230 PICKETT: EMBRYO-SAC OF ARISAEMA TRIPHYLLUM 
OVULE AND MEGASPORE MOTHER CELL 
The ovules occur in groups of two to six arising from the basal 
placenta in each ovary. As it first appears, each ovule is a 
rounded conical mass of cells covered by a clearly marked epi- 
dermal layer, but otherwise undifferentiated (FIG. 4). Soon after 
the ovule has reached the stage shown in FIG. 4, or about the time 
of the appearance of the inner integument, the primary arche- 
sporial cell, or megaspore mother cell or cells, may be found at the 
apex of the nucellus beneath the epidermis. These cells are hypo- 
dermal in origin, and are clearly differentiated from surrounding 
cells by their greater size, denser cytoplasm, and larger nuclei. 
The time of the differentiation of the sporogenous cells is only 
approximately the same as that of the development of the first 
integument as will be seen by a comparison of FIG. I, 2, 3, 6, 8. 
Fic. 1 and 2 show two and three clearly differentiated megaspore 
mother cells, and FIG. 6 one such cell with complete periclinal 
divisions of the epidermal cells, while the integument in each of 
these ovules shows less growth than in FIG. 3 where no such cell 
can be with certainty distinguished. From one to four megaspore 
mother cells have been observed in a single nucellus (FIG. 1, 7, 8, 9" 
11). These are usually in the hypodermal layer at first, although 
in a few instances some were beneath that (FIG. 2, 11, 19). The 
cells in FIG. If and 19 have been removed from the surface by 
periclinal division of epidermal cells. In every specimen examined 
the megaspore mother cells were contiguous, but in no case was 
there found direct evidence that they had been formed by a division 
of a primary archesporial cell as suggested by Mottier (2, p. 259): 
and by Gow (5, p. 40). The position of the cells in FIG. 19 and of 
the two cells at the left in FIG. 11 indicates that such an origin is 
possible; but the failure to find other than very rare cases of 
nuclear activity in cells which have not arisen by the division of 
epidermal cells discredits the probability of such an origin. The 
only form of nuclear activity found in clearly differentiated sporo- 
genous cells was the synaptic or later stages in the tetrad division. 
The fact that in different ovules showing the same general struc- 
ture, more than one megaspore mother cell may be found, suggests 
the origin of such multiple sporogenous cells by simultaneous 
differentiation from hypodermal cells (see FIG. 1, 2, 8). That 
