1: %\documentclass[preprint2]{aastex}

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

3: 

4: \documentclass[preprint2]{emulateapj}

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

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

7: 

8: \newcommand{\HI}{\ion{H}{1}}

9: \newcommand{\OII}{[\ion{O}{2}]}

10: \newcommand{\kms}{km\,s$^{-1}$}

11: \newcommand{\Msun}{M$_\sun$}

12: 

13: \shorttitle{Radio continuum observations of  VCC128} \shortauthors{Buyle, De

14: Rijcke, Debattista, Ferreras, Pasquali, Seth, Morelli}

15: 

16: \begin{document}

17: 

18: \title{Radio continuum observations of the candidate supermassive

19: black hole in the dwarf elliptical VCC128}

20: 

21: \author{Pieter Buyle\altaffilmark{1}, Sven De Rijcke\altaffilmark{1},

22: Victor P. Debattista\altaffilmark{2,3}, Ignacio Ferreras\altaffilmark{4},

23: Anna Pasquali\altaffilmark{5}, Anil Seth\altaffilmark{6},

24: Lorenzo Morelli\altaffilmark{7}}

25: 

26: \altaffiltext{1}{Sterrenkundig Observatorium, Universiteit Gent,

27: Krijgslaan 281, S9, B-9000, Ghent, Belgium; Pieter.Buyle@UGent.be,

28: Sven.DeRijcke@UGent.be} 

29: \altaffiltext{2}{Centre For Astrophysics

30: University of Central Lancashire

31: Preston, UK PR1 2HE; vpdebattista@uclan.ac.uk} 

32: \altaffiltext{3}{RCUK Academic Fellow} 

33: \altaffiltext{4}{Physics Department, King's College

34: London, Strand, London WC2R 2LS, UK; ferreras@star.ucl.ac.uk}

35: \altaffiltext{5}{Max-Planck-Institut f\"ur Astronomie, Koenigstuhl 17,

36: D-69117 Heidelberg, Germany; pasquali@mpia-hd.mpg.de}

37: \altaffiltext{6}{Harvard-Smithsonian Center for Astrophysics, 60 Garden St., MS20, Cambridge, MA 02138, US; aseth@cfa.harvard.edu}

38: \altaffiltext{7}{Dipartimento di astronomia, Universita' di Padova, Vic. Osservatorio 3, I-35122 Padova, Italy; lorenzo.morelli@unipd.it}

39: 

40: \begin{abstract}

41: The presence of black holes (BHs) at the centers of dwarf elliptical

42: galaxies (dEs) has been argued both theoretically and

43: observationally. Using archival HST/WFPC2 data, we found the Virgo

44: cluster dwarf elliptical galaxy VCC128 to harbor a binary nucleus, a

45: feature that is usually interpreted as the observable signature of a

46: stellar disk orbiting a central massive black hole. \citet{db06}

47: estimated its mass M$_\bullet \sim 6 \times 10^6-5\times

48: 10^7$~M$_\odot$. One of the most robust means of verifying the

49: existence of a BH is radio continuum and/or X-ray emission, however

50: because of the deficiency of gas in dEs, radio continuum emission

51: forms the best option here. We have tried to detect the X-band radio

52: emission coming from the putative black hole in VCC128 when it

53: accretes gas from the surrounding ISM. While we made a positive

54: 4$\sigma$ detection of a point source 4.63$''$ south-west of

55: the binary nucleus, no statistically significant evidence for emission

56: associated with the nuclei themselves was detected.

57: 

58: This implies either that VCC128 has no massive central black hole,

59: which makes the nature of the binary nucleus hard to explain, or, if

60: it has a central black hole, that the physical conditions of the ISM

61: (predominantly its density and temperature) and/or of the surrounding

62: accretion disk do not allow for efficient gas accretion onto the black

63: hole, making the quiescent black hole very hard to detect at radio

64: wavelengths.

