This thesis describes the use of centrifugal photosedimentation, for the particle size analysis of soil clays and clay minerals. The instrument that was used is the Horiba, CAPA 500. It has a lower limit to particle size analysis for clays of 20nm (Stokes' diameter) and gives a high degree of size resolution. The initial output is a relationship between turbidity and Stokes' diameter over a limited range of particle sizes. Use of the centrifugal photosedimentometer with materials containing a broad range of particle sizes, such as clays, by combining results from runs with overlapping size ranges is described. In order to maximize sensitivity it was necessary to use higher initial suspension concentrations when analysing for small particles, compared to those used when analysing for larger particles. Typically clays contain particles that are both greater and smaller than the wavelength of light in the dispersing medium (400 nm in this case). It is therefore important to correct for the variation in light extinction with particle size and those aspects of light scattering and absorption theory that are relevant to photosedimentation data analysis are reviewed. For particles that are smaller than 100nm absorption makes a significant contribution to light extinction. Therefore the absorption coefficients of a range of contrasting clays were estimated from reflectance measurements taken from diluted clays (in powder form). These samples provided a set of standards that allowed the absorption coefficients of other clays to be estimated by visual comparison. In very force fields disc-like particles sediment so that they are oriented with their minimum dimension in the direction of motion. Light scattering is sensitive to particle orientation, hence photosedimentation was used to estimate the orientation state of sedimenting particles. Estimates of the maximum advisable centrifugal speed for use with the various particle sizes are given. These were obtained from both simple theory and from experimental results. Although repeatable results were obtained from photosedimentation, the particle size distributions calculated from the relationship between turbidity and Stokes' diameter did not agree with those from gravimetric size analysis, when traditional theories (e.g. Mie) were used to correct the experimentally obtained turbidity data prior to the calculation of particle size distribution. However, satisfactory semi-empirical scattering efficiencies are suggested for the 100nm to 10 μm size range, but they need further testing. Photosedimentation analysis produced the same shape of particle size distribution for TEM size analysed samples, however the positions of these size distributions in terms of mean particle size did not always coincide. Despite the various sources of uncertainty in the particle size distributions that are obtained from photosedimentation data analysis, the high degree of size resolution and good repeatability of the instrument make it suitable for many applictions in Soil Science. Centrifugal photosedimentation was found to be particularly useful for studying the effect of saturating cation on the particle size distributions of dispersed bentonite. Further work is suggested to extend the range of semi-empirical values of extinction efficiency that are given and also to quantify the biasing that can result from samples which have a variation of axial ratio and/or chemical composition with particle size.