40 ' BOTANICAL GAZETIE [yuLy 
Meanwhile the lower of the two cells formed by the first 4 
division of the hypodermal cell has undergone considerable — 
growth. It rapidly becomes several times its original size, and 
extends nearly to the base of the nucellus; but the increase in 
breadth is not so great. Just previous to the first division of its + 
nucleus the cell is unusually long and narrow, but is completely 
filled with cytoplasm (fig. 29). 4 
The cell gradually becomes longer and longer until at length _ 
a division takes place whereby two daughter cells are formed 
(fig. 31). After a short time both of these daughter cells divide _ 
again simultaneously. As a result we have an axial row of four _ 
cells reaching from the base to the apex of the nucellus (fg. 
32). In the upper three cells a change is noticed almost at 
once, whereby the cytoplasm becomes denser and darker, the 
nuclei less chromatic, and the cells in fact show evident signs of — 
disintegration (fig. 33). The lower cell by its continued growth 
gradually compresses the other three, which very soon are all 
crowded into a small disorganized mass in the micropylar region — | 
( fig. 34)-0-A large number of sections was obtained showing — 
all stages of this process in the plainest manner. The lower- — 
most cell alone finally takes part in the formation of the embryo- E 
sac. The process in Canna, therefore, is exactly in accord with 4 
that in the Iridacez, Rosacee, Polygonacez, and many Ranun- 
culaceze and Liliacez as described by Strasburger and others. — 
Owing to the extremely narrow cavity and small nuclei, the 
changes within the embryo-sac are very difficult to follow. 
Humphrey” has already given a full discussion of the literature 
on the embryo-sac of Canna; and also the results of his owl 
investigations on the same plant. In this case the antipodal 
cells were not discovered, and he came to the conclusion, as did 
also Guignard*} that these cells although always formed must 
disintegrate immediately, 
Sie todeiahds The development of the seed in the Scitamines. Annals of Bot 
10:1, 1896. 
*3GUIGNARD: Recherches sur le sac embryonnaire des Phanerogames Angio- 
Spermes. Ann. Sci. Nat, Bot. VI. 13: 136. 1882. 
‘ 