65: \end{abstract}

66: 

67: \keywords{galaxies: evolution, galaxies: elliptical, galaxies: ISM, galaxies: fundamental parameters}

68: 

69: \section{Introduction}\label{intro}

70: 

71: The masses of the massive central black holes observed in many

72: galaxies exhibit a variety of scaling relations, such as the

73: M$_\bullet - \sigma$ relation between black-hole mass, M$_\bullet$,

74: and central velocity dispersion, $\sigma$ \citep{g00,fm00} and the

75: $v_{\rm circ} - \sigma$ relation between circular velocity, $v_{\rm

76: circ}$, and central velocity dispersion

77: \citep{f02,b03,pi05,bfg06}. The latter can be interpreted, via the

78: M$_\bullet - \sigma$ relation, as a relation between the mass of the

79: central black hole and the total mass of the host galaxy. These

80: correlations suggest a strong coupling between the formation of

81: massive central black holes and the formation of galaxies.

82: 

83: Literature abounds with scenarios for producing massive central black

84: holes. Models designed to explain the properties of QSOs in massive

85: galaxies typically produce central seed black holes with a minimum

86: mass of M$_\bullet \gtrsim 10^6$M$_\odot$, e.g. through the collapse

87: of a massive central gaseous disk or a supermassive stellar object

88: \citep{lr94,h98,sr98}. This seed can grow even more massive by feeding

89: from a surrounding gas disk. These models are motivated by the

90: necessity of producing supermassive black holes very rapidly after the

91: Big Bang and therefore may be biased to large seed masses.  The

92: gravitational collapse of a relativistic star cluster, supposedly born

93: during a starburst, might produce black hole seeds with masses up to

94: $10^4$~M$_\odot$ \citep{s04}. Scenarios for growing less massive black

95: holes, aptly called intermediate-mass black holes, also exist in the

96: literature \citep{mr01,mh02,pm02}. Their applicability seems to be

97: restricted to very dense systems, such as, for instance, globular

98: clusters or galactic nuclear star clusters, because they rely on

99: dynamical friction to sink stellar-mass seed black holes, possibly

100: originating from Pop{\sc iii} objects as suggested by \citet{mr01}, to

101: the center of the system on short enough timescales.

102: 

103: The picture becomes even more complicated if the fact that galaxies

104: grow via a process of hierarchical merging is taken into

105: account. Recently, the impact of the co-evolution of central black

106: holes and their host galaxies has been incorporated in semi-analytical

107: models (SAMs) of galaxy evolution. E.g., \citet{kh00}, \citet{m07} and

108: \citet{ka05} allow massive central black holes to grow by two

109: processes:~{\em (i)} the influx of gas into the galaxy center due to

110: the starburst and the strongly non-axisymmetric forces that accompany

111: a galaxy-galaxy merger, and {\em (ii)} merging of the black holes

112: after the host galaxies have merged. These SAMs can reproduce the

113: aforementioned scaling relations between the black-hole mass and the

114: properties of the host galaxy. Using a high-resolution small-volume

115: cosmological N-body/SPH simulation, \citet{mhbs07} show they can grow

116: M$_\bullet \sim 10^6$~M$_\odot$ central black holes in Milky-Way size

117: halos by a redshift of $z \sim 6$ by allowing central black holes to

118: merge if their host galaxies merge and by gas accretion during

119: mergers.

120: 

121: However, when the members of a binary black-hole system merge, there

122: can be a considerable beaming of the gravitational radiation emitted

123: during the final stages of the plunge. This results in a recoil of the

124: newly formed black hole. Fully relativistic calculations of black-hole

125: binary mergers have shown that this recoil velocity can be as high as

126: a few hundred~km~s$^{-1}$, depending on the holes' spin, orientation,

127: and orbital eccentricity \citep{p07}. This could potentially hamper

128: the retention of black holes in merging low-mass protogalaxies and by

129: extension the build-up of supermassive black holes.

130: 

131: Clearly, the search for massive central black holes in low-mass

132: galaxies is of considerable interest for the study of the symbiosis of

133: host galaxies and their central black holes and for constraining the

134: minimum mass of seed black holes. Nuclear activity in dwarf galaxies

135: has turned up a number of central black holes with estimated masses in

136: the range M$_\bullet \sim 10^4-10^6$~M$_\odot$

137: \citep{fh03,b04,gh04}. Dynamical modeling of ground-based stellar

138: kinematics of dwarf elliptical galaxies ruled out the presence of

139: central black holes with masses M$_\bullet \gtrsim 10^7$~M$_\odot$

140: \citep{g02}.

