Temperature maps are presented of the 9 largest clusters in the mock catalogues of Muanwong et al. for both the Preheating and Radiative models. The maps show that clusters are not smooth, featureless systems, but contain a variety of substructure which should be observable. The surface brightness contours are generally elliptical and features that are seen include cold clumps, hot spiral features, and cold fronts. Profiles of emission-weighted temperature, surface brightness and emission-weighted pressure across the surface brightness discontinuities seen in one of the bimodal clusters are consistent with the cold front in Abell 2142 observed by Markevitch et al.

We have extracted over 400 clusters, covering more than 2 decades in mass, from three simulations of the TCDM cosmology. This represents the largest, uniform catalogue of simulated clusters ever produced. The clusters exhibit a wide variety of density-profiles. Only a minority are well-fit in their outer regions by the widely used density profile of Navarro, Frenk & White (1977). Others have steeper outer density profiles, show sharp breaks in their density profiles, or have significant substructure. If we force a fit to the NFW profile, then the best-fit concentrations decline with increasing mass, but this is driven primarily by an increase in substructure as one moves to higher masses. The measured temperature-mass relations for properties measured within a sphere enclosing a fixed overdensity all follow the self-similar form, T\propto M^{2/3}, however the normalisation is lower than in observed clusters. The temperature-mass relations for properties measured within a fixed physical radius are significantly steeper then this. Both can be accurately predicted using the NFW model.

The alignment of clusters of galaxies with their nearest neighbours and between clusters within a supercluster is investigated using simulations of 512^{3} dark matter particles for \LambdaCDM and \tauCDM cosmological models. Strongly significant alignments are found for separations of up to 15h^{-1}Mpc in both cosmologies, but for the \LambdaCDM model the alignments extend up to separations of 30h^{-1}Mpc. The effect is strongest for nearest neighbours, but is not significant enough to be useful as an observational discriminant between cosmologies. As a check of whether this difference in alignments is present in other cosmologies, smaller simulations with 256^{3} particles are investigated for 4 different cosmological models. Because of poor number statistics, only the standard CDM model shows indications of having different alignments from the other models.

The effects of differences between the local and global values of the Hubble parameter on the cosmologies consistent with studies of high-redshift Type Ia supernov\ae are discussed. It is found that with a local Hubble parameter around 10 per cent higher than the global value then open cosmological models (such as $\Omega_{M} = 0.3, \Omega_{\Lambda}=0$) are prefered and if the local value is around 20 per cent higher then standard cosmological models ($\Omega_{M} = 1, \Omega_{\Lambda}=0$) can be recovered. Even in the case where the Hubble parameter ratio is 1, low $\Omega_M$ open cosmologies with $\Omega_{\Lambda} = 0$ are not rejected at the 95 per cent confidence level.