150 BOTANICAL GAZETTE [OCTOBER 
Should the apical cell also divide by the two series of vertical walls, 
the apex of the embryo will consist of four octohedral cells (fig. 33). 
This is the situation as described by Haupt. Lerirces describes a 
similar condition in Blasia. Owing to the curvature of the embryo 
two of these will be nearer the neck of the archegonium than the 
other two. 
APICAL CELL.—The apical cell of the row of four may divide ver- 
tically, or, as certain embryos suggest, it may divide again trans- 
versely (fig. 29) before vertical division takes place, although in 
SQOVO 
@ GOS 
I9Q-27 —Figs «19-23, eeripc at PEACO rh +1 1 
tion of quad walls ‘s one another in successive segments: sen ee is basal cae 
fig. 23, apex; fig. 24, transverse section of apical segment of embryo, x is neck of 
archegonium; figs. 25-27, transverse section of embryo with segments in quadrants; 
compare fig. 27 with fig. 42, noting position of walls; 570. 
the absence of a mitosis one cannot be absolutely positive on this 
point. It may also divide by an oblique wall, whether its basal wall 
is oblique or transverse (figs. 31, 32, 34, 35). WOODBURN shows a 
similar situation, to which Haupt takes exception, claiming that in 
his investigation ‘‘a truly median section has never revealed the 
presence of a triangular apical cell.” Lerrces found that in 
Blasia the apical cell may divide by oblique walls, and he figures 
several embryos with a triangular apical cell. The writer’s observa- 
tions of Reboulia confirm WoopBuRN’s statement. In fact, owing 
to the curvature of the embryo, a truly median section is the one 
