Martin placed the lateral embryo, which occurs only in grasses, adjacent to the broad embryo at the base of his family tree of seed phylogeny. Since Poales and Poaceae are derived monocots, we questioned the evolutionary relationship between the lateral embryo and other kinds of monocot embryos. Information was compiled on embryo and seed characteristics for the various families of monocots, kind of embryogenesis for families in Poales and germination morphology of families with lateral (only Poaceae) and broad embryos. The kinds of monocot embryos are broad, capitate, lateral, linear fully developed, linear underdeveloped and undifferentiated, but only broad and lateral embryos are restricted to Poales. Asterad embryogenesis occurs in Poaceae with a lateral embryo and in Eriocaulaceae, Rapataceae and Xyridaceae with a broad embryo. In developing grass seeds, the growing scutellum (cotyledon) pushes the coleoptile, mesocotyl and coleorhiza to the side. In the organless broad embryo, the cotyledonary sector is larger than the epicotyledonary sector. During germination of grass seeds, the coleorhiza and then the coleoptile emerge, while in a seed with a broad embryo the elongating cotyledon pushes the epicotyledonary sector outside the seed, after which a root–shoot axis is differentiated at a right angle to the cotyledon inside the seed. Broad and lateral embryos are closely related; however, the lateral embryo is more advanced in seed/embryo traits and germination morphology than the other kinds of monocot embryos, suggesting that its position on the family tree of seed phylogeny should be higher than of the other monocot embryos.

Rudimentary, broad and small linear embryos occur among members of the most primitive (basal) extant angiosperms, collectively called the ANITA grade (i.e. Amborella, Nymphaeales and Austrobaileyales). Amborella (rudimentary) and Austrobaileyales (rudimentary in Austrobaileyaceae, Illiciaceae and Schisandraceae and small linear in Trimeniaceae) have kinds of embryos that are known to be underdeveloped; consequently, they must grow inside the seed prior to radicle emergence (germination). On the other hand, it is not known if broad embryos need to grow before radicles can emerge, and whether they are underdeveloped or fully developed. Thus, we addressed the question: ‘Is the broad embryo of Nymphaeales also underdeveloped?’. Although the embryo length : seed length ratios in Nymphaea Albert Greenburg, N. capensis var. zanzibariensis and N. immutabilis were 0.311, 0.349 and 0.234, respectively, embryos did not grow prior to radicle emergence. Thus, they are fully developed at seed maturity. If Amborella and Nymphaeales are equally the most basal angiosperms, as some molecular phylogenetic studies indicate, then we must conclude that the broad and rudimentary embryos are equally primitive.