1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2: %% Trim Size: 9.75in x 6.5in

3: %% Text Area: 8in (include Runningheads) x 5in

4: %% ws-ijmpd.tex   :     19-4-07

5: %% Tex file to use with ws-ijmpd.cls written in Latex2E.

6: %% The content, structure, format and layout of this style file is the

7: %% property of World Scientific Publishing Co. Pte. Ltd.

8: %% Copyright 1995, 2002 by World Scientific Publishing Co.

9: %% All rights are reserved.

10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

11: %

12: 

13: %\documentclass[draft]{ws-ijmpd}

14: \documentclass{ws-ijmpd}

15: 

16: \begin{document}

17: 

18: \markboth{Mark J. Reid}

19: {}

20: 

21: %%%%%%%%%%%%%%%%%%%%% Publisher's Area please ignore %%%%%%%%%%%%%%%

22: %

23: \catchline{}{}{}{}{}

24: %

25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

26: 

27: \title {IS THERE A SUPERMASSIVE BLACK HOLE\\

28: AT THE CENTER OF THE MILKY WAY?}

29: 

30: %\runninghead{Supermassive Black Hole}

31: 

32: \author{Mark J. Reid}

33: 

34: \address{Harvard--Smithsonian Center for Astrophysics\\

35:          60 Garden Street, Cambridge, MA 02138, USA\\

36:          mreid@cfa.harvard.edu}

37: 

38: 

39: %% manuscript produces a one-column, double-spaced document:

40: % \documentclass[manuscript]{aastex}

41: %% preprint2 produces a double-column, single-spaced document:

42: % \documentclass[preprint2]{aastex}

43: 

44: \def\Rsch           {\ifmmode {R_{\rm Sch}} \else  {$R_{\rm Sch}$} \fi}

45: \def\MoverR         {\ifmmode {M/R} \else  {$M/R$} \fi}

46: \def\Mh             {\ifmmode {M_h} \else  {$M_h$} \fi}

47: \def\rhoh           {\ifmmode {\rho_h} \else  {$rho_h$} \fi}

48: \def\Msun           {M$_\odot$}

49: \def\Msunpercpc     {M$_\odot$~pc$^{-3}$}

50: \def\kms            {km~s$^{-1}$}

51: \def\etal           {et al.}

52: \def\ie             {i.e.}

53: \def\eg             {e.g.}

54: \def\NeII           {Ne~{\small II}}

55: \def\um             {$\mu$m}

56: \def\SgrA           {Sgr~A*}

57: \def\SgrAwest       {Sgr~A-West}

58: \def\SgrAeast       {Sgr~A-East}

59: \def\p              {\phantom{>}}

60: \def\q              {\phantom{0}}

61: 

62: \maketitle

63: 

64: \begin{history}

65: \received{2008 August 18}

66: \revised{}

67: \comby{Managing Editor}

68: \end{history}

69: 

70: \begin{abstract}

71: This review outlines the observations that now 

72: provide an overwhelming scientific case that the center of our

73: Milky Way Galaxy harbors a supermassive black hole. 

74: Observations at infrared wavelength trace stars that 

75: orbit about a common focal position and require a central mass ($M$) 

76: of $4\times10^6$~\Msun\ within a radius of 100~AU.  Orbital 

77: speeds have been observed to exceed 5,000~\kms.  At the

78: focal position there is an extremely compact radio source (\SgrA),

79: whose apparent size is near the Schwarzschild radius ($2GM/c^2$).

80: This radio source is motionless at the $\sim1$~\kms\ level at the 

81: dynamical center of the Galaxy.  The mass density required by these 

82: observations is now approaching the ultimate limit of a supermassive 

83: black hole within the last stable orbit for matter near the event horizon.  

84:  

85: \end{abstract}

86: 

87: \keywords{supermassive black hole; \SgrA; Galactic center.}

88: 

89: \vskip 0.4truecm

90: \centerline {\bf Table of Contents}

91: \vskip 0.2truecm

92: 1. Introduction  \hfill{2.}

93: \vskip 0.0truecm

94: 2. Evidence for Supermassive Black Holes (SMBHs) \hfill{2.}

95: \vskip 0.0truecm

96: 3. Our Galactic Center \hfill{5.}  

97: \vskip 0.0truecm

98: ~~~~~3.1  The Discovery of \SgrA\  \hfill{5.}

99: \vskip 0.0truecm

100: ~~~~~3.2  The Growing Case for a Dark Mass at the Galactic Center   \hfill{7.}

101: \vskip 0.0truecm  

102: 4. Recent Overwhelming Evidence that \SgrA\ is a SMBH \hfill{8.}

103: \vskip 0.0truecm

104: ~~~~~4.1  Stars Orbiting an Immense Unseen Mass Concentration  \hfill{9.}

105: \vskip 0.0truecm

106: ~~~~~4.2  The Unseen Mass is Centered on \SgrA\  \hfill{10.}

107: \vskip 0.0truecm

108: ~~~~~4.3  \SgrA's Emission is Extremely Compact  \hfill{13.}

109: \vskip 0.0truecm

110: ~~~~~4.4  \SgrA\ is Motionless at Dynamical Center of Milky Way   \hfill{14.}

111: \vskip 0.0truecm

112: 5. Excluding Alternatives to a SMBH \hfill{17.}

113: \vskip 0.0truecm

114: 6. Other Evidence for a SMBH \hfill{18.}

115: \vskip 0.0truecm

116: 7. Broader Implications of SMBHs at Centers of Galaxies  \hfill{19.} 

117: \vskip 0.0truecm

118: 8. Future Observations \hfill{19.} 

119: \vskip 0.0truecm

120: 9. Summary  \hfill{20.} 

121: \newpage

122: 

123: 

124: \section {Introduction}

125: 

126: In the late 18$^{th}$ century, the English naturalist John Michell and

127: the French mathematician Pierre Simon Laplace considered what would 

128: happen if a huge mass were placed in an incredibly small volume.

129: They conjectured that gravitational forces might

130: not allow anything, even light, to escape.  Two centuries later,

131: Albert Einstein's theory of General Relativity provided the theoretical 

132: foundation for such conjectures, and in the 1960s John Archibald 

133: Wheeler introduced the term ``black hole'' to describe the effects of 

134: mass at such an extreme density.

135: 

136:         The concept of a black hole formed by the explosive

137: collapse of a dying star is astounding.  The possibility that matter 

138: from millions and even billions of stars can condense into a single 

139: supermassive black hole (SMBH) is even more fantastic.  

140: Yet we are now confident that supermassive black holes exist and in 

141: fact are commonplace, occupying the centers of many, if not all,

142: of the $\sim10^{11}$ galaxies in the Universe.    

143: Indeed, SMBHs may hold more than 0.01\% of the baryonic mass of the 

144: Universe.

145: 

146: 

147: \section   {Evidence for Supermassive Black Holes} \label{section:evidence}

148: 

149: Early evidence for SMBHs closely paralleled the development

150: of radio astronomy.  Very strong sources of radio waves were 

151: discovered in the early years of radio astronomy.

152: Accurate positions of these sources revealed that they were often

153: centered on distant galaxies.  In the 1950s, radio interferometers

154: revealed a totally unexpected picture of these ``radio galaxies.''  

155: The radio waves did not come from the galaxy itself, 

156: but from two giant ``lobes'' symmetrically placed about, 

157: but well separated from, the parent galaxy (see Fig.~\ref{fig:cyga}).  

158: These lobes can be among the largest structures in the Universe, 

159: hundreds of times the size of the parent galaxy.

160: 

161: \begin{figure*}

162: \begin{center}

163: \begin{tabular}{cc}

164: \includegraphics[width=0.97\textwidth]{f1.eps} \\

165: \end{tabular}

166: \end{center}

167: \caption{Image of the radio galaxy Cygnus A made with the

168: Very Large Array.  The two large lobes of synchrotron emitting

169: plasma dominate.  In some radio galaxies these lobes can span

170: $\sim10^6$~parsecs (1 pc $\approx3\times10^{16}$~m $\approx 3.3$ 

171: light years), and the minimum energy to power the lobes can be up to 

172: $10^7~{\rm M}_\odot c^2$~!  Note the thin ``jets'' that connect 

173: the small bright nucleus at the center with the lobes and terminate with

174: ``hot spots'' at the outer edges of the lobes.  An optical image of the 

175: central galaxy would be comparable in size to a hot spot.

176: All of the energy to power the lobes comes from the nucleus, in fact 

177: from a region smaller than 1~parsec, the distance from the Sun to the 

178: nearest star.  Image courtesy of NRAO/AUI. 

179:         }

180: \label{fig:cyga}

181: \end{figure*}

182: 

183: 	How are immense radio lobes energized?  Their symmetrical

184: placement about galaxies indicated a galactic link.  In the 1960s,

185: sensitive radio interferometers confirmed the previously 

186: circumstantial case by detecting faint trails (called ``jets'') of 

187: radio emission from the lobes back toward the parent galaxy.  

188: Radio jets often lead back to a point-like source at the precise 

189: center of the galaxy.  These point-like sources were found to

190: be variable on timescales less than one year, implying sizes

191: less than a light-year\cite{Dent:65,Maltby:65}.  

192: Martin Rees, in a remarkably prescient paper, showed that the

193: source energetics and variability required synchrotron emitting 

194: plasma that has been accelerated to bulk relativistic speeds\cite{Rees:66}.  

195: One {\it predicted} consequence was that such

196: sources would display apparent motions on the sky that were

197: faster than the speed of light (so called super-luminal motion discussed

198: below).

199: 

200:       In the late 1960s, Very Long Baseline Interferometry (VLBI) 

201: extended baselines to the size of the Earth, achieving

202: resolution ($\approx\lambda/D$, where $\lambda$ is the observing 

203: wavelength and $D$ is the maximum separation of interferometer elements) 

204: better than 0.001~arcseconds\cite{Cohen:68}.  

