THE EMBRYO-SAC OR FEMALE GAMETOPHYTE. 171 
rise to the axial row of four cells, correspond, as in other Gymno- 
sperms, precisely with the first and second divisions in the microspore 
mother-cell of this plant. In my own opinion the only legitimate 
conclusion to be drawn from this morphological and cytological evi- 
dence is that the macrospore mother-cell of Zarzx is homologous with 
that of Heleborus and other Angiosperms in which the embryo-sac 
develops similarly. 
In the development of the embryo-sac, as typified by Zzléum and 
many other monocotyledonous plants, the hypodermal cell does not 
produce an axial row of four cells, but becomes at once the functional 
macrospore. With the growth of the nucellus this hypodermal cell 
increases greatly in size, as does also its nucleus (Fig. 71). The 
nucleus, after its characteristic period of 
growth, divides heterotypically. The two 
resulting daughter-nuclei lie in the ends of 
the cell. No cell-division follows this 
nuclear division, although the thickening of { 
the connecting fibers in the equatorial region { 
seems to indicate that a tendency toward cell- | 
division existed (Fig. 72, A). The macro- 
spcre continues its growth, and the daughter- 
nuclei divide. This division is homotypic 
and corresponds exactly to the second mitosis 
in the pollen mother-cell. The four resulting 
nuclei have, as a rule, the orientation shown 
‘ Fic. 71 —Embryo-sac mother-cell 
in Fig. 72, B. Very frequently no vacuole of Lilium martagon with nu- 
cleus showing beginning of 
is present at this stage, and the four nuclei are SEsphANs GF duvteton, 
connected with each other and with the plasma 
membrane by systems of kinoplasmic radiations and connecting fibers. 
The increase of the cell in length is now rapid, and, as a result, one 
or more large vacuoles are formed at the center or near the micropylar 
end of the sac. Two of the four nuclei which are sisters move into 
the upper, and the other two into the lower end of the cell. In normal 
cases the nuclei in each end divide so that a group of four nuclei occu- 
pies each end. The four nuclei in the micropylar end are arranged 
either in a plane, or nearly so, or in the form of a tetrad (Fig. 73, 
A,B). The arrangement and behavior of the nuclei in the chalazal 
end of the sac is more variable (Mottier, ’97). 
As a rule the two nuclei in the micropylar end of the sac, and it is 
with these that we are especially concerned, divide simultaneously, and, 
before cell-plates are laid down, the four resulting nuclei are connected 
