1: %\documentstyle[11pt,newpasp,twoside,epsf]{article}

2: \documentclass[11pt,twoside]{article}

3: \usepackage[dvips]{graphics}

4: \usepackage{star2000}

5: 

6: \markboth{Gilmore \& Wyse}{The Thick Disk-Halo Interface}

7: 

8: \pagestyle{myheadings}

9: %\nofiles

10: 

11: \begin{document}

12: 

13: \title{The Thick Disk-Halo Interface}

14: 

15: \author{Gerard Gilmore}

16: \affil{Institute of Astronomy, Cambridge, UK}

17: 

18: \author{Rosemary F.G. Wyse}

19: \affil{Johns Hopkins University, Baltimore, MD, USA}

20: 

21: 

22: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

23: 

24: \begin{abstract}

25: The star formation history of a galaxy, explicitely here our Milky Way

26: Galaxy, where the most detailed information is attainable, is the

27: convolution of two functions. One function describes the rate of

28: formation of the stars which are today in the Galaxy. The second

29: describes the assembly of those stars into the present Galactic

30: potential well. There is direct evidence that this assembly continues

31: today, with both stars and gas being assembled into, or at least

32: rearranged in, the Galactic potential. But is this recent accretion

33: significant? Was the last significant accretion the formation

34: of the thick disk, some 10Gyr ago?. Recent spectroscopic studies

35: support this unexpected result, while dynamical studies find

36: increasing numbers of specific examples of smaller scale more recent

37: accretion. We present early results for one specific such survey, the

38: Anglo Australian Old Stellar populations Survey, to illustrate current

39: studies. 

40: \end{abstract}

41: 

42: 

43: 

44: \section{Old Stellar Populations: the Context}

45: 

46: While present stellar populations are a undoubtedly a manifestation of the fossil

47: record of Galactic evolution, quantification and interpretation of

48: that fossil record remains a subject of lively debate and rapid progress. 

49: Among the key issues are the places and times of formation of the

50: oldest stellar populations : the halo, thick disk and bulge - and

51: their overlaps and evolutionary relationships, if any. Analysis of

52: these stellar populations will in principle quantify the history of

53: merging and accretion in a typical galaxy, of great importance in

54: determining galactic evolution, and constraining cosmological theories

55: of galaxy formation.

56: 

57: The metallicity and kinematic distribution functions of complete

58: samples of long-lived stars have long been recognised as providing

59: unique constraints on the early stages of chemical evolution of the

60: Galaxy.  The main sequence lifetime of F/G dwarf stars is greater than

61: the age of the Galaxy and hence the chemical-abundance distribution

62: function of such stars provides an integrated record of the

63: chemical-enrichment history without the need for model-dependent

64: corrections for dead stars ( van den Bergh 1962; Tinsley

65: 1980).  Pioneering studies focussed on the only

66: reasonably-complete sample available, which is that for stars in the

67: immediate solar neighborhood; in effect stars within about 30pc of the

68: Sun.  These samples have been sufficiently small that reliable study

69: of those stellar populations whose kinematics are such that member

70: stars spend only a small fraction of an orbit in the solar

71: neighborhood has necessarily been difficult. This is potentially a

72: serious restriction, as such stars might in principle be a major

73: contributor to the stellar population in a valid, representative

74: volume of the Galaxy.  In addition, intrinsically-rare stellar

75: populations are missed entirely.

76: 

77: Thus, it is important, in deriving a reliable determination of

78: Galactic structure and evolution, that one consider the joint

79: distributions functions over chemical abundance and kinematics.

80: 

81: The observational situation has been improved recently in two ways: by

82: collection and analysis of spectroscopic data for all-sky samples of

83: stars extending to somewhat greater, but still essentially local,

84: distances (Norris, Bessel \& Pickles 1985; Carney etal

85: 1990; Beers etal 1999; Carney, this volume; Chiba \& Beers 2000), 

86: and by deeper pencil-beam surveys, to isolate {\sl in

87: situ} samples of old disk (Kuijken \& Gilmore 1989a), thick

88: disk (Gilmore, Wyse, \& Jones 1995) and halo stars. The

89: combination of the large but local samples with the small but distant

90: samples has allowed the deconvolution, to first order, of the

91: abundance distribution functions, and the mean velocity dispersions,

92: of the dominant Galactic populations. While our understanding of

93: Galactic structure and evolution has advanced considerably of late,

94: extension of these analyses has become limited by the intrinsic

95: breadth and overlap of the population distribution functions and by

96: the small size of the available {\sl in situ} samples.