205: This was achieved by replacing

206: cables between antennas of the interferometer with tape recorders

207: synchronized with atomic oscillators.

208: Radio images made from VLBI observations revealed that the 

209: sources at the centers of radio galaxies are incredibly small,

210: even smaller than the distance between the Sun and the nearest 

211: star\cite{Kellermann:68}.

212: 

213: Simple calculations of the {\it minimum} energy needed to power 

214: giant radio lobes require the {\it total} 

215: conversion of up to $10^7$ stars into energy\cite{Kellermann:88}!  

216: Since nuclear reactions convert less than 1\% of mass 

217: to energy, trying to explain a radio galaxy with nuclear power would 

218: require channeling more than $10^9$ stars through a region smaller 

219: than the distance between the Sun and the nearest star.  

220: Because of these requirements, astronomers began considering a more 

221: efficient energy source: a supermassive black hole.

222: 

223: The small measured sizes of radio sources at the centers of galaxies

224: were consistent with those inferred 

225: from the rapid variability.  Additionally, VLBI data taken months

226: to years apart indicated that ``blobs'' of synchrotron emitting

227: plasma appeared to move across the sky at speeds exceeding that

228: of light\cite{Whitney:71,Cohen:71}.  

229: This phenomenon, called super-luminal motion, has now 

230: been well documented in numerous sources\cite{Zensus:97}.  

231: Super-luminal motion is simply explained as an ``optical illusion'' that 

232: occurs when light emitting plasma moves toward us, nearly along our

233: line of sight, at speeds near, but below, that of light. 

234: This phenomenon finds a natural explanation in the acceleration of 

235: material from regions very close to the event horizon of a black 

236: hole\cite{Blandford:77}. 

237: 

238: In 1990s, images made with the Hubble Space Telescope\cite{Ford:94,Harms:94}

239: showed that material located within 18 parsecs\footnote{1 parsec (pc) = 

240: 3.26 light-years, approximately the distance from the Sun to the nearest 

241: stars.} of the center of the galaxy M~87 was moving with speeds of about 

242: 750 \kms.  Assuming the motions are from material 

243: in orbit about a large mass concentration, the Virial theorem implies a 

244: central mass of $\approx2.4\times10^9$~\Msun\ (1 solar mass or \Msun

245: $=2\times10^{30}$ kg) and an average mass density

246: of $10^5$~\Msunpercpc.  While this density is very high, 

247: it is not above stellar densities in dense clusters and 

248: does not require a supermassive black hole.  

249: 

250: \begin{figure*}

251: \begin{center}

252: \begin{tabular}{cc}

253:  \includegraphics[width=0.90\textwidth]{f2.eps}

254: \\

255: \end{tabular}

256: \end{center}

257: \caption{Water vapor masers in a thin rotating disk at the center

258: of the galaxy NGC~4258 from Herrnstein \etal\ (1999). 

259: {\it Top:} Schematic drawing of the disk superposed with a radio frequency 

260: image ({\it contours}) of synchrotron emitting plasma jets emanating from 

261: the center.  Water masers are shown as black triangles in the disk.  

262: {\it Bottom:} An observed spectrum (flux density versus Doppler velocity).

263: {\it Inset:} Plot of maser Doppler velocity versus distance from the center.

264: A Keplerian ($v \propto 1/\sqrt{r}$) rotation curve is shown with

265: the curved solid lines.  The straight line shows the changing line of sight 

266: projection of masers from the center of the spectrum that lie on a 

267: fixed radius in a small arc in front of the disk center.

268: (Reprinted by permission of Macmillan Publishers Ltd: Nature, {\bf 400}, 

269: 1999, 539)

270:         }

271: \label{fig:ngc4258}

272: \end{figure*}

273: 

274: In the 1980s, maser emission was discovered toward the centers of nearby 

275: galaxies\cite{Claussen:84}.  This emission comes from clouds of gas

276: containing trace amounts of water molecules whose level populations for 

277: the $6_{16}-5_{23}$ transition at 22 GHz become inverted.  In the mid-1990s, 

278: water masers in the galaxy NGC~4258 were discovered to have internal motions 

279: exceeding 1000 \kms~\cite{Nakai:93}.  Subsequent VLBI images of the 

280: masers showed that they came from a rotating disk of 

281: material\cite{Miyoshi:95,Greenhill:95,Herrnstein:99} 

282: (see Fig.~\ref{fig:ngc4258}).  

283: The variation of velocity ($v$) with

284: radius ($r$) followed $v\propto1/\sqrt{r}$ to better than 1\% accuracy,

285: indicating gravitational orbits about a compact central mass. 

286: The rotation speed was about 1000 \kms\ at a radius

287: of about 0.13 parsecs, requiring a central gravitational mass of 

288: $4\times10^7$~\Msun.   The corresponding enclosed mass density is  

289: $4\times10^9$~\Msunpercpc.  Were one to place $4\times10^7$ stars inside a 

290: radius of 0.13 parsecs, the system would be dynamically unstable with less massive

291: stars being expelled (``evaporated'') and more massive stars sinking to the 

292: center, colliding and possibly forming a black hole.  The timescale for the 

293: cluster to evaporate would be fairly short, $\sim10^9$~years,

294: making it unlikely that a cluster of stars could provide the central 

295: gravitational mass\cite{Maoz:98}.  However, the densities were not extreme 

296: enough to conclusively rule out clusters of some types of objects or more 

297: exotic speculations involving dense condensations of elementary particles.  

298: 

299: \begin{figure*}

300: \begin{center}

301: \begin{tabular}{cc}

302: \includegraphics[width=0.95\textwidth]{f3.eps} \\

303: \end{tabular}

304: \end{center}

305: \caption{Fluorescence Fe K$\alpha$ line toward the galaxy 

306: MCG-6-30-15 from Tanaka \etal\ (1995).  The line is extremely

307: broad ($\sim10^5$ \kms), asymmetric, and mostly redshifted

308: from the rest energy of about 6.4 keV.  This line probably 

309: arises form material in an accretion disk inside of $10 \Rsch$

310: of a SMBH where Doppler and gravitational redshifts are large.

311: (Reprinted by permission of Macmillan Publishers Ltd: Nature, 

312: {\bf 375}, 1995, 659)

313:         }

314: \label{fig:mcg-6-30-15}

315: \end{figure*}

316: 

317: In 1994, the ASCA satellite's x-ray telescope recorded a spectrum of

318: the nucleus of the galaxy MCG-6-30-15, showing a broad line at 6.4 keV

319: from iron K$\alpha$ emission\cite{Tanaka:95}.  

320: As shown in Fig.~\ref{fig:mcg-6-30-15}, the full width at zero intensity 

321: was $\sim10^5$~\kms, or about 30\% of the speed of light!  

322: The line is asymmetric with most of the 

323: flux appearing redshifted with respect to the rest motion of the galaxy.

324: The simplest interpretation of this line is that it comes from

325: fluorescence of iron atoms in an accretion disk irradiated by a halo

326: of hot gas and all within $\sim10\Rsch$ of a massive black hole,

327: where $\Rsch=2GM/c^2$ is the Schwarzschild radius. 

328: Thus, by the mid-1990s extremely strong, but perhaps not overwhelming, 

329: evidence for supermassive black holes existed.

330: 

331: \section {Our Galactic Center}

332: 

333:      Surveys of the sky at radio wavelengths in the 1950s revealed

334: a strong radio source in the constellation Sagittarius toward

335: the center of the Milky Way.  This radio source was

336: named Sagittarius~A (Sgr~A), where the letter ``A'' denoted the strongest

337: source in the constellation.  Early radio telescopes did not have

338: the angular resolution to resolve this source, and its nature

339: remained a mystery for some time.  With the advent of radio

340: interferometry, Sgr~A was revealed to contain multiple 

341: components\cite{Morris:96}.

342: Fig.~\ref{fig:sgra} shows recent images of this complicated region made 

343: the Very Large Array radio interferometer.  The left-hand image reveals 

344: at least two strong and extended sources: \SgrAeast, a non-thermal 

345: (synchrotron emitting) supernova remnant and \SgrAwest, 

346: a ``spiral-shaped'' source of thermal bremsstrahlung emission

347: from ionized gas. 

348: 

349: \begin{figure*}

350: \begin{center}

351: \begin{tabular}{cc}

352:  \includegraphics[width=0.47\textwidth]{f4a.eps} &

353:  \includegraphics[width=0.47\textwidth]{f4b.eps}

354: \\

355: \end{tabular}

356: \end{center}

357: \caption{Radio wavelength negative images of the center  

358: of the Milky Way with East toward the left and North toward the top.

359: {\it Left Panel:} The multi-component source Sgr A.  The diffuse

360: elliptical shell-like source \SgrAeast, a supernova remnant from an

361: exploded star, fills most of the left side of this panel, which measures

362: $\approx7$ parsecs across.  The bright spiral-shaped emission 

363: toward the right-center of the panel is called \SgrAwest\ and comes from 

364: plasma spiraling inward to the center.

365: {\it Right Panel:} Expanded view of the central 0.25 parsecs, showing the

366: central region of the \SgrAwest\ spiral.  Also seen at the center of 

367: the panel is a point-like source called \SgrA, a candidate for a 

368: supermassive black hole.  (Images courtesy of J.-H. Zhao.)

369:         }

370: \label{fig:sgra}

371: \end{figure*}

372: 

373: \subsection {The Discovery of \SgrA} \label{section:sgra_discovery}

374: 

375: In 1974, a very compact radio source, smaller than 1 arcsecond in 

376: diameter, was discovered toward Sagittarius~A

377: and later given the name Sagittarius~A* (\SgrA)\cite{Balick:74}. 

378: Early VLBI observations established that \SgrA\  was extremely compact -- 

379: less than the size of our solar system when observed at centimeter 

380: wavelengths\cite{Lo:85}.  However, no obvious optical, infrared, or 

381: X-ray counterpart to \SgrA\  could be easily identified, and its nature 

382: remained a mystery.   

