434 BOTANICAL GAZETTE [DECEMBER 
od 
Unknown intervening plants derived from g and undergoing reduction in 
the number of the sporangial loculi, with thinning of the synangial wall, or 
An unknown form much like g, but with a simpler type of pollinial organi- 
zation. (This is the more probable member of the direct series, but bot 
may be conceived of as having once existed.) 
A campanula derived from either # or i; in which reduction to the angio- 
spermous pollen sac has taken place, and in which elongation of the stami- 
neal pedicels or filaments is going on (cf. fig. 4). 
The staminate campanula of the conyolvulaceous and many other gamo- 
petalous flowers. 
=. 
=, 
2 
Thus may we derive the staminate zone of gamopetalous forms 
by a series of readily conceivable and closely united reduction stages, 
the principal members of which are in the larger sense already 
known. But let us now see if it be possible to go on and establish 
a plausible ovulate correlation; for if this cannot be done it is more 
than superfluous to say that K is in the extended sense not proven 
to be a true member of the evolutionary series A’, A-j. : 
That an apical series of spirally inserted carpophylls giving Tse 
to a central ovulate cone played the chief role in the development 
of the Magnoliaceae, as in all the conifers and doubtless many other 
forms, is evident enough. But that such a cone was in all cases 
organized, or much less that a terminal group of diffuse carpophylls 
was present in all the ancestral phyla of the angiosperms, is after a 
little consideration seen to be a cumbrous hypothesis. By what 
other means then than by the reduction of the carpophylls and of 
cones may we conceive of the cognate origin of the ovulate region in 
angiosperms? 
Perhaps the Draceneae may be made to give an initial anew: 
Take for instance the cultivated maguey (Agave) of Mexico, with its 
six immense versatile anthers borne on their long projecting filaments 
as seated in the interior surfaces of the six fused sporophylls, three 
of which are smaller, and the other three of which are alternately of 
larger size. On cutting this flower open and viewing it from the 
interior, is not the structure in reality that very diagrammatically 
indicated in fig. 5? 
Now, was not this floral structure derived from a cone 
and changing whorl of six primitive bisporangiate fronds with bass 
megaspores and apical microspores? The simple method of deriva- 
