312 Botanical Society uf Edinburgh. 



minous, as in Bignoniacece, whereas in those of the Myoporacece the 

 emhryo is always contained within albumen. 



After the comparison of these several circumstances, the author is 

 unable to perceive the existence of any marked affinity between 

 Oxycladus and any genus of the Myoporacece, and therefore sees no 

 reason to alter the conclusion at which he formerly arrived, that this 

 genus, although deviating from the usual form of its fruit and seed, 

 bears in every essential respect all the characteristic features of a 

 member of the family of the Bignoniacece. It is not however in the 

 singularity of the large fleshy cotyledons, or the wingless state of 

 the seed, that Oxycladus is remarkable, for Mr. Miers has shown 

 that these occur in other genera of the Bignoniacece ; its peculiarity 

 consists in the development of only one of its many ovules, and in 

 the shape of its cotyledons, which in most other instances are 

 deeply cordate, or almost bipartite at each extremity, with the 

 radicle placed between the lobes : in this genus, however, they are 

 entire, oval, and plano-convex ; in Rhigozum they are likewise 

 fleshy, orbicular, and entire. 



The limits of many genera of Bignoniacece appear, Mr. Miers 

 adds, ill-defined, and the characters derivable from the seeds much 

 neglected. Fenzl and DeCandolle have done much in extending 

 our knowledge of the family, but the subject still requires farther 

 investigation, for he has observed many singular deviations from 

 recorded structure that have not yet been noticed. Should it be 

 found desirable to class Rhigozum with Oxycladus, the character 

 suggested for this tribe in his former paper would require modifica- 

 tion. In that of the Crescentiece, this name ought to be suppressed, 

 and that of Tanceciece substituted, with the same character there 

 indicated : all the genera of this section of DeCandolle's Prodromus 

 strictly coincide with the Bignoniacece in their completely 2-locular 

 ovarium, and in the development of their ovules on the surface of 

 the dissepiment, and they agree also with the genus Bignonia in the 

 form of their embryo : Crescentia and Kigelia, however, present 

 characters wholly at variance with the Order, because of their 

 parietal placentation. He is not, however, persuaded of the pro- 

 priety of establishing a separate order for these two genera, which 

 has been done upon high authority, when they might so well form a 

 good tribe of the Cyrtandrai ex. Crescentia, with its large amygdaloid 

 embryo, does not differ more widely from the Cyrtandracece, than 

 Adenocabjmna does from Bignonia : in habit and in floral structure 

 the two last-mentioned genera are scarcely distinguishable. 



BOTANICAL SOCIETY OF EDINBURGH. 



July 