Sphcerocarpales . 83 
formation of simple or branched lobes or leaves. Apical growth is 
always effected by a single apical cell; in the leafy forms this cell 
is generally a three-sided pyramid, in the thalloid forms it varies in 
shape. In apical branching, the apical cell of the branch is 
constituted of a segment cut from the parent apical cell, and the 
branching is typically monopodial. There are no tuberculate 
rhizoids. The sexual organs are usually developed in groups, but 
never on special stalked receptacles. In the young antheridium 
vertical divisions set in at a very early stage, usually after a single 
transverse division of the mother-cell. The neck of the arche- 
gonium consists of five rows of cells. The embryo typically divides 
by repeated transverse walls, the lower (hypobasal) of the two cells 
formed by the first division usually taking no further part in the 
development of the sporophyte; there is no octant stage. The 
sporogonium usually has a relatively long seta, and the capsule 
nearly always opens by four valves, never by a lid. The spore 
mother-cell becomes deeply four-lobed before the tetrad division 
occurs. 
Though these two groups are, on the whole, sharply distin¬ 
guished, there are interesting cross-affinities between the lowest 
members of the two groups. The Ricciaceas, the lowest of the Mar- 
chantiales, present the simplest type of sporophyte, in which 
sterilisation is (except in Tessellina) confined to a single peripheral 
layer of cells, forming an epidermis or capsule-wall around the 
wholly sporogenous central tissue. In this respect the Ricciaceae 
are sharply marked off from all other Bryophytes, but in the 
structure of the thallus and of the sexual organs they agree closely 
with the higher Marchantiales. 
The small families Sphaerocarpoideae and Rielloideas, which 
Schiffner (27) places at the base of the Jungermanniales, are 
regarded by Goebel and by Lotsy as belonging to the Marchan¬ 
tiales, and as a matter of fact they were included in the Ricciaceae 
by the earlier systematists. These two families agree with the 
Jungermanniales in having no air-chambers in the thallus, in having 
no tuberculate rhizoids, and in showing a marked tendency towards 
the development of leaves; but they agree with the Marchantiales 
in the development of the sexual organs, in the early stages of embryo- 
geny, and in having a single-layered capsule wall. Finally, they 
show certain peculiarities not found elsewhere in the Liverworts— 
the special envelope around each antheridium and archegonium in 
both families, and the curious wing of Riella. It would seem 
