Very local patterns of adaptation and differentiation are now known from many plant populations. These patterns contribute to the great ability of plants to persist in differing physical and biotic environments (cf. Clausen, 1951; Kruckeberg, 1951; Antonovics et al., 1971; Jain, 1976; Warwick and Briggs, 1978; Turkington and Harper, 1979). The ability of a plant population to track such local differences is a function of its genetic structure, determined in part by an interaction among gene flow, selection, stochastic processes in the environment, and aspects of the mating system. The relative roles of these interacting factors in allowing plant populations to persist in different microhabitats remain to be understood (Jain and Bradshaw, 1965; Raven, 1979), because they are difficult to unravel where many factors may coincide during any one year in shaping gene frequency changes for later years. Very localized pollen flow appears to be a major component of local adaptation in plants (Epling, 1947; Proctor and Yeo, 1972; Levin and Kerster, 1974; Jain, 1976; Levin, 1979, 1981). This has been measured by both the direct methods of mapping progeny genotypes and measuring the amount of exogenous pollen grains on stigmas (Bateman, 1947a, 1947b; Nieuwhof, 1963; Levin and Kerster, 1969; Richards and Ibraham, 1978; Schaal, 1980), and by the indirect methods of tracking the flight patterns of pollinators and labelling pollen (Levin and Kerster, 1968, 1974; Handel, 1976; Reinke and Bloom, 1979). The specific pattern of pollen flow changes with weather, taxa in the pollinator pool, and the floral biology of the plants (Free, 1966; Linhart, 1973; Levin and Kerster, 1974; Beattie, 1978; Levin, 1979; Schmitt, 1980). Differences in pollen dispersal dynamics among populations could be a factor in subsequent evolutionary change. Some such cases are known where adjacent conspecific populations have different breeding systems or pollen flow dynamics (Watson, 1969; Antonovics, 1971; Eisikowitch and Woodell, 1975). In the cases of heavy metal-tolerant populations, for example, strong selection is implicated in creating the different gene flow patterns. I have recorded pollen flow dynamics in two adjacent plantings of the same species where there is no difference in the physical environment, to test whether pollen flow dynamics as measured in one population are sufficient to describe pollen flow in an adjacent population. In addition to using a test garden situation where one can standardize arrangement of plants and cultivation conditions, I have also used plants of a single cultivar to minimize differences in phenology, physiology and growth among the plants. Differences and similarities in pollen flow as factors creating local variation in this experiment would not be affected by such biotic and physical variables. Consequently, pollen movement as an element of the microevolutionary process can be isolated, then used to interpret findings from natural populations where the influence of pollen flow dynamics is often obscured by other ecological components of evolutionary change.