A chance section of the young seed of Pontederia made several years ago presented a peculiar appearance in the antipodal end of the endosperm that led me to make an investigation of seed development in this family. I have been able to study Pontederia cordata and Heteranthera limosum in some detail, and Eichhornia crassipes to a less extent. Collections of Pontederia have been made in South Carolina and New York for several years. All my Heteranthera and Eichhornia material was collected in Jamaica during August i900. Examination of the literature of this family shows but one title dealing with its embryology. In i898 SMITH (14) published a rather full account of seed development in Eichhornia and Pontederia, and gave one figure of the embryo sac of Heteranthera. Oddly enough, he completely overlooked the interesting peculiarity that appears in the endosperm of all three of the above genera, but otherwise, where parallel, my work, with a few exceptions, agrees with his. As SMITH did not attempt to give the development of Heteranthera, I have made a careful study of its ovular growth as a comparison with his work on Eichhornia and Pontederia. In the last two genera I have confined myself to the origin and development of the endosperm, which has the same peculiarity in all three types. HOFMEISTER (g) devotes a few lines to the development of Pontederia, but does not mention two sorts of endosperm.