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

2: \usepackage{natbib,epsfig}

3: 

4: \begin{document}

5: 

6: \newcommand{\NGC}{\mbox{\protect\small NGC\hspace{.15em}\normalsize}}

7: \newcommand{\hii}{{\rm H}\,{\sc ii}}

8: \newcommand{\uchii}{UCH\,{\sc ii}}

9: \newcommand{\etal}{{\it et. al.}}

10: \newcommand{\hcoplus}{HCO$^{+}$ ${(1-0)}$}

11: \newcommand{\co}{$^{12}$C$^{16}$O}

12: \newcommand{\cyano}{HC$_3$N}

13: \newcommand{\siomaser}{SiO$(J=1\rightarrow0,\nu=1)$}

14: \newcommand{\sioone}{SiO$(J=1\rightarrow 0)$}

15: \newcommand{\siotwo}{SiO$(J=2\rightarrow 1)$}

16: \newcommand{\siothree}{SiO$(J=3\rightarrow 2)$}

17: \newcommand{\siofive}{SiO$(J=5\rightarrow 4)$}  

18: \newcommand{\siofi}{SiO(5-4)}  

19: \newcommand{\sioeight}{SiO$(J=8\rightarrow 7)$}  

20: \newcommand{\sot}{SO$_2$}

21: \newcommand{\nhone}{NH$_3$(1,1)}

22: \newcommand{\nhtwo}{NH$_3$(2,2)}

23: \newcommand{\nhthree}{NH$_3$(3,3)}

24: \newcommand{\ammonia}{NH$_3$}

25: \newcommand{\methanol}{CH$_3$OH}

26: \newcommand{\methanolline}{CH$_3$OH $5_{1,4}-4_{2,2}$E}

27: \newcommand{\methylcyanide}{CH$_3$CN}

28: \newcommand{\thcoone}{$^{13}$CO ${(J=1\rightarrow0)}$}

29: \newcommand{\thco}{$^{13}$CO}

30: \newcommand{\coone}{CO(1-0)}

31: \newcommand{\cotwo}{CO ${(J=2\rightarrow1)}$}

32: \newcommand{\ceioone}{C$^{18}$O ${(J=1\rightarrow0)}$}

33: \newcommand{\ceio}{C$^{18}$O}

34: \newcommand{\dcn}{DCN(3-2)}

35: \newcommand{\hcccn}{HCCCN(25-24)}

36: \newcommand{\hcctcn}{HCC$^{13}$CN(24-23)}

37: \newcommand{\hctccn}{HC$^{13}$CCN(24-23)}

38: \newcommand{\htwosline}{H$_2$S $2_{2,0}-2_{1,1}$}

39: \newcommand{\htwos}{H$_2$S}

40: \newcommand{\sotwo}{SO$_2 \, 22_{2,20}-22_{1,21}$}

41: \newcommand{\thfsotwo}{$^{34}$SO$_2 \, 11_{1,11}-10_{0,10}$}

42: \newcommand{\htwoo}{H$_2$O}

43: \newcommand{\htw}{H$_2$}

44: \newcommand{\nh}{NH$_3$}

45: \newcommand{\msun}{M$_{\sun}$}

46: \newcommand{\lsun}{L$_{\sun}$}

47: \newcommand{\cs}{C$^{34}$S ${(J=3\rightarrow2)}$}

48: \newcommand{\um}{$\mu$m}

49: \newcommand{\percc}{cm$^{-3}$}

50: \newcommand{\persqcm}{cm$^{-2}$}

51: \newcommand{\kms}{km~s$^{-1}$}

52: \newcommand{\vlsr}{\hbox{${\rm v}_{\rm lsr}$}}

53: \newcommand{\cstwo}{CS $(J=2\rightarrow1)$ }

54: \newcommand{\Msun}{M$_\odot$}

55: \newcommand{\jyperbeam}{\rm{Jy beam$^{-1}$}}

56: \newcommand{\mjyperbeam}{\rm{mJy beam$^{-1}$}}

57: \newcommand{\jyperbeamkms}{\rm{Jy Beam$^{-1}$}km~s$^{-1}$}

58: \newcommand{\arcseconds}{''}

59: \newcommand{\hthonea}{H$31\alpha$}

60: \newcommand{\htha}{H$30\alpha$}

61: \newcommand{\htwsixa}{H$26\alpha$}

62: \newcommand{\htwonea}{H$21\alpha$}

63: 

