a 
eee athe as lek', 
38 BOTANICAL GAZETTE [juLy 
of development it may, together with the embryo, become — 
entirely detached, so that the whole mass is then free (jig. 26). : 
The wall increases slightly in thickness, and is always thicker 
than any other wall in the young embryo. . 
The first division of the egg is soon followed by a second and — 
later by a third. The embryo then consists of a row of four cells ; 
of which the three upper have nearly the same size. Occasionally | 
only three cells are present at this stage, but the four-celled — 
Stage predominates. The next division is vertical, and in the i 
uppermost cell (fig. 25). This is again followed by one or two ~ 
more oblique walls, thereby dividing the terminal cell into several — 
Sectors, each extending from the basal line to the periphery. 
Periclinal and anticlinal walls form successively in the different 
sectors, as a result of which the upper cell soon becomes a mass 
of tissue (fig. 26). a 
Meanwhile divisions have occurred in the next lower cell. a 
This first formed an oblique wall, after which several divisions a 
took place in various directions. The third cell from the apex . 
has divided once or perhaps twice by vertical walls. The a 
embryo now is nearly spherical, a fact which makes it very diffi | 
cult to trace the development farther with the idea of determin- a 
ing just what portions of the mature embryo are derived from — | 
the various primary cells. The oldest stage at which one can 
with certainty distinguish the four original cells is shown im 
Z. 20. 
The subbasal cell has here und 
two vertical divisions. 
ergone a transverse as well - 
No further divisions took place in this — 
embryo w 
pee together with the basal cell form the true suspensor (fg 
29), 
The fate of the other 
be inferred from their 
the upper cell and the a 
two cells is immediately lost. It caf 
position that the cotyledon arises from 
xis from the subapical. We cannot be 
