5 Cephalozia. 



In those genera, however, that shew a frontal compression of the 

 leafy stem — a great majority of which have either incubous or transverse 

 leaves — and a perianth whose primary angles are derived from the 

 medial /oZcZ, or keel, (and not from the marginal sutures) of the flower- 

 jeaves, whenever those angles are reduced to three, the third angle is 

 constantly postical, and the perianth is flattened (not keeled) in front. 

 To this law there is no exception, as may be seen throughout the large 

 genera Lejeimea and Frullania ; also in Lepidosia, Micropterygium, Her- 

 bcrta (Sendtnera) and many others. Even in Cei^halozia, where the 

 great majority of the species have succubous leaves, the same law 

 obtains. 



In Scapania and Radula, where the perianth is so much compressed 

 frontally as often to bring the upper and under faces into contact — at 

 least in the upper half, and there are externally no angles visible besides 

 the two lateral (or marginal) ones, a transverse section will often shew 

 a very slight and obtuse, yet distinct, postical keel. It is hardly ne- 

 cessary to observe that the flattening of the perianth in these two genera, 

 as also in the few species of Lejeimea where it exists, is exactly at right 

 angles to that of PlagiocMla, Leioscyphus, &c. 



The whole of the leafy JuNGEEMANiDEiE thus divide themselves into 

 two great groups, whereof the one (Epigonianth^) has the third (or odd) 

 angle of the XDerianth in front, or antical ; and the other (Hypogonianthte) 

 has it at the back, or postical. I do not, however, propose them as 

 primary divisions, for there are cases — chiefly in the genus Jungermania, 

 as it remains after the elimination of Cejyhalozia and a few other smaller 

 groups — where the leading characters of the two divisions seem to com- 

 bine, or their distinction to fade away. For instance, a small section 

 oi Jungermania, com.'^vismg J. J. pumila, riparia, cordifolia, &c., has a 

 furrow instead of a ridge at the antical suture of the perianth, and thus 

 recedes from J. sp)herocarpa and other plainly epigonianthous species. 



It is only, in fact, by a judicious combination of all the characters 

 which an extensive study of species reveals to us, that a perfectly natural 

 grouping of the genera of Jungermanidem can be arrived at ; and that is 

 not the task I propose to myself to-day. 



A recapitulation of the main characters of Cephalozia, as I now un- 

 derstand it, is needed to render clear what I shall have further to say 

 of its subgenera and allied genera. 



