Experiments with Medicago tribuloides and Lactuca sativa confirm Sedgley's conclusion that the wetted area of contact is a factor controlling germination of the seed. The evidence suggests that this is of consequence at matric potentials near that of free water and is most important for germination of the last seeds in each population. Matric potential is shown to be an important factor in seed germination over and above its effect in controlling the wetted contact area of seed and medium. Prewetting the seed at matric potentials near that of free water for a few minutes increases the germination rate markedly and minimizes the normal effect of matric potential. The results and experimental conclusions are discussed and are shown to indicate the need for further investigation. They show that it is practicable to design experiments to distinguish the effects of the various facets of soil water energy on seed germination behaviour.