Amongst some maize seedlings grown for class use, one was observed to have two embryos emerging from a single caryopsis (P1. i A, B). Because its appearance immediately caused speculation-as to the exact relationship of the embryos to each other, it was pickled and sectioned (Text-fig. i A-J). The two embryos appear to be completely separate except for the scutellum, which offers an interesting condition. At the top and bottom (Text-fig. i A, B, C and J) it is completely separated into two parts, each being quite normal in appearance. In the region of theunion of the epicotyl (Text-fig. i D, E, F and G) it is entire but indented, the epidermis being itifolded to a considerable degree. At one point (Text-fig. i H) there is absolutely no sign of any bisection. From the section illustrated in J in Text-fig. i, it appears that both radicles were originally enclosed in a single coleorhiza. Although the right-hand embryo is placed a little lower in the caryopsis (P1. i A, and Text-fig. i A, E and G) the two appear to have been amazingly evenly balanced in their development and subsequent germination, and to&have proved very well-matched rivals for the food supplies. Even as late in development as when photographed (P1. i A) they were still practically mirror images. This even development, the way in which the scutellum seems to have arisen as an essentially single unit and the obvious singleness of the endosperm, testa and pericarp, suggest that the twinning arose early in the life history, probably being caused by longitudinal division or constriction of the more or less spherical mass of cells constituting the 'pro-embryo' stage of Soueges (1924). There appears to be only a very limited literature bearing on the structure of twin seedlings in grasses. Kempton (I913) deals, with a number of maize caryopses which have obviously arisen from two ovaries showing various degrees of 'fusion'. In extreme cases he shows examples of caryopses bearing embryos 'back to back' so that a germ appears on both sides of the grain. These have come from a single spikelet having two flowers, one of which is normally aborted but can, as on this occasion, be' fertile. He thinks of his grains as resulting from the fusion of the two ovaries after fertilization, but his illustrations, even of the mature caryopses, would suggest that the union, if such it ever is, has arisen much earlier. He concluded, however, that 'thedevelopment of the two ovaries in one spikelet must be simultaneous, as a large number of cases have been found where the two seeds from one spikelet have grown together with a single pericarp. These connate seeds had been fertilized through a double silk which was attached to the pericarp near the union of the two seeds. Connate seed& are a distinct phenomenon from single seeds with