383: 

384: Interestingly, even before the discovery of a compact radio source 

385: at the center of the Milky Way, 

386: it had been speculated that the center of our Galaxy contained a 

387: supermassive black hole\cite{Lynden-Bell:71}.  This speculation

388: was based primarily on two arguments: 1) the presumed similarity 

389: between the nuclei of radio (and other active) galaxies and that of our

390: Galactic center and 2) the possibility of explaining most of the 

391: luminosity arising from the Galactic center as owing ultimately

392: to accretion of material onto a black hole.  While argument 1) has

393: been borne out, argument 2) has not, since most of the luminosity

394: of the Galactic center can be traced to stellar origins.  

395: Unlike the centers of active galaxies believed

396: powered by supermassive black holes, \SgrA\ is extremely 

397: under-luminous\cite{Ozernoy:96}.

398: 

399: \subsection {The Growing Case for a Dark Mass at the Galactic Center} 

400: 

401:      In the late 1970s, infrared observations of gas motions provided the 

402: first clues for a very large mass concentration at the center of the 

403: Milky Way.  Analysis of the center velocities of fine-structure lines of

404: singly-ionized neon atoms (at a wavelength of 12.8~\um) across \SgrAwest\  

405: showed differences of $\pm260$~\kms.  

406: Assuming that the gas clouds were in circular orbits indicated ``a central 

407: point-like mass of several $\times 10^6$~\Msun\ in addition to several 

408: $\times 10^6$~\Msun\ of stars within 1~parsec of the center''\cite{Lacy:80}.

409: This proved to be a remarkably accurate inference.

410: While these observations provided the first solid observational evidence for

411: a supermassive ``object'' at the center of the Milky Way, other possibilities 

412: were recognized.  Firstly, while a point-mass component fit the data best, 

413: a distributed mass of $\sim 10^7$~stars could not be ruled out.  Secondly,

414: gas is susceptible to non-gravitational forces, and the assumption of

415: gravitationally bound orbits was questioned.

416: Finally, the ``point-mass'' component needed only to be smaller than

417: $\approx0.2$~parsecs, and as discussed later in \S\ref{section:alternatives}, 

418: this does not require a SMBH.

419:   

420: While gas is susceptible to non-gravitational forces, stars are not, and soon

421: measurements of radial velocities (Doppler shifts) of stars confirmed the 

422: presence of several 

423: million solar masses located within $\sim0.1$~parsecs of the center of the Milky 

424: Way\cite{McGinn:89,Sellgren:90,Krabbe:95,Haller:96,Genzel:97}.  

425: There is indeed a dense cluster of stars at the center of the Milky Way,

426: which cannot be seen at optical wavelengths because visible light is totally 

427: absorbed by dust between the Galactic center and the Sun.

428: However, these stars can be seen at infrared wavelengths, where $\sim10$\%

429: of the 2~\um\ wavelength light is received.  

430: Using novel techniques that allow diffraction-limited imaging in the 

431: infrared, groups led by Reinhard Genzel in Germany and Andrea Ghez in the 

432: USA have been measuring positions of these stars for more than a decade

433: \cite{Eckart:96,Eckart:97,Ghez:98,Genzel:00}.  These results

434: showed that stars projected very close to the position of \SgrA\  

435: were moving very rapidly across the sky.  

436: 

437: Fig.~\ref{fig:enclosed_mass} plots the stellar velocity dispersions available 

438: in 1998 and the inferred enclosed mass as a function of projected distance 

439: from the position of \SgrA.  Measurements of stellar radial velocity sample 

440: projected distances from 4 to 0.1 parsecs and proper motions

441: (motions on the plane of the sky) sample from 0.3 to 0.01 parsecs.

442: Between projected radii of 0.2 and 0.01 parsecs, the velocity dispersion,

443: $\sigma$, increases as $\sigma \propto 1/\sqrt{r}$.  

444: The enclosed mass estimated for virialized material is indeed nearly 

445: constant between these radii, greatly strengthening the case for a large 

446: ``point mass.''  The implied mass density, while very high (comparable to 

447: that inferred from the water masers in the galaxy NGC~4258 discussed in

448: \S\ref{section:evidence}), still could not rule some alternatives to a SMBH 

449: (see \S\ref{section:alternatives}).

450: 

451: \begin{figure*}

452: \begin{center}

453: \begin{tabular}{cc}

454: \includegraphics[width=0.49\textwidth]{f5a.eps} & 

455: \includegraphics[width=0.47\textwidth]{f5b.eps} 

456: \\

457: \end{tabular}

458: \end{center}

459: \caption{Measured stellar velocity dispersions {\it (left panel)} and 

460: inferred enclosed mass {\it (right panel)} versus projected distance 

461: from the Galactic center (after Eckart \& Genzel 1997 and Ghez et al 1998).  

462: The line through the data corresponds to Keplerian orbits for a point-like 

463: mass of $2.5\times10^6$~\Msun\ and fits the data well for radii smaller than

464: 0.2~pc.  Beyond this radius, the velocity dispersions exceed that

465: expected from a point mass, owing to the contribution to the enclosed mass

466: from an observed dense cluster of stars.

467:         }

468: \label{fig:enclosed_mass}

469: \end{figure*}

470: 

471: \section  {Recent Overwhelming Evidence that \SgrA\  is a SMBH}

472: 

473: Evidence for the existence of SMBHs, especially for \SgrA, has been 

474: steadily growing over the years, but recently observational constraints 

475: have become so strong that there can be almost no doubt that \SgrA\ is a 

476: supermassive black hole.  

477: 

478: \subsection  {Stars Orbiting an Immense Unseen Mass Concentration}

479: 

480: \begin{figure*}

481: \begin{center}

482: \begin{tabular}{cc}

483: \includegraphics[width=0.95\textwidth]{f6.eps} 

484: \end{tabular}

485: \end{center}

486: \caption{False color infrared image taken with by the European Southern

487: Observatory's Very Large Telescope of the central few parsecs of the Milky Way.  

488: Superposed with a 100-times finer scale

489: is the orbital track of one star named S2.  The orbital period of 

490: S2 is 15.8 years, and recently a complete and closed elliptical orbit 

491: has been observed.  The orbit requires an unseen mass of 

492: $\approx4\times 10^6$~\Msun\ at the focal position, 

493: indicated by the arrow.  The focal position is coincident with the position 

494: of the compact radio source \SgrA\ as discussed in 

495: \S\ref{section:sgra_position}.

496: (Image courtesy R. Genzel.)

497: }

498: \label{fig:S2_orbit}

499: \end{figure*}

500: 

501: Continued monitoring of the positions of stars, with increasing

502: positional accuracy, led to clear detections of acceleration

503: (ie, curving motions on the sky)\cite{Ghez:00,Eckart:02}.  Importantly, the

504: directions of the acceleration vectors ``pointed'' to a common central

505: gravitational source very close to the position of \SgrA.  

506: Recently these observations culminated in the discovery that

507: stars are executing elliptical paths (orbits) on the sky

508: \cite{Ghez:00,Schoedel:02,Ghez:03,Schoedel:03,Ghez:05}.  

509: One star named S2 (a.k.a. S0-2) and has now been observed over one 

510: complete 15.8-year elliptical orbit (see Fig.~\ref{fig:S2_orbit}).

511: All stellar orbits are well fit by a single enclosed mass and  

512: focal position (see Fig.~\ref{fig:stellar_orbits}).  

513: Two stars have been observed to approach within 100~AU

514: \footnote{1 Astronomical Unit (AU), the mean distance from the Earth 

515: to the Sun, equals $1.5\times10^{11}$~m or $5\times10^{-6}$ parsecs.}

516: of the focal position, moving at nearly $10^4$ \kms!  

517: {\it The orbital solutions leave no doubt that the stars are responding 

518: to an unseen compact mass of $\approx4\times10^6$~\Msun.}  

519: 

520:    We note that the central mass estimated from stellar velocity 

521: dispersions ($\approx2.5\times10^6$~\Msun) has been lower than the 

522: value obtained from fitting Keplerian orbits ($\approx4\times10^6$~\Msun).

523: Transforming one-dimensional velocity dispersions to enclosed masses can 

524: be accomplished with a standard Virial analysis or using alternative 

525: statistical mass estimators: \eg\ Bahcall-Tremaine\cite{Bahcall:81}.  

526: These methods are usually based on an assumption of isotropic motions.  

527: While this is well justified for old stars orbiting the Galactic center, 

528: it does not work well for young stars that still ``remember'' the orbital 

529: plane of the gas cloud from which they recently formed.  At least one such 

530: collection of young stars has been discovered\cite{Genzel:00,Lu:08}.

531: An alternative statistical mass estimator, the 

532: Leonard--Merritt\cite{Leonard:89} approach, does not depend on

533: isotropic distributions of stellar velocities.  However, as for all 

534: projected mass estimators, it assumes measurements over all radii.  

535: For the central stellar cluster, which is distributed as a power-law 

536: over the range of observed radii\cite{Genzel:03b,Schoedel:07}, 

537: normalization problems arise.  

538: Empirical corrections to the Leonard--Merritt estimator, for the limited 

539: range of projected orbital radii of stars that have been measured, 

540: increase the central mass estimate and resolve the discrepancy with 

541: orbital fitting\cite{Genzel:00}.

542: 

543: \begin{figure*}

544: \begin{center}

545: \begin{tabular}{cc}

546: \includegraphics[width=0.95\textwidth]{f7.eps} \\

547: \end{tabular}

548: \end{center}

549: \caption{

550: Stars within the 0.02 parsecs of the Galactic center orbiting an unseen mass.

551: Yearly positions of seven stars are indicated with filled colored circles.

552: Both curved paths and accelerations (note the non-uniform spacings between 

553: yearly points) are evident. Partial and complete elliptical orbital fits for 

554: these stars are indicated with lines.   All orbital fits require 

555: the same central mass of $\approx4\times10^6$~\Msun\ and a common 

556: focus at the center of the image, the position of the radio source \SgrA.  

