EMBRYO-SAC 9 
THE EMBRYO-SAC 
The development of the embryo-sac of Arisaema triphyllum 
was first studied by Strasburger in 1879 (26). Later Mottier (18) 
and Gow (13) worked out most of the details in the development of 
this structure. In 1913 (20) the writer reviewed the earlier work 
and made some additions and corrections. For the sake of con- 
tinuity the findings of that work will be briefly restated here. 
One to four megaspore mother cells are formed independently 
in the hypodermal layer of the ovule tip. Each of these may 
produce a tetrad of potential megaspores. One or more of these 
megaspores may germinate and produce a typical eight-celled 
embryo-sac. More than one embryo-sac may be formed in each 
nucellus. The fusion of the polar nuclei is doubtful. The anti- 
podal cells rarely develop fully as in typical angiosperms. 
The points of this part of the history worthy of further atten- 
tion are as follows. As already stated, a regular tetrad of mega- 
spores, variously placed, is formed. Later work has shown that 
these megaspores are potentially the same, and that even when 
more than one tetrad is formed many from the plural groups may 
germinate (PLATE 1, FIG. E). Thereisno rule of precedence in later 
development, the matter of quickness of starting and rapidity of 
growth giving some one spore, as a rule, advantage over the others. 
One striking example of the precedence of the lowest spore of each 
of two tetrads has been observed (PLATE 1, FIG. 2), but the consider- 
ation of this as a regular order is prevented by the many cases 
where the uppermost or some of the intermediate spores develop 
into the embryo-sac. That one megaspore usually develops first 
and at the expense of the others is, of course, indicated by the 
common formation of but one embryo-sac; while occasional ap- 
pearance of plural sacs in a nucellus proves the possibility of the 
growth of more than one spore. 
At the time of publishing the earlier paper investigations were 
under way to determine definitely the action of the polar nuclei 
and the fate of the antipodal cells. The flowers of another season 
have been examined since that time, more than two hundred 
preparations showing approximately mature embryo-sacs having 
been made. The following seems to be the usual course of events. 
The two polar nuclei float about for some time in the embryo-sac 
