PICKETT: EMBRYO-SAC OF ARISAEMA TRIPHYLLUM 233 
anything between the first nuclear division of the megaspore (FIG. 
20) and the mature embryo-sac (FIG. 20). The daughter nuclei 
of the first division remain centrally located in the cell until after 
the second division (F1G. 26). Later a normal egg cell and two 
synergidae are formed close in the apex of the embryo-sac, and in 
some cases three antipodals are present (FIG. 20). The antipodal 
cells are poorly developed however, only occasionally showing 
typical angiosperm structure, and never showing the remarkable 
development described by Campbell (3, p. 11) and Rosendahl (6, 
p. 5) for other members of the Araceae. As far as the writer has 
been able to determine, the polar nuclei show no such fusion as 
described by Gow (5, p. 41), but remain separate as in Anthurium 
(3, p. 15). In every case where a polar nucleus was observed after 
fertilization of the egg, the lower one was found separate, greatly 
enlarged, and often disintegrating near the antipodals. Further 
details concerning the fate of the polar nuclei and of the antipodal 
structure will be more fully dealt with in a subsequent paper. 
Where more than one tetrad of potential megaspores are formed 
in one nucellus, the usual course of events is for some one cell 
from one tetrad to germinate while the other spores are crushed 
or broken down. In rare cases more than one embryo-sac have 
been found developing in the same nucellus. In one ovule two 
well-developed embryo-sacs were found side by side and of approxi- 
mately the same size, but unfortunately one lay directly beneath 
the other in the section. The only preparation showing two such 
structures in the same section is shown in FIG, 21. These present 
a marked difference in size as well as deficiencies in antipodal 
structure. As stated above, such imperfect development of anti- 
podals is not rare in otherwise normal embryo-sacs. 
SUMMARY 
The present study has verified most of the findings of other 
investigators, but has given results at variance with their reports 
in regard to the following points. 
(a) The origin of the several megaspore mother cells from a 
single primary archesporial cell is doubtful. 
The first division in the formation of the tetrad has been prob- 
ably mistaken by earlier investigators for a division of a primary 
archesporial cell into embryo-sac initials. 
