In a recent paper, Wallace (1966) re-examined data on the dispersion of Drosophila pseudoobscura reported by Dobzhansky and Wright (1943). He found that "the log of the number of flies recaptured at various distances from the point of release (both for capture of individual days following release and, for practical purposes, the cumulative captures of all days) decreases linearly with the square root of the distance." There is serious inconsistency with the results of the 1943 analysis, which is not immediately obvious and not referred to by the author. In each of four experiments (I-IV), several thousand orange-eyed flies were released at the center of a cross-shaped array of traps, 20 m apart, in an area inhabited by wild red-eyed flies of the same species. The numbers of both sorts, caught in the traps, were recorded daily. The flies dispersed so widely that it became necessary on the second or third day to drop two arms of the cross in order to extend the others. After demonstrating that temperature was a major cause of differences in activity, both in percentage trapped and amount of dispersion, an attempt was made to obtain some idea of the nature of the dispersion for each day in each experiment by calculating the kurtosis. The formula used was