10 Cephalozia. 



either entire, or very shortly bifid, or subdenticulate at the apes.. The 

 main difference, however, is in the female bracts, which scarcely differ 

 from the stem-leaves except in being slightly larger and longer, but are 

 quite conformable to them at the apes, and not at all more deeply di- 

 vided (as is usual in most other Cephalozia). C. Boschiana (Luc.) is 

 esactly intermediate between Lembidium and Odontoscldsma, having the 

 peculiar bracts and the boatshaped leaves of the former; but the loosely- 

 imbricated, obliquely-inserted and succubous leaves of the latter. The 

 trigonous j)erianth and its included ofgans, ard the monandrous male 

 bracts are esactly as in Odontosclmma and Eueephalozia. 



If we start from Eueephalozia in another direction, it brings us to 

 Alobiella nobis : a subgenus confined, so far as hitherto known, to 

 tropical America, where I have gathered 4 species, one of which (Jung. 

 Husnoti Gottsche) has been found also in the Antilles, by M. Husnot. 

 Here also, as in OdontoscMsma, the leaves are normally entire, and only 

 by rare exception cloven at the apes; bat, instead of the concave and 

 rather closely reticulate leaves of Odontoschisma, we find nearly flat, 

 oblong or lanceolate leaves, with large pellucid elongate cells, i/ie— 1/12 °^™ 

 long, and haK as broad, which give the plants at first sight more the 

 aspect of Kantia Trichomanis than of Cephalozia. In C. Alobiella integri- 

 folia n. sp., indeed, an incubous leaf is sometimes (though very rarely) 

 interposed among the normal succubous leaves, which makes the resem- 

 blance to Kantia more striking. The very long perianths are mostly 

 laciniate at the constricted mouth, and the laciais cUiiform. Three of 

 the species are cladocarpous, but the fourth [C. Al. acroscypha n. sp.) is 

 acrocarpous, having the perianth consfcantly terminal on the main stem ; 

 yet in every essential feature it is a Cephalozia. C. Al. macella n. sp., 

 by its habit of slender C. bicuspidata, and by the variable leaf- apex — 

 rounded, obliquely acute, or bidentate — unites this subgenus to Eu- 

 cepihalozia. 



A minute Amazonian Eiicej^halozia [C. micromera n. sp.) diverges 

 from the type by the minute leaves, having the antical lobe much 

 smaller than the postical, and not unfrequently quite obsolete, so that 

 the leaves become simple and acuminate ; but the globose cells — fequi- 

 latero-hesagona^] by mutual pressure — forbid its being placed in Alobiella. 

 It affords, however, a direct transition to the curious subgenus Zoopsis 

 Hook. f. et Tayl., through C, Z. monodactyla n= sp. 