97: 

98: 

99: The theoretical situation has also become more specific.  Though the

100: many dynamical, structural and chemical evolution questions one poses

101: concerning galactic evolution may seem well-defined and relatively

102: distinct, it is now clear that the answers are intimately

103: interrelated. For instance, galaxies probably accrete their

104: neighbours, so that the place of origin of a star may be far from its

105: present location; dynamical instabilities in disks result in the

106: mixing through phase space of stellar populations, further blurring

107: the relation between a star's present location and its birthplace. Bar

108: instabilities are also likely to cause significant gas transport, and

109: may drive star bursts and possibly nuclear non-thermal phenomena.

110: Major mergers may thicken disks. Bulges may be accreted, or created

111: during mergers.

112: 

113: Modern models of Galaxy formation make fairly specific predictions

114: concerning each of these possibilities. A detailed review is provided

115: by Silk \& Wyse (1993) where further

116: discussion may be found. For example, fashionable Cold Dark Matter

117: models, which contain aspects of both the monolithic (`ELS') and the

118: multi-fragment (`Searle-Zinn') pictures often discussed in chemical

119: evolution models, `predict' growth of the Galaxy about a central core,

120: which should contain the oldest stars. Later accretion of material

121: forms the outer halo and the disks, while continuing accretion will

122: continue to affect the kinematic structure of both the outer halo and

123: the thin disk.  Considerable phase-space substructure should be

124: detectable, when one looks sufficiently far from the Plane and the

125: Galactic centre that dynamical timescales are long (Ibata, Gilmore, \&

126: Irwin 1995; Arnold \& Gilmore 1992), and has

127: recently been seen with plausible significance (Helmi etal 1999).

128: 

129: Dissipational models for thick-disk formation predict observable

130: abundance gradients (cf. Burkert, Hensler and Truran 1992),

131: and similar scale lengths for the thick and thin disks (Ferrini

132: etal 1994).  Specific column-integral abundance

133: distributions can be calculated (numerically) for some of these models

134: and compared to observations.

135: Satellite merger models for thick disk formation require the stars

136: from the satellite to be detectable, as a tail in the thick disk

137: distribution functions below [Fe/H]=$-1$ (Silk \& Wyse, 1993)

138: `Continuum' models of thick disk formation from the

139: thin disk require that an accurately defined joint distribution

140: function over chemical abundance and kinematics for the oldest stars

141: be smooth and continuous (Norris \& Ryan, 1991).

142: Alternative models, such as the discrete merger model, can then be

143: distinguished by their prediction that the distibutions overlap in

144: abundance, and perhaps velocity dispersion, but not in angular

145: momentum (Gilmore, Wyse \& Kuijken 1989).

146: Most detailed models make specific predictions concerning the

147: abundance distribution function in a cylinder, through the Galactic

148: disk - the `G-dwarf problem' -- which remains widely studied, and a

149: valuable diagnostic of early accretion and gas flows in the disk

150: (Pagel \& Patchett 1975).

151: 

152: That is, quantitative study of the essential physics of galaxy

153: evolution, requires that one must study the distributions over

154: chemistry, kinematics and spatial structure of the oldest stars (eg

155: Sandage \& Fouts 1987).

156: Determination of the wings of the distribution functions, and their

157: separation or deconvolution, is however feasible, given adequate

158: samples.  One such project, which we introduce here, is the 

159: Anglo-Australian Old Stellar Populations Survey, with joint UK (the

160: present authors) and Australian (J Norris, K Freeman) involvement.

161: 

162: 

163: \section{Old Populations: what should one observe?}

164: 

165:  The ideal tracer of Galactic Structure is one which is selected

166: without any biases, does not suffer from stellar age-dependent

167: selection effects, is representative of the underlying populations,

168: and is easily observable.  Historically, the need for easy observation

169: restricted studies to the immediate solar neighborhood. The primary

170: limitation of the nearby star sample is its small size.  This

171: inevitably means that stars which are either intrinsically rare --

172: such as halo population subdwarfs -- and stars which are common but

173: whose spatial distribution is such that their local volume density is

174: small -- such as thick disk stars -- are poorly represented.  Most

175: recent and current efforts to extend present local volume-limited

176: samples to include minority populations have, for practical

177: observational reasons, utilized kinematically-selected samples defined

178: in the solar neighborhood, following the pioneering work of Eggen,

179: Lynden-Bell and Sandage (1962).  Subsequent correction for the

180: kinematic biases inherent in these samples requires careful modelling

181: (Norris and Ryan 1991).

