The standard structure formation scenario is successful on linear scales. Several apparent problems affect it however at galactic scales, such as the small scale problems at low redshift and more recent issues involving early massive galaxy and black hole formation. As these problems arise where complex baryonic physics becomes important, the associated unknowns are often assumed to be behind the problems. But the same scales are also those where the primordial spectrum is relatively unconstrained, and there are several ways in which it can be modified. We focus on that arising from effects possibly associated with the crossing of high energy cutoff scale by fluctuation modes during inflation. Elementary arguments show that adiabatic evolution cannot modify the near scale invariance, we thus discuss a simple model for the contrary extreme of sudden transition. Numerical calculations and simple arguments suggest that its predictions, for parameters considered here, are more generic than may be expected, with significant modifications requiring a rapid transition. We examine the implications of such a scenario, in this simplest form of sudden jump, on the matter power spectrum and halo mass function in light of the limitations imposed by particle production. We show that enhancement and oscillation in the power spectrum on currently nonlinear scales can potentially simultaneously alleviate both the apparent problem of early structure formation and, somewhat counterintuitively, problems at low redshift such as the apparent dearth of dwarf galaxies. We discuss consequences that can observationally constrain the scenario and its parameters, including an inflationary Hubble scale $\lesssim 10^{-8} M_{\rm Pl}$, while touching on the possibility of simultaneous modification of power on the largest scales.

Fuzzy dark matter of ultra-light axions has gained attention, largely in light of the galactic scale problems associated with cold dark matter. But the large de Broglie wavelength, believed to possibly alleviate these problems, also leads to fluctuations that place constraints on ultra-light axions. We adapt and extend a method, previously devised to describe the effect of gaseous fluctuations on cold dark matter cusps, in order to determine the imprints of ultra-light axion haloes on the motion of classical test particles. We first evaluate the effect of fluctuations in a statistically homogeneous medium of classical particles, then in a similar system of ultra light axions. In the first case, one recovers the classical two body relaxation time (and diffusion coefficients) from white noise density fluctuations. In the second situation, the fluctuations are not born of discreteness noise but from the finite de Broglie wavelength; correlation therefore exists over this scale, while white noise is retained on larger scales, elucidating the correspondence with classical relaxation. The resulting density power spectra and correlation functions are compared with those inferred from numerical simulations, and the relaxation time arising from the associated potential fluctuations is evaluated. We then apply our results to estimate the heating of disks embedded in axion dark haloes. We find that this implies an axion mass $m \ga 2 \times 10^{-22} {\rm eV}$. We finally apply our model to the case of the central cluster of Eridanus II, confirming that far stronger constraints on $m$ may in principle be obtained, and discussing the limitations associated with the assumptions leading to these.

We simultaneously study the dynamics of the growth of errors and the question of the faithfulness of simulations of $N$-body systems. The errors are quantified through the numerical reversibility of small-$N$ spherical systems, and by comparing fixed-timestep runs with different stepsizes. The errors add randomly, before exponential divergence sets in, with exponentiation rate virtually independent of $N$, but scale saturating as $\sim 1/\sqrt{N}$, in line with theoretical estimates presented. In a third phase, the growth rate is initially driven by multiplicative enhancement of errors, as in the exponential stage. It is then qualitatively different for the phase space variables and mean field conserved quantities (energy and momentum); for the former, the errors grow systematically through phase mixing, for the latter they grow diffusively. For energy, the $N$-variation of the `relaxation time' of error growth follows the $N$-scaling of two-body relaxation. This is also true for angular momentum in the fixed stepsize runs, although the associated error threshold is higher and the relaxation time smaller. Due to shrinking saturation scales, the information loss associated with the exponential instability decreases with $N$ and the dynamical entropy vanishes at any finite resolution as $N \rightarrow \infty$. A distribution function depending on the integrals of motion in the smooth potential is decreasingly affected. In this sense there is convergence to the collisionless limit, despite the persistence of exponential instability on infinitesimal scales. Nevertheless, the slow $N$-variation in its saturation points to the slowness of the convergence.

We propose an $f(T)$ teleparallel gravity theory including a torsional infrared (IR) correction. We show that the governing Friedmann's equations of a spatially flat universe include a phantom-like effective dark energy term sourced by the torsion IR correction. As has been suggested, this phantom phase does indeed act as to reconcile the tension between local and global measurements of the current Hubble value $H_0$. The resulting cosmological model predicts an electron scattering optical depth $\tau_e\thickapprox 0.058$ at reionization redshift $z_{re} \sim 8.1$, in agreement with observations. The predictions are however in contradiction with baryon acoustic oscillations (BAO) measurements, particularly the distance indicators. We argue that this is the case with any model with a phantom dark energy model that has effects significant enough at redshifts $z \lesssim 2$ as to be currently observable. The reason being that such a scenario introduces systematic differences in terms of distance estimates in relation to the standard model; e.g., if the angular diameter distance to the recombination era is to be kept constant while $H_0$ is increased in the context of a phantom scenario, the distances there are systematically overestimated to all objects at redshifts smaller than recombination. But no such discrepancies exist between $\Lambda$CDM predictions and current data for $z \lesssim 2$.

The cold dark matter model of structure formation faces apparent problems on galactic scales. Several threads point to excessive halo concentration, including central densities that rise too steeply with decreasing radius. Yet, random fluctuations in the gaseous component can 'heat' the centres of haloes, decreasing their densities. We present a theoretical model deriving this effect from first principles: stochastic variations in the gas density are converted into potential fluctuations that act on the dark matter; the associated force correlation function is calculated and the corresponding stochastic equation solved. Assuming a power law spectrum of fluctuations with maximal and minimal cutoff scales, we derive the velocity dispersion imparted to the halo particles and the relevant relaxation time. We further perform numerical simulations, with fluctuations realised as a Gaussian random field, which confirm the formation of a core within a timescale comparable to that derived analytically. Non-radial collective modes enhance the energy transport process that erases the cusp, though the parametrisations of the analytical model persist. In our model, the dominant contribution to the dynamical coupling driving the cusp-core transformation comes from the largest scale fluctuations. Yet, the efficiency of the transformation is independent of the value of the largest scale and depends weakly (linearly) on the power law exponent; it effectively depends on two parameters: the gas mass fraction and the normalisation of the power spectrum. This suggests that cusp-core transformations observed in hydrodynamic simulations of galaxy formation may be understood and parametrised in simple terms, the physical and numerical complexities of the various implementations notwithstanding.

Unified dark matter models are appealing in that they describe the dark sector in terms of a single component. They however face problems when attempting to account for structure formation: in the linear regime, density fluctuations can become Jeans stable and oscillate rather than collapse, though it is possible that this difficulty may be circumvented by invoking nonlinear clustering. Here we examine the behaviour in the fully nonlinear regime, of collapsed objects that should mimic standard dark matter haloes. It is shown that the pressure gradient associated with the unified dark matter fluid should be significant in the outer parts of galaxies and clusters, and its effects obervable. In this case, no flat or falling rotation curve is possible for any (barotropic) equation of state with associated sound speed decreasing with density (a necessary condition if the fluid is to behave as pressureless matter at high density). The associated density profile is therefore also incompatible with that inferred in the outer part of clusters. For the prototypical case of the generalised Chaplygin gas, it is shown that this limits the values of the equation of state index $\alpha$ that are compatible with observations to $\alpha \la 0.0001$ or $\alpha \ga 2$. This is in line from what is deduced from linear analysis. More generally, from the expected properties of dark matter haloes, constraints on the sound speed are derived. For the particular case of the generalised Chaplygin gas, this further constrains the index to $\alpha \la 10^{-9}$ or $\alpha \ga 6.7$. For a unified dark matter fluid to mimic dark halo properties, therefore, it needs to have an equation of state such that the pressure gradients are either minimal or which decrease fast enough so as to be negligible at densities characteristic of the outer parts of haloes.

We show that a standard model gauge singlet fermion field, with mass of order keV or larger, and involved in the inverse seesaw mechanism of light neutrino mass generation, can be a good warm dark matter candidate. Our framework is based on B-L extension of the Standard Model. The construction ensures the absence of any mixing between active neutrinos and the aforementioned dark matter field. This circumvents the usual constraints on the mass of warm dark matter imposed by X-ray observations. We show that over-abundance of thermally produced warm dark matter (which nevertheless do not reach chemical equilibrium) can be reduced to an acceptable range in the presence of a moduli field decaying into radiation --- though only when the reheat temperature is low enough. Our warm dark matter candidate can also be produced directly from the decay of the moduli field during reheating. In this case, obtaining the right amount of relic abundance, while keeping the reheat temperature high enough as to be consistent with Big Bang nucleosynthesis bounds, places constraints on the branching ratio for the decay of the moduli field into dark matter.

The origin of equilibrium gravitational configurations is sought in terms of the stability of their trajectories, as described by the curvature of their Lagrangian configuration manifold of particle positions --- a context in which subtle spurious effects originating from the singularity in the two body potential become particularly clear. We focus on the case of spherical systems, which support only regular orbits in the collisionless limit, despite the persistence of local exponential instability of $N$-body trajectories in the anomalous case of discrete point particle representation even as $N \rightarrow \infty$. When the singularity in the potential is removed, this apparent contradiction disappears. In the absence of fluctuations, equilibrium configurations generally correspond to positive scalar curvature, and thus support stable trajectories. A null scalar curvature is associated with an effective, averaged, equation of state describing dynamically relaxed equilibria with marginally stable trajectories. The associated configurations are quite similar to those of observed elliptical galaxies and simulated cosmological halos, and are necessarily different from the systems dominated by isothermal cores, expected from entropy maximization in the context of the standard theory of violent relaxation. It is suggested that this is the case because a system starting far from equilibrium does not reach a 'most probable state' via violent relaxation, but that this process comes to an end as the system finds and (settles in) a configuration where it can most efficiently wash out perturbations. We explicitly test this interpretation by means of direct simulations.

The origin of equilibrium gravitational configurations is sought in terms of the stability of their trajectories, as described by the curvature of their Lagrangian configuration manifold. We focus on the case of spherical systems, which are integrable in the collisionless (mean field) limit despite the apparent persistence of local instability of trajectories even as $N \rightarrow \infty$. It is shown that when the singularity in the potential is removed, a null scalar curvature is associated with an effective, averaged, equation of state describing dynamically relaxed equilibria with marginally stable trajectories. The associated configurations are quite similar to those of observed elliptical galaxies and simulated cosmological halos. This is the case because a system starting far from equilibrium finally settles in a state which is integrable when unperturbed, but where it can most efficiently wash out perturbations. We explicitly test this interpretation by means of direct simulations.

If dark matter (DM) annihilation accounts for the tantalizing excess of cosmic ray electron/positrons, as reported by the PAMELA, ATIC, HESS and FERMI observatories, then the implied annihilation cross section must be relatively large. This results, in the context of standard cosmological models, in very small relic DM abundances that are incompatible with astrophysical observations. We explore possible resolutions to this apparent conflict in terms of non-standard cosmological scenarios; plausibly allowing for large cross sections, while maintaining relic abundances in accord with current observations.

We present measurements of the two-point galaxy angular correlation function w(\theta) in the COSMOS field. Independent determinations of w(\theta) as a function of magnitude limit are presented for both the HST ACS catalog and also for the ground-based data from Subaru and the CFHT. Despite having significantly different masks, these three determinations agree well. At bright magnitudes (IAB<22), our data generally match very well with existing measurements and with mock catalogs based on semi-analytic galaxy formation calculations of Kitzbichler and White from the Millennium Simulation. The exception is that our result is at the upper end of the expected cosmic variance scatter for \theta > 10 arcmin, which we attribute to a particularly rich structure known to exist at z~0.8. For fainter samples, however, the level of clustering is somewhat higher than reported by some previous studies: in all three catalogues we find w(\theta=1')~0.014 at a median IAB magnitude of 24. At these very faintest magnitudes, our measurements agree well with the latest determinations from the Canada-France Legacy Survey. This level of clustering is approximately double what is predicted by the semi-analytic catalogs (at all angles). The semi-analytic results allow an estimate of cosmic variance, which is too small to account for the discrepancy. We therefore conclude that the mean amplitude of clustering at this level is higher than previously estimated.

We present the first identification of large-scale structures (LSS) at z $< 1.1$ in the Cosmic Evolution Survey (COSMOS). The structures are identified from adaptive smoothing of galaxy counts in the pseudo-3d space ($\alpha,\delta$,z) using the COSMOS photometric redshift catalog. The technique is tested on a simulation including galaxies distributed in model clusters and a field galaxy population -- recovering structures on all scales from 1 to 20\arcmin without {\it a priori} assumptions for the structure size or density profile. Our procedure makes {\bf no} {\it a priori} selection on galaxy spectral energy distribution (SED, for example the Red Sequence), enabling an unbiased investigation of environmental effects on galaxy evolution. The COSMOS photometric redshift catalog yields a sample of $1.5\times10^5$ galaxies with redshift accuracy, $\Delta z_{FWHM}/(1+z) \leq 0.1$ at z $< 1.1$ down to I$_{AB} \leq 25$ mag. Using this sample of galaxies, we identify 42 large-scale structures and clusters. abstract truncated for astroph 25 line limit -- see preprint

This paper aims at quantifying discreetness effects, born of finite particle number, on the dynamics of dark matter haloes forming in the context of cosmological simulations. By generalising the standard calculation of two body relaxation to the case when the size and mass distribution are variable, and parametrising the time evolution using established empirical relations, we find that the dynamics of a million particle halo is noise-dominated within the inner percent of the final virial radius. Far larger particle numbers (~ 10^8) are required for the RMS perturbations to the velocity to drop to the 10 % level there. The radial scaling of the relaxation time is simple and strong: t_relax ~ r^2, implying that numbers >> 10^8 are required to faithfully model the very inner regions; artificial relaxation may thus constitute an important factor, contributing to the contradictory claims concerning the persistence of a power law density cusp to the very centre. The cores of substructure haloes can be many relaxation times old. Since relaxation first causes their expansion before recontraction occurs, it may render them either more difficult or easier to disrupt, depending on their orbital parameters. It may thus modify the characteristics of the subhalo distribution and effects of interactions with the parent. We derive simple closed form formulas for the characteristic relaxation times, as well as for the weak N-scaling reported by Diemand et al. when the main contribution comes from relaxing subhaloes (abridged).

We investigate a model of galaxy clusters in which the hot intracluster gas is efficiently heated by dynamical friction (DF) of galaxies. We allow for both subsonic and supersonic motions of galaxies and use the gravitational drag formula in a gaseous medium presented by Ostriker (1999). The energy lost by the galaxies is either redistributed locally or into a Gaussian centered on the galaxy. We find that the condition of hydrostatic equilibrium and strict energy balance yields a trivial isothermal solution T_iso, independent of radius, or rising temperature distributions provided T_iso/gamma < T < T_iso, where gamma is the adiabatic index of the gas. The isothermal temperature corresponds to the usual scaling relation between the gas temperatures and the velocity dispersions of galaxies. However the minimal temperature associated with the rising solutions is ~ 0.5 T_vir, larger than that inferred from observations, the radial distribution of galaxy masses notwithstanding. Heating by supersonically moving galaxies cannot suppress thermal instability, although it can lengthen the growth time up to the level comparable to the ages of clusters when Mach number of galaxies is less than about two. We show using numerical hydrodynamic simulations that DF-induced heating is generally unable to produce stable equilibrium cores by evolving arbitrary non-equilibrium clusters, although it can lengthen the cooling time. We conclude that DF-induced heating alone is an unlikely solution to the cooling flow problem, although it can still be an important heat supplier, considerably delaying cooling catastrophe. We discuss other potential consequences of DF of galaxies in galaxy clusters.

Simple simulations suggest that the phase space structure of haloes identified in cosmological calculations is invariant under the dynamics induced by sinking substructure satellites -- the background expands so as to leave the total distribution unchanged. We use a Fokker-Planck formulation to show that this is the case. The solutions obtained are valid for any mass spectrum of substructure, because the governing equation is linear in their mass weighed phase space distribution. If the clumps are considered solid, an exchange of poulations takes the form of an exponential instability, with characteristic timescale of the order of the dynamical friction time, leading to a low energy cutoff in the distribution function of the background and a constant density core. We show that there are long lived solutions with such a cutoff. They approximate a situation whereas the clumps are made of dense baryonic material. When stripping is important, as in the case of dissipationless substructure, it is likely that this situation is reversed -- the cutoff is now in the clump distribution function. A simple description suggests that this renders equilibria even more long lived. The results of this paper suggest that halo profiles similar to those found in dissipationless cosmological simulations are approximately invariant under the interaction induced by the presence of substructure satellites -- a necessary condition for the observed `universality'. In addition, the total profile, including baryons, should also be invariant; provided the latter are initially in the form of dense clumps, whose distribution follows that of the dark matter.

Sand, Treu, & Ellis (2002) have measured the central density profile of cluster MS2137-23 with gravitational lensing and velocity dispersion and removed the stellar contribution with a reasonable M/L. The resulting dark matter distribution within r<50 kpc was fitted by a density cusp of r^{-beta} with beta=0.35. This stands in an apparent contradiction to the CDM prediction of beta~1, and the disagreement worsens if adiabatic compression of the dark matter by the infalling baryons is considered. Following El-Zant, Shlosman & Hoffman (2001), we argue that dynamical friction acting on galaxies moving within the dark matter background counters the effect of adiabatic compression by transfering the orbital energy of galaxies to the dark matter, thus heating up and softening the central density cusp. Using N-body simulations of massive solid clumps moving in clusters we show that indeed the inner dark matter distribution flattens (with beta approx 0.35 for a cluster like MS2137-23) when the galaxies spiral inward. We find as a robust result that while the dark matter distribution becomes core-like, the overall mass distribution preserves its cuspy nature, in agreement with X-ray and lensing observations of clusters.

We revisit the notion that galaxy motions can efficiently heat intergalactic gas in the central regions of clusters through dynamical friction. For plausible values of the galaxy mass-to-light ratio, the heating rate is comparable to the cooling rate due to X-ray emission. Heating occurs only for supersonic galaxy motions, so the mechanism is self-regulating: it becomes efficient only when the gas sound speed is smaller than the galaxy velocity dispersion. We illustrate with the Perseus cluster, assuming a stellar mass-to-light ratio for galaxies in the very central region with the dark-matter contribution becoming comparable to this at some radius $r_s$. For $r_s \la 400 {\rm kpc} \sim 3 r_{\rm cool}$--corresponding to an average mass-to-light ratio of $\sim10$ inside that radius--the dynamical-friction coupling is strong enough to provide the required rate of gas heating. The measured sound speed is smaller than the galaxy velocity dispersion, as required by this mechanism. With this smaller gas temperature and the observed distribution of galaxies and gas, the energy reservoir in galactic motions is sufficient to sustain the required heating rate for the lifetime of the cluster. The galaxies also lose a smaller amount of energy through dynamical friction to the dark matter implying that non--cooling-flow clusters should have flat-cored dark-matter density distributions.

A substantial fraction of disk galaxies is double-barred. We analyze the dynamical stability of such nested bar systems by means of Liapunov exponents,by fixing a generic model and varying the inner (secondary) bar mass. We show that there exists a critical mass below which the secondary bar cannot sustain its own orbital structure, and above which it progressively destroys the outer (primary) bar-supporting orbits. In this critical state, a large fraction of the trajectories (regular and chaotic) are aligned with either bar, suggesting the plausibility of long-lived dynamical states when secondary-to-primary bar mass ratio is of the order of a few percent. Qualitatively similar results are obtained by varying the size of the secondary bar, within certain limits, while keeping its mass constant. In both cases, an important role appears to be played by chaotic trajectories which are trapped around (especially) the primary bar for long periods of time.

The masses of supermassive black holes (SBHs) show correlations with bulge properties in disk and elliptical galaxies. We study the formation of galactic structure within flat-core triaxial haloes and show that these correlations can be understood within the framework of a baryonic component modifying the orbital structure in the underlying potential. In particular, we find that terminal properties of bulges and their central SBHs are constrained by the destruction of box orbits in the harmonic cores of dark haloes and the emergence of progressively less eccentric loop orbits there. SBH masses, M_SBH, should exhibit a tighter correlation with bulge velocity dispersions, sigma_B, than with bulge masses, M_B, in accord with observations, if there is a significant scatter in the M_H-sigma_H relation for the halo. In the context of this model the observed M_SBH-sigma_B relation implies that haloes should exhibit a Faber-Jackson type relationship between their masses and velocity dispersions. The most important prediction of our model is that halo properties determine the bulge and SBH parameters. The model also has important implications for galactic morphology and the process of disk formation.

We investigate the stability of trajectories in barred galaxies with mildly triaxial halos by means of Liapunov exponents. This method is perfectly suitable for time-dependent 3D potentials where surfaces of sections and other simple diagnostics are not applicable. We find that when halos are centrally-concentrated most trajectories starting near the plane containing the bar become chaotic. Moreover, the shape of many of the remaining regular trajectories do not match the bar density distribution, being too round. Therefore, time-independent self-consistent solutions are highly unlikely to be found. When the non-rotating non-axisymmetric perturbation in the potential reaches 10%, almost all trajectories integrated are chaotic and have large Liapunov exponents. No regular trajectories aligned with the bar have been found. Hence, if the evolution of the density figure is directly related to the characteristic timescale of orbital instability, bar dissolution would take place on a timescale of few dynamical times. The slowly rotating non-axisymmetric contribution to the potential required for the onset of widespread chaotic behavior is remarkably small. Systems consisting of centrally-concentrated axisymmetric halos and stellar bars thus appear to be structurally unstable, and small (1%) deviations from perfect axisymmetry should result in a bar dissolution on a timescale significantly smaller than the Hubble time. Since halos found in CDM simulations of large scale structure are both centrally-concentrated and triaxial it is unlikely that stellar bars embedded in such halos would form and survive unless the halos are modified during the formation of the baryonic component.

Coupling the dark matter (DM) and baryonic components in galaxies via dynamical friction (by allowing the latter to be clumpy) leads to energy input into the DM. The resulting expanding tendency in the DM component overcomes the competing effect of adiabatic contraction, resulting from the shrinking gas component, and leads to a nearly constant density core. The baryonic mass inflow also leads to the coupling of the structure at the kpc scale and below and that of the structure of the DM at the 10 kpc scale and above, with rate of mass inflow being dependent on the initial halo central concentration. The latter also determines the final halo core radius.

The stability of the dynamical trajectories of softened spherical gravitational systems is examined, both in the case of the full $N$-body problem and that of trajectories moving in the gravitational field of non-interacting background particles. In the latter case, for $N \ge 10000$, some trajectories, even if unstable, had exceedingly long diffusion times, which correlated with the characteristic e-folding timescale of the instability. For trajectories of $N \approx 100000$ systems this timescale could be arbitrarily large --- and thus appear to correspond to regular orbits. For centrally concentrated systems, low angular momentum trajectories were found to be systematically more unstable. This phenomenon is analogous to the well known case of trajectories in generic centrally concentrated non-spherical smooth systems, where eccentric trajectories are found to be chaotic. The exponentiation times also correlate with the conservation of the angular momenta along the trajectories. For $N$ up to a few hundred, the instability timescales of $N$-body systems and their variation with particle number are similar to those of the most chaotic trajectories in inhomogeneous non-interacting systems. For larger $N$ (up to a few thousand) the values of the these timescales were found to saturate, increasing significantly more slowly with $N$. We attribute this to collective effects in the fully self-gravitating problem, which are apparent in the time-variations of the time dependent Liapunov exponents. The results presented here go some way towards resolving the long standing apparent paradoxes concerning the local instability of trajectories of gravitational systems (abridged).

N-body simulations and analytical calculations of the gravitational collapse in an expanding universe predict that halos should form with a diverging inner density profile, the cusp. There are some observational indications that the dark matter distribution in galaxies might be characterized by a finite core. This `core catastrophe' has prompted a search for alternatives to the CDM cosmogony. It is shown here that the discrepancy between theory and observations can be very naturally resolved within the standard CDM model, provided that gas is not initially smoothly distributed in the dark matter halo, but rather is concentrated in clumps of mass $\geq 0.01 %$ the total mass of the system. Dynamical friction acting on these lumps moving in the background of the dark matter particles, dissipates the clumps orbital energy and deposits it in the dark matter. Using Monte-Carlo simulations, it is shown that the dynamical friction provides a strong enough drag, and that with realistic baryonic mass fractions, the available orbital energy of the clumps is sufficient to heat the halo and turn the primordial cusp into a finite, non-diverging core --- overcoming the competing effect of adiabatic contraction due to gravitational influence of the shrinking baryonic component. Depending on the initial conditions, the total density distribution may either become more or less centrally concentrated. Possible consequences of the proposed mechanism for other problems in the CDM model and for the formation and early evolution of the baryonic component of galaxies are also briefly discussed.

Due to the clumpy nature of the self gravitating gas composing the interstellar medium, it is not clear whether galactic gas dynamics can be discussed in terms of standard hydrodynamics. Nevertheless, it is clear that certain generic properties related to orbital structure in a given potential and the effect of dissipation can be used to qualitatively understand gas motion in galaxies. The effect of dissipation is examined in triaxial galaxy potentials with and without rotating time dependent components. In the former case, dissipative trajectories settle around closed loop orbits when these exist. When they do not, e.g., inside a constant density core, then the only attractor is the centre and this leads to mass inflow. This provides a self regulating mechanism for accession of material towards the centre --- since the formation of a central masses destroys the central density core and eventually stops the accession. In the case when a rotating bar is present, there are usually several types of attractors, including those on which long lived chaotic motion can occur (strange attractors). Motion on these is erratic with large radial and vertical oscillations.

The evolution of closed gravitational systems is studied by means of $N$-body simulations. This, as well as being interesting in its own right, provides insight into the dynamical and statistical mechanical properties of gravitational systems: the possibility of the existence of stable equilibrium states and the associated relaxation time would provide an ideal situation where relaxation theory can be tested. Indeed, these states are found to exist for single mass $N$-body systems, and the condition condition for this is simply that obtained from elementary thermodynamical considerations applied to self-gravitating ideal gas spheres. However, even when this condition is satisfied, some initial states may not end as isothermal spheres. It is therefore only a necessary condition. Simple considerations also predict that, for fixed total mass, energy and radius, stable isothermal spheres are unique. Therefore, statistically irreversible perturbations to the density profile caused by the accumulation of massive particles near the centre of multimass systems, destroy these equilibria if the aforementioned quantities are kept fixed. The time-scale for this to happen was found to be remarkably short (a few dynamical times when $N= 2500$) in systems undergoing violent relaxation. The time taken to achieve thermal equilibrium depended on the initial conditions and could be comparable to a dynamical time (even when the conditions for violent relaxation were not satisfied) or the two body relaxation time. The relaxation time for velocity anisotropies was intermediate between these two time-scales, being long compared to the dynamical time but much (about four times) shorter than the time-scale of energy relaxation.

Using the Ricci and scalar curvatures of the configuration manifold of gravitational N-body systems, we study the exponential instability in their trajectories. It is found that the exponentiation time-scale for isotropic Plummer spheres varies very little with particle number if the softening is small. Large softening on the other hand has a marked effect and, if large enough, can cause the curvatures to become positive. This last result confirms the previous observations for self gravitating sheets and suggests that the qualitative behaviour of large-N and continuum systems may be different, and that their equivalence is only obtained in the limit of infinite $N$ and finite softening. It is also found that the presence of a large fraction of the kinetic energy in rotational motion increases the exponentiation time-scales significantly --- an effect that should be expected given the regular nature of nearly circular motion. In the light of the results of this and of previous studies, it is suggested that the exponential instability may arise from low order resonances between the period of the variation of the gravitational field due to distant encounters and the orbital period of a test particle. For periods long compared to the exponentiation time but short compared to the diffusion time-scales of the action variables, the standard picture of collisionless dynamics may be valid in an averaged sense. Instead it is suggested that, at least for systems with well defined final states, the relaxation time should scale as $\sim N^{1/2}$.

A geometric method, using the Ricci (or mean) curvature of the Lagrangian (configuration) manifold, is used to study the relative instability of evolving $N$-body gravitating systems with massive centres. The performed numerical experiments show that the Ricci curvature decreases with the increase in central concentration, thus indicating an increasing instability for systems with massive centre --- either in the form of a point mass or a core formed during the evolution of single and multimass systems. Simulations of such systems do indeed suggest that they have faster gravothermal evolution than less concentrated ones.

Much of standard galaxy dynamics rests on the implicit assumption that the corresponding N-body problem is (near) integrable. This notion although leading to great simplification is by no means a fact. It is therefore important to develop methods characterizing the stability of motion of N-body systems. A method based on the geometry of a system's phase space is examined and compared with more standard techniques. We also present the results of high precision N-body simulations of the dynamics of systems of 231 point particles over a few dynamical times. The scalar and Ricci (or mean) curvatures are calculated along the trajectories and the relative instabilities of different systems are thus derived along with the corresponding ``mixing'' timescales. The effect of softening is also examined. We compare the predictions obtained from the calculations above with the spatial evolution of the different systems and deduce that this is well described.

We study the stability of trajectories near the disk plane of galaxy models with a triaxial dark matter halo component. We also examine the effect of weak discreteness noise, rapidly rotating bar perturbations and weak dissipation on these trajectories. If the matter distribution is triaxial and has a constant density core, dissipation leads to inflow of material inside the core radius to the centre. This leads to the formation of central masses which in turn destabilise the trajectories of any stars formed in these regions. In particular, even if the gas settles by dissipation into a flat disk, stars formed in that disk will later form a bulge like dstribution the extent of which would be related to the core radius of the halo and the original asymmetry in the plane. The process of gas inflow is regulated by the fact that too large and condensed a central mass leads to the creation of stable closed loop orbits in the central area around which the gas can move. This would appear to stop the accumulation of central mass before it becomes large enough for rapid loss of halo triaxiality. It was found that weak discreteness noise can change the fraction of such trajectories significantly and can therefore have important consequences for the modelling of galaxies. The addition of rapidly rotating bar perturbations also increases the degree of instability dramatically. So if bars can form in triaxial haloes they are likely to be quickly destroyed leaving a bulge like structure behind (which may explain the absence of bars in surveys of high redshift galaxies).

Complications arising from the non-compact nature of the phase space of N-body systems prevent any asymptotic characterization of chaotic behaviour (since no equilibrium final states can exist). This leads us to revisit some of the old results concerning geodesic stability on Lagrangian manifolds. These cannot be applied directly but may be useful when properly reinterpreted. A method for doing that is described along with some tests of its validity and possible future applications.