141: 

142: \begin{figure*}

143: \includegraphics[width=18cm]{f1.eps}

144: %\caption{V-band image of VCC128. The inset shows a $12.5''\times

145: %  12.5''$ region centered on the binary nucleus, which is just

146: %  resolved in this image, overlayed with a contourplot of the VLA

147: %  radio observations. In increasing order, the contourlevels are

148: %  3$\sigma$, 4$\sigma$, 5$\sigma$, \ldots with $1\sigma$ =

149: %  $30\mu$Jy. The detected source, at $4\sigma$ significance, can be

150: %  seen SW of the center (encircled).

151: %\label{ima_128}}

152: \caption{Contourplot of the VLA radio observations on top of a V-band

153:   image of the center of VCC128. The figure shows the $12''\times

154:   12''$ region centered on the binary nucleus, which is just

155:   resolved in this image. The white contours correspond in increasing

156:   order to the contourlevels 3$\sigma$, 4$\sigma$, 5$\sigma$, \ldots

157:   with $1\sigma$ = $30\mu$Jy, while the gray contours correspond to

158:   $-3\sigma$. The detected source, at $4\sigma$ significance, can be

159:   seen SW of the center (encircled). The synthesized beam is shown in

160:   the bottomleft corner. A V-band image of entire VCC128 is shown in

161:   the topright corner.

162: \label{ima_128}}

163: \end{figure*}

164: 

165: Obtaining and modeling high-quality stellar kinematics of faint dwarf

166: galaxies is, unfortunately, a very time-consuming way of searching for

167: inactive central black holes. This prompted us to look for a more

168: efficient, photometry-based method. Using archival HST images, we

169: found the Virgo Cluster dwarf elliptical galaxy (or dE) VCC128 to

170: harbour a binary nucleus \citep{db06}. At the time, only two other

171: galaxies with binary nuclei were known: M31 and NGC4486B

172: \citep{l93,l96,l05}. In both cases, the binary nucleus was interpreted

173: as the observable signature of a stellar disk orbiting a central

174: massive black hole \citep{t95,l96}. This suggested that the binary

175: nucleus in VCC128 might also be the hallmark of a central massive

176: black hole.

177: 

178: However, the presence of a central massive black hole in VCC128 still

179: needs to be established by independent means. An observationally

180: rather undemanding method for detecting massive central black holes is

181: through the radio emission they produce when they accrete gas. In

182: Section \ref{rce}, we give a short discussion of this method, followed

183: by an account of our radio continuum observations of VCC128 with the

184: Very Large Array (VLA) in Section \ref{obs}. We end by presenting our

185: results and conclusions in Section \ref{results}.

186: 

187: \section{Radio flux density prediction for an accreting black hole} \label{rce}

188: 

189: In order to predict the expected radio continuum flux density of an

190: actively accreting BH, we need an estimate of the BH's mass. In the

191: case of VCC128 we can make use of the optical photometry from Hubble

192: of the double nuclei. By comparing the observed SED of the nuclei with

193: population synthesis models, we estimated the combined nuclear stellar

194: mass at $\sim 10^6$~M$_\odot$. Assuming that the disk to black-hole

195: mass ratios of M31 (M$_{\rm disk}$/M$_\bullet=0.16$) and NGC4486B

196: (M$_{\rm disk}$/M$_\bullet=0.019$) are typical, \citet{db06} estimated

197: the mass of the putative black hole in VCC128 at M$_\bullet \sim 6

198: \times 10^6-5\times 10^7$~M$_\odot$. As a verification they used the

199: Faber-Jackson relation of dEs in order to estimate the galaxy's

200: central velocity dispersion, $\sigma$, at $\sigma \sim

201: 35-65$~km~s$^{-1}$. This result is in good agreement with the

202: empirical scaling M$_\bullet - \sigma$ relation \citep{mf01,w06}.

203: 

204: \citet{mhd03} observationally derived a fundamental plane relation

205: between 5~GHz radio luminosities $L_{\rm cont}$, X-ray luminosities

206: $L_X$ and BH masses M$_\bullet$ of Galactic and extragalactic

207: BHs. Unfortunately, an X-ray luminosity does not exist for VCC128.

208: Still, the fundamental plane relation and its projections provide us

209: with a useful tool to estimate the radio continuum flux density

210: produced by VCC128 if an actively accreting BH were present (see

211: \citet{bash08} for a similar approach). From Fig. 2 in \citet{mhd03},

212: for M$_\bullet \gtrsim 6 \times 10^6$~M$_\odot$ one would expect a

213: 5~GHz radio luminosity well above $L_{\rm cont} \sim 10^{35}$

214: erg~s$^{-1}$.

215: 

216: An alternative method to obtain the radio flux $F_{\rm cont}$ without

217: knowing $L_X$ is proposed by \cite{ma04}, assuming a flat radio

218: spectrum and the Fundamental-Plane relationship between radio

219: luminosity, X-ray luminosity, and black hole mass

220: \citep{mhd03,fa04,dbd06}. For a given gas accretion rate onto the

221: black hole, expressed as a fraction of the Bondi accretion rate which

222: depends on the gas density and temperature, the X-ray luminosity can

223: be calculated, assuming that 10~\% of the rest mass energy of the

224: infalling matter is converted into radiation. 

225: %, and that the total

226: %luminosity $L_{\rm tot}$ is related to the X-ray luminosity as given

227: %by $L_{\rm tot} = 6\times 10^{-3} L_X^{0.5} + L_X$ (using the

228: %Eddington luminosity as unit). This relation takes into account the

229: %energy of a relativistic jet \citep{fe03}. This way, the radio

230: %continuum flux $F_{\rm cont}$ can be expressed as a function of the

231: %black hole mass $M_\bullet$, the distance $d$, and the gas density $n$

232: %and temperature $T$.

233: Unfortunately, no H$\alpha$ imaging or 21~cm radio line emission data

234: are available for VCC128, leaving us without direct constraints on the

235: gas density and temperature or even an indication that gas is at all

236: present in this dwarf galaxy.

237: 

238: % It is well known that while dwarf

239: %ellipticals have old stellar populations and reside in dense cluster

240: %environments where they are vulnerable to ram-pressure stripping, some

241: %have been able to retain a sizable reservoir of neutral and ionised

242: %gas clouds \citep{c03,sd03,m04,bdm05}. \citet{g03} failed to detect

243: %ionised gas in two Local Group dwarf spheroidal galaxies, placing a

244: %$3\sigma$ upper limit of $\sim 10^5$~M$_\odot$ on the diffuse ionised

245: %ISM of these systems. These authors note these dwarf spheroidals could

246: %very well contain substantial amounts of diffuse interstellar gas

247: %coming from stellar mass loss without violating these upper

248: %limits. Moreover, numerical simulations of dwarf galaxies subjected to

249: %ram-pressure stripping show that even the most lightweight galaxies

250: %are able to hold on to part of their gas reservoir, especially near

251: %the galaxy center, over gigayear timescales \citep{mb00,mmw06}. In the

252: %following, we will assume a temperature $T = 10^4$~K and a gas density

253: %$n \sim 0.1$~cm$^{-3}$, which is a typical value for globular clusters

254: %and for the warm diffuse ionised gas that is a major phase of the ISM

255: %of our Galaxy, and about what is expected from stellar mass loss

256: %(\citet{fr01,ma05,m06}).

257: 

258: %\begin{figure}

259: %\includegraphics[angle=-90,width=10.5cm]{mass.eps}

260: %\caption{The relation between black-hole mass and radio flux for

261: %  different accretion rates, parameterized as fractions of the Bondi

262: %  rate, and gas densities. The horizontal dotted and full line

263: %  indicate the $1\sigma$ and $3\sigma$ radio flux levels,

264: %  respectively. The full lines trace the M$_\bullet-{\rm F}_{\rm

265: %    cont}$ relation for a gas density $n = 0.1$~cm$^{-3}$, typical for

266: %  the diffuse ionised ISM, and different accretion rates. The dashed

267: %%  lines trace the same relations but now for a density

268: %  $n=100$~cm$^{-3}$, more typical of Galactic H{\sc ii}

269: %%  regions. Clearly, without strong constraints on the 

270: %\label{mass}}

271: %\end{figure}

272: 

