}

1460: {Martins}, F., {Schaerer}, D., \& {Hillier}, D.~J. 2005, \aap, 436, 1049

1461: 

1462: \bibitem[{{McKee} {et~al.}(1984){McKee}, {Van Buren}, \& {Lazareff}}]{mckee84}

1463: {McKee}, C.~F., {Van Buren}, D., \& {Lazareff}, B. 1984, \apjl, 278, L115

1464: 

1465: \bibitem[{Megeath} {et~al.}(2004){Megeath}, {Allen}, {Gutermuth},  {Pipher}, 

1466: {Myers}, {Calvet}, {Hartmann}, {Muzerolle}, \& {Fazio}]{megeath04}

1467: {Megeath}, S.~T., et~al.\ 2004, \apjs, 154, 367

1468: 

1469: \bibitem[{{Moffat}(1972)}]{moffat72}

1470: {Moffat}, A.~F.~J. 1972, \aaps, 7, 355

1471: 

1472: \bibitem[{{Moffat} {et~al.}(1998){Moffat}, {Marchenko}, {Seggewiss}, {van der

1473:   Hucht}, {Schrijver}, {Stenholm}, {Lundstrom}, {Setia Gunawan}, {Sutantyo},

1474:   {van den Heuvel}, {de Cuyper}, \& {Gomez}}]{moffat98}

1475: {Moffat}, A.~F.~J., et~al.\ 1998, \aap, 331, 949

1476: 

1477: \bibitem[{{Motte} {et~al.}(2007){Motte}, {Bontemps}, {Schilke}, {Schneider}, {Menten}, {Brogui{\`e}re}}]{motte07}

1478: {Motte}, F., {Bontemps}, S., {Schilke}, P., {Schneider}, N., {Menten}, K.~M., \& {Brogui{\`e}re}, D. 2007, \aap, 476, 1243

1479: 

1480: \bibitem[{{Muzerolle} {et~al.}(2004){Muzerolle}, {Megeath}, {Gutermuth},

1481:   {Allen}, {Pipher}, {Hartmann}, {Gordon}, {Padgett}, {Noriega-Crespo},

1482:   {Myers}, {Fazio}, {Rieke}, {Young}, {Morrison}, {Hines}, {Su}, {Engelbracht},

1483:   \& {Misselt}}]{muzerolle04}

1484: {Muzerolle}, J., et~al.\ 2004, \apjs, 154, 379

1485: 

1486: \bibitem[{{Najita}, {Strom} \& {Muzerolle}(2007)}]{najita07}

1487: {Najita}, J.~R., {Strom}, S.~E., \& {Muzerolle}, J. 2007, \mnras, 378, 369 

1488: 

1489: \bibitem[{{Nakano} {et~al.}(2008){Nakano}, {Sugitani}, {Niwa}, {Itoh}, \& {Watanabe}}]{nakano08}

1490: {Nakano}, M., {Sugitani}, K., {Niwa}, T., {Itoh}, Y., \& {Watanabe}, M. 2008, {\it astro-ph} 0208.2624

1491: 

1492: \bibitem[{{Normandeau} {et~al.}(1996){Normandeau}, {Taylor}, \&

1493:   {Dewdney}}]{normandeau96}

1494: {Normandeau}, M., {Taylor}, A.~R., \& {Dewdney}, P.~E. 1996, \nat, 380, 687

1495: 

1496: \bibitem[{{Poveda} {et~al.}(1967){Poveda}, {Ruiz}, \& {Allen}}]{poveda67}

1497: {Poveda}, A., {Ruiz}, J., \& {Allen}, C. 1967, Bol. Obs. Ton. y Tac., 4, 86

1498: 

1499: \bibitem[{{Price} \& {Murdock}(1983)}]{price83}

1500: {Price}, S.~D., \& {Murdock}, T.~L. 1983, NASA STI/Recon Technical Report N,

1501:   84, 15005

1502: 

1503: \bibitem[{{Reipurth} (2000)}]{reipurth00}

1504: {Reipurth}, B. 2000, \aj, 120, 3177

1505: 

1506: \bibitem[{{Rieke} {et~al.}(2004){Rieke}, {Young}, {Engelbracht}, {Kelly},

1507:   {Low}, {Haller}, {Beeman}, {Gordon}, {Stansberry}, {Misselt}, {Cadien},

1508:   {Morrison}, {Rivlis}, {Latter}, {Noriega-Crespo}, {Padgett}, {Stapelfeldt},

