50 
The Origin of Flowering Plants. 
The ovule of Gnetum consists of a nucellus which is surrounded 
by an envelope with six micropylar lobes and is thus radially 
symmetrical. The nucellus and lobed inner integument appear to 
me on grounds given in detail elsewhere, to be equivalent to a 
synangium. Adopting for the moment this view we may say that 
the female synangium of Gnetum is further surrounded by two 
envelopes, the inner of which probably represents the outer integu¬ 
ment of the Angiospermic seed and the outer, as has frequently 
been pointed out, the carpel. 
Parallel parts are present in the so-called “male flower.” 
We find a synangium (the anther) surrounded by its envelope. 
This envelope appears comparable with those of the female 
synangium. If the latter with its envelopes is a megasporophyll, 
we may safely regard the former with its envelope as a micro- 
sporophyll or primitive stamen. If we now call to mind the so- 
called “ inflorescence ” of Gnetum we find it consists of axes on 
which are borne, in verticils, sterile bracts, and closely approximating 
to these whorls the structures which may, and often have been, 
interpreted as mega-, or microsporophylls respectively. If so, their 
position in the axils of the bracts may be entirely due to the need 
for protection in the young state and be correlated with a loss of 
the internodes next below them. 
Suppose now the internodes above the fertile leaves to be like¬ 
wise suppressed, and the whole spike reduced to a conical torus with 
the loss of the now unnecessary bracts, we should then have the 
structure of the central part of such an Angiospermic flower as 
Liriodendron. 
Moreover, as some of the “spikes” are androgynous, with the 
carpels occupying the uppermost nodes, we may even obtain the 
ambi-sporangiate flower by the same process of reduction. In 
Gnetum we know that the so-called “female flowers” terminating 
an androgynous spike are sterile, but that they represent ancestrally 
fertile organs would be granted by most morphologists. 
In Welwitscliia the telescoping has been accomplished and we 
have the outlines of the Angiospermic flower for the first time laid 
down. 
Here again the central capellary organ is sterile, but we cannot 
doubt that it was fertile in the ancestors of Welwitscliia. The 
similarity of the stamen of the latter with the so-called “ male 
flower” of Gnetum strongly corroborates this view of their homology. 
When we add to these considerations the similarity in the structure 
