146 FisHerR: SEED DEVELOPMENT IN THE GENUS PEPEROMIA 
tion of the primary archesporial cell (Fic. 20) to the practically 
mature fruit, is essentially the same as in P. reflexa, except for some 
anomalous cases, which will be described later. The nucleus of 
the embryo sac mother-cell shows synapsis before it divides. 
Evanescent walls frequently more or less completely separate the 
nuclei in the two-nucleate and four-nucleate stages of the embryo 
sac. The mature embryo sac contains sixteen nuclei (FIG. 21). 
An egg and one synergid are cut off from the rest of the sac. Of 
the remaining fourteen nuclei, at least six are individually cut off 
as peripherals, and at least six others fuse together to form the 
endosperm nucleus. There can be very little doubt that the 
combined number of peripheral and endosperm nuclei is fourteen 
as it is in P. reflexa, and the number of each probably varies, as it 
does in that species. 
As in P. reflexa, relatively many more four-nucleate embryo 
sacs were found than two-nucleate or eight-nucleate ones, thus 
indicating that there is a resting stage after the sac becomes 
four-nucleate. 
The abnormal lobing of the ovules is the most interesting 
feature in the development of this species. A very large percent- 
age of the ovules do not develop in the manner typical for this 
genus, but become more or less lobed (Fics. 22-27). The carpel, 
which encloses an abnormal lobed ovule, seems altogether normal. 
The irregularity is confined to the nucellus and integument. 
Instead of the simple, ellipsoidal ovule, usual in Peperomia, we 
often have a two-lobed structure, one lobe developing an embryo 
sac and the other usually remaining sterile (Fics. 23-27). The 
notch between the fertile and sterile lobes does not extend clear 
to the base of the ovule, but only about one half to three fourths 
of the distance (Fics. 23, 27)—so that this lobing may be looked 
upon as a forking or branching of the ovule, somewhat similar 
to that reported by Hofmeister (29) in Morus alba. 
The sterile lobe is generally larger than the fertile lobe, so 
that less than half of the tissue of the ovule takes part in the forma- 
tion of the seed. The fertile lobe is the only part of the ovule which 
develops an integument (Fic. 25), and this integument may vary 
greatly. In normal, unlobed ovules the integument is two cells 
thick except at the apex where it is thicker around the micropyle. 