1509:   {Hines}, {Egami}, {Muzerolle}, {Alonso-Herrero}, {Blaylock}, {Dole}, {Hinz},

1510:   {Le Floc'h}, {Papovich}, {P{\'e}rez-Gonz{\'a}lez}, {Smith}, {Su}, {Bennett},

1511:   {Frayer}, {Henderson}, {Lu}, {Masci}, {Pesenson}, {Rebull}, {Rho}, {Keene},

1512:   {Stolovy}, {Wachter}, {Wheaton}, {Werner}, \& {Richards}}]{rieke04}

1513: {Rieke}, G.~H., et~al.\ 2004, \apjs, 154, 25

1514: 

1515: \bibitem[{{Schneider} {et~al.}(2006){Schneider}, {Bontemps}, {Simon}, {Jakob}, {Motte}, {Miller}, {Kramer}, \& {Stutzki}}]{schneider06}

1516: {Schneider}, N., {Bontemps}, S., {Simon}, R., {Jakob}, H., {Motte}, F., {Miller}, M., {Kramer}, C., \& {Stutzki}, J. 2006, \aap, 458, 855

1517: 

1518: \bibitem[{{Seiden} \& {Schulman}(1990)}]{seiden90}

1519: {Seiden}, P.~E., \& {Schulman}, L.~S. 1990, Advances in Physics, 39, 1

1520: 

1521: \bibitem[{{Skrutskie} {et~al.}(2006){Skrutskie}, {Cutri}, {Stiening},

1522:   {Weinberg}, {Schneider}, {Carpenter}, {Beichman}, {Capps}, {Chester},

1523:   {Elias}, {Huchra}, {Liebert}, {Lonsdale}, {Monet}, {Price}, {Seitzer},

1524:   {Jarrett}, {Kirkpatrick}, {Gizis}, {Howard}, {Evans}, {Fowler}, {Fullmer},

1525:   {Hurt}, {Light}, {Kopan}, {Marsh}, {McCallon}, {Tam}, {Van Dyk}, \&

1526:   {Wheelock}}]{skrutskie06}

1527: {Skrutskie}, M.~F., et~al.\ 2006, \aj, 131, 1163

1528: 

1529: \bibitem[{{Smith} {et~al.}(2006){Smith}, {Hora}, {Marengo}, \&

1530:   {Pipher}}]{smith06}

1531: {Smith}, H.~A., {Hora}, J.~L., {Marengo}, M., \& {Pipher}, J.~L. 2006, \apj,

1532:   645, 1264

1533: 

1534: %\bibitem[{{Smith} {et~al.}(2002){Smith}, {Norris}, \& {Crowther}}]{smith02}

1535: %{Smith}, L.~J., {Norris}, R.~P.~F., \& {Crowther}, P.~A. 2002, \mnras, 337, 1309

1536: 

1537: \bibitem[{{Smith} \& {Brooks}(2007)}]{smith07}

1538: {Smith}, N., \& {Brooks}, K.~J. 2007, \mnras, 379, 1279

1539: 

1540: \bibitem[{{Stahler} \& {Palla}(2005)}]{stahler05}

1541: {Stahler}, S.~W., \& {Palla}, F. 2005, {The Formation of Stars} (The Formation

1542:   of Stars, by Steven W.~Stahler, Francesco Palla, pp.~865.~ISBN

1543:   3-527-40559-3.~Wiley-VCH , January 2005)

1544: 

1545: \bibitem[{{Stern} {et~al.}(2005){Stern}, {Eisenhardt}, {Gorjian}, {Kochanek},

1546:   {Caldwell}, {Eisenstein}, {Brodwin}, {Brown}, {Cool}, {Dey}, {Green},

1547:   {Jannuzi}, {Murray}, {Pahre}, \& {Willner}}]{stern05}

1548: {Stern}, D., et~al.\ 2005, \apj, 631, 163

1549: 

1550: \bibitem[{{Sterzik} \& {Durisen}(1995)}]{sterzik95}

1551: {Sterzik}, M.~F., \& {Durisen}, R.~H. 1995, \aap, 304, L9

1552: 

1553: \bibitem[{{Stetson}(1987)}]{stetson87}

1554: {Stetson}, P.~B. 1987, \pasp, 99, 191

1555: 

