tgrt] BROWN & SHARP—EPIPACTIS AAI 
pylar end distinct walls and nuclear membranes are present. 
The wall which appears on the fibers of the double spindle cuts 
off one nucleus in the base of the sac, and leaves one free in the cyto- 
plasm (fig. 15). In these cases the latter nucleus apparently fuses 
with the polar nucleus from the typical micropylar group, while 
the one cut off by the wall later disorganizes (fig. 16). That sucha 
six-nucleate condition offers no hindrance to fertilization is evi- 
denced by the sac represented in fig. 16, in which a two-celled 
embryo is present. 
In fig. 11 the chalazal spindles have taken up a nearly parallel 
position, but at too late a stage to coalesce, since membranes are 
already present about the four nuclei. It is probable that in this - 
case a continuous wall would be formed across the base of the sac, 
cutting off two nuclei and leaving two free in the cytoplasm, but 
no sac was observed in which such an end had been reached. These 
Phenomena appear in most cases to be in some way associated 
with a narrow configuration of the chalazal end of the sac at this 
time, and a consequent diminution in the amount of cytoplasm 
present there. 
In other cases the fate of the megaspore mother cell is quite 
different from that described in the foregoing account. After 
enlarging somewhat the nucleus divides, the spindle lying at about 
the center of the cell, so that the thin wall formed upon the fibers 
Separates the mother cell into two nearly equal daughter cells 
(fig. 17). The wall, however, soon disappears, leaving the two 
nuclei in a single cell cavity which is to form the embryo sac. 
Between the nuclei vacuolation occurs, so that the center of the 
Sac comes to be occupied by a single large vacuole, the two nuclei 
taking up positions at opposite ends of the sac, where the greater 
part of the cytoplasm lies (fig. 18). Aside from the conspicuously 
larger size of the sac and its nuclei, this stage is closely similar to 
the corresponding one of a sac derived from a single megaspore. 
It is important to note that here the epidermal layer of the nucellus 
could be seen to be everywhere in contact with the sac, degenerat- 
ing cells being clearly absent (cf. figs. 6 and 18). 
The two nuclei again divide, and delicate walls appear on the 
spindle fibers between each pair of resulting nuclei. Later both 