64: \title{Mapping the Outflow from G5.89-0.39 in \siofive}

65: 

66: \author{

67: P. K. Sollins\altaffilmark{1,4}

68: T. R. Hunter\altaffilmark{1},

69: J. Battat\altaffilmark{1},

70: H. Beuther\altaffilmark{1},

71: P. T. P. Ho\altaffilmark{1,2},

72: J. Lim\altaffilmark{2},

73: S. Y. Liu\altaffilmark{2},

74: N. Ohashi\altaffilmark{3},

75: T. K. Sridharan\altaffilmark{1},

76: Y. N. Su\altaffilmark{2},

77: J.-H. Zhao\altaffilmark{1},

78: Q. Zhang\altaffilmark{1}

79: }

80: 

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

82: Street, Cambridge, MA, 02138}

83: 

84: \altaffiltext{2}{ASIAA, No. 1, Roosevelt Road, Section 4, Taipei 106,

85: Taiwan, R.O.C.}

86: 

87: \altaffiltext{3}{ASIAA, 645 North A'ohoku Place, Hilo, HI, 96720}

88: 

89: \altaffiltext{4}{psollins@cfa.harvard.edu}

90: 

91: \begin{abstract}

92: 

93: We have mapped the ultracompact \hii\ region, G5.89-0.39, and its

94: molecular surroundings with the Submillimeter Array at $2\arcseconds.8

95:  \times 1\arcseconds.8$ angular resolution in 1.3 mm

96: continuum, \siofive, and eight other molecular lines. We have resolved

97: for the first time the highly energetic molecular outflow in this

98: region. At this resolution, the outflow is definitely bipolar and

99: appears to originate in a 1.3 mm continuum source. The continuum

100: source peaks in the center of the \hii\ region. The axis of the

101: outflow lines up with a recently discovered O5V star.

102: 

103: \end{abstract}

104: 

105: \keywords{stars: formation --- ISM: jets and outflows --- ISM: individual (G5.89-0.39) --- \hii\, regions}

106: 

107: \section{Introduction} \label{section:int}

108: 

109: G5.89-0.39 (G5.89) is a shell-like ultracompact (UC) \hii\ region

110: \citep{woo89b} which was recently found to contain an O5V star via

111: near-IR imaging \citep{fel03}. \citet{aco98} have used proper motion

112: measurements of the expansion of G5.89 to determine its distance,

113: $\sim 2$ kpc, its size, 0.01 pc, and its dynamical age, 600

114: years. Associated with G5.89 is a molecular outflow \citep{har88} with

115: a mass of 77 \msun\ and an energy of $5 \times 10^{47}$ ergs

116: \citep[hereafter, AWC]{aco97}. While many \uchii\ regions have been

117: observationally associated with molecular outflows

118: \citep{sne90,she96a}, the question of whether an \uchii\ region can

119: itself be or contain the source of a bipolar outflow remains in

120: dispute.  Bipolar outflows are generally understood to be a signpost

121: of ongoing accretion, while the presence of an \uchii\ region has been

122: understood by some as a sign that accretion has ceased

123: \citep{gar99}. Recent observational and theoretical results have shown

124: that the presence of an \uchii\ region does not necessarily shut off

125: accretion \citep{ket02a,ket02b}. But since massive stars and their

126: \uchii\ regions tend to form in very crowded fields, high resolution

127: imaging is necessary to associate the origin of any particular bipolar

128: molecular outflow with an \uchii\ region, and not a distinct, nearby

129: protostar. This determination is possible with and well suited to the

130: capabilities of the new Submillimeter Array\footnote{The Submillimeter

131: Array is a joint project between the Smithsonian Astrophysical

132: Observatory and the Academia Sinica Institute of Astronomy and

133: Astrophysics, and is funded by the Smithsonian Institution and the

134: Academia Sinica. For more on the SMA and its specifications see

135: \citet{ho03}.} (SMA) on Mauna Kea.

