1: % mn2esample.tex
2: %
3: % v2.1 released 22nd May 2002 (G. Hutton)
4: %%
5: %\documentclass[useAMS,usenatbib]{mn2e}
6: \documentclass[usenatbib]{mn2e}
7: \usepackage{times}
8: \usepackage{epsfig}
9:
10: %
11: % The usenatbib command allows the use of Patrick Daly's natbib.sty for
12: % cross-referencing.
13: %
14: % If you wish to typeset the paper in Times font (if you do not have the
15: % PostScript Type 1 Computer Modern fonts you will need to do this to get
16: % smoother fonts in a PDF file) then uncomment the next line
17: % \usepackage{Times}
18:
19: %%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
20: % Personal Macros created by HGK
21: \newcommand{\eg}{e.$.\!$g.\ }
22: \newcommand{\be}{\begin{equation}}
23: \newcommand{\ee}{\end{equation}}
24: \newcommand{\etal}{et al.\ }
25: \newcommand{\chandra}{{\it Chandra} }
26: \newcommand{\xmm}{{\it XMM-Newton} }
27: \newcommand{\ciao}{{\it CIAO} }
28: \newcommand{\tbc}{ TO BE COMPLETED\ }
29: \newcommand{\tbd}{ TO BE DECIDED\ }
30: \newcommand{\ie}{i$.\!$e$.\!$, }
31: \newcommand{\hubble}[1]{H_0={#1}\unit{km\,sec^{-1}\,Mpc^{-1}}}
32: \newcommand{\hubbleunit}{\unit{km\,sec^{-1}\,Mpc^{-1}}}
33: \newcommand{\ki}{$\chi^2$\,}
34: \newcommand{\qz}{q_0}
35: \newcommand{\labequn}[1]{\label{eq:#1}}
36: \newcommand{\labtab}[1]{\label{tab:#1}}
37: \newcommand{\labfig}[1]{\label{fig:#1}}
38: \newcommand{\labsecn}[1]{\label{sec:#1}}
39: \newcommand{\labsubsecn}[1]{\label{subsecn:#1}}
40: \newcommand{\mean}[1]{\left\langle#1\right\rangle}
41: \newcommand{\equn}[1]{Equation~\ref{eq:#1}}
42: \newcommand{\tab}[1]{Table~\ref{tab:#1}}
43: \newcommand{\fig}[1]{Figure~\ref{fig:#1}}
44: \newcommand{\secn}[1]{Section~\ref{sec:#1}}
45: \newcommand{\subsecn}[1]{subsection~\ref{subsecn:#1}}
46: \newcommand{\hi}{$h^{-1}$~}
47: \newcommand{\kms}{~km~s$^{-1}$}
48: \newcommand{\kmsm}{{\rm \ km\> s^{-1}}}
49: \newcommand{\nid}{\noindent}
50: \newcommand{\cen}{\centerline}
51: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
52: \title[Fossil groups and formation of BCGs]
53: {The central elliptical galaxy in fossil groups and formation
54: of BCGs}
55: \author[Khosroshahi, Ponman \& Jones]{
56: Habib G. Khosroshahi\thanks{E-mail:
57: habib@star.sr.bham.ac.uk (HGK)}, Trevor. J. Ponman \& Laurence R. Jones \\
58: School of Physics and Astronomy, The University of Birmingham,
59: Birmingham B15 2TT, UK}
60:
61: \begin{document}
62:
63: \date{Accepted, Received}
64:
65: \pagerange{\pageref{firstpage}--\pageref{lastpage}} \pubyear{2004}
66:
67: \maketitle
68:
69: \label{firstpage}
70:
71: \begin{abstract}
72:
73: We study the dominant central giant elliptical galaxies in ``Fossil
74: groups'' using deep optical (R-band) and near infrared (Ks-band)
75: photometry. These galaxies are as luminous as the brightest cluster
76: galaxies (BCGs), raising immediate interest in their link to the
77: formation of BCGs and galaxy clusters. However, despite apparent
78: similarities, the dominant fossil galaxies show non-boxy isophotes, in
79: contrast to the most luminous BCGs. This study suggests that the
80: structure of the brightest group galaxies produced in fossil groups
81: are systematically different to the majority of BCGs. If the fossils
82: do indeed form from the merger of major galaxies including late-types
83: within a group, then their disky nature is consistent with the results
84: of recent numerical simulations of semi-analytical models which
85: suggest that gas rich mergers result in disky isophote ellipticals.
86:
87: We show that fossils form a homogeneous population in which the
88: velocity dispersion of the fossil group is tightly correlated with the
89: luminosity of the dominant elliptical galaxy. This supports the
90: scenario in which the giant elliptical galaxies in fossils can grow to
91: the size and luminosity of BCGs in a group environment. However, the
92: boxy structure of luminous BCGs indicate that they are either not
93: formed as fossils, or have undergone later gas-free mergers within the
94: cluster environment.
