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

2: \documentclass{emulateapj}

3: \usepackage{times}

4: %\usepackage[figuresright]{rotating}

5: 

6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

7: % AUTHOR'S MACROS

8: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

9: \newcommand{\beq}{\begin{equation}}

10: \newcommand{\eeq}{\end{equation}}

11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

12: % AUTHOR'S SYMBOLS

13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

14: \newcommand{\gapprox}{$\stackrel {>}{_{\sim}}$}

15: \newcommand{\lapprox}{$\stackrel {<}{_{\sim}}$}

16: \def\alp{\mbox{$\alpha$}}

17: \def\farcd{\hbox{$.\mkern-4mu^\circ$}}

18: \def\farcm{\hbox{$.\mkern-4mu^\prime$}}

19: \def\farcs{\hbox{$.\!\!^{\prime\prime}$}}

20: \def\earth{\hbox{$\oplus$}}

21: \def\arcmin{\hbox{$^\prime$}}

22: \def\arcsec{\hbox{$^{\prime\prime}$}}

23: \def\solar{\mbox{$_{\normalsize\odot}$}}

24: \def\fs{\hbox{$.\!\!^{\rm s}$}}

25: \def\deg{\hbox{$^\circ$}}

26: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

27: \newcommand{\ttbs}{\char'134}

28: \newcommand{\AmS}{{\protect\the\textfont2

29:   A\kern-.1667em\lower.5ex\hbox{M}\kern-.125emS}}

30: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

31: \newcommand{\lsim}{\ \raise

32: -2.truept\hbox{\rlap{\hbox{$\sim$}}\raise5.truept\hbox{$<$}\ }}

33: \newcommand{\gsim}{\ \raise

34: -2.truept\hbox{\rlap{\hbox{$\sim$}}\raise5.truept\hbox{$>$}\ }}

35: \newcommand{\simsim}{\ \raise

36: -2.truept\hbox{\rlap{\hbox{$\sim$}}\raise5.truept\hbox{$\sim$}\ }}

37: \newcommand{\FeH}{\mbox{[Fe/H]}\,}

38: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

39: 

40: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

41: % add words to TeX's hyphenation exception list

42: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

43: \hyphenation{author another re-commend-ed Post-Script}

44: 

45: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

46: %% You can insert a short comment on the title page using the command

47: %% below.

48: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

49: %\slugcomment{Accepted for Publication in ApJL October 5, 2007}

50: \slugcomment{Accepted for Publication in the Astrophysical Journal}

51: 

52: %\righthead{}

53: %\lefthead{Gouliermis D., et al.}

54: 

55: \shorttitle{Triggered Star Formation in NGC~346/N~66 in the SMC}

56: \shortauthors{D. A. Gouliermis et al.}

57: 

58: \begin{document}

59: 

60: \title{NGC~346 in The Small Magellanic Cloud. IV.  Triggered Star

61: Formation in the {\sc H~ii} Region N~66}

62: 

63: %\altaffilmark{1}}

64: 

65: %\footnotetext[]{Research supported by the Deutsche Forschungsgemeinschaft 

66: %(German Research Foundation)}

67: 

68: %\author{Dimitrios A. Gouliermis, Thomas Henning, Eva Hennekemper, 

69: %Wolfgang Brandner, and Felix Hormuth}

70: %\affil{Max Planck Institute for Astronomy, K\"onigstuhl 17, 69117 

71: %Heidelberg, Germany}

72: 

73: %\email{dgoulier@mpia.de; henning@mpia.de; hennekemper@mpia.de; 

74: %brandner@mpia.de; hormuth@mpia.de}

75: 

76: \author{Dimitrios A. Gouliermis\altaffilmark{1}, 

77:         You-Hua Chu\altaffilmark{2}, 

78:         Thomas Henning\altaffilmark{1},\\

79:         Wolfgang Brandner\altaffilmark{1}, 

80:         Robert A. Gruendl\altaffilmark{2}, 

81:         Eva Hennekemper\altaffilmark{1}, and 

82:         Felix Hormuth\altaffilmark{1}

83: }

84: 

85: \email{dgoulier@mpia.de, chu@astro.uiuc.edu, henning@mpia.de,

86: brandner@mpia.de, gruendl@astro.uiuc.edu, hennekemper@mpia.de,

87: hormuth@mpia.de}

88: 

89: %\altaffiltext{1}{Research supported by the Deutsche 

90: %Forschungsgemeinschaft (German Research Foundation)}

91: \altaffiltext{1}{Max Planck Institute for Astronomy, K\"onigstuhl 17,

92: 69117 Heidelberg, Germany}

93: \altaffiltext{2}{Department of Astronomy, University of Illinois, 1002

94: West Green Street, Urbana, IL 61801}

95: 

96: %\altaffiltext{1}{Research supported by the Deutsche

97: %Forschungsgemeinschaft (German Research Foundation)}

98: %\author{Dimitrios A. Gouliermis}

99: %\affil{Max Planck Institute for Astronomy, K\"onigstuhl 17, 69117

100: %Heidelberg, Germany; dgoulier@mpia.de}

101: %\author{You-Hua Chu}

102: %\affil{Department of Astronomy, University of Illinois, 1002

103: %West Green Street, Urbana, IL 61801; chu@astro.uiuc.edu}

104: %\and

105: %\author{Thomas Henning, Wolfgang Brandner, Eva Hennekemper, Felix

106: %Hormuth} \affil{Max Planck Institute for Astronomy, K\"onigstuhl 17,

107: %69117 Heidelberg, Germany;\\ [henning], [brandner], [hennekemper],

108: %[hormuth]@mpia.de}

109: 

110: 

111: \begin{abstract}

112: 

113: Stellar feedback, expanding {\sc H~ii} regions, wind-blown bubbles, and

114: supernovae are thought to be important triggering mechanisms of star

115: formation.  Stellar associations, being hosts of significant numbers of

116: early-type stars, are the loci where these mechanisms act. In this part

117: of our photometric study of the star-forming region NGC~346/N~66 in the

118: Small Magellanic Cloud, we present evidence based on previous and recent

119: detailed studies, that it hosts at least two different events of

120: triggered star formation and we reveal the complexity of its recent star

121: formation history.  In our earlier studies of this region (Papers I,

122: III) we find that besides the central part of N~66, where the bright OB

123: stellar content of the association NGC~346 is concentrated, an arc-like

124: nebular feature, north of the association, hosts recent star formation.

125: This feature is characterized by a high concentration of emission-line

126: stars and Young Stellar Objects, as well as embedded sources seen as

127: IR-emission peaks that coincide with young compact clusters of low-mass

128: pre-main sequence stars. All these objects indicate that the northern

129: arc of N~66 encompasses the most current star formation event in the

130: region. We present evidence that this star formation is the product of a

131: different mechanism than that in the general area of the association,

132: and that it is triggered by a wind-driven expanding {\sc H~ii} region

133: (or bubble) blown by a massive supernova progenitor, and possibly other

134: bright stars, a few Myr ago. We propose a scenario according to which

135: this mechanism triggered star formation away from the bar of N~66, while

136: in the bar of N~66 star formation is introduced by the photo-ionizing OB

137: stars of the association itself.

138: 

139: \end{abstract}

140: 

141: \keywords{Magellanic Clouds --- stars: winds, outflows --- ISM: bubbles 

142: --- HII regions --- supernova remnants --- clusters: individual (NGC~346)}

143: 

144: \section{Introduction}

145: 

146: Massive OB stars, not having an optically visible pre-main sequence

147: (PMS) contraction phase, appear almost immediately after their birth on

148: the main sequence (e.g. Stahler \& Palla 2005). They are mostly grouped

149: in stellar associations, loose concentrations of stars, which host also

150: significant numbers of intermediate- and low-mass PMS stars (see review

151: by Brice\~{n}o et al. 2007). When the far-UV radiation of the bright OB

152: stars reaches the surface of the parental molecular cloud, a

153: photo-dissociated region (PDR) develops. Nebula LHA~115-N~66 or in short

154: N~66 (Henize 1956), the brightest {\sc H~ii} region in the Small

155: Magellanic Cloud (SMC), being very rich in early-type stars, is

156: certainly an excellent example of an extragalactic PDR. The stellar

157: association NGC~346, located in the central part of the nebula, hosts

158: the largest sample of spectroscopically confirmed OB stars in the SMC

159: (Massey et al. 1989), which have been the subject of several previous

160: investigations (Niemela et al.  1986; Walborn et al. 2000; Evans et al. 

161: 2006).

162: 

163: These massive stars evolve rapidly, and immediately start to ionize the

164: cloud, blowing its material away. The degree of ionization decreases

165: outwards, and a thin barrier develops that segregates the ionized from

166: the atomic gas. Molecular hydrogen is not fully ionized behind this

167: front, but partly dissociated, while CO molecules which are located a

168: bit deeper into the cloud are more easily dissociated than H$_{2}$ by

169: absorbing UV photons. A correlation between H$_{2}$ infrared emission

170: and CO lines, characteristic of a PDR, has been found for the region of

171: NGC~346/N~66 by Contursi et al. (2000) and Rubio et al. (2000). These

172: authors define the ``bar'' of N~66 as the oblique bright emission

173: region, extending from southeast to northwest centered on NGC~346. They

174: suggest that {\sl star formation has taken place as a sequential process

175: in the bar of N~66}, resulting in several embedded sources, seen as

176: IR-emission peaks in the 2.14~\micron\ H$_{2}$ line and the ISOCAM LW2

177: band (5 - 8 \micron). These peaks are alphabetically numbered from ``A''

178: to ``I'' (Contursi et al. 2000), with the association NGC~346 itself

179: coinciding with peak ``C''.

180: 

181: Recent studies reconstruct the star formation history in nearby Galactic

182: OB associations, and provide observational evidence for sequential and

183: triggered star formation in their vicinity. Preibisch \& Zinnecker

184: (2007) conclude from their study of the Scorpius-Centaurus OB

185: association that the formation of whole OB subgroups (each consisting of

186: several thousand stars) requires large-scale triggering mechanisms such

187: as shocks from expanding wind and supernova driven super-bubbles

188: surrounding older subgroups\footnote{According to these authors, other

189: triggering mechanisms, like radiatively driven implosion of globules,

190: also operate, but seem to be secondary processes, forming only small

191: stellar groups rather than whole OB subgroups with thousands of stars.}.

192: Since the low-mass stellar members of associations remain in their PMS

193: phase for \lsim~30~Myr, they play a key r\^{o}le in the understanding of

194: star formation in the vicinity of these systems (Brice\~{n}o et al.

195: 2007). Consequently, the recent discovery of a plethora of low-mass PMS

196: stars in the region of NGC~346/N~66 with photometry from the {\sl 

197: Advanced Camera for Surveys} (ACS) onboard HST (Nota et al. 2006;

198: Gouliermis et al. 2006, hereafter Paper~I) can contribute significantly

199: to the clarification of the mechanisms that may act in this

200: extraordinary star-forming region.

