370 
DUNCAN S. JOHNSON 
larger, denser mass of cytoplasm that usually lies near the center of 
the embryo sac (figs. 82, 84). Thus far there is no indication, with 
one or two possible exceptions to be mentioned later, of the separation 
of any of these 14 nuclei from their fellows by cell walls. Not even 
cell plates have been found. But it must be noted that no cell plates 
were actually seen in the cases of the divisions forming the egg and 
synergid, whose walls probably arise from a cell plate, as Brown (1908) 
has found them to do in the case of Peperomia sintenisii. It is there- 
fore possible that embryo sacs of just the proper stage would shov\r 
more or less definite cell plates on the spindles of the sac in this 
last division. Such walls, if present, would correspond to the cell 
plates that form the walls of the peripheral cells of Peperomia pel- 
lucida (Johnson, 1900a) and of P. sintenisii (Brown, 1908). Such 
peripheral cells are rarely formed in our species, only 3 or 4 instances 
being noted (figs. 85, 105). 
3. Fertilization and formation of the endosperm nucleus. 
At the stage just described with two micropylar cells and the more 
central group of 14 nuclei, the embryo sac is mature and ready for 
fertilization (figs. 78, 80). The 14 nuclei may possibly not become 
closely grouped until after the male nucleus has entered the egg. 
At least, since the earlier phases of fertilization were not observed, it 
can only be said that all cases where the male nucleus was present in 
the egg the endosperm nuclei of the sac were closely grouped. The 
finding of structures in the style which can be identified with nothing 
but pollen tubes makes it practically certain that fertilization is 
accomplished in the ordinary way by the entrance of one male nucleus 
into the egg. A few cases were noted in which two nuclei were 
present in the egg after the penetration of the pollen tube into the micro- 
pyle. Many other eggswere seen with anucleus containing two nacleoli 
(figs. 82, 83, 84, 98, 104) or a single nucleolus of double size (figs. 88, 
89, 102). The nucleolus is always single except in these presumably 
fertilized eggs and in the primary endosperm nucleus or its deriva- 
tives. All these facts indicate that fertilization is normal in this 
Peperomia. No evidence was discovered that the second male nucleus 
enters into the endosperm-forming complex. This phenomenon has 
been sought for in each species studied but without discovering it in 
any of them. 
As the nuclei become compacted into the group that form the 
endosperm nucleus the cytoplasm immediately about them grows 