95:
96: \end{abstract}
97:
98: \begin{keywords}
99: galaxies: clusters: general - galaxies: elliptical - galaxies: haloes -
100: intergalactic medium - X-ray: galaxies - X-rays: galaxies: clusters
101: \end{keywords}
102:
103: \section{Introduction}
104:
105: It is believed that most of the large and luminous elliptical galaxies
106: have formed via mergers of disk galaxies
107: \citep{toomre72,searle73}. This has been suggested by morphology density
108: relation \citep{dressler80}, the observed frequency of merging
109: galaxies at high redshift and also extensively in computer simulations
110: \citep{barnes89}. Many of these luminous
111: ellipticals ($M_B\le$ -21) are found in rich galaxy clusters. This,
112: however, does not imply that they are formed in cluster
113: environment. The brightest cluster galaxies (BCGs) are of special
114: interest as they reside close to the centroid of cluster X-ray
115: emission \citep{jf84}, and the centre of the dark matter distribution
116: in clusters, as inferred from gravitational lensing \citep{smith05} --
117: implying that they lie at the minimum of the cluster potential
118: well. They also show various correlations with cluster properties.
119:
120: In general, two main formation modes could be assumed for the
121: hierarchical formation of BCGs according to their formation
122: environment: {\bf 1)} BCGs formed in the high velocity environment of
123: clusters. Although the effectiveness of dynamical friction in bringing
124: individual galaxies to the cluster centre via orbital decay will be
125: reduced by the high velocity dispersion, infalling groups will still
126: suffer rapid orbital decay if they survive long enough, and can then
127: deposit their brightest galaxies in the cluster core, where they can
128: merge and form a bright elliptical galaxy
129: \citep{mohr04,hausman78}. {\bf 2)} BCGs are formed in the low velocity
130: environment of groups, where dynamical friction causes the orbits of
131: individual galaxies to decay, resulting in the merger of all large
132: galaxies, if the group forms early and is left undisturbed for a
133: sufficiently long period \citep{dubinski98,ponman94,jones03}. The group
134: containing this `ready made' BCG then provides the nucleus around which
135: a cluster forms.
136:
137: Elliptical galaxies show fine structures and are more complex than
138: originally thought. These structures take the form of hidden
139: disks, shells and bars, departures from pure elliptical isophotes
140: \citep{bender88} and variations in radial surface brightness profiles
141: \citep{kwkm00}, some of which are found to be environment
142: dependent \citep{habib04}. BCGs formed in the above two modes should
143: display several observational signatures of how they formed.
144: A useful probe is provided by galaxy morphology --
145: isophotal shapes, radial surface brightness profiles and the presence
146: or not of multiple-nuclei. Different star forming histories are also
147: expected. Some of these studies require space
148: resolution data, but some can be studied using ground based
149: observations, including the isophotes of elliptical galaxies.
150:
151: Based on their isophotal shapes, elliptical galaxies can be classified as
152: disky or boxy \citep{bender88}. Low mass ellipticals, which are fast
153: rotating, are usually disky isophote galaxies with positive
154: fourth-order Fourier coefficient $B_4>0$. Some disky-isophote
155: ellipticals might contain faint disks similar to S0 galaxies
156: \citep{scorza95}.
157: With negative $B_4$, boxy isophote ellipticals are
158: less rotationally supported. They generally contain flat cores
159: \citep{faber97,laine03} and show complex internal kinematics
160: \citep{emsellem04}. The observations of \citet{rest01} show that it
161: is very unlikely to find disky ellipticals which are also core galaxies.
162: The majority of the BCGs are found to be core galaxies \citep{laine03}.
163: The distinct observed properties of disky and
164: boxy isophotes of elliptical galaxies, such as their radio properties
165: show that they are more than just an artifact of viewing angle or the
166: projection on the plane of sky \citep{bender89}.
167:
168: It is important to understand the origin of isophotal shapes before
169: they can be used to trace the formation history of ellipticals.
170: \citet{naab03} performed a large survey of dissipationless
171: merger simulations of disk galaxies and found that unequal-mass
172: 3:1 to 4:1 mergers lead to fast rotating disky ellipticals while
173: equal-mass 1:1 to 2:1 mergers produce slowly rotating, pressure-supported
174: ellipticals.
