1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%5

2: %\batchmode

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

4: %\documentclass[manuscript]{aastex}

5: %\documentclass[10pt,preprint2]{aastex}

6: %\documentstyle[emulateapj]{article}

7: 

8: \slugcomment{Accepted by ApJ Letters}

9: 

10: \shorttitle{Nonaxisymmetric Collapse in Magnetic Clouds}

11: \shortauthors{Basu \& Ciolek}

12: 

13: 

14: %\input psfig.sty

15: 

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

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

18: \newcommand{\bl}[1]{\mbox{\boldmath$ #1 $}}

19: \newcommand{\few}{\rm few~}

20: \newcommand{\cs}{c_{\rm s}}

21: %\newcommand{\cs}{C}

22: \newcommand{\cmc}{~{\rm cm}^{-3}}

23: \newcommand{\cms}{~{\rm cm}^{-2}}

24: \newcommand{\Alf}{Alfv\'en\ }

25: \newcommand{\Msun}{~M_{\odot}}

26: \newcommand{\tff}{t_{\rm ff}}

27: \newcommand{\cm}{~\rm cm}

28: \newcommand{\kms}{~\rm km~s^{-1}}

29: \newcommand{\s}{~\rm s}

30: \newcommand{\pc}{~\rm pc}

31: \newcommand{\erg}{~\rm erg}

32: \newcommand{\K}{~\rm K}

33: \newcommand{\muG}{~\mu{\rm G}}

34: \newcommand{\mH}{m_{\rm H}}

35: \newcommand{\Pe}{P_{\rm ext}}

36: \newcommand{\mui}{\mu_0}

37: \newcommand{\muc}{\mu_{\rm core}}

38: \newcommand{\muenv}{\mu_{\rm env}}

39: \newcommand{\mucrit}{\mu_{\rm crit}}

40: \newcommand{\tnii}{\tilde{\tau}_{\rm ni,0}}

41: \newcommand{\Pext}{\tilde{P}_{\rm ext}}

42: \newcommand{\rhon}{\rho_{\rm n}}

43: \newcommand{\nn}{n_{\rm n}}

44: \newcommand{\nnc}{n_{\rm n,c}}

45: \newcommand{\nni}{n_{\rm n,0}}

46: \newcommand{\nion}{n_{\rm i}}

47: \newcommand{\sign}{\sigma_{\rm n}}

48: \newcommand{\signc}{\sigma_{\rm n,c}}

49: \newcommand{\signi}{\sigma_{\rm n,0}}

50: \newcommand{\Beq}{B_{z,\rm eq}}

51: \newcommand{\Beqc}{B_{z,\rm eq,c}}

52: \newcommand{\Beqi}{B_{z,\rm eq,0}}

53: \newcommand{\Bref}{B_{\rm ref}}

54: \newcommand{\vn}{v_{\rm n}}

55: \newcommand{\vnx}{v_{{\rm n},x}}

56: \newcommand{\zhat}{\mbox{\boldmath$ \hat{z}$}}

57: \newcommand{\lamcrit}{\lambda_{\rm T,cr}}

58: \newcommand{\lammax}{\lambda_{\rm T,m}}

59: 

60: 

61: \begin{document}

62: 

63: %\title{Formation and Nonaxisymmetric Collapse of Protostellar Cores

64: %in Planar Magnetic Interstellar Molecular Clouds}

65: \title{Formation and Collapse of Nonaxisymmetric Protostellar Cores

66: in Planar Magnetic Molecular Clouds}

67: \author{Shantanu Basu\altaffilmark{1} and Glenn E. Ciolek\altaffilmark{2}}

68: %\affil{Department of Physics and Astronomy, University of

69: %Western Ontario, London, Ontario N6A~3K7, Canada}

70: %\author{Glenn E. Ciolek}

71: %\affil{New York Center for Studies on the Origins of Life (NSCORT),

72: %and Department of Physics, Applied Physics, and Astronomy,

73: %Rensselaer Polytechnic Institute, 110 W. 8th Street, Troy, NY 12180}

74: %\email{basu@astro.uwo.ca}

75: %\email{cioleg@rpi.edu}

76: \altaffiltext{1}{Department of Physics and Astronomy, University of

77: Western Ontario, London, Ontario N6A~3K7, Canada; basu@astro.uwo.ca.}

78: \altaffiltext{2}{New York Center for Studies on the Origins of Life (NSCORT),

