8 BOTANICAL GAZETTE [JULY 
occupied by a granular mass, apparently of some firm substance 
deposited by the protoplast as an addition to the protective 
layers of the fruit and seed, or possibly connected with the 
absorption of water by these cells (fig. 75). 
The subtending bract increases but little in size after the 
macrospore is formed, and as the fruit ripens the bract withers 
and is squashed down by the swelling carpel (dr. figs. 1, 4). 
In comparing the foregoing with Campbell’s results it will 
be seen that my observations confirm his in regard to the origin. 
of the macrospore and its development to a sixteen-nucleate ripe 
embryo-sac. Campbell thinks that one of the upper of these 
nuclei goes to the egg, and one to each of the two naked syner- 
gids; while eight others, which he interprets as probably 
antipodals, temporarily collect at the base of the embryo-sac, 
but later disperse and become indistinguishable. The other five 
nuclei play no prominent part, there being according to his 
observations no nuclear fusion analogous to that of the polar 
nuclei of the ordinary angiosperm embryo-sac. 
In my own work I have seen but a single synergid which is 
long persistent and has a distinct wall. The nuclei of the group 
which Campbell interprets as possible antipodals I find are 
ultimately fused together into one endosperm nucleus, there 
being no special basal (antipodal) group of sterile cells or 
nuclei. 
Again, Campbell says that the at first flattened embryo finally : 
fills the whole embryo-sac and that there is no endosperm what- 
The meaning of these very striking peculiarities of the 
embryo-sac of Peperomia pellucida (and other species of the same 
genus) is not easy to determine. The extra division of the 
embryo-sac is quite unique, and so also is the lack of a basal 
group of sterile cells or antipodals, Finally the fusion of s0_ 
large a number of nuclei into one, in forming the endosperm 
nucleus, is approached only by the cases of fusion, at quite 2 