175: However, \citet{khochfar05} showed that the above scenario is not
176: able to reproduce the observation that the fraction of boxy and disky
177: ellipticals depends on galaxy luminosity. They argued that equal-mass
178: mergers lead to boxy ellipticals and unequal-mass mergers produce
179: disky ellipticals. However, major mergers between bulge-dominated
180: galaxies result in boxy ellipticals, independent of the mass ratio,
181: while merger remnants that subsequently accrete gas, leading to a
182: secondary stellar disk with more than 20 per cent of the total stellar
183: fraction, are always disky. More recently they showed that mergers of
184: spiral galaxies alone cannot reproduce the kinematic and photometric
185: properties of very massive elliptical galaxies \citep{naab06}, nor can
186: they reproduce the observed correlation between isophotal shapes and the
187: luminosity of ellipticals.
188:
189: Here we study the isophotes of brightest group galaxies (BGGs)
190: in fossils. A brief introduction to fossils is given in section 2
191: where we also describe the sample and observations. Our analysis and
192: the results are presented in section 3. A discussion and concluding
193: remarks are in section 4. We assume $H_0=70$ km~s$^{-1}$~Mpc$^{-1}$ and
194: $\Omega_m=0.3$ with cosmological constant $\Omega_\Lambda=0.7$
195: throughout.
196:
197: \begin{figure}
198: \center
199: \epsfig{file=new-j1119.ps,width=1.3in,height=1.2in}
200: \epsfig{file=new-j1256.ps,width=1.3in,height=1.2in}
201: \epsfig{file=new-j1331.ps,width=1.3in,height=1.2in}
202: \epsfig{file=new-j1340.ps,width=1.3in,height=1.2in}
203: \epsfig{file=new-j1416.ps,width=1.3in,height=1.2in}
204: \epsfig{file=new-j1552.ps,width=1.3in,height=1.2in}
205: \epsfig{file=new-n6482.ps,width=1.3in,height=1.1in}
206: \caption{Ellipse fits to the surface brightness distribution of the
207: central galaxy in fossil groups. All the images are from the Ks-band
208: observations with the exception of the most distant fossil, RX J1256.0+2556,
209: which is observed in R-band. 5 arcsec scale bars are also shown.}
210: \label{chansoft}
211: \end{figure}
212:
213: \section{Fossil galaxy groups}
214:
215: In the class of galaxy groups known as ``fossil groups'',
216: the group is dominated optically by a single luminous elliptical
217: galaxy at the centre of extended luminous X-ray emission similar to
218: that seen in bright X-ray groups. The X-ray emission in fossils is
219: regular and symmetric, indicating the absence of recent group
220: merging. The dominant giant elliptical galaxy has an
221: optical luminosity similar to BCGs. A cD galaxy has also been reported
222: in a fossil group \citep{mendes05}. The observed
223: properties of fossils, and their absence of $L^\star$ galaxies, suggest
224: that they must be old galaxy groups. These properties are discussed in
225: recent studies \citep{kmpj06,kpj06} where we report a higher dark
226: matter concentration in fossils, compared to non-fossils groups and
227: clusters with similar masses, which is consistent with an early
228: formation epoch.
229:
230: Observationally a galaxy group is classified as a fossil if \citep{jones03} it
231: has an X-ray luminosity of $L_{X,bol} \geq 10^{42} h^{-2}_{50}$ erg
232: s$^{-1}$ spatially extended to few 100 kpc, and the dominant galaxy
233: is at least 2 magnitudes brighter (in R-band) than the second ranked galaxy
234: within half the projected virial radius of the group.
235: The X-ray criterion guarantees
236: the existence of a group size galaxy halo while the optical criterion
237: assures that the $M^\star$ galaxies are absent within the given radius
238: which corresponds to the radius for orbital decay by dynamical friction
239: \citep{binney87}. No upper limit is
240: placed on the X-ray luminosity or temperature, and recently a fossil galaxy
241: cluster was found \citep{kmpj06}.
242:
243: \subsection{The sample}
244:
245: This study makes use of a flux-limited sample of fossils found in
246: the catalogue of spatially extended X-ray sources compiled by
247: Wide Angle ROSAT Pointed Survey (WARPS) project. Details of the
248: fossil identification and sample selection is given in \citet{jones03}.
249: This is the largest statistical sample
250: of fossil groups studied to date. In addition, the nearest known fossil
251: group NGC 6482 and the first discovered fossil group, RX J1340.5+4017, are
252: included in our sample. The detailed X-ray analysis
253: of the sample is the subject of a separate study \citep{kpj06}.
254:
255: \subsection{Optical and near-IR observations}
256:
257: The above sample was observed using the observational facilities of
258: Issac-Newton Group of Telescopes (ING) and Kitt-peak National
259: Observatory (KPNO). R-band images were obtained using the INT 2.5m wide
260: field imager. Unfortunately the conditions were not photometric, and so
261: further R-band imaging was obtained, in photometric conditions, with the
262: 8k mosaic camera at the University of Hawaii 2.2-m telescope, and in INT
263: wide-field camera service time. The resultant photometric accuracy
264: for all the systems is $\le$0.05 mag. R-band
265: observation of NGC 6482 was performed with the KPNO-0.9m in April 2005.
266: Spectroscopic observations of the sample were also obtained, using slit
267: spectroscopy on the KPNO 4m, to examine the optical membership
268: of these groups. This is discussed in \citet{kpj06}.
269:
270: $K_s$-band observations of the sample using UIST/UKIRT were performed
271: in 2004. The seeing was measured to be $\sim 1.0''$. The
272: data were reduced using the ORAC data reduction package
273: (http://www.oracdr.org/).
274: Where multi-snaps were taken, the images were co-added to increase the
275: signal-to-noise. Figure \ref{chansoft} shows the $K_s$ images and their
276: ellipse fits (section 3.3) for the central fossil galaxies in the sample
277: with the exception of RX J1256.0+2556 for which only the R-band data was
278: available.
279:
280: \section{Analysis and results}
281:
282: Elliptical galaxies are usually single-component galaxies with
283: radial surface brightness profiles described by a de Vaucouleurs
284: law ($r^{1/4}$). It has been shown that the Sersic profile ($r^{1/n}$)
285: gives a better fit, in general, for ellipticals with a wide luminosity range
286: and in different environments (Trujillo \etal 2001; Khosroshahi \etal 2004). Ellipticals
287: are also divided based on their isophotal shapes. For this study we
288: concentrate on the radial surface brightness profile,
289: the ellipticity profile and the fourth order Fourier coefficient
290: ($B_4$), which is an indicator of boxy and disky isophotes.
291: The analysis is performed on both the R-band and $K_s$-band images using
292: the well-known IRAF/ellipse task.
293:
294: \subsection{Radial surface brightness profile}
295:
296: The Sersic profile, was used to model the R-band surface
297: brightness distribution of the giant elliptical galaxies, using a two
298: dimensional bulge/disk decomposition method. The values of $n$ and the
299: half-light radius, along with other parameters, are given in Table
300: 1. The analysis shows that the fossil central galaxies are best
301: modelled with $<n>= 4.1 \pm 0.7$. While this agrees in general with
302: the surface brightness profiles of the remnants of collisionless
303: disk mergers, these simulations are not able to produce galaxies
304: as large as the dominant fossil galaxies \citep{naabtru06}.
305:
306:
307: Our ground-based observations are inadequate for probing the radial
308: surface brightness profiles within the central $\sim 1$ kpc, which is
309: necessary for a power-law/core classification.
310: As a result we limit our investigation to the Sersic fit
311: to the galaxy.
312:
313: \begin{figure}
314: \center
315: \epsfig{file=isophot-j1119b.ps,width=3.2in,height=0.8in}
316: \epsfig{file=isophot-j1256b.ps,width=3.2in,height=0.8in}
317: \epsfig{file=isophot-j1331b.ps,width=3.2in,height=0.8in}
318: \epsfig{file=isophot-j1340b.ps,width=3.2in,height=0.8in}
319: \epsfig{file=isophot-j1416b.ps,width=3.2in,height=0.8in}
320: \epsfig{file=isophot-j1552b.ps,width=3.2in,height=0.8in}
321: \epsfig{file=isophot-n6482b.ps,width=3.2in,height=0.8in}
322: \caption{Profiles of radial surface brightness, ellipticity and the
323: fourth Fourier coefficient, $B_4$, for
324: RX J1119.7+2126, RX J1256.0+2556, RX J1331.5+1108, RX J1340.5+4017,
325: RX J1416.4+2315, RX J1552.2+2013 and NGC 6482. The profiles extracted from
326: R-band and $K_s$-band data are shown in dark (black) and grey (red),
327: respectively.}
328: \label{isophots}
329: \end{figure}
330:
331: \subsection{Ellipticity profile}
332:
333: The ellipticity profiles presented in Fig \ref{isophots} show a
334: general pattern in which the ellipticity of the isophotes increases
335: with the radius, with the exception of RX J1552.2+2013. Galaxies with
336: high quality data show
337: ellipticity increasing to 0.4-0.6 exceeding the ellipticity of the
338: X-ray halo which is usually less than 0.3 \citep{buote96}. The
339: central galaxy in fossils are
340: aligned with the underlying dark matter confirmed both in X-ray and
341: lensing studies.
342: Similar alignment has been noted for the BCGs in many clusters\citep{fuller99}.
343:
344:
345: \begin{table*}
346: \begin{center}
347: \caption{Photometric properties of the sample galaxies.
348: \label{table1}
349: }
350: \begin{tabular}{lllcccccc}
351: \hline
352: Group & R.A. & Dec. & z & M$_R$ & $a_4/a$ & $n$ & $r_e$ & kpc/arcsec\\
353: & (J2000) &(J2000) & & mag & $\times100$ & & arcsec \\
354: \hline
355: RX J1119.7+2126 & 11:19:43.6& +21:26:51 & 0.061 &-22.1 & 0.1 & 5.1 & 8.9 & 1.14\\
356: RX J1256.0+2556 & 12:56:03.4 & +25:56:48 & 0.232 &-24.1 & 0.1 & 3.1 & 7.3 & 3.73\\
357: RX J1331.5+1108 & 13:31:30.2 & +11:08:04 & 0.081 &-22.9 & 0.3 & 4.3 & 7.2 & 1.53\\
358: RX J1340.5+4017$^a$ & 13:40:33.4 & +40:17:48 & 0.171 &-23.0 & irr & 4.2 & 7.6 & 2.92\\
359: RX J1416.4+2315 & 14:16:26.9 & +23:15:32 & 0.137 &-24.3 & 0.7 & 3.6 & 12.1 &2.44\\
360: RX J1552.2+2013 & 15:52:12.5 & +20:13:32 & 0.135 &-24.0 & 0.5 & 4.6 & 15.8 &2.40\\
361: NGC 6482$^b$ & 15:52:12.5 & +20:13:32 & 0.013 &-22.9 & 1.3 & 3.8 & 16.0 &0.26\\
362:
363: \hline
364: \end{tabular}
365: \end{center}
366: $^a$ This system is the first confirmed fossil group and not part of the
367: flux-limited sample of fossils. $^b$ This group is known to be the nearest fossil system
368: \citep{kjp04} and is not part of the flux-limited sample.
369: \end{table*}
370:
371: \subsection{Isophotal analysis}
372:
373: Fig \ref{isophots} shows the results of the ellipse fits.
374: In order to quantify the shape
375: of the isophotes and to be able to make a direct comparison with
376: similar analyses in the literature, we calculate $a_4/a$, which is based on
377: the measured fourth Fourier coefficient, $B_4$ \citep{jorgensen99}.
378: Similarly to \citet{bender89}, $a_4/a=\frac{\sqrt{1-\epsilon}B_4}
379: {adI/da}$ is quantified at its peak value. In the absence of a peak
380: the $a_4/a$ is quantified at $r_e$.
381: Here $\epsilon$, $a$ and $I$ are the ellipticity, semi-major axis length
382: and the surface brightness of the isophotes, respectively.
383: As seen in Fig.\ref{isophots} none of the galaxies have predominantly
384: boxy isophotes, except in the outskirts were the statistics are
385: very poor and the values of $B_4$ are not well constrained.
386:
387: \begin{figure}
388: \center
389: \epsfig{file=magbgga4.ps,width=2.6in,height=2.2in}
390: \caption{The variation in isophotal shapes of early-type
391: brightest group or cluster galaxies, and
392: those in fossil groups (crosses), with the optical luminosity of the brightest
393: galaxy. The comparison sample is a combination of early-type
394: BGGs (triangles) and BCGs (circles) from \citet{ellis06}, for
395: which $a_4/a$ values were available.}
396: \label{a4}
397: \end{figure}
398:
399: We compare the values of $a_4/a$ with those of the brightest galaxies in
400: groups and clusters, BGGs and BCGs (Fig \ref{a4}). It is clear from this
401: plot that none of the fossil's dominant galaxies have prominent boxy
402: isophotes. Indeed some, including the nearest fossil, NGC 6482 \citep{kjp04},
403: and the fossil cluster, RX J1416.4+2315 \citep{kmpj06}, are highly disky
404: isophote galaxies. An earlier study \citep{faber97} gives an even higher
405: value for the diskyness of NGC 6482. The comparison sample is a
406: combination of early-type BGGs and BCGs from \citet{ellis06} for
407: which $a_4/a$ values were available from earlier studies and therefore
408: it is not a complete sample.
409:
410:
411: \section{Discussion and conclusions}
412:
413: This analysis shows that, despite apparent similarities, the dominant
414: giant elliptical galaxy in fossil groups are different in their
415: isophotal shapes, compared to the brightest central
416: galaxies in non-fossil systems, especially in rich clusters. Luminous
417: elliptical galaxies in non-fossil groups and clusters do not
418: present disky isophotes. Less luminous BGGs
419: show disky and boxy isophotes in similar
420: proportions. However, the observed frequency of boxy-isophote dominant fossil
421: galaxies is apparently zero.
422:
423: If the central galaxies of fossil groups have indeed been formed from the
424: merger of all major galaxies within the inner regions of the group, as
425: we suppose, then some of these mergers would have been gas-rich, as
426: group spirals are incorporated into the central merger-remnant.
427: The disky character of central fossil galaxies is
428: then consistent with the findings of numerical simulations, that
429: disky isophotes result from gas rich mergers.
430: \citet{khochfar05} highlight the importance of the role of gas
431: in galaxy mergers and show that the isophotal shapes of merger
432: remnants are sensitive to the morphology of their progenitors and
433: to subsequent gas infall.
434: In contrast, boxy isophote ellipticals
435: are formed by equal-mass mergers of bulge dominated galaxies. Such
436: mergers are likely to occur at the core of clusters where most of the
437: galaxies are gas poor.
438:
439:
440: Almost 90\% of BCGs studied by \citet{laine03} are core galaxies, ie.
441: with flatter slope near the nucleus in their radial surface brightness
442: profiles. The
443: study by \citet{faber97} shows that a large fraction ($\sim$70\%) of core
444: galaxies have boxy isophotes, with as low as $\sim$10\% with disky isophotes,
445: implying a strong association of boxy isophotes with core galaxies.
446: Taking into account the conversion of cuspy cores into flat
447: low-density cores by black hole merging, \citet{khochfar05} find that
448: disky ellipticals should contain central density cusps whereas boxy
449: ellipticals should in general be characterised by flat cores. Only
450: rare low-luminosity boxy ellipticals, resulting from equal-mass
451: mergers of disk galaxies, could have power-law cores.
452: Space resolution data is needed to study the core of the galaxies to
453: verify the core and power- law property of dominant fossil galaxies.
454: If future observations show that fossil group dominant galaxies are
455: power-law galaxies, as expected from their isophotes in light of the
456: above argument, then they will be the first of such entities to grow
457: to the size of BCGs.
458:
459: \begin{figure}
460: \center
461: \epsfig{file=lbggsigma.ps,width=2.6in,height=2.3in}
462: \caption{A comparison between the correlation of group/cluster
463: velocity dispersion and absolute B-magnitude of the dominant
464: galaxy. The data points from Table 1 (except RX J1119.7+2126) are
465: shown with filled circles. Diamonds represent two of the fossil candidate
466: OLEGs, out of four, in \citet{yoshioka04}.
467: The correlation is much tighter in fossils than in non-fossil
468: groups (circles) and clusters (squares). The comparison non-fossil
469: groups are GEMS X-ray selected groups with early-type BGG and with
470: group scale X-ray emission (G-sample in \citet{osmond04}). The cluster
471: sample is selected from \citet{girardi02} for which the luminosity
472: of the central galaxy was available in \citet{mohr04}.}
473: \label{sigma}
474: \end{figure}
475:
476: The link between fossil galaxy groups and the BCGs is further
477: motivated by observational space density estimates of fossils which is
478: found to be 8\% to 20\% of X-ray luminous systems \citep{jones03} and
479: as large as the space density of poor and rich galaxy clusters combined.
480: This means that there are enough fossils to provide BCGs to clusters.
481: A recent theoretical study \citep{milos06} predicts 5\%-40\% of galaxy
482: groups and 1\%-3\% of galaxy clusters to be fossils. In the context of BCG
483: formation, fossils appear to be suitable environments for the formation of
484: luminous giant ellipticals around which clusters
485: can form. This is a different formation mode to the one in
486: which the BCG forms via mergers of brightest group galaxies during the
487: cluster collapse.
488: Numerical simulations suggest that the isophotes of the BCGs formed in the
489: latter case will be predominantly boxy, characteristic of gas free
490: (``dry'') mergers.
491:
492: The BGGs and BCGs with non-boxy isophotes in Fig.\ref{a4} could easily
493: originate as fossil group central galaxies which have been incorporated
494: into larger structures. The difference in photometric structure seen
495: clearly in many BCGs and BGGS, especially the most luminous ones,
496: does not rule out the possibility that they originated in fossil groups.
497: This could still be the case provided that they have undergone later
498: gas-free mergers within the cluster environment -- for example, with
499: the BGGs of infalling galaxy groups.
500: About 40\% of BCGs contain at least one secondary nucleus\citep{laine03},
501: which strongly suggests the action of late mergers within the
502: cluster environment. If fossils are indeed old and undisturbed systems
503: then they should be found to have a very {\it low} incidence of
504: multiple nuclei.
505:
506: In support of the above argument we show (Fig \ref{sigma}) a tight
507: correlation between the luminosity of the central galaxy in fossils
508: and the underlying gravitational mass probed by the group velocity
509: dispersion. This shows that fossils form a homogeneous population
510: in which the luminosity of the central galaxy is strongly tied to the
511: property of its parent group. This property of the fossils is
512: consistent with their early formation epoch and absence of recent
513: merger. In contrast the large scatter in the distribution of the
514: non-fossil galaxy groups and clusters, on the same plane, is
515: understood to be merger driven. Absence of a recent major merger helps
516: the dominant fossil galaxy preserve its original structure.
517:
518: We conclude that there is high chance that disky
519: BCGs are formed in fossil groups. Boxy BCGs, could still result from
520: progenitor fossils, but would need to have
521: undergone a dry merger within the cluster, probably
522: as a result of merger with BGGs of infalling groups.
523:
524:
525: We would like to thank Doug Burke, for his involvement in the
526: UH 88'' observations, and the INT for the service observations,
527: Ewan O'Sullivan for his involvement in the KPNO observations
528: of NGC 6482, Graham P. Smith for his valuable input
529: which helped to improve the discussion, and Arif Babul for his
530: useful comments.
531:
532: \begin{thebibliography}{99}
533: \bibitem[\protect\citeauthoryear{Barnes}{1989}]{barnes89}
534: Barnes J. E., 1989, Nat, 338, 123
535: \bibitem[\protect\citeauthoryear{Bender}{1988}]{bender88}
536: Bender, R. 1988, A\&A, 193, 7
537: \bibitem[\protect\citeauthoryear{Bender}{1989}]{bender89}
538: Bender R., Surma P., Doebereiner S., Moellenhoff C., Madejsky R.,
539: 1989, A\&A, 217, 35
540: \bibitem[\protect\citeauthoryear{(Bender, Burstein \& Faber}{1992}]{bender92}
541: Bender R., Burstein D., Faber S. M., 1992, ApJ, 399, 462
542: \bibitem[\protect\citeauthoryear{Binney \& Tremaine}{1987}]{binney87}
543: Binney K., Tremaine S., 1987, Galactic Dynamics. Princeton Univ. Press,
544: Princeton, NJ
545: \bibitem[\protect\citeauthoryear{Buote \& Canizares}{1996}]{buote96}
546: Buote D. A., Canizares C. R., 1996, ApJ, 457, 565
547: \bibitem[\protect\citeauthoryear{(Dressler}{1980}]{dressler80}
548: Dressler A., 1980, ApJ, 236, 351
549: \bibitem[\protect\citeauthoryear{Dubinski}{1998}]{dubinski98}
550: Dubinski, J. 1998, ApJ, 502, 141)
551: \bibitem[\protect\citeauthoryear{Ellis \&O'Sullivan}{2006}]{ellis06}
552: Ellis S. C., O'Sullivan E., 2006, MNRAS, 367, 627
553: \bibitem[\protect\citeauthoryear{Emsellem \etal}{2004}]{emsellem04}
554: Emsellem E., Cappellari M., Peletier R. F., McDermid R. M. 2004, MNRAS, 352, 721
555: \bibitem[\protect\citeauthoryear{Faber \etal}{1997}]{faber97}
556: Faber S. M., Tremaine S., Ajhar E. A., Byun Y. \etal, 1997, AJ, 114, 1771
557: \bibitem[\protect\citeauthoryear{Fuller, West \& Bridges}{1999}]{fuller99}
558: Fuller T. M., West J. W., Bridges T. J., 1999, 519, 22
559: \bibitem[\protect\citeauthoryear{Girardi \etal}{2002}]{girardi02}
560: Girardi M., Manzato P., Mezzetti M., Giuricin G., Limboz F., 2002, ApJ, 569, 720
561: \bibitem[\protect\citeauthoryear{Trujillo \etal}{2001}]{trujillo01}
562: Trujillo I., Graham A. W. and Caon N. 2001, MNRAS, 326, 869
563: \bibitem[\protect\citeauthoryear{Hausman \& Ostriker}{1978}]{hausman78}
564: Hausman, M. A., \& Ostriker, J. P. 1978, ApJ, 224, 320
565: \bibitem[\protect\citeauthoryear{Jorgensen \etal}{1999}]{jorgensen99}
566: Jorgensen I., Franx, M., Hjorth, J., van Dokkum P. G., 1999, MNRAS, 308, 833
567: \bibitem[\protect\citeauthoryear{Jones \& Forman}{1984}]{jf84}
568: Jones, C., \& Forman, W. 1984, ApJ, 276, 38
569: \bibitem[\protect\citeauthoryear{Jones \etal}{2003}]{jones03}
570: Jones L. R., Ponman T. J., Horton A., Babul A., Ebeling H., Burke D. J.,
571: 2003, MNRAS, 343, 627
572: \bibitem[\protect\citeauthoryear{Jones \etal}{2000}]{jones00}
573: Jones L. R., Ponman T. J., Forbes D.A., 2000, MNRAS, 312, 139
574: \bibitem[\protect\citeauthoryear{Khochfar \& Burkert}{2005}]{khochfar05}
575: Khochfar S., Burkert A., 2005, MNRAS, 359, 1379
576: \bibitem[\protect\citeauthoryear{Khosroshahi \etal}{2004}]{habib04}
577: Khosroshahi H. G., Raychaudhury S., Ponman T. J., Miles T. A. Forbes D.,
578: 2004, MNRAS, 349, 524
579: \bibitem[\protect\citeauthoryear{Khosroshahi, Jones \& Ponman}{2004}]{kjp04}
580: Khosroshahi H. G., Jones L. R. and Ponman T. J., 2004, MNRAS, 349, 1240
581: \bibitem[\protect\citeauthoryear{Khosroshahi \etal}{2006}]{kmpj06}
582: Khosroshahi H. G., Maughan B., Ponman T. J., and Jones L. R., 2006,
583: MNRAS, to be published (astro-ph/0603606)
584: \bibitem[\protect\citeauthoryear{Khosroshahi, Ponman \& Jones}{2006}]{kpj06}
585: Khosroshahi H. G., Ponman T. J., and Jones L. R., 2006, MNRAS, submitted
586: \bibitem[\protect\citeauthoryear{Khosroshahi \etal}{2000}]{kwkm00}
587: Khosroshahi H. G., Wadadekar Y., Kembhavi A., Mobasher B., 2000, ApJ, 531, L103
588: \bibitem[\protect\citeauthoryear{Laine \etal}{2003}]{laine03}
589: Laine S., \etal, 2003, AJ, 125, 478
590: \bibitem[\protect\citeauthoryear{Mendes de Oliveira, Cypriano \& Sodre Jr.}
591: {2005}]{mendes05}
592: Mendes de Oliveira C., Cypriano E. S., Sodre Jr. L., 2005, astro-ph/0509884
593: \bibitem[\protect\citeauthoryear{Milosavljevic \etal}{2006}]{milos06}
594: Milosavljevic M., Miller C. J., Furlanetto S. R., Cooray A., 2006, ApJ, 637, L9
595: \bibitem[\protect\citeauthoryear{Lin \& Mohr}{2004}]{mohr04}
596: Lin Y.-T. \& Mohr J. J., 2004, ApJ, 617, 879
597: \bibitem[\protect\citeauthoryear{Naab \& Burkert}{2003}]{naab03}
598: Naab T., Burkert A., 2003, ApJ, 597, 893
599: \bibitem[\protect\citeauthoryear{Naab, Khochfar \& Burkert}{2006}]{naab06}
600: Naab T., Khochfar S., Burkert A., 2006, ApJ, 636, L81
601: \bibitem[\protect\citeauthoryear{Naab \& Burkert}{2003}]{naabtru06}
602: Naab T., Trujillo I., 2006, MNRAS, 369, 625
603: \bibitem[\protect\citeauthoryear{Osmond \& Ponman}{2004}]{osmond04}
604: Osmond J.P.F., Ponman T.J., 2004, MNRAS, 350, 1511
605: \bibitem[\protect\citeauthoryear{Ponman \etal}{1994}]{ponman94}
606: Ponman T. J., Allan D. J., Jones L. R., Merrifield M., MacHardy I. M., 1994,
607: Nature, 369, 462
608: \bibitem[\protect\citeauthoryear{Rest \etal}{2001}]{rest01}
609: Rest A., \etal 2001, AJ, 121, 2431
610: \bibitem[\protect\citeauthoryear{Scorza \& Bender}{1995}]{scorza95}
611: Scorza C., Bender R., 1995 A\&A, 293, 20
612: \bibitem[\protect\citeauthoryear{Searle, Sargent \& Bagnuolo}{1973}]{searle73}
613: Searle L., Sargent W. L. W., and Bagnuolo W. G., 1973, ApJ, 179, 427
614: \bibitem[\protect\citeauthoryear{Smith \etal}{2005}]{smith05}
615: Smith G. P. Kneib J., Smail I., Mazzotta P., Ebeling H., Czoske, O. , 2005,
616: MNRAS, 359, 417
617: \bibitem[\protect\citeauthoryear{Toomre \& Toomre}{1972}]{toomre72}
618: Toomre A., and Toomre J., 1972, ApJ, 178, 623
619: \bibitem[\protect\citeauthoryear{Sun \etal}{2003}]{sun03}
620: Sun M., Forman W., Vikhlinin A., Hornstrup A., Jones C., Murray S. S., 2003, ApJ, 598, 250
621: \bibitem[\protect\citeauthoryear{Yoshioka \etal}{2004}]{yoshioka04}
622: Yoshioka T., Furuzawa A., Takahashi S., Tawara Y., Sato S., Yamashita K.,
623: Kumai Y., 2004, Adv. in Space Res, 34, 2525
624:
625: \end{thebibliography}
626:
627: \bsp
628:
629: \label{lastpage}
630: \end{document}
631: