402 CHURCH: DEVELOPMENT OF EMBRYO SAC AND EMBRYO 
formation of a two-celled suspensor. Further cell divisions in the 
suspensor have not been observed. A number of examples were 
found where the suspensor consisted of but a single cell. Following 
the division of the basal cell, the free terminal cell divides first 
in a plane parallel with the longitudinal axis of the proembryo, 
forming the two-celled embryo (Fic. 11). The four-celled stage 
results from radial walls perpendicular to the first (F1G. 12). 
From the four-celled stage by repeated division arises a spher- 
ical mass of cells (FIGS. 13, 14), on one side of which develops a 
niche or indentation (F1G. 15). This niche is the point where the 
shoot primordium will arise, while the cotyledon differentiates 
off from the rounded top, and the hypocotyl from the cells just 
above the suspensor of this immature embryo. In one instance 
this niche in an early stage was covered with a thin cell-like 
transparent membrane (Fic. 16). Later stages are shown in F's. 
19 and 20. 
The structure of the mature embryo may be seen in Fics. 
22 and 23. In relief a scarcely protruding pocket protects the 
leaf primordia by surrounding their outer surface as they lie 
against the base of the cotyledon. This is the ‘“‘cotylar” sheath, 
corresponding to the sheath of a monocotyledonous leaf. This 
sheath has no vascular system and is very simple in structure. 
The long spindle-shaped cotyledon is succulent and projects into 
the mass of the endosperm (Fic. 21), where like a haustorium 
it absorbs food for the development of the embryo by means of a 
thin-walled row of parenchymatous cells running parallel with the 
longitudinal axis of the embryo. This strong development of the 
cotyledon is assumed to be in itself indicative of richness of 
endosperm development, and is, according to Goebel (3), char- 
acteristic of plants such as Leucojum and Narcissus. 
The suspensor, varying in the number and arrangement of its 
cells, still holds the embryo in position at maturity. 
The number of leaf primordia developed at the time the seed 
is ripe depends on the individual embryo, influenced as it is by 
various conditions of internal and external factors. No cases 
have been noted where the first leaf is undeveloped at the end of 
the same period, but the development and size attained by such 
leaf primordia in a mature embryo is not at all constant. The 
