Evans : Hepaticae of Puerto Rico 5 



of the stem ; they are suborbicular in outline, bifid to the middle or 

 less, with acute divisions, and their margins are entire or nearly so. 



The female inflorescence is sometimes borne on a leading 

 branch and sometimes on a short branch, and in the latter case 

 the number of branch-leaves may be reduced to one or two pairs. 

 Variations in the length of the branch are often to be observed in 

 a single individual. In certain species the branches bearing peri- 

 anths are invariably simple ; in other species subfloral innova- 

 tions are constantly produced. The bracts are larger than the 

 leaves which immediately precede them, but are sometimes ex- 

 ceeded in size by the leaves on robust and sterile shoots ; they 

 are unequally bifid and usually distinctly complicate, the lobe 

 being broad and rounded, and the lobule narrower and com- 

 monly acute. The bracteole is similar to the underleaves, but 

 is larger and less deeply bifid. The perianth is distinctly com- 

 pressed, the lateral keels being sharp ; the antical face is plane or 

 nearly so y and the postical keel is either rounded or two-angled ; 

 the apex is broad, varying from rounded to slightly retuse, and the 

 beak is distinct but sometimes very short. The antheridial spike 

 sometimes occupies a short branch and is sometimes terminal on a 

 longer branch ; the antheridia usually occur in pairs, but the bracts 

 themselves offer no generic characters of importance. 



As thus emended, Clieilolejeunea is apparently more closely 

 related to Euosmolejeunea than to any other genus. Typical. spe- 

 cies of the two genera are, to be sure, amply distinct. Such a 

 Euosmolejeunea, for example, as E. trifaria (Nees) Schiffn. is char- 

 acterized by its yellowish-green color, by its large and cordate 

 underleaves, by the constant presence of a subfloral innovation, 

 and by its sharply keeled perianth. Unfortunately there are other 

 species of the genus in which all of these differential characters are 

 not clearly exhibited. The underleaves, for example, may be small 

 and cuneate at the base, and the antical keel of the perianth may 

 be low or even obsolete. It is sometimes necessary, therefore, as 

 in other similar cases, to rely on a combination of characters rather 

 than on a single generic difference. 