79: and Department of Physics, Applied Physics, and Astronomy,

80: Rensselaer Polytechnic Institute, 110 W. 8th Street, Troy, NY 12180;

81: cioleg@rpi.edu.}

82: 

83: 

84: 

85: \begin{abstract}

86: We extend our earlier work on ambipolar diffusion induced formation of

87: protostellar cores in isothermal sheet-like magnetic interstellar

88: clouds, by studying nonaxisymmetric collapse for the physically

89: interesting regime of magnetically critical and supercritical model

90: clouds ($\mui \geq 1$, where $\mui$ is the initial mass-to-magnetic

91: flux ratio in units of the critical value for gravitational collapse).

92: Cores that form in model simulations are effectively triaxial, with

93: shapes that are typically closer to being oblate, rather than prolate. Infall

94: velocities in the critical model ($\mui = 1$) are subsonic; in contrast,

95: a supercritical model ($\mui = 2$) has extended supersonic infall that

96: may be excluded by observations. For the magnetically critical model,

97: ambipolar diffusion forms cores that are supercritical ($\muc > 1$) and

98: embedded within subcritical envelopes ($\muenv < 1$).

99: Cores in our models have

100: density profiles that eventually merge into a near-uniform background,

101: which is suggestive of observed properties of cloud cores.

102: \end{abstract}

103: 

104: \keywords{diffusion --- ISM: clouds --- ISM: kinematics and dynamics

105:  --- ISM: magnetic fields --- MHD --- stars: formation}

106: \section{Introduction}

107: Magnetic fields play an important role in star formation, especially in

108: the early stages of core formation and collapse; measured mass-to-flux

109: ratios of molecular clouds yield an average that is $\sim 1 - 2$ 

110: times greater than the critical value for collapse (Crutcher

111: 1999). However, observational biases tend to push toward higher values

112: of measured mass-to-flux ratio (Crutcher 2003, private communication),

113: so that moderately subcritical cloud regions are not ruled out.

114: Dense cores within molecular clouds are the sites of star formation,

115: with detected infall 

116: %motions of 

117: up to $\approx 0.5 \, \cs \approx

118: 0.1 \kms$ on scales $\lesssim 0.1 \pc$, (e.g., in L1544, Tafalla et al.

119: 1998; Williams et al. 1999), where $\cs$ is the isothermal sound speed.

120: Ciolek \& Basu (2000) have fit the main features of the observed

121: velocity and density profiles in L1544, modeling it as an axisymmetric

122: supercritical core embedded in a moderately subcritical envelope.

123: 

124: Axisymmetry is clearly an idealization to real cores. Observations

125: suggest a typical projected axis ratio of 0.5 for cores (Myers et al.

126: 1991), and deprojections of the distribution of shapes imply

127: intrinsically triaxial but nearly oblate cores (Jones, Basu,

128: \& Dubinski 2001). Polarized emission measurements from dense cores also

129: imply triaxiality (Basu 2000). More generally, detailed submillimeter

130: maps of star-forming regions reveal significant irregular structure and

131: multiple cores (Motte, Andr\'e, \& Neri 1998).

132: Theoretical nonaxisymmetric magnetic models of the collapse and

133: fragmentation of a single core were presented by Nakamura \& Hanawa

134: (1997) and Nakamura \& Li (2002), without and with ambipolar diffusion,

135: respectively, using the magnetic thin-disk approximation (Ciolek \&

136: Mouschovias 1993, hereafter CM93). The early stages of core formation in

137: a nonaxisymmetric infinitesimally thin subcritical sheet (including the

138: effects of magnetic tension but ignoring magnetic pressure) were studied

139: by Indebetouw \& Zweibel (2001). Here, we also study a planar cloud that

140: is perpendicular to the mean magnetic field, focusing on the case of

141: either exactly critical or decidedly supercritical cores; these cases

142: are shown to lead to observationally distinguishable outcomes. We again

143: use the thin-disk approximation, which allows for finite thickness

144: effects, and explicitly includes the effects of both magnetic pressure

145: and tension.

146: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

147: \vspace{-4ex}

148: \section{Physical and Numerical Model}

149: The fundamental equations we use to model molecular clouds as

150: self-gravitating, partially-ionized, isothermal thin sheets or disks

151: have been presented in axisymmetric form in CM93 and Basu \& Mouschovias

152: (1994, hereafter BM94). In this study, the condition of axisymmetry is no

153: longer employed, and instead we model clouds as thin sheets of infinite

154: extent [with vertical half-thickness $Z(x,y,t)$] in a cartesian

155: coordinate system ($x,y$). Magnetohydrostatic equilibrium in the

156: $z$-direction (i.e., balance of thermal-pressure, gravitational, and

157: magnetic pinching forces) is assumed at all times. The evolution is

158: followed in a square region using periodic boundary conditions. To find

159: the gravitational field components in the sheet, we calculate the

160: gravitational potential $\psi(x,y,t)$. For this situation, solving

161: Poisson's equation (see eq. [29] of CM93) relates $\psi$ to the column

162: density $\sign(x,y,t)$ ($= \int_{-Z}^{Z} \rhon(x,y,t)dz$; $\rhon$ is the

163: mass density) by

164: \beq

165: \label{FTeq}

166: %{\cal F} [\psi] = - \frac{2 \pi G {{\cal F}[\sign]}}{k_{z}}  ,

167: {\cal F} [\psi] = - 2 \pi G {{\cal F}[\sign]}/k_{z}  ,

168: \eeq

169: where ${\cal F}[f]$ is the Fourier transform of a function $f$, $k_{z}$

170: ($= [k_{x}^{2} + k_{y}^{2}]^{1/2}$) is the wavenumber, and $G$ is the

171: gravitational constant. In our governing equations, the effects of

172: magnetic pressure and magnetic tension are both included. The magnetic

173: potential $\Psi(x,y,t)$ that is used to determine the $x$- and $y$-

174: components of the magnetic field at the surface of the sheet (necessary

175: to calculate magnetic tension forces) is also given by equation

176: (\ref{FTeq}), by substituting $\psi$ with $\Psi$, and $2 \pi G \sign$

177: with $-(\Beq - \Bref)$; for details see CM93.

178: $\Beq(x,y,t)$ is the vertical magnetic field strength in the sheet, and

179: $\Bref$ is the uniform magnetic field of the background (reference)

180: state, as well as the assumed constant magnetic field as

181: $|z| \rightarrow \infty$.

182: In the present study, we neglect magnetic braking due to a finite

183: density external medium (BM94), and dust grains and UV ionization (CM93,

184: Ciolek \& Mouschovias 1995).

185: 

186: In our formulation, velocities are normalized to the isothermal speed of

187: sound $\cs = 0.188 \, (T/10~{\rm K})^{1/2}\kms$ (where $T$ is the

188: temperature and we have used a mean molecular mass $m = 2.33 \mH$, in

189: which $\mH$ is the hydrogen atom mass), column densities are in units of

190: $\signi$, the uniform column density of the background state. The time

191: unit is $t_0 = \cs/2\pi G \signi$, the length unit is

192: $L_0 = \cs^{2}/2 \pi G \signi$, and the magnetic field unit is

193: $B_0 = 2 \pi G^{1/2} \signi$.

194: As discussed in CM93 and BM94, model clouds are distinguished by

195: three basic dimensionless parameters, namely, the initial mass-to-magnetic

196: flux ratio (in units of the critical value for gravitational collapse)

197: $\mui \equiv B_0/\Bref = 2 \pi G^{1/2} \signi/\Bref$, the initial

198: dimensionless neutral-ion collision time $\tnii \equiv \tau_{\rm ni}/t_0$, and

199: the normalized external pressure acting on the disk, $\Pext \equiv

200: \Pe/(\frac{\pi}{2} G \signi^2)$. The ion number density $\nion$ is

201: calculated using the scaling law $\nion \propto \nn^{1/2}$ (as done by

202: BM94; $\nn$ is the neutral number density), which is a reasonable

203: approximation for $\nn \lesssim 10^{6} \cmc$ (e.g., Ciolek \& Mouschovias

204: 1998). For a background number density $\nni = 3 \times 10^3 \, \cmc$

205: and $\Pext = 0.1$, we find that $\signi = 5.98 \times 10^{-3}$ g cm$^{-2}$

206: (therefore $N_{\rm n,0} \equiv \signi/m = 1.54 \times 10^{21} \cms$),

207: $L_0 = 4.57 \times 10^{-2}$ pc, and $t_0 = 2.38 \times 10^5$ yr.

208: 

209: We use the method of lines technique (e.g., Schiesser 1991); the

210: system of partial differential equations of two-fluid MHD are converted

211: to a system of ordinary differential equations (ODE's) through second-order

212: spatial finite differencing, and use of the van Leer (1977) advection scheme.

213: Time integration of the ODE's is performed using the

214: implicit Adams-Bashforth-Moulton method. Details of

215: our basic technique can be found in Morton, Mouschovias, \& Ciolek

216: (1994). Fast Fourier transform (FFT) routines are used to solve for the

217: gravitational and magnetic potentials obtained through equation

218: (\ref{FTeq}). The computational region has extent

219: $4 \lammax$ on each side, where $\lammax=4 \pi L_0$ is

220: the wavelength with maximum growth rate of gravitational instability

221: in a non-magnetic (thermal) infinitesimally thin disk.

222: Models presented in this paper are run on a uniform grid

223: of $128 \times 128$ points.

224: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

225: \vspace{-4ex}

226: %\section{Core Formation in Model Clouds}

227: \section{Results}

228: We present the evolution of two model clouds;

229: one model is exactly critical,

230: with $\mui =1$, and the other is supercritical with $\mui =2$.

231: Both models have $\tnii = 0.16$ (as adopted in the standard model of

232: BM94) and $\Pext = 0.1$.

233: Since the background

234: (reference) state is characterized by a uniform column density $\signi$ and

235: magnetic field $\Bref \zhat$ (where $\Bref = B_0/\mui$)

236: the gravitational and magnetic forces in the reference state

237: are identically zero.

238: %, similar to that in the standard Jeans stability analysis.

239: To initiate evolution, we superimpose

240: a set of random, small-amplitude (the rms is 3\% of the background)

241: column density perturbations $\delta \sign$. The magnetic field is

242: %simultaneously 

243: also

244: perturbed, with $\delta \Beq \propto \delta \sign$, 

245: %so as

246: to maintain flux-freezing in the initial state of each model. The

247: spectrum of perturbations is flat (white noise), with damping so that

248: wavelengths of twice the grid spacing and smaller are negligible; we

249: choose this particular type of spectrum so as not to preferentially

250: excite modes with the maximum growth rate.

251: The evolution of each model is followed until the maximum

252: column density $\approx 10 \, \signi$. Beyond this

253: enhancement, gravitational instability cannot be spatially resolved, and

254: the evolution is relatively very rapid in a small region near the

255: density peaks, making a simulation of the larger cloud impractical even

256: at higher resolution. Since the vertical balance along field lines is

257: primarily between gas pressure $\rhon \, \cs^2$ and self-gravitational

258: pressure $\frac{\pi}{2} G \sign^2$, the density has increased by

259: a factor $\approx 100$.

260: 

261: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

262: \subsection{Critical model ($\mui = 1$)}

263: Figure 1 exhibits

264: color images of the normalized column density $\sign(x,y)/\signi$ and

265: the local mass-to-flux ratio $\mu(x,y) = 2 \pi G^{1/2} \sign/\Beq$

266: over the entire computational region for the critical model

267: at its final output time, $t = 133.9 \, t_0$.

268: The time is rather long due to the very

269: small-amplitude perturbations and because magnetic forces exactly balance

270: gravity in the flux-freezing limit, requiring instability to

271: be initiated by random ion-neutral drifts.

272: Two major density peaks have formed by this time. The column density

273: structure shows that the lowest density contours are the most

274: irregular and elongated. Henceforth,

275: we define a dense core to be the region within the $\sign/\signi = 2$ contour,

276: corresponding to the observationally significant threshold

277: $\nn \approx 10^4 \cmc$. The masses of the two dense cores in

278: this simulation are $5.3 M_{\odot}$ (lower left of Fig. $1a$)

279: and $4.2 M_{\odot}$ (upper right of Fig. $1a$), respectively, using

280: the standard values of units given in \S\ 2.

281: For comparison, the enclosed mass in our periodic computational domain

282: is $151 M_{\odot}$.

283: %Supercritical cores (with $\muc > 1$) have formed by this time;

284: The shape of the cores are generally nonaxisymmetric, a result that was

285: first seen in the magnetically subcritical planar ambipolar diffusion

286: models of Indebetouw \& Zweibel (2000).

287: %Low density contours are the

288: %most irregular and elongated. If we identify a dense core as existing

289: %within the

290: %contour level $\sign/\signi = 2$ (corresponding to $\nn \approx 10^4 \cmc$),

291: The mean value of $Z \equiv \sign/(2 \rhon)$ implies one oblate core

292: with relative axis lengths 0.15:1:1, and another that is triaxial with

293: relative lengths 0.1:0.7:1, making it more nearly oblate than prolate.

294: The cores formed in this model always have their shortest axis aligned

295: normal to the sheet (parallel to the mean magnetic field). Although one

296: core is effectively oblate in this particular model realization, our

297: simulations reveal they are most commonly triaxial and more nearly

298: oblate than prolate, in accordance with statistical analyses of observed

299: cores (Jones et al. 2001).

300: 

301: Velocity vectors of the neutrals are also displayed in Figure $1a$.

302: Calculation of the rms speeds of neutrals ($v_{\rm n,rms}$) and ions

303: ($v_{\rm i,rms}$) in the entire computational region yield

304: $v_{\rm n,rms} = 0.114 \, \cs = 1.14 \, v_{\rm i,rms}$; ions lag

305: behind neutrals because the motions are ultimately

306: {\em gravitationally driven}. Furthermore,

307: {\em all of the infall speeds in this model are subsonic}. The maximum

308: infall speed is $|\vn| = 0.54 \, \cs$, and speeds of this order occur only

309: within the cores. As we shall see in \S~3.2, subsonic infall is a

310: {\em distinct} feature of ambipolar-diffusion initiated collapse in clouds

311: with $\mui \lesssim 1$. This trend is consistent with multi-line studies

312: that find subsonic gas speeds in dense protostellar cores (Tafalla et

313: al. 1998; Williams et al. 1999), and with axisymmetric models of core

314: formation in subcritical clouds (Ciolek \& Basu 2000).