136: 

137: We have used the SMA to map the region of G5.89-0.39 in the \siofive\

138: line, known to trace outflows and this outflow in particular

139: \citep[AWC]{cod99}. Because of the spectral capabilities of the SMA,

140: we were simultaneously able to map the region in nine spectral lines

141: and broadband continuum. The aim of the project was to improve on the

142: previous $18\arcseconds$ resolution map in \siofi\ of AWC, and also

143: to detect in dust and molecular lines any other massive protostars in

144: the vicinity which might be sources of the outflow.

145: 

146: \section{Observations} \label{section:observations}

147: 

148: G5.89-0.39 was observed with the SMA on the night of 2003 13 July with

149: 160 minutes on source. The pointing center of the observations was

150: $\alpha(2000)=\rm{18^h00^m30^s.32}, \, \delta(2000)=-24^o

151: 04'00''.48$. Five antennas were used in the array, giving baselines

152: with projected lengths ranging from 10.2 meters to 116 meters

153: resulting in a $2\arcseconds.8 \times 1\arcseconds.8$ synthesized

154: beam. The FWHP of the primary beam was $50\arcseconds$. We observed

155: the quasar NRAO530 for phase calibration every 35 minutes ($\sim

156: 13$~degrees from the source, $S_{217 \rm{GHz}} \sim 2.4$~Jy), and the

157: data were self-calibrated for phase correction at 1 minute intervals

158: using the continuum as a model. We observed Mars and Neptune for

159: bandpass and flux calibration respectively. The average system

160: temperature was 180~K. We had 1~GHz (1300~\kms) spectral coverage in

161: each sideband with 0.8125~MHz (1.1~\kms) spectral resolution. The

162: resulting noise in the 227 GHz (1.3 mm) continuum map was $\sim 20$

163: \mjyperbeam\ with a bandwidth of 900 MHz. The theoretical noise limit

164: for this map is 1.7 \mjyperbeam, so the actual noise level is probably

165: limited by dynamic range. The noise in a single spectral channel map

166: was $\sim 0.12$ \jyperbeam, which is only a factor of two worse than

167: the theoretical limit of 0.057 \jyperbeam. We smoothed our data to 3.3

168: \kms\ resolution for most of the analysis presented below.

169: 

170: \section{Results} \label{section:results}

171: 

172: 	\subsection{\siofive\ Map and Other Lines} \label{subsection:lines}

173: 

174: Our map of the integrated \siofi\ line emission in Figure

175: \ref{figure:linemaps} clearly resolves and separates the red- and

176: blue-shifted lobes of the outflow. The extent of the \siofi\ emission

177: is consistent with the unresolved \siofi\ emission of AWC. We detect

178: \siofi\ emission which is much stronger and more extended than the

179: \sioone\ of AWC, probably due to excitation effects in the

180: outflow. Figure \ref{figure:linemaps} also shows the location of the

181: O5 star detected by \citet{fel03} and a line connecting the peaks of

182: the outflow lobes. Both lobes are more than two synthesized beams

183: across in the direction perpendicular to the outflow axis. At the

184: peaks of the red- and blue-shifted lobes, the velocity integrated flux

185: in our \siofi\ spectra are 256~K~\kms and 306~K~\kms,

186: respectively. Assuming LTE and optically thin emission with a uniform

187: rotation temperature of 100~K, from AWC's analysis of SiO, CO and

188: \ammonia, we calculate that the column density of SiO at the peak of

189: the red- and blue-shifted lobes are $5.4 \times 10^{14}$~\persqcm\ and

190: $6.4 \times 10^{14}$~\persqcm\ respectively. These are higher than the

191: column densities determined by AWC, probably because we have much

192: higher spatial resolution and are able to avoid beam dilution. The

193: brightest channel in either of the two spectra considered is 16.4~K,

194: in which channel the optical depth would be less than 0.2, so our

195: assumptions in this calculation are self-consistent.

196: 

197: Figure \ref{figure:posvel} shows a position-velocity (P-V) diagram of

198: the \siofi\ emission. The P-V cut goes through the peaks of the

199: \siofi\ lobes. There is no obvious dependence of velocity on

200: position within each lobe, as is seen in some bipolar outflows from

201: low- to intermediate-mass young stars \citep{lad96,arc01}. The peak

202: emission is at low relative velocities, $\pm 5$ \kms, and tails off

203: towards higher relative velocities. The \siofi\ emission exceeding

204: $3\sigma$ ranges from \vlsr = -20 \kms\ to +40 \kms. Red- and

205: blue-shifted \coone\ and \hcoplus\ emission have been detected out to

206: $\sim 30\arcseconds$ from the \uchii\ region \citep{wat02}, but the

207: \siofi\ emission extends only $\sim 5\arcseconds$ in either

208: direction along the outflow giving a dynamical time-scale of 1600

209: years (separation divided by velocity). This number, however, is

210: uncertain, since the line-of-sight velocity is a lower limit on the

211: three dimensional velocity and the projected separation is a lower

212: limit on the three dimensional separation.