1556: \bibitem[{{Stutzki} {et~al.}(1988){Stutzki}, {Stacey}, {Genzel}, {Harris},

1557:   {Jaffe}, \& {Lugten}}]{stutzki88}

1558: {Stutzki}, J., {Stacey}, G.~J., {Genzel}, R., {Harris}, A.~I., {Jaffe}, D.~T.,

1559:   \& {Lugten}, J.~B. 1988, \apj, 332, 379

1560: 

1561: \bibitem[{{Sugitani} {et~al.}(1991){Sugitani}, {Fukui}, \&

1562:   {Ogura}}]{sugitani91}

1563: {Sugitani}, K., {Fukui}, Y., \& {Ogura}, K. 1991, \apjs, 77, 59

1564: 

1565: \bibitem[{{Thompson} {et~al.}(2004){Thompson}, {White}, {Morgan}, {Miao},

1566:   {Fridlund}, \& {Huldtgren-White}}]{thompson04}

1567: {Thompson}, M.~A., {White}, G.~J., {Morgan}, L.~K., {Miao}, J., {Fridlund},

1568:   C.~V.~M., \& {Huldtgren-White}, M. 2004, \aap, 414, 1017

1569: 

1570: %\bibitem[{{Thornley} {et~al.}(2000){Thornley}, {Schreiber}, {Lutz}, {Genzel}, R. {Spoon}, {Kunze}, \& {Sternberg}}]{thornley00}

1571: %{Thornley}, M.~D., {Schreiber}, N.~M.~F., {Lutz}, D., {Genzel}, R., {Spoon}, H.~W.~W., {Kunze}, D., \& {Sternberg}, A. 2000, \apj, 539, 641

1572: 

1573: \bibitem[{{Vall{\'e}e} {et~al.}(1979){Vall{\'e}e}, {Hughes}, \&

1574:   {Viner}}]{vallee79}

1575: {Vall{\'e}e}, J.~P., {Hughes}, V.~A., \& {Viner}, M.~R. 1979, \aap, 80, 186

1576: 

1577: \bibitem[{{Weaver} {et~al.}(1977){Weaver}, {McCray}, {Castor}, {Shapiro}, \& {Moore}}]{weaver77}

1578: {Weaver}, R., {McCray}, R., {Castor}, J., {Shapiro}, P., \& {Moore}, R. 1977, \apj, 218, 377

1579: 

1580: \bibitem[{{Westerhout}(1958)}]{westerhout58}

1581: {Westerhout}, G. 1958, \bain, 14, 215

1582: 

1583: \bibitem[{{Whitney} {et~al.}(2003){Whitney}, {Wood}, {Bjorkman}, \&

1584:   {Cohen}}]{whitney03}

1585: {Whitney}, B.~A., {Wood}, K., {Bjorkman}, J.~E., \& {Cohen}, M. 2003, \apj,

1586:   598, 1079

1587: 

1588: \bibitem[{{Whitworth} {et~al.}(1994){Whitworth}, {Bhattal}, {Chapman},

1589:   {Disney}, \& {Turner}}]{whitworth94}

1590: {Whitworth}, A.~P., {Bhattal}, A.~S., {Chapman}, S.~J., {Disney}, M.~J., \&

1591:   {Turner}, J.~A. 1994, \mnras, 268, 291

1592: 

1593: \bibitem[{{Wilking} {et~al.}(1984){Wilking}, {Doering}, {Harvey}, {Lada}, \&

1594:   {Joy}}]{wilking84}

1595: {Wilking}, B.~A., {Doering}, C.~R., {Harvey}, P.~M., {Lada}, C.~J., \& {Joy},

1596:   M. 1984, \apj, 279, 291

1597: 

1598: \bibitem[{{Xu} {et~al.}(2006){Xu}, {Reid}, {Zheng}, \& {Menten}}]{xu06}

1599: {Xu}, Y., {Reid}, M.~J., {Zheng}, X.~W., \& {Menten}, K.~M. 2006, Science, 311,

1600:   54

1601: 

1602: \bibitem[{{Zavagno} {et~al.}(1999){Zavagno}, {Lagage}, \& {Cabrit}}]{zavagno99}

1603: {Zavagno}, A., {Lagage}, P.~O., \& {Cabrit}, S. 1999, \aap, 344, 499

1604: 

1605: \bibitem[{{Zinnecker} \& {Yorke}(2007)}]{zinnecker07}

1606: {Zinnecker}, H., \& {Yorke}, H.~W. 2007, \araa, 45, 481

1607: 

1608: \end{thebibliography}

1609: 

1610: \end{document}

1611: