We study random bubble lattices which can be produced by processes such as first order phase transitions, and derive characteristics that are important for understanding the percolation of distinct varieties of bubbles. The results are relevant to the formation of topological defects as they show that infinite domain walls and strings will be produced during appropriate first order transitions, and that the most suitable regular lattice to study defect formation in three dimensions is a face centered cubic lattice. Another application of our work is to the distribution of voids in the large-scale structure of the universe. We argue that the present universe is more akin to a system undergoing a first-order phase transition than to one that is crystallizing, as is implicit in the Voronoi foam description. Based on the picture of a bubbly universe, we predict a mean coordination number for the voids of 13.4. The mean coordination number may also be used as a tool to distinguish between different scenarios for structure formation.

We analyze the core dynamics of critically coupled, superheavy gauge vortices in the (2+1) dimensional Einstein-Abelian-Higgs system. By numerically solving the Eistein and field equations for various values of the symmetry breaking scale, we identify the regime in which static solutions cease to exist and topological inflation begins. We explicitly include the topological winding of the vortices into the calculation and extract the dependence on the winding of the critical scale separating the static and inflating regimes. Extrapolation of our results suggests that topological inflation might occur within high winding strings formed at the Grand Unified scale.

We analyze the window functions for the spherical harmonic mode estimators of all--sky, volume limited surveys considering evolutionary effects along the past light--cone which include the deviation of the distance scale from a linear relationship with redshift, linear peculiar velocity corrections, and linear evolution of the density perturbations. The spherical harmonic basis functions are considered because they correspond most closely to the symmetries of typical survey geometries and of the light--cone effects we consider. Our results show substantial broadening of the windows over that expected by ignoring light--cone effects, indicating the difficulty of measuring the power spectrum independently from cosmology. We suggest that because of light--cone effects, deep redshift surveys should either be analyzed in conjunction with CMBR data which determines the cosmological parameters, or by using a Bayesian likelihood scheme in which varying cosmological parameters and a simple parameterization of the primordial power spectrum are assumed as the priors, so that observed data can be mapped from redshift to real space. The derived power spectrum can then be compared to underlying models of fluctuation generation and growth in structure formation to evaluate both these models and the cosmological priors.

The multipole moments of the power spectrum of large scale structure, observed in redshift space, are calculated for a finite sample volume including the effects of both the linear velocity field and geometry. A variance calculation is also performed including the effects of shot noise. The sensitivity with which a survey with the depth and geometry of the Sloan Digital Sky Survey (SDSS) can measure cosmological parameters $\Omega_0$ and $b_0$ (the bias) or $\lambda_0$ (the cosmological constant) and $b_0$ is derived through fitting power spectrum moments to the large scale structure in the linear regime in a way which is independent of the evolution of the galaxy number density. We find that for surveys of the approximate depth of the SDSS no restrictions can be placed on $\Omega_0$ at the 99% confidence limit when a fiducial open, $\Omega_0 = 0.3 $ model is assumed and bias is unconstrained. At the 95% limit, $\Omega_{0} < .85$ is ruled out. Furthermore, for this fiducial model, both flat (cosmological constant) and open models are expected to reasonably fit the data. For flat, cosmological constant models with a fiducial $\Omega_{0} = 0.3$, we find that models with $\Omega_{0} > 0.48$ are ruled out at the 95% confidence limit regardless of the choice of the bias parameter, and open models cannot fit the data even at the 99% confidence limit.

We consider the gravitational lensing produced by long cosmic strings formed in a GUT scale phase transition. We derive a formula for the deflection of photons which pass near the strings that reduces to an integral over the light cone projection of the string configuration plus constant terms which are not important for lensing. Our strings are produced by performing numerical simulations of cosmic string networks in flat, Minkowski space ignoring the effects of cosmological expansion. These strings have more small scale structure than those from an expanding universe simulation - fractal dimension 1.3 for Minkowski versus 1.1 for expanding - but share the same qualitative features. Lensing simulations show that for both point-like and extended objects, strings produce patterns unlike more traditional lenses, and, in particluar, the kinks in strings tend to generate demagnified images which reside close to the string. Thus lensing acts as a probe of the small scale structure of a string. Estimates of lensing probablity suggest that for string energy densities consistant with string seeded structure formation, on the order of tens of string lenses should be observed in the Sloan Digital Sky Survey quasar catalog. We propose a search strategy in which string lenses would be identified in the SDSS quasar survey, and the string nature of the lens can be confirmed by the observation of nearby high redshift galaxies which are also be lensed by the string.

The gravitational lensing by long, wiggly cosmic strings is shown to produce a large number of lensed images of a background source. In addition to pairs of images on either side of the string, a number of small images outline the string due to small-scale structure on the string. This image pattern could provide a highly distinctive signature of cosmic strings. Since the optical depth for multiple imaging of distant quasar sources by long strings may be comparable to that by galaxies, these image patterns should be clearly observable in the next generation of redshift surveys such as the Sloan Digital Sky Survey.

We calculate the deflection of a light ray caused by the gravitational field of a cosmic string loop in the weak field limit and reduce the problem to a single quadrature over a time slice of the loop's world sheet. We then apply this formalism to the problem of gravitational lensing by cosmic string loops. In particular, we find an analytic solution for the special case of a circular loop perpendicular to the optical axis. As examples of more complicated loops, we consider two loops with higher frequency Fourier modes. The numerical analysis illustrates the general features of loop lenses. Our estimates, using typical parameters for GUT scale loops, show that the stringy nature of loop lenses can be observed for lensing systems involving high redshift galaxies ($z \sim 2$), and we suggest that gravitational lensing can confirm the existence of GUT scale strings if they are the seeds for large scale structure formation.

We examine Gaussian isocurvature hot dark matter (massive neutrino) models for large-scale structure in which the initial density perturbations are produced in the baryons with a power--law spectrum $P_B(k) = Ak^{n_B}$. We calculate the linearly-evolved power spectrum and cosmic microwave fluctuations. We find that models with only isocurvature perturbations are inconsistent with observations of damped Ly$\alpha$ systems and COBE constraints on the power index. However, models which contain a mixture of adiabatic and isocurvature perturbations can be made consistent with COBE, galaxy surveys and damped Ly$\alpha$ systems.Isocurvature hot dark matter models also produce a bias between baryons and neutrinos even in the linear regime. We find that this ``natural bias'' can increase the baryon fraction in small scale objects like damped Ly$\alpha$ systems, but it has no effect on cluster scales.

We give an improved formalism for calculating the evolution of density fluctuations and temperature perturbations in flat universes. Our equations are general enough to treat the perturbations in collisionless relics like massive neutrinos. We find this formulation to be simpler to use than gauge dependent and other gauge-invariant formalisms. We show how to calculate temperature fluctuations (including multipole moments) and transfer functions, including the case of collisionless relics like massive neutrinos. We call this formalism "quasi-Newtonian" because the equations for the potential and cold matter fluctuation evolution have the same form as the Newtonian gravitational equations in an expanding space. The density fluctuation variable also has the same form inside and outside of the horizon which allows the initial conditions to be specified in a simple intuitive way. Our sample calculations demonstrate how to use these equations in cosmological models which have hot, cold, and mixed dark matter and adiabatic (isentropic) or isocurvature modes. We also give an approximation which may be used to get transfer functions quickly.

We consider the growth of density perturbations in the presence of self--interacting dark matter, SIDM, proposed by Carlson, Machacek and Hall (1992). We determine the range of values for the coupling constant $\lambda$ and the particle mass $m^\prime$, for which the power spectrum lies in the ``allowed" range based on constraints from the IRAS galaxy survey and damped Lyman--$\alpha $ systems. Our results show that no combination of parameters can meet both limits. We consider constraints on the $\2-2$ scatterings which keep the SIDM particles in pressure equilibrium, and we show that if such interactions maintain pressure equilibrium down to the present, they will be strong enough to disrupt galaxy mergers and may lead to stripping of galaxy halos as galaxies move through the dark matter background of these particles. Hence, we also investigate the evolution of large-scale structure in the SIDM model when the particles drop out of pressure equilibrium at some higher redshift. The resulting free-streaming leads to an additional suppression of small-scale perturbations, but it does not significantly affect our results.