Evolution of Monocotyledons. 
307 
The other two papers deal with the origin of the monocotyle- 
donous condition. 
Miss Farrell (2), working on the ovary and embryo of Cyrtanthus 
sanguineus, concludes that the sheath of monocotyledons is probably 
the result of fusion of two or more cotyledons. In this species, the 
youngest observed stages of the embryo, show the stem-tip 
surrounded by a four-lobed sheath ; the growth of this sheath (1, p. 
516) produces the cotyledonary ring in which the four lobes or 
growing points persist for a time, but ultimately grow together in 
pairs producing two cotyledonary rudiments. Finally the cells of 
one of the cotyledons cease to divide, and an apparently single 
terminal cotyledon results. Miss Farrell considers her investigation 
to furnish a last proof of the theory of the derivation of monocotyly 
from dicotyly. 
The work of Coulter and Land (1) on Agapanthus umbellatus is 
in agreement with that of Miss Farrell with regard to the origin 
of the cotyledonary sheath. Their observations form the basis of a 
new theory of the origin of monocotyly. Agapanthus umbellatus 
may produce monocotyledonous or dicotyledonous embryos, and in 
either case, the cotyledons arise as lateral structures from a 
peripheral zone surrounding the stem-apex of the proembryo. The 
proembro is massive, a condition which Coulter and Land consider 
to be primitive for Angiosperms, and at its broad, somewhat 
flattened shoot-end the meristemic activity of the peripheral cells 
produces the cotyledonary zone, in which two or more growing 
points appear. By continued growth of the whole zone a sheath is 
formed surrounding the depressed stem-apex, and equal growth of 
the two primordia produces a dicotyledonous embryo. In other 
cases, however, only one growing-point continues to divide, so that 
apparently only one cotyledon is developed. Coulter and Land 
maintain that, since the sheath and the cotyledons are all one 
structure, the production of a so-called single cotyledon is due to the 
“growth of the whole cotyledonary zone under the guidance of a 
single growing point” (1, p. 515), instead of the distribution of 
activity between two separate primordia. In polycotyly, more than 
two primordia appear and persist in growth. The authors compare 
the zonal development of the cotyledonary sheath with its varying 
number of growing points, with the zonal development of sympe¬ 
talous corollas, which according to them, provides a parallel case. 
In view of the importance attached by Miss Sargant and other 
