We present a technique to organize gauge invariant sums of tree level Feynman diagrams into string-like and suggestively gauge invariant expressions. String-like amplitudes for scalar QED are reviewed and the technique is extended to pure and scalar QCD processes. This allows us to explore gauge symmetry in a new light and understand some of the efficient organization of String theory amplitudes. These ideas can be used in the future to derive string-like expressions for multiloop processes with potential to guide String theory into unexplored territories and to enhance our understanding of gauge symmetries.

The purpose of this work is to use the spinor helicity technique in conjunction with supersymmetry to generate tree-level supersymmetric Ward identities for component fields in the pure gauge sector of a supersymmetric non-Abelian vector field. Using the operator formalism with supersymmetric transformations, relations are found among processes containing different numbers of gluinos. Explicit calculations are made for the four-point functions as a demonstration of this approach. To further separate these relations, off-shell calculations are made from which a diagrammatic algorithm is derived. This algorithm is shown to reproduce the standard relations among processes. Explicit calculations using the algorithm are then presented for the four-point functions. In addition, the algorithm is shown to separate the standard relations into subsets of identities.

The substantial recent advances in early universe observational cosmology have confronted theoretical physics with a number of deep conceptual questions. The standard model of particle physics and its natural extensions have proven to be insufficient for explaining the recent observations. String theory as the main candidate for a complete theory of quantum gravity provides us with the possibility of resolving this discrepancy. In this dissertation, I begin by an overview of thermodynamics in string theory. The issue of stability and the existence of the thermodynamic limit have been addressed. I will describe the string gas cosmology in detail, as the leading candidate of early universe cosmology. Successes and shortcomings of the most recent model of string gas cosmology in matching with observations are addressed. I conclude with comments on interesting directions for future study and a possible link with black hole physics. / A la lumiere des plus recentes observations cosmologiques, la physique theorique se trouve confrontee a de nombreuses questions conceptuelles. Le Modele Standard de physique des particules et ses plus simples extensions se sont averes insuffisants pour expliquer les recentes observations. La theorie des cordes, qui est presentement la principale candidate pour une theorie complete de la gravite quantique, nous offre la possibilite de resoudre ces lacunes du Modele Standard. Nous debutons cette these par un apercu de la thermodynamique dans la theorie des cordes. Nous nous penchons en particulier sur la stabilite et sur l'existence d'une limite thermodynamique en theorie des cordes. Nous presentons par la suite les details de la cosmologie des gaz de cordes et discutons des succes et des problèmes de cette approche pour decrire l'evolution cosmologique du debut de l'univers. Nous concluons en commentant sur les possibles travaux futurs et sur un lien entre ce sujet et la physique des trous noirs.

Data taken with the ARGUS detector at DESY, Hamburg, specifically tt pairs produced from e+e -- collisions in the energy range (10.4--10.6)GeV, are used to make a precision measurement of the tau lifetime. A new method is introduced which is independent of the beam position and envelope, and applicable to tau events having one-three topology. Applied to ARGUS data the method yields a value for the tau lifetime of tt=287+/-11st atistical+/-8systemat icfs.

It is found experimentally that total cross sections increase with energy. Partonic total cross sections can be calculated in perturbative QCD provided the coupling constant is small, though multiloop diagrams must be included at high energies. Such calculations are difficult and can usually be carried out only in the leading-log approximation. The resulting BFKL Pomeron violates the Froissart bound, which forbids total cross sections asymptotically to grow faster than the square of the logarithm of energy. To restore unitarity and the Froissart bound, subleading contributions of all orders must be included. Unfortunately existing techniques for computing Feynman diagrams prove to be inadequate for this difficult task. The purpose of this thesis is to develop new techniques capable of solving this problem. Since factorization in the impact-parameter space is a main ingredient needed to reach unitarity, such a technique must be capable of implementing factorization in an efficient way. We introduce the non-abelian cut diagrams for that purpose. We use it to compute quark-quark scattering amplitude to the two-loop order and show how the new technique can overcome the inadequacies of the existing method. We are also able to use this method to prove factorization of a class of Feynman diagrams, which we shall refer to as 's-channel ladder diagrams', though the proof of general factorization is not attempted in this thesis. We have also developed a more efficient method to calculate the high-energy dependences of individual Feynman and non-abelian cut diagrams. This method relies on a systematic study of the paths used by momenta to flow through the diagram, and will thus be referred to as the path method for flow diagrams. These new techniques can be used to implement unitarity and to restore the Froissart bound, but this final goal is not carried out here.

