In this thesis we make a new measurement of the primordial deuterium abundance, and analyse five other systems selected as possible D/H candidates. We also undertake an investigation of systematic errors in a system where an existing deuterium measurement has been made. We measure the number of hydrogen components and their velocity distributions in a moderate and high redshift sample of Lyman limit systems in one Angstrom resolution spectra. We present a new measurement of the deuterium-to-hydrogen ratio, D/H, at z=3.2560 in a newly-discovered low metallicity absorption system towards the quasar PG1937-1009. We attempt to account for any systematic effects that could influence the D/H measurement. We find a 1 sigma range for D/H*1e5 of 1.6 (+0.25) (-0.30). Using high resolution spectra from the Hubble Space Telescope and the Very Large Telescope, we assess the suitability of five D/H candidate absorption systems. These systems were selected as candidates using lower resolution spectra. We measure the neutral hydrogen column density, identify metal lines and analyse the velocity structure of each system, and show them to be unsuitable for measuring D/H. We also investigate the systematic errors in the absorber at z=0.701 towards quasar PG1718-4801 that was initially thought to show a high primordial D/H value. We analyse the dependence of the putative deuterium line's parameters on wavelength calibration errors in the HST spectra and present a revised deuterium measurement. We examine the velocity widths of two samples of one Angstrom resolution quasar spectra showing Lyman limit absorption systems. The first sample is at high redshifts, taken from the Sloan quasar data release 3 catalogue. The second is at intermediate redshifts, compiled from a survey for UV quasar absorption systems taken with the Hubble Space Telescope. We use a modified version of the Voigt profile fitting program, VPFIT, to estimate the number of hydrogen velocity components and column density in the Lyman limit systems. We compare the velocity distributions of the higher and lower redshift samples. We find the distributions are consistent with other measures of the velocity spread in absorption systems, and find no compelling evidence for evolution between the redshift samples.

In this thesis we investigate the large-scale distribution of Ly alpha forest absorption, the effect of ionizing radiation from QSOs on their surrounding intergalactic medium and the primordial abundance of deuterium. We develop a new technique for detecting structure on Mpc scales in the Ly alpha forest. This technique does not rely on identifying individual absorption lines but is rather based on the statistics of the transmitted flux. We demonstrate that the new method is significantly more sensitive to the presence of large-scale structure in the Ly alpha forest than a two-point correlation function analysis. We apply this method to 2 A resolution spectra of ten QSOs which cover the redshift range 2.2 < z < 3.4. The QSOs form a closely spaced group on the sky and are concentrated within a 1-deg^2 field. We find evidence for large-scale structure in the distribution of Ly alpha forest absorption at the > 99 per cent confidence level. Along the line of sight we find overdense Ly alpha absorption on scales of up to 1200 km s^-1. There is also strong evidence for correlated absorption across line of sight pairs separated by < 3 h^-1 Mpc. For larger separations the cross-correlation signal becomes progressively less significant. Using the same technique and dataset we confirm the existence of the proximity effect. We derive a value for the mean intensity of the extragalactic background radiation at the Lyman limit of J = (3.6^+3.5_-1.3) x 10^-22 ergs s^-1 cm^-2 Hz^-1 sr^-1. This value assumes that QSO redshifts measured from high ionization lines differ from the true systemic redshifts by Delta v = 800 km s^-1. Allowing for known QSO variability we find evidence at a level of 2.1 sigma that the significance of the proximity effect is correlated with QSO Lyman limit luminosity. From the complete sample we find no evidence for the existence of a foreground proximity effect, implying either that J > 20 x 10^-22 ergs s^-1 cm^-2 Hz^-1 sr^-1 or that QSOs emit at least a factor of 1.4 less ionizing radiation in the plane of the sky than along the line of sight to Earth. We do, however, find one counter-example where a foreground QSO apparently depletes the absorbing gas in four surrounding lines of sight. We discuss the feasibility of pre-selecting absorption systems from low resolution data for a measurement of the primordial deuterium abundance. We present a new, low resolution spectroscopic survey of 101 high redshift QSOs aimed at identifying candidate D/H systems. We further present an echelle spectrum of a Lyman limit system at z = 2.917. We find that this system is most likely heavily contaminated and does not yield an interesting limit on D/H.