248 BOTANICAL GAZETTE [OCTOBER 
It does not seem likely that a primitive embryo sac, as CAMPBELL 
believes the sixteen-nucleate type to be, has been retained in plants so 
far examined only in such distantly related genera as Peperomia, Pan- 
danus (CAMPBELL ’09), Gunnera (ERNST ’08), Euphorbia (MopILEw- 
SKI 709), and the Penaeaceae (STEPHENS 708); and especially since 
most of these genera are anything but primitive in other respects, and 
the embryo sac is in most cases derived from a spore mother cell which, 
as CAMPBELL says, can hardly be regarded as a primitive feature. 
ENDOSPERM 
NaWASCHIN (700) finds that in Phajus and Arundina there is no 
fusion of the polar or second male nuclei, and he attributes the lack 
of endosperm to this cause. STRASBURGER (?00) has shown that in 
several European orchids this fusion may or may not take place, but . 
in either case there is no division to form endosperm. He concludes 
from this that the lack of endosperm is not due to the failure of the 
nuclei to fuse. The condition in Habenaria, where there is no endo- 
sperm, although the fusion nucleus is of constant occurrence, 15 4 
confirmation of this view. 
In contrast to Habenaria, endosperm may be formed in Lemna 
(CALDWELL ’99) without a fusion of the polar nuclei. , 
In the aposporous embryo sac of Hieracium (ROSENBERG 06) 
polar nuclei with the diploid number of chromosomes may fuse 1 
form the endosperm nucleus. In all known sixteen-nucleate sacs 
all of the nuclei not cut off by walls fuse to form the primary endo- 
sperm nucleus. In Peperomia hispidula (JOHNSON ’07) pees 
fourteen of these fusing nuclei; while in P. pellucida (JOHNSON 00) 
and P. sintensii (BROWN ’08) there are eight; in the Penaeaceae 
(STEPHENS ’08) and Euphorbia procera (MopILEwskI ’09) there are 
four; and in Gunnera (ERNsT ’09) seven. 
The fact that in Habenaria the fusion of the polar and second male 
nuclei does not result in the formation of endosperm, while in eRe 
endosperm is formed without this fusion, taken together with the 2 
that the primary endosperm nucleus may be formed by the fusion 0" @ 
variable number of nuclei or of nuclei with either the diploid oT ai 
loid number of chromosomes, seems to strengthen Se 
(?05) view that the endosperm is not a sexually produced embryo 
