( 350) 
and is separated from the epidermis by three tabular tapetal 
cells. The nucellus at this time grows rapidly and the in- 
teguments cover only the lower half of it. These arise in 
the usual manner and by the time the embryo-sac is ready 
for fertilization they have closed over the top of the nucellus. 
Until some time after fertilization the outer integuments are 
relatively thinner than those of the other cases described, 
except Fevillca. The inner integument is always thin. 
The division of the spore-mother-cell to form the usual 
potential megaspores was not observed, though there is evi- 
dence that it occurs. In stages somewhat later, certain dis- 
integrating masses were observed above the young embryo- 
sac, which were taken to be non-functional spores (fg. 65). 
An embryo-sac of two cells was observed, but none between 
this condition and fertilization. When the odsphere is mature 
the embryo-sac differs in some respects from the preceding 
forms. Figures 66 and 66a@ represent parts of the same 
embryo-sacs. The synergids are much less vacuolated and 
are decidedly hyaline and striated in their upper portions. 
The odsphere is likewise more compact and the whole egg- 
apparatus is larger than in most other cases examined. No 
antipodals could with certainty be determined. If formed at 
all they must disappear quickly. 
The endosperm nucleus is formed by the fusion of two 
polar nuclei. The nucleus thus formed is conspicuous for 
its size and chromatic content. Each polar nucleus has a 
very large nucleolus and these evidently become merged, as 
only one could be found in the endosperm nucleus. After 
fertilization the endosperm rapidly increases and the cells of 
the nucellus are gradually displaced. No starch was observed 
in the endosperm of Momordica at any time. As the endo- 
sperm grows, most of its substances become peripherally dis- 
tributed with numerous small nuclei scattered through it. 
This condition is maintained for some time, while the expand- 
ing surface of the endosperm is gradually replacing the sur- 
rounding tissue. Sometimes a thin line of cytoplasm may be 
seen penetrating the tissue toward the chalaza (jig. 68). 
