534 
ripe, forming twelve nuclei for each pollen 
mother cell, instead of eight; and thus the 
homology falls. 
7. Another view is that the embryo-sac, 
at the eight-nucleated stage, is homologous 
with the endosperm of the gymnosperms or, 
in other words, that the eight free cells are 
homologous with the endosperm of the 
gymnosperms. This was proposed by Stras- 
burger as early as 1879. One of the cells 
of the endosperm forms the archegonium 
which here is very much more reduced 
than in the gymnosperms, being reduced 
to a single cell. A suggestion of such a 
reduced archegonium is found in Welwit- 
schia, etc., where a single endosperm cell 
without division elongates to form the 
simple archegonium. Guinard (’81, Ann. 
d. Sci. Nat. Bot.) upholds this view. 
8. Very closely allied to this view is 
the one which interprets all the cells at 
the eight-nuclear stage of the embryo-sac 
as potential eggs. Dodel (’91) found in 
Tris siberica that the synergids were some- 
times fertilized, and developed embryos. 
He suggests that the synergids are poten- 
tial eggs. Chamberlain, ’95, suggests that 
one of the antipodals has all the appearance 
of an egg ready for fertilization. Stras- 
burger (Ang. und Gyn., ’79) states that 
while in Santalum the normal number of 
eggs is two, there are sometimes three, and 
one of these may be the upper polar nucleus, 
when the endosperm is developed from the 
lower polar nucleus. Overton (’92) records 
a case of fertilization of one of the synergids 
of Lilium martagon; Guinard (’81) the de- 
velopment of embryos from two synergids 
in Mimosa denhartvi, and Tretjakow (95, 
Ber. deut. Bot. Ges.), the development of 
embryos from the antipodals of Alliwm 
odorum. Tretjakow interprets this as a 
case of apogamy, since he regards the 
antipodal cells as representing the vegeta- 
tive portion of the prothallium. 
In view of all the facts, Strasburger’s hy- 
SCIENCE. 
[N. S. Von. XIII. No. 327. 
pothesis that the eight cells of the embryo- 
sac are homologous with the endosperm of 
the gymnosperms, seems the more reason- 
able one. If the synergids, then, can be 
fertilized and produce embryos, they too 
represent archegonia reduced to a single 
cell each. This would confirm the view 
first proposed by Hofmeister that the syner- 
gids and egg are ‘ germinal,’ or ‘embryonal 
vesicles.’ Tretjakow holds that the anti- 
podals represent the vegetative part of the 
prothallium, while the synergids and egg 
represent archegonia, which is very like the 
earlier views held by Strasburger, Prings- 
heim and others. 
But if the eight cells of the embryo-sac 
are homologous with the endosperm of the 
gymnosperms, and the egg is an archegon- 
ium reduced to a single cell, it would seem 
that all the cells of the embryo-sac are po- 
tential eggs or potential archegonia. 
Itis probable that all the peripheral cells 
of the endosperm in the Abietinez, for ex- 
ample, at a certain stage of development, a 
short time prior to fertilization, are poten- 
tial archegonia. While the archegonia are 
usually developed from superficial cells at 
the micropylar end of the endosperm, they 
are frequently formed from superficial cells 
down on the side some distance from the 
end, giving to the endosperm in longitudi- 
nal section the appearance of a comb. 
Archegonia are in some cases developed at 
the opposite end of the endosperm which 
would correspond in position to the antipo- 
dals of then agiosperm-embryo-sac. Fur- 
thermore, archegonia are rarely developed 
from internal cells of the endosperm. This 
would indicate that prior to the time for 
fertilization all the cells of the endosperm 
are potential archegonia. 
And this too seems reasonable since up 
to this time the course of development in 
all parts of the gymnosperm embryo-sac 
have been the same, all parts bear the same 
nutritive relation to the surrounding nucel- 