273: %The massive central black hole of the Milky Way and its homologues in

274: %nearby galaxies seem to accrete much more slowly than the Bondi rate,

275: %with $\dot{\rm M}_\bullet < 10^{-2} \dot{\rm M}_{\bullet,\rm Bondi}$

276: %\citep{qg00,yqn03,di01}. The low number of Galactic neutron stars that

277: %were detected in the ROSAT All-Sky Survey can be explained if they are

278: %accreting at a rate of roughly 0.1~\% of the Bondi rate \citep{pe03},

279: %although, of course, a neutron star accreting from the interstellar

280: %medium is a far cry from a central black hole. Theoretical models of

281: %the accretion of matter with non-zero angular momentum onto a black

282: %hole show a reduction of the accretion rate down to the level of a few

283: %percent of the Bondi rate \citep{kr05,pb03}. In the following, we will

284: %assume an accretion rate in the range $ 10^{-5} \dot{\rm

285: %  M}_{\bullet,\rm Bondi} < \dot{\rm M}_\bullet < 10^{-2} \dot{\rm

286: %  M}_{\bullet,\rm Bondi}$.

287: 

288: Thus, it is in principle possible to detect massive central black

289: holes and to estimate their masses (or place broad upper limits

290: thereon), allowing for a large uncertainty due to the combination of

291: several empirical scaling relations and the unknown properties of the

292: accreting gas, using straightforward and relatively short radio

293: observations. We wish to stress that the detection of a radio

294: continuum signal from the center of VCC128 would provide strong

295: evidence for the presence of a central black hole, independent of the

296: reliability and accuracy of the methods with which we can estimate its

297: mass.

298: 

299: \section{Observations and data reduction}\label{obs}

300: 

301: We observed VCC128 with the Very Large Array (VLA) in New Mexico (USA)

302: around midnight of June 28, 2007 (project number AB1248). The VLA is

303: being replaced gradually by the Expanded VLA (EVLA) and, therefore, at

304: the time of our observations consisted of 16 antennas. The radio

305: continuum emission in a region around the center of the galaxy was

306: mapped in the X-band (8.6~GHz), which is at the VLA's maximum

307: sensitivity. The VLA was in its A configuration at that time, yielding

308: a spatial resolution of $0.24''$ corresponding to 19~pc at the

309: distance of VCC128 (16.5~Mpc). The total project time was $3~h$. At

310: startup, the flux calibrator 1331+305 was observed for 7 minutes,

311: followed by alternating observations of the phase calibrator 1222+042

312: (with an angular separation of 5.5$^\circ$ from VCC128) with a 2

313: minute exposure time and the galaxy VCC128 with a 15 minute exposure

314: time. This gives a total on-source integration time of 129

315: minutes. Standard flagging and calibration of the $u-v$ data was

316: performed with the Astronomical Image Processing Software (AIPS). We

317: preferred to use MIRIAD \citep{sault95} to derive a natural weighted

318: map. This map was cleaned (1000 iterations) and yields a final image

319: beam size of 0.19$'' \times $0.18$''$.

320: 

321: \section{Results and conclusions}\label{results}

322: 

323: We created a radio continuum map of the central $12''\times 12''$

324: region of VCC128 with an rms noise of $30~\mu$Jy/beam. This map is

325: presented in Fig. \ref{ima_128}, overplotted on a V-band image of

326: VCC128 obtained with FORS1 mounted on the VLT (ESO program

327: 079.B-0632(B)). The data reduction and analysis of the optical imaging

328: of VCC128 will be reported in detail elsewhere. A 4$\sigma$ detection

329: of a point source was made at R.A. $12h~14m~59.66s$, DEC. $+9^\circ~

330: 33'~50.94''$ (J2000). This is 4.63$''$ SW of the binary

331: nucleus. The source has a flux density of $1.66 \pm

332: 0.05$~mJy. Unfortunately, no emission was found at the optical

333: position of the nucleus.

334: 

335: Since we do not detect any radio emission coming from the center of

336: VCC128, we can only place a $S_\nu \approx 90~\mu$Jy $3\sigma$ upper

337: limit on the radio continuum emission of the putative central black

338: hole. Assuming a flat radio spectrum, this corresponds to a $3\sigma$

339: upper limit on the 5~GHz radio luminosity of $L_{\rm cont} = \nu L_\nu

340: \approx 1.5 \times10^{35}$~erg~s$^{-1}$. This is lower than the expected

341: radio luminosity of an actively accreting BH with a mass M$_\bullet

342: \gtrsim 10^6$~M$_\odot$.

