1: \documentclass[usegraphicx,usenatbib]{mn2e}
2:
3: \newcommand{\kms}{km s$^{-1}$}
4: \newcommand{\mh}{M_{\bullet}}
5: \newcommand{\milos}{Milosavljevi\'c\ }
6: \newcommand{\mdef}{M_{\rm def}}
7:
8: \def\gap{\;\rlap{\lower 2.5pt
9: \hbox{$\sim$}}\raise 1.5pt\hbox{$>$}\;}
10: \def\lap{\;\rlap{\lower 2.5pt
11: \hbox{$\sim$}}\raise 1.5pt\hbox{$<$}\;}
12:
13: \begin{document}
14: \title[Galaxy Cores]{Galaxy Cores as Relics of Black Hole Mergers}
15: \author[\milos {\it et al.}]
16: {Milo\v s Milosavljevi\' c$^1$, David Merritt$^1$, Armin Rest$^2$, and Frank C. van den Bosch$^3$ \\
17: $^1$Department of Physics and Astronomy, Rutgers University, Piscataway, NJ, 08854, USA \\
18: $^2$ Department of Astronomy, University of Washington, Seattle, WA, 98195, USA \\
19: $^3$ Max-Planck Institut f\"ur Astrophysik, Karl-Schwarzschild Strasse 1, Postfach 1317, 85741 Garching, Germany}
20:
21: \maketitle
22:
23: \begin{abstract}
24: We investigate the hypothesis that the cores of elliptical galaxies and
25: bulges are created from the binding energy liberated by the coalescence of
26: supermassive binary black holes during galaxy mergers.
27: Assuming that the central density profiles of galaxies were
28: initially steep power laws, $\rho\sim r^{-2}$, we define the
29: ``mass deficit'' as the mass in stars that had to be removed from
30: the nucleus in order to produce the observed core.
31: We use nonparametric deprojection to compute the mass deficit in a
32: sample of 35 early-type galaxies with high-resolution imaging data.
33: We find that the mass deficit correlates well with the mass of the nuclear
34: black hole, consistent with the predictions of merger models.
35: We argue that cores in halos of non-interacting dark matter
36: particles should be comparable in size to those observed in the stars.
37: \end{abstract}
38: \begin{keywords}
39: black holes: binary black holes--- galaxies: elliptical and
40: lenticular, cD--- galaxies: interactions--- galaxies: nuclei
41: \end{keywords}
42:
43: \section{Introduction}
44: \citet{fer94} and \citet{lau95}
45: divide elliptical galaxies
46: into two classes based on their nuclear properties, which Lauer {\it et al.}
47: call ``core'' and ``power-law'' galaxies.
48: Core galaxies exhibit a definite break in the surface brightness profile
49: at some radius $R_b$; inward of this break, the logarithmic slope
50: gently decreases in a manner that mimics a constant-density core.
51: Power-law galaxies show essentially a single power-law profile
52: throughout their inner regions, $\Sigma(R)\sim R^{-\Gamma}$,
53: $\Gamma\approx -0.8\pm 0.2$.
54: The brightest galaxies, $M_V\lap -21$, are exclusively core galaxies
55: while faint galaxies, $M_V\gap -16$ always exhibit power laws;
56: galaxies of intermediate luminosity can exhibit either type of profile
57: \citep{geb96}.
58: While the two categories were initially seen as distinct,
59: nonparametric deprojection revealed that even the ``core'' galaxies
60: exhibit power laws in their central space densities,
61: $\rho\sim r^{-\gamma}$, with $\gamma\lap 1$ \citep{mef95}.
62: Power-law galaxies have $1\lap\gamma\lap 2.5$ \citep{geb96}.
63: Furthermore the distribution of de-projected slopes is essentially
64: continuous as a function of galaxy luminosity in the larger samples
65: now available \citep{rav01, res01}.
66:
67: Here we assume that the steep central density cusps of faint ellipticals
68: and bulges, $\rho\sim r^{-2}$, are characteristic of the earliest generation
69: of galaxies, and ask: How do the low-density cores associated with bright
70: galaxies form?
