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

2: 

3: %\documentclass[preprint,11pt]{aastex}

4: % Uncomment next line for full ApJ emulation if you have this package

5: %\usepackage{emulateapj5}

6: % General purpose macros

7: 

8: \newcommand{\ee}[1]{\mbox{${} \times 10^{#1}$}}% scientific number format

9: \newcommand{\eten}[1]{\mbox{$10^{#1}$}}% power of ten

10: 

11: %macros for RA and Dec

12: %\newcommand{\h}{\mbox{{$^h$}}

13: %\newcommand{\m}{\mbox{{$^m$}}

14: %\newcommand{\s}{\mbox{{$^s$}}

15: \newcommand{\degree}{\mbox{$^{\circ}$}}

16: \newcommand{\am}{\mbox{\arcmin}}

17: \newcommand{\as}{\mbox{\arcsec}}

18: %macros for distance, volume, speed

19: \newcommand{\kms}{\mbox{km s$^{-1}$}}% km/s

20: \newcommand\cmv{\mbox{cm$^{-3}$}}

21: \newcommand\cmc{\mbox{cm$^{-2}$}}

22: \newcommand\cmdv{\mbox{cm$^{-2}$ (\kms)$^{-1}$}}

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

24: %\newcommand{\micron}{$\mu$m}

25: %macros for commonly used symbols

26: \newcommand{\x}{\mbox{${}\times{}$}}

27: %\def\lsim {$\rlap{\raise.4ex\hbox{$<$}}\lower.55ex\hbox{$\sim$}\,$}

28: \newcommand\tto{\mbox{$\rightarrow$}}

29: \newcommand\about{\mbox{$\sim$}}

30: %macros to avoid typing headache and needless acronyms simultaneously

31: \newcommand{\iras}{\mbox{\it IRAS}}

32: \newcommand{\iso}{\mbox{\it ISO}}

33: \newcommand{\ISO}{\mbox{\it ISO}}

34: \newcommand{\mm}{millimeter}

35: \newcommand\submm{submillimeter}

36: \newcommand\smm{submillimeter}

37: \newcommand\fir{far-infrared}

38: \newcommand\mir{mid-infrared}

39: \newcommand\nir{near-infrared}

40: \newcommand\uv{ultraviolet}

41: \newcommand{\sfr }{\mbox{$\dot M_{\star}$}}

42: \newcommand\sed{spectral energy distribution}

43: \newcommand{\lsun}{\mbox{L$_\odot$}}% Lsun

44: \newcommand{\msun}{\mbox{M$_\odot$}}% Msun

45: \newcommand{\ta}{\mbox{$T_A^*$}}

46: \newcommand{\tex}{\mbox{$T_{\rm ex}$}}

47: \newcommand{\tmb}{\mbox{$T_{\rm mb}$}}

48: \newcommand{\tr}{\mbox{$T_R$}}

49: \newcommand{\tk}{\mbox{$T_K$}}

50: \newcommand{\td}{\mbox{$T_d$}}

51: \newcommand{\lbol}{\mbox{$L_{bol}$}} % bolometric luminosity

52: \newcommand{\tbol}{\mbox{$T_{bol}$}} % bolometric temperature

53: \newcommand{\dv}{\mbox{$\Delta v$}}

54: \newcommand{\n}{\mbox{$n$}}

55: \newcommand{\nbar}{\mbox{$\overline{n}$}}

56: \newcommand{\mv}{\mbox{$M_V$}} % virial mass

57: \newcommand{\mc}{\mbox{$M_N$}} % column density mass

58: \newcommand{\mn}{\mbox{$M_n$}} % density mass

59: \newcommand{\meanl}{\mbox{$\langle l \rangle$}} % mean size

60: \newcommand{\meandev}{\mbox{$\langle \delta \rangle$}} % mean deviation

61: \newcommand{\meanar}{\mbox{$\langle a/b \rangle$}} % mean aspect ratio

62: \newcommand{\mean}[1]{\mbox{$\langle#1\rangle$}} %generic mean for defined qu.

63: \newcommand{\opacity}{\mbox{$\kappa(\nu)$}} % opacity as func. of freq.

64: \newcommand{\av}{\mbox{$A_V$}} % Visual Extinction

65: \newcommand{\bperp}{\mbox{$B_{\perp}$}} % Projection of B on plane of sky

66: \newcommand{\rinf}{\mbox{$r_{inf}$}} % infall radius

67: \newcommand{\fsmm}{\mbox{$L_{smm}/L_{bol}$}} % submm lum over bol. luminosity

68: \newcommand{\lsmm}{\mbox{$L_{smm}$}} % luminosity longward of 350 mic.

69: \newcommand{\alphanir}{\mbox{$\alpha_{NIR}$}} % spectral index 2-20 mic.

70: \newcommand{\isrf}{\mbox{\rm{ISRF}}}

71: %macros for molecule names

72: \newcommand{\hh}{\mbox{{\rm H}$_2$}}

73: \newcommand{\form}{H$_2$CO}

74: \newcommand{\water}{H$_2$O}

75: \newcommand{\ammonia}{\mbox{{\rm NH}$_3$}}

76: \newcommand{\coo}{$^{13}$CO}

77: \newcommand{\cooo}{C$^{18}$O}

78: \newcommand{\coooo}{C$^{17}$O}

79: \newcommand{\hcop}{HCO$^+$}

80: \newcommand{\hcopi}{H$^{13}$CO$^+$}

81: \newcommand{\dcop}{DCO$^+$}

82: \newcommand{\nthp}{N$_2$H$^+$}

83: \newcommand{\jj}[2]{\mbox{$J = #1\rightarrow#2$}}

84: 

85: %%% macros for figures in postscript format

86: %   syntax: \psfig{vertical size}{ps file name}{caption}

87: %\input{epsf}

88: 

89: %\newcommand{\psfig}[3]

90: %            {{\begin{figure}[tbp]

