Evolution of the Angiosperms . 505 
formed by the megasporangium itself; an inference supported by the 
presence, in the three genera, of a long, tubular micropyle, the adaptation 
of which is obvious. Consequently in plants pollinated in this manner 
where reduction has been carried to extreme limits, it is not surprising to 
find that the megasporophyll has totally disappeared, and that the ovule 
is now borne on the axis. On the other hand, in the most reduced 
Angiospermous flowers, in which cauline ovules occur, the closed mega- 
sporophyll must persist to continue the task of pollen-collection. 
Female Gametophyte (Embryo- sac). 
In recent years attention has been largely focussed on the embryo-sac. 
The papers of Lotsy , 1 Land , 2 Pearson , 3 and Berridge and Sanday , 4 have 
shown that, on the one hand there is a great reduction in the gametophyte, 
when compared with other Gymnosperms, and on the other a notable 
dissimilarity between the embryo-sacs of all three genera. Such facts have 
been interpreted as evidence that the three genera are not so closely allied 
to one another as has been supposed, and they have further been regarded 
as likely to throw light on the puzzling homologies of the stereotyped 
form of the Angiospermous embryo-sac. In Ephedra the gametophyte 
is decidedly the least reduced. Typical Gymnospermous archegonia are 
present. There is, however, some morphological distinction between 
the upper (reproductive) and the lower (vegetative) region of the mature 
embryo-sac. 
In Gnetum Gnemon the sac is also markedly differentiated into two 
parts, but no archegonia have as yet been found at the upper (micropylar) 
end. Here free nuclei only occur, which apparently are all potentially 
eggs. The lower (antipodal) portion consists of a more compact tissue, 
which continues its growth after fertilization. In other species of 
Gnetum , Karsten 5 has observed further stages in the reduction of the 
gametophyte, and has shown that the ‘ endosperm ' is not developed until 
after fertilization — a striking parallelism to Angiosperms, and a contrast 
to other Gymnosperms. 
The embryology of Welwitschia has been considered to be more or less 
intermediate between Ephedra and Gnetum Gnemon , but Pearson’s re- 
searches 6 do not favour this view. The embryo-sac in Welwitschia appears 
to be highly modified on somewhat different lines. 
In respect to the development of the embryo, Coulter and Chamber- 
lain 7 regard Ephedra as standing nearest to the other Gymnosperms. 
Bower , 8 however, from his study of Gnetum Gnemon> concludes that c In 
1 Lotsy (’99). 2 Land (’04) and (’07). 3 Pearson (’06). 
4 Berridge and Sanday (’07). 5 Karsten (’93). 6 Pearson (’06),. p. 297 . 
7 Coulter and Chamberlain (’01), p. 132 . 8 Bower (’82), p. 298 . 
