Astron. Astrophys. 359, 447-456 (2000)

Appendix A: nuclear strength versus ellipticity

Although there is apparently no correlation between the nuclear offset distance and the ellipticity of the underlying galaxy (see above), there is clearly such a correlation between nuclear strength and ellipticity. The data are available since years, but the correlation has never been shown in the literature. As the nuclear strengths, just as the nuclear offsets, must have a bearing on the mass of the nuclei and the structure of the host galaxies, we show this in the present appendix.

By nuclear strength we mean the magnitude of the central luminosity excess of a nucleated dE above a suitable luminosity profile fitted to the underlying galaxy. Such nuclear magnitudes are provided by BC91/93 for a subsample of their dENs, where the galaxy luminosity profile was represented by a King model profile (King 1966). In Fig. 8 we have plotted the ellipticity of the underlying galaxy (also taken from BC93) versus the nuclear magnitude, , from the King model fitting of B93. B93 do not give a mean uncertainty for their nuclear magnitudes, but we estimate it to be as large as 0.5 mag. Recall first the absolute magnitude scale: with ( = 31.5 (D=20 Mpc) an apparent magnitude of =20 corresponds to , which is roughly met by the very brightest globular clusters of M87 (however, being several magnitudes brighter than an ordinary globular cluster of the Milky Way galaxy; see BC91).

Fig. 8. Magnitude of the nucleus , calculated as central light excess above a King model profile fitted to the underlying dE galaxy, versus the ellipticity of the underlying galaxy. Data taken from BC93.

There is a rapid decline in the number of nuclei beyond =21, which is most likely due to incompleteness: fainter nuclei are obviously much harder to detect. The interesting part of Fig. 8 is the bright end. We notice a pronounced tendency of the brightest nuclei (19.5) to live in round (0.2) galaxies. In Fig. 9 we show the same for the relative nuclear strength, i.e. the luminosity ratio , which in magnitudes is simply . A value of =2.5 means that the nucleus comprises 10% of the total light of the galaxy; a value of 5 corresponds to 1% of the total galaxy, etc. There is clearly the same effect here in Fig. 9: the relatively brightest nuclei are found in round galaxies. Nuclei that are brighter than 4% of the total galaxy light (3.75) are found in galaxies with ellipticities (there is only one outlier in Fig. 9, VCC1826).

Fig. 9. Nuclear magnitude relative to the total galaxy magnitude (including the nucleus), , versus the ellipticity of the underlying galaxy. Data taken from BC93.

Nucleated dEs are known to be rounder than the non-nucleated dEs on average by about half an ellipticity class (Binggeli & Popescu 1995). What we have shown here is that this is probably due to the distinct roundness of the host galaxies of the very brightest nuclei. There are at least three possible explanations for this effect. (1) The more elongated dwarfs might have higher internal extinction towards the nucleus due to our viewing angle. This is rather unlikely, as there is no evidence for a great amount of dust in dwarf ellipticals. (2) The built-up of the nucleus could be linked to gas infall which, in turn, is likely governed by angular momentum and, if dEs are rotation supported, by ellipticity (e.g., Davies & Phillipps 1988). However, the very sparse kinematic data for dEs available so far seem to indicate that dEs are not rotationally supported (see Ferguson & Binggeli 1994; however, Jerjen et al. 2000). (3) the most plausible explanation is that the presence of the nucleus, if it is massive enough, has changed the orbital structure of the galaxy over a Hubble time. Norman et al. (1985) have calculated and simulated the effect of a black hole put in the center of a triaxial elliptical, with a mass of a few tenths of the core mass of the galaxy, and found a significant roundening of the underlying galaxy out to at least five core radii. The physical cause for this is the disruption of box orbits by scattering of stars off the central density cusp. A similar result, with a less massive and less effective central compact object, was obtained by Gerhard & Binney (1985).

Based on the work of Norman et al. (1985), Norman (1986) actually predicted that nucleated dEs should be rounder than non-nucleated ones. This is certainly confirmed by our observations. Indeed, the effect that a nucleus of only a few percent of the total galaxy light should cause a nearly perfect roundness of the underlying galaxy out to a faint surface brightness of 26 B arcsec^-2 (the level at which the ellipticities were determined in BC93) may appear even stronger than expected. Could it be that some of these nuclei harbour a black hole, thus comprising a more significant fraction of the total mass of the galaxy? Without further, detailed work on the dynamics of nucleated dEs, including the effects of dark matter, it is impossible to draw any quantitative conclusions.

© European Southern Observatory (ESO) 2000

Online publication: July 7, 2000
helpdesk.link@springer.de