A hypothesis is presented which states that flower production in hermaphroditic flowering plants is primarily controlled by male function. The male function hypothesis predicts a lower fruit—to—flower ratio for hermaphrodites as compared to monoecious or dioecious plants. The hypothesis also predicts that self—compatible hermophrodites should exhibit a higher percent fruit—set than self—incompatible hermaphrodites. These predictions are supported by fruit—set data complied from the literature. An alternative hypothesis relating fruit—set to the probability of self—fertilization also predicts low fruit—set for hermaphrodites as compared to monoecius or dioecious plants. The self—incompatibility hypothesis is tested and rejected on the basis of fruit—set patterns in self—incompatible andromonoecious, self—incompatible monoecious, and self—compatible monoecious species. The effect of the male function hypothesis on current ideas concerning low fruit—set in hermaphrodites is then examined.
