and Fertilization in Oenothera. 
3 i 3 
uppermost embryo sac is derived from a different mother-cell, while the 
middle and lowest sacs have a common origin from another one. 
Summary. 
1. The embryo sac arises from either micropylar or chalazal one 
of the tetrad; often both of them simultaneously develop into complete 
embryo sacs. 
2. The embryo sac is tetranucleate, lacking the antipodals and one of 
the pole nuclei. 
3. The pollen-tube enters the synergid through the filiform apparatus 
and the mixed plasma flows out through the synergid and spreads over 
the oosphere. 
4. The plasma of the pollen grain is impregnated with an immense 
number of minute fusiform starch grains, which migrate through the pollen- 
tube, enter the synergid, and finally become entirely consumed. 
5. The male nucleus is enclosed in a distinct plasma sheath, until 
it reaches the oosphere. 
6. One of the male nuclei fuses with the pole nucleus, and gives rise to 
the endosperm nucleus with diploid number of chromosomes. 
7. Synergids and upper two-thirds of the oosphere have a distinct 
cellulose membrane which sometimes contains also pectic substance. The 
lower part of the egg-cell acquires a cellulose membrane after fertilization. 
8. Self-sterility of some hybrids is due to a feeble growth of the 
pollen-tube. 
9. Besides Oenothera , Ludwigia , Ganra , Godetia , and Circaea also 
produce tetranucleate embryo sacs, which is a diagnostic character of the 
Onagraceae. On the other hand, Trapa , which has a normal 8-nucleate 
embryo sac, would better be separated from the Onagraceae. 
10. The chalazal tissue is nearly occupied by some chromatic substance 
which appears to be some ultimate substance. Many starch grains accumu¬ 
late in the young nucellar tissue before the chromatic substance makes an 
appearance. 
11. In other genera examined of the Onagraceae, the nucellar tissue is 
nearly or entirely devoid of deposition of starch grains, correlated with the 
absence of the chromatic substance. 
12. Tetranucleate condition in the Onagraceae may have been pro¬ 
duced by mutation, but not by adaptation. No relation exists between the 
tetranucleate condition and the presence of the chromatic substance. 
13. The species with 4- or 16-nucleate embryo sacs are mostly 
herbaceous. 
The writer is deeply indebted to Prof. K. Fujii and Prof. K. Shibata 
for their kind advice and suggestions. Acknowledgements are especially 
