Oephalozia. 



is often adnate to the perianth, either externally, or more frequently in- 

 ternally, and usually by only one edge, so as to form an inner wing to 

 the postical suture. It is, however, not very rare to find a floral under- 

 leaf connate with the side leaves into a trigonous perianth, and then the 

 diagnosis from Lophocolea may become rather difficult; especially where 

 the stem leaves of the Plagiochila are bifid at the apex (as in Lophocolea), 

 which is a not unfrequent character of several Plagiochilce. For in both 

 genera the recurvation of the antical leaf-margin is a constant feature ; 

 and, in both, the perianths are often terminal on the stem, as well as 

 on the branches. The branches of Lophocolea, whether floriferous or 

 not, are never truly postical (as has been affirmed of them) — never 

 spring from the axil of an underleaf — sometimes, indeed, stand midway 

 between two underleaves. Their true origin is contiguous to, and at 

 least half way within, the under angle of a sideleaf. The branches of 

 Plagiochila have in many species a similar origin, but in some they are 

 more nearly mid-axillary. Pinnately-branched species are rare in both 

 genera, but do exist, and then the branches spring exactly from the 

 mid-axil of the side-leaves, as in Plagiochila abietina N., Lophocolea 

 trachyopa Tayl.! dc. In Leioscyphus and Mylia, where the perianth is 

 compressed laterally, as in Plagiochila, but apparently never winged at 

 the sutures, the floral underleaf is sometimes included ; very rarely is it 

 connate at both edges with the sideleaves into a trigonous prism, with 

 the postical face much the narrowest. In Leioscyphus, however, I have 

 found the branches constantly axillary to the underleaves — i.e. postical, 

 although subfloral innovations axillary to the side leaves are occasionally 

 present ; and there are cases where the ramification is almost the only 

 , character to be relied on for distinguishing this genus from Lophocolea. 



In other genera allied to Lopihocolea, but with a pluricarinate pe- 

 rianth, although the number of keels or angles may vary in nearly every 

 species, yet, wherever the angles are reduced to only three, the third angle is 

 invariably antical ; of which we have examples in Eu-Jungermania sphce- 

 rocarpa, lurida, amaina, dc, and in Nardia (Eucalyx) obovata (N.J, N. 

 (Eucalyx) succulenta (Lehm. et L.) Spruce, &c. The lateral flattening 

 of the leaves to the stem in all these genera is visible enough at the 

 apex of the stem and branches and in the inflorescence. In some cases 

 it is so pronounced as to render the leaves laterally complanate, or ac- 

 cumbent. 



