RADULA. 315 
stanter ad folia viva. Plereque planitiem et montes humiliores, 
paucze loca subalpina diligunt. 
Distr. Per orbem terrarum, in terris tropicis frequentiores. Ipse 
in sylvis Amazonicis et Andinis species 13 carpsi, quarum 8 
novas videntur. 
Radule are about equally abundant in the forest-plains of the Amazon 
and on the wooded slopes of the Andes; but are usually of different 
species, those of the plains rarely ascending the hills. Few of the 
montane species reach an altitude of 2000 metres, but two of the finest, 
f. ramulina Tayl. and Rk. Jamesoni Tayl. are-almost subalpines, growing 
sometimes up to 3000 metres. 
Radula is apparently separated by a wide interval from every other 
genus. It has one unique character, viz., the attachment of the radicles 
to the under lobe of the side leaves, and not to the under leaves, or the 
underside of the stem, asin all other hepaticze. In the branches spring- 
ing from the outer base of the leaves it agrees with Lejewnea, and with no 
other. It has also, like Lejewnea, complicate bilobed leaves, with the 
under lobe smaller than the upper; but this character it shares with 
Porella, Micropterygium and a few others. Radula flaccida, the only 
species that grows on living leaves of trees, is externally so like certain 
Lejeunee which prefer the same habitat as to be easily mistaken for a 
species of that genus, especially for one of those that have no under 
leaves ; but there the resemblance ends. In its most essential features 
it differs altogether from Leewnea ; ¢.g., in the polygynous @ flowers ; 
the macrostomous perianth ; the stout pedicel composed of 6 to 8 (not 
of only 2) concentric layers of alternate cells ; and (above all) the very 
numerous, long, slender, 2-spired deciduous elaters. 
In the European R&. complanata (Transoubat, Pyrenees) I find the 
capsule lined with a thin brown membrane down to the very base, even 
over the pedicel. This lining is of linear subflexuose brownish cells, on 
transverse section 4 or 5 times thinner than those of the outer layer, and 
there are certainly in some places 2 inner layers, both of which are apt to 
secede in fragments, although by no means so deciduous as in Porella. 
It seems elateriferous throughout, and the faint transverse strive of the 
cells are possibly the remains of an annular (or spiral) fibre. In R. ramu- 
lina (Quitenian Andes) the base of the capsule, over the pedicel-apex, is 
composed of large, pale, unequal-sided 5-6-angular cells,and is not covered 
by the continuation of the valvular lining, yet seems equally elateriferous. 
Suscenus L.—CLADORADULA. 
Flores 2 cladogeni, ramulo brevissimo simplice constantes. Capsula 
oblongo-globosa, perpaulo longior quam lata. 
1. &. Gottscheana. Elata laxe bipinnata. F. caulina ad basin fere usque 
bipartita ; lobo ovato-orbiculato ; lobulo duplo breviore, a caule 
sublibero, bifido (7.e. auricula equimagna aucto); cell. minutulis. Br. 
floris 2 lobis oblongo-lingulatis. Per. campanulata, quadrangula, 
parum compressa, ore 4-loba. 