71: An appealing hypothesis links cores to nuclear black holes (BHs):
72: in a galactic merger, the BHs will fall to the center of the merger remnant
73: and form a bound pair, releasing their binding energy to the surrounding
74: stars \citep{bbr80,ebi91}.
75: High-resolution $N$-body simulations verify that this process can
76: convert a steep power-law cusp, $\rho\sim r^{-2}$, into a shallow
77: power-law cusp, $\rho\sim r^{-1}$, within the radius of gravitational
78: influence of the BHs \citep{mme01}.
79: Successive mergers would presumably lower the density of the core still
80: more.
81: In this model, power-law galaxies are those which have not experienced
82: a major merger since the era of peak BH growth, or which
83: have re-generated their cusps via star formation \citep{mme01}.
84:
85: Preliminary tests of the cusp-disruption model were presented by
86: \citet{fab97} and \citet{mme01}.
87: The former authors plotted core properties (break radius,
88: core luminosity) versus global properties
89: in a sample of 19 early-type galaxies and noted a rough proportionality.
90: In the paradigm investigated here,
91: core properties should correlate more fundamentally with BH
92: mass, since the mass of stars ejected by a decaying binary BH is expected
93: to be of order the BHs' mass.
94: \citet{mme01} used the new empirical relation between
95: galaxy velocity dispersion and BH mass, the $\mh-\sigma$ relation
96: \citep{fem00,geb00}, to estimate BH masses in the Faber
97: {\it et al.} sample. They found that a rough dynamical estimate of the
98: ``mass deficit'' -- the mass that would need to
99: be removed from an initially $r^{-2}$ density cusp in order to produce
100: the observed profile -- correlated well with $\mh$.
101:
102: In this paper we present the most careful test to date of the BH
103: merger hypothesis for the formation of galaxy cores.
104: We use nonparametric deprojection
105: to compute the mass deficit in a sample of galaxies with high-resolution
106: imaging data from HST (\S2). We find a strong correlation
107: between this mass and the mass of the nuclear BH; typical
108: ejected masses are $\sim$ several $\mh$ (\S3).
109: We argue (\S4) that this result is consistent with the formation of
110: cores via hierarchical mergers of galaxies containing pre-existing BHs.
111: Cusps of non-interacting dark matter particles should behave in
112: the same way as cusps of stars in response to heating by binary BHs,
113: and we argue that the damage done to stellar cusps by this mechanism
114: is a reasonable guide to the damage that would be done to dark matter cusps.
115:
116: \section{Data and Method}
117: \label{sec_data}
118: Our data set is drawn from a sample of 67 surface brightness profiles of
119: early-type galaxies observed with HST/WFPC2 by \citet{res01},
120: and three additional galaxies: NGC~4472 and 4473, observed with
121: WFPC1 by \citet{fer94}; and a WFPC2 F547M image of M87
122: \citep{jor02}.
123: The Rest {\it et al.} sample was selected from the set of all early-type
124: galaxies with radial velocities less than 3400 km s$^{-1}$,
125: absolute $V$-band magnitudes less than $-18.5$, and absolute galactic
126: latitude exceeding 20 degrees.
127: From this sample we excluded 13 galaxies for which central velocity
128: dispersions were not available in literature; as discussed below,
129: velocity dispersions were needed to compute BH masses in most of the galaxies.
130: For specifics of the image-data reduction we refer the reader to the
131: sources cited above. Surface brightness profiles used in this study were
132: major-axis profiles. We applied a crude correction for the apparent
133: ellipticity of the galaxies by multiplying the volume-integrated quantities,
134: defined below, by $(1-\epsilon_b)$, where $\epsilon_b$ is the
135: ellipticity of the isophote at the break radius.
136:
137: The intrinsic luminosity profiles $\nu(r)$ were obtained by deprojecting
138: the PSF-deconvolved surface brightness profiles $\Sigma(R)$ using the
139: non-parametric MPL technique \citep{met94}.
