224 FISHER: SEED DEVELOPMENT IN PEPEROMIA 
spore nuclei divide, e. g. in Cercis Siliquastrum, Phaseolus multi- 
florus, and Erythrina Crista-galli—all three investigated by Guig- 
nard (24). Another interesting case of this kind was found in 
Scilla (McKenney, 46). 
(7) That all four nuclei take part in the formation of one 
embryo sac can not constitute a valid objection to the view, because 
there are cases where megaspores are undoubtedly formed and in 
which all four participate in the formation of one embryo sac, as 
in Smilacina stellata (McAllister, 45), and sometimes in Epipactis 
(Brown & Sharp, 5), the walls following the first two divisions 
of the megaspore mother-cell being evanescent in both cases. 
In the former case the megaspores are frequently arranged in an 
axial row, but as was mentioned above, the arrangement is very 
often an intermediate condition between the axial row and the 
tetrahedral arrangement, while in Epipactis the megaspores always 
have the linear arrangement. 
In Euphorbia procera and E. palustris (Modilewski, 49, 52) 
and in the Penaeaceae (Stephens, 70), the quadripolar grouping of 
the nuclei in the mature embryo sac furnishes a strong bit of 
evidence that the embryo sac in each of these cases is a composite 
structure derived from the equivalents of four megaspore nuclei. 
The same may be true of Peperomia, which has the same number 
of nuclei in the mature embryo sac, even though the nuclei do 
not have the quadripolar grouping. 
(8) Another significant piece of evidence is that there is a 
resting period following the second division in the embryo sac— 
that is, following the four-nucleate stage. This was observed in 
every species of Peperomia which was sufficiently studied. As 
is well known, a resting stage following the formation of the four 
megaspores is almost universal in plants. So, this phenomenon 
is in harmony with what would be expected. 
(9) One of the strongest pieces of evidence favoring this view 
is the appearance of evanescent walls following the first and second 
divisions in the embryo sac of Peperomia—together with the 
fact that these walls have never been seen in the eight-nucleate 
embryo sac—that is, following the third division. These walls 
were first reported by Brown (4) in Peperomia Sintenisii and in 
P. arifolia. They have been seen by the writer in seven other 