201: 

202: The subsequent investigation of these PMS stars in NGC~346/N~66 (Sabbi et 

203: al. 2007, hereafter SSN07; Hennekemper et al. 2008, hereafter Paper~III) 

204: demonstrated that indeed they are clustered in several compact 

205: concentrations, some of them coinciding with the IR-emission peaks of 

206: Contursi et al. (2000; detected also with {\sl Spitzer}, see \S 2.1), 

207: verifying the existence of stellar subgroups in the region, similarly to 

208: galactic associations. Not being able to resolve differences in age 

209: smaller than 1 - 2 Myr in the individual CMDs of these sub-clusters, SSN07 

210: suggest that ``all sub-clusters appear coeval with each other''. According 

211: to these authors, this coevality is a signature of the star formation 

212: conditions predicted by the hierarchical fragmentation of a turbulent 

213: molecular cloud model (Klessen \& Burkert 2000; Bonnell \& Bate 2002; 

214: Bonnell et al. 2003).

215: 

216: However, our analysis on the clustering properties of the PMS stars in

217: NGC~346/N~66 (Paper~III) showed that {\sl there are significant age

218: differences between some sub-clusters and the association itself}. 

219: Specifically, three compact PMS clusters located to the north of the bar

220: of N~66 are found to be not older than 2.5~Myr, while the CMD of the

221: main body of the association NGC~346 show indications of an underlying

222: older PMS population. This clearly suggests that these northern clusters

223: are probably formed {\sl after} the central stellar association. As we

224: discuss later, triggering mechanisms for star formation such as those

225: described by models of ionization shock fronts from OB stars (e.g.

226: Kessel-Deynet \& Burkert 2003) or wind-driven shock waves (e.g. Vanhala

227: \& Cameron 1998) may explain better their formation. Indeed, while a

228: sequential star formation mechanism has been suggested to take place

229: {\sl in} the bar of N~66, around the association (Rubio et al. 2000),

230: the existence of young compact PMS clusters {\sl away} from it to the

231: north of the association, does not quite fit in the hypothesis that this

232: mechanism propagated from the center of N~66 along the bar. The

233: triggering agent of these clusters may well be located outside of the

234: bar.

235: 

236: In this paper we consider the findings from earlier and recent

237: comprehensive investigations of the region NGC~346/N~66 to attempt a

238: clearer understanding of the mechanisms that shape the recent star

239: formation of this outstanding extragalactic star-forming region. Our aim

240: is to provide answers to two important questions: i) {\sl Is the star

241: formation away from the association NGC~346 the product of triggered

242: fragmentation of the cloud alone?} ii) {\sl Is the photo-dissociation by

243: the early-type stars of NGC~346 in the center of N~66 the only

244: triggering mechanism in the region?} In section~2 we present evidence

245: that this region has a far more complicated recent star formation

246: history than what was previously considered, and that the most recent

247: star formation event could not have been triggered by the central

248: association. In section~3 we propose a scenario for the recent star

249: formation in the northern part of the region NGC~346/N~66 and we

250: indicate a nearby massive supernova progenitor, located at the northeast

251: of the bar of N~66, as the triggering agent. We also discuss the

252: possibility that other nearby massive stars away from the association

253: may have also contributed to this mechanism. We present supporting

254: evidence to this scenario using analyses of massive stellar content and

255: gas kinematics of this region. Concluding remarks are given in

256: section~4.

257: 

258: %Hierarchical fragmentation of a turbulent molecular cloud:

259: %----------------------------------------------------------

260: %According to this model, the fragmentation of the cloud is due to

261: %supersonic turbulent motions present in the gas. The turbulence induces

262: %the formation of shocks in the gas and produce filamentary structures

263: %(Bate et al. 2003). The chaotic nature of the turbulence increases the

264: %density in the filamentary structures locally. When regions of high

265: %density become self-gravitating, they start to collapse to form stars.

266: %Simulations show that SF occurs simultaneously at several different

267: %location in the cloud (Bonnell et al. 2003), as appears to be the case

268: %for NGC 346.

269: 

270: \section{Star Formation in the region of NGC~346/N~66}

271: 

272: \subsection{Star Formation away from the bar of N~66}

273: 

274: Recent {\sl Spitzer} IR observations of NGC~346/N~66 revealed a rich

275: sample of embedded sources, identified as candidate Young Stellar

276: Objects (YSOs; Bolatto et al. 2007; Simon et al. 2007). A large

277: concentration of these sources are located in the bar of N~66, in

278: agreement with the distribution of the PMS stars found with our ACS

279: photometry (Paper~III), and they coincide with the emission peaks of

280: this area identified by Rubio et al. (2000; peaks A, B, C, D, E, H and

281: I; see also their Fig. 8).  However, another important concentration of

282: YSOs is located to the north of the bar, coinciding with two additional

283: emission peaks (peaks F and G) and with a second concentration of PMS

284: stars, which forms an arc-like feature, extending from southwest to

285: northeast. The ACS image of the region in H\alp\ (Paper~III), the

286: 8\micron\ {\sl Spitzer} image (Simon et al. 2007), as well as the maps

287: of the CO(2$-$1) line emission (Rubio et al. 2000) give further support

288: for the existence of this feature\footnote{The description of this

289: feature as an ``arc-like'' is based on the gas emission in the region

290: seen in H\alp\ and [{\sc O~iii}] mostly from its southern part. In CO

291: and 8\micron-emission, which dominates its northern part the morphology

292: of this feature is more linear, but still curved, and closer to

293: ``arm-like''.}. Three compact clusters of PMS stars are identified by us

294: (Paper~III) and SSN07 in this arc-like feature (clusters 2a, 2b, and 3

295: in Paper~III, and Sc-13, 14 and 15 in SSN07) to coincide with both

296: IR-emission peaks. The PMS stars in these clusters are found to be

297: younger than those of the bar, implying that they are more recently

298: formed. Moreover, this northern feature is also characterized by a high

299: concentration of emission-line stars found with our ACS photometry in

300: H{\alp} (Paper~III). These facts clearly suggest that indeed star

301: formation currently takes place in an extended feature away from the bar

302: of N~66, northeast of NGC~346.

303: 

304: This feature can be easily distinguished in the color-composite image 

305: shown in the left panel of Fig.~\ref{fig-imaxoi}. We constructed this 

306: image from observations of the general area around NGC~346/N~66 with 

307: XMM-Newton in X-rays (blue), ESO NTT in [{\sc O~iii}] (green) and {\sl 

308: Spitzer}/IRAC in the 8\micron\ band (red). The XMM-Newton observations 

309: have ID 0110000201 (PI: J. Bleeker), and they have been used in 

310: variability studies of X-ray sources (Naz\'{e} et al. 2004) and in the 

311: investigation of high mass X-ray binaries (Sasaki, Pietsch \& Haberl 2003; 

312: Shtykovskiy \& Gilfanov 2005) in the region of NGC~346/N~66. The [{\sc 

313: O~iii}] (501.1 nm) image of NGC~346/N~66 was taken with NTT/EMMI within 

314: the ESO Program 56.C-0379 (PI: I. J. Danziger) and it was also presented 

315: in the analysis by Rubio et al. (2000).  The 8\micron\ imaging data have 

316: been obtained with the Infrared Array Camera (IRAC; Fazio et al. 2004) 

317: onboard the {\sl Spitzer Space Telescope} within the GTO Science Program 

318: 63 (PI: J. R. Houck) on 2004 April 20, and it has been used for the 

319: detection of candidate YSOs in the region (Simon et al. 2007).

320: 

321: The association NGC~346 is located within the bar of N~66, almost at the

322: center of the image of Fig.~\ref{fig-imaxoi} (left panel). From this

323: image, as well as from the ones taken with {\sl HST}/ACS (see Nota et

324: al. 2006) and {\sl Spitzer}/IRAC (see Simon et al. 2007) it can be seen

325: that there is a relation between NGC~346 and a southern dusty arc

326: feature. This one seems to outline the ionization front of the remaining

327: cloud, due to the powerful winds of the OB stars of the association.

328: However, the shape and orientation of the star-forming feature away and to

329: the north of NGC~346, located at the top left part of the image,

330: suggests that it is triggered probably by the same kind of mechanism,

331: but this process {\sl could not be produced} by the photo-dissociation

332: of the cloud by the ionizing stars of the association.  If it was so,

333: this feature would not face to the east, but rather to the south in a

334: symmetrical manner to the southern arc. Since, this is not the case, the

335: source that triggered star formation in the northern extended feature

336: {\sl should not be the central association, but it should be rather

337: located to the east of the field}.

338: 

339: %Therefore, second generation star formation was not just triggered by

340: %the ionizing radiation of the OB stars in NGC~346.

341: 

342: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

343: %\clearpage

344: \begin{figure*}[t!]

345: \epsscale{1.15}

346: \plottwo{f1a.eps}{f1b.eps}

347: %\includegraphics[scale=0.475]{f1.eps}

348: \caption{{\sl Left}: Color-composite image of the area of NGC~346/N~66

349: from combination of X-ray observations with XMM-Newton (blue), [{\sc O

350: iii}] (501.1~nm) with NTT (green) and at 8\micron\ with {\sl Spitzer}

351: (red). The image covers an area about 6\arcmin~$\times$~6\arcmin\ ($\sim

352: 100 \times 100$~pc$^2$) wide centered on NGC~346. North is up, East is

353: to the left. {\sl Right}: Schematic representation of the proposed star

354: formation scenario, according to which the massive progenitor of

355: SNR~B0057$-$724 triggered current star formation at the north of the

356: N~66 bar through its wind-blown bubble.  The contribution of the most

357: massive stellar object in the area, the LBV HD~5980, is also considered

358: important. The colored iso-density contour map, constructed from star

359: counts of both bright main sequence and faint PMS stars of the region,

360: is shown in the background to demonstrate the geometry of the youngest

361: stellar concentrations. The locations of the SNR, HD~5980, the O-star

362: Sk~80 and significant PMS clusters and {\sc H~ii} regions are also

363: indicated. Two important dusty features, also visible in the left panel,

364: are drawn with red thick curves. The one to the south, south-west of the

365: bar of N~66 is considered to be the product of the energy output from

366: the OB stars of the association NGC~346. The other to the north and east

367: of the bar coincides very well with a shock-wave coming from the

368: direction of the SNR~B0057$-$724.  The synergy of both these strong

369: energy sources seems to define the recent star formation history of the

370: region. \label{fig-imaxoi}}

371: \end{figure*} 

372: %\clearpage

373: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

374: 

375: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

376: %%%%%%%%%%%%%%%    PR CAPTION / DESCRIPTION

377: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

378: %{\sl Left}: Color-composite image from multi-wavelength observations of

379: %the general area of NGC~346/N~66. North is up, East is to the left.  The

380: %image covers an area about 6\arcmin~$\times$~6\arcmin\ ($\sim 100 \times

381: %100$~pc$^2$) wide centered on NGC~346. Combination of X-ray observations

382: %with XMM-Newton (blue), [{\sc O iii}] (501.1~nm) with NTT (green) and

383: %the 8 \micron\ IRAC channel of {\sl Spitzer} (red) is shown. Two

384: %circular X-ray emission features are visible at the left of the image;

385: %the upper, at the north, coincides with the SNR~B0057$-$724 and the

386: %triple LBV HD~5980 (Sk~78), while the lower one, to the south, with the

387: %bright O-type star Sk~80 (see \S~2.2). A photo-ionizing front, appearing

388: %greenish and embracing the northern X-ray emission, is brighter to the

389: %west of the SNR.  Further away, at least two distinct shock-like

390: %features can be seen in [{\sc O iii}]. The red component of the image

391: %exhibits the locations where ongoing star formation is present, as seen

392: %from the mid-infrared dust emission.  Two major reddish features are

393: %apparent. The one to the south, south-west of the central part of the

394: %bar of N~66 is considered to be the product of the energy output from

395: %the OB stars of the association NGC~346. The other to the north,

396: %north-east of the bar coincides very well with a shock-wave coming from

397: %the direction of the SNR. The synergy of both these strong energy

398: %sources seems to define the recent star formation history of the region.