140: We opted for one-step deprojection via maximization of the penalized
141: likelihood functional
142: \begin{equation}
143: \label{eq_mpl} {\mathcal L}_\lambda [\nu] =
144: \sum_i \frac{\left(\Sigma_i - P_i[\nu] \right)^2}{\left(\Sigma^{err}_i\right)^2}
145: -\lambda \int_0^\infty \left[\frac{d^2\log\nu}{(d\log r)^2}\right]^2 d\log r.
146: \end{equation}
147: The first term compares the observed surface brightness data
148: $\Sigma_i\equiv\Sigma(R_i)$ to the projections $P_i[\nu]$ of the intrinsic luminosity estimate $\nu$ given by the operator
149: \begin{equation}
150: P_i[\nu] = 2 \int_{R_i}^\infty
151: \frac{\nu(r) r dr}{\sqrt{r^2-{R_i}^2}}.
152: \end{equation}
153: The second term in equation \ref{eq_mpl} is the penalty function which assigns
154: zero penalty to any power-law $\nu(r)$;
155: hence our estimate of $\nu$ is unbiassed if $\nu$ is an unbroken power law
156: and should be minimally biassed if $\nu$ is approximately a power law.
157: The relative strength of the penalty term is regulated through the parameter
158: $\lambda$ which was chosen by eye and equal to $0.01$ for all galaxies;
159: integrated quantities like the mass deficit
160: defined below are only weakly dependent on $\lambda$.
161:
162: Distances for 26 of the galaxies were drawn from the SBF survey \citep{ton01}.
163: For the remaining 34 galaxies we adopted distances computed assuming
164: a pure Hubble expansion with $H_0=80\textrm{ km s}^{-1}\textrm{ Mpc}^{-1}$
165: corrected for Virgo-centric infall \citep{res01}.
166: Luminosity densities were converted to mass densities
167: $\rho(r)=\Upsilon\nu(r)$ using the individual mass-to-light ratios
168: $\Upsilon_V$ quoted in \citet{mag98}
169: or their best-fit relation $\log(\Upsilon_V/\Upsilon_{\odot,V})=
170: -1.11\pm0.33 +(0.18\pm0.03)\log(L_V/L_{\odot,V})$ for galaxies not
171: included in that study.
172:
173: We define $\gamma\equiv-d\log\rho(r)/d\log(r)$ as the local, negative
174: logarithmic slope of the deprojected density profile.
175: Power-law galaxies are defined as those in which $\gamma\ge 2$ at
176: all radii; typically the profiles of such galaxies show no clear
177: feature that can be identified as a ``break radius'' and are unlikely
178: candidates for cusp destruction by binary BHs.
179: In the remaining 35, ``core'' galaxies, the slope varies from $\gamma>2$
180: at large radii to $\gamma<2$ at small radii; the radius at which
181: the slope crosses $\gamma=2$ in the positive sense ($d\gamma/dr>0$) is
182: called here the ``break radius'' $r_b$ (Figure \ref{fig_defin}). This definition has
183: little in common with the more standard definition
184: based on fitting of the {\it surface brightness} profile to an ad hoc
185: parametric function.
186: In four galaxies the slope crosses $\gamma=2$ in the positive sense at
187: more than one radius and thus the definition of $r_b$ is ambiguous.
188: In such cases, we select the crossing toward larger radius
189: from the largest dip of the slope below $\gamma=2$.
190:
191: \begin{figure}
192: \includegraphics[width=8.0cm]{figure1.eps}
193: \caption{Definition of the mass deficit, illustrated using the profile
194: of NGC~5903.
195: Left panel: density in $M_\odot\textrm{pc}^{-3}$ as a function of radius
196: in parsecs (thick line); hypothetical original density cusps
197: for fiducial slopes $\gamma_0=2, 1.75, 1.5$ (thin lines).
198: Dots indicate the corresponding break radii.
199: Shaded region is the mass deficit for $\gamma_0=1.5$.
200: Right panel: the negative logarithmic derivative
201: of the density profile, $\gamma\equiv-d\log\rho/d\log r$.
