332 BOTANICAL GAZETTE [May 
small compared with the cotyledon, and is usually more or less 
protected by it. 
In the third period the stem bud becomes the ascending axis, 
and the cotyledon—even if persistent—is dwarfed by its rapid 
development. In the early part of its third period the plant is 
still very far from maturity. The epicotyl of monocotyledons 
contains at this time a single ring of collateral bundles, which’ 
may even show traces of a cambium. The stem, in fact, resem- 
bles that of a dicotyledon. In some monocotyledons the mature 
rhizome and the nodal regions of the aerial stem are character- 
ized by amphivasal bundles. These are not found in the seedling 
(Jeffrey, 2. 2.). 
There is nothing forced about the explanation of these facts 
given by Professor Jeffrey. As the characters of the young 
stem recall those of a dicotyledon, the features of monocotyle- 
dons appearing later in life, it is to be concluded that the stem 
anatomy of the common ancestor was essentially dicotyledonous. 
As has been said already, the anatomy of fossil cryptogams 
establishes a presumption that the primitive angiosperm possessed 
secondary thickening. The traces of the process found in the 
immature stem of monocotyledons strongly confirms this pre- 
sumption, 
For some years I have been employed on a comparative. 
examination of monocotyledons soon after germination—as a 
rule during what I have called their second period of growth. 
About this epoch the vascular system of the immature structure 
is first completely defined by the lignification of the xylem. 
This work has led me to the conclusion that monocotyledons 
are descended from an ancestor with two cotyledons, and that 
the single cotyledon which distinguishes them is a member 
formed by the fusion of the pair. For the nature and strength 
of the evidence I must refer to a previous paper.** It is sufficient 
to say here that the absence of a true midrib is a very general 
character in the cotyledon of monocotyledons, that its place is 
commonly supplied by two equivalent bundles, and that in the 
“@SARGANT, E., A theory of the origin of monocotyledons. Annals of Botany 
r7 21. 1903. 
