653 
certain Orders of the Ranales. 
tyledonous seedlings would exhibit two cotyledons or a bifurcation of 
the single one ; but as far as I am aware none such have ever been found ; 
whereas in Dicotyledonous seedlings cases in which the two cotyledons 
have fused together into a single organ are very common, and probably, in 
my opinion, represent reversions to the ancestral condition. And the con- 
verse phenomenon of bifurcation of the cotyledons is still commoner, showing 
that this is the direction in which modification is proceeding, and probably 
has proceeded in the past in the course of formation of the dicotyledonary 
character. The above facts and ideas have never been faced by those who 
hold the opposite view to the one I am here promulgating. 
Hence I hold that the presence of two cotyledons in Dicotyledons 
is merely an illusive appearance. There is only one cotyledon, as in Mono- 
cotyledons ; this has become so deeply bifurcated in the great majority of 
Text-fig. i.—A. Shoot of Polyalthia, illustrating sympodial structure, i.e. each leaf terminates 
a stem segment. B. Embryo of Monocotyledon, showing similar sympodial structure of plant 
(after Celakovsky). 
cases as to give the appearance of two distinct leaves. In confirmation of 
this idea I have recently observed a seedling Wallflower in which one of the 
cotyledons had bifurcated, the two resulting leaves assuming positions with 
regard to each other precisely like those assumed by the two cotyledons of 
a normal seedling. Hence the artificial distinction which has always been 
drawn between the two so-called ‘ classes } of Angiosperms has not met 
with much support from the facts of seedling morphology. 
I assume, therefore, that every stem or shoot is essentially a sympodinm 
in the sense that each leaf in its turn constitutes the termination of the main 
axis, the shoot as a whole being continued and built up by means of 
a succession of branches, each arising (alternately right and left) from the 
base of the terminal leaf. And this is precisely what actually happens 
in all Monocotyledonous seedlings at the earliest stage or stages. 
I feel more and more inclined to hold the view that Angiosperms have 
developed directly from an ancestor belonging to the Bryophytic level and 
that they have not come from either Gymnosperms, Pteridosperms, or 
