‘ 
514 BOTANICAL GAZETTE é [JUNE 
It seems evident that the variability in the number of cotyle- 
dons could appear only in massive proembryos. The conception 
of monocotyledonous and dicotyledonous embryos has become a 
somewhat rigid one because it has been based upon such pro- 
embryos as those of Alisma and Capsella as types. It has become 
customary to regard these filamentous proembryos as primitive 
and typical of the two groups, and the massive proembryos that 
occur in both Monocotyledons and Dicotyledons as modifications. 
It is our belief that massive proembryos represent the primitive 
condition of proembryos in Angiosperms, and that only from such 
a proembryo could the monocotyledonous and dicotyledonous 
conditions have differentiated. After this differentiation, the 
difference has become relatively fixed by the reduction of pro- 
embryos to filaments. While massive proembryos occur in all the 
three great divisions of Angiosperms, they are notably present 
among the Ranales, from which the monocotyledonous branch 
seems to have arisen; and they are also retained by many of the 
Monocotyledons, notably the Arales and Liliales, and in these 
groups one may expect to find occasional dicotyledony or even 
polycotyledony. 
The sequence of events in the development of the embryo of 
Agapanthus is as follows. As the massive proembryo enlarges, the 
root end elongates, thus remaining narrow and pointed; while the 
shoot end widens, becoming relatively broad and flattish. At this 
broad and flat end the peripheral cells remain more actively meriste- 
matic than do the central cells. It is this peripheral meristematic 
zone that is the cotyledonary zone. In this zone two more active 
points or primordia appear and begin to develop. Soon the whole 
zone is involved in more rapid growth, resulting in a ring or tube, 
but with the primordia still evident. The sequence of these two 
stages must be kept distinctly in mind, namely, (1) the appearance 
of primordia, and (2) zonation. The cotyledonary zone continues 
its growth until a tube of considerable length is developed, leaving 
the apex of the proembryo depressed. In the case of the dicotyle- 
donous embryo of Agapanthus the two primordia on the rim of the 
tube continue to develop equally, the growth of the whole cotyle- 
donary zone being shared equally by the two cotyledons. In the 