343: 

344: We conclude that either {\em (i)} VCC128 does not contain a central

345: massive black hole or, alternatively, {\em (ii)} VCC128 might contain

346: a central massive black hole with a mass of the order of a few

347: $10^6$~M$_\odot$ but the gas near the galaxy center is either too hot

348: or too rarefied or conditions near the black hole are unsuitable for

349: efficient accretion. In the former case, one needs to rethink the

350: nature of the binary nucleus. As detailed in \citet{db06}, all other

351: explanations for the binary nucleus other than it being the

352: observational signature of a stellar disk rely on rather contrived

353: chance configurations. 

354: 

355: %Given the very shallow gravitational potential

356: %of VCC128 it would then be very unclear how this disk can remain

357: %stable at the galaxy center without some central, unseen mass

358: %concentration.

359: 

360: %Although no detection is made, we can still use the M$_\bullet-{\rm

361: %  F}_{\rm cont}$ relation of \citet{ma04} to derive very broad

362: %constraints on the physical conditions near the putative BH in VCC128

363: %that would prevent it from emitting more radio continuum emission:

364: %\begin{itemize}

365: %\item the gas density must be below $\sim 0.005$~cm$^{-3}$, for

366: %$T=10^4$~K and an accretion rate of $10^{-3}$ times the Bondi rate,

367: %\item the accretion rate must to well below $10^{-5}$ times the Bondi

368: %rate, for $T=10^4$~K and $n=0.1$~cm$^{-3}$,

369: %\item the temperature must be higher than $\sim 10^5$~K, for

370: %$n=0.1$~cm$^{-3}$ and an accretion rate of $10^{-3}$ times the Bondi

371: %rate.

372: %\end{itemize}

373: %In short, our VLA radio continuum observations of VCC128 did not yield

374: %a detection of an accreting massive central black hole with a mass of

375: %%the order M$_\bullet \sim 6 \times 10^6-5\times 10^7$~M$_\odot$ (as

376: %claimed by \citet{db06}) in this Virgo cluster dwarf elliptical

377: %galaxy. Our observations however are not conclusive as many different

378: %physical conditions are involved in the determination of the accretion

379: %rate of a BH.

380: 

381: \acknowledgements PB and SDR are post-doctoral fellows of the Fund for

382: Scientific Research - Flanders, Belgium (F.W.O.). LM is supported by

383: grant (CPDR061795/06) by Padova University. The Very Large Array is a

384: facility instrument of the National Radio Astronomy Observatory, which

385: is operated by Associated Universities, Inc. under contract from the

386: National Science Foundation.

387: 

388: \begin{thebibliography}{}

389:   