91: %             \vbox to#1{\epsfxsize=\textwidth\epsfbox{#2.ps}}

92: %         \caption{#3}

93: %         \end{figure}}}

94: %  \newcommand{\psscaledfig}[4]

95: %             {{\begin{figure}[tbp]

96: %         \centerline{\vbox to#1{\epsfxsize=#2\epsfbox{#3.ps}}}

97: %         \caption{#4}

98: %         \end{figure}}}

99: %\def\plotfiddle#1#2#3#4#5#6#7{\centering \leavevmode

100: %\vbox to#2{\rule{0pt}{#2}}

101: %\special{psfile=#1 voffset=#7 hoffset=#6 vscale=#5 hscale=#4 angle=#3}}

102: %Macros for this paper

103: 

104: \newcommand{\Inu}{\mbox{$I_{\nu}(b)$}}

105: \newcommand{\Snu}{\mbox{$S_{\nu}$}}

106: \newcommand{\Bnu}{\mbox{$B_{\nu}(\Td)$}}

107: \newcommand{\nInu}{\mbox{$I^{norm}(b)$}}

108: \newcommand{\nInum}{\mbox{$I^{norm}_{mod}(b)$}}

109: \newcommand{\kappanu}{\mbox{$\kappa_{\nu}$}}

110: \newcommand{\Td}{\mbox{$T_{d}$}}

111: \newcommand{\Tk}{\mbox{$T_{K}$}}

112: \newcommand{\Tdr}{\mbox{$T_{d}(r)$}}

113: \newcommand{\Tkr}{\mbox{$T_{K}(r)$}}

114: \newcommand{\rhor}{\mbox{$\rho (r)$}}

115: \newcommand{\ppc}{pre-protostellar core}

116: \newcommand{\m}{\mbox{$m$}}

117: \newcommand{\p}{\mbox{$p$}}

118: \newcommand{\nc}{\mbox{$n_c$}}

119: \newcommand{\chisq}{\mbox{$\chi_r^2$}}

120: \newcommand{\beam}{\mbox{$\theta_{mb}$}}

121: \newcommand{\router}{\mbox{$r_o$}}

122: \newcommand{\rinner}{\mbox{$r_i$}}

123: \newcommand{\rflat}{\mbox{$r_{flat}$}}

124: \newcommand{\lfir}{\mbox{$L_{IR}$}}

125: \newcommand{\lmol}{\mbox{$L_{mol}$}}

126: \newcommand{\lhcn}{\mbox{$L_{HCN}$}}

127: \newcommand{\lco}{\mbox{$L_{CO}$}}

128: \newcommand{\lunit}{\mbox{$L_{unit}$}}

129: \newcommand{\lmin}{\mbox{$L_{min}$}}

130: \newcommand{\mstar}{\mbox{$M_{\star}$}}

131: \newcommand{\lstar }{\mbox{$L_{\star}$}}

132: \newcommand{\mstarmax}{\mbox{$M_{\star}(max)$}}

133: \newcommand{\mdense}{\mbox{$M(dense)$}}

134: \newcommand{\mvir}{\mbox{$M_{vir}$}}

135: 

136: %Use the following to give date of each draft; comment for submitted version

137: %\slugcomment{\footnotesize {\LaTeX}ed at \number\time\ min., \today}

138: %Uncomment this for submitted version

139: %\slugcomment{\footnotesize {}}

140: \shorttitle{Inflow in Massive Regions}

141: \shortauthors{Wu \& Evans }

142: %\received{2003 April 30}

143: \begin{document}

144: %%%%%%%%%%%%%%%%%% title %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

145: \title {\bf Connecting Dense Gas Tracers of Star Formation in our Galaxy to 

146: High-z Star Formation }

147: \author {Jingwen Wu$^{1}$, Neal J. Evans II$^{1}$, Yu Gao$^{2}$, 

148:  Philip M. Solomon$^{3}$, Yancy L. Shirley$^{4,5}$, Paul A. Vanden Bout$^{6}$ }

149: \altaffiltext{1}{Department of Astronomy, The University of Texas at Austin,

150:         1 University Station, C1400, Austin, Texas 78712--0259,

151:         jingwen@astro.as.utexas.edu, nje@astro.as.utexas.edu}

152: %\altaffiltext{1}{Secondary Affiliation if relevant}

153: %\email{jingwen@astro.as.utexas.edu}

154: %\email{nje@astro.as.utexas.edu}

155: 

156: %\author{Yu Gao$^{2}$}

157: \altaffiltext{2}{Purple Mountain Observatory, Chinese Academy of Sciences, 

158: 2 West Beijing Road, Nanjing 210008. P.R.China, yugao@pmo.ac.cn}

159: %\email{yugao@pmo.ac.cn}

160: 

161: %\author{Philip M. Solomon$^{3}$}

162: \altaffiltext{3}{Department of Physics and Astronomy, SUNY at Stony Brook, Stony Brook, NY 11974, Philip.Solomon@sunysb.edu}

163: %\email{Philip.Solomon@sunysb.edu}

164: 

165: %\author{Yancy L. Shirley$^{4,5}$}

166: \altaffiltext{4}{Steward Observatory, University of Arizona, Tucson, AZ 85721.

167:  yshirley@as.arizona.edu}

168: \altaffiltext{5}{Bart J. Bok Fellow.}

169: %\email{yshirley@as.arizona.edu}

170: 

171: %\author {Paul A. Vanden Bout$^{6}$}

172: \altaffiltext{6}{National Radio Astronomy Observatory, 520 Edgemont Rd., 

173: Charlottesville, VA 22903. pvandenb@nrao.edu}

174: %\email{pvandenb@nrao.edu}

175: 

176: 

177: 

178: %Use the following to give date of each draft; comment for submitted version

179: %\slugcomment{\footnotesize {\LaTeX}ed at \number\time\ min., \today}

180: %comment this for submitted version

181: 

182: \pagebreak

183: 

184: \begin{abstract}

185: 

186: Observations have revealed prodigious amounts of star formation in starburst 

187: galaxies as traced by dust and molecular emission, even at large redshifts. 

