Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~marel/abstracts/psdir/yale01_geha1.ps Äàòà èçìåíåíèÿ: Tue Jul 10 02:54:39 2001 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 14:32:03 2007 Êîäèðîâêà:

Keck Spectroscopy of Dwarf Elliptical Galaxies in the Virgo
Cluster
M. Geha, P. Guhathakurta
UCO/Lick Observatory, Santa Cruz, CA 95064, USA
E­mail: mgeha, raja@ucolick.org
R. van der Marel
STScI, 3700 San Martin Drive, Baltimore, MD 21218, USA
E­mail: marel@stsci.edu
Keck spectroscopy is presented for four dwarf elliptical galaxies in the Virgo Clus­
ter. At this distance, the mean velocity and velocity dispersion are well resolved as
a function of radius between 100 to 1000 pc, allowing a clear separation between
nuclear and surroundinggalaxy light. We find a variety of dispersion profiles for the
inner regions of these objects, and show that none of these galaxies is rotationally
flattened.
1 Introduction
Dwarf elliptical galaxies (dEs) are among the poorest studied galaxies due
to their faint luminosities, M V
>
# - 17, and characteristic low e#ective surface
brightness µ e (V ) > 22 mag arcsec -2 (Ferguson & Binggeli 1994). The defining
characteristic of dEs is an exponential surface brightness profile. The majority
of dEs brighter than M V = -16 have compact nuclei typically containing
5 to 20% of the total galaxy light; most dEs fainter than M V = -12 show
no sign of a nucleus (Sandage et al. 1985). Although the sample of dEs
with measured internal kinematics is small (Bender & Nieto 1990; Peterson &
Caldwell 1993), these observations have provided strong evidence that dwarf
and classical ellipticals evolve via very di#erent physical processes.
Here we present Keck spectroscopy for four Virgo dEs. Velocity and veloc­
ity dispersion profiles are measured out to # 1 kpc, assuming a Virgo Cluster
distance of 16.1 Mpc (Kelson et al. 2000). These are the initial results of a
larger project to study the dynamics of dwarf elliptical galaxies.
2 Keck Observations
Four Virgo dEs were observed with the Echelle Spectrograph and Imager (ESI)
on the Keck II telescope in March 2001. The spectra were obtained through a
0.75 ## â20 ## slit placed along the major axis of each galaxy with wavelength cov­
erage ##3900-9500 š A and resolution of 23 km s -1 (Gaussian sigma). As shown
in Table 1, the observed galaxies cover a range of ellipticities and three of the
1

Figure 1: Mean velocity and velocity dispersion profiles for four Virgo dEs. The bar at the
lower left of each panel indicates the seeing FWHM during each observation. At the distance
of the Virgo Cluster, 1 ## # 100 pc.
four are nucleated dwarfs (dE,N). These objects lie near the bright end of the
dE luminosity function and were selected to have archival WFPC2 imaging.
Mean radial velocities and velocity dispersions were determined using a pixel
space # 2 minimization scheme described in van der Marel (1994). The data
were spatially rebinned to achieve a S/N > 5 at all radii. Velocities are mea­
sured relative to a K0III template star using the Mg b region, ##5000-5400š A;
an analysis of the full wavelength region will be presented in a forthcoming
paper. Tests show that the galaxies' internal velocity dispersions are recovered
accurately down to the instrumental resolution of 23 km s -1 .
3 Discussion
3.1 Anisotropic Dispersion Versus Rotational Flattening
The observed shapes and kinematics of elliptical galaxies between -20 < MB <
-18 are consistent with rotational flattening. This trend does not appear to
extend to lower luminosity classical ellipticals and the three Local Group dwarf
2

Figure 2: The ratio of the upper limit on the rotation velocity vmax to observed velocity
dispersion # plotted versus mean ellipticity for four Virgo dwarf ellipticals. The solid line is
the expected relation for an oblate, isotropic galaxy flattened by rotation.
ellipticals (Davies et al. 1983; Bender & Nieto 1990). The four Virgo dEs
presented here are also not rotationally flattened. For each galaxy, an average
ellipticity # was determined by standard ellipse fitting of archival WFPC2
V ­band images between radii of 1 ## - 20 ## (see Table 1). From the velocity
profiles shown in Figure 1, we estimate an upper limit to the maximum rotation
velocity, v max . An average velocity dispersion # is determined for each galaxy
beyond r > 1 ## to avoid nuclear contamination.
The ratio v max /# is plotted against ellipticity in Figure 2 and is compared
to the ratio expected from an isotropic, rotationally flattened body (Binney
& Tremaine 1987). The upper limits on v max /# determined for these galaxies
are 2 to 8 times smaller than expected if the observed flattenings were due
to rotation. Thus, we conclude that these dEs are primarily flattened by
anisotropic velocity dispersions.
3.2 Velocity Dispersion Profiles and dE Nuclei
Although the mean velocity profiles presented in Figure 1 are qualitatively
similar, the velocity dispersion profiles are more heterogeneous. The veloc­
ity dispersion of the non­nucleated dE VCC 917 decreases smoothly towards
the galaxy center in contrast to the three nucleated dwarfs, which vary more
3

Table 1: Observed Virgo Cluster Dwarf Elliptical Galaxies
Galaxy Name MB Type #
VCC 917 -16.4 dE6 0.54
VCC 1073 -17.3 dE3,N 0.24
VCC 1254 -16.4 dE0,N 0.05
VCC 1876 -16.8 dE5,N 0.45
abruptly in the central few arcseconds. The nuclear velocity dispersions of
two dE,Ns, VCC 1073 and VCC 1876, are lower than the surrounding galaxy,
whereas the nuclear velocity dispersion of VCC 1254 is higher. The origin of
nuclei in dEs is largely unknown, but their presence has been correlated to
global galaxy parameters such as shape and specific globular cluster frequency
(Ryden & Terndrup 1994; Miller et al. 1998). A favored hypothesis is that the
nuclei are dense star clusters, possibly remnants of larger stripped or harassed
objects (Moore et al. 1998). More work is needed to determine whether the
kinematic profiles presented here are consistent with these scenarios.
Dynamical mass modeling of classical ellipticals has placed strong con­
straints on their origin and evolution. We are in the process of modeling these
dE kinematic data using techniques similar to those described in van der Marel
(1994). We will investigate the variation of M/L ratio across our sample of
galaxies and as a function of galactic radius within each galaxy. In addition,
we plan to study the position of these dEs in the Fundamental Plane. These
results will be presented in an forthcoming paper.
References
1. R. Bender & J. Nieto, A&A 239, 97 (1990).
2. J. Binney & S. Tremaine, Galactic Dynamics, Princeton Univ. Press,
1987.
3. R. Davies et al., ApJ 266, 41 (1983).
4. H. Ferguson & B. Bingelli, A&A Rev 6, 67 (1994).
5. D. Kelson et al., ApJ 529, 768 (2000).
6. B. Miller et al., ApJ 508, L133 (1998).
7. B. Moore et al., ApJ 495, 139 (1998).
8. R. Peterson & N. Caldwell, AJ 105, 1411 (1993).
9. B. Ryden & D. Terndrup, ApJ 425, 43 (1994).
10. A. Sandage et al., AJ 90, 1759 (1985).
11. R. van der Marel, MNRAS 271, 270 (1994).
4