390: \bibitem[Baes et al. (2003)]{b03} Baes, M., Buyle, P., Hau, G. K. T.,

391: \& Dejonghe, H., 2003, MNRAS, 341, L44

392: \bibitem[Barth et al. (2004)]{b04} Barth, A. J., Ho, L. C., Rutledge,

393: R. E., \& Sargent, W. L. W., 2004, \apj, 607, 90

394: \bibitem[Bash et al. (2008)]{bash08} Bash, F. N., Gebhardt, K., Goss,

395: W. M., \& Vanden Bout, P. A., 2008, \aj, 135, 182

396: %\bibitem[Buyle et al. (2005)]{bdm05} Buyle, P., De Rijcke, S.,

397: %Michielsen, D., Baes, M., Dejonghe, H., 2005, \mnras, 360, 853

398: \bibitem[Buyle et al. (2006)]{bfg06} Buyle, P., Ferrarese, L.,

399: Gentile, G., Dejonghe, H., Baes, M., Klein, U., 2006, \mnras, 373, 700

400: %\bibitem[Conselice et al. (2003)]{c03} Conselice, C. J., O'Neil, K.,

401: %Gallagher, J. S., Wyse, R. F. G., 2003, \apj, 591, 167

402: \bibitem[Debattista et al. (2006)]{db06} Debattista, V. P., Ferreras, I.,

403: Pasquali, A., Seth, A., De Rijcke, S., Morelli, L., 2006, \apj, 651, L97

404: %\bibitem[De Rijcke et al. (2003)]{sd03} De Rijcke, S., Zeilinger,

405: %W. W., Dejonghe, H., Hau, G. K. T., 2003, \mnras, 339, 225

406: \bibitem[De Rijcke et al. (2005)]{sd05} De Rijcke, S., Michielsen, D.,

407: Dejonghe, H., Zeilinger, W. W., \& Hau, G. K. T., 2005, \aap, 438, 491

408: \bibitem[De Rijcke, Buyle, Dejonghe (2006)]{dbd06} De Rijcke, S.,

409: Buyle, P., Dejonghe, H., 2006, \mnras, 368, L43

410: %\bibitem[Di Matteo, Carilli, Fabian (2001)]{di01} Di Matteo,

411: %T., Carilli, C. L., Fabian, A. C., 2001, \apj, 547, 731

412: \bibitem[Falcke, K\"ording, Markoff (2004)]{fa04}

413: Falcke, H., K\"ording, E., Markoff, S., 2004, \aap, 414, 895

414: %\bibitem[Fender, Gallo, Jonker (2003)]{fe03} Fender, R. P., Gallo, E.,

415: %Jonker, P., 2003, MNRAS, 343, L99

416: \bibitem[Ferrarese \& Merritt (2000)]{fm00} Ferrarese, L. \& Merritt,

417: D., 2000, \apj, 539, L9

418: \bibitem[Ferrarese (2002)]{f02}  Ferrarese, L. 2002, \apj, 578, 90

419: \bibitem[Filippenko \& Ho (2003)]{fh03} Filippenko, A. V., \& Ho,

420: L. C., 2003, \apj, 588, L13

421: %\bibitem[Freire et al. (2001)]{fr01} Freire, P. C., Kramer, M., Lyne,

422: %A. G., Camilo, F., Manchester, R. N., D'Amico, N., 2001, \apj, 557,

423: %L105

424: %\bibitem[Gallagher et al. (2003)]{g03} Gallagher, J. S., Madsen,

425: %G. J., Reynolds, R. J., Grebel, E. K., Smecker-Hane, T. A., 2003,

426: %\apj, 588, 326

427: \bibitem[Gebhardt (2000)]{g00} Gebhardt, K., et al,. 2000, \apj, 539,

428: L13

429: \bibitem[Geha et al. (2002)]{g02}  Geha, M., Guhathakurta, P., \& van der Marel, R. P., 2002, \aj, 124, 3073

430: \bibitem[Greene \& Ho (2004)]{gh04}  Greene, J. E., \& Ho, L. C., 2004, \apj, 610, 722

431: \bibitem[Haehnelt et al. (1998)]{h98} Haehnelt, M. G., Natarajan, P.,

432: Rees, M. J., 1998, \mnras, 300, 817

433: \bibitem[Kauffmann \& Haehnelt (2000)]{kh00} Kauffmann, G. \& Haehnelt,

434: M., 2000, \mnras, 311, 576

435: \bibitem[Kazantzidis et al. (2005)]{ka05} Kazantzidis, S., et al., 2005, \apj, 623, L67

436: %\bibitem[Krumholz, McKee, Klein (2005)]{kr05} Krumholz, M. R., McKee,

437: %C. F., Klein, R. I., 2005, \apj, 618, 757

438: \bibitem[Lauer et al. (1993)]{l93} Lauer, T. R., et al., 1993, \aj, 106, 1436

439: \bibitem[Lauer et al. (1996)]{l96} Lauer, T. R., et al., 1996, \apj, 482, L139

440: \bibitem[Lauer et al. (2005)]{l05} Lauer, T. R., et al., 2005, \aj, 129, 2138

441: \bibitem[Loeb \& Rasio (1994)]{lr94} Loeb, A. \& Rasio, F. A., 1994,

442: \apj, 432, 52

443: \bibitem[Maccarone (2004)]{ma04} Maccarone, T. J.,

444: 2004, \mnras, 351, 1049

445: %\bibitem[Maccarone, Fender, Tzioumis (2005)]{ma05} Maccarone, T. J.,

446: %Fender, R. P., Tzioumis, A. K., 2005, \mnras, 456L, 17

447: \bibitem[Madau \& Rees (2001)]{mr01} Madau, P. \& Rees, M. J., 2001,

448: \apj, 551, L27

449: %\bibitem[Madsen, Reynolds, Haffner (2006)]{m06} Madsen, G. J.,

450: %Reynolds, R. J., Haffner, L. M., 2006, \apj, 652, 401

451: \bibitem[Malbon et al. (2007)]{m07} Malbon, R. K., Baugh, C. M.,

452: Frenk, C. S., Lacey, C. G., 2007, \mnras, 382, 1394

453: %\bibitem[Mayer et al. (2006)]{mmw06} Mayer, L., Mastropietro, C.,

454: %Wadsley, J., Stadel, J., Moore, B., 2006, \mnras, 369, 1021

455: \bibitem[Merloni, Heinz, di Matteo (2003)]{mhd03}

456: Merloni, A., Heinz, S., di Matteo, T., 2003, \mnras, 345, 1057

457: \bibitem[Merritt \& Ferrarese (2001)]{mf01} Merrit, D. \& Ferrarese,

458: L., 2001, \apj, 547, 140

459: %\bibitem[Michielsen et al. (2004)]{m04} Michielsen, D., De Rijcke, S.,

460: %Zeilinger, W. W., Prugniel, P., Dejonghe, H., Roberts, S., 2004,

461: %\mnras, 353, 1293

462: \bibitem[Micic et al. (2007)]{mhbs07} Micic, M., Holley-bockelmann,

463: K., Sigurdsson, S., Abel, T., 2007, \mnras, 380, 1533

464: \bibitem[Miller \& Hamilton (2002)]{mh02} Miller, M. C. \& Hamilton,

465: D. P., 2002, \mnras, 330, 232

466: %\bibitem[Mori \& Burkert (2000)]{mb00} Mori, M. \& Burkert, A., 2000,

467: %\apj, 538, 559

468: %\bibitem[Perna et al. (2003)]{pe03} Perna, R.,

469: %Narayan, R., Rybicki, G., Stella, L., Treves, A., 2003, ApJ, 594, 936

470: \bibitem[Pizzella et al. (2005)]{pi05} Pizzella, A.,

471: Corsini, E.M., Dalla Bont\'a, E., Sarzi, M., Coccato, L., Bertola, F., 2005, ApJ, 631, 785

472: \bibitem[Pollney et al. (2007)]{p07} Pollney, D. et al., 2007,

473: Phys. Rev. D 76, 124002

474: \bibitem[Portegies Zwart \& McMilan (2002)]{pm02} Portegies Zwart,

475: S. F., McMillan, S. L. W., 2002, \apj, 576, 899

476: %\bibitem[Proga \& Begelman (2003)]{pb03} Proga, D. \&

477: %Begelman, M. C., 2003, \apj, 582, 69

478: %\bibitem[Quataert \& Gruzinov (2000)]{qg00} Quataert,

479: %E. \& Gruzinov, A., 2000, \apj, 545, 842

480: \bibitem[Sault, Teuben \& Wright (1995)]{sault95} Sault, R.~J.,

481:   Teuben, P.~J., Wright, M.~C.~H., 1995, in Shaw R., Payne H.~E.,

482:   Hayes J.~J.~E., eds., ASP Conf. Ser. Vol. 77, Astronomical Data

483:   Analysis Software and Systems IV. Astron. Soc. Pac., San Fransisco,

484:   p.433 360, 587

485: \bibitem[Shapiro (2004)]{s04} Shapiro, S. L., 2004, \apj, 613, 1213

486: \bibitem[Silk \& Rees (1998)]{sr98} Silk, J. \& Rees, M. J., 1998,

487: \aap, 331, L1

488: \bibitem[Tremaine (1995)]{t95} Tremaine, S., 1995, \aj, 110, 628

489: \bibitem[Wyithe (2006)]{w06}  Wyithe, J. S. B. 2006, \mnras, 365, 1082

490: %\bibitem[Yan, Quataert, Narayan (2003)]{yqn03} Yuan, F., Quataert, E.,

491: %Narayan, R., 2003, \apj, 598, 301

492: \end{thebibliography}

493: 

494: \end{document}

495: 

496: %%% Local Variables: 

497: %%% mode: latex

498: %%% TeX-master: t

499: %%% End: 

500: