a 
Proceedings 55 
atically remote from them. I have examined the seedlings 
of Eranthis hiemalis, Podophyllum peltatum, Delphinium 
nudicaule, Anemone Coronaria, and in all these species— 
but especially in the two first—the vascular skeleton very 
closely resembled that of Anemarrhena asphodeloides, the 
species which in my view best represents the primitive Lily- 
type. The two midribs remain distinct in the solid tube 
formed by the united cotyledons. They resemble the twin 
bundles of Anemarrhena in their position in tranverse sec- 
tion, in orientation, in their internal structure, They join 
the bundles of the axis in the same way. 
The importance of this comparison is obvious. Before 
making it we could only say that the vascular skeleton of 
the primitive Lily type suggested the fusion of two lateral 
members. Comparison of this skeleton with that of forms 
in which two cotyledons are undoubtedly united has re- 
vealed an almost identical structure. 
The case for considering the cotyledon of Liliaceous 
seedlings as a fusion of two lateral members has now been 
shortly stated. What ground is there for extending this 
view to other Monocotyledons ? 
The vascular skeletons of the seedlings I have examined 
from other families are of two kinds. Among Amaryllids 
Irids, Aroids, are types already familiar within the Lilia- 
cee, and presumably to be interpreted in the same way, 
The Palms and Scitaminez show distinct types in which 
the dual symmetry is very clearly marked. The vascular 
skeleton of the Grass seedling presents many difficulties, 
but it is most satisfactorily interpreted as a variant on that 
of the Scitaminez. 
If we suppose the origin of the Monocotyledonous coty- 
ledon from two lateral members to be established as a 
working hypothesis, further questions arise. Can any cause 
be assigned for the original fusion of the cotyledons? And 
would the same cause favour the development of other 
Monocotyledonous characters ? 
The vast majority of Dicotyledons have two distinct 
