374 Stephens . — - The Embryo-Sac and Embryo of 
The polar nuclei are homologized with ventral canal cells, and as the 
ventral canal cell has been shown capable of fertilization, the fusion of the 
second male nucleus with them can be regarded as a kind of fertilization, 
and the endosperm as of the nature of a modified embryo. Ernst, applying 
this c Archegontheorie 5 to the Penaeaceae, considers then that each of the 
four peripheral groups represents an archegonium, and that the four nuclei 
fusing to form the primary endosperm nucleus are the four ventral canal 
cells of these archegonia. This would hold equally well if each group were 
regarded as formed from a macrospore, only in this case the sac would 
be reduced rather than primitive, as Ernst interprets it. But it is difficult 
to imagine that a gametophyte so greatly reduced as that of the Angio- 
sperms could continue to form the rudiments of such specialized structures 
— structures which in the Gymnosperms appear only at a much later stage 
in the life-cycle, and moreover have disappeared in the highest group of 
Gymnosperms, so that in Welwitschia and Gnetum the naked cells of the 
prothallus function as eggs. Moreover, a view that regards the synergids as 
homologous with neck-cells affords no explanation of such a case as that 
described for Naias major } where a synergid is often fertilized and gives 
rise to an embryo. Porsch, it is true, regards such ‘ Synergidenbefruchtung ’ 
as of no more significance than the fact that embryos can arise from nucellus 
or even integuments , 2 but the cases are surely very different, the latter 
being merely sporophytic budding. For these reasons, the present writer 
is inclined to look for a more probable explanation for the peculiarities 
of this embryo-sac. 
A suggestion which seems of service here is that offered by Pearson 
regarding possible homologies in the embryo-sac of the primitive angiosperm. 
In an abstract of some recent observations on W elwitschiaj he describes 
all the nuclei found in the embryo-sac at the end of free-nuclear division as 
being potential gametes. Some of these remain free and can function as 
gametes ; the remainder fuse to form the primary endosperm nuclei. He 
suggests that the endosperm of the primitive angiosperms was homologous 
with that of Welwitschia. If the sixteen nuclei formed in the free-nuclear 
stage in the Penaeaceae can be similarly regarded as potential or reduced 
gametes, and the endosperm as physiologically homologous with that 
formed by the fusion-nuclei in Welwitschia , the chief peculiarities 
noted — the apparent formation of an egg-apparatus in each peripheral 
group ; the equality in development and structure of these groups and their 
component cells ; the formation of a normal endosperm by a fusion of four 
(exceptionally more) nuclei ; the exceptional case of division of the primary 
endosperm nucleus before fertilization — all find an explanation. This 
s uggestion thus seems to the writer to meet the facts of the present case. If 
this embryo-sac can be regarded as more primitive than the normal type, 
1 Guignard, 1901. 2 Porsch, 1907, p. 25. 3 Pearson, 1909. 
