— 19 — 



tibus. Marginibus incrassatis. Segmente antico sorifero pin- 

 nularum superiorum integro, in sorum singulum terminanti- 

 bus, eodem pinnularum inferiorum bifido, ramo antico sori- 

 fero, postico lobo sterili integro vel interdum dentato, 

 angusto, ssepe curvato. Indusio exserto vel basi anguste 

 marginato, ssepe incurvo vel recurvato, cylindrico, vel potius 

 angnste turbinato, ore integro; receptaculo incluso. 



Madagascar, leg. Humblot n° 280. (Muséum d'Historié 

 naturelle, Paris). 



Tbis most remarkable species was by Baker (Journ. Bot. 

 1884, p. 140) referred to T. rigidum Sw. as a variety; if 

 the spécimens under Humblot n" 280 examined by Baker 

 are identical witb that described above, his détermination 

 is very surprising. Our nev^ species is very remotely related 

 to T. rigidum^ and to other species of the group of T. 

 rigidum, a group represented by several species in Mada- 

 gascar and the Mascarenes. T. cornutum is, I am inclined 

 to believe, rather a species of the section Cephalomanes, 

 and possibly related to T. auriculatum Bl. Still it differs 

 from that species as well as from ail other known species 

 of Cephalomanes by its fuUy bipinnate lamina, by its vena- 

 tion, which is more flabellate than pinnate with the outer- 

 most branches of a pair characteristically convergent; the 

 margins are distinctly thickened, and the tips of the veins 

 are often united with the margin. The basai upper pinnules 

 bear generally two, rarely three, fertile segments on the 

 anterior side, while ail other pinnules are constantly uniso- 

 rous. I hâve not found a trace of an exserted réceptacle, 

 and it seems improbable that they ail are broken o£f; ail 

 known species of Cephalomanes hâve the réceptacle exserted 

 and very long. The spécifie name alludes to the stérile seg- 

 ment below the sori, which is distinct in the lower pin- 

 nules becoming gradually smaller in the upper pinnules aud 

 finally not at ail developed in the 2 — 3 last pairs of pin- 

 nules. Unfortunately the rhizome is unknown to me; it is 

 probably erect, thus agreeing with most species of Cephalo- 

 manes, not with T. auriculatum that has a scandent rhizome. 



