Populations of I. elegans (Vander L.) were studied during 1965, 1966, and 1970 by means of the capture-recapture method at three ponds in Dunham Park, North Cheshire, England. The results seem to suggest that neighbouring imaginal populations between which there is little interchange may each have their own specific survival rates, which may be largely determined by the geographical features, vegetation and specific predators, and which are not directly related to the numbers of imagines present.