In this thesis, we examine the evolution of a quantum system through background matter in order to determine the conditions of validity required for a description in terms of an effective Hamiltonian. General scattering rate equations are obtained for coherent diffuse scattering. We find, as expected, that the diffuse scattering rate is small compared to the coherent scattering rate in a medium with low fluctuations allowing an effective Hamiltonian description. However, for a more turbulent background, it is still possible to have more significant contributions to the diffuse scattering which may affect the existing theory of neutrino flavor conversion (MSW effect).

Within the framework of perturbative QCD, two processes are considered: heavy flavor production (of open $Q bar Q,Q$ = c or b, and of heavy quarkonia) and large-$p sb{ Upsilon}$ direct photon production--both in proton-proton collisions with polarized beam and target. For the first process, the differential and total cross sections for the subprocesses $ vec g vec g to Q bar Q$ and $ vec q vec{ bar q} to Q bar Q$ are determined to order $ alpha sbsp{s}{2}$. For both processes, asymmetries are computed using two sets of polarized parton distributions differing in the magnitude of the gluon distribution. It is shown that, for both processes, experiment can well distinguish between a large and a small polarized gluon distribution.

Relativistic heavy ion collisions aim to study the quark-gluon plasma, a hot and dense state of matter in which quarks and gluons, which are the elementary constituents of nuclei, are set free from the forces that normally confine them. Because of the very short lifetime of the produced plasmas, to study them it is necessary to use probes produced by the collisions themselves. Particles carrying exceptionally large energies, so-called hard probes, are promising candidates to allow a precise reconstruction of the complete history and properties of the evolving plasma. In this thesis, by extending to next-to-leading order at weak coupling the calculation of scattering rates of high-energy jets, we identify an important source of theoretical uncertainties in existing models of their propagation through the medium, and we discuss how this could be improved in the near future. Considering energy loss of heavy quarks, we propose a novel nonperturbative method for quantifying it in the large mass limit, using numerical lattice simulations. Finally, we consider the prospects that hard processes in supersymmetric theories could enjoy a simplified description compared to the non-supersymmetric case, thus making these theories into useful mathematical tools. / Les collisions ultrarelativistes d'ions lourds visent à étudier le plasma de quarks et gluons, un état extrêment dense et chaud de la matière dans lequel les quarks et les gluons, les constituants élémentaires des noyaux atomiques, sont libérés des forces qui les confinent habituellement. Les plasmas produits ont une durée de vie très brève et doivent être étudiés à l'aide des particules produites lors de la collision elle-même. Les particles exceptionellement énergétiques, ou sondes dures, offrent la possibilité de reconstruire de façon précise l'évolution du plasma. Dans cette thèse, en étendant au second ordre perturbatif le calcul du taux de collision pour un jet de haute énergie, nous identifieront une source importante d'incertitude dans les théories existantes concernant la propagation des jets, et nous discuterons de possibles améliorations à moyen terme. Considérant l'énergie perdue par des quarks lourds, nous proposerons une méthode non-perturbative pour quantifier leur perte d'énergie dans la limite de grande masse, au moyen de simulations numériques en théorie de jauge sur réseau. Finallement, nous considérerons la possibilité que les processus durs à température finie dans les théories supersymétriques bénéficient d'une description simplifiée, par rapport au cas non supersymétrique, ce qui leur conférerait un intérêt mathématique accru.

A comparison of the CDF (Collider Detector at Fermilab) central calorimeter electron response between 1991 test beam electron data and the CDF full detector simulation, using six variables that are used to identify electrons in CDF, is performed. Possible parameters that could be used to tune the simulation are also pointed out.

We present an analysis of the requirements of a 2--3 exchange symmetry in the neutrino mass matrix in comparison to the experimentally allowed ranges of neutrino oscillation parameters. The symmetric matrix, being defined at an energy scale appropriate to a right-handed neutrino in a See-saw scheme of mass suppression, is subject to evolution under Supersymmetric Renormalisation Group Equations, in order to interpret the matrix at experimental energies. By way of motivation we discuss the status of neutrino mass in the Standard Model and justify the context of the analysis by examining the mechanisms and evidence for oscillations. We then review the See-saw mechanism and also the process of renormalisation and its implications for bridging disparate energy scales. We present the relevant Renormalization Group Equations and demonstrate the parameterisation of pertinent Renormalization Group effects. Finally, we review previous work analysing this symmetry before updating some of these results and extending the analysis to its global behaviour in the space of both solar and atmospheric neutrino oscillation parameters.