454 BOTANICAL GAZETTE [DECEMBER 
nucleus divides to two, but afterward the nuclei derived from the 
three megaspores nearest the chalazal end degenerate, while those 
from the one nearest the micropyle form the sac. The condition in 
Crucianella approaches that in Peperomia, and it may be that the 
shape of the sac, which is much less elongated in Peperomia than in 
Crucianella, gives the megaspores a better chance to develop together 
in Peperomia. 
That physiological conditions do play a part in the structure of 
the embryo sac of Peperomia may be indicated by the fact that while 
the megaspore nucleus which is cut off by a cell wall is generally at the 
chalazal end, it is always the one nearest the micropyle which forms 
the functional egg apparatus. 
WrEcGAND (’00) reports an evanescent wall in the first division 
of the embryo sac of Convallaria. Here the mother cell forms the 
embryo sac directly and this wall may represent a megaspore wall. 
That more than one megaspore may possess the potentialities for 
development is indicated by the number of plants in which more 
than one has been reported as dividing (CouLTER and CHAMBERLAIN 
703). 
The similarity in the fate of the four megaspores of P. Sintensii 
is striking. Each gives rise to two nuclei of the endosperm and two 
cut off against the embryo sac wall. : 
If we were to accept the view of Porscx (’07) that an egg 4pPat 
atus represents an archegonium, we might conceive of the em id bd 
sac of Peperomia as really composed of four sacs, each of which gives 
rise to a single archegonium. © The relationship of the nuclei of each 
egg apparatus, as previously described, is the same as that found in 
the egg apparatus of an ordinary angiosperm, if we assume that one 
synergid fuses with the nucleus which usually fuses to form the endo- 
sperm, that is, with the sister of the egg, as is the case in the foe 
nucleate sac of Cypripedium (Pace ’08). That there may sch 
Peperomia four potential egg apparatuses is indicated by the fact 
that the nucleus of the four-nucleate stage, which is cut off and there- 
fore resembles the functional megaspore of the ordinary angiosperms 
does not usually form the functional egg apparatus. It must be 
remembered, however, that no such. similarity in the fate of the meg 
spore nuclei as has been described for P. Sindensii exists in P. MS 