202: Break radii are defined as the radii where $\gamma=\gamma_0$.}
203: \label{fig_defin}
204: \end{figure}
205:
206: We define the {\it mass deficit} as the difference in integrated mass
207: between the deprojected density profile $\rho(r)$ and a $\gamma=\gamma_0=2$
208: profile extrapolated inward from the break radius:
209: \begin{equation}
210: \mdef\equiv 4\pi (1-\epsilon_b) \int_0^{r_b}
211: \left[\rho(r_b) \left(\frac{r}{r_b}\right)^{-\gamma_0} -\rho(r)\right] r^2 dr.
212: \end{equation}
213: Our choice of an $r^{-2}$ density profile to characterize the
214: ``undisrupted'' core is to a certain extent arbitrary;
215: adiabatic growth of BHs can produce cusps with $1.5\lap\gamma\lap 2.5$
216: depending on initial conditions, and the faintest ellipticals with
217: measured cusp slopes exhibit a similar range of $\gamma$'s (e.g.,
218: \citet{geb96}).
219: To test the sensitivity of our results to the assumed initial profile,
220: we repeated the analysis using fiducial slopes of $\gamma_0=1.75$ and
221: $1.5$.
222: Values of $\gamma_0>2$ were found to exclude all but a few galaxies.
223:
224: BH masses for a few of the galaxies in our sample are
225: available from spatially-resolved kinematical studies
226: \citep{mef01}.
227: For all other galaxies we estimated $\mh$ via the $\mh-\sigma$ relation,
228: \begin{equation}
229: \label{eq_msigma}
230: \mh\approx 1.4\times10^8 M_\odot
231: \left(\frac{\sigma_c}{200\textrm{ km s}^{-1}}\right)^{4.8\pm0.5}
232: \end{equation}
233: \citep{fem00}, where $\sigma_c$ is the central
234: velocity dispersion corrected to an aperture of $r_e/8$ with $r_e$
235: the effective radius.
236: We used the aperture corrections of \citet{jor95}
237: to compute $\sigma_c$ from published values of $\sigma$ in
238: \citet{dav87}, \citet{ton81} and \citet{din95}.
239:
240: \section{Results}
241: Results are given in Tables \ref{tab_galaxies} and \ref{tab_slopes} and Figure \ref{fig_data};
242: Table \ref{tab_galaxies} gives
243: mass deficits only for $\gamma_0=2$ while Figure \ref{fig_data} shows
244: $\mdef$ computed using all three values of the fiducial slope,
245: $\gamma_0$=(2, 1.75, 1.5).
246: In the first panel of Figure \ref{fig_data}
247: we have also plotted ``dynamical'' estimates of
248: $\mdef$ for a sample of galaxies from \citet{geb96},
249: using equation (41) from \citet{mme01},
250: $M_{\rm dyn}\equiv2(2-\gamma)/(3-\gamma)\sigma^2R_b/G$;
251: $M_{\rm dyn}$ depends on the density profile only through $R_b$, the
252: break radius of the surface bightness profile, and $\gamma$.
253: When calculated for the deprojected galaxies in our sample,
254: $M_{\rm dyn}$ was consistent within the scatter with $\mdef$.
255:
256: \begin{figure}
257: \includegraphics[width=8.0cm]{figure2.eps}
258: \caption{Mass deficit vs. BH mass for three different values of $\gamma_0$,
259: the assumed logarithmic slope of the density cusp before energy input
260: from BHs.
261: (a) $\gamma_0=2$; (b) $\gamma_0=1.75$; (c) $\gamma_0=1.5$.
262: Filled circles are elliptical galaxies and empty circles are lenticulars.
263: Squares show dynamical estimates of $\mdef$ for the sample
264: of galaxies considered in \citet{mme01}.
265: Solid line is the $\mdef=\mh$ relation.
266: Dotted line in panel (a) is a linear regression
267: fit to all galaxies (see Table \ref{tab_slopes}).}
268: \label{fig_data}
269: \end{figure}
270:
271: For $\gamma_0=2$,
272: the mass deficits are clustered about a linear relation defined by
273: $\langle \log (\mdef/\mh) \rangle = 0.92$, $1.0$, and
274: $0.65$, respectively, for Es and S0s; Es only; and S0s only,
275: corresponding to $\mdef\sim (8.4,10,4.5)\mh$.