315: \begin{figure}

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

317: \caption{Critical model ($\mui = 1$). The data are shown when the maximum

318: column density $\approx 10 \, \signi$. ({\it a, Left}.) Image and

319: contour plot of normalized column density $\sign(x,y)/\signi$. The

320: contour lines are spaced in multiplicative increments of $2^{1/2}$. Also

321: shown are velocity vectors of the neutrals; the distance between tips of

322: vectors corresponds to a speed $0.5 \, \cs$. ({\it b, Right.}) Image and

323: contours of $\mu(x,y)$, the mass-to-flux ratio in units of the critical

324: value for collapse. Regions with $\mu < 1$ are black. The contour lines

325: are spaced in additive increments of 0.1.}

326: \end{figure}

327: Figure $1b$ shows another interesting result: {\em the cores are

328: supercritical ($\muc > 1$) and are surrounded by magnetically subcritical

329: envelopes ($\muenv < 1$)}. This is a natural consequence of ambipolar

330: diffusion, which redistributes mass in magnetic flux tubes (Mouschovias 1978).

331: Due to the initial precise balance between gravitational and magnetic

332: forces in this model, evolution occurs as ambipolar diffusion effects a

333: drift of mass through essentially stationary magnetic field lines. In

334: time, this leads to the formation of both supercritical cores

335: and a subcritical envelope. This can also occur in clouds with 

336: %a value of 

337: $\mui$ 

338: %that is 

339: slightly above unity.

340: 

341: Figure 2 shows the column density $\sign/\signi$, vertical magnetic

342: field $\Beq/B_0$, and $x$-component of neutral velocity $\vnx/\cs$

343: along the $x$-axis for a line that cuts through one of the cores shown

344: in Figure 1. The core stands

345: out as a well-defined high density region with an eventual merger into a

346: lower-density background (the remnant of the initial uniform state) that

347: surrounds it; this is reminiscent of the mid-infrared

348: absorption maps of dense cores made by Bacmann et al. (2000).

349: The magnitude of $\vnx$ increases inward toward the core

350: center (but remains subsonic throughout) before dropping to zero at

351: the core center.

352: %The magnitude of the infall speed is subsonic throughout, with the largest

353: %values occurring near the core center. At the actual core center,

354: %$\vnx$ goes to zero and changes sign on opposite sides, because

355: %the motion is directed towards the gravitational point of attraction.

356: \begin{figure}

357: \epsscale{0.5}

358: \plotone{f2.eps}

359: \caption{Physical quantities in the $\mui = 1$ model, along a line

360: parallel to the $x$-axis that cuts through a supercritical core (shown

361: in Figure 1) centered at $x = 0.64 \, \lammax, y = 0.70 \, \lammax$.

362: Solid curve: column density $\sign/\signi$. Dashed curve: magnetic field

363: strength $\Beq/B_0$. Dotted curve: $x$-component of the neutral

364: velocity, $\vnx/\cs$.}

365: \end{figure}

366: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

367: \subsection{Supercritical model ($\mui = 2$)}

368: Figure $3a$ exhibits a color image and contour plot of $\sign/\signi$,

369: as well as velocity vectors for the supercritical model at

370: its final output time, $t = 17.6 \, t_0$.

371: The time is significantly lesser than in

372: the previous model due to the relative ease of gravitational instability

373: in a supercritical cloud. Because the evolution is more rapid and

374: dynamical in the supercritical model, ambipolar diffusion doesn't have

375: as much time to operate.

376: Hence, the mass-to-flux ratios of the high-density

377: cores in this model ($\muc \simeq 2.1$) are only slightly greater than

378: the initial value. The core shapes are again nonaxisymmetric,

379: either near-oblate or decidedly triaxial.

380: For this model we find $v_{\rm n,rms} = 0.222 \, \cs = 1.05 \, v_{\rm i,rms}$;

381: these results differ from the critical model due to the more dynamical

382: evolution and greater ability of neutrals to drag ions and magnetic

383: field with them. The maximum infall speed is $1.19 \, \cs$, and speeds of

384: this order typically occur within distances $\approx 0.1$ pc from the core

385: centers. In contrast to the critical model, {\it the infall speeds

386: within the cores in this model are often supersonic, and significant

387: motions $\approx 0.5 \, \cs$ are also seen throughout the

388: cloud, and well outside the cores}. Figure $3b$ shows $\sign/\signi$,

389: $\Beq/B_0$, and $\vnx/\cs$ along the $x$-axis for a line passing

390: through one of the cores in Figure $3a$.

391: The collapsing core again has a density profile that eventually

392: merges into the background.

393: 

394: Our model shows that the extent of the supersonic flow region is well

395: within the resolution capability of current observations (scales

396: $\lesssim 0.1~{\rm pc}$, $\nn \approx 10^{4}-10^{5}\cmc$). Since

397: supersonic infall has not been detected over these length scales and

398: densities, and models with $\mui \gtrsim 2$ would have even greater

399: infall speeds, this suggests that {\em molecular clouds that are

400: supercritical by a factor $\gtrsim 2$ are incompatible with 

401: %present-day

402: observations of protostellar cores.}

403: \begin{figure}

404: \epsscale{1.0}

405: \plottwo{f3a.eps}{f3b.eps}

406: \caption{Supercritical model ($\mui = 2$). The data are shown when the

407: maximum column density $\approx 10 \, \signi$. ({\it a, Left}.) Image

408: and contour plot of normalized column density $\sign(x,y)/\signi$. The

409: contour lines are spaced in multiplicative increments of $2^{1/2}$. Also

410: shown are velocity vectors of the neutrals; they are normalized in the same

411: way as in Fig. $1a$. ({\it b, Right.}) Physical

412: quantities along a line parallel to the $x$-axis that cuts through a

413: supercritical core centered at

414: $x = 1.42 \, \lammax, y = -0.33 \, \lammax$. Solid curve: column density

415: $\sign/\signi$. Dashed curve: magnetic field strength $\Beq/B_0$.

416: Dotted curve: $x$-component of the neutral velocity, $\vnx/\cs$.}

417: \end{figure}

418: 

419: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

420: \section{Conclusions}

421: Ambipolar diffusion leads to a nonuniform distribution of mass-to-flux

422: ratio, in a natural extension of the process described by Mouschovias

423: (1978). Stars form preferentially in the most supercritical regions.

424: Core shapes are somewhat triaxial, and usually

425: more nearly oblate than prolate. The core column density

426: eventually merges into a near-uniform

427: background value. In the critical ($\mui = 1$) model, a surrounding region is

428: established which is mildly magnetically subcritical

429: (due to flux redistribution) and infall motions both inside and

430: outside cores are subsonic. Conversely, the

431: supercritical ($\mui = 2$) model exhibits supersonic motions within cores, and

432: extended rapid motions outside them. The critical model requires a

433: significantly longer time to develop gravitational instability; however,

434: we caution that the growth time for both models are likely upper limits

435: due to the possibility of nonlinear perturbations in more realistic

436: situations. We also note that all motions in these models are

437: fundamentally gravitationally driven; the neutral speeds are

438: typically greater than those of the ions.

439: 

440: %We have demonstrated that ambipolar diffusion acting in a cloud of

441: %initially uniform mass-to-flux ratio leads to a nonuniform distribution

442: %of the same, in a natural extension of the process described

443: %by Mouschovias (1978).

444: %Ambipolar diffusion naturally leads to a nonuniform distribution of

445: %mass-to-flux ratio; stars form preferentially in the most supercritical

446: %regions (e.g., Mouschovias 1987; CM93; BM94). As demonstrated here,

447: %If the background value $\mui$ is close to unity, flux redistribution leads

448: %to subcritical envelopes and the magnetic field may ultimately prevent

449: %their dynamical collapse onto newly formed protostars. This may provide

450: %at least a partial explanation for the observed inefficiency of star

451: %formation.

452: %useful framework within which to understand the inefficiency of star

453: %formation.

454: 

455: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

456: \begin{acknowledgements}

457: We thank Craig Markwardt for providing some key IDL routines.

458: SB is supported by a grant from the Natural Sciences and Engineering

459: Research Council of Canada. GC is supported by NASA grant

460: NAG 5-7589 to the New York Origins of Life Center (NSCORT).

461: \end{acknowledgements}

462: 

463: \begin{thebibliography}{}

464: 

465: \bibitem[]{Bac00} Bacmann, A., Andr\'{e}, P., Puget, J.-L., Abergel, A.,

466: Bontemps, S., \& Ward-Thompson, D. 2000, \aap, 361, 555

467: \bibitem[]{B00}  Basu, S. 2000, \apjl, 540, L103

468: \bibitem[]{BM94} Basu, S., \& Mouschovias, T. Ch. 1994, \apj, 432, 720 (BM94)

469: \bibitem[]{CB00} Ciolek, G. E., \& Basu, S. 2000, \apj, 529, 925

470: \bibitem[]{CM93} Ciolek, G. E., \& Mouschovias, T. Ch. 1993, \apj, 454,

471: 194 (CM93)

472: \bibitem[]{CM95} Ciolek, G. E., \& Mouschovias, T. Ch. 1995, \apj, 418, 774

473: \bibitem[]{CM98} Ciolek, G. E., \& Mouschovias, T. Ch. 1998, \apj, 504, 280

474: \bibitem[]{C99} Crutcher, R. M. 1999, \apj, 520, 706

475: %\bibitem[]{FM93} Fiedler, R. A., \& Mouschovias, T. Ch. 1993, \apj, 415, 680

476: \bibitem[]{IZ00} Indebetouw, R., \& Zweibel, E. G. 2000, \apj, 532, 361

477: \bibitem[]{JBD01} Jones, C. E., Basu, S., \& Dubinski, J. 2001, \apj, 551, 387

478: %\bibitem[]{LN02} Li, Z.-Y., \& Nakamura, F. 2002, \apj, 578, 256

479: \bibitem[]{MMC94} Morton, S. A., Mouschovias, T. Ch., \& Ciolek, G. E. 1994,

480: \apj, 421, 561

481: \bibitem[]{MAN98} Motte, F., Andr\'{e}, P., \& Neri, R. 1998, \aap, 336, 150

482: \bibitem[]{M78} Mouschovias, T. Ch. 1978, in Protostars \& Planets, ed. T.

483: Gehrels (Tucson: Univ. Arizona), 209

484: %\bibitem[]{M87} Mouschovias, T. Ch. 1987, in Physical Processes in

485: %Interstellar Clouds, ed. G. Morfil \& M. Scholer (Dordrecht: Reidel), 453

486: \bibitem[]{M91} Myers, P. C., Fuller, G. A., Goodman, A. A., \& Benson,

487: P. J. 1991, \apj, 376, 561

488: \bibitem[]{NH97} Nakamura, F., \& Hanawa, T. 1997, \apj, 480, 701

489: \bibitem[]{NL02} Nakamura, F., \& Li, Z.-Y. 2002, \apj, 566, L101

490: \bibitem[]{Sch91} Schiesser, W. E. 1991, The Numerical Method of Lines:

491: Method of Integration of Partial Differential Equations (San Diego:

492: Academic)

493: \bibitem[]{Ta98} Tafalla, M., Mardones, D., Myers, P. C., Caselli, P.,

494: Bachiller, R., \& Benson, P. J. 1998, \apj, 504, 900

495: \bibitem[]{van77} van Leer, B. 1977, JCP, 23, 276

496: \bibitem[]{Wil99} Williams, J. P., Myers, P. C., Wilner, D. J.,

497: \& DiFrancesco, J. 1999, \apj, 513, L61

498: 

499: \end{thebibliography}

500: 

501: 

502: 

503: \end{document}

504: 

505: