STRUCTURE AND DEVELOPMENT OF DENDROCEROS. 473 
wall; this is followed by a second transverse wall in each of the 
terminal cells, and in each of the resulting eight cells there then 
is formed a periclinal dividing it into an inner and an outer cell 
(fig. 9). The inner cells, by rapid divisions in three planes, 
give rise to the central mass of small, nearly cubical sperm-cells, 
while the outer layer of cells divides no further by periclinals, 
but remains permanently but a single cell in thickness. Each 
of these outer cells contains a single chloroplast as in the other 
genera. There is very little displacement of the cell-walls in the 
inner mass of sperm-cells, and these are arranged in nearly 
cubical masses, corresponding to the early divisions in the central 
cells (fig. 12). 
The sperm-cells are too small to make possible a satisfactory 
study of the spermatozoids, and as there was no indication of any 
variation from the type of structure found in the other Hepatice, 
no attempt was made to trace the development of the spermato- 
zoids. The latter are small, but were not studied in the living 
material owing to lack of facilities when the material was 
collected. 
The ripe antheridium, as Leitgeb showed, is very large and 
the stalk extremely long so that it is coiled up within the cavity 
containing the antheridium. The upper wall of the cavity 
projects strongly above the surface of the thallus, forming a 
nearly globular protuberance easily seen with the naked eye, and 
resembling the similar swellings caused by the Nostoc-colonies. 
The Sporophyte. 
As might be expected, from its intermediate character in 
other respects, the sporogonium of Dendroceros is also, to a 
certain extent, intermediate in structure between that of Noto- 
thylas and Anthoceros. The structure of the sporogonium in the 
former genus has been studied by Mottier *, whose results agree 
with those of the writer made upon the same genus, and do not 
confirm Leitgeb’s contention t that the origin of the columella 
may be secondary, but show that, as in Anthoceros, the columella 
is formed by the first periclinals in the upper cells of the embryo. 
In Notothylas, as in most Hepatice, the first wall in the 
embryo is transverse and divides the embryo into nearly equal 
Parts in which longitudinal walls then arise, dividing the embryo 
* D. M. Mottier, in Ann. Bot., viii. Dec. 1894. 
+ Leitgeb, Unters. üb. Lebermoose, Heft 5, p. 7. 
