Feb., 1908.] Polar Conjugation in the Angiosperms. 
2 57 
(Italics not in the original.) The pollentube is so remarkably 
distinct in Sagittaria and produces such marked changes that I 
am certain I could not have overlooked it in the stages where 
polar conjugation takes place, especially since the study of the 
pollentube was one of the main objects of this investigation. 
Thus it is certain that in some plants the polar nuclei have 
learned to conjugate without the influence of the second sperm 
nucleus or even the pollentube. The question now arises as to 
how the polars acquired this remarkable ability if they do not 
represent opposite sexes. For it seems excluded that one could 
think of the Angiosperm female gametophyte as being a direct 
descendant of an hermaphrodite thallus, the polar nuclei being 
descendants from male and female gametes. It is altogether 
probable that the Angiosperms passed through the Heterospo- 
rous Pteridophyte stage before becoming seed plants. Porsch's 
view, therefore, seems the correct one, that the triple conjugation 
results from the essentially female character of the polars. If 
therefore a conjugation takes place without the presence of the 
second sperm, this must be looked upon as a special sort of 
parthenogenetic development. All polar conjugations, accord- 
nig to this view, had their origin in the original conjugation of 
one or both polars with the second sperm, typically in the second 
w r ay through triple fusion. 
Now the question arises as to whether there is a triple fusion 
in Sagittaria and other such cases. Does the second sperm 
come down later and fuse with the polars acquired the property 
or function of conjugating with each other through their common 
attraction to the second sperm with the first upper endosperm 
nucleus after the partition wall is formed at the end of the 
division of the definitive nuclues? This division takes place 
about the same time as the first division of the oospore, and such 
a possibility is suggested by the following facts: The second 
sperm seems to remain in the tube for some time after the first 
one escapes to unite with the egg; the upper endosperm nucleus, 
immediately after the division of the definitive nucleus, begins to 
travel upwards; the lower endosprem nucleus presents a remark¬ 
ably different development from the upper one. But no weight 
is to be attached to the suggestion until further investigations 
are made. 
True endosperm, as has been suggested by several investi¬ 
gators, may be present even in Gymnosperm archegonia. A 
true endosperm might originate from the division of a ventral 
canal cell without conjugation of the second sperm with the 
ventral canal cell. In Angiosperms an endosperm might result 
from the conjugation of either polar nucleus with the second 
sperm; from the conjugation of both polars with the second 
sperm, which seems to be the usual mode; or through partial 