557: (Image courtesy A. Ghez.)

558:         }

559: \label{fig:stellar_orbits}

560: \end{figure*}

561: 

562: The possibility of a combination of a point mass and an extended ($>100$ AU) 

563: distribution of mass has been considered\cite{Mouawad:05,Ghez:08,Gillessen:08}.

564: Based on the small deviations from an elliptical orbit for star S2 allowed

565: by measurement uncertainty, any extended component within 0.01 parsecs of

566: the center must be $<10$\% of the point mass.  Limits on deviations of orbits

567: from those responding to a pure point-mass will undoubtedly improve rapidly

568: as the star S2 proceeds on a second cycle and other stellar orbits are

569: better traced.  Because the orbital paths are almost perfect ellipses,

570: most of the unseen mass must be contained within a radius of about 

571: 100~AU (0.0005 pc).  This implies a mass density of 

572: $>8\times10^{15}$~\Msunpercpc, which is so great that one can rule out 

573: the exotic speculation that a ``ball'' of Fermions with rest energies 

574: of $\sim15$~keV might provide the extreme central mass for \SgrA\ and other, 

575: more massive, galactic nuclei (see \S\ref{section:alternatives}) .  

576: sss

577: \subsection {The Unseen Mass is Centered on \SgrA} \label{section:sgra_position}

578: 

579:      The infrared results just described are beautifully complemented

580: by observations at radio wavelengths.  It is crucial to know the position

581: of \SgrA\ on infrared images.  However, while \SgrA\ is a strong radio

582: source, it is extremely dim at infrared wavelengths.  In general, the infrared

583: coordinate system is not known to better than about 0.1 arcseconds relative

584: to the International Celestial Reference Frame determined by VLBI observations

585: at radio wavelengths.  Unfortunately, there are many stars in the 

586: Galaxy's central 0.1 arcseconds, so this level of position accuracy

587: is inadequate to locate and identify \SgrA\ on crowded infrared images.

588: 

589: How can one transfer \SgrA's radio position to infrared images with better 

590: than the 0.01 arcseconds accuracy needed to clearly identify candidates?   

591: The key is to find sources visible at both radio and infrared wavelengths.

592: This has been accomplished with red giant stars that are bright

593: in the infrared and have molecular maser emission at radio wavelengths

594: from their circumstellar material.  Fig.~\ref{fig:sgra_sio} shows an infrared

595: image of the central $\pm20$~arcseconds of the Galaxy and the positions

596: and measured motions of SiO maser emission at radio 

597: wavelengths for nine stars.  

598: 

599: \begin{figure*}

600: \begin{center}

601: \begin{tabular}{cc}

602: \includegraphics[width=0.90\textwidth]{f8.eps}  \\

603: \end{tabular}

604: \end{center}

605: \caption{Infrared negative image at  2.2~\um\ wavelength of the 

606: central $\pm20$ arcseconds of the Milky Way from Reid et al. (2007).

607: The positions ({\it circles}) and measured motions ({\it arrows}) of red

608: giant stars that have detected radio emission from circumstellar 

609: SiO masers are indicated.  These stars were used to accurately transfer 

610: the position of \SgrA, measured in the radio, to the infrared.

611: (Reproduced by permission of AAS.)

612:         }

613: \label{fig:sgra_sio}

614: \end{figure*}

615: 

616: This novel combination of infrared and radio observations has allowed the 

617: position of the compact radio source at the center of the Galaxy, \SgrA, 

618: to be transfered accurately ($\pm0.01$~arcseconds or $\pm80$~AU) 

619: to infrared images\cite{Menten:97,Reid:03,Reid:07}, where \SgrA\ is 

620: extremely dim.  The position of \SgrA, determined in this manner, 

621: is circled at the center Fig.~\ref{fig:sgra_sio}.

622: Fig.~\ref{fig:sgra_ir} shows one 

623: infrared frame covering the inner-most $\pm1$~arcsecond 

624: ($\pm0.04$~parsecs) of the Galactic center.  Many stars, including some 

625: that are known to orbit \SgrA, are visible.  However, at the position of 

626: \SgrA\ no steady emission was detected, although on this frame the 

627: diffraction-limited image of star S3 overlaps slightly with \SgrA.

628: Recently, using improved infrared observing techniques, some weak

629: varying emission has been seen toward the position of \SgrA.

630: (see \S\ref{section:other}).  

631: {\it However, the infrared emission of \SgrA\ is much weaker than individual stars!}

632: 

633: \begin{figure*}

634: \begin{center}

635: \begin{tabular}{cc}

636: \includegraphics[width=0.90\textwidth]{f9.eps}  \\

637: \end{tabular}

638: \end{center}

639: \caption{An infrared frame from July 1995 covering the 

640: inner $\pm1$ arcsecond from Reid \etal\ (2003).  

641: Three of the stars that orbit the Galactic center, S1, S2, and S3, are labeled,

642: as is the position of \SgrA.  The image of star S3 overlaps slightly 

643: with \SgrA, whose emission is much weaker than a single star.

644: (Reproduced by permission of AAS.)

645:         }

646: \label{fig:sgra_ir}

647: \end{figure*}

648: 

649: 

650: The radio position of \SgrA\ transfered to infrared images matches 

651: the focal position determined from the stellar

652: orbital fits to within the 0.01~arcseconds measurement accuracy.  

653: Could \SgrA\  be projected toward the Galactic

654: center, but in reality not be located there?  

655: The fact that \SgrA, a nearly unique source in the 

656: Milky Way, lies in projection within 0.01~arcseconds

657: (ie, within a solid angle of $\sim10^{-14}$ steradians) of the gravitational 

658: focus of the stellar orbits makes it extraordinarily unlikely that

659: it is simply a projection effect.  In addition, \SgrA's apparent motion  

660: with respect to distant quasars is consistent with it being at the 

661: dynamical center of the Milky Way (see \S\ref{section:motionless}), 

662: ensuring that it is indeed at that location.

663: 

664: \subsection  {\SgrA's Emission is Extremely Compact}

665: \label{section:sgra_size}

666: 

667: The cm-wavelength emission from \SgrA\  is strongly affected by

668: scattering from interstellar electrons, which increases its apparent size.

669: At mm-wavelengths the scattering is reduced, and the true size of

670: the source can be measured.  Many groups have analyzed VLBI data for 

671: \SgrA\  and find that the intrinsic size is less than 1 AU at mm-wavelengths

672: \cite{Rogers:94,Krichbaum:98,Doeleman:01,Bower:04,Shen:05}.

673: However, the absence of strong refractive scintillations for \SgrA\ 

674: provides a lower limit of $\sim0.1$~AU for the intrinsic size

675: of the cm-wave emission\cite{Gwinn:91}. 

676: Thus, the intrinsic radius of the radio emission from \SgrA\  

677: near 1~cm wavelength is between 1 and 6 Schwarzschild radii 

678: for a $4\times10^6$~\Msun\ black hole.

679: 

680: \subsection  {\SgrA\ is Motionless at Dynamical Center of Milky Way}

681: \label{section:motionless}

682: 

683: How much of the unseen mass in the central 100 AU region can be tied

684: directly to \SgrA?   Were \SgrA\  just a stellar mass object in the 

685: central stellar cluster, it would be moving at $\sim10^4$ \kms\ 

686: in the strong gravitational potential as stars are observed to do.  

687: Only if \SgrA\ is extremely massive could it move slowly.

688: Early measurements of \SgrA's motion revealed it to be moving $<20$ \kms\ 

689: \cite{Backer:99,Reid:99}.

690:   

691: Fig.~\ref{fig:sgra_motion} shows the latest data on the apparent motion

692: of \SgrA\ on the sky relative to a background quasar.  This includes

693: the data published by Reid \& Brunthaler in 2004\cite{Reid:04} and new data 

694: taken in 2007 that confirms the published results.  The apparent movement 

695: is as expected for an object that is stationary at the dynamical center of 

696: the Milky Way and viewed from the Sun-Earth system, which orbits the 

697: Galaxy with a period of $\approx210$ My at a radius of about 8~kpc and 

698: a speed of 240~\kms.  It is amazing that VLBA can detect this motion

699: in only a few weeks time.

700: 

701: The Sun's orbit is almost entirely in the plane of the Galaxy, but  

702: the orbital speed in the plane is not easily measured from within the 

703: Galaxy and is uncertain by $\approx20$ \kms.  However, the small component 

704: of the Sun's motion {\it perpendicular} to the plane of the 

705: Galaxy is very accurately known ($7.16\pm0.38$~\kms) from observations of 

706: $10^4$ stars in the solar neighborhood by the astrometric satellite 

707: HIPPARCOS\cite{Dehnen:98}.  Thus, the contribution of the Sun's motion

708: perpendicular to the plane of the Galaxy can be removed from the apparent 

709: motion of \SgrA\  with very high accuracy.  When this is done,

710: \SgrA's intrinsic motion perpendicular to the plane of the Galaxy is 

711: $-0.4\pm0.9$ \kms~\cite{Reid:04}.  The extremely small intrinsic motion 

712: (essentially an upper limit) for \SgrA\  is close to the expected motion 

713: for a SMBH in the presence of its dense cluster of surrounding stars.

714: 

715: \begin{figure*}

716: \begin{center}

717: \begin{tabular}{cc}

718: \includegraphics[width=0.9\textwidth]{f10.eps}

719: \\

720: \end{tabular}

721: \end{center}

722: \caption{The {\it apparent} motion on the sky of the compact radio source \SgrA\

723: relative to a distant quasar (J1745--283).  The dashed line is the

724: variance-weighted best-fit motion of $0.006379\pm0.000024$

725: arcseconds per year for the data published by Reid \& Brunthaler through 

726: 2004.  Recent data from 2007 shown here confirms the published result.

727: All of the {\it apparent} motion of \SgrA\ can be accounted for by the 

728: $\approx210$ My period orbit of the Sun about the Galactic center.

