80 BOTANICAL GAZETTE [ FEBRUARY 
38-42, in which may be noted also the constant development oi 
a large vacuole separating the micropylar and antipodal nuclei 
The preparations of R. multifidus were much more favorable 
for a study of the mature condition of the embryo sac than thos 
of R. septentrionalis. The series shown in figs. 43-47 may be taket 
as fairly representative of the mature sac in FR. multifidus. The 
strong development of the antipodal cells is a notable featur, — 
distinct walls being developed, the cells enlarging in size and 
giving evidence of great activity until late in the endosperm for 
mation. The formation of the very large definitive nucleus 
plainly shown, and its placing near the oosphere. The general 
insignificance of the nuclei of the egg apparatus as compared 
with the others, and their relatively late organization into cells, 
is at once remarked, synergid nuclei being especially small. Io 
figs. 43-46 the fact that the synergids are sister cells is evident, 
and the shifting position of the micropylar nuclei can be traced. 
The prominence and activity of the antipodal cells is mort 
marked in certain other ranunculaceous plants than in Ranut 
culus itself. Thalictrum purpurascens and Hepatica acutiloba (fg. 
48), can be taken as illustrations. Associated with the grow! 
of these cells is the usual abundant nuclear division, which | 
although in every case apparently mitotic, is more OF less inte , 
ular, as indicated in jigs. 49, 50, both of which represent tht | 
_ antipodal cells of Hepatica. i 
It has long since become evident that the most variad 
region of the embryo sac in dicotyls, preceding fertilization, * | 
the antipodal region. The old formula of three dwindling - 
evanescent cells, with or without walls, is far from adequt ” 
Such a statement seems to be more true of monocotyls than df 
dicotyls. In the latter 1 am able to distinguish at least four as 
tinct types of antipodal development, with the vast nao | 
ee a ee 
