360 BOTANICAL GAZETTE [may 
tion in this region to demonstrate the egg-like naturé of these 
cells. Tretjakow (1895) reports an instance in which he finds an 
embryo in the antipodal region, but as he did not find fertiliza- 
tion he believed the embryo to have arisen apogamously. 
Of the two views held as to the homologies of the antipodals, 
namely, that they represent either a vegetative prothallus or that 
they are potential eggs, the discovery of the fertilization of an 
antipodal egg adds substantial evidence to the theory that all 
the cells of the embryo sac arepotentialeggs. If the antipodals 
are potential eggs, then it would seem that the fusion of one 
sperm nucleus with the two polar nuclei is, as Nawaschin and 
some other investigators believe, a true act of fertilization. 
After fertilization the egg nucleus becomes spherical but does 
not divide until sometime subsequent to the first division of the 
endosperm nucleus. The first division of the egg is transverse, 
as usual, and the subsequent development of the embryo differs 
little from that of other Compositae. 
SUMMARY. 
The embryo sac arises from the lowest cell of an axial row of 
four cells. 
The eight-nucleate embryo sac is formed in the usual manner. 
Great variation is shown in the form of embryo sacs, both in 
different species and in the same species. 
The polar nuclei fuse ordinarily before fertilization, but may 
fuse after it. 
The endosperm nucleus is marked by its very large size, its 
almost spherical shape, its large nucleolus, and its position below 
and in contact with the oosphere. 
The antipodals show a remarkable growth before the time of 
fertilization and persist very late in the development of the 
embryo. 
From one to nineteen nuclei were found in a single antipodal 
cell. 
The lowest antipodal cell in A. undulatus and A. Novae- 
Anglae is often very much enlarged. 
The fertilization of an antipodal egg is clearly demonstrated 
in Aster, 