399: %{\sl Right}: Schematic representation of the proposed star formation

400: %scenario, according to which the massive progenitor of SNR~B0057$-$724

401: %triggered current star formation at the north of the N~66 bar through

402: %its wind-blown bubble.  The contribution of the two most massive stellar

403: %objects in the area, HD~5980 and Sk~80 may have also been important. The

404: %colored iso-density contour map, constructed from star counts of both

405: %bright main sequence and faint PMS stars of the region, is shown in the

406: %background to demonstrate the geometry of the youngest stellar

407: %concentrations. The southern gaseous arc produced by the photo-ionizing

408: %action of the OB stars of the association and the arc located to the

409: %east of NGC~346 shaped by the wind-blown bubble, are drawn with red

410: %thick curves. The locations of the SNR, HD~5980, Sk~80 and significant

411: %PMS clusters and {\sc H~ii} regions are also indicated. 

412: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

413: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

414: 

415: \subsection{A Multi-wavelength Image of NGC~346/N~66}

416: 

417: Two X-ray bright features can be seen (in blue) in Fig.~\ref{fig-imaxoi}

418: ({\sl left}), located to the east and east-northeast of NGC~346

419: respectively (left part of the image). In this image an extraordinarily

420: good agreement between the northern (upper) X-ray nebula (in blue) and

421: gas presumably photo-ionized seen in [{\sc O~iii}] (in green) is

422: present. Although the X-ray emission region appears in

423: Fig.~\ref{fig-imaxoi} to be small and almost circular, it is actually

424: much more extended, as {\sl Chandra} ACIS observations show (Guerrero \&

425: Chu 2008; their Fig. 1c), with its western boundary reaching the

426: northern edge of the bar of N~66, where [{\sc O~iii}] emission provides

427: more evidence of heated gas. The northern arc-like nebular feature (seen

428: in green-red) is perfectly outlined by YSOs, IR peaks, and young PMS

429: stars (discussed in \S~2.1), and the 8\micron\ mid-infrared component of

430: the color image (shown in red) gives further information about dust

431: emission from the star-forming arm.  Specifically, its northern part

432: (which includes two PMS clusters and one IR-emission peak), suggests

433: that there is probably still undisrupted cloud material in this part of

434: the region (this is verified by the CO map by Rubio et al. 2000). The

435: location of this part of the star-forming feature as seen in

436: Fig.~\ref{fig-imaxoi} ({\sl left}) is not compatible with a sequential

437: star formation in the bar of N~66, but rather with a mechanism outside

438: the bar. In order to clarify how this feature may have been actually

439: formed we discuss in the following sections the two most important

440: energy sources known in this specific area.

441: 

442: \subsection{HD~5980: A Peculiar Massive Object in N~66}

443: 

444: Embedded in the northernmost diffuse X-ray emission of

445: Fig.~\ref{fig-imaxoi} is a point source centered on the remarkable

446: massive triple system HD~5980, located at 00$^{\rm h}$59$^{\rm

447: m}$26.55$^{\rm s}$, $-$72\deg09\arcmin53.8\arcsec\ (J2000).  This

448: object, which underwent a luminous blue variable (LBV) type

449: eruption\footnote{LBVs are thought to be evolved He-burning stars which

450: may develop into Wolf-Rayet stars.} in 1994 is also known as Sk 78

451: (Sanduleak 1968)\footnote{The southernmost X-ray bright source is known

452: as Sk 80, and it is a bright massive star classified as O7Iaf (see also

453: Evans et al. 2006).}. Its spectral variations are unique: Before 1980

454: this object was classified as WN+OB, then it changed to WN3+WN4 for the

455: period 1980 - 1983, and to WN6 in 1992.  In 1994, after the

456: LBV-eruption, it turned to WN11. The high X-ray emission of HD 5980

457: makes it comparable to the X-ray-brightest single WN stars (Wessolowski

458: 1996) and the brightest WR+OB binaries (Pollock 2002) of the Galaxy. The

459: nature of HD~5980 is investigated in detail with {\sl Chandra}

460: observations of NGC~346 by Naz\'{e} et al. (2002), who note that the

461: fast wind from the post-eruptive phase, which collides with the slow

462: wind ejected during the eruption, causes the high X-ray surface

463: brightness. Bright diffuse X-ray emission has been observed in other

464: LBVs, i.e. $\eta$ Carinae in the Milky Way. However, in contrast to

465: $\eta$ Carinae, HD~5980 has not had time to develop a LBV nebula around

466: it yet, and so another proposed explanation for the high X-ray

467: luminosity of HD~5980 is the collision of the winds from its close

468: massive stellar components.

469: 

470: Naturally, this massive object may be associated, due to its energetic 

471: nature, with the northern feature in the region of NGC~346/N~66, 

472: and indeed, Walborn (1978) notes that the arcs seen in H$\alpha$ and [{\sc 

473: O~iii}] are an indication that HD~5980 interacts with its environment. 

474: However, from comparisons between X-ray and visible data, Naz\'{e} et al. 

475: (2002) argue that {\sl the bright, extended X-ray emission around HD 5980 

476: is most probably due to a nearby core-collapse supernova remnant (SNR) 

477: whose progenitor is unknown}. These authors describe the shape of this 

478: emission as rectangular with an extension to the northeast.  Its 

479: brightness is rather uniform with the exception of some dark arcs. Its 

480: size is $\simeq$ 37~$\times$~29~pc. They also note that the spatial 

481: coincidence of this extended X-ray emission with the peculiar massive star 

482: implies an association between the two objects. We explore this hypothesis 

483: in the following section, where the known information about this SNR is 

484: presented.

485: 

486: 

487: %We provide evidence that the star formation at the northern part of the

488: %N~66 nebula was triggered by a mechanism related to the progenitor of

489: %this SNR and/or other bright stars in the area like HD~5980.

490: 

491: \subsection{A Core-Collapse Supernova Remnant in N~66}

492: 

493: Massive stars in stellar associations affect their environment by ionizing 

494: radiation, stellar winds, and, finally, supernova explosions (e.g. Chu 

495: 1997). Massive supernova progenitors do not migrate far from their 

496: birthplaces in their short lifetimes, and when they explode the produced 

497: shock-waves can trigger cloud collapse, enhancing star formation in the 

498: vicinity of their environment. Numerical studies (e.g. Vanhala \& Cameron 

499: 1998; Fukuda \& Hanawa 2000) show that shock-waves with velocities in the 

500: range of 15~-~45~km~s$^{-1}$, corresponding to supernova explosions at a 

501: distance between 10 pc and 100 pc from the molecular cloud, can induce 

502: such a collapse.

503: 

504: Earlier studies show that the general area of NGC~346/N~66 is 

505: characterized by at least three known SNRs (Ye et al. 1991; Reid et al. 

506: 2006)\footnote{Reid et al.  (2006) suggest the existence of another SNR 

507: candidate in the region at 00$^{\rm h}$59$^{\rm m}$32.4$^{\rm s}$, 

508: $-$72\deg08\arcmin42.4\arcsec (J2000), based on observations from the DBS 

509: 2.3-m telescope at the Siding Spring Observatory.} located away from the 

510: center of the region: SNR B0056$-$724 (also known as SNR 0056$-$72.4), 

511: located to the southwest of the region, SNR B0056$-$725 (or SNR 

512: 0056$-$72.5) to the west, and SNR B0057$-$724 (or SNR 0057$-$72.2) to the 

513: east. The first two SNRs are located to the south and west, outside the 

514: field-of-view shown in Fig.  \ref{fig-imaxoi}. The third SNR 

515: (B0057$-$724), though, is located closer to the central part of 

516: NGC~346/N~66, and actually {\sl is the SNR associated with the 

517: northernmost extended X-ray emission} seen in Fig.~\ref{fig-imaxoi} ({\sl 

518: left}). Indeed, Reid et al. (2006) demonstrated that the extended X-ray 

519: emission surrounding HD~5980 coincides exactly with the radio non-thermal 

520: emission, confirming ``beyond a doubt its true nature as a SNR''.

521: 

522: SNR~B0057$-$724 has been previously studied in H$\alpha$ with echelle

523: spectroscopy by Chu \& Kennicutt (1988), who observed high-velocity,

524: shock-accelerated material within the boundary of the SNR that is

525: distinct from the low-velocity quiescent material in the {\sc H~ii}

526: region. It was also identified in the radio by Ye et al. (1991), who

527: offer an explanation of why SNRs are expected to be found near the edge

528: of {\sc H~ii} regions based on the argument that sequential star

529: formation is expected to occur about 2.5 Myr after and 10 - 15 pc away

530: from the previous generations, as massive stars move away from their

531: birthplaces (Elmegreen \& Lada 1977).  UV analyses with the {\sl 

532: International Ultraviolet Explorer} (de Boer \& Savage 1980), {\sl HST}

533: STIS (Koenigsberger et al. 2000), and {\sl Far-Ultraviolet Spectroscopic

534: Explorer} ({\sl FUSE}; Hoopes et al. 2001) have also confirmed the

535: presence of an expanding structure. Danforth et al.  (2003) measured the

536: velocity of the SNR emission with {\sl FUSE}, and they found

537: high-velocity features with increasing velocity towards the center,

538: across the whole face of SNR B0057$-$724, {\sl typical for an expanding

539: shell}. The fastest emission ($v_{\rm LSR}\sim+335~{\rm km}~{\rm

540: s}^{-1}$) is found at about 4 pc south of HD~5980 and is consistent with

541: the peak in X-ray brightness.  From X-ray and radio images these authors

542: set the center of the SNR at 00$^{\rm h}$59$^{\rm m}$27$^{\rm s}$,

543: $-$72\deg10\arcmin15\arcsec (J2000).