188: Recent work shows that for both nearby spiral galaxies and 

189: distant starbursts, the global star formation rate, as indicated by the 

190: infrared 

191: luminosity, has a tight and almost linear correlation with the amount of 

192: dense gas as traced by the luminosity of HCN. Our surveys of Galactic dense 

193: cores in HCN 1$-$0 emission show that this correlation continues 

194: to a much smaller scale, with nearly the same ratio of infrared

195: luminosity to HCN luminosity found over 7-8 orders of magnitude in \lfir, 

196: with a lower cutoff around $10^{4.5}$ \lsun\ of infrared luminosity.

197: The linear correlation suggests that we may understand distant star 

198: formation in terms of the known properties of local star-forming 

199: regions. Both the correlation and the luminosity cutoff

200: can be explained if the basic unit of star formation in galaxies is

201: a dense core, similar to those studied in our Galaxy.

202: 

203: 

204: \end{abstract}

205: \keywords{stars: formation  --- ISM: molecules }

206: 

207: \clearpage

208: 

209: \section{Introduction}

210: 

211: Recent work has revealed large amounts of dust and molecules in 

212: starburst galaxies, even at large redshifts (e.g. Isaak et al. 2002; 

213: Reuland et al. 2003; Greve et al. 2005; Solomon \& Vanden Bout 2005). 

214: Understanding star formation in galaxies at high redshift is a critical step

215: in understanding the formation of galaxies in the early Universe.

216: 

217: The simplest and most widely used relations between the star formation rate 

218: and a property of the interstellar medium (ISM) are the so-called

219: ``Schmidt laws". Schmidt (1959) proposed that the star formation rate was

220: proportional to $\rho^2$, where $\rho$ is the gas volume density. 

221: In their modern form, these ``laws" relate the {\it surface density} of

222: star formation to the {\it surface density} of gas:

223: \begin{equation}

224:  \Sigma_{SFR} = A \Sigma^{N}_{gas} 

225: \end{equation}

226: (e.g., Kennicutt 1998).

227: The index $N$ has been inferred by various authors to be in the range of

228: 1 to 2 (Kennicutt 1997). Measurements of HI, CO, and H$\alpha$ on a large 

229: sample of normal spiral galaxies and starburst galaxies (Kennicutt 1998) 

230: could be fitted  with $N = 1.4$. It should be noted that for many of these 

231: galaxies, particularly the starburst galaxies, in the Kennicutt sample the 

232: surface density or surface brightness was not measured but obtained only by 

233: dividing the luminosity by a characteristic size often obtained from another 

234: parameter. 

235: 

236: On a global scale, including luminous and ultraluminous starburst galaxies,

237: there is a correlation between the total luminosity of far-infrared emission,

238: which traces the star formation rate (e.g. Sanders and Mirabel 1996; Kewley 

239: et al. 2002),

240: and the total luminosity of CO, tracing the molecular gas mass. 

241: However, this relation is not linear; the ratio of \lfir\ to \lco\ increases

242: with increasing \lfir\  ( Sanders and Mirabel 1996; Solomon et al. 1997; 

243: Gao \& Solomon 2004a, 2004b; Solomon \& Vanden Bout 2005).

244: Does this variation reflect an increase in ``efficiency" of star formation,

245: or an increasing failure of CO to trace the gas that is relevant to star 

246: formation?

247: 

248: The latter possibility is suggested by data on HCN $J = 1-0$ in galaxies.

249: A recent survey of HCN $J=1-0$ emission in 65 

250: normal spiral and starburst galaxies found that the star formation rate, 

251: as traced by the infrared luminosity, has a tight and linear correlation 

252: with the luminosity of HCN (Gao \& Solomon 2004a, 2004b). 

253: Those authors argued that CO is not linearly correlated with star formation 

254: because it traces only the low density GMC envelopes, not the really active 

255: star-forming part, the dense cores. 

256: The critical molecular parameter that measures 

257: star formation rates in galaxies is the amount of dense molecular gas,

258: measured by the HCN luminosity.  Because

259: HCN traces the dense gas better than CO,  L$_{HCN}$ has a tighter correlation 

260: with 

261: $L_{IR}$ than does $L_{CO}$.  The correlation remains linear over a factor 

262: of $10^3$ in luminosity for both normal galaxies and extreme starbursts,

263: like luminous and ultraluminous infrared galaxies (LIRGs and ULIRGs, see 

264: Sanders and Mirabel, 1996). Gao and Solomon (2004a) therefore argue that both 

265: normal galaxies and starbursts should have the same star formation rate per 

266: amount of {\it dense} gas. 

267: 

268: At lower luminosities, the relation for galaxies between \lfir\ and \lco\ is 

269: linear, but the ratio between \lfir\ and \lco\ has a dispersion of 

270: an order of magnitude. 

271: An even larger variation in \lfir/\lco\ (several orders of magnitude) is seen 

272:  in Galactic clouds (Mooney \& Solomon 1988; Evans 1991; Mead et al. 1990).

273: The dispersion in the relation for galaxies is less if \lhcn\ is used instead 

274: of CO; the same is true for other tracers of dense gas in Galactic cores 

275: (e.g. Mueller et al. 2002, Shirley et al. 2003), 

276: suggesting that studies

277: of Galactic star formation can shed light on the trends in other galaxies. 

278: 

279: In the Milky Way, star formation is dominated by 

280: clustered star formation (Carpenter 2000; Lada \& Lada 2003 ). 

281: Clustered star formation produces stars with a range of masses, 

282: but massive stars form nearly exclusively in clusters within massive 

283: dense cores. Since massive stars dominate the luminosity, they 

284: are the stars directly probed in studying star formation in other galaxies. 

285: Thus, understanding the relation between star formation and massive,

286: dense cores in our Galaxy may shed light on the starburst phenomenon.

287: 

288: These cores are dense, turbulent, and dusty. They are well identified by 

289: the continuum emission from dust (e.g., Mueller et al. 2002) and line emission 

290: from molecular dense gas tracers like CS (Plume et al. 1992, 1997; Shirley 

291: et al. 2003) and HCN (Wu \& Evans 2003). In our previous work on a large 

292: sample of massive cores, we found that the bolometric (almost all far-infrared)

293: luminosity of the cores is roughly proportional to the mass inferred from the

294: dust emission (Mueller et al. 2002) and the virial mass determined from CS

295: (Shirley et al. 2003). This result suggests that a relation between

296: \lfir\ and \lhcn\ may exist in Galactic dense cores, possibly similar to that 

297: in starburst galaxies.

298: 

299: One difficulty is that systematic studies of the Galactic dense cores have used

300: dust continuum emission and CS, while HCN has been more commonly used for 

301: studies of

302: galaxies. To facilitate comparison with the HCN $J = 1-0$ galaxy survey, 

303: we have mapped the HCN $J = 1-0$ transition in a sample of 

304: 47 Galactic star forming cores (Wu et al. 2005, in preparation). 

305: In this paper we will 

306: summarize the results from this survey in connection with the HCN surveys of

307: other galaxies.

308: 

309: The sample mapped in HCN $J = 1-0$ is mostly a subset of a larger sample of 

310: Galactic massive dense cores that have been mapped in CS and dust emission

311: (Shirley et al. 2003; Mueller et al. 2002). The sources in this category 

312: have infrared luminosities ranging from 10$^{3}$ - 10$^{7}$ L$_{\odot}$ 

313: and most contain compact or ultracompact H $\amalg$ (UCH$\amalg$) regions.

314: To extend the sample towards lower luminosities, 

315: we selected 14 IRAS sources from outflow surveys (Zhang et al. 2005; Wu et al. 

316: 2004) and a few from other publications. 

317: 

318: \section{Observation \& Data Analysis}

319: 

320: Observations of HCN \jj10\ (88.6318473 GHz) on Galactic dense cores 

321: were made with the 

322: 14-m telescope of the Five College Radio Astronomy Observatory (FCRAO) 

323: in 2004 April, December and 2005 February. The FWHM of the beam for 

324: this frequency is 58\arcsec. 

325: The 16-element focal plane array (SEQUOIA) was used, with typical system 

326: temperatures 100-200K. A velocity resolution of 0.1 \kms\ was achieved with 

327: the 25 MHz bandwidth on the dual channel correlator (DCC). 

328: We convert the measured \ta\ to \tr\ via $\tr = \ta/(\eta_{FSS}\eta_c)$, with

329: $\eta_{FSS}=0.7$. The value of $\eta_{c}$ depends on source size; for

330: the typical map in this study ($\sim 10\arcmin$), $\eta_{C}=0.7$.

331: The map size was extended until the edge of the HCN \jj10\ emission was 

332: reached, typically at the $2\sigma$ level 

333: (mean $\sigma$ $\sim$ 0.3 K km/s), so we could get the total 

334: HCN luminosity.

335: 

336: Maps of HCN 1-0 emission were obtained for 47 Galactic star-forming cores. 

337: Detailed results will be presented elsewhere (Wu et al. 2005, in preparation),

338: but we give a few properties of the sample here, 

339: which are relevant to this paper. 

340: More than 90\% of the cores were well resolved by the maps. We focus on 

341: these resolved sources in this paper. 

342: 

343: The size of the core is characterized by the nominal core radius after

344: beam deconvolution, 

345: $R_{HCN}$, the radius of a circle that has the same area as the half 

346: peak intensity contour:

347: $$R_{HCN}=D(\frac{A_{1/2}}{\pi}-\frac{\theta^{2}_{beam}}{4})^{1/2}, $$ 

348: where $A_{1/2}$ is the area within the contour of half peak intensity.

349: The median $R_{HCN 1-0}$ of the sample is 0.48 pc. 

350: 

351: The HCN line luminosity of each core, assuming a gaussian 

352: brightness distribution for the source and a gaussian beam, is: 

353: \begin{equation}

354: L_{HCN}=23.5\times 10^{-6}\times D^{2}\times (\frac{\pi\times\theta_{s}^{2}}

355: {4ln 2})\times(\frac{\theta_{s}^{2}+\theta_{beam}^{2}}{\theta_{s}^{2}})

356: \times \int T_{R}dv 

357: \end{equation}

358: Here D is the distance in the unit of kpc, $\theta_{s}$ and $\theta_{beam}$ 

359: are the angular size of the source and beam in arcsecond. 

360: This method is parallel to that of Gao \& Solomon (2004b), but adapted

361: to Galactic cores. \lhcn\ ranges from 0.4 to 8000 

362: K \kms\ pc$^{2}$, with the median value of 80 K \kms\ pc$^{2}$ . 

363: 

364: The total infrared luminosity (8-1000 $\mu$m) was calculated 

365: based on the 4 IRAS bands (Sanders and Mirabel 1996), as was done for

366: the galaxy sample of Gao \& Solomon (2004a):

367: \begin{equation}

368:  \lfir = 0.56\times D^{2}\times (13.48\times f_{12}+5.16\times f_{25}+

369: 2.58\times f_{60}+f_{100}), 

370: \end{equation}

371: where $f_x$ is the flux in band $x$ from the four IRAS bands in the units 

372: of Jy, D in kpc, and L$_{IR}$ in \lsun. 

373: 

374: 

375: \section{Comparison of Milky Way and galactic Relations}

376: 

377: The derived \lfir\ and \lhcn\ are plotted in a log-log diagram in 

378: Fig. 1 to compare with data on galaxies from Gao \& Solomon (2004a). 

379: The correlation of \lfir--\lhcn\ 

380: extends from galaxy scales to the much smaller scales of Galactic molecular 

381: cores. The fit for Gao \& Solomon's 

382: galaxy sample is log$(\lfir)=1.00\times {\rm log}(\lhcn)+2.9$,

383:  without a few galaxies that only have upper limits to HCN emission. 

384: This linear correlation continues to the Galactic massive cores, but a decline

385: in \lfir\ occurs at $10^{4.5}$ \lsun,

386: below which the slope of the correlation becomes much steeper. 

387: When fit to Galactic cores with $\lfir > 10^{4.5}$ \lsun, the linear 

388: least squares fit gives 

389: log($\lfir)=1.02(\pm0.06)\times {\rm log}(\lhcn)+2.79(\pm0.16)$. 

390: This relation agrees remarkably well with the relation for galaxies, as seen 

391: in fig. 1a. In fig. 1b, we fit simultaneously the data of galaxies from 

392: Gao \& Solomon (2004a) and Galactic cores with $\lfir > 10^{4.5}$ \lsun. 

393: The result is log$(\lfir)=1.01\times {\rm log}(\lhcn)+2.83$, with 

394: a correlation coefficient of 0.99.

395: Understanding the physics behind this linear correlation will lead to

396: a better understanding of star formation on galactic scales. 

397: 

398: The linear log($\lfir)-{\rm log}(\lhcn)$ 

399: correlation and the turnoff can be seen more 

400: clearly from fig. 2, where the distance independent ratio 

401: \lfir/\lhcn\ has been plotted versus \lfir\ (top) and against \lmol\ (bottom).

402: A constant mean value of \lfir/\lhcn\ is seen over 8 orders of magnitude

403: in \lfir, from galaxies to Galactic cores, as long as $\lfir > L(min)$,

404: with $L(min)$ being around $10^{4.5}$ \lsun. The corresponding cut-off

405: value for \lhcn\ is $\lunit = 10^{1.5}$ K \kms\ pc$^{2}$.

406: For Galactic cores,

407: $\mean{\lfir/\lhcn}= 911\pm 227$, with median 380. For galaxies, 

408: $\mean{\lfir/\lhcn}= 950\pm 76$, with median 855. The uncertainties are

409: the standard deviation of the means, which are remarkably similar, though

410: the dispersion is much higher for Galactic cores. The median 

411: for dense cores is significantly smaller than its mean, which indicates that 

412: the mean is dominated by a few quite large values. The logarithmic mean and 

413: median are 2.64$\pm$0.53 and 2.54$\pm$0.56 for dense cores with 

414: $\lhcn > \lunit$, and 2.91$\pm$0.24 and 2.93$\pm$0.25 for galaxies.

415:  

416: 

417: For comparison, we added CO data on Galactic cores (Mooney \& Solomon 1988), 

418: galaxies (Gao \& Solomon 2004a) and high-z molecular emission line galaxies 

419: (Solomon \& Vanden Bout 2005) in fig. 2.

420: The ratio, \lfir/\lco, changes by two orders of magnitude as \lco\ increases

421: from Galactic cores to distant galaxies,

422: confirming Gao \& Solomon's conclusion that CO is 

423: not as good a tracer of star-forming gas as is HCN, especially for very 

424: luminous starburst galaxies.

425:  

426: 

427: \section{Discussion}

428: 

429: The fact that \lfir/\lhcn\ is similar, on average, for star-forming

430: cores in the Galaxy, normal spirals, starbursts, and ULIRGs

431: suggests the possibility of interpreting intense high-z star formation in

432: terms of nearby high mass star forming regions. 

433: Before we can exploit this possibility, we must understand some key points.

434: Why does \lfir/\lhcn\ rise steeply with \lhcn\ and

435: then remain constant for $\lhcn > \lunit$? 

436: And why is the ratio, \lfir/\lco, NOT constant for starbursts?

437: 

438: As a first step, we seek a more physical basis for the relations.

439: We have so far discussed \lhcn\ as a measure of the mass of dense gas, but

440: can we quantify this assumption? A roughly linear correlation between the 

441: mass of dense gas and bolometric luminosity has been found by 

442: our work on CS and dust emission (Shirley et al. 2003; Mueller et al. 2002). 

443: To see whether this applies also to HCN \jj10,

444: we have calculated the virial mass of the dense gas

445: (\mdense) using the most optically thin line (C$^{34}$S \jj54) to measure the 

446: linewidth 

447: and compared \mdense\ to \lhcn. Based on the 31 cores with available

448: C$^{34}$S \jj54 data, we obtained the correlation: 

449: log($M_{vir}(R_{HCN1-0})$)=(0.81$\pm$0.03)$\times$ log($L_{HCN1-0}$)

450: +1.29($\pm$0.09). The correlation is roughly linear. The 

451: $M_{vir}$-$L_{HCN1-0}$ plot and details of the observations 

452: of C$^{34}$S \jj54 will be presented by Wu et al. (2005, in preparation).

453: Leaving out one peculiar source (G35.58-0.03), we get $\mean{\mdense/\lhcn} = 

454: 11\pm 2$~\msun/K \kms\ pc$^{2}$, where the uncertainty is the standard 

455: deviation of the mean; the median value is 6 ~\msun/K \kms\ pc$^{2}$, 

456: indicating that the mean is affected by some quite large

457: values. The logarithmic mean is 7$\pm$2 ~\msun/K \kms\ pc$^{2}$. 

458: Some of the scatter in the ratio may be caused by distance 

459: uncertainties because the virial mass depends linearly on distance, while

460: $\lhcn \propto D^2$.

461: 

462: 

463: Even after establishing that \lhcn\ traces the mass of dense gas,

464: it is not at all clear why \lfir/\lhcn\ should be constant, since 

465: the luminosity of a cluster is typically dominated by its most massive 

466: members, and $\lstar \propto \mstar^\alpha$, with $\alpha \approx 3$ to 4.

467: Indeed, below the cutoff of $10^{4.5}$ \lsun, \lfir\ does rise rapidly

468: with \mdense. What causes the transition to a constant value?

469: 

470: To solve this puzzle, we propose the existence of a basic unit of cluster 

471: formation. For

472: \mdense\ less than the mass of this unit, \lfir/\mdense\ rises rapidly

473: with \mdense, as higher mass stars can form. For \mdense\ greater than the

474: mass of this unit, the IMF is reasonably sampled and further increases

475: in mass produce more units, but no further change in \lfir/\mdense.

476: If we suppose that larger scale cluster formation is built up 

477: by adding more and more such units,

478: then the linear correlation between the total \lfir\ and 

479: \mdense\ is a natural result. In that case, the only difference between star 

480: formation on different scales and in different environments--big clusters, 

481: normal galaxies, massive ULIRGs--is just how {\it many} such cores 

482: they contain. This is of course somewhat of a simplification because

483: the cores have a range of properties. The fact that the mean is roughly

484: twice the median for both \lfir/\lhcn\ and \mdense/\lhcn\ for Galactic cores

485: suggests that the most extreme cases may dominate when averaged over a whole

486: galaxy.  Since most of our cores contain compact H $\amalg$ or 

487: UCH $\amalg$ regions, which trace the most massive stars, it is interesting

488: to compare to the luminosity function of these regions.

489: A study of Galactic UCH $\amalg$ regions 

490: did find a peak luminosity ($\sim$10$^{5}$\lsun ) that is close to  that of

491: a basic unit (Cassassus et al. 2000). 

492: The detailed discussion of this model will be presented in a 

493: separate paper (Wu et al. 2005, in preparation).

494: 

495: 

496: Another question is why does \lfir/\lco\ increase as we move from Galactic 

497: Cores or normal spiral galaxies to starburst galaxies. 

498: In Galactic clouds, CO can be used as a tracer of the overall

499: mass of molecular clouds, even though it is optically thick and thermalized.

500: However, CO does not specifically trace the mass of dense cores. For that, dust

501: continuum emission, or

502: molecules, like HCN, that are only excited at higher densities are required.

503: In the Galaxy, these dense cores occupy a small fraction of the mass

504: of a cloud (typically a  few percent).

505: If the fraction of dense gas in the overall cloud

506: stays roughly constant, the relation between \lfir\ and \lco\ can stay 

507: linear, but this relation is secondary to that between \lfir\ and \mdense.  

508: For starburst galaxies, the fraction of the molecular gas concentrated in 

509: dense cores increases, causing (Gao \& Solomon 2004a) the secondary relation 

510: between \lfir\ and \lco\ to become non-linear.

511: 

512: 

513: These considerations lead us to offer some new versions of ``Schmidt

514: Laws":

515: \begin{equation}

516: \sfr (\msun yr^{-1}) \sim 1.4\ee{-7} \lhcn ( K \kms\ pc^2)

517: \end{equation}

518: \begin{equation}

519: \sfr (\msun yr^{-1}) \sim 1.2\ee{-8} \mdense (\msun)

520: \end{equation}

521: where \sfr\ is the star formation rate, and we have assumed that 

522: $\sfr (\msun yr^{-1}) =  2.0\ee{-10} \lfir (\lsun)$

523: (see Gao \& Solomon (2004a)) and we use the fit to both dense cores

524: and galaxies. The coefficients (1.4\ee{-7} and 1.2\ee{-8}) are very 

525: similar to, but slightly less than, those values (1.8\ee{-7} and 1.8\ee{-8}) 

526: given by Gao \& Solomon (2004a), based only on the galaxies.

527: 

528: Given these simpler, linear

529: relations, how do we understand the usual star formation law (Kennicutt 1998) 

530: relations, with a steeper dependence 

531: ($\Sigma_{SFR} \propto \Sigma^{1.4}_{gas}$) of \sfr\ on gas mass? 

532: We suggest that the 

533: steeper dependence reflects the dependence of \mdense\ on the

534: surface density, or total mass, of gas. All the tracers of gas

535: used by Kennicutt trace lower density components, not the actual

536: gas that is directly involved in star formation. Once it is clear that it is 

537: the dense gas mass that indicates the star formation rate, it becomes clear 

538: why the total surface density of gas may not be a clean star formation 

539: indicator. For example, there is no evidence that HI emission in galaxies 

540: correlates at all with the star formation rate deduced form the far-infrared 

541: luminosity, so its contribution to the surface density may have no effect on 

542: the star formation rate.  Attempts to provide

543: a theoretical framework (e.g., Li et al. 2005, Krumholz \& McKee 2005)

544: for the Kennicutt relations should also be

545: able to explain the new relations.

546: 

547: 

548: It will be interesting to see how far these ideas can be extended.

549: One future project is to look at HCN in more nearby galaxies, 

550: especially in individual regions forming super-star clusters, which 

551: may be the building blocks of star formation in starburst regions 

552: like ULIRGs and LIRGs (Ho 1997).

553: ALMA will allow detailed study of HCN in other galaxies, including

554: the \jj32\ transition.

555: Another is to explore how far toward higher z these relations

556: can be pushed, and to understand how the relations depend on

557: metallicity and chemistry. Theoretical work by Lintott et al. (2005) suggest

558: that HCN may trace very early star formation, depending on the

559: nucleosynthetic yield of the earliest stars. 

560: 

561: 

562: \acknowledgements

563: 

564: We are grateful to Mark Heyer and other FCRAO staff for assistance with the 

565: observations. We thank the referee for helpful comments. 

566: We thank Dan Jaffe, John Scalo, Shardha Jogee, John Kormendy, Michael Dopita,

567: and Luis Ho for helpful discussions. 

568: This work was supported by NSF Grant AST-0307250 to the

569: University of Texas at Austin and by the State of Texas. YG acknowledges

570: support of the NSF of China grants 10425313 \& 10333060.

571: 

572: 

573: 

574: \clearpage

575: \begin{thebibliography}{}

576: %\bibitem[Beuther and Schilke (2004)]{Beuther and Schilke (2004}

577: %Beuther, H and Schilke, K. 2004, Science, 303, 1167

578: \bibitem[Carilli et al. (2005)]{Carilli et al. (2005)}

579: Carilli, C.\ L., Solomon., P.\ M., Vanden Bout, P.\ A., Walter, F.,

580: Beelen, A., and Cox, P., Bertoldi, F., Menten, K.\ M., Isaak, K.\ G.,

581: Chandler, C. J., \& Omont, A. 2005, \apj, 618, 586

582: \bibitem[Carpenter (2000)]{Carpenter (2000)}

583: Carpenter, J.\ M. 2000, \aj, 120, 3139

584: \bibitem[Casassus et al. (2000)]{Casassus et al. (2000)}

585: Casassus, S., Bronfman, L., May, J., \& Nyman, L. -A. 2000, \aap,

586: 358, 514

587: \bibitem[Evans (1991)]{Evans (1991)}

588: Evans, N.\ J.\ II. 1991, in Frontiers of Stellar Evolution, ed. 

589: D.\ L.\ Lambert, (San Francisco: Astro. Soc. Pacific), 45 

590: \bibitem[Gao and Solomon (2004a)]{Gao and Solomon (2004a)}

591: Gao, Y., \& Solomon, P.\ M., 2004a, \apj, 606, 271

592: \bibitem[Gao and Solomon (2004b)]{Gao and Solomon (2004b)}

593: Gao, Y., \& Solomon, P.\ M., 2004b, \apjs, 152, 63

594: \bibitem[Greve et al. (2005)]{Greve et al. (2005)}

595: Greve, T.\ R., Bertoldi, F., Smail, Ian, Neri, R., Chapman, S.\ C., Blain, 

596: A.\ W., Ivison, R.\ J., Genzel, R., Omont, A., Cox, P., Tacconi, L., 

597: Kneib, J. -P. 2005, \mnras, 359, 1165. 

598: \bibitem[Ho (1997)]{Ho (1997)}

599: Ho, L.\ C. 1997, in Rev. Mexicana Astron. Astrofis. Conf., Vol. 6, 

600: 1st Guillermo Haro Conference on Astrophysics: Starburst Activity in Galaxies,

601:  5

602: %\bibitem[Hoyle (1953)]{Hoyle (1953)}

603: %Hoyle, E. 1953, \apj, 118, 513

604: \bibitem[Isaak et al. (2002)]{Isaak et al. (2002)}

605: Isaak, K.\ G., Priddey, R.\ S., Mcmahon, R.\ G., Omont, A., Peroux, C., 

606: Sharp, R.\ G., and Withington, S. 2002, \mnras, 329, 149

607: %\bibitem[Johnson (2004)]{Johnson (2004)}

608: %Johnson, K. \ E. 2004. The Formation and Evolution of Massive Young Star 

609: %Clusters, ASP Conference Series, Vol. 322. Edited by H.J.G.L.M. Lamers, L.J. 

610: %Smith, and A. Nota. San Francisco. p339.

611: %\bibitem[Larson (1982)]{Larson (1982)}

612: %Larson, R.\ B. 1982, \mnras, 200, 159

613: \bibitem[Kennicutt (1997)]{Kennicutt (1997)}

614: Kennicutt,R.\ C. 1997, in The Interstellar Medium in Galaxies, ed. J.\ M.\ 

615: van der Hulst (Dordrecht: Kluwer), 171

616: \bibitem[Kennicutt (1998)]{Kennicutt (1998)}

617: Kennicutt,R.\ C. 1998, \apj, 498, 541

618: \bibitem[Kewley et al. (2002)]{Kewley et al. (2002)}

619: Kewley, L.\ J., Geller, M.\ J., Jansen R.\ A. \& Dopita, M.\ A.

620: 2002, \aj, 124, 3135

621: \bibitem[Krumholz \& McKee (2005)]{Krum05} 

622: Krumholz, M. R. \& McKee, C. F.  2005, \apj, 630, 250

623: \bibitem[Lada and Lada (2003)]{Lada and Lada (2003)}

624: Lada, C.\ J. \& Lada, E.\ A. 2003,  Ann. Rev. Ast. \& Astrophys., 41, 57

625: \bibitem[Li et al. (2005)]{lietal05} 

626: Li, Y., Mac Low, M.-M., \& Klessen, R. S. 2005, \apj, 626, 823

627: \bibitem[Lintott et al. (2005)]{Lintott et al. (2005)}

628: Lintott, C.\ J., Viti, S., Williams, D.\ A., Rawlings, J.\ M.\ C., Ferreras, I.

629: 2005, \mnras, 360, 1527

630: \bibitem[Mead et al. (1990)]{Mead et al. (1990)}

631: Mead, K.\ N., Kutner, M.\ L. and Evans, N.\ J.\ II. 1990, \apj, 354, 492

632: \bibitem[Mooney and Solomon (1988)]{Mooney and Solomon (1988)}

633: Mooney, T.\ J., \& Solomon, P.\ M. 1988, \apjl, 334, 51

634: \bibitem[Mueller et al. (2002)]{Mueller et al. (2002)}

635: Mueller, K.\ E., Shirley, Y.\ L., Evans, N.\ J.\ II, \& Jacobson, H. R.

636: 2002, \apjs, 143, 469

637: %\bibitem[Myers et al. (1986)]{Myers et al. (1986)}

638: %Myers, P.\ C., Dame, T.\ M., Thaddeus, P., Cohen, R.\ S., Silverberg, R.\ F., 

639: %Dwek, E, and Hauser, M.\ G. 1986, \apj, 301, 398

640: %\bibitem[Paglione et al. (1997)]{Paglione et al. (1997)}

641: %Paglione, T. A. D., Jackson, J. M., \& Ishizuki, S. 1997, \apj, 484, 656

642: \bibitem[Plume et al. (1992)]{Plume et al. (1992)}

643: Plume, R., Jaffe, D.\ T., Evans, N.\ J.\ II. 1992, \apjs, 78, 505

644: \bibitem[Plume et al. (1997)]{Plume et al. (1997)}

645: Plume, R., Jaffe, D.\ T., Evans, N.\ J.\ II, Martin-Pintado, J., \&

646: Gomez-Gonzalez, J. 1997, \apj, 476, 730

647: \bibitem[Reuland et al. (2003)]{Reuland et al. (2003)}

648: Reuland, M., R\"{o}ttgering, H., and van Breugel, W. 2003, 

649: New Astronomy Reviews, 47, 303

650: \bibitem[Sanders and Mirabel (1996)]{Sanders and Mirabel (1996)}

651: Sanders, D.\ B., \& Mirabel, I.\ F. 1996, Ann. Rev. Ast. 

652: \& Astrophys., 34, 749

653: \bibitem[Shirley et al. (2003)]{Shirley et al. (2003)}

654: Shirley, Y.\ L., Evans, N.\ J.\ II, Mueller, K.\ E., Knez, C., Jaffe, D. T.

655: 2003, \apjs, 149, 375.

656: %\bibitem[Shu et al. (1987)]{Shu et al. (1987)} 

657: %Shu, F.\ H., Adams, F.\ C., and Lizano, S. 1987, \apj 1987,

658: %Ann. Rev. Ast. \& Astrophys., 25, 23

659: %\bibitem[Solomon et al. (1987)]{Solomon et al. (1987)}

660: %Solomon., P.\ M., Rivolo, A.\ R., Barrett, J. and Yahil, A. 

661: %1987, \apj, 319, 730

662: \bibitem[Solomon et al. (1997)]{Solomon et al. (1997)}

663: Solomon, P.\  M., Downes, D., Radford, S. \ J. \ E., Barrett, J. \ W.

664: 1997, \apj, 478, 144

665: \bibitem[Solomon et al. (2003)]{Solomon et al. (2003)}

666: Solomon., P.\ M., Vanden Bout, P.\ A.,Carilli, C.\ L., \& Guelin, M.

667: 2003, Nature, 426, 636

668: \bibitem[Solomon and Vanden Bout (2005)]{Solomon and Vanden Bout (2005)}

669: Solomon., P.\ M. \& Vanden Bout, P.\ A. 2005, Ann. Rev. Ast. 

670: \& Astrophys. 43, 677.

671: \bibitem[Schmidt (1959)]{Schmidt (1959)}

672: Schmidt, M. 1959, \apj, 129, 243

673: \bibitem[Vanden Bout et al. (2004)]{Vanden Bout et al. (2004)}

674: Vanden Bout, P.\ A., Solomon., P.\ M., \& Maddalena, R. J. 

675: 2004, \apj, 614, 97

676: \bibitem[Wu and Evans (2003)]{Wu and Evans (2003)}

677: Wu, J., \& Evans, N.\ J.\ II, 2003, \apj, 592, L79

678: %\bibitem[Wu et al. (2005a)]{Wu et al. (2005a)}

679: %Wu, J. et al. 2005a, in preparation

680: %\bibitem[Wu et al. (2005b)]{Wu et al. (2005b)}

681: %Wu. J. et al. 2005b, in preparation

682: \bibitem[Wu et al. (2004)]{Wu et al. (2004)}

683: Wu, Y., Wei, Y., Zhao, M., Shi, Y., Yu, W., Qin, S., and Huang, M.

684: 2004, \aap, 426, 503

685: \bibitem[Zhang et al. (2005)]{Zhang et al. (2005)}

686: Zhang, Q., Hunter, T.\ R., Brand, J., Sridharan, T. K., Cesaroni, R., 

687: Molinari, S., Wang, J., and Kramer, M. 2005. ApJ, 625, 864

688:  

689: 

690: 

691: \end{thebibliography}

692: 

693: \clearpage

694: 

695: %%%%%%%Begin Figures%%%%%%%%%

696: 

697: \begin{figure}[hbt!]

698: %\plotfiddle{fig2.ps}{6.5in}{-90}{65}{65}{-300}{400}

699: %\plotone{fig2.ps}{6.5in}{0}{65}{65}{-200}{-20}

700: %\plotone{f1.eps}

701: \rotatebox{-90}{\plotone{f1.eps}}

702: %\epsscale{0.55}

703: %\plotone{f1.ps}

704: %}

705: \caption{logL$_{IR}$ - logL$'_{HCN1-0}$ correlation for Galactic and

706:  extragalactic sources. Fig.1a (above) indicates the linear least squares 

707:  fit for Galactic cores (with $\lfir > 10^{4.5}$ L$_{\odot}$, points above 

708:  the dashed line) and for 

709:  galaxies, separately.  Fig. 1b (down) shows a overall fit for both parts.  

710:  The three isolated solid squares are high z HCN 1-0 points from 

711:  Solomon et al. (2003), Vanden Bout et al. (2004) and Carilli et al. (2005); 

712:  they are not included in the fit because the sources are 

713:  QSOs and the contribution from the AGN to \lfir\ is not yet clear. 

714:  }

715: 

716: \label{f1}

717: \end{figure}

718: 

719: \clearpage

720: \begin{figure}[hbt!]

721: \epsscale{0.75}

722: \rotatebox{-90}{\plotone{f2.eps}}

723: %\plotone{f2.eps}

724: \caption{Above: Correlation between L$_{IR}$ and L$_{IR}$/L$_{HCN1-0}$ 

725: for galaxies (solid squares) and Galactic star forming cores (filled circles). 

726:  L$_{IR}$/L$_{HCN1-0}$ is constant for a large range of L$_{IR}$ utill 

727:  a turn off around L$_{IR}$=10$^{4.5}$ L$_{\odot}$.

728: Below: L$_{IR}$ per unit of molecular gas vs. molecular line luminosity. 

729:  The star formation rate per amount of CO gas changes a lot from Galactic GMCs

730: (the heavy line, with a dashed line boundary to indicate the variation), to

731: galaxies (filled triangles) and high-z CO galaxies (hollow 

732: triangles). }

733: 

734: \label{f2}

735: \end{figure}

736: 

737: 

738: \end{document}

739: 

740: 

741: 

742: 

743: