Ovule and seed development in six species of Nymphaeales was examined. In the Cabombaceae the two species studied resemble some extant basal angiosperms by having a hood-shaped outer integument. A micropyle-hilum complex results when the outer integument and derived testa are lacking between the micropyle and the funiculus, thus the hood-shaped appearance. In the Nymphaeaceae the outer integument is annular at an early stage and then cup-shaped though it is semiannular at initiation in Nupar japonicum and Nymphaea alba. The micropyle and hilum are separated by an intervening testa. Developmental data on the formation of the outer integument, from semiannular to hood-shaped vs. from annular to cup-shaped, are useful for inferring the morphology of the outer integument from the relative position of the micropyle to the hilum in seed fossils. The oldest (early Cretaceous) probable nymphaealean seeds had the micropyle-hilum complex, suggesting that the hood-shaped outer integument may be primitive in the Nymphaeales. This needs to be tested by examination of this feature in other groups of basal angiosperms.Key words: developmental morphology; evolution; integument; micropyle-hilum complex; ovules; Nymphaeales; seeds.Bitegmic and anatropous ovules are common in basal angiosperms and are usually considered to be primitive in flowering plants as a whole (Stebbins, 1974;Bouman, 1984;Cronquist, 1988;Takhtajan, 1991;Johri, Ambegaokar, and Srivastava, 1992). Only a few authors regard anatropy as derivative and orthotropy as ancestral (Bocquet and Bersier, 1960). It has been suggested that the outer integuments are cup shaped and asymmetrical (Bouman, 1984). However, recent developmental and morphological studies have shown that in many basal angiosperms the outer integument is sharply asymmetric and hood shaped, that is, it is lacking on the concave (funicular) side of the ovule (Matsui, Imaichi, and Kato, 1993;Umeda, Imaichi, and Kato, 1994;Imaichi, Kato, and Okada, 1995;Igersheim, 1997, 1999;Endress, 1997, 1998; see also Johri, Ambegaokar, and Srivastava, 1992).The origin of bitegmy has been variously interpreted in contrast to the agreement on the origin of the first (inner) integument by fusion of telomes (Herr, 1995). Crane (1985Crane ( , 1986 and Doyle and Donoghue (1986, 1987) postulated that the outer integument is derived from the leafy lamina of a Caytonialike ancestor, while the carpel enclosing ovules is derived by lateral expansion of its leaf axis. It is also hypothesized that the bitegmic ovules have been derived from unitegmic and orthotropous ovules of glossopterid-like gymnosperms by the enclosure of the unitegmic ovules by the ovuliferous leaf to form the outer integument (Stebbins, 1974;Dahlgren, 1983;Kato, 1990;Stewart and Rothwell, 1993;Umeda, Imaichi, and Kato, 1994;Imaichi, Kato, and Okada, 1995;Doyle, 1996). Kato (1991) also speculates that the anatropy is an extreme modification of the hyponastic curvature of leaves (ovuliferous leaf). However, the origin of the outer integument and the origin ...