W. C. Worsdell. 
130 
capsule of a number of entirely new organs, a manoeuvre to which 
the capsule would have been up to that time a complete stranger ! 
Bower also, I think, places the cart before the horse in taking as his 
model for the protrusion-theory the ontogenetic origin of lateral 
leaf-members on the axis of the higher plants. Surely ontogeny 
must take its clue from phylogeny, and not vice-versa. 
The facts of embryology yield us all the clue we need for dis¬ 
covering the truth. In Jungermanniaceae the epibasal quadrants 
in the embryo give rise to the capsule or sporogonial head; in 
Monocotyledons the same two quadrants give rise to the cotyledon, 
which is therefore terminal. In Ccratopteris, a Fern, the epibasal 
quadrants also are used up in forming the terminal cotyledon. 
Hence it is clear that sporogonial capsule and the cotyledon of the 
Monocotyledons and Ccratopteris are each and all homologous 
structures. By dichotomous branching, the Dicotyledonous 
embryo produces a cotyledon from each of the epibasal quadrants 
or from the whole embryonic head 1 . This leads me to remark 
that it is certain that the Dicotyledons are not so far removed 
from the Monocotyledons as to make the embryological history of 
both is anything but similar. It can hardly be so different in the 
two classes as is usually supposed. We know now several 
Dicotyledons which have a single entire cotyledon or in which the 
single cotyledon is forked more or less deeply 1 ; in some cases also 
it is in a direct line with the hypocotyl. In almost all cases, in 
both classes, the cotyledons arise long before the existence of any 
axis on which they could be borne as lateral appendages: this 
agrees with the phyton- or sporogonial theory, but not with the lateral 
appendage hypothesis ; and this axis does not assert itself until 
quite a late period in the Monocotyledons. If the development of 
the Dicotyledonous embryo is to be regarded as resembling in any 
degree that of the Monocotyledonous embryo, the single cotyledon 
must, in the former group, have dichotomized to form the normal 
two members and the plumular shoot must be considered as 
arising in a position lateral to both cotyledons, and therefore, 
exactly between the two ; yet it is the dichotomized cotyledon, 
and not the plumular shoot, which is to be regarded as primarily 
terminal to the hypocotyl. The actual ontogenetically terminal 
position of the plumule is a secondary adaptation; in some plants 
we see also how it becomes enclosed in the basal sheathing portion 
of a single, tubular cotyledon, which is terminal to the hypocotyl. 
1 This is exactly comparable to the forking of the sporogonial 
head observed in Diphyscium Joliosum. 
