We present a dynamical study of NGC 4697, an almost edge-on, intermediate-luminosity, E4 elliptical galaxy, combining new surface brightness photometry, new as well as published long-slit absorption line kinematic data, and published planetary nebulae (PNe) velocity data. The combined kinematic data set extends out to ~= 5' ~= 4.5 R_e and allows us to probe the galaxy's outer halo. For the first time, we model such a dataset with the new and flexible Chi^2-made-to-measure particle code NMAGIC. We extend NMAGIC to include seeing effects, introduce an efficient scheme to estimate the mass-to-light ratio, and incorporate a maximum likelihood technique to account for discrete velocity measurements. For modelling the PNe kinematics we use line-of-sight velocities and velocity dispersions computed on two different spatial grids, and we also use the individual velocity measurements with the likelihood method, in order to make sure that our results are not biased by the way we treat the PNe measurements. We generate axisymmetric self-consistent models as well models including various dark matter halos. These models fit all the mean velocity and velocity dispersion data with Chi^2/N<1, both in the case with only luminous matter and in potentials including quite massive halos. The likelihood analysis together with the velocity histograms suggest that models with low density halos such that the circular velocity v_c ~< 200 km/s at R_e are not consistent with the data. A range of massive halos with v_c ~= 250 km/s at 5R_e fit the PN data best. To derive stronger results would require PN velocities at even larger radii. The best fitting models are slightly radially anisotropic; the anisotropy parameter beta ~= 0.3 at the center, increasing to beta ~= 0.5 at radii >~ 2R_e.

We describe a made-to-measure algorithm for constructing N-particle models of stellar systems from observational data (Chi-Squared-M2M), extending earlier ideas by Syer and Tremaine. The algorithm properly accounts for observational errors, is flexible, and can be applied to various systems and geometries. We implement this algorithm in a parallel code NMAGIC and carry out a sequence of tests to illustrate its power and performance: (i) We reconstruct an isotropic Hernquist model from density moments and projected kinematics and recover the correct differential energy distribution and intrinsic kinematics. (ii) We build a self-consistent oblate three-integral maximum rotator model and compare how the distribution function is recovered from integral field and slit kinematic data. (iii) We create a non-rotating and a figure rotating triaxial stellar particle model, reproduce the projected kinematics of the figure rotating system by a non-rotating system of the same intrinsic shape, and illustrate the signature of pattern rotation in this model. From these tests we comment on the dependence of the results from Chi-Squared-M2M on the initial model, the geometry, and the amount of available data.

We have analysed the magnitudes, kinematics and positions of a complete sample of 320 PNs in the elliptical galaxy NGC 4697. We show (i) that the PNs in NGC 4697 do not constitute a single population that is a fair tracer of the distribution of all stars. The radial velocity distributions, mean velocities, and dispersions of bright and faint subsamples differ with high statistical confidence. (ii) Using the combined data for PNs brighter than 26.2, we have identified a subpopulation of PNs which is azimuthally unmixed and kinematically peculiar, and which thus neither traces the distribution of all stars nor can it be in dynamical equilibrium with the galaxy potential. (iii) The planetary nebula luminosity functions (PNLF) of two kinematic subsamples in NGC 4697 differ with 99.7% confidence, ruling out a universal PNLF. We estimate that the inferred secondary PN population introduces an uncertainty in the bright cutoff magnitude of ~ 0.15 mag for this galaxy. -- We argue that this secondary PN distribution may be associated with a younger, >~ 1 Gyr old stellar population, perhaps formed in tidal structures that have now fallen back onto the galaxy, as has previously been suggested for the X-ray point sources in this galaxy, or coming from a more recent merger/accretion with a red galaxy. The use of PNs for extragalactic distance determinations is not necessarily compromised, but their use as dynamical tracers of dark halos will require deep observations and careful analysis of large PN samples.

The transverse velocities of steadily rotating, non--axisymmetric patterns in flat galaxies may be determined by a purely kinematical method, using two dimensional maps of a tracer surface brightness and radial current density. The data--maps could be viewed as the zeroth and first velocity moments of the line--of--sight velocity distribution, which is the natural output of integral--field spectrographs. Our method is closely related to the Tremaine--Weinberg method of estimating pattern speeds of steadily rotating patterns, when the tracer surface brightness satisfies a source--free continuity equation. We prove that, under identical assumptions about the pattern, two dimensional maps may be used to recover not just one number (the pattern speed), but the full vector field of tracer flow in the disc plane. We illustrate the recovery process by applying it to simulated data, and test its robustness by including the effects of noise.

We present stellar dynamical models of the lopsided, double-peaked nucleus of M31, derived from Hubble Space Telescope (HST) photometry. A Schwarzscild-type method, in conjunction with Richardson-Lucy deconvolution, was employed to construct steadily rotating, hot, stellar disks. The stars orbit a massive dark object, on prograde and retrograde quasi-periodic loop orbits. Our results support Tremaine's eccentric disk model, extended to include a more massive disk, non zero pattern speed ($\Omega$), and different viewing angle. Most of the disk mass populated prograde orbits, with $\simeq 3.4%$ on retrograde orbits. The best fits to photometric and kinematic maps were disks with $\Omega\approx 16\kmspc $. We speculate on the origins of the lopsidedness, invoking recent work on the linear overstability of nearly Keplerian disks, that possess even a small amount of a counter-rotating component. Accretion of material-no more massive than a globular cluster-onto a preexisting stellar disk, will account for the mass in our retrograde orbits, and could have stimulated the lopsidedness seen in the nucleus of M31.

We investigate the orbital structure of a model triaxial star cluster, centered around a supermassive black hole (BH), appropriate to galactic nuclei. Sridhar and Touma (1999) proved that the presence of the BH enforces some regularity in the dynamics within the radius of influence of the BH. We employ their averaging method to reduce the degrees of freedom from three to two. Numerical orbit integrations, together with Poincar\'e surfaces of section allow us to draw a global portrait of the orbital structure; in our calculations we employ a model cluster potential that is triaxial and harmonic. The averaged dynamics of the axisymmetric case is integrable, and we present a detailed comparison of orbits in oblate and prolate axisymmetric potentials. Both cases support resonant orbits with fixed values of eccentricity, inclination, and periapse, whose lines of nodes rotates steadily. We then systematically explore significantly triaxial potentials, possessing small oblateness, or prolateness. Resonant orbits and their families are studied both numerically, and through secular perturbation theory. Chaos appears to be suppressed for all the cases we studied, and we obtain effective third integrals. Some of the orbits appear to reinforce the shape of the potential; we provide phase space, as well as real space portraits of these orbits. A particularly promising resonant orbit exists in highly prolate, triaxial potentials.

The twin peaks in the nucleus of M31 have been interpreted by Tremaine as a thick, eccentric, disk of stars orbiting a massive dark object; the required alignment of the apoapsides of the stellar orbits could be maintained by self-gravity, and the whole structure might be a discrete, nonlinear eigenmode. The pattern speed of this mode could, in principle, be determined by the Tremaine-Weinberg (TW) method, which requires measurements of the surface brightness, and radial velocity along a strip parallel to the line of nodes. However, spectroscopic observations along the line of nodes are not available. We propose a variant of the TW method, which exploits a basic feature of the eccentric disk model, to extract estimates of the pattern speed from {\it Hubble Space Telescope} spectroscopic data, taken along the line joining the two peaks. Within limitations imposed by the data, we estimate that the pattern rotates in a prograde manner and, for an assumed disk inclination of $77^{\circ}$, the pattern speed $|\Omega_p| < 30 {\rm km s^{-1} pc^{-1}}$, or period more than $200,000$ years.