Abstract
Data from lizards on population-size variation over time are compared to similar data from other kinds of organisms surveyed by Connell and Sousa (1983). Criteria of turnover of individuals similar to that used by Connel and Sousa, but which employ survivorship data only from the population in question, are proposed and used here. Lizard populations are substantially less variable through time than most of Connell and Sousa''s populations, the main exception being those of birds and of parasites. The significance of empirical population-size variability for evaluating certain theoretical constructs is briefly discussed.