The distributions of some statistics arising from the comparison of two populations share a common structure.Consequences of this structure are derived and their usefulness indicated.The examples of statistics to which the theory applies include various correlation coefficients.