213: 

214: In total we detected 9 spectral lines. \siofi, \htwosline, \hcccn,

215: \sotwo, and \thfsotwo, all show the outflow to varying degrees.

216: \hctccn, \hcctcn, and \methanolline, show only marginally resolved

217: emission all at the same location in the Northeast lobe of the outflow

218: and all at the ambient velocity of \vlsr=10 \kms. \dcn\ was also

219: detected but was difficult to map, probably due to missing flux.

220: 

221: 	\subsection{227 GHz Continuum} \label{subsection:continuum}

222: 

223: The first panel of Figure \ref{figure:linemaps} shows our 227 GHz

224: continuum emission map. 900 MHz of line-free channels from the upper

225: sideband were averaged to obtain the continuum visibilities. The map

226: shows the full extent of the detected 1.3 mm continuum emission. We

227: detect no other 1.3 mm continuum sources anywhere in our

228: $50\arcseconds$ field of view down to a $3\sigma$ detection limit of

229: 60 mJy. The second panel in Figure \ref{figure:linemaps} shows the

230: 8.46 GHz continuum map of \citet{woo89b} in grayscale overlaid with

231: our 1.3 mm map in contours, and the stellar source of

232: \citet{fel03}. The peak in 1.3 mm emission is about $1\arcseconds.5$

233: removed from the star. The strongest 1.3 mm emission is extended

234: nearly perpendicular to the outflow axis. The faint extensions of the

235: 1.3 mm emission follow the axis of the outflow and are probably

236: emission from dust in the outflowing molecular gas, as has been

237: observed in L1157 \citep{gue03}.

238: 

239: The total flux detected in the 1.3 mm map is 8.8 Jy, integrated over a

240: central square of $16\arcseconds \times 16\arcseconds$. Continuum data

241: from 21 cm to 350 \um\ are plotted in Figure \ref{figure:sed} showing

242: the SED of this region. Using the distance of 2.0 kpc determined by

243: \citet{aco98}, we did a least squares fit of a standard free-free

244: spectrum \citep{gor88} to the data below 50 GHz. We get an electron

245: temperature of 8000 K, angular diameter of $3\arcseconds.1$

246: corresponding to a physical size of 6000 AU (0.03 pc), an electron

247: density of $1.4 \times 10^5$ \percc, and an emission measure of $6

248: \times 10^8$ pc cm$^{-6}$. For the dust-dominated, short-wavelength

249: regime, we simply extended the existing ``cold-component'' fit from

250: \citet{hun00}. The sum of the two spectra is the solid line in Figure

251: \ref{figure:sed} and fits our data point well. Based on these fits, we

252: predict 5.6 Jy from free-free emission and 3.2 Jy from dust

253: emission. Following \citet{hun00}, assuming a dust emissivity of

254: $Q(\nu=227 \rm{GHz}) = 2.0 \times 10^{-5}$, a dust temperature of 100

255: K as above, and a total mass to dust mass ratio of $M_{total}/M_{dust}

256: = 100$ \citep{tes98}, we calculate a total mass of 83 \msun\

257: associated with the dust emission. If we define the 2 \mjyperbeam\

258: contour of the free-free map as a boundary, the 1.3 mm map contains

259: 8.2 Jy of flux inside the boundary and 0.7 Jy outside. Since the flux

260: outside is largely from the extensions which follow the outflow, we

261: estimate the mass of gas associated with the dust in the outflow to be

262: at least 18 \msun. This number is an approximate lower limit since

263: much of outflow overlaps with the \uchii\ region, but since the

264: effects of the outflow on abundance are unknown it is not a hard lower

265: limit. The remaining 65 \msun\ is projected against the \uchii\

266: region.

267: 

268: \section{Discussion} \label{section:discussion}

269: 

270: We believe the origin of the outflow is the 1.3 mm continuum source

271: for two reasons. First, the slightly extended 1.3 mm continuum peak

272: lies on the projected outflow axis, halfway between the peaks of the

273: two outflow lobes. Second, the 1.3 mm continuum source seems to show a

274: physical connection to both outflow lobes in the form of the

275: extensions of the continuum along the outflow axis. Apparently the

276: dust is being swept up in both lobes of the outflow (as seen in L1157

277: by \citet{gue03}), and thus it seems that the outflow is emanating

278: from the 1.3 mm continuum source.

279: 

280: The question of whether the outflow actually originates {\it inside}

281: the \uchii\ region is somewhat thornier. One possibility is that the

282: 1.3 mm continuum source lines up with the \uchii\ region by chance,

283: and is a distinct high-mass protostar with its own dusty envelope, not

284: physically connected to the \uchii\ region, but part of the same young

285: cluster. Alternatively, the dust and free-free continuua could be from

286: a single source, one dusty, hot molecular core surrounding an embedded

287: \uchii\ region.  In this case, it seems the outflow would have to come

288: from inside the \uchii\ region in order for both sides of the outflow

289: to be seen in the dust continuum. We cannot distinguish between these

290: two cases, and neither seems unlikely. However, we can say that the

291: known O5 star is a bad candidate to be the source of the outflow since

292: it is not equidistant from the outflow lobes.

293: 

294: We cannot eliminate any model chronologies by comparing

295: timescales. The outflow has a dynamical age of 1600 years, compared to

296: just 600 for the \uchii\ region. This might lead one to believe that

297: the outflow is truly older than the \uchii\ region. However, any

298: ``quenching'' of the \uchii\ region by infalling material would tend

299: to make the dynamical age of the \hii\ region an underestimate of its

300: true age, perhaps by a great deal \citep{ket02b}. The gas in the

301: outflow was probably accelerated before the \uchii\ region began to

302: expand, but not necessarily before the \uchii\ region was created.

303: 

304: The central unknown about the outflow is its inclination. The fact

305: that the two lobes hardly overlap at all on the plane of the sky and

306: the fact that the \siofi\ emission is extended perpendicular to the

307: outflow axis might lead one to believe that the outflow has a wide

308: opening angle, and lies primarily in the plane of the sky. But these

309: two facts could just as easily be interpreted as a highly collimated

310: outflow, like those seen around low-mass proto-stars \citep{bac91},

311: whose axis is largely along the line of sight. Including only the

312: line-of-sight velocity, the outflow is already quite energetic, so

313: adding a large component in the plane of the sky would make this a

314: remarkable outflow, indeed. Some studies of regions of massive star

315: formation have found poor collimation of outflows

316: \citep{ric00,rid01}. But \citet{beu02a} found that 15 outflows with

317: apparently low collimation were consistent with the very high

318: collimations seen in low mass cases, with the discrepancy due to low

319: resolution. Higher resolution interferometer studies have confirmed

320: the presence of highly collimated outflows in some regions of

321: high-mass star formation \citep{beu02b}.

322: 

323: The velocity structure of the \siofi\ is intriguing and may be a clue

324: as to the outflow orientation. In both lobes, the spectra peak at low

325: relative velocity and tail off towards much higher relative

326: velocity. In the case of outflows from low-mass protostars, the SiO

327: emission is concentrated immediately behind the shock in the highest

328: velocity molecular gas \citep{van98}. So perhaps the fact that the low

329: relative velocity peaks in the \siofi\ emission, only $\pm 5$ \kms, is

330: an indication that the outflow is mainly in the plane of the sky. This

331: outflow, however, is far more energetic than an outflow from a

332: low-mass protostar. There could be highly excited SiO emission which

333: follows behind the shock more closely than the \siofi\ emission. We

334: plan to observe \sioeight, and lines of several CO isotopes with the

335: SMA allowing calculations of excitation and abundance in the outflow

336: to see if higher excitation gas occurs at higher line of sight

337: velocities, as we might expect for an outflow largely along the line

338: of sight. These observations will also have higher angular resolution

339: which should clarify the outflow opening angle.

340: 

341: On an equally hypothetical note, consider the fact that while some

342: dust emission appears to come from the outflow, the free-free emission

343: shows no preference for the direction of the outflow, and is more

344: extended perpendicular to the outflow. Recent results have shown that

345: massive protostars ($\sim 10$\msun) can have disk-like structures akin

346: to the disks seen in low-mass protostars \citep{zha98b,zha02}. If such

347: a disk were to photoevaporate, as was modeled by \citet{hol94}, that

348: would naturally produce free-free emission which is extended

349: perpendicular to the outflow, and possibly ring-like. However, if the

350: outflow does not originate in the \uchii\ region, the more traditional

351: limb-brightened-shell interpretation of the free-free emission may

352: make more sense. Future work to investigate the geometry and the

353: velocity structure of the ionized gas could be carried out with the

354: SMA using low quantum number radio recombination lines (RRLs) of

355: hydrogen, including \htwonea, \htwsixa, \htha, and \hthonea. These

356: lines suffer less from pressure broadening than higher quantum number

357: RRLs, and could be used to probe the velocity structure of the densest

358: ionized gas in the \uchii\ region.

359: 

360: \bibliographystyle{apj}

361: \bibliography{bib_entries}

362: 

363: \begin{figure}

364: \epsscale{1.0}

365: \plotone{f01.eps}

366: \caption{227 GHz (1.3 mm) and 8.46 GHz continuum maps are shown in

367: grayscale, and 227 GHz continuum and integrated line emission maps are

368: shown in contours. The first panel is 227 GHz continuum in linear

369: grayscale from -0.1 to 2.6 \jyperbeam\ and contours in multiples of

370: 0.25 \jyperbeam. The second panel shows the same 227 GHz continuum in

371: contours, and the 8.46 GHz continuum map of \citet{woo89b} in linear

372: grayscale from -5 \mjyperbeam\ to 100 \mjyperbeam. The remaining

373: panels all show the 227 GHz continuum in the same grayscale as the

374: first panel, and contour maps of the velocity integrated brightness in

375: a variety of molecular lines. The \methanolline\ map is contoured in

376: multiples of 1 \jyperbeamkms\ and the other three contour maps are

377: contoured in multiples of 3 \jyperbeamkms. The \siofi\ maps are

378: integrated from \vlsr\ = -25 \kms\ to 5 \kms\ for the blue lobe and

379: from \vlsr = 15 \kms\ to 45 \kms for the red lobe. All other lines are

380: integrated over their full extent in velocity. In all panels the white

381: line segment connects the positions of the \siofi\ peaks, the white

382: star represents the position of the O5 star detected by \citet{fel03},

383: and the beam is in the lower right.\label{figure:linemaps}}

384: \end{figure}

385: 

386: \begin{figure}

387: \epsscale{1.0}

388: \plotone{f02.eps}

389: \caption{A position-velocity diagram of the \siofi\ emission in

390: grayscale ranging linearly from -0.1 to 3.3 \jyperbeam\ and contoured

391: at 1 to 10 $\times 0.3$ \jyperbeam\ in the first panel, and in the

392: second panel, the \siofive\ in the same grayscale and \htwos\ emission

393: in contours at 1 to 10 $\times 0.3$ \jyperbeam. The data are hanning

394: smoothed to 3.3 \kms\ spectral resolution, with the resolutions along

395: both axes shown by the ``beam'' in the lower right. The cut goes

396: through both of the integrated line emission peaks in the \siofi\ map

397: in Figure \ref{figure:linemaps} with positive direction in position

398: being Northeast as shown. \label{figure:posvel}}

399: \end{figure}

400: 

401: \begin{figure}

402: \epsscale{1.0}

403: \plotone{f03.eps}

404: \caption{The spectral energy distribution of G5.89-0.39. Points shown

405: are from the VLA and ATCA in the free-free part of the spectrum, and

406: from the SMA, OVRO, JCMT and CSO in the dust dominated part of the

407: spectrum. The fit to the free-free spectrum is shown as a dotted

408: curve, the fit to the dust emission is dashed, and the sum of the two

409: is solid. Error bars are shown were they are known. The fits include

410: data from \citet{ake96,san94,aco98,bec94,woo89b,hun97}. The datum from

411: this work is a solid square. Other data are open

412: squares. \label{figure:sed}}

413: \end{figure}

414: 

415: \end{document}

416: