PDF《compact?》

Cimatti et al. 2008;

van Dokkum et al. 2008;

Buitrago et al. 2008;

Damjanov et al. 2009;

Saracco et al. 2009;

Rettura et al. 2008). ETGs with similar stellar densities appear to be extremely rare in the local Universe (Trujillo et al. 2009), although it has been argued that such compact high redshift ETGs have survived as the cores of present-day massive spheroids (Hopkins et al. 2009). It has also been argued that the Re ? M? relation for ETGs would keep evolving also from z ?

1 all the way to z ? 0, with a steady increase of Re for given M? (van der Wel et al. 2008;

Bernardi 2009). The formation of very compact ETGs at high redshift may not be a problem: submillimetre galaxies appear to have comparable masses (largely in gas form) and radii, hence may likely be precursors to compact ETGs (Cimatti et al.

2 C. Mancini , et al. 2008;

Tacconi et al. 2008), turning into them upon expul- sion of their residual gas. However, no generally accepted explanation has yet been established on how such compact ETGs would evolve into their present descendants, i.e., how they can in?ate to reach 2C4 times larger e?ective radii. ETG-ETG (dry) merging soon appeared to be a plausi- ble mechanism (e.g., Khochfar &

Silk 2006), but most such events would result in an increase of both mass and radius such to move galaxies parallel to the local Re ? M? rela- tion, rather than towards it (Cimatti et al. 2008;

Saracco et al. 2009). Moreover, major dry merging events may be too rare anyway, as the vast majority of high-z ETGs do not appear to have close ETG companions of similar brightness (Cimatti et al. 2008). Having excluded major mergers, ac- cretion of many satellites (i.e., minor dry mergers) were then considered, either adding an extended envelope to the com- pact core (Cimatti et al. 2008), or expanding such core by gravitationally heating (Naab et al. 2007). It has also been suggested that the expansion of high-z ETGs would be the result of AGN feedback driving the rapid expulsion of the residual gas (Fan et al. 2008), but such expansion must take place in at most a few dynamical times (?

108 yr), hence only very few objects could be caught as already passive and still compact. On the other hand, the possibility that the apparent small radii may not be real was not completely excluded. This could be either the e?ect of part of the light com- ing from a centrally concentrated source (e.g., an AGN or central starburst), or to some systematic bias not having been taken in full account (D05). Recently, La Barbera et al. (2009) have argued for the presence of a radial age gradient in local ETGs, that would result in an apparent decrease of Re with increasing redshift, as the younger, more cen- trally concentrated population di?erentially brightens w.r.t. the older and broader stellar component. However, there ap- pears to be no strong colour gradients in the few cases in which both optical and near-IR HST imaging is available for high-z ETGs (e.g. Toft et al. 2007), or within di?erent ACS bands (D05). In this letter we present the results of two-dimensional (2D) surface brightness pro?le ?tting of a complete sam- ple of

12 extremely massive ETGs (M? >

2.5 *

1011 M⊙) at z 1.4, using the HST+ACS/WFC images of the COS- MOS

2 deg2 ?eld (Koekemoer et al. 2007). Selection of high- z ETGs over such a large area allows us to pick the most massive/luminous high-z ETGs, and tackle the size issue with the highest S/N ratio.

2 DATA, SAMPLE SELECTION, AND SED FITTING A sample of extremely massive ETGs was extracted from the catalogue of K-selected galaxies in the