M. COEEEA DE MELLO ON SOME BRAZILIAN PLANTS. 267 



Similar to that of C diffusa^ and that it must Fig. i. 



belong to the genus Cayaponia of Manso. 

 To this same genus belong probably Der- 



mopliylla pendulina^ Manso, Z. c. 31, and 

 Bryonia ternata^ Vellozo, L c. t. 91. As to 

 the position of the genus Cayaponia in the 

 Order, although it cannot be included in the Transverse section of 



tribe Abobre^, Naud. (Benth. etHook. Gen. %Tsa!VL^o?S^^nt 

 PI. i. 819), in which the number of ovules is fit^» Thefunideofeach 



limited to two in each cell at the most, it can- ger^ba^th^^ 



not, nevertheless, cease to be placed next to ner one, raising the 



Trianosperma, from which it does not essen- ra%™eS^' !^ct^^^^^^ 

 tially differ, except in the greater number of the outer ones appear 



ovules in each cell of the ovary. I have said ^^"^"' ^^^^^ ^^' ^"^""• 

 above that I suppose the C. diffusa of Manso to be identical 

 with the Bryonia pilosa of Yellozo. I have been led to this 

 conclusion from the excessive diversity of shapes represented 

 in the leaves of C. diffusa^ sometimes scarcely angular, and very 

 similar to those of B. pilosa oft. 86, Fh Flum,, sometimes more 

 or less deeply 3-5- or 7-lobed, with broad or with almost linear 

 lobes; and I have seen leaves of all these different shapes on 

 one and the same individual ; but when they are seen separate 

 on different specimenSj one is led to suppose the existence of 

 three or more distinct species. 



The HypantTiera of Manso is probably identical with Fevillea, 

 Linn. The II. Guepira, Manso, is the same as P. co7*difoUa, 

 Veil. n. Plum. ic. x. t. 1 & 2, which, according to Martins, is 

 the F. trUolata, Linn. ' The anthers of Fevillea are 2-celled, but 

 BO organized that, after the emission of the pollen (which takes 

 l)lace through a single slit in the anterior face of the anther), 

 they appear to be 1-celled ; but if they are examined in the bud it 

 will be seen that the connective is produced into the poUiniferous 

 cavity, so as to divide it longitudinally into two perfect cells, and 

 that after the emission of the pollen it withdraws itself, so as to give 

 the unilocular appearance, a structiu-e identical with that which 

 has been so well observed by Dr. G. Dickie (Joum. Liim. 

 Soc. X. 54) in the anthers of Ganna speciosa, and which occura 

 also in the three species of Ganna which inhabit this place. 



With regard to Fevillea^ the question arises whether the 

 ovary is really, as described, 3-celled, and the seeds in the fruit 

 inserted on a trigonous axis. The margins of the three carpels 



