APRIL 5, 1901.] 
lus. No part functions particularly as the 
vegetative part, or protonema, as is the 
case with most of the pteridophytes. The 
true vegetative function of the endosperm 
appears later. The fact that the archegonia 
do usually arise at the micropylar end of 
the endosperm is probably acquired or 
hereditary, since archegonia there are more 
certain to be fertilized. 
If, therefore, the eight-celled stage of the 
angiosperm embryo-sac is homologous with 
the endosperm of the gymnosperms, then 
all the cells are potential eggs. The func- 
tional egg is at the micropylar end of 
the embryo-sac, because the chances for its 
being fertilized here are greater. This is 
true of the synergids also when they become 
functional eggs. The antipodals probably 
appear in this rdle very rarely. Neverthe- 
less, potentially they are eggs, or greatly re- 
duced archegonia. If this be so, then an 
embryo developing from an unfertilized 
autipodal cell would fall in the category of 
parthenogenesis, instead of apogamy. 
This, however, may be drawing the line 
very fine, and I am not so much concerned 
with that fine distinction as I am with the 
probable origin of the embryo-sac. Since the 
embryo-sac has been recognized as the fe- 
male prothallium of angiosperms, it is natu- 
ral that there should be an effort to interpret 
it as a derivative froma macrospore. It is 
derived, according to the different interpre- 
tations of its homologies, from a different 
number of macrospores. 
According to the Vesque theory, it con- 
sists of eight macrospores derived from two 
cells of the axile row, a fusion of two-spore 
mother cells. According to the theory pro- 
posed by Mann it consists of four macro- 
spores and is derived from a single cell of 
the axile row, regarded as a spore mother 
cell (sporocyte). According to the theory 
suggested by Marshall Ward, it consists of 
two macrospores, each representing a pro- 
thallium of four nuclei, the two spores de- 
SCIENCE. 
530 
rived from the third cell in the axile row by 
a division in which no cross wall is formed. 
This cell of the axile row then would rep- 
resent a one-half spore mother cell, or if the 
number of divisions from the primary mother 
cell be taken into account it would represent a 
single spore. 
The theory that it consisted of eight 
spores was shown long ago by Guinard 
(’81) to be untenable. It necessitated the 
fusion of two spore mother cells. Vesque 
had little in support of his theory, since he 
considered the embryo-sac to be formed by 
a fusion of the two upper cells of the axile 
row, while as a matter of fact.the embryo- 
sac arises from the lower cell in all except 
a very few cases. Although Marshall Ward 
found no evidence of a cell wall separating 
the two nuclei in the first division of the 
embryo-sac mother cell, he regarded these 
as representing two cells of the axile row. 
There have been recent attempts to show 
that the embryo-sac in some cases is de- 
rived by a fusion of two cells of the axile 
row, where a weak or temporary cell wall 
is formed after the division of the nucleus 
(Wiegand, ’98, in OConvallaria). This, 
however, is more properly to be classed 
with the phenomenon so often exhibited in 
free-cell formation, where a temporary cell 
plate is laid down to be soon dissolved, and 
as often occurs in the first division of 
the pollen mother cell of different plants. 
Guinard (’81), in addition to citing this 
phenomenon as evidence that the embryo- 
sac is not formed by a fusion of two cells, 
adduces other strong evidence against it. 
In Agraphis patula (Mellink, ’80) the axile 
row consists of two cells and the upper cell 
forms the embryo-sac, as Campbell (’99) has 
recently shown to be the case in Peperomia 
pellucida, while in Caltha, according to 
Mottier (95), with a three-celled axile 
row, the upper one sometimes forms the 
embryo-sac. In Narcissus tazetta (Mellink, 
80) there are two cells in the axile row. 
