1902] EMBRYOLOGY OF LIMNOCHARIS (217 
distinguishable. Schaffner says that the endosperm is not 
abundant in Alisma; but as he does not follow the development 
of the embryo far, might it not be, as in the present case, that it 
- finally completely fills the sac ? 
Before the polar nucleus, which forms the endosperm, has 
_ gone through the first division, fertilization has taken place; for . 
_ Mtthis time the egg nucleus has divided so that there are now 
‘wo nuclei in the young embryo (fig. 72), and the remains of the 
_ pollen tube can be seen in many cases. Actual fusion of sperm 
nucleus and egg was not observed. It is evident that fertiliza- 
_ “on takes place very soon after pollination, from the fact that 
“Material killed within eighteen hours after pollination showed 
the embryo in a two-celled stage, with the disintegrating 
edded in watch glasses for sectioning. 
The first division of the egg after fertilization is transverse, 
oy macs the cell into suspensor and embryo-cell (fg. 13). 
op peal Suspensor cell increases very much in size (figs. 
), has an exceptionally large nucleus, and never divides 
an es polyembryony. In this case it divides and sub- 
e form an embryogenic mass, from which grow out as | 
eral young embryos ( figs. r7 a, 6,c), somewhat as in — 
americanum and Tulipa Gesneri. Whether these all — 
- 9 not I cannot say, for polyembryony is not very fre- 
and | _was unable to find any older stages than those _ 
47 4,6, ¢, although I observed a number of younger : 
ee “uspensor in the older embryo is not wholly composed 
non SUSPensor cell, but ay iates some of the cells 
; ite the first terminal segment until it contains some- 
a ri €ven five cells (figs. 28, 29). The second divi- 
ae <i ansverse ( fig. rg), while the third division takes 
‘ree different directions, transversely, vertically, and 
Ne Ne ee 
