148 BOTANICAL GAZETTE [OCTOBER 
three cells; Haupt states that it is in the apical cell. According to 
both workers the first division of the hypobasal cell does not take 
place until after the formation of the row of four cells, and then 
usually by a vertical wall. Haupt says that occasionally this verti- 
cal wall formation takes place before the third transverse division, 
that is, when the embryo consists of but three cells. Neither Woop- 
BURN nor Haupt show mitotic figures definitely proving this sequence 
of division. The sequence of these early divisions may hold a definite 
relation to the later differentiation of the sporophyte into foot, stalk, 
and capsule regions. The writer’s preparations show that transverse 
division may take place in both hypobasal and epibasal cells (fig. 9), 
and that the division of the hypobasal cell precedes that of the epiba- 
sal cell (fig. 7). The division of the hypobasal cell may be completed 
before the épibasal cell begins to divide, resulting in a three-celled 
filament (fig. 8). Such embryos are probably quite rare, however, 
the writer having found but a single case. It is more probable that 
the division of the epibasal cell is generally initiated before that of 
the hypobasal cell is complete (fig. 9), and the four-celled embryo 
results with the completion of the two division processes (fig. 10). 
The writer concludes, therefore, that at this stage the embryo con- 
sists typically of a row of four cells with parallel walls, as a result of 
the transverse division of both hypobasal and epibasal cells, the 
division of the former preceding slightly that of the latter. Not a 
single case was found suggestive of the quadrant, as claimed by 
Cavers, who probably based his interpretation on later stages with- 
out having observed these early ones, as he shows no figures of early 
embryo development. 
Variations from the typical situation are of interest. For 
example, the transverse walls are often more or less curved, with the 
concave side toward the apex, the curvature often being especially 
pronounced in case of the apical cell (figs. 10, 17); or the division 
of the epibasal cell may be by an oblique wall (figs. 11-13) whose 
inclination to the perpendicular may show considerable variation. 
Such a division of the epibasal cell is more likely to occur when the 
first division has been an oblique one (figs. 11, 12), although the 
inclination of this wall may be independent of the first wall (figs- 
13, 16). Woopsurn figures several embryos showing oblique walls 
