54 Proceedings 
forms of skeleton. Such a type as this is clearly older than 
any one of the variations on it. 
Now whenever reasoning of this kind led me to consider 
a type of vascular structure as relatively primitive, I found 
its cotyledon possessed two distinct bundles. In other 
words the evidence pointed to the development of the 
uninerved from the binerved cotyledon. The most primi- 
tive type of all is one which I believe to represent the 
vascular structure of the seedling in the ancestral Lilia- 
ceous stock. In this type the cotyledon possesses two 
massive bundles symmetrically placed opposite each other 
in the oval tranverse section of the cotyledon. They are 
inserted on the bundles of the axis at opposite sides, just 
as the midribs of two opposite cotyledons would be 
inserted. The structure at once suggested that the single 
cotyledon of this type had arisen from the fusion of two 
opposite cotyledons, each with its own midrib. 
This conclusion naturally leads to the hypothesis that 
Monocotyledons are derived from a dicotylous race by 
gradual fusion of two cotyledons into one, The importance 
of this result makes it imperative to criticise the evidence 
on which it is based, and to seek for corroboration. 
First as to the value of the evidence, Experience has 
shown that the systematic value of any character can be 
tested only by experience. But if a collection of embryo- 
logical characters—such as the number and course of the 
vascular bundles in the young seedling—do indicate affin- 
ities,*there is some reason for attaching special importance 
to them. Such characters have always been ranked high 
by systematists. The number of the cotyledons is itself an 
embryological character. 
The most important corroborative evidence is the struc- 
ture of certain Dicotyledons in which the Cotyledons are 
united almost to the apex. Such forms are not uncommon 
among the Ranales, and are found also in families system- 
* The evidence for supposing that they do, will be found in my paper, 
‘“‘A Theory of the Origin of Monocotyledons, founded on the Struc- 
ture of their Seedlings.’’ Ann, of Bot. 1903, p. I. 
_ ee 
