5.2. Results from Analyses Including PN or GC Velocities
Because the dark matter fraction inside 2Re is still modest, and the orbit structure in the outer main bodies of ellipticals is not well-constrained by data that end at 2Re, it is important to include discrete velocity data from planetary nebulae (PN) or globular clusters (GC) that reach to larger radii.
The results so far are mixed. PNe velocity dispersions in several intermediate luminosity ellipticals were found to decline with radius, and to require little if any dark matter at 2 - 5Re [2, 15]. Two of these (NGC 3379, 4494) were also part of the sample of [4, 31], but were consistent with constant M/L also in that study. NGC 3379 has been argued to be a face-on S0 and a weakly triaxial system viewed face-on [35]. This would seem to be not a likely explanation for several galaxies at the same time, unless selecting round galaxies introduces a significant bias in this sense. It is comforting that the near-edge-on, similar luminosity NGC 4697 also has little evidence for dark matter [2]. See Romanowsky's paper for further discussion [19]. In Cen A, the observed PN kinematics do require a dark halo, consistent with the GC velocities, but the implied M / LB is noticeably low [18, 36] when compared with values determined from the hot gas in X-ray bright ellipticals. A similar, mild gradient in M / LB was inferred from PN velocities in NGC 1316 [14]. In several further elliptical galaxies, globular cluster velocities were used to estimate host galaxy masses, increasing outwards to several effective radii, but no M / LB values were given [37].
Two interesting cases are the central galaxies of the Fornax and Virgo clusters, NGC 1399 and M87. In NGC 1399, the dynamical models for the ALS kinematics, predicting significantly increased M / LB already at 1 - 2Re [4], the PN and GC velocities [5, 18, 38], and the ROSAT and ASCA X-ray data [39] imply a steady outward rise in M / LB. The PN and GC velocities are in the right radial range to allow a study of the transition between the potential of the central NGC 1399 galaxy and the potential of the Fornax cluster. Similarly, from a study of the stellar kinematics and the GC velocities around M87, and a comparison to the X-ray mass profile, Romanowsky & Kochanek [5, see also 40] find a rising circular velocity curve, and suggest that the potential of the Virgo cluster may already dominate at r ~ 20 kpc from the center of M87. The dark matter distribution inferred by them is in agreement with that inferred from the X-ray data [6].