729: The solid line gives the orientation of the Galactic plane, and the 

730: difference in orientation of the two lines is caused by the 7.16 \kms\ 

731: motion of the Sun perpendicular to the Galactic plane.  The residual,

732: intrinsic motion of \SgrA\ perpendicular to the Galactic plane is

733: extremely small: $-0.4\pm0.9$ \kms.  A SMBH perturbed by stars orbiting 

734: within its gravitation sphere of influence is expected to move $\approx0.2$

735: \kms\ in each coordinate.  Only a supermassive object could

736: be motionless in the presence of the $4\times10^6$ \Msun\ known from

737: infrared stellar orbital fits to occupy this region.

738:         }

739: \label{fig:sgra_motion}

740: \end{figure*}

741: 

742: The discovery that \SgrA\  is nearly stationary at the Galactic center

743: requires that \SgrA\  must contain a significant fraction of the unseen mass 

744: indicated by the orbiting stars.  The calculation of the expected motion of a 

745: massive object within its bound stellar cluster (which allows the 

746: limit on the motion of \SgrA\  to give a minimum mass) 

747: can be cast as a gravitational Brownian-motion problem

748: \cite{Chatterjee:02,Dorband:03,Merritt:07}.  

749: Such a system comes into ``thermal'' equilibrium,

750: resulting in equipartition of kinetic energy among its constituents.

751: For the Galactic center case in particular, both an analytical result, 

752: with only a few simplifications, and a numerical result for fully 

753: realistic cases allow estimation of the minimum expected motion of a SMBH 

754: surrounded by a dense stellar cluster\cite{Reid:04}.  

755: For a $4\times10^6$~\Msun\ black hole, one expects $\approx0.2$~\kms\

756: for each component of its motion.  This is a very conservative estimate

757: as it does not include possible clustering of the perturbing stars

758: nor any contribution from a cluster of dark stellar remnants, which

759: are expected to have accumulated in this region\cite{Morris:93}.  

760: Including either of these effects would increase the expected motion of 

761: the black hole, bringing it very close to the measured limit.

762: 

763: Both analytical and numerical estimates of the mass of the motionless

764: radio source \SgrA\ yield $>4\times10^5$ \Msun.  

765: Assuming this mass is contained within 

766: the observed source size (see \S\ref{section:sgra_size}), 

767: the mass density is a staggering $>7\times10^{21}$ \Msunpercpc.  

768: This density is within about three 

769: orders of magnitude of the ultimate limit of $4\times10^6$ \Msun\ within 

770: its Schwarzschild radius, $\Rsch$.  For the simplest case of a non-rotating 

771: black hole, stable gravitational orbits exist only outside of $3\Rsch$.  

772: Any material inside $3\Rsch$ cannot orbit and falls rapidly into the hole.  

773: Thus, the effective volume of such a black hole is $3^3$-times that 

774: defined by the Schwarzschild radius, resulting in a corresponding decrease 

775: in the ``ultimate''  mass density.  In this case,

776: the measured density is within only two orders of magnitude of the

777: black hole limit.  

778: {\it At this density the evidence is overwhelming that Sgr~A* is a supermassive black hole}

779: (see \S\ref{section:alternatives}).  

780: 

781: \begin{table}[ph]

782: \tbl{Density Limits for SMBH Candidates.}

783: {

784: \begin{tabular}{@{}lccccc@{}} 

785: \toprule

786: Object & Mass    & Radius & $M/R$   & Density   &Reference \\

787:        & (\Msun) & (AU)   & (kg/m)       & (\Msunpercpc)        \\

788: \colrule

789: Globular Cluster&$\p1\times10^6$ &$2\times10^5$&$\p7\times10^{19}$ &$\p2\times10^5\q$ & \\

790: M~87           &$\p2\times10^9$ &$4\times10^6$ &$\p7\times10^{21}$ &$\p1\times10^{5\q}$ &\cite{Ford:94,Harms:94} \\

791: NGC~4258       &$\p4\times10^7$ &$2\times10^4$ &$\p3\times10^{22}$ &$\p1\times10^{10}$ &\cite{Miyoshi:95} \\

792: \SgrA\ (orbits)&$\p4\times10^6$ &$<100$        &$>5\times10^{23}$ &$>8\times10^{15}$   

793:   &\cite{Schoedel:02,Ghez:03,Schoedel:03,Ghez:05} \\

794: \SgrA\ (motion)&$>4\times10^5$  &$<0.5$        &$>1\times10^{25}$ &$>7\times10^{21}$   &\cite{Reid:04} \\

795: SMBH (3\Rsch)  &$\p4\times10^6$ &$0.24$        &$\p2\times10^{26}$ &$\p7\times10^{23}$ &   \\

796: \botrule

797: \end{tabular} 

798: \label{table:densities}

799: }

800: \end{table}

801: 

802: For comparison, Table~\ref{table:densities} lists observed densities,

803: in standard astronomical units, for dense stellar clusters 

804: (\ie, globular clusters), some SMBH candidates, and a Schwarzschild 

805: SMBH within its inner-most stable orbit.  Also listed are the mass-to-radius 

806: ($M/R$) ratios, in standard physical units.  For a black hole, 

807: $M/3\Rsch = c^2/6G$.

808: 

809: The above density calculation assumes that 

810: the size of the emitting region is equal to or greater than

811: the size of the mass in \SgrA.  This is true for almost all astrophysical

812: sources.  Notable exceptions for radio sources are solar flares

813: and pulsars.  However, these sources 1) are sporadic (either flares

814: or pulses), 2) are characterized by highly-polarized gyrosynchroton 

815: emission and 3) have spectra with flux density falling very rapidly with

816: increasing observing frequency.  \SgrA\ shares none of these 

817: characteristics; its radio emission is generally slowly varying,

818: almost unpolarized, and has a rising spectrum.

819: 

820: \section   {Excluding Alternatives to a SMBH} \label{section:alternatives}

821: 

822: Are there alternatives to a SMBH that are consistent with

823: the extraordinarily high mass density of \SgrA?   

824: The most obvious possibility is a cluster of dark stars.

825: As a point of reference, globular clusters are spherical

826: collections of upwards of $10^6$ stars within a radius of $\sim1$~parsec.

827: These are long-lived systems, some nearly the age of the Universe, 

828: and their stellar densities are truly astounding -- imagine placing 

829: $\sim10^6$ stars between the Sun and the nearest star!  

830: However, while such a density is extremely high,

831: it is nearly a factor of $10^{18}$ times less than 

832: for a SMBH of \SgrA's mass.  (Also, normal, luminous stars are easily 

833: ruled out by the dearth of infrared emission from the position of 

834: \SgrA.)

835: 

836: Dense clusters of stars undergo significant gravitational interactions, 

837: including core-collapse, collisions and evaporation of 

838: stars\cite{Binney:87}.  Both the evaporation or collisional 

839: timescales can provide approximate upper limits for the lifetimes of 

840: such systems\cite{Maoz:98}.  If the cluster members have typical 

841: masses of $\sim1$~\Msun, then the evaporation timescale for a cluster

842: that satisfies the Galactic center stellar orbital data would be $<10^6$~years.  

843: The existence of a cluster with such a short lifetime is extremely 

844: implausible.   

845: 

846: Evaporation timescales are approximately inversely proportional to 

847: the typical mass of a cluster member.  Thus, by postulating very low 

848: member masses, the evaporation timescale can be made arbitrarily long.  

849: In order to have an evaporation timescale be a reasonable fraction of the 

850: age of the Galaxy ($\approx13\times10^9$ years), member masses would need 

851: to be $<0.001$~\Msun.  Thus, one would need $\sim10^{10}$ planets like 

852: Jupiter to make a long-lived $4\times10^6$~\Msun\ cluster.  

853: This would seem impossible to arrange.

854: 

855: The argument for excluding dense stellar clusters based on their rapid 

856: evaporation timescales is based on the implicit assumption of 

857: an isolated system.  Perhaps a quasi-steady state condition could

858: occur for which the cluster is fed stars from the outside at a 

859: rate comparable to the evaporation rate.  A $4\times10^6$~\Msun\ 

860: cluster that evaporates in $10^6$~years would require the addition

861: of only a few stars per year to offset evaporation losses.  

862: This possibility seems unlikely but has yet to be critically analyzed.

863: 

864: Are there other possibilities for extreme concentrations of matter

865: at the Galactic center?  Hypothetical concentrations of exotic dark

866: particles have been considered as alternatives to supermassive

867: black holes in the centers of galaxies.   For example, a ``ball'' 

868: of heavy Fermions supported by degeneracy pressure has been proposed

869: \cite{Viollier:93,Tsiklauri:96,Munyaneza:98,Bilic:99}.  

870: One attractive aspect of this alternative to supermassive black holes 

871: was that Fermions of rest energy $\sim15$~keV could naturally explain 

872: the range of masses ($\sim3\times10^6$ to $\sim3\times10^9$~\Msun) 

873: observed at the centers of galaxies.  However, a Fermion ball cannot achieve 

874: extraordinarily high central densities and its gravitational potential

875: is softer than that of a black hole and flattens at the center.  

876: In 2002, Munyaneza \& Viollier\cite{Munyaneza:02} 

877: showed that the then existing stellar motion data could only

878: be fit by a restricted range of parameter space for a supermassive 

879: black hole, but could be easily fit by Fermion ball models.  

880: Subsequent observations of stellar orbits\cite{Ghez:03,Schoedel:03,Ghez:05} 

881: showed that the supermassive black hole predictions were indeed met,

882: and that the stars moved with speeds greater than allowed by their 

883: Fermion ball model.

884: 

885: We note that the density limit employed in this section was the one obtained 

886: only from the orbits of stars.  The additional constraints

887: from the radio observations provide nearly a factor of $10^6$ more

888: stringent limit on density, making all of the above arguments

889: vastly stronger.   Since the density limits are now within only

890: about two orders of magnitude of that of a SMBH 

891: (see Table~\ref{table:densities}), any alternatives to a SMBH must

892: allow similarly high densities.  At such

893: densities it is difficult to avoid gravitational collapse and

894: proposed  alternatives, such as ``boson stars''\cite{Torres:00} and 

895: ``magnetospheric eternally collapsing objects,''\cite{Robertson:03} or

896: others, possibly involving``new physics,'' would be even more fantastic 

897: than a ``mundane'' SMBH.

898: 

899: Finally, there is now strong evidence for the existence of an

900: event horizon in both stellar-mass black holes\cite{Garcia:01,McClintock:04}

901: and for SMBHs.  Many galactic nuclei, in which high mass-densities

902: indicative of SMBHs are observed, are under-luminous and best

903: explained with radiatively inefficient accretion flows in which

904: energy can vanish through the event horizon\cite{Narayan:08}.  

905: Without an event horizon, energy liberated by the accretion process

906: cannot vanish.  Indeed, \SgrA\ is such an under-luminous object, 

907: and evidence from the dearth of infrared emission and an extremely small 

908: intrinsic size strongly point to an event horizon\cite{Doeleman:08,Broderick:08}.

909: 

910: \section   {Other Evidence for a SMBH} \label{section:other}

911: 

912:     While quiescent emission from \SgrA\ has been difficult to detect

913: outside of cm to sub-millimeter wavelengths, it does produce 

914: detectable ``flares'' of short duration at 

915: radio\cite{Herrnstein:04,Mauerhan:05,Marrone:06} and

916: infrared\cite{Genzel:03,Ghez:04,Hornstein:07,Do:08} wavelengths and

917: x-ray energies\cite{Baganoff:01}.  This flaring is thought to be

918: associated with material occasionally spiraling inward,

919: converting a significant fraction of its total energy into heat, and

920: radiating profusely near the inner-most stable orbit around a

921: black hole.  Of particular interest is the hint of a quasi-periodic 

922: flaring at infrared wavelengths\cite{Genzel:03,Do:08}, which could arise during 

923: the final few orbits of material falling into a black hole.

924: Since, the radius of the inner-most stable orbit depends on black hole

925: spin and whether the material is orbiting in a prograde or retrograde 

926: sense\cite{Misner:73}, this may provide an observational 

927: approach to measure black hole spin\cite{Genzel:03,Shafee:06,McClintock:06}.

928: 

929: 

930:     In 1988, J. G. Hills\cite{Hills:88} considered the fate of 

931: tightly bound binary stars that encounter a supermassive black 

932: hole at the Galactic center.  If the encounter is close enough, 

933: he predicted that one of the stars could become bound to the 

934: black hole, while the other could be ejected at high speeds of

935: up to 4000~\kms.  Detailed calculations of ejection rates have now

936: been made\cite{Yu:03}.  Recently, such hyper-velocity stars have been

937: discovered leaving the Milky Way\cite{Brown:05}, confirming Hills' 

938: prediction.  While this may be a strong confirmation of the

939: existence of a supermassive black hole in the Galactic center, 

940: it is probable that large numbers of stellar-mass black holes have

941: migrated to the inner 0.1 parsec of the center and hyper-velocity

942: stars can also be ejected from interactions with these lower mass black

943: holes\cite{OLeary:08}. 

944: 

945: 

946: \section   {Broader Implications of SMBHs at Centers of Galaxies}

947: 

948: The highly complementary discoveries at infrared wavelengths (of

949: stars orbiting an unseen massive object) and at radio wavelengths

950: (that the stellar orbital focus coincides with \SgrA, that

951: the size of \SgrA\ is comparable to \Rsch and that \SgrA\ is essentially 

952: motionless at the dynamical center of the Galaxy) establish with near certainty

953: that the center of the Milky Way is anchored by a supermassive

954: black hole.  Of course the Milky Way cannot be unique in the Universe

955: in having a supermassive black hole at its center, and, indeed,

956: the Hubble Space Telescope finds evidence for a SMBH at the

957: center of all galaxies nearby enough so that the telescope can

958: resolve the gravitational sphere of influence of the SMBH

959: \cite{Ferrarese:02}.  While

960: the centers of most, if not all, galaxies contain a SMBH, this is not to

961: say that all galaxies are identical.  The mass and spin of the black hole 

962: and the density of stars and gas within its gravitational sphere of 

963: influence lead to a rich variety of phenomena, which are collectively 

964: called active galactic nuclei.  \SgrA, being the nearest SMBH, 

965: serves as the archetypal source for understanding other 

966: galaxies.  

967: 

968: Astronomers now believe that when supermassive black holes accrete 

969: matter at a high rate they become extremely luminous and can outshine 

970: their entire host galaxy of $\sim10^{11}$ stars (ie, the quasar phenomenon).  

971: Quasars have been found at high redshifts\cite{Fan:01}, 

972: indicating they existed as early as

973: 1 billion years after the Big Bang, when the Universe was less than

974: 8\% of its current age.  While it is not yet understood how supermassive

975: black holes form so quickly, they now appear to be an integral part of 

976: the generation of structure in the early Universe\cite{Haiman:01}.  

977: 

978: Recent observations show a strong correlation between the masses of 

979: supermassive black holes and the masses and motions of stellar bulges 

980: which protrude from disk galaxies\cite{Magorrian:98,Ferrarese:00}. 

981: These observations and large computational simulations of how galaxies

982: form indicate that supermassive black holes may shape the evolution 

983: of galaxies\cite{Springel:05,Robertson:06}.  

984: Thus, supermassive black holes are not only 

985: fascinating objects from the perspective of fundamental physics, 

986: they also are exceptionally important from an astrophysical perspective 

987: for determining the nature of galaxies.  

988: 

989: \section    {Future Observations}

990: 

991:    We know that galaxies interact gravitationally and sometimes

992: collide and merge.  The merger of two galaxies, each containing

993: a SMBH, would be followed by orbital decay of the two nuclei,

994: owing first to dynamical friction and ultimately to gravitational 

995: radiation (when a tight black hole binary is reached).  The final

996: merging of two SMBHs produces very strong gravitational waves,

997: which might be detected with the future space-based gravity-wave 

998: detector LISA\cite{LISA:08}.

999: 

1000:    Returning to the nearest SMBH, \SgrA, an obvious goal for astronomers 

1001: is to image the region within the inner-most stable orbit with 

1002: resolution of $\sim\Rsch$.

1003: As this review was being assembled, interferometer fringes were reported

1004: toward \SgrA\ at a wavelength of 1.3~mm and a fringe

1005: spacing of $0.00005$~arcseconds\cite{Doeleman:08}.  This demonstrates that

1006: a significant fraction of the mm-wave emission of \SgrA\ comes from

1007: a region with a radius $<0.2$~AU, which is $<3\Rsch$.  When ``snapshot''

1008: images can be made at slightly shorter wavelengths (eg, 0.8~mm),

1009: they should reveal the ``shadow'' caused by extreme bending of

1010: space around a black hole\cite{Falcke:00} and might show highly dynamic 

1011: activity associated with the final in-spiral of plasma toward the

1012: event horizon\cite{Broderick:06}.  

1013: 

1014:    Astrometric observations of the region within $\sim10\Rsch$ can

1015: also be made with infrared interferometry.  Plans for such an interferometer 

1016: experiment, called GRAVITY\cite{Eisenhauer:08}, are well underway.  

1017: Since the radius, and hence period, of the inner-most stable orbit 

1018: depends on the black hole spin, measuring the in-spiraling material

1019: motion can yield determinations of spin.

1020: 

1021:    Stars orbiting with short periods ($\sim1$ yr) in highly elliptical 

1022: orbits would approach very close to \SgrA\ at pericenter and should display 

1023: general relativistic precession of their orbital planes.  Diffraction-limited 

1024: imaging of the Galactic center in the infrared with future extremely large 

1025: ($>30$~m diameter) telescopes should be able to make such 

1026: observations\cite{Weinberg:05} and also allow a test of the ``no hair''

1027: theorem for a black hole\cite{Will:08}

1028: 

1029: \section {Summary}

1030: 

1031: The major observational results that provide overwhelming evidence

1032: that \SgrA\ is a SMBH are as follows:

1033: 

1034: \begin{itemlist}[(ii)]

1035: \item Stars near \SgrA\ move on elliptical orbits with a common

1036:       focal position.      

1037: \item The required central mass is $4\times10^6$~\Msun\ within a

1038:       radius of 100~AU.

1039: \item The position of \SgrA\ agrees with the orbital focus to within

1040:       measurement uncertainty of $\pm80$~AU.

1041: \item The infrared emission from \SgrA\ is far less luminous

1042:       than a single star.  

1043: \item The intrinsic size of \SgrA\ at mm-wavelengths is $<6\Rsch$.

1044: \item \SgrA\ is intrinsically motionless at the \kms\ level at the

1045:       dynamical center of the Galaxy. 

1046: \end{itemlist}

1047: 

1048: The great impact of these discoveries is their simplicity and elegance.

1049: Elliptical orbits for stars provide an absolutely clear and unequivocal

1050: proof of a great unseen mass concentration.  The discoveries that the

1051: compact radio source is at the position of the unseen mass and is 

1052: motionless provide even more compelling evidence for a supermassive black hole.

1053: {\it Together they form a simple, unique demonstration that the fantastic 

1054: concept of a supermassive black hole is, with a high degree of certainty,

1055: a reality.}

1056: 

1057: \section*{Acknowledgements}

1058: I thank R. Genzel and A. Ghez for comments and suggestions on a draft 

1059: of this review.

1060: 

1061: \begin{thebibliography}{000}  %for 3 digits

1062: 

1063: \bibitem{Dent:65} W. A. Dent, {\it Science} {\bf 148} (1965) 1458.

1064: \bibitem{Maltby:65} P. Maltby \& A. T. Moffet, {\it Science} {\bf 150} 

1065:   (1965) 63.

1066: \bibitem{Rees:66} M. J. Rees, {\it Nature} {\bf 211} (1966) 468.

1067: 

1068: \bibitem{Cohen:68} M. H. Cohen, D. L. Jauncey, K. I. Kellermann

1069:   \& Clark, B. G. {\it Science} {\bf 162} (1968) 88.

1070: \bibitem{Kellermann:68} K. I. Kellermann \etal, {\it Ap. J. Lett.} {\bf 153} 

1071:   (1968) L209.

1072: 

1073: \bibitem{Kellermann:88} K. I. Kellermann \& F. N. Owen, 

1074:   Radio Galaxies and Quasars,  

1075:   in {\it Galactic and Extragalactic Radio Astronomy, Second Edition}, 

1076:   eds. G. L. Verschuur \& K. I. Kellermann 

1077:   (Springer-Verlag, New York, 1988), p. 563. 

1078: \bibitem{Whitney:71} A. R. Whitney \etal, {\it Science} {\bf 173} (1971) 225.

1079: \bibitem{Cohen:71} M. H. Cohen, W. Cannon, G. H. Purcell, D. B. Shaffer,

1080:   J. J. Broderick, K. I. Kellermann \& D. L. Jauncey, 

1081:   {\it Ap. J.} {\bf 170} (1971) 207.

1082: \bibitem{Zensus:97} A. Zensus, {\it Ann. Rev. Ast. Astroph.} {\bf 35} (1997) 607.

1083: 

1084: \bibitem{Blandford:77} R. D. Blandford \& R. L. Znajek, 

1085:   {\it Mon. Not. Roy. Aston. Soc.} {\bf 179} (1977) 433.

1086: 

1087: \bibitem{Ford:94} H. C. Ford \etal, {\it Ap. J. Lett.} {\bf 435} (1994) L27.

1088: \bibitem{Harms:94} R. J. Harms \etal, {\it Ap. J. Lett.} {\bf 435} (1994) L35.

1089: 

1090: \bibitem{Claussen:84} M. J. Claussen, G. M. Heiligman \& K. Y. Lo,  

1091:   {\it Nature} {\bf 310} (1984) 298.

1092: \bibitem{Nakai:93} N. Nakai, M. Inoue \& M. Miyoshi, {\it Nature}

1093:   {\bf 361} (1993) 45.

1094: \bibitem{Miyoshi:95} M. Miyoshi, J. Moran, J. Herrnstein, L. Greenhill, 

1095:   N. Nakai, P. Diamond \& M. Inoue, {\it Nature} {\bf 373} (1995) 127.

1096: \bibitem{Greenhill:95} L. J. Greenhill, D. R. Jiang, J. M. Moran, 

1097:   M. J. Reid, K. Y. Lo \& M. J. Claussen, {\it Ap. J.} {\bf 440} (1995) 619.

1098: \bibitem{Herrnstein:99} J. Herrnstein \etal, {\it Nature} {\bf 400} (1999) 539.

1099: 

1100: 

1101: \bibitem{Maoz:98} E. Maoz {\it Ap. J. Lett.} {\bf 494} (1998) L181.

1102: 

1103: \bibitem{Tanaka:95} Y. Tanaka \etal, {\it Nature} {\bf 375} (1995) 659. 

1104: 

1105: \bibitem{Morris:96} M. Morris \& E. Serabyn {\it Ann. Rev. Ast. Astroph.} {\bf 34}

1106:   (1996) 645.

1107: 

1108: \bibitem{Balick:74} B. Balick \& R. L. Brown, {\it Ap. J.} {\bf 194} (1974) 265.

1109: \bibitem{Lo:85} K. Y. Lo, D. C. Backer, R. D. Ekers, K. I. Kellermann,

1110:   M. J. Reid \& J. M. Moran, {\it Nature} {\bf 315} (1985) 124.

1111: 

1112: \bibitem{Lynden-Bell:71} D. Lynden-Bell \& M. J. Rees {\it Mon. Not. Roy. Aston. Soc.}

1113:   {\bf 152} (1971) 461.

1114: \bibitem{Ozernoy:96} L. M. Ozernoy \& R. Genzel,

1115:   Is Sgr A* Underfed, Underefficient, or Underdone?,

1116:   in {\it Unsolved Problems of the Milky Way, 

1117:   Proceedings of the 169th Symposium of the International Astronomical Union},

1118:   eds. L.Blitz and P.Teuben (Kluwer, Dordrecht, 1996), p. 181.

1119: 

1120: \bibitem{Lacy:80} J. H. Lacy, C. H. Townes, T. R. Geballe \& 

1121:   D. J. Hollenbach, {\it Ap. J.} {\bf 241} (1980) 132.

1122: 

1123: \bibitem{McGinn:89} M. T. McGinn, K. Sellgren, E. E. Becklin \& 

1124:   D. N. B. Hall, {\it Ap. J.} {\bf 338} (1989) 824. 

1125: \bibitem{Sellgren:90} K. Sellgren, M. T. McGinn, E. E. Becklin \& 

1126:   D. N. Hall, {\it Ap. J.} {\bf 359} (1990) 112.

1127: \bibitem{Krabbe:95} A. Krabbe, {\it Ap. J. Lett.} {\bf 447} (1995) L95.

1128: \bibitem{Haller:96} J. W. Haller, M. J. Rieke, G. H Rieke, P. Tamblyn,

1129:   L. Close \& F. Melia, {\it Ap. J.} {\bf 456} (1996) 194.

1130: \bibitem{Genzel:97} R. Genzel, A. Eckart, T. Ott \& F. Eisenhauer 

1131:  {\it Mon. Not. Roy. Ast. Soc.} {\bf 291} (1997) 219.

1132: 

1133: \bibitem{Eckart:96} A. Eckart \& R. Genzel, {\it Nature} {\bf 383} (1996) 415.

1134: \bibitem{Eckart:97} A. Eckart \& R. Genzel, {\it Mon. Not. Roy. Ast. Soc.} {\bf 284} 

1135:  (1997) 576.

1136: \bibitem{Ghez:98} A. M. Ghez, B. L. Klein, M. Morris \& E. E. Becklin

1137:  {\it Ap. J.} {\bf 509} (1998) 678.

1138: \bibitem{Genzel:00} R. Genzel, C. Pichon, A. Eckart, O. E. Gerhard \&

1139:  T. Ott {\it Mon. Not. Roy. Ast. Soc.} {\bf 317} (2000) 348.

1140: 

1141: \bibitem{Ghez:00} A. M. Ghez, M. Morris, E. E. Becklin, A. Tanner \& 

1142:  T. Kremenek {\it Nature} {\bf 407} (2000) 349.

1143: \bibitem{Eckart:02} A. Eckart, R. Genzel, T. Ott \& R. Sch\"odel, 

1144:   {\it Mon. Not. Roy. Ast. Soc.} {\bf 331} (2002) 917.

1145: 

1146: \bibitem{Schoedel:02} R. Sch\"odel \etal, {\it Nature} {\bf 419} (2002) 694.

1147: \bibitem{Ghez:03} A. M. Ghez, E. Becklin, G. Duch\^ene, S. Hornstein,

1148:   M. Morris, S. Salim \& A. Tanner, {\it Astronomische Nachrichten} {\bf 324} 

1149:   (2003) 527.

1150: \bibitem{Schoedel:03} R. Sch\"odel, T. Ott, R. Genzel, A. Eckart,

1151:   N. Mouawad \& T. Alexander, {\it Ap. J.} {\bf 596} (2003) 1015. 

1152: \bibitem{Ghez:05} A. M. Ghez \etal, {\it Ap. J.} {\bf 620} (2005) 744.

1153: 

1154: \bibitem{Bahcall:81} J. N. Bahcall \& S. Tremaine

1155:   {\it Ap. J.} {\bf 244} (1981) 805.

1156: \bibitem{Lu:08} J. R. Lu, A. M. Ghez, S. D. Hornstein, M. R. Morris,

1157:   E. E. Becklin \& K. Matthews, to appear in {\it Ap. J.}.

1158: \bibitem{Leonard:89} P. J. T. Leonard \& D. Merritt

1159:   {\it Ap. J.} {\bf 339} (1989) 195.

1160: \bibitem{Genzel:03b} R. Genzel \etal, {\it Ap. J.} {\bf 594} (2003) 812.  

1161: \bibitem{Schoedel:07} R. Sch\"odel \etal, {\it Astron. Astroph.} {\bf 469} (2007) 

1162:   125.

1163: 

1164: \bibitem{Mouawad:05} N. Mouawad, A. Eckart, S. Pfalzner, R. Schödel, 

1165:   J. Moultaka \& R. Spurzem {\it Astronomische Nachrichten} {\bf 326} (2005) 83.

1166: \bibitem{Ghez:08} A. Ghez \etal, to appear in {\it Ap. J.}.

1167: \bibitem{Gillessen:08} S. Gillessen, F. Eisenhauer, S. Trippe, 

1168:   T. Alexander, R. Genzel, F. Martins \& T. Ott, to appear in {\it Ap. J.}.

1169: 

1170: \bibitem{Menten:97} K. M. Menten, M. J. Reid, A. Eckart \& R. Genzel 

1171:   {\it Ap. J.} {\bf 475} (1997), L111.

1172: \bibitem{Reid:03} M. J. Reid, K. M. Menten, R. Genzel, T. Ott, R. Sch\"odel

1173:   \& Eckart, A. {\it Ap. J.} {\bf 587} (2003), 208.

1174: \bibitem{Reid:07} M. J. Reid, K. M. Menten, S. Trippe, T. Ott \& R. Genzel

1175:   {\it Ap. J.} {\bf 659} (2007), 378.

1176: 

1177: \bibitem{Rogers:94}  A. E. E. Rogers \etal, {\it Ap. J. Lett.} {\bf 434} (1994) L59.

1178: \bibitem{Krichbaum:98} T. P. Krichbaum \etal, {\it Astron. Astroph. Lett.} 

1179:  {\bf 335} (1998) L106.

1180: \bibitem{Doeleman:01} S. S. Doeleman \etal, {\it Astron. J.} {\bf 121} 

1181:  (2001) 2610.

1182: \bibitem{Bower:04} G. C. Bower, H. Falcke, R. M. Herrnstein,

1183:   J.-H. Zhao, W. M. Goss \& D. C. Backer, {\it Ap. J.} {\bf 524} (2004) 805.

1184: \bibitem{Shen:05} Z.-Q. Shen, K. Y. Lo, M.-C. Liang, P. T. P. Ho \&

1185:   J.-H. Zhao, {\it Nature} {\bf 438} (2005) 62.

1186: \bibitem{Gwinn:91} C. R. Gwinn, R. M. Danen, J. Middleditch, L. M. Ozernoy

1187:   \& T. Kh. Tran {\it Ap. J. Lett.} {\bf 381} (1991) L43.

1188: 

1189: \bibitem{Backer:99} D. C. Backer \& R. A. Sramek, {\it Ap. J.} {\bf 524} (1999) 

1190:   805.

1191: \bibitem{Reid:99} M. J. Reid, A. C. S. Readhead, R. C. Vermeulen \& 

1192:   R. N. Treuhaft, {\it Ap. J.} {\bf 524} (1999) 816.

1193: 

1194: \bibitem{Reid:04} M. J. Reid \& A. Brunthaler, {\it Ap. J.} {\bf 616} (2004) 872.

1195: \bibitem{Dehnen:98} W. Dehnen \& J. Binney, {\it Mon. Not. Roy. Ast. Soc.} {\bf 298} 

1196:   (1998) 387.

1197: 

1198: \bibitem{Chatterjee:02} P. Chatterjee, L. Hernquist \& A. Loeb, 

1199:   {\it Ap. J.} {\bf 572} (2002) 371.

1200: \bibitem{Dorband:03} E. N. Dorband, M. Hemsendorf \& D. Merritt,

1201:   {\it  J. Comput. Phys.} {\bf 185} 484.

1202: \bibitem{Merritt:07} D. Merritt, S. Harfst \& G. Bertone {\it Astron. J.} {\bf 6133} 

1203:   (2007) 553.

1204: 

1205: \bibitem{Morris:93} M. Morris, {\it Ap. J.} {\bf 408} (1993) 496.

1206: 

1207: \bibitem{Binney:87} J. Binney \& S. Tremaine, {\it Galactic Dynamics}

1208:  (Princeton Univ. Press, Princeton, 1987).

1209: 

1210: \bibitem{Viollier:93} R. D. Viollier, D. Trautmann \& G. B. Tupper,

1211:   {\it Phys. Lett. B}  {\bf 306} (1993) 79.

1212: \bibitem{Tsiklauri:96} D. Tsiklauri \& R. D. Viollier, {\it Mon. Not. Roy. Ast. Soc.} 

1213:   {\bf 282} (1996) 1299.

1214: \bibitem{Munyaneza:98} F. Munyaneza, D. Tsiklauri\& R. D. Viollier, {\it Ap. J.} 

1215:   {\bf 509} (1998) 105.

1216: \bibitem{Bilic:99} N. Bili\'c, F. Munyaneza \& R. D. Viollier,

1217:   {\it Phys. Rev. D} {\bf 59} 4003.

1218: \bibitem{Munyaneza:02} F. Munyaneza \& R. D. Viollier, {\it Ap. J.} {\bf 564} 

1219:   (2002) 274.

1220: 

1221: \bibitem{Torres:00} D. F. Torres, S. Capozziello \& G. Lambiase,

1222:   {\it Phys. Rev. D} {\bf 62} (2000) 104012.

1223: \bibitem{Robertson:03} S. Robertson \& D. Leiter, {\it Ap. J. Lett.} {\bf 596} 

1224:   (2003) L203. 

1225: 

1226: \bibitem{Garcia:01} M. R. Garcia, J. E. McClintock, R. Narayan, P. Callanan, 

1227:   D. Barret \& S. S. Murray, {\it Ap. J. Lett.} {\bf 553} (2001) L47. 

1228: \bibitem{McClintock:04} J. E. McClintock, R. Narayan \& G. B. Rybicki,

1229:   {\it Ap. J.} {\bf 615} (2004) 402.

1230: \bibitem{Narayan:08} R. Narayan \& J. E. McClintock, {\it New Astron. Rev.}

1231:   {\bf 51} (2008) 733.

1232: 

1233: \bibitem{Doeleman:08} S. S. Doeleman \etal, to appear in {\it Nature}

1234: \bibitem{Broderick:08} A. E. Broderick \& R. Narayan, {\it Ap. J. Lett.} {\bf 638} 

1235:   (2008) L21. 

1236: 

1237: \bibitem{Herrnstein:04} R. M. Herrnstein, J.-H. Zhao, G. C. Bower \&

1238:   W. M. Goss, {\it Astron. J.} {\bf 127} (2004) 3399.

1239: \bibitem{Mauerhan:05} J. C. Mauerhan, M. Morris, F. Walter \& F. K.

1240:   Baganoff, {\it Ap. J. Lett.} {\bf 623} (2005) L25.

1241: \bibitem{Marrone:06} D. P. Marrone, J. M. Moran, J.-H. Zhao \& R. Rao,

1242:   {\it Ap. J.} {\bf 640} (2006) 308.

1243: \bibitem{Genzel:03} R. Genzel \etal, {\it Nature} {\bf 425} (2003) 934.

1244: \bibitem{Ghez:04} A. M. Ghez \etal, {\it Ap. J. Lett.} {\bf 601} (2004) L159. 

1245: \bibitem{Hornstein:07} S. D. Hornstein \etal, {\it Ap. J.} {\bf 667} (2007) 900.

1246: \bibitem{Do:08} T. Do \etal, to appear in {\it Ap. J.}.

1247: \bibitem{Baganoff:01} F. K. Baganoff \etal, {\it Nature} {\bf 413} (2001) 45.

1248: \bibitem{Misner:73} W. Misner, K. S. Thorne \& J. A. Wheeler,

1249:   {\it Gravitation}, (Freeman, New York, 1973)

1250: 

1251: \bibitem{Shafee:06} R. Shafee, J. E. McClintock, R. Narayan, S. W. Davis, 

1252:   L.-X. Li \& R. A. Remillard, {\it Ap. J. Lett.} {\bf 636} (2006) L113. 

1253: \bibitem{McClintock:06} J. E. McClintock, R. Shafee, R. Narayan, R. A. Remillard,

1254:   S. W. Davis \& L.-X. Li, {\it Ap. J.} {\bf 652} (2006) 518.

1255: 

1256: \bibitem{Hills:88} J. G. Hills, {\it Nature} {\bf 331} (1988) 687.

1257: \bibitem{Yu:03} Q. Yu \& S. Tremaine, {\it Ap. J.} {\bf 599} (2003) 1129.

1258: \bibitem{Brown:05} W. R. Brown, M. J. Geller, S. J. Kenyon \& M. J. Kurtz,

1259:   {\it Ap. J. Lett.} {\bf 622} (2005) L33.

1260: \bibitem{OLeary:08} R. M. O'Leary \& A. Loeb, {\it Mon. Not. Roy. Ast. Soc.} 

1261:   {\bf 383} (2008) 86.

1262: 

1263: \bibitem{Ferrarese:02} L. Ferrarese \& D. Merritt, {\it Physics World}

1264:   {\bf 15N6} (2002) 41.

1265: 

1266: \bibitem{Fan:01} X. Fan \etal, {\it Astron. J.} {\bf 122} (2001) 2833.

1267: \bibitem{Haiman:01} Z. Haiman \& A. Loeb, {\it Ap. J.} {\bf 552} (2005) 459.

1268: 

1269: \bibitem{Magorrian:98} J. Magorrian \etal, {\it Astron. J.} {\bf 115} (1998) 2285.

1270: \bibitem{Ferrarese:00} L. Ferrarese \& D. Merritt, {\it Ap. J. Lett.} {\bf 539} 

1271:   (2000) L9.

1272: \bibitem{Springel:05} V. Springel, T. Di Matteo \& L. Hernquist	

1273:   {\it Mon. Not. Roy. Ast. Soc.} {\bf 361} (2005) 776.

1274: \bibitem{Robertson:06} B. Robertson, L. Hernquist, T. J. Cox, T. Di Matteo,

1275:   P. F. Hopkins, P. Martini \& V. Springel, {\it Ap. J.} {\bf 641} (2006) 90.

1276: 

1277: \bibitem{LISA:08} http://lisa.nasa.gov/

1278: 

1279: \bibitem{Falcke:00} H. Falcke, F. Melia \& E. Agol, {\it Ap. J. Lett.} {\bf 528} 

1280:   (2000) L13.

1281: \bibitem{Broderick:06} A. Broderick \& A. Loeb, {\it J. Phys.: Conf. Ser.} 

1282:   {\bf 54} (2006) 448.

1283: 

1284: \bibitem{Eisenhauer:08} F. Eisenhauer \etal, 

1285:   GRAVITY: microarcsecond astrometry and deep interferometric imaging 

1286:   with the VLTI, in {\it A Giant Step: from Milli- to Micro-arcsecond Astrometry, 

1287:   Proceedings of the International Astronomical Union, IAU Symposium, 

1288:   Volume 248},  eds. W. J. Jin, I. Platais and M. A. C. Perryman

1289:   (Cambridge Univ. Press, Cambridge, 2008), p. 100.

1290: 

1291: \bibitem{Weinberg:05} N. N. Weinberg, M. Milosavljevi\'c \& A. M. Ghez,

1292:   {\it Ap. J.} {\bf 622} (2005) 878.

1293: \bibitem{Will:08} C. M. Will, {\it Ap. J. Lett.} {\bf 674} (2008) L25.

1294: 

1295: \end{thebibliography}

1296: 

1297: \end{document}

1298: