no Campbell —The Embryo-Sac of Peperomia. 
very much the appearance of being made up of several 
fused nuclei. 
The dense cytoplasm of these cells shows the coarse, 
vacuolate appearance seen in the undivided embryo-sac. The 
subsequent division-walls are mostly radial, so that the young 
embryo is surrounded by a single layer of very large endosperm- 
cells. The contents of these cells become more uniformly 
granular as they grow older, and the vacuoles mostly dis- 
appear. Later, there may be formed a small number of 
periclinal divisions ; but even in the ripe seed many of the 
endosperm-cells extend from the embryo to the periphery 
of the embryo-sac. Johnson states that there may be forty 
or more endosperm-cells in the ripe seed. 
The Accessory Nuclei. 
The nuclei which do not fuse to form the endosperm- 
nucleus are usually closely appressed to the wall of the 
embryo-sac, but this is not always the case. In Fig. 6, B, is 
shown one of these nuclei which has remained very like the 
endosperm-nuclei, and has surrounded itself with a cell-wall, 
forming a very conspicuous cell projecting into the embryo- 
sac. In this case there was a similar conspicuous cell 
occupying the position corresponding to the single synergid 
described by Johnson; and, as the accessory nuclei regularly 
develop a cell-wall, become filled with deeply stained cyto- 
plasm, and are very similar to the supposed synergid, it is 
probable they are homologous with it. 
While the group of unfused endosperm-nuclei, with the 
surrounding cytoplasm, often occupies the antipodal end of 
the embryo-sac and looks very much like a group of antipodal 
cells, their subsequent behaviour shows that they are not the 
equivalents of the antipodals of the typical angiospermous 
embryo-sac. These must be sought in the separated cells, 
developed about the accessory nuclei, which here are not 
united in a group, but are distributed singly about the peri- 
phery of the embryo-sac. While they can hardly be spoken 