182: An {\it in situ\/} sample, truly representative of the

183: dominant stellar population far from the Sun, circumvents these large,

184: model-dependent corrections.

185: 

186: Several surveys of tracer stars which can be observed {\it in situ}

187: are available.  Intrinsically luminous tracers are {\it a priori\/}

188: favored in terms of telescope time, but the likely candidates have

189: other characteristics that diminish their suitability: RR Lyrae stars

190: have intrinsic age and metallicity biases in that only stars of a

191: given range in metallicity and age exist in this evolutionary stage;

192: the accessible globular clusters are few in number; bluer horizontal

193: branch stars are also rare, and their color distribution depends on

194: chemical abundance and on the unidentified `second

195: parameter(s)'. K-giants are the most representative {\it evolved\/}

196: tracers of the spheroid, and have been used extensively.  However, one

197: must first identify giant stars from among the substantially larger

198: number of foreground K dwarfs with similar apparent magnitudes and

199: colors, and even with that selection, reliable determination of the

200: distance of a halo K giant has proven to be extremely difficult.

201: 

202: A desirable solution to these limitations, which has become feasible

203: with current multi-object spectroscopic systems and large-scale

204: photometric surveys, is to identify and study F/G dwarfs to

205: significant distances from the Sun.  This is the solution which we

206: have adopted.  Chemical abundances for these stars provide the

207: integrated record of the star formation and enrichment history during

208: the early stages of Galaxy formation, analogous to the local G-dwarf

209: distribution. Radial velocities allow discrimination between stellar

210: populations, when combined with abundances and spatial distributions.

211: 

212: Thus, the AAOSPS project is optimised to provide the next stage of

213: quantitative analysis of the structure, contents and evolution of the

214: early Milky Way, building from current observational and theoretical

215: expertise developed in pre-cursor phases of this study.

216: 

217: 

218: The scientific aim of the AAOSPS project is to determine the

219: distribution functions over metallicity, kinematics, and spatial

220: distributions of the oldest stellar populations, with particular

221: emphasis on the overlapping wings of each distribution function. To

222: achieve this, the primary technical requirement is that each of the

223: three variables - metallicity, radial velocity, distance - be

224: determined to a precision which is less than the intrinsic `cosmic'

225: dispersion in that parameter. 

226: 

227: 

228: \section{The Anglo Australian Old Stellar populations Survey:AAOSPS}

229: 

230: 

231: Is the thick disk a merger remnant?  How does it overlap the halo?

232: What are the systems that merge, how frequently does this happen over

233: a Hubble time, and with what consequence? Could we identify stars from

234: the intruder, and from our own early disk?  

235: 

236: To address these questions, we are using the two-degree-field

237: multi-object spectrograph (2dF) on the Anglo-Australian Telescope,

238: which provides 400 spectra simultaneously, to measure the radial

239: velocities and chemical abundances for F/G main sequence stars at

240: distances from the Sun of 3--7kpc down several key lines-of-sight. 

241: 

242: \subsection{AAOSPS: the specific goals}

243: 

244: 

245: Mergers and strong interactions between galaxies happen, as evidenced

246: today by the Sagittarius dwarf spheroidal galaxy (Ibata, Gilmore \&

247: Irwin 1994, 1995; Ibata et al.  1997).  The occurrence of a `minor

248: merger' between the Milky Way and a small satellite galaxy provides an

249: attractive explanation for the thick disk (see Gilmore \& Wyse 1985;

250: Gilmore, Wyse \& Kuijken 1989; Freeman 1993; Majewski 1993; Walker,

251: Mihos \& Hernquist 1996; Huang \& Carlberg 1997; Velazquez \& White

252: 1999).

253: 

254: What are the systems that merge, how frequently does this

255: happen over a Hubble time, and with what consequence?

256: 

257: Depending on the mass, density profile and orbit of the merging

258: satellite, `shredded-satellite' stars will leave a kinematic

259: signature, distinct from the canonical thick disk that will result

260: from the heated thin disk.  Satellites on prograde (rather than

261: retrograde) orbits couple better to the disk and provide more heating,

262: and thus are favoured to form the thick disk (e.g. Velazquez \& White

263: 1999). Thus one might expect a signature to be visible in the mean

264: orbital rotational velocity of stars, and for a typical satellite

265: orbit, lagging the Sun by more than does the canonical thick disk.

266: The relative number of stars in the `shredded satellite' versus the

267: heated-thin disk (now the thick disk) depends on the details of the

268: shredding and heating processes, and is a diagnostic of them, and may

269: well vary strongly with location.

270: 

271: We are measuring radial velocities and abundances for F/G main

272: sequence stars at distances from the Sun of 3--7kpc.  The sample is

273: selected on the basis of colour and apparent magnitude (V=19-20; 0.5 <

274: B-V < 0.9).  The 2dF spectra have 1 Angstrom/pixel and cover

275: 3700--4600 Angstroms. The external velocity accuracy of the data is

276: around 10--15 km/s, from repeat observations of program stars and from

277: a globular cluster standard.

278: We (mostly work by Norris) have extended the abundance

279: determination techniques developed by Beers et al. (1999).  As well as

280: abundances based on Ca II K, values are being determined from the

281: G-band of CH.  For stars with sufficient S/N we can derive

282: metallicities from an autocorrelation Fourier method, which uses all

283: the weak lines in the available spectral range.  Standard star data

284: show that abundances with precision better than 0.3dex are already

285: obtainable.  

286: 

287: Our primary targets are fields towards and against Galactic rotation,

288: to provide optimal halo/thick disk discrimination through orbital

289: angular momentum. We measure the time-integrated structure of the halo

290: and thick disk, evolved over many orbital times at these

291: Galactocentric distances.  This is the only statistical survey

292: targeting fields that probe the angular momentum, far from the thin

293: disk and without strong metallicity bias.

294: 

295: Our earlier multi-object (AUTOFIB) survey

296: demonstrated that there is a negligible fraction of

297: stars in the canonical thick disk that are younger than the globular

298: clusters of the same metallicity (Wyse \& Gilmore 1995) strengthening

299: the earlier inferences from kinematically-biased local surveys;

300: Gilmore \& Wyse 1985, Carney et al 1989.  This result limits the time

301: of the last significant merger event to be very early in the history

302: of the Galaxy (Gilmore, Wyse \& Jones 1995), challenging standard CDM

303: cosmologies (cf Wyse 2001).

304: 

305: \begin{figure}[!ht]

306: \plotfiddle{GG_fig1.eps}{3.25in}{270}{40}{40}{-155}{240}

307: \caption{Scatter plot of iron abundance {\it vs\/} B-V colour for

308: thick disk F/G stars, selected {\it in situ\/} in the South Galactic Pole 

309: at 1-2kpc above the

310: Galactic Plane (stars), together with the 14~Gyr turnoff colours

311: (crosses) from VandenBerg \& Bell (1985; Y=0.2) and 15~Gyr turnoff colours

312: (asterisks) from VandenBerg (1985; Y=0.25).  The open circle represents the

313: turnoff colour (de-reddened) and metallicity of 47~Tuc (Hesser et

314: al.~1987).  The vast majority of thick disk stars lie to the red of

315: these turnoff points, indicating that few, if any, stars in this

316: population are younger than this globular cluster.  This figure is

317: from Wyse (2001).}

318: \end{figure}

319: 

320: \subsubsection{First Results:}

321: 

322: We have detected a substantial population of low metallicity stars

323: with disk-like kinematics, intermediate between those of the canonical

324: thick disk and the canonical (non-rotating) stellar halo.  Figure 1

325: shows the radial velocity histogram for around 900 stars with high

326: signal-to-noise spectra, in a line-of-sight for which, at these

327: distances, radial V-velocity is approximately 0.8 times the rotational

328: lag behind the Sun's orbit. Our efficient selection against thin disk

329: contamination -- radial velocity near zero -- is apparent.

330: The smooth Gaussian represents a smooth halo in this

331: line-of-sight, appropriately normalised to fit the high-velocity data.

332: The canonical thick disk has a rotational lag of some 40km/s.  Thus

333: the large number of stars with radial velocity around 100km/s is not

334: expected, and probably traces a new kinematic component of the Milky

335: Way Galaxy.

336: 

337: \begin{figure}[htb!]

338: %\plotone{GG_fig2.ps}

339: \plotfiddle{GG_fig2.ps}{7.5cm}{0}{40}{40}{-120}{-60}

340: \caption{Radial velocity histogram for around 900 stars in a

341: line-of-sight chosen to probe orbital rotation.  The smooth curve

342: represents the stellar halo, and while it clearly is a reasonable

343: description of the shape of the distribution of the 

344: highest velocity stars, it

345: fails to describe the majority of the stars.  The canonical thick disk

346: provides the stars with radial velocity of less than 100km/s; the

347: broad shoulder between 100km/s and 200km/s is not expected. }

348: \end{figure}

349: 

350: 

351: The 100km/s stars are best interpreted as being the actual debris of the

352: satellite. The debris `stream' is  detected in

353: widely-separated lines-of-sight, but requires a larger

354: statistically-significant sample for confirmation and to allow

355: quantification of the properties of the former satellite galaxy.  This

356: quantification would be a strong constraint on hierarchical models of

357: galaxy formation.  

358: 

359: 

360: 

361: \begin{figure}[htb!]

362: \begin{center}

363: \plotfiddle{GG_fig3.ps}{7.5cm}{0}{100}{100}{-270}{-180}

364: \caption{2dF metallicities vs heliocentric radial velocity, for our

365: stars with sufficient S/N, in one line of sight.  Radial V velocity at

366: these distances in this line-of-sight is approximately Galactic

367: rotational velocity, with 180km/s approximately zero net orbital

368: rotation. The many stars with low metallicity and disk-like (small-V)

369: velocities are the identification of a low metallicity tail of the

370: thick disk. There is no metal-rich halo.  Lumpiness in the figure is

371: apparent: if real it will indicate phase-space structure, allowing

372: quantification of the past merger history.}

373: \end{center}

374: \end{figure}

375: 

376: 

377: \subsubsection{Metallicity and Phase Space Structure?}

378: 

379: The combination of kinematics and metallicity provides the best

380: constraints on stellar populations.  We are quantifying the

381: distributions of the Galaxy's populations in metallicity-velocity

382: space, to higher precision than simple Gaussian fits: it is these

383: distributions which encode galaxy formation.

384: 

385: Our first results (figure 2) show substantial numbers of stars with

386: very low metallicities, and very high angular momentum: that is, we

387: have discovered the much sought metal-weak thick disk, perhaps the

388: remnants of the Milky Way's last big merger.

389: 

390: 

391: 

392: Our velocity accuracy is adequate to identify any high-frequency

393: phase-space structure which may exist.  The clumpiness in the

394: metallicity vs velocity diagram shown, and the spikiness in the number

395: vs Galactic rotation velocity data, are kinematically resolved: our

396: goal now is to obtain sufficient numbers of adequate quality spectra to

397: quantify the statistical significance of these features, and the scale

398: length on the sky with which they are associated.

399: 

400: The key result is apparent from figure 2.  At every scale

401: we see mildly significant structure, and deviations from Gaussians:

402: are the groupings of stars in phase space, and the deviations from

403: kinematic smoothness, physical?  Statistical tests show that

404: substructure is significant, but only marginally, and only when the

405: data are restricted by angular scale length on the sky. This is

406: just what some spaghetti models predict (eg. Helmi \& White 1999; Helmi

407: et al 1999; Harding et al 2001). We are continuing to investigate its

408: reality.

409: 

410: \section{Conclusions:}

411: 

412: Modern large area surveys, complemented by deeper multi-object

413: facilities, are obtaining and analysing the combination of metallicity

414: and kinematic data for large samples of Galactic stars in the thick

415: disk -- halo interface.  Our preliminary results for one such study,

416: AAOSPS, show intriguing, but low statistical significance, deviations

417: from canonical distributions.  These may signal the remnant of the

418: satellite whose merger with the young Milky Way formed the thick disk,

419: the last high-impact merger that our Galaxy experienced.  Further, the

420: amplitude of small-scale lumpiness constrains the more recent merger

421: history, with recent detections from HIPPARCOS local data again being

422: complemented by more distant studies.  We are on the way to

423: deciphering the fossil record of the physical processes in the

424: formation and evolution of a typical large disk galaxy, the Milky Way.

425: 

426: 

427: \begin{references}

428: \reference 

429: Arnold, R \& Gilmore, G 1992 \mnras\ 257 225

430: \reference

431: Beers, T et al 1999, \aj\, 117, 981

432: \reference

433: Burkert, Hensler and Truran 1992 \apj\ 391, 651

434: \reference

435: Carney, B et al 1989, \aj\, 97, 423

436: \reference Chiba,M., \& Beers, T., 2000 \aj\ 119, 2843 

437: \reference

438: Eggen, O. Lynden-Bell, D. and Sandage, A. 1962 \apj\ 136 748

439: \reference

440: Ferrini, F. etal 1994 \apj\ 427 745

441: \reference

442: Freeman, K., 1993, ASP Conf series 49, eds Majewski, p12.

443: \reference

444: Gilmore, G \& Wyse, R. 1985, \aj\ 90, 2015

445: \reference

446: Gilmore,G., Wyse, R., \& Kuijken,K. 1989, ARAA, 27, 555

447: \reference

448: Gilmore,G.,  Wyse, R.,  \& Jones,B., 1995, \aj\ 109, 1095

449: \reference

450: Harding et al, 2001, \aj\ in press, astro-ph/0012307

451: \reference

452: Helmi, A., \& White, S. 1999, \mnras\ 307, 495 

453: \reference

454: Helmi,A.,  White,S., de Zeeuw, T. \& Zhao, H. 1999, Nature, 402, 55 

455: \reference

456: Huang, S. \& Carlberg, R. 1997, \apj\ 480, 503

457: \reference

458: Ibata,R., Gilmore, G., \& Irwin,M. 1994, Nature, 370, 191

459: \reference

460: Ibata,R., Gilmore, G \& Irwin,M. 1995, \mnras\  277, 781

461: \reference

462: Ibata,R., Wyse,R., Gilmore,G., Irwin,M. \& Suntzeff,N. 1997, \aj\ 113, 634

463: \reference

464: Kuijken,K. \& Gilmore, G. 1989 \mnras\ 239 571

465: \reference

466: Majewski, S., 1993, ARAA,  31, 575

467: \reference

468: Norris, J., 1996, ASP Conf Series 92, eds Morrison and Saradajeni, p14

469: \reference

470: Norris, J., Bessel, M. \& Pickles, A. 1985 ApJS 58 463

471: \reference

472: Norris, J. \& Ryan,S. 1991, \apj\ 340 739

473: \reference

474: Pagel, B. \& Patchett,B. 1975, \mnras\ 172, 13

475: \reference

476: Silk,J. \& Wyse,R. 1993, Physics Reports 231 293

477: \reference

478: Tinsley, B.  1980 Fund Cosmic Phys 5 287

479: \reference

480: van den Bergh, S. 1962 \aj\ 67 486

481: \reference VandenBerg, D. 1985, ApJS, 58, 711

482: \reference VandenBerg, D. \& Bell, R. 1985, ApJS, 58, 561

483: \reference

484: Velazquez,A., \& White, S. 1999, \mnras\ 304, 254

485: \reference

486: Walker, C., Mihos,C. \& Hernquist, L. 1996, \apj\ 460, 121

487: \reference

488: Wyse, R. 2001, astro-ph/0012270, to appear in Galactic Disks and Disk

489: Galaxies.

490: \reference

491: Wyse, R., \& Gilmore, G. 1995 \aj\ 110, 2771

492: \end{references}

493: 

494: 

495: 

496: 

497: 

498: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

499: \end{document}

500: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

501: 

502: 

503: \begin{figure}[htb!]

504: %\plotfiddle{w6_lagr.ps}{7.5cm}{90}{45}{45}{200}{-10}

505:   \caption{Evolution of mass shells (Lagrange radii) for the model

506:    $(W_{0,i},\Omega_{0,i})=(6.0,0.60)$. Shown are the radii for mass

507:    columns containing the indicated percentage of total mass

508:   \label{fig:lag}

509:    in the direction of the $\theta = 54.74^{\circ}$-angle,

510:    the tidal radius $r_{\rm tid}$ determined from

511:    ($\phi(R,z)=E_{\rm tid}$),

512:    and the core radius

513:    $r_{c}=  ( 9\sigma^{2} / (4\pi G \rho )^{\frac{1}{2}} $.

514:   }

515: \end{figure}

516: 

517: