Corrections of a paper by Wallace, 1966, Amer. Nat. 100: 551-563, a reexamination of data of Dobzhansky and Wright, 1943, Genetics 23: 304-340. Dispersion of released orange-eyed flies showed marked kurtosis after one day ([beta]2 = 8.9) which declined almost to normality ([beta]2 = 3) after 6 days as expected if dispersion is at random on each day. This is incompatible with the dispersion curve of constant kurtosis (about 22) assumed by Wallace. His combination of data from separate experiments with widely different rates of dispersion (in accordance with temperature) was probably responsible for the erroneous result. Another correction concerns the statement that the authors assumed normal dispersion in estimating the population density of the wild (red-eyed) flies of the region. The actual dispersion was used.