Bot. — Vol. I . ] CA NNON—A VENA FA TUA . 3 5 1 



capable of being absorbed by the embryo in its germination. 

 My attention has been called to the great similarity in 

 appearance and in function between the epithelial cells of 

 the scutellum of grasses and the cells at the tips of haustoria 

 in certain phanerogamic parasites. The likeness between 

 them and the epidermal cells in the haustorium of the Yel- 

 low Dodder is striking (Peirce, 1893). Both are epidermal 

 cells modified to extract food from foreign tissues, and both 

 exude solvents which can dissolve carbohydrates and pro- 

 teids. A further and more careful comparison would be of 

 importance, and might lead to some definite knowledge as 

 to the reason of the peculiar parasitic habit assumed by the 

 cotyledon of grasses. 



As the region below the stem-apex increases in length, 

 the epiblast, which was at first connected intimately with 

 the plumule-sheath, becomes more and more removed from 

 it toward the radicle end of the embryo. The epiblast 

 increases somewhat in size although it is never as conspicu- 

 ous an organ in Avena as in many other grass embryos. 

 The epiblast is composed wholly of large parenchymatous 

 cells and has no traces of conducting tissue. 



VII. Summary. 



The principal points in this paper may be summarized as 

 follows : — 



I. (a). The spikelet branches sympodially, the flower 

 and finally the ovary terminating the parent 

 axis. 



(b). The lodiculae and the upper palet originate in 

 the same rudiment. 



(c). The awn of the lower glume appears before 

 its lamina, the latter being an outgrowth 

 from it. Thus the origin of the parts of the 

 lower glume resemble that of the blade and 

 ligule of the vegetative leaf. 



(d). There is but a single carpel. This originates 

 on the anterior side of the floral apex, which 

 it encircles and finally encloses. 