544: 

545: Danforth et al. (2003) propose a topology, according to which a roughly

546: spherical SNR is located on the near side of N~66, basically closer to

547: us than NGC~346 itself. Consequently the rear side of the SNR is

548: propagating into relatively denser ionized nebula, while the near side

549: is propagating through the more diffuse SMC ISM, and is harder to

550: detect. The X-ray emission arises over a spherical cap, where the shock

551: has encountered and heated the denser material, for which Danforth et

552: al. (2003) estimate a pre-shock density of $\sim 6 \pm 2$~cm$^{-3}$.

553: According to these authors the remnant is not yet even halfway submerged

554: in the N~66 material, and thus, the observed radius of X-ray emission

555: (as observed by Naz\'{e} et al. 2002 and seen in the left panel of Fig.

556: \ref{fig-imaxoi}) is smaller than the actual blast wave radius. HD~5980

557: and other UV-bright stars in the NGC~346 association are embedded within

558: N~66, but HD~5980 lies behind the remnant, while NGC~346 is outside the

559: SNR\footnote{Danforth et al. suggest that Sk~80 is probably inside the

560: SNR.}. 

561: 

562: %NOTE: The combined stellar winds and thermal pressure have formed a 

563: %slowly expanding envelope around NGC~346.

564: 

565: Danforth et al. (2003) also discuss the relationship between SNR

566: B0057$-$724 and N~66, and they conclude that the detection of {\sc O~vi}

567: emission together with the H$\alpha$ kinematics imply the presence of a

568: strong shock in N~66 associated with SNR B0057$-$724.  Although this

569: SN-driven shock affects its direct surroundings, it cannot serve as the

570: triggering mechanism of the observed star-forming event at the north of

571: the bar of N~66 for the following reason.  The expansion velocity of the

572: SNR is $v = 175$ km~s$^{-1}$ (Danforth et al.\ 2003).  Assuming the

573: Sedov solution for the SNR (e.g. Tang \& Wang 2005), the travel time

574: for the shock front to reach the observed star-forming region ($r \sim

575: 18$ pc) is $t = 2r/5v = 0.04$ Myr, much shorter than the ages of the

576: compact PMS clusters located in the nebular arc, as estimated by us

577: (Paper~III; $\tau$~\lsim~2.5~Myr), and SSN07 ($\tau\approx3\pm1$~Myr).

578: Consequently, while the relative position of the SNR and the east-facing

579: feature seems to support a relation between them (Fig. \ref{fig-imaxoi};

580: left) SNR~B0057$-$724 {\sl cannot be} the triggering agent for the star

581: formation in the northern part of N~66. On the other hand, its massive

582: progenitor could have played a very important r\^{o}le to this event. 

583: 

584: %%%%% OLD VERSION %%%%%%

585: %Danforth et al. (2003) also discuss the relationship between

586: %SNR~B0057$-$724 and N~66, and they conclude that the detection of [{\sc

587: %O vi}] emission together with the H$\alpha$ kinematics imply {\sl the

588: %presence of a strong shock in N~66 associated with SNR~B0057$-$724}. 

589: %However, although this SN-driven shock affects its direct surroundings,

590: %it {\sl cannot} serve as the triggering mechanism of the observed

591: %star-forming event at the north of the bar of N~66. If we consider the

592: %distance of the star-forming feature from the SNR ($r\sim$~18~pc) and

593: %the radial velocity of [{\sc O~iv}] and H{\alp} emission observed by

594: %Danforth et al. (2003) at the X-ray edge of the SNR ($v_{\rm

595: %LSR}\sim$~225~km/s), assuming the Sedov solution (e.g. Tang \& Wang

596: %2005) we have for the velocity of the shock $v(t) = 2r/5t$, and

597: %therefore we can estimate that the SN explosion should have taken place

598: %$t\sim$~0.03 Myr ago. Since this measurement is based on the radial

599: %velocity at the edge of the SNR, where most of the expansion velocity

600: %is perpendicular to the line-of-sight, one may consider the expansion

601: %velocity near the SNR center of $v_{\rm LSR}\sim$~175~km/s (Danforth et

602: %al. 2003). In this case the time since the SN explosion occurred is a

603: %bit longer and within the typical expected visibility timescale for an

604: %isolated SNR of \lsim~0.1~Myr (e.g.  Williams et al. 2004), but it is

605: %much shorter than the ages of the compact PMS clusters located in the

606: %nebular arc, as estimated by us (Paper~III; $\tau$~\lsim~2.5~Myr), and

607: %SSN07 ($\tau\approx3\pm1$~Myr). Consequently, while the relative

608: %position of the SNR and the east-facing feature seems to support a

609: %relation between them (Fig. \ref{fig-imaxoi}; left) SNR~B0057$-$724 {\sl 

610: %cannot be} the triggering agent for the star formation in the northern

611: %part of N~66. On the other hand, its massive progenitor could have

612: %played a very important r\^{o}le to this event. 

613: 

614: 

615: 

616: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

617: %\clearpage

618: \begin{figure*}[t!]

619: \epsscale{1.15}

620: \plotone{f2.eps}

621: %\includegraphics[scale=0.75]{f2.eps}

622: \caption{Chart of all bright stars in the region of NGC~346/N~66 (left)

623: from photometry with the 0.9-m CTIO Telescope by Massey et al. (1989),

624: and the corresponding $U-B$, $B$ CMD (right).  The limits of the bubble

625: as it is outlined by the extended X-ray emission in the area within a

626: radius of $\sim$~18~pc is also plotted over the stellar chart. The stars

627: included by these limits are plotted with thick symbols in the CMD.  The

628: brightest sources in the region of the bubble, HD~5980 and Sk~80, are

629: indicated in both the chart and the CMD. A young isochrone of

630: $\log{\tau}\simeq 6.6$ (the youngest available) for the grid of

631: evolutionary models by Girardi et al. (2002) with no extinction applied 

632: is overplotted on the CMD for an assumed distance modulus of 

633: $m-M\simeq18.95$~mag.\label{fig-mapcmd}}

634: \end{figure*}

635: %\clearpage

636: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

637: 

638: \section{A Star Formation scenario}

639: 

640: During their main sequence phase massive stars photo-ionize {\sc H~ii}

641: regions or blow bubbles through their winds. Shocks associated with the

642: ionization fronts or expanding bubbles can trigger cloud collapse, and

643: this process lasts for a few Myr, the lifetime of the star during the

644: main sequence phase. Such a process has been observed in the Milky Way,

645: e.g. at the periphery of the {\sc H~ii} regions Sh2-219 (Deharveng et

646: al. 2006) and RCW~79 (Zavagno et al. 2006). This mechanism may produce

647: young compact star clusters. Indeed, Lee \& Chen (2007) presented

648: recently their diagnosis of the r\^{o}le that massive stars play in the

649: formation of low- and intermediate-mass stars in the $\lambda$~Ori

650: region, Ori~OB1, and Lac~OB1 associations in the Milky Way. Their study

651: supports the radiation-driven implosion (e.g. Kessel-Deynet \& Burkert

652: 2003) as the triggering mechanism, where the Lyman continuum photons

653: from a luminous O star create expanding ionization fronts to evaporate

654: and compress nearby clouds into bright-rimmed clouds. Implosive pressure

655: then causes dense clumps to collapse, prompting the formation of

656: intermediate-mass Herbig Ae/Be (HAeBe) stars somewhat deeper in the

657: cloud and low-mass (T~Tauri) PMS stars on the cloud surface (i.e., the

658: bright rim).

659: 

660: Bubbles are formed by the stellar winds, not the ionization energy of

661: massive stars. The expansion of an {\sc H~ii} region (not yet a bubble)

662: is caused by its higher thermal pressure due to the high temperature

663: compared to the pre-ionized state.  Typical expansion velocities are

664: around 10~km/s at temperatures $\sim 10^4$~K. On the other hand, the

665: expansion of a bubble is driven by the thermal pressure of the shocked

666: stellar wind in the central cavity. The velocity of this expansion

667: ranges from 10 to 100~km/s (Drake 1986), depending on the wind strength

668: and ambient density. When the massive star explodes as a supernova, it

669: explodes inside the bubble blown by itself during its main sequence

670: phase, and the produced SNR will have a size similar to the bubble blown

671: by the star. This will create the picture of a SNR with star formation

672: taking place on its edge, similar to what we observe at the northern

673: part of N~66. However, it is the massive progenitor that triggered star

674: formation and not the SN. The importance of stellar winds as a feedback

675: mechanism to star formation over supernovae has been demonstrated by

676: Leitherer et al. (1992), who present models for the integrated output of

677: mass, momentum, and energy from a concentration of massive stars,

678: computed for typical starburst parameters. In general, the most massive

679: stars evolve from the main sequence (MS) into blue supergiants (BSG) and

680: luminous blue variables (LBV) to become then Wolf-Rayet (WR) stars.

681: Leitherer et al. (1992) found that the (post-main sequence) OB phase,

682: together with the WR phase is the dominant contributor to the total wind

683: output. These authors considered the cases of an instantaneous and of a

684: continuous starburst and they found that in both cases the LBV phase is

685: unimportant, as a consequence of its short life time ($< 10^5$~yr; e.g.

686: Lamers 1989). A total power output between $10^{39.5}$ and

687: $10^{40.5}$~ergs~s$^{-1}$ is estimated by these authors for a

688: concentration of stars following a Salpeter IMF with masses between 1

689: and 120~M{\solar} throughout the epoch of a continuous starburst at

690: $10^{6} < T < 10^{7}$~yr. 

691: 

692: %(between $10^{52}$ and $10^{53}$~UV~photons~s$^{-1}$)

693: 

694: %The mechanical energy output due to mass loss of such a stellar

695: %concentration, as estimated by us with {\tt Starburst99}

696: %(Ver.~5.1)\footnote{{\tt Starburst99} is available at

697: %http://www.stsci.edu/science/starburst99/} amounts to about

698: %$10^{55}$~ergs in total. 

699: 

700: \subsection{Wind-Triggered Star Formation in N~66}

701: 

702: Considering the geometry of the north-northeastern part of the region of

703: NGC~346/N~66 as seen in Fig.~\ref{fig-imaxoi} ({\sl left}), the

704: hypothesis of a wind-triggered star formation event seems to explain

705: best the observed recent star formation in this area. We propose a

706: scenario according to which the massive progenitor of SNR~B0057$-$724

707: has triggered or contributed significantly to the triggering mechanism

708: of star formation through its expanding bubble, producing the northern

709: gaseous arc seen in [{\sc O~iii}] and H\alp. A schematic representation

710: of this scenario is shown in the right panel of Fig.~\ref{fig-imaxoi},

711: where the two important gaseous features, the southern part of the N~66

712: bar shaped by the photo-dissociation of the central OB stars and the

713: northern feature delineating the boundary of the SNR X-ray emission, are

714: drawn with red thick curves. If indeed the star-forming northern nebular

715: arc {\sl is} related to the activity at east, it would be interesting to

716: estimate how much mechanical energy could have been dumped in the area

717: to produce this event.

718: 

719: %Tenorio-Tagle et al. (2006) modeled the effect of thermalized winds from

720: %several very massive stars, assuming a young massive star cluster

721: %surrounded by a distribution of cloudlets, which are exposed to both

722: %photo-ionization and stellar winds. These authors assume a major star

723: %formation burst with a stellar mass of 3~$\times$~10$^6$~M{\solar} and a

724: %Salpeter IMF for stars between 1 and 100~M{\solar}, all concentrated

725: %within a radius of 5~pc. The initial power of such a cluster is

726: %$10^{41}$~ergs~s$^{-1}$ ($10^{54}$~UV~photons~s$^{-1}$), comparable to

727: %the models by Leitherer et al. (1992). Their calculations show an

728: %evolutionary sequence in which two competing hydrodynamic processes

729: %define the timescale: Photo-ionization aims at establishing a constant

730: %density medium everywhere, while wind-driven mechanical energy escapes

731: %out of the cloudlet distribution. In all cases that consider both

732: %effects, the ensemble of clouds ends up being driven to compose an

733: %almost constant density, low-velocity ragged shell that acts as a

734: %barrier to most of the shocked wind.

735: 

736: Naturally the existence of HD~5980 and Sk~80 in the same region with the

737: SNR may not be considered as a coincidence, and therefore one may

738: question if the progenitor is the only triggering agent or not. Indeed,

739: the massive object HD~5980 could contribute to the triggering process

740: through its own bubble blown during its main sequence phase, and another

741: hypothesis is that the cumulative action of even more massive stars in

742: the region, in addition to the SN-progenitor, may have triggered the

743: formation of the northern extended feature. This implies the existence

744: of a massive cluster in the area of the SNR as the triggering agent. In

745: order to test this hypothesis we used the most complete photometric

746: catalog of bright stars in the region (Massey et al. 1989) to search for

747: such a cluster. The stellar chart of all bright sources in this catalog

748: and the corresponding CMD, shown in Fig.~\ref{fig-mapcmd}, indicate that

749: no specific concentration of massive stars can be observed within the

750: bubble limits.  Support to the lack of a star cluster of any kind in the

751: region is also provided by the catalog of faint sources down to

752: $V\simeq$~28~mag from our ACS/WFC photometry of the whole NGC~346 region

753: (Paper~I), where it can be seen that no specific concentration of

754: low-mass stars exist in the area of the SNR either. It is most probable

755: that the bright stars seen in the CMD of Fig.~\ref{fig-mapcmd} belong to

756: the outskirts of NGC~346 itself or to the field. Nevertheless, the only

757: bright sources, which could provide the necessary kinetic energy for a

758: triggered star formation event, in addition to the SN-progenitor, are

759: HD~5980 (Sk~78) and Sk~80, both located within the SNR

760: (Fig.~\ref{fig-mapcmd}).

761: 

762: Single stars between 8 and 20~M{\solar} produce type II-P SNe when they

763: are red supergiants (Crowther \& Smartt 2007). Stellar structure models

764: of more massive stars predict that such stars will have final

765: carbon-oxygen cores of more than 3 M{\solar}, and hence form black holes

766: (Heger et al.  2003), suggesting that such objects form very faint SNe.

767: Recent discoveries suggest that very massive stars may explode as SNe in

768: their LBV stage. Pastorello et al. (2007) found that the very energetic

769: SN2006jc was a type Ic event, but embedded within a He-rich

770: circumstellar envelope and that it was coincident with an LBV-like

771: outburst just 2 yrs before collapsing. According to these authors it

772: could have been an originally very massive LBV which lost the last of

773: its H and He envelopes in energetic mass ejection episodes and collapsed

774: as a stripped Wolf-Rayet star. Recently Smith et al. (2007) reported the

775: analysis of the most luminous SN ever recorded, the type~IIn SN2006gy in

776: NGC 1260, which reached an absolute magnitude of $R \simeq -22$~mag. The

777: combination of optical and X-ray monitoring suggests it was the

778: explosion of a very massive star of about 100 M{\solar}, which failed to

779: shed its hydrogen envelope. According to these authors SN2006gy implies

780: that some of the most massive stars can explode prematurely during the

781: LBV phase, never becoming Wolf-Rayet stars.

782: 

783: Since there is no information about the nature of the progenitor of

784: SNR~B0057$-$727, and in order to have an upper-limit estimation of the

785: wind-blown energy expected by it, we may assume that this progenitor was

786: an object as massive as the estimate of Smith et al. (2007) for

787: SN2006gy, meaning roughly an early O-type BSG. Typical ionizing photon

788: fluxes for such stars are $F_{\rm Ly} \simeq

789: 10^{50}$~UV~photons~s$^{-1}$ (Vacca et al.  1996; Schaerer \& de Koter

790: 1997). O-type stars have mass-loss rates $\dot{M}$ of about

791: $10^{-7}$~-~$10^{-5}$~M{\solar}~yr$^{-1}$ and terminal velocities

792: $v_{\rm W}$ in the range 1~-~4~$\times 10^{3}$~km~s$^{-1}$ (Chlebowski

793: \& Garmany 1991; Lamers et al. 1995), thus injecting large amounts of

794: energy to the neighboring ISM. The corresponding mechanical luminosity

795: can be derived as $L_{\rm W} = \dot{M}v_{\rm W}^{2}/2$ (e.g. Cappa \&

796: Benaglia 1998). Consequently, an O3-type star with a mass loss rate of

797: $10^{-5}$~M{\solar}~yr$^{-1}$ and a wind velocity of 3750~km~s$^{-1}$

798: (Puls et al. 1996) would have a stellar wind mechanical luminosity of

799: $4.5 \times 10^{37}$~ergs~s$^{-1}$. This mechanical input is lower than

800: the assumed energy contribution of a massive cluster in the models by

801: Leitherer et al.  (1992), but since the density of the original cloud is

802: unknown, one can only speculate about the energy that would be needed to

803: trigger star formation in the region.  Nevertheless, the two additional

804: energy sources in the region should be also considered. Sk~80, being an

805: O7~I giant would contribute around $3\times 10^{36}$~ergs~s$^{-1}$ (Puls

806: et al. 1996). 

807: 

808: HD~5980 has been suggested to be a triple system (Koenigsberger et al.

809: 2000) with the first component being a 50~M{\solar} WN6 star, the second

810: a 28~M{\solar} WN2 (with high uncertainty) and the third component being

811: a O4~-~O7 giant. Koenigsberger et al. (2000) found that a wind-wind

812: collision shock cone winds tightly around the second companion, and

813: using the relation given by Volk \& Kwok (1985), under a two-wind

814: approximation, they provide a rough estimate of $2 \times

815: 10^{-4}$~M{\solar}~yr$^{-1}$ for the current mass-loss rate of the

816: system with a terminal velocity of $\sim$~2000~km~s$^{-1}$. This

817: mass-loss rate measurement, which is higher than typical rates of

818: Galactic WR stars (2~-~10~$\times 10^{-5}$~M{\solar}~yr$^{-1}$ with

819: terminal velocities of 1000~-~2500~km~s$^{-1}$; van der Hucht 1992),

820: yields a mechanical luminosity of $2.6\times 10^{38}$~ergs~s$^{-1}$ for

821: HD~5980. As far as the ionizing flux of HD~5980 is concerned, Law et al.

822: (2002) found that the Lyman continuum flux of WN stars appears to be

823: independent of spectral type, and lower bound estimates tend to cluster

824: around $10^{48}$~UV~photons~s$^{-1}$. This would be the ionizing energy

825: contributed by each of the WN components of HD~5980, while the third

826: component, being an O4~-~O7 giant, would provide an additional

827: $\sim$~$10^{49.5}$~UV~photons~s$^{-1}$.

828: 

829: From the mechanical luminosities of each of the involved objects in the

830: region, one can estimate that the total mechanical power that was

831: supposedly dumped in the region is of the order of that produced by

832: HD~5980 alone, $L_{\rm W}^{\rm Tot} \simeq 3\times

833: 10^{38}$~ergs~s$^{-1}$, implying a small contribution from the massive

834: progenitor of SNR~B0057$-$724 and Sk~80 in the process. However, this

835: result stands only if the SN-progenitor did not go through a major

836: mass-loss event. According to the standard evolution of O-type stars

837: with masses \gsim~60~M{\solar}, these stars will go through a WNH or LBV

838: phase that removes via line-driven winds the remaining H-envelope to

839: yield a \lsim~20~M{\solar} WR star. Moreover, the large number of

840: observed giant eruptions of LBVs, which can remove around 10~M{\solar}

841: in a few years (Smith \& Owocki 2006), clearly suggests that massive

842: O-type stars do not shed most of their initial mass by the end of the

843: main sequence phase, but there is still enough mass remaining to power

844: major mass-loss during the WNH or LBV phase. Indeed, recent

845: observational evidence suggests that mass-loss rates for O stars need to

846: be revised downward (Smith 2007), in line with recent observed mass-loss

847: rates for clumped winds (e.g. Fullerton et al. 2006). 

848: 

849: Could the progenitor of SNR~B0057$-$724 have lost most of its initial

850: mass via an eruptive mass-loss event (during e.g. a LBV phase),

851: triggering the observed star formation? The answer is positive. The

852: primary star in the $\eta$~Car system, which is suggested to have a

853: present-day mass \gsim~100~M{\solar} has lost 20~-~30~M{\solar}, in

854: violent LBV eruptions in just a few thousand years (Smith et al. 2003),

855: in addition to its steady wind. Consequently a SN-progenitor alone is

856: capable of triggering star formation event, as has been recently

857: observed with the {\sl AKARI} infrared satellite by Koo et al. (2008).

858: These authors report the discovery of a star-forming loop around the

859: young, Crab-like SNR~G54.1+0.3 in the Galaxy, and they propose that star

860: formation was triggered by the progenitor star of G54.1+0.3, which has a

861: mass of \lsim\ 15 M{\solar}. The triggering must have occurred near the

862: end of the progenitor's life, possibly after it had evolved off the main

863: sequence. Under these circumstances, the massive progenitor of

864: SNR~B0057$-$724 alone may have provided the necessary mechanical energy

865: to trigger star formation to the north, while the contribution of

866: HD~5980 may have significantly enhanced this event.

867: 

868: 

869: %Under these circumstances, the massive progenitor of SNR~B0057$-$724

870: %alone may not have provided the necessary mechanical energy to trigger

871: %star formation to the north during its main-sequence phase, while the

872: %contribution of HD~5980 to this process is found to be quite

873: %significant. 

874: 

875: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

876: %\clearpage

877: \begin{figure*}[t!]

878: \epsscale{1.}

879: \plotone{f3.eps}

880: %\includegraphics[scale=0.75]{f2.eps}

881: \caption{Images of the region of NGC~346/N~66 from the {\sc H~i} mosaic

882: constructed with {\sl ATCA} and {\sl Parkes} (greyscale) with the H$\alpha +$

883: continuum from MCELS overlaid (contours). The most significant velocity

884: planes are shown, with the corresponding heliocentric velocity indicated

885: on the top-left corner. Each image covers an area about

886: 0.35\deg~$\times$~0.35\deg\ ($\sim 357 \times 357$~pc$^{2}$) wide,

887: centered on N~66, which is indicated by the H\alp\ isopleths. North is

888: up, East is to the left. The thick cross-symbol in the image of the

889: 168~km~s$^{-1}$ velocity plane indicates the position of

890: SNR~B0057$-$724. A hole in the {\sc H~i} east of N~66 and possibly

891: related to the area of the SNR can be seen in the velocity planes

892: between 165 and 178~km~s$^{-1}$, with its larger extend seen in

893: the 168~km~s$^{-1}$ velocity plane.  This velocity coincides almost

894: perfectly with the systematic velocity of the SNR found with echelle

895: spectroscopy by Chu \& Kennicutt (1988), indicating that the neutral

896: hydrogen {\sl is possibly} kinematically related to the ionized

897: gas.\label{fig-hi}}

898: \end{figure*}

899: %\clearpage

900: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

901: 

902: \subsection{Gas Kinematics in the Area of the SNR}

903: 

904: Additional information on the interstellar environment of

905: SNR~B0057$-$724 can certainly be provided by the kinematic study of the

906: surrounding gas. In Fig. \ref{fig-hi} we show a set of images centered

907: on NGC~346/N~66, constructed from observations of atomic and ionized

908: gas. The observations from the neutral hydrogen emission survey of the

909: SMC constructed with the {\sl Australia Telescope Compact Array} (ATCA)

910: and {\sl Parkes Observatory} at the {\sl Australia Telescope National

911: Facility} (ATNF) are shown as greyscale images\footnote{{\sl Parkes}

912: single-dish observations were combined with an aperture synthesis mosaic

913: of 320 separate pointings of the 375-m array from {\sl ATCA}

914: interferometry. These observations are sensitive to angular scales

915: between 98{\arcsec} (30~pc) and 4{\deg} (4~kpc) over a field of 20

916: deg$^2$ with velocity channels $\sim$1.65~km~s$^{-1}$ wide, spanning

917: between heliocentric velocities of $\sim$88.5 and

918: $\sim$215.5~km~s$^{-1}$.} (Staveley-Smith et al. 1997; Stanimirovi\'{c}

919: et al. 1999). The original data-cubes were kindly provided by Snezana

920: Stanimirovi\'{c}. Overplotted are the contours from the H\alp\ image of

921: the same general area of N~66 obtained within the {\sl Magellanic Cloud

922: Emission Line Survey}\footnote{MCELS is accessible at {\tt

923: http://www.ctio.noao.edu/mcels/}} (MCELS; Smith et al. 2000). Only

924: isopleths, which correspond to intensity higher than 3$\sigma$ above the

925: background, where $\sigma$ is the standard deviation of the background

926: noise, are shown. 

927: 

928: %Both the contours and the greyscale images are constructed from the

929: %original FITS files of the H\alp\ image and the low-resolution {\sc H~i}

930: %data-cube of the interferometry by ATCA, available from Lister

931: %Staveley-Smith at {\tt ftp.atnf.csiro.au/pub/data/smc}. 

932: 

933: %taken at Siding Spring Observatory by G. D. Bothun. 

934: 

935: In Fig. \ref{fig-hi} the 12 most significant velocity planes from the HI

936: interferometric data between 155 and 192 km~s$^{-1}$ (heliocentric

937: velocities) are shown. Each plane is 3.3~km~s$^{-1}$ wide and it is the

938: average of two planes in the processed cube. The central compact H\alp\

939: concentration in each image is the nebula N~66 and the area of the SNR

940: is directly to its left (east). It is very interesting to see that from

941: these sequential images, indeed, {\sl a possible relation between the

942: gas kinematics and the area of SNR~B0057$-$724} is revealed through {\sl 

943: a gas bubble visible in the planes between 165 and 178 km~s$^{-1}$}. 

944: However, the hole seen in {\sc H~i} seems to be much larger than the SNR

945: itself, which is about 2\arcmin\ in diameter (Chu \& Kennicutt 1988).

946: This implies that this hole should have been blown also by HD~5980 and

947: perhaps with the help of other stars. It should also be mentioned that

948: the location of the SNR is at the edge of the {\sc H~i} hole, not its

949: center. From the quantitative point-of-view, the echelle data of Chu \&

950: Kennicutt (1988) indicate a systemic velocity for the SNR of $v_{\rm

951: LSR} \sim$~158~km~s$^{-1}$, which translates to heliocentric velocity of

952: $v \approx v_{\rm LSR} + 11~{\rm km~s}^{-1} \simeq 169$~km~s$^{-1}$.

953: This value {\sl is in very good agreement with the velocity plane where

954: the hole in the atomic gas at the area of the SNR is more clearly

955: visible} in the images of Fig. \ref{fig-hi} (velocity of

956: 168~km~s$^{-1}$). This result provides evidence that indeed the atomic

957: gas {\sl is possibly} kinematically related to the ionized gas and

958: therefore possibly to the SNR. 

959: 

960: However, it is not clear whether the {\sc H~i} hole, seen at

961: 168~km~s$^{-1}$, is indeed a bubble. In order to clarify if this hole

962: actually represents an {\sc H~i} bubble, we examine the spatio-kinematic

963: structure of the neutral hydrogen in the vicinity of N~66 by

964: constructing position-velocity (PV) diagrams. Each PV diagram consists

965: of a east-west slice (slit) through the {\sc H~i} data cube where we

966: have averaged over 1\arcmin\ in the north-south direction. Multiple PV

967: diagrams were constructed, each after stepping sequentially by 1\arcmin\

968: in the north-south direction to cover the whole area of N~66 from east

969: to west. Twelve PV diagrams are shown in Fig.~\ref{fig-hiewcuts}, where

970: the positions of the corresponding slits are overlayed on the maps of

971: the area in H{\alp} (from MCELS) and {\sc H~i} (from ATCA and Parkes).

972: The vicinity of N~66 is fully covered in the H{\alp} image, the features

973: of which are easily comparable to the color-composite image of

974: Fig.~\ref{fig-imaxoi}. The star-forming feature north of the bar of N~66

975: is also easily detectable in this map. Above this feature, at slit

976: position 07, the {\sc H~i} velocity does not show any split. At slit

977: position 08, though, the {\sc H~i} at $\sim$~170~km~s$^{-1}$ starts to

978: split, and further south, this velocity split becomes more clear, giving

979: possible evidence of an expanding shell-structure. The PV diagrams of

980: Fig.~\ref{fig-hiewcuts} indicate an expansion velocity of

981: $\sim$~10~km~s$^{-1}$. It is interesting to note that the line split

982: does not occur over N~66, at slit positions 08 and 09. This indicates

983: that N~66 does not seem to be responsible for the expanding shell.

984: 

985: Considering a distance of 60~kpc and a correction factor for He of 1.4

986: we integrated the {\sc H~i} column density over an area roughly

987: 12\arcmin~$\times$~12\arcmin wide, centered on the shell, between

988: velocities of 158 and 188 km~s$^{-1}$ and derived a total mass of

989: 1.52~$\times$~10$^6$~M{\solar} in the {\sc H~i} expanding shell. This

990: mass corresponds to a total kinetic energy of about

991: 1.5~$\times$~10$^{51}$~erg. Taking into account the total mechanical

992: power estimated in \S~3.1 for the SN-progenitor, HD~5980 and Sk~80

993: (equal to 3~$\times$~10$^{38}$~erg~s$^{-1}$), one can see that these

994: objects have dumped a larger amount of mechanical energy than the

995: estimated kinetic energy in the shell in less than 1~Myr. Indeed, it

996: seems that it is typical for expanding bubbles that the kinetic energy

997: retained in the shell is only a small fraction of the total stellar

998: energy injected. This is observed for example in the superbubble N~51D

999: in the Large Magellanic Cloud, where the {\sc H~i} kinetic energy is

1000: found to be almost an order of a magnitude lower that the total stellar

1001: wind and supernova energy input (Chu et al. 2005). Based on our

1002: calculations of \S~3.1, the total energy provided by the SN-progenitor,

1003: HD~5980 and Sk~80 for the last couple Myr is indeed an order of a

1004: magnitude higher than the kinetic energy in the shell. 

1005: 

1006: %Consequently, these objects possibly provided the necessary energy for

1007: %the expansion of the {\sc H~i} shell.

1008: 

1009: %It would be interesting to compare the kinetic energy of the {\sc H~i}

1010: %gas with the energy input by the massive stars in the region of the SNR.

1011: 

1012: %Consequently, SNR~B0057$-$724 seems to be the most important feature at 

1013: %the eastern part of the region NGC~346/N~66, since its progenitor should 

1014: %be considered as the most probable agent that triggered star formation at 

1015: %the northern arc feature.

1016: 

1017: %The only spectral lines attributable to the second close companion are

1018: %the very extended P~Cygni absorption components of the strong UV lines,

1019: %present only during the eclipse of the eruptor, implying that the

1020: %wind-wind collision shock cone winds tightly around this companion.

1021: 

1022: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 

1023: %\clearpage 

1024: \begin{figure}[t!]

1025: \epsscale{1.15} 

1026: \plotone{f4.eps} 

1027: \caption{Position-velocity diagrams from ATCA and Parkes HI observations

1028: for 12 selected slits covering the whole area of N~66 from east to west

1029: (long-slit spectra). The positions of the slits are shown on the maps of

1030: the area in H{\alp} (from MCELS) and HI (from ATCA and Parkes). The HI

1031: velocity of $\sim$~170~km~s$^{-1}$ shows a split at slit position 08 and

1032: southwards, indicating that the shell probably expands. This expanding

1033: shell structure continues south.\label{fig-hiewcuts}}

1034: \end{figure} 

1035: %\clearpage

1036: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1037: 

1038: \section{Final Remarks}

1039: 

1040: %Considering that the formation of the young compact PMS clusters of the

1041: %northern star-forming arc is induced by an {\sc H~ii} region or a

1042: %wind-blown bubble, the individual study of these clusters will provide

1043: %further information on the products of triggered star formation in the

1044: %region. However, the available ACS data, being taken with the Wide Field

1045: %Channel, are not deep enough to allow us an accurate measurement of the

1046: %individual characteristics of the clusters, such as their Initial Mass

1047: %Function and dynamical status. Data taken with the High Resolution

1048: %Channel of ACS, which have been recently released (HST GO Program

1049: %10542), as well as near-IR imaging previously taken with VLT/ISAAC (ESO

1050: %Program 63.I-0329) are included as part of an on-going follow-up

1051: %investigation of the properties of these young clusters.

1052: 

1053: In conclusion, based on the evidence presented here, we propose an

1054: expanding {\sc H~ii} region or bubble blown by the winds of the massive

1055: progenitor of SNR~B0057$-$724, possibly the massive object HD~5980 and

1056: maybe the O7 giant Sk~80 - all located at the eastern vicinity of

1057: NGC~346/N~66, as the secondary mechanism that shapes the recent star

1058: formation in this region, in addition to the photo-ionizing process of

1059: the OB stars of the association. This mechanism is similar to shell-like

1060: {\sc H~ii} regions, with a central young cluster in an evacuated cavity,

1061: and with ongoing star formation triggered around their periphery, and

1062: therefore is a rather quiescent process. Typical examples for such

1063: regions in the MCs are LH~9/N~11 (Walborn \& Parker 1992) and 30 Doradus

1064: (Walborn et al. 2002) in the LMC and NGC~602/N~90 (Gouliermis et al.

1065: 2007; Carlson et al. 2007) in the SMC. It should be noted, however, that

1066: within the X-ray bubble associated with the SNR, which most probably

1067: outlines the wind-blown bubble of its progenitor, no signature of any

1068: underlying cluster is found.

1069: 

1070: The induced star formation in this part of the region of NGC~346/N~66 is

1071: evidenced by the existence of PMS clusters, candidate YSOs and

1072: IR-emission peaks at a projected distance $\sim$ 18 pc and possibly at

1073: even larger distances from the X-ray bubble. All these objects are

1074: preferably located in an extended feature north of the bar of N~66,

1075: which is outlined by the emission from the gas, observed in H\alp\ and

1076: [{\sc O~iii}], as well as from the dust, as it is seen in CO, H$_2$ and

1077: 8\micron\ maps. Lee \& Chen (2007) findings in the $\lambda$~Ori region,

1078: Ori~OB1, and Lac~OB1 associations suggest that intermediate-mass HAeBe

1079: stars are formed in the denser parts of the cloud, while the low-mass

1080: T~Tauri stars are formed at the outskirts near the photo-evaporating

1081: cloud layers. Our photometry in H{\alp} (Paper~III) showed that indeed

1082: stars with H{\alp}-excess (signature of candidate HAeBe stars) are

1083: located in the area of the northern compact PMS clusters. Based on the

1084: aforementioned observations in different wavelengths, the extended

1085: feature north of the bar of N~66, which hosts this star formation, may

1086: be considered as part of a large arc centered on the SNR

1087: (Fig.~\ref{fig-imaxoi} {\sl right}), which extends also to the south and

1088: in the bar. In this case, one may conclude that its northern part is

1089: projected closer to us, outside N~66, since the dust is well outlined by

1090: the 8\micron\ emission (seen in red in Fig.  \ref{fig-imaxoi}; left),

1091: while its southern part is embedded {\sl in} the nebula (as suggested by

1092: the schematic of Fig.~\ref{fig-imaxoi}; right). 

1093: 

1094: On the other hand, considering that the northern part of this feature as

1095: seen in CO and IR maps appears to be more linear with small curvature

1096: (see e.g. maps by Contursi et al. 2000; Rubio et al. 2000; Simon et al.

1097: 2007), one may argue that it may not bear a morphological connection to

1098: the triggering sources to the east. Against this argument stands the

1099: fact that this linearity is confined only in the northernmost part of

1100: the structure, while its southern part shows a curvature to the east,

1101: and fits very well to the alignment of emission from both gas and dust

1102: towards the south (see e.g. maps by Contursi et al. 2000; Rubio et al.

1103: 2000; Simon et al. 2007; Paper~III). The latter appears to be well

1104: related to the positions of the {\sc H~ii} regions and IR peaks, which

1105: are located {\sl in} the south-eastern part of the bar of N~66.

1106: Consequently, the northern feature can be considered to be extended to

1107: the south and to become {\sl embedded} in the south-eastern part of the

1108: bar, forming a large arc-like feature, which covers the whole eastern

1109: part of N~66 nebula (Fig.\ref{fig-imaxoi}), and which is centered on the

1110: SNR. Further support to this argument provides the outstanding

1111: coincidence of this arc structure and the non-thermal emission related

1112: to the SNR~B0057$-$724, shown in the H\alp\ and {\sc [S~ii]}

1113: continuum-subtracted images by Reid et al. (2006; their figures 1 and

1114: 2). Under these circumstances the dust emission from the northernmost

1115: linear part of the structure may not come from only a ``strip'' of

1116: molecular material, but from a whole curved surface of the wind-blown

1117: bubble, projected in front of the bar of N~66. It should be noted,

1118: though, that an equally valid description of the morphology of the

1119: 8\micron\ emission is that of a partial arc with mild curvature pointing

1120: west, which appears connected to the N~66 bar through a clump of

1121: emission.

1122: 

1123: In any case, the proposed scenario draws the picture of a unique case of

1124: complexity in the star formation history of NGC~346/N~66, the brightest

1125: {\sc H~ii} region in the SMC. According to this scenario it is the

1126: collaborative (or competitive) action of two major energy sources that

1127: shapes the current star formation process in this extraordinary region.

1128: The first source, located at the center of the nebula N~66, is the

1129: bright stellar content of the association NGC~346 and affects mostly the

1130: bar of the nebula. The second source, located at the eastern part of the

1131: region, is associated with the bright X-ray bubble of a SNR. We propose

1132: that the massive progenitor and possibly two other massive stellar

1133: objects have triggered the most recent star formation at the northern

1134: part of the region away from the bar of N~66 within that last few Myr. 

1135: Observations of the gas kinematics provide further support to this

1136: hypothesis. Under these circumstances, the identified compact PMS

1137: sub-clusters in the region {\sl cannot} all be the product of

1138: spontaneous cloud collapse and fragmentation (see e.g. Elmegreen \& Lada

1139: 1977). They are rather the product of star formation inside {\sl and}

1140: outside the bar of N~66, induced by wind-shock waves (e.g. Vanhala \&

1141: Cameron 1998) or ionization shock fronts (e.g. Kessel-Deynet \& Burkert

1142: 2003) driven by OB stars into the surrounding cloud.

1143: 

1144: %The results of these authors, according to which an age spread of

1145: %several Myr exists between the star groups that formed in the sequence,

1146: %are in contrast to the scenario proposed by Elmegreen \& Lada (1977) and

1147: %Lada (1987) for which massive stars are formed in shocked cloud layers

1148: %by triggering, whereas low-mass stars are formed spontaneously via cloud

1149: %collapse and fragmentation. 

1150: 

1151: \acknowledgements We wish to thank Lars Christensen (ESA) for producing

1152: the beautiful color-composite image of NGC~346 presented in Figure

1153: \ref{fig-imaxoi} (left). We are also grateful to Snezana

1154: Stanimirovi\'{c} for providing the original {\sl ATCA} and {\sl Parkes}

1155: combined {\sc H~i} data-cubes of the SMC. D.A. Gouliermis kindly

1156: acknowledges the support of the German Research Foundation (Deutsche

1157: Forschungsgemeinschaft - DFG) through the individual grant GO 1659/1-1.

1158: Based on observations made with ESO Telescopes at the La Silla

1159: Observatories under program ID 56.C-0379, with XMM-Newton, an ESA

1160: science mission with instruments and contributions directly funded by

1161: ESA Member States and NASA, and with the Spitzer Space Telescope, which

1162: is operated by the Jet Propulsion Laboratory, California Institute of

1163: Technology under a contract with NASA. The Australia Telescope Compact

1164: Array and the Parkes Telescope are part of the Australia Telescope,

1165: which is funded by the Commonwealth of Australia for operation as a

1166: National Facility managed by CSIRO.

1167: 

1168: %%%%%%%%%%%%%%%%%%%%%%%%%%%  BIBLIOGRAPHY  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1169: %\newpage

1170: \begin{thebibliography}{}

1171: 

1172: \bibitem[]{906} Bolatto, A., et al. 2007,ApJ, 655, 212

1173: 

1174: \bibitem[]{908} Bonnell, I. A., \& Bate, M. R. 2002, MNRAS, 336, 659

1175: 

1176: \bibitem[]{910} Bonnell, I. A., Bate, M. R., \& Vine, S. G. 2003, MNRAS,

1177: 343, 413

1178: 

1179: \bibitem[]{913} Brice\~{n}o, C., et al. 2007, in ``Protostars and Planets V'', 

1180: Eds. B. Reipurth, D. Jewitt, and K. Keil, University of Arizona Press, 

1181: Tucson, p. 345

1182: 

1183: \bibitem[]{914} Cappa, C.~E., \& Benaglia, P.\ 1998, \aj, 116, 1906

1184: 

1185: \bibitem[]{915} Carlson, L.~R., et al.\ 2007, \apjl, 665, L109

1186: 

1187: \bibitem[]{916} Chlebowski, T., \& Garmany, C.~D.\ 1991, \apj, 368, 241 

1188: 

1189: \bibitem[]{917} Chu, Y.-H., \& Kennicutt, R.~C., Jr.\ 1988, \aj, 95, 1111 

1190: 

1191: \bibitem[]{919} Chu, Y.-H.\ 1997, \aj, 113, 1815 

1192: 

1193: \bibitem[]{920} Chu, Y.-H., et al.\ 2005, \apjl, 634, L189

1194: 

1195: \bibitem[]{921} Contursi, A., et al. 2000, A\&A, 362, 310

1196: 

1197: \bibitem[]{923} Crowther, P., \& Smartt, S.\ 2007, Astronomy and

1198: Geophysics, 48, 35

1199: 

1200: \bibitem[]{926} Danforth, C.~W., Sankrit, R., Blair, W.~P., Howk, J.~C., \&

1201: Chu, Y.-H.\ 2003, \apj, 586, 1179

1202: 

1203: \bibitem[]{929} de Boer, K. S., \& Savage, B. D. 1980, ApJ, 238, 86

1204: 

1205: \bibitem[]{931} Deharveng, L., Lefloch, B., Massi, F., Brand, J., Kurtz,

1206: S., Zavagno, A., \& Caplan, J.\ 2006, A\&A, 458, 191

1207: 

1208: \bibitem[]{932} Drake, S.~A.\ 1986, Cool Stars, Stellar Systems and the

1209: Sun, 254, 369

1210: 

1211: 

1212: \bibitem[]{934} Elmegreen, B.~G., \& Lada, C. J.\ 1977, AJ, 214, 725

1213: 

1214: \bibitem[]{936} Evans, C.~J., Lennon, D.~J., Smartt, S.~J., \& Trundle, C.\ 

1215: 2006, A\&A, 456, 623

1216: 

1217: \bibitem[]{939} Fazio, G., et al. 2004, ApJS, 154, 10

1218: 

1219: \bibitem[]{940} Fukuda N. \& Hanawa T. 2000, ApJ, 533, 911

1220: 

1221: \bibitem[]{941} Fullerton, A.~W., Massa, D.~L., \& Prinja, R.~K.\ 2006,

1222: \apj, 637, 1025

1223: 

1224: \bibitem[]{942} Girardi, L., Bertelli, G., Bressan, A., Chiosi, C.,

1225: Groenewegen, M.~A.~T., Marigo, P., Salasnich, B., \& Weiss, A.\ 2002,

1226: A\&A, 391, 195

1227: 

1228: \bibitem[]{943} Gouliermis, D.~A., Dolphin, A.~E., Brandner, W., \& Henning, 

1229: T.\ 2006, ApJS, 166, 549 (Paper~I)

1230: 

1231: \bibitem[]{946} Gouliermis, D.~A., Quanz, S.~P., \& Henning, T.\ 2007, ApJ, 

1232: 665, 306

1233: 

1234: \bibitem[]{947} Guerrero, M.~A., \& Chu, Y.-H.\ 2008, ApJS, 177, 216

1235: 

1236: \bibitem[]{949} Heger, A., Fryer, C. L., Woosley, S. E., Langer, N., \&

1237: Hartmann, D. H. 2003, ApJ, 591, 288

1238: 

1239: \bibitem[]{952} Henize, K.~G.\ 1956, ApJS, 2, 315   

1240: 

1241: \bibitem[]{954} Hennekemper, E., et al.\ 2008, ApJ, 672, 914 (Paper~III)

1242: 

1243: \bibitem[]{956} Hoopes, C. G., Sembach, K. R., Howk, J. C., \& Blair, W. P. 

1244: 2001, ApJ, 558, L35

1245: 

1246: \bibitem[]{959} Kessel-Deynet, O., \& Burkert, A.\ 2003, MNRAS, 338, 545 

1247: 

1248: \bibitem[]{961} Klessen, R. S., \& Burkert, A. 2000, ApJS, 128, 287

1249: 

1250: \bibitem[]{963} Koenigsberger, G., et al.\ 2000, ApJ, 542, 428

1251: 

1252: \bibitem[]{965} Koo, B.-C., et al.\ 2008, ApJL, 673, L147 

1253: 

1254: \bibitem[]{967} Lamers, H.~J.~G.~L.~M.\ 1989, IAU Colloq.~113: Physics of

1255: Luminous Blue Variables, 157, 135

1256: 

1257: \bibitem[]{969} Lamers, H.~J.~G.~L.~M., Snow, T.~P., \& Lindholm, D.~M.\

1258: 1995, \apj, 455, 269

1259: 

1260: \bibitem[]{970} Law, D.~R., DeGioia-Eastwood, K., \& Moore, K.~L.\ 2002,

1261: ApJ, 565, 1239

1262: 

1263: \bibitem[]{973} Lee, Hsu-Tai, \& Chen W. P. 2007, ApJ, 657, 884

1264: 

1265: \bibitem[]{974} Leitherer, C., Robert, C., \& Drissen, L.\ 1992, \apj,

1266: 401, 596

1267: 

1268: \bibitem[]{975} Massey, P., Parker, J. W., Garmany, C. D. 1989, AJ, 98, 1305

1269: 

1270: \bibitem[]{977} Naz\'{e}, Y., et al. 2002, ApJ, 580, 225

1271: 

1272: \bibitem[]{979} Naz\'{e}, Y., Manfroid, J., Stevens, I. R., Corcoran, 

1273: M. F., Flores, A.\ 2004, ApJ, 608, 208

1274: 

1275: \bibitem[]{982} Niemela, V. S., Marraco, H. G., \& Cabanne, M. L. 1986, PASP,

1276: 98, 1133

1277: 

1278: \bibitem[]{985} Nota, A., et al.\ 2006, ApJL, 640, L29

1279: 

1280: \bibitem[]{987} Pastorello, A., et al.\ 2007, Nature, 447, 829

1281: 

1282: \bibitem[]{989} Pollock, A. M. T. 2002, in ASP Conf. Ser., 260, Interacting

1283: Winds from Massive Stars, ed. A. F. J. Moffat \& N. St Louis (San

1284: Francisco: ASP), 363

1285: 

1286: \bibitem[]{993} Preibisch, T., \& Zinnecker, H.\ 2007, IAU Symposium, 237,

1287: 270

1288: 

1289: \bibitem[]{994} Puls, J., et al.\ 1996, \aap, 305, 171 

1290: 

1291: \bibitem[]{996} Reid, W.~A., Payne, J.~L., Filipovi{\'c}, M.~D., Danforth, 

1292: C.~W., Jones, P.~A., White, G.~L., \& Staveley-Smith, L.\ 2006, MNRAS, 

1293: 367, 1379

1294: 

1295: \bibitem[]{1000} Rubio, M., Contursi, A., Lequeux, J., Probst, R.,

1296: Barb\'{a}, R., Boulanger, F., Cesarsky, D., Maoli, R. 2000, A\&A, 359,

1297: 1139

1298: 

1299: \bibitem[]{1004} Sabbi, E., et al. 2007, AJ, 133, 44 (erratum 133, 2430;

1300: SSN07)

1301: 

1302: \bibitem[]{1006} Sanduleak, N.\ 1968, AJ, 73, 34

1303: 

1304: \bibitem[]{1008} Sasaki, M., Pietsch, W., Haberl, F.\ 2003, A\&A, 403, 901

1305: 

1306: \bibitem[]{1010} Schaerer, D. \& de Koter, A. 1997, A\&A, 322, 598 

1307: 

1308: \bibitem[]{1012} Simon, J.~D., et al.\ 2007, ApJ, 669, 327

1309: 

1310: \bibitem[]{1013} Shtykovskiy, P. \& Gilfanov, M.\ 2005, MNRAS, 362, 879

1311: 

1312: \bibitem[]{1014} Smith, C., Leiton, R., \& Pizarro, S.\ 2000, Stars, Gas

1313: and Dust in Galaxies: Exploring the Links, 221, 83

1314: 

1315: \bibitem[]{1015} Smith, N., et al. 2003, AJ, 125, 1458 

1316: 

1317: \bibitem[]{1016} Smith, N., \& Owocki, S. P. 2006, ApJ, 645, L45

1318: 

1319: \bibitem[]{1017} Smith, N., et al.\ 2007, ApJ, 666, 1116

1320: 

1321: \bibitem[]{1018} Smith, N.\ 2007, in Clumping in Hot-Star Winds,

1322: Proceedings of an International Workshop held in Potsdam, Germany, 18-22

1323: June 2007, ed. W.-R. Hamann, A. Feldmeier \& L. Oskinova, Electronic

1324: publication of the University Potsdam: 

1325: http://nbn-resolving.de/urn:nbn:de:kobv:517-opus-13981, p. 27

1326: (arXiv:0710.3430)

1327: 

1328: \bibitem[]{1019} Stahler, S. W., \& Palla, F. 2005, The Formation of Stars,

1329: ISBN 3-527-40559-3, Wiley-VCH

1330: 

1331: \bibitem[Stanimirovic et al.(1999)]{1999MNRAS.302..417S}

1332: Stanimirovi\'{c}, S., Staveley-Smith, L., Dickey, J.~M., Sault, R.~J.,

1333: \& Snowden, S.~L.\ 1999, \mnras, 302, 417

1334: 

1335: \bibitem[Staveley-Smith et al.(1997)]{1997MNRAS.289..225S} Staveley-Smith, 

1336: L., Sault, R.~J., Hatzidimitriou, D., Kesteven, M.~J., 

1337: \& McConnell, D.\ 1997, \mnras, 289, 225 

1338: 

1339: \bibitem[]{1021} Tang, S., \& Wang, Q.~D.\ 2005, ApJ, 628, 205

1340: 

1341: \bibitem[]{1023} Vacca, W. D., Garmany, C. D., \& Shull, J. M. 1996, ApJ,

1342: 460, 914

1343: 

1344: \bibitem[]{1024} van der Hucht, K.~A.\ 1992, \aapr, 4, 123 

1345: 

1346: \bibitem[]{1026} Vanhala, H. A. T., \& Cameron, A. G. W. 1998, ApJ, 508, 291

1347: 

1348: \bibitem[]{1027} Volk, K., \& Kwok, S. 1985, A\&A, 153, 79 

1349: 

1350: \bibitem[]{1028} Walborn, N. R. 1978, APJ, 224, L133

1351: 

1352: \bibitem[]{1030} Walborn, N. R., \& Parker, J. W. 1992, ApJ, 399, L87

1353: 

1354: \bibitem[]{1032} Walborn, N. R., et al. 2000, PASP, 112, 1243

1355: 

1356: \bibitem[]{1034} Walborn, N.~R., Ma{\'{\i}}z-Apell{\'a}niz, J., \&

1357: Barb{\'a}, R.~H.\ 2002, AJ, 124, 1601

1358: 

1359: \bibitem[]{1037} Wessolowski, U. 1996, in MPE Rep. 263, R\"{o}ntgenstrahlung

1360: from the Universe, ed. H. U. Zimmerman, J. E. Tr\"{u}mper, \& H. Yorke,

1361: 75

1362: 

1363: \bibitem[]{1044} Ye, T., Turtle, A. J., \& Kennicutt, R. C., Jr. 1991,

1364: MNRAS, 249, 722

1365: 

1366: \bibitem[]{1045} Zavagno, A., Deharveng, L., Comer{\'o}n, F., Brand, J.,

1367: Massi, F., Caplan, J., \& Russeil, D.\ 2006, A\&A, 446, 171

1368: 

1369: %\bibitem[]{1047} Lada, C. J. 1987, in IAU Symp. 115, Star Forming Regions,

1370: %ed. J. Jugaku \& M. Peimbert (Dordrecht: Reidel), 1

1371: 

1372: %\bibitem[]{1050} Kontizas, E., Kontizas, M., Gouliermis, D., Dapergolas, A.,

1373: %Korakitis, R., Morgan, D. H. 1999, IAUS, 190, 410

1374: 

1375: %\bibitem[]{1053} Preibisch, T., \& Zinnecker, H.\ 1999, AJ, 117, 2381

1376: 

1377: %\bibitem[]{1055} Vel\'{a}zquez, P. F., Koenigsberger, G., \& Raga, A. C.

1378: %2003, ApJ, 584, 284

1379: 

1380: %\bibitem[]{1058} Heydari-Malayeri, M., Royer, P., Rauw, G., \& Walborn, N.

1381: %R.  2000, A\&A, 361, 877 (erratum 364, 923)

1382: 

1383: %\bibitem[]{1041} Williams, R.~M., Chu, Y.-H., Dickel, J.~R., Gruendl, R.~A.,

1384: %Shelton, R., Points, S.~D., \& Smith, R.~C.\ 2004, ApJ, 613, 948

1385: 

1386: \end{thebibliography}

1387: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1388: 

1389: \end{document}

1390: 

1391: