Of the several theories for the origin of the ovule advanced in this century, based largely on fossil evidence, the telome concept and double integument theory are accepted as the most plausible. Derivation of the integument through fusion of sterile telomes is the salient feature. For all theories, a common point of agreement is that the entire nucellus is a megasporangium. Evidence from both Paleozoic and extant ovules indicates that some of the integumentary telomes were fertile; that the nucellus is a sporangiophore of fertile telome origin; and that among all features cited to characterize ovules, the unique nature of the megaspores alone is definitive. Changes in the megaspore that extended the period of nutrient absorption over the course of female gametophyte development were probably achieved in a Late Devonian population of progymnosperms, indicating a monophyletic origin for seed plants. Ancient primitive features in ovule structure recur in genetically transformed plants and in natural populations of advanced taxa as well. In ovules of the bel1-3 mutant of Arabidopsis thaliana, a single integument takes the form of two or three columnar axes structurally similar to the integumentary axes of Devonian ovules. A single homeotic recessive gene derived by mutation of its dominant allele expresses this departure from the typical bitegmic form. The phenomenon suggests that expression of primitive form may be suppressed over the course of ovule evolution and later reinstated by homeotic gene mutation.