A crogynous Jungermanniales. 
293 
resembling the leaves in form and size; perianth with 
longitudinal folds and often concrescent with involucre. 
About a dozen genera, including Ptilidium, Isotacliis, 
Hcrberta, Polyotus, Tricliocolea, Blepharostoma. 
7. — Cephaloziace/e ( TrigonantJiece, Spruce). Leaves 
usually incubous ; leaf-margin rarely entire, usually 
lobed or toothed ; underleaves usually present, but 
smaller than the leaves and of different form; perianth 
usually hypogonanthous {i.e., triangular in cross- 
section, with two dorsal angles and a ventral angle). 
About thirty genera, including Cephnlozin, Protocephn- 
lozia, Pteropsiella, Zoopsis, Lepidozin, Anomoclada, 
Mytilopsis, Micropterygium, Arachniopsis, Bazzania, 
Adelantlius, Mcirsupidium, Kantia, etc.. 
8. Lophoziace^e ( Epigonanthece , Spruce. Leaves suc- 
cubous or transversely inserted, entire or two-Iobed; 
underleaves absent or small; perianth usually epigo- 
nanthous {i.e., triangular in cross-section, with a dorsal 
angle and two ventral angles). About thirty genera, 
including Lopliozia, Haplozia, Nardia, Gymnomitrium, 
Marsupella, Plagiochila, Notoscyphus, Southbya, 
Arnellia, Tylimanthus, Letliocolea, Symphyornitra, 
Calypogeia, Saccogyna, Acrobolbus, etc. 
In the foregoing scheme, which is admittedly to a large extent 
artificial and provisional, I have suggested the elevation of the eight 
groups, which Schiffner (68) treats as sub-orders of Jungermanniaceae, 
into independent orders. The only sharply marked and really 
natural families are the Lejeuneaceae, Porellaceae, Pleuroziaceae, 
and Radulaceae ; the remaining four families are closely connected 
with each other and therefore difficult to characterise. This is 
especially the case with the Lophoziaceae and the Cephaloziaceae, 
which should, perhaps, be merged into one family. In these two 
families the perianth is quite as often as not a smooth or folded 
cylinder, neither distinctly “epigonanthous” nor “hypogonanthous”; 
the leaves are often quite as distinctly bilobed in Lophoziaceae as in 
Cephaloziaceae ; the acrocarpous or pleurocarpous position of the 
archegonial group (and therefore of the sporogonium) varies in 
closely allied genera and even in species of the same genus (e.g., 
Cephnlozia). That is, the two families cannot be kept apart by any 
of the characters generally used to distinguish them. The difficulty 
