We introduce the flattening equation, which relates the shape of the dark halo to the angular velocity dispersions and the density of a tracer population of stars. It assumes spherical alignment of the velocity dispersion tensor, as seen in the data on stellar halo stars in the Milky Way. The angular anisotropy and gradients in the angular velocity dispersions drive the solutions towards prolateness, whilst the gradient in the stellar density is a competing effect favouring oblateness. We provide an efficient numerical algorithm to integrate the flattening equation. Using tests on mock data, we show that the there is a strong degeneracy between circular speed and flattening, which can be circumvented with informative priors. Therefore, we advocate the use of the flattening equation to test for oblateness or prolateness, though the precise value of the flattening $q$ can only be measured with the addition of the radial Jeans equation. We apply the flattening equation to a sample extracted from the Sloan Digital Sky Survey of $\sim 15000$ halo stars with full phase space information and errors. We find that between Galactocentric radii of 5 and 10 kpc, the shape of the dark halo is prolate, whilst even mildly oblate models are disfavoured. Strongly oblate models are ruled out. Specifically, for a logarithmic halo model, if the asymptotic circular speed $v_0$ lies between $210$ and 250 kms$^{-1}$, then we find the axis ratio of the equipotentials $q$ satisfies $1.5 \lesssim q \lesssim 2$.

A family of spherical halo models with flat circular velocity curves is presented. This includes models in which the rotation curve has a finite central value but declines outwards (like the Jaffe model). It includes models in which the rotation curve is rising in the inner parts, but flattens asymptotically (like the Binney model). The family encompasses models with both finite and singular (cuspy) density profiles. The self-consistent distribution function depending on binding energy $E$ and angular momentum $L$ is derived and the kinematical properties of the models discussed. These really describe the properties of the total matter (both luminous and dark). For comparison with observations, it is better to consider tracer populations of stars. These can be used to represent elliptical galaxies or the spheroidal components of spiral galaxies. Accordingly, we study the properties of tracers with power-law or Einasto profiles moving in the doubloon potential. Under the assumption of spherical alignment, we provide a simple way to solve the Jeans equations for the velocity dispersions. This choice of alignment is supported by observations on the stellar halo of the Milky Way. Power-law tracers have prolate spheroidal velocity ellipsoids everywhere. However, this is not the case for Einasto tracers, for which the velocity ellipsoids change from prolate to oblate spheroidal near the pole. Asymptotic forms of the velocity distributions close to the escape speed are also derived, with an eye to application to the high velocity stars in the Milky Way. Power-law tracers have power-law or Maxwellian velocity distributions tails, whereas Einasto tracers have super-exponential cut-offs.

We present a model for producing tidal streams from disrupting progenitors in arbitrary potentials, utilizing the idea that the majority of stars escape from the progenitor's two Lagrange points. The method involves releasing test particles at the Lagrange points as the satellite orbits the host and dynamically evolving them in the potential of both host and progenitor. The method is sufficiently fast to allow large-dimensional parameter exploration using Monte Carlo methods. We provide the first direct modelling of 6-D stream observations -- assuming a stream rather than an orbit -- by applying our methods to GD-1. This is a kinematically cold stream spanning $60^{\circ}$ of the sky and residing in the outer Galaxy $\approx 15$ kpc distant from the centre. We assume the stream moves in a flattened logarithmic potential characterised by an asymptotic circular velocity $v_0$ and a flattening $q$. We recover values of normalisation $v_0$ = $227.2^{+15.6}_{-18.2}$ kms$^{-1}$ and flattening $q$ = $0.91^{+0.04}_{-0.1}$, if the stream is assumed to leading, and $v_0$ = $226.5^{+17.9}_{-17.0}$ kms$^{-1}$, $q$ = $0.90^{+0.05}_{-0.09}$, if it is assumed to be trailing. This can be compared to the values $v_0 = 224 \pm 13$ kms$^{-1}$ and $q= 0.87^{+0.07}_{-0.04}$ obtained by Koposov et al (2010) using the simpler technique of orbit fitting. Although there are differences between stream and orbit fitting, we conclude that orbit fitting can provide accurate results given the current quality of the data, at least for this kinematically cold stream in this logarithmic model of the Galaxy.

We demonstrate that the asymmetric distribution of M31 satellites cannot be produced by tides from the Milky Way as such effects are too weak. However, loosely bound associations and groups of satellites can fall into larger haloes and give rise to asymmetries. We compute the survival times for such associations. We prove that the survival time is always shortest in Keplerian potentials, and can be ~ 3 times longer in logarithmic potentials. We provide an analytical formula for the dispersal time in terms of the size and velocity dispersion of the infalling structure. We show that, if an association of ~10 dwarfs fell into the M31 halo, its present aspect would be that of an asymmetric disk of satellites. We also discuss the case of cold substructure in the Andromeda II and Ursa Minor dwarfs.

We present a set of highly flattened galaxy models with asymptotically constant rotation curves. The mass density in the equatorial plane falls like (distance)$^{-1}$ at large radii. Although the inner equidensity contours may be spherical, oblate or prolate, the outer parts are always severely flattened. The elongated shape is supported by rotation or tangential velocity anisotropy. The models are thickened Mestel discs, and form a previously undiscovered part of the Miyamoto & Nagai sequence of flattened galaxies. The properties of the models -- axis ratios, velocity dispersions, streaming velocities and distribution functions -- are all discussed in some detail. We pose the question: are extremely flattened or disk-like haloes possible for the Milky Way galaxy? This has never been examined before, as very flattened halo models were not available. We fit the rotation curve and the vertical kinematics of disc stars in the solar neighbourhood to constrain the overall shape of the Galaxy. Denoting the ratio of polar axis to major axis by $q$, we show that models with $q\lesssim 0.57$ cannot simultaneously reproduce the in-plane and out-of-plane constraints. The kinematics of the Sagittarius galaxy also strongly disfavour models with high flattening, as the orbital plane precession is too great and the height reached above the Galactic plane is too small. At least for our Galaxy, the dark halo cannot be flatter than E4 (or axis ratio $q \sim 0.57$) at the Solar circle. Models in which the dark matter is accounted for by a massive baryonic disc or by decaying neutrinos are therefore ruled out by constraints from the rotation curve and the vertical kinematics.

We present a simple formula for the Hamiltonian in terms of the actions for spherically symmetric, scale-free potentials. The Hamiltonian is a power-law or logarithmic function of a linear combination of the actions. Our expression reduces to the well-known results for the familiar cases of the harmonic oscillator and the Kepler potential. For other power-laws, as well as for the singular isothermal sphere, it is exact for the radial and circular orbits, and very accurate for general orbits. Numerical tests show that the errors are always small, with mean errors across a grid of actions always less than 1 % and maximum errors less than 2.5 %. Simple first-order corrections can reduce mean errors to less than 0.6 % and maximum errors to less than 1 %. We use our new result to show that :[1] the misalignment angle between debris in a stream and a progenitor is always very nearly zero in spherical scale-free potentials, demonstrating that streams can be sometimes well approximated by orbits, [2] the effects of an adiabatic change in the stellar density profile in the inner regions of a galaxy weaken any existing 1/r density cusp, which is reduced to $1/r^{1/3}$. More generally, we derive the full range of adiabatic cusp transformations and show how to relate the starting cusp index to the final cusp index. It follows that adiabatic transformations can never erase a dark matter cusp.

We construct simple triaxial generalisations of Navarro-Frenk-White haloes. The models have elementary gravitational potentials, together with a density that is cusped like 1/r at small radii and falls off like 1/r^3 at large radii. The ellipticity varies with radius in a manner that can be tailored to the user's specification. The closed periodic orbits in the planes perpendicular to the short and long axes of the model are well-described by epicyclic theory, and can be used as building blocks for long-lived discs. As an application, we carry out the simulations of thin discs of satellites in triaxial dark halo potentials. This is motivated by the recent claims of an extended, thin disc of satellites around the M31 galaxy with a vertical rms scatter of ~12 kpc and a radial extent of ~ 300 kpc (Ibata et al. 2013). We show that a thin satellite disc can persist over cosmological times if and only if it lies in the planes perpendicular to the long or short axis of a triaxial halo, or in the equatorial or polar planes of a spheroidal halo. In any other orientation, then the disc thickness doubles on ~5 Gyr timescales and so must have been born with an implausibly small vertical scaleheight.