
778 THE AMERICAN NATURALIST. | [Vor. XXXVI. 
case, and these have encouraged further researches upon the 
simpler types of angiosperms. Some of these have yielded 
interesting results, and already give some hints of the possible 
origin of the structures of the angiospermous embryo sac. 
While it cannot be said that this very important point is 
likely to be explained satisfactorily in the near future, never- < 
theless it can be said that not unim- 
portant advances have been made, and 
the purpose of this sketch is to bring 
together the most noteworthy results 
of these recent investigations, and to 
indicate their bearing upon the ques- 
tion of the origin of the angiosperms. 
In the great majority of investigated 
angiosperms the embryo sac, as is well 
known, exhibits extraordinary uniform- 
ity in its structure. In most instances 
the embryo sac arises from a subepi- 
dermal cell (Fig. 2 A), which may either 
develop at once into the embryo sac, 
or, as is more commonly the case, it 
divides into several cells, one of which 
becomes the embryo sac. 
With the first division of the nucleus 
of the young embryo sac, the polarity 
which is so marked a feature becomes 
established. Of the two nuclei result- 
mui ing from the division of the primary 
stabs “hs de saco er nDüclns one: moves to the apical 
is the egg apparatus, consisting of ^ (micropylar) end of the sac, the other 
he two synergidz,. 
s; at the chalazal end, the three to the basal (chalazal) end. Two 
Sec X ill shows du peo nuclear divisions follow, resulting in 
nucleoli of the polar nuclei of four micropylar and four chalazal nuclei. 
which it is formed. 
From each of these groups one nucleus, 
the polar nucleus, moves toward the center of the sac, where 
it unites with the corresponding nucleus from the opposite end. 
1e three remaining apical nuclei, with their accompany ing 
cytoplasm, constitute the egg apparatus ; the three basal ones, 

