In an area of felled-woodland subsere, where biotic plagio-climax had become established, ant conditions had probably been favourable for about 20 yrs. The four spp. present were distr. in different proportions on and off rotting stumps, and that characteristic spp. positions on the stumps were also established. By placing artificial sites[long dash]stones, slates, or banks or turf[long dash]in such a way as to encourage reaction, it was established that a sp. replacement order existed, only reversed in cases of extreme disparity in colony size or periodic unsuitablity of the habitat. Replacement was either by direct aggression (in the nest site, or by seige parties from a distance) or by substitution during temporary absence caused by either drought or cold. Different methods were characteristic of different spp., but some latitude existed. Even the foundress queens had a sp. replacement order, but the degree to which this operated in normal conditions was probably slight. The period of vegetational stability had lasted long enough to enable a development from an initially near-random dist. in the favourable regions-primary loci of colonization[long dash]to a sp. differentiated distr. over the whole habitable area. This arose in a heterovalent environment through the existence of a replacement order, differences in architectural capacity, and differences in climatic tolerance.