276: Decreasing $\gamma_0$ decreases $\mdef$ (cf. Fig. 1b) and $\mdef$
277: becomes negative/undefined in galaxies when the minimum pointwise
278: slope $\gamma_{\rm min}$ approaches $\gamma_0$.
279: We do not cite values of $\langle\log(\mdef/\mh)\rangle$ for these low values
280: of $\gamma_0$ since the mean depends strongly on which galaxies
281: are defined as having ``cores.''
282: However Figures \ref{fig_data}b,c shows that $\mdef$ remains
283: of order $\mh$ or greater
284: for many of the galaxies even when $\gamma_0<2$.
285: We speculate that the lower mean value of $M_{\rm def}$ for the
286: S0s may indicate a role for gaseous dissipation in the
287: re-formation of cusps following mergers. We emphasize, however,
288: that the mass deficit of lenticulars alone appears to be completely
289: uncorrelated with the BH mass.
290:
291: Table \ref{tab_slopes} lists linear regression fits to the data
292: that were carried out using the routine of \citet{akb96}.
293: For the \citet{res01} ellipticals,
294: as well as for the complete sample, the fitted power-law indices
295: of the $\mdef-\mh$ relation calculated with
296: $\gamma_0=2.0$ and $1.75$ are statistically consistent with unity.
297: To test the sensitivity of the fitting parameters to the assumed
298: power-law index $\alpha$ of the $\mh-\sigma$ relation
299: (equation \ref{eq_msigma}),
300: we recomputed the fits with the \citet{geb00} value $\alpha=3.75\pm0.3$
301: in addition to the \citet{fem00} value $\alpha=4.85\pm0.5$
302: that is standard throughout the current paper. The effect of changing to
303: a shallower version of the $\mh-\sigma$ relation is
304: a steepening of the $\mdef-\mh$ relation,
305: from $d\log\mdef/d\log\mh\approx 0.91$ to $d\log\mdef/d\log\mh\approx 1.16$.
306:
307: We note that in most galaxies the break radius defined via
308: $\gamma(r_b)=\gamma_0=2$ is close to the radius at which the
309: density profile exhibits a visual break, i.e., where the curvature is
310: greatest. The break radius, however, is not a good predictor
311: of the mass deficit; in particular, some galaxies with large break radii,
312: $r_b\gg r_\bullet$, with $r_\bullet=G\mh/\sigma^2$ the dynamical
313: radius of the BH, have mass deficits below the mean.
314: We speculate below that these large break radii may be produced
315: by mechanisms other than BH mergers.
316:
317: Although estimation of errors in Figure \ref{fig_data} is difficult,
318: we believe that the scatter is at least partly intrinsic.
319: Uncertainties in $\log\mh$ are due primarily to uncertainties in
320: $\sigma$ and are of order $\log 2$. Uncertainties in $\log\mdef$ are
321: also roughly $\log 2$ based on variances between the redshift-based and SBF
322: distances.
323: A scatter greater than that due to measurement uncertainties would
324: be reasonable given the different merger histories of galaxies
325: with given $\mh$. Similarly, if the progenitors of the galaxies
326: exhibited a range of different $\gamma_0$, this could in itself explain
327: the observed scatter in Figure \ref{fig_data}.
328:
329: \section{Discussion}
330: The mass ejected by a decaying BH binary is
331: \begin{equation}
332: M_{\rm ej} \approx J M_{12} \ln\left({a_h\over a_{gr}}\right)
333: \end{equation}
334: where $M_{12}=M_1+M_2$ is the binary mass,
335: $a_h$ is the semi-major axis when the binary first becomes hard, and
336: $a_{gr}$ is the separation at which the rate of energy loss to gravitational
337: radiation equals the rate of energy loss to the stars \citep{qui96}.
338: $J$ is a dimensionless mass-ejection rate;
339: for equal-mass binaries, $J\approx 0.5$ \citep{mme01}.
340: \citet{qui96} claims that $J$ is nearly independent of $M_2/M_1$
341: even for extreme mass ratios, implying that a mass in stars
342: of order $M_1+M_2$ is ejected during {\it every} accretion event,
343: even when $M_2$ is tiny compared with $M_1$.
344: This non-intuitive result is due to Quinlan's ejection criterion,
345: which for $M_2\ll M_1$ includes stars that would not have gained
346: sufficient energy to escape from the binary.
347: If instead we equate the change in energy of the binary
348: with the energy carried away by stars that are ejected
349: with $v\gap V_{bin}$,
350: we find $M_{ej}\approx M_2$.
351: This argument suggests a relation
352: \begin{equation}
353: M_{\rm ej} \approx M_2 \ln\left({a_h\over a_{gr}}\right)
354: \end{equation}
355: which is consistent with equation (5) when $M_1\approx M_2$.
356:
357: Using equation (6),
358: and adopting Merritt's (2000) semi-analytic model
359: for decay of a binary in a power-law cusp, we find
360: \begin{equation}
361: {a_{gr}\over a_h}\approx A\left|\ln A\right|^{0.4}, \ \ \
362: A \approx 7.5\left({M_1\over M_2}\right)^{0.2}{\sigma\over c}
363: \end{equation}
364: and
365: \begin{equation}
366: M_{ej}\approx 4.6 M_2\left[1 + 0.043\ln\left({M_2\over M_1}\right)\right], \ \ \ M_1\le M_2.
367: \label{eq_mej}
368: \end{equation}
369: Thus $M_{ej}/M_2$ varies only negligibly with $M_1/M_2$.
370: Henceforth we adopt $M_{ej}\approx 5M_2$.
371:
372: If a BH grows by sequential accretion of smaller BHs, this
373: result implies a mass deficit of order five times the
374: final BH mass.
375: However if the BH grows via a merger hierarchy of comparably-massive
376: BHs, we expect $M_{def}$ to be larger.
377: The idea here is that the damage done to cusps is {\it cumulative}:
378: a merger of two galaxies whose cusps had previously been destroyed
379: by binary BHs, will produce a shallower profile than a merger
380: between two galaxies with initially steep cusps, even if the final
381: BH mass is the same.
382: Galaxies with masses $M\gap 10^{11}M_{\odot}$,
383: including most of the galaxies plotted on Figure \ref{fig_data},
384: are believed
385: to have undergone at least one major merger since a redshift of
386: $1$ (e.g. Kauffmann, Charlot \& Balogh 2001). Thus we predict $\mdef\gap 5\mh$,
387: consistent with Figure \ref{fig_data} if $\gamma_0\gap1.5$.
388:
389: Our interpretation of the mass deficit depends critically on the
390: assumption that all of the change in $\gamma$ during a merger
391: can be attributed to the BHs,
392: i.e., that cusp slopes remain unchanged the absence of BHs.
393: This is known to be the case in equal-mass mergers between galaxies with
394: power-law cusps \citep{bar99, mme01},
395: though in mergers with extreme mass ratios, features can appear in the
396: density profile that are not due to BHs \citep{mec01}.
397: We speculate that the break radii in some of the galaxies in our sample
398: may be due to this process, particularly those galaxies (e.g., NGC~3640, 4168)
399: where $r_b$ greatly exceeds the radius of gravitational influence of
400: the BH.
401: $N$-body simulations of cumulative mergers with unequal BH masses
402: will be needed to assess this hypothesis.
403:
404: Our model presents an interesting contrast to
405: that of \citet{vdm99}, who proposed that cores
406: (in the sense of constant-density regions) were present
407: in all galaxies ab initio, and that power-law cusps were
408: generated by the growth of the BHs -- roughly the opposite
409: of our model in which BHs destroy pre-existing cusps. Van der Marel
410: assumed that core mass correlated initially with bulge
411: luminosity as $M_{\rm core}\sim L^{1.5}$ and that
412: $M_{\bullet}\propto L$; hence $M_{\rm core}\propto M_{\bullet}^{1.5}$,
413: consistent with the correlation in Figure \ref{fig_data} if we identify
414: $M_{\rm core}$ with $\mdef$. We believe that this agreement is coincidental.
415: Van der Marel's model relates core mass to BH mass via an ad hoc
416: postulate, while the model discussed here contains a mechanism for
417: core formation.
