446 BOTANICAL GAZETTE [DECEMBER 
if it should be shown that some normal sacs are formed without 
the original determining condition, for cases apparently quite 
similar to this are known among animals. A striking example 
is the lens of the amphibian eye (SPEMANN 14), which in some 
species requires the presence of the optic cup for its development, 
while in others it develops even if the optic cup is removed. 
From the foregoing discussion it does not follow that the nuclei 
play a passive part in development; for the external conditions 
which influence them may in turn be due to the nuclei from 
which these have been derived, i.e., the vegetative nuclei of the 
plant; and if a nucleus were other than it is, it probably could 
not react to external conditions to produce the structures which it 
does. 
Any analysis of the conditions determining the course of devel- 
opment of the embryo sac must at present be incomplete and largely 
tentative, but a comparison of the conditions under which various 
types of sacs are formed may be worth while, as it is likely to sug- 
gest new ways of looking at their origin and development. The 
first point to be considered is the production of polarity. Before 
the megaspore mother cell divides, it has the general shape of the 
mature sac, and an enlargement of the whole nucellus without 
further change would preserve this shape. The formation of the 
megaspores in rows in most angiosperms, and the elongation of the 
nucellar cells in a direction parallel to this row would indicate that 
the elongated shape of the functional megaspore and the sac is 
connected with the direction of greatest pressure in the nucellus. 
When the nucleus of the mother cell divides, the daughter nuclei, 
as is usually the case, tend to be evenly distributed in the cyto- 
plasm. After vacuolization, a continuation of this same tendency 
would carry the nuclei to the two ends of the sac, where surface 
tension would cause the accumulation of the cytoplasm. The 
conclusion that this is the explanation of polarity is supported by 
the development of the sixteen-nucleate sacs. In Peperomia 
sintenesit (BROWN 1), where the sac would seem to be derived from 
four megaspores, the mother cell and embryo sac are both rounded, 
and there is no polarity. The same thing is true in the Penaeaceae, 
where Miss STEPHENS (15) believes that the embryo sac is derive 
