1 Introduction

   

Table 1: Setups of the spectroscopic runs. Column 1 gives the run number, Col. 2 the date, Col. 3 the telescope used, Col. 4 the instrument, Col. 5 the detector, Col. 6 the wavelength range, Col. 7 the scale in arcsec/pixel, Col. 8 the slit width in arcsec, Col. 9 the instrumental resolution in km s^-1

Run	Date	Telescope1	Instrument2	Detector	$\lambda$ - range	Scale	Slit-	$\sigma_{\rm inst}$
					[Å]	[''/pix]	width	[km s-1]
1	31.10-2.11.1992	NTT	EmmiRed	Tek 1024 $\times$ 1024 19$\mu$	4890 - 5430	0 $\hbox{$.\!\!^{\prime\prime}$ }$44	1 $\hbox{$.\!\!^{\prime\prime}$ }$	70
2	4-7.6.1994	NTT Remote	EmmiRed	Tek 2048 $\times$ 2048 24$\mu$	4826 - 5474	0 $\hbox{$.\!\!^{\prime\prime}$ }$268	3 $\hbox{$.\!\!^{\prime\prime}$ }$	85
3	26-30.9.1994	3.5 CA	TwinRed	Tek 1024 $\times$ 1024 24$\mu$	4760 - 5640	0 $\hbox{$.\!\!^{\prime\prime}$ }$896	3 $\hbox{$.\!\!^{\prime\prime}$ }$6	85
4	6-8.6.1994	NTT Remote	EmmiRed	Tek 2048 $\times$ 2048 24$\mu$	4826 - 5474	0 $\hbox{$.\!\!^{\prime\prime}$ }$268	3 $\hbox{$.\!\!^{\prime\prime}$ }$	85
5	26-29.8.1995	3.5 CA	TwinRed	Tek 1024 $\times$ 1024 24$\mu$	4760 - 5640	0 $\hbox{$.\!\!^{\prime\prime}$ }$896	3 $\hbox{$.\!\!^{\prime\prime}$ }$6	85
6	10-12.3.1996	NTT Remote	EmmiRed	Tek 2048 $\times$ 2048 24$\mu$	4826 - 5474	0 $\hbox{$.\!\!^{\prime\prime}$ }$268	3 $\hbox{$.\!\!^{\prime\prime}$ }$	85
(1) NTT: ESO 3.5 m New Technology Telescope; 3.5 CA: 3.5 Calar Alto Telescope.
(2) EmmiRed: Red Arm of EMMI; TwinRed: Red arm of the twin spectrograph.

Much less is known about the mass distributions of elliptical galaxies than for spirals. Ellipticals are hot stellar systems, and their circular velocity curves cannot be directly measured, but must be indirectly inferred from fitting dynamical models to the kinematic data. Has the luminous matter segregated dissipatively in the halo potential? Is there a "conspiracy'' between luminous and dark matter to produce a flat rotation curve, like in spiral galaxies? How do the mass-to-light ratio, the slope of the circular velocity curve, or the orbital anisotropy scale with luminosity? The purpose of this paper is to construct a sample of elliptical galaxies for which these quantities are well-enough known to address these questions.

Evidence for dark matter in ellipticals comes from X-ray data on their hot gas atmospheres (e.g., Matsushita et al. [1998]; Loewenstein & White [1999]), from a few cases with cold gas rings (Bertola et al. [1993]), from gravitational lensing analyses (Keeton et al. [1998]) and from stellar-dynamical work. Saglia et al. ([1992]) studied the velocity dispersion profiles of a sample of 10 bright ellipticals using anisotropic two-component models, finding that the amount of dark matter inside the half-luminosity radius is of the order of the luminous mass. Similar results were obtained modeling the extended profiles of Saglia et al. ([1993]) and Bertin et al. ([1994]), but these authors also showed that some of their galaxies could be modelled with constant mass-to-light ratio when more complicated, tangentially anisotropic models were employed.

Van der Marel ([1991]) constructed axisymmetric two-integral models for a sample of 37 bright ellipticals. He found that these models predicted too much motion on the major axis and interpreted this as evidence for anisotropy in the form $\sigma_{\rm r}>\sigma_\theta$. From the model fits he also derived mass-to-light ratios and concluded that these rise approximately as $\propto L^{0.35}$. These results were based on velocity dispersion and rotation measurements. Both modelling approaches described above are subject to some uncertainty, because of the well-known degeneracy between anisotropy and mass distribution, given only rotation and velocity dispersion data.

In the meantime, both the quality of the available kinematic data and of the dynamical modelling has improved greatly. Line profile shape measurements are now available for many ellipticals (e.g., Bender et al. [1994]), and in a number of cases these measurements now extend to sufficiently large radii ( $\sim1-2R_{\rm e}$) that they begin to constrain the halo mass distribution (e.g., Carollo et al. [1995]). Theoretical work has shown that such absorption line profile data contain sufficient information to extract constraints on both anisotropy and mass distribution (Gerhard [1993]; Merritt [1993]). In several recent studies, modern data together with sophisticated modelling methods were used to determine improved mass-to-light ratios and to estimate M(r) and the anisotropy structure in individual elliptical galaxies (Rix et al. [1997]; Gerhard et al. [1998]; Emsellem et al. [1999]; Cretton & van den Bosch [1999]; Saglia et al. [2000]; Matthias & Gerhard [1999]; Gebhardt et al. [2000]).

The goal of this paper is to extend this work to a sample of elliptical galaxies. We have selected galaxies that (i) cover as wide a range in total magnitudes and $R_{\rm e}$ as possible, in order to probe different regions of the Fundamental Plane, and (ii) are nearly round (E0-E2), with small rotational velocities, so as to allow us to use spherical, non-rotating models in analyzing the data. We have obtained new kinematic data to $\sim1-2R_{\rm e}$ for eight galaxies. In addition, we have used extended kinematic data for NGC 1399 from Saglia et al. ([2000]), for NGC 2434 from Carollo & Danziger ([1994]), for NGC 3379 from Statler & Smecker-Hane ([1999]), and for NGC 6703 from Gerhard et al. ([1998]). To these we have added a second sample of ellipticals with kinematic data to smaller radii, from Bender et al. ([1994], hereafter BSG94), in order to extend the total sample to less luminous galaxies and to increase the sample size. Dynamical models for all of these are shown in this paper. Based on the results of these models we will be able to investigate the family properties of ellipticals regarding mass-to-light ratios and anisotropies; this will be described in a forthcoming paper.

This paper is organised as follows: New spectroscopic observations are presented in Sect. 2. In Sect. 3 we describe the photometric data used for the modelling, as well as the distances employed in calculating masses and luminosities. Our modelling technique, which follows Gerhard et al. ([1998], hereafter G+98) and Saglia et al. ([2000], hereafter S+2000), is briefly discussed in Sect. 4. Results are given in Sect. 5, and in Sect. 6 we present our conclusions.