418: Van der Marel also ignored the effects of mergers.
419: Nevertheless, van der Marel's model shows that our interpretation
420: is not unique.
421:
422: If stellar cusps are destroyed by binary BHs, the same should
423: be true of dark-matter cusps, like those predicted in
424: CDM theories of structure formation (e.g., \citet{nfw96, moo98, bul01}).
425: Destruction of dark matter cusps could be very efficient if
426: supermassive BHs were present in dark matter halos
427: at large redshifts (e.g. \citet{fan01, hal01, mhn01})
428: due to the cumulative effect mentioned above;
429: furthermore, dark matter cusps would not be regenerated the way
430: that stellar cusps might be via star formation.
431: If our model for the formation of stellar cores is correct,
432: we would predict tha cores of non-interacting CDM
433: should be about as
434: large as those observed in the stars, and perhaps larger.
435:
436: We thank L. Ferrarese for her generous help with the data
437: reduction and interpretation. The image of M87 used for
438: calculating the mass deficit of that galaxy (Section \ref{sec_data})
439: was kindly provided
440: by P. C\^ot\'e and A. Jord\'an in advance of publication.
441: This work was supported by NSF grant AST 00-71099 and NASA grants
442: NAG5-6037 and NAG5-9046 to DM.
443:
444: \bibliographystyle{mn2e.bst}
445: \bibliography{cores.bbl}
446: %end bibliography
447:
448:
449: \begin{table}
450: \begin{tabular}{lcccccc}
451: Galaxy &
452: $D$ &
453: $M_B$ &
454: $r_b$ &
455: $\gamma_{\rm min}$ &
456: $\log \mh$ &
457: $\log \mdef$ \\
458: \hline
459: NGC~2549 & 12.6 & -18.5 & 0.22 & 1.59 & 7.91 & 8.46 \\
460: NGC~2634 & 33.4 & -19.9 & 0.27 & 1.73 & 7.90 & 8.62 \\
461: NGC~2986 & 26.8 & -20.5 & 0.41 & 0.99 & 8.79 & 9.51 \\
462: NGC~3193 & 34.0 & -20.9 & 0.23 & 0.84 & 8.15 & 9.23 \\
463: NGC~3348 & 38.5 & -21.1 & 0.49 & 0.75 & 8.47 & 9.79 \\
464: NGC~3414 & 25.2 & -20.1 & 0.14 & 1.61 & 8.71 & 8.66 \\
465: NGC~3613 & 29.1 & -20.7 & 0.52 & 0.80 & 8.20 & 9.39 \\
466: NGC~3640 & 27.0 & -21.0 & 0.68 & 0.94 & 7.82 & 9.51 \\
467: NGC~4121 & 28.3 & -17.9 & 0.12 & 1.76 & 7.06 & 7.94 \\
468: NGC~4128 & 32.4 & -19.8 & 0.18 & 1.64 & 8.32 & 8.83 \\
469: NGC~4168 & 30.9 & -20.4 & 1.25 & 1.01 & 7.90 & 9.58 \\
470: NGC~4291 & 26.2 & -19.8 & 0.17 & 0.49 & 8.28 & 9.22 \\
471: NGC~4365 & 20.4 & -21.1 & 0.57 & 0.71 & 8.50 & 9.79 \\
472: NGC~4472 & 16.3 & -21.7 & 0.63 & 0.90 & 8.74 & 10.03 \\
473: NGC~4473 & 15.7 & -19.9 & 0.24 & 1.22 & 7.90 & 9.25 \\
474: NGC~4478 & 18.1 & -19.1 & 0.30 & 1.42 & 7.55 & 8.75 \\
475: NGC~4486 & 16.1 & -21.5 & 1.12 & 0.74 & 9.55 & 10.49 \\
476: NGC~4503 & 17.6 & -19.2 & 0.25 & 1.56 & 7.11 & 8.70 \\
477: NGC~4564 & 15.0 & -18.9 & 0.11 & 1.69 & 7.76 & 8.22 \\
478: NGC~4589 & 22.0 & -20.0 & 0.18 & 1.05 & 8.22 & 8.78 \\
479: NGC~5077 & 34.0 & -20.4 & 0.54 & 1.15 & 8.78 & 9.51 \\
480: NGC~5198 & 34.1 & -20.0 & 0.11 & 0.92 & 8.09 & 8.63 \\
481: NGC~5308 & 26.6 & -19.7 & 0.10 & 1.84 & 8.37 & 8.01 \\
482: NGC~5370 & 41.3 & -19.0 & 0.32 & 1.56 & 7.49 & 8.63 \\
483: NGC~5557 & 42.5 & -21.2 & 0.44 & 0.87 & 8.64 & 9.62 \\
484: NGC~5576 & 25.5 & -20.3 & 0.24 & 1.37 & 8.00 & 9.19 \\
485: NGC~5796 & 36.5 & -20.7 & 0.18 & 1.18 & 8.60 & 9.15 \\
486: NGC~5812 & 26.9 & -20.3 & 0.25 & 1.71 & 8.15 & 8.83 \\
487: NGC~5813 & 32.2 & -21.1 & 0.43 & 0.31 & 8.40 & 9.66 \\
488: NGC~5831 & 27.2 & -19.9 & 0.22 & 1.42 & 7.72 & 8.67 \\
489: NGC~5898 & 29.1 & -20.4 & 0.33 & 1.34 & 8.30 & 9.20 \\
490: NGC~5903 & 33.9 & -20.9 & 0.78 & 0.77 & 8.41 & 9.57 \\
491: NGC~5982 & 39.3 & -20.9 & 0.45 & 0.49 & 8.71 & 9.58 \\
492: NGC~6278 & 37.1 & -19.7 & 0.10 & 1.53 & 7.64 & 8.59 \\
493: UGC~4551 & 23.6 & -18.7 & 0.24 & 1.36 & 8.02 & 8.87 \\
494: \hline
495: \end{tabular}
496: \caption{Galaxies with $\gamma_{\rm min}\leq 2$;
497: $D$ is distance in Mpc; $r_b$ is break radius in kpc;
498: $\gamma_{\rm min}$ is the minimum logarithmic slope;
499: $\mh$ and $\mdef$ are in solar masses.}
500: \label{tab_galaxies}
501: \end{table}
502:
503: \begin{table}
504: \begin{tabular}{ccccc}
505: $\gamma_0$ &
506: E &
507: S0 &
508: $\alpha=4.8$ &
509: $\alpha=3.75$ \\
510: \hline
511: $2.00$ & $0.93\pm0.10$ & $-0.10\pm0.14$ & $0.91\pm0.09$ & $1.16\pm0.12$ \\
512: & $(10.5\pm1.8)$ & $(3.8\pm 0.86)$ & $(9.75\pm1.78)$ & $(10.7\pm1.7)$ \\
513: $1.75$ & $1.07\pm0.15$ & $-0.78\pm0.46$ & $0.86\pm0.24$ & $1.10\pm0.30$ \\
514: & $(2.75\pm 0.78)$ & $(0.35\pm0.19)$ & $(2.17\pm0.50)$ & $(2.36\pm0.56)$ \\
515: $1.50$ & $1.56\pm0.37$ & & $1.58\pm0.35$ & $2.02\pm 0.45$ \\
516: & $(0.49\pm0.26)$ & & $(0.46\pm0.23)$ & $(0.55\pm0.25)$ \\
517: \hline
518: \end{tabular}
519: \caption{Linear regression fits to the
520: ($\log\mh,\log\mdef$) data for three
521: values of the fiducial logarithmic slope $\gamma_0$.
522: Values in parentheses are
523: $\mdef/\mh$ interpolated from
524: the fit at $\mh=10^8M_\odot$. Fourth and fifth columns are, respectively,
525: fits of the entire data set (including the galaxies
526: with dynamical estimates of $\mdef$) using the \citet{fem00}
527: and the \citet{geb00} values of
528: the $\mh-\sigma$ relation exponent $\alpha$.}
529: \label{tab_slopes}
530: \end{table}
531:
532: \end{document}
533:
534:
535:
536:
537:
538:
539:
540:
541:
542:
543:
544:
545:
546: