232 FISHER: SEED DEVELOPMENT IN PEPEROMIA 
Representatives of all four families of the order Piperales have 
been examined in an attempt to ascertain whether any species 
showed an intermediate stage between the embryo sac of Pepero- 
mia and that of the typical eight-nucleate embryo sac, and to see 
whether any feature of the development of the embryo sac con- 
firmed the view that Peperomia is primitive. 
Of the Piperaceae, besides the genus Peperomia, four species 
of Piper have been examined—three by Johnson (32, 35) and 
one by the writer and reported in this paper. All four have 
been shown to have eight-nucleate sacs, which develop directly 
from the definitive archesporial cell, no non-functional megaspores 
being formed. No other condition has been found in this genus. 
As was pointed out, the reduction in the number of generations, 
from the definitive archesporial cell to the mature embryo sac, 
has proceeded one step further in this genus than it has in the 
Peperomia. 
Of the Saururaceae, species of Saururus, Anemiopsis, and 
Houttuynia have been examined by Johnson (31, 33), and Houwt- 
tuynia by Shibata & Miyake (66), and it is found that one or more 
non-functional megaspores are cut off and that in all cases typical 
eight-nucleate embryo sacs are formed from single megaspores. 
Of the Lacistemaceae, Lacistema (Johnson, 33) shows a typical 
eight-nucleate embryo sac derived from a single megaspore. 
Of the Chloranthaceae, Hedyosmum (Johnson, 33) and Chlo- 
ranthus (Armour, 1) have been examined, each of which shows a 
typical eight-nucleate embryo sac which arises from a single mega- 
spore. 
A mode of origin and development of the embryo sac, which 
would confirm the idea that Peperomia is primitive, has not yet 
been found in any other genus of the order Piperales. 
Finally, when we consider, (1) that the primary archesporial 
cell of Peperomia is single—a condition probably derived; (2) that 
the first four nuclei of the embryo sac are probably homologous 
with megaspores, certainly not a primitive feature; (3) that the 
peculiar origin of the endosperm nucleus here probably represents 
a derived condition; (4) that an endosperm which is cellular from 
the start can not be regarded as primitive; and (5) that none of 
the close relatives of Peperomia furnishes any indication, from the 
