474 PROF. D. H. CAMPBELL ON THE 
into nearly equal octants, as in the Marchantiacew. In Anthoceros 
(at least in A. fusiformis, Mont., which was specially investigated 
by the writer *), the first wall in the embryo is vertical, and is 
followed by a transverse wall in each cell, the resulting cells 
being of very unequal size. (See ‘ Mosses and Ferns,’ fig. 59.) 
The youngest embryo met with in Dendroceros is shown in Pl.22. 
fig. 13. The arrangement of the walls indicated that the primary 
wall here is longitudinal, as in Anthoceros, but the first transverse 
wall is nearly median, and the lower cells are thus decidedly 
larger than in the latter genus, and it seems probable that the foot 
is determined by this first transverse division, as in Votothylas, and 
does not involve the second tier of cells as is the case in Antho- 
ceros. All three genera agree in having the embryo in this 
stage made up of three tiers of cells, each composed of four cells 
arranged quadrant-wise in cross-section (figs. 13, 15). 
The next divisions in each of these three tiers separate four 
central cells (endothecium) from usually eight peripheral ones 
(amphithecium) (Pl. 22. fig. 16). In Dendroceros, as in all the 
Anthocerotacez,it is from theamphithecium that thearchesporium 
develops, the whole of the endothecium in the upper segments of 
the embryo going to form the columella. The archesporium is 
determined by the first set of periclinals in the amphithecial 
cells, and at first forms a single layer of cells extending over the 
columella and reaching to the base of the middle one of the three 
primary segments of the embryo (PI. 22. fig. 14). In this respect 
Dendroceros differs markedly from Anthoceros, where the forma- 
tion of the archesporial cells is confined to the terminal segments 
(see fig. 60 in ‘Mosses and Ferns’). Leitgeb's figure of an 
embryo of Dendroceros (Heft 5, pl. 3. fig. 10) agrees closely with 
Pl. 22. fig. 14, given here. In Notothylas there is usually à 
zone of tissue between the foot and capsule in which the 
archesporium cannot be made out, and it is not impossible that 
this may also be the case, sometimes, in Dendroceros. 
In regard to the development of the sporogenous tissue, 
Dendroceros is the most reduced genus of the order. While in 
Notothylas the archesporium becomes very massive, being three oF 
four cells in thickness, and in Anthoceros regularly two cells 
thick, in Dendroceros the periclinal walls in the archespor Y 
are very irregular and may, at certain points, be completely 
* Campbell, * Mosses and Ferns,’ pp. 127-136. 
