34 BOTANICAL GAZETTE {JULY 
proved to be merely that characteristic of an ovule which is destined 
to be abortive. 
It is uncertain whether three or four potential megaspores are 
formed, but it is quite certain that only one becomes functional. 
As shown in fig. 12, the surrounding cells begin to disintegrate so 
quickly that the other cells derived from the functional mother cell 
cannot be recognized. There is some indication that it is not the 
lowest megaspore which develops, as is usually the case, but no defi- 
nite statement can be made. The wall of the mature megaspore 
is so thin that it can be distinguished only with great difficulty. 
In fig. 13 a cell immediately above the megaspore shows one of 2 
the “brown bodies” very distinctly, and a part of this cell is drawn 
on a larger scale in fig. 14. Fig. 13 indicates that polarity may be 
established in the embryo sac before the first division of the megaspore 
nucleus, and many other preparations show the same, but by no means 
all (compare figs. 12 and 15). 
4. The female gametophyte* 
After the first division of the megaspore nucleus, the walls of the 
embryo sac and disintegrating mother cells can scarcely be distin- 
guished. Their thickness has been unavoidably exaggerated in 
fig.15. In the preparation figured (the only one showing a binucleate 
sac), no polarity is evident, nor are the nuclei arranged quite parie- 
tally. After the next division, however, the four nuclei are imbedded 
in (or rather protrude inward from) a very thin parietal layer of 
cytoplasm, and there is some indication of polarity, two nuclei being — 
situated near the upper end and two near the lower end of the sac, the 
two pairs being symmetrically placed with regard to one another 
(fig. 16). This is distinctly interesting in view of the recently revived - 
opinion that the embryo sac of angiosperms is derived from that of — 
gymnosperms by a series of reductions, although it is to be noticed - 
that certain “primitive” angiosperms do not show the polarity sO _ 
characteristic of the vast majority. The growth of the embryo sac 
(prothallus) proceeds rapidly now toward the apex of the nucellus, — 
5 As pointed out by CouLTer and CHAMBERLAIN (2), the bitary of a gameto- : 
phy te should start from the mother cell rather than from the spore, but practically 
it is more convenient to consider sporogenesis separately. 
