260 BOTANICAL GAZETTE | APRIL 
Even plants of the same species and different parts of the 
same plant, when treated in exactly the same way, will react 
quite differently, and one must practically invent a new process 
for every form studied if reliable results are to be obtained. 
THE ANTIPODAL REGION. 
As already stated, the three antipodals are surrounded at an 
early stage by very definite cell walls, although in certain cases 
the formation of walls may be delayed for some time. When 
the embryo is two-celled, the embryo sac is still quite narrow, 
and tapers very gradually into the antipodal region (figs. 36, 37, 
39); but after this it widens out greatly, leaving the antipodal 
region, which in the meantime has developed very thick cell 
walls, as a sort of vermiform appendix to the lower part of the 
sac (figs. go, 41, 42, 43). The antipodal region, with its three 
nuclei, persists even in the fully formed ovule, where it always 
produces a very striking appearance because of the constancy 
with which it preserves its original shape and dimensions. In 
these late stages, however, the antipodal nuclei stain a very 
deep color, so that they appear almost entirely homogeneous 
(jigs. 40, 41, 42,43, 44, 73). It is probable that this persistence 
of the antipodal region may be much more common than is 
generally supposed, and that it may often have been overlooked 
and reported as disappearing when it was actually present and 
persisting even in the mature ovule. 
DEVELOPMENT OF THE ENDOSPERM. 
In the development of its endosperm Sagittaria presents 
some very interesting peculiarities. The first division of the 
E jaueas nucleus takes. place about the same time as the first 
division of the oospore; and what is most remarkable, at this 
: division a cell plate is formed between the daughter nuclei, 
ae which cuts the embryo. sac transversely into two compartments 
(figs. 36, 37+ 38, 39). This transverse wall will be called the 
partiti on ie 
: . = pe male, “ wid ambiguity one on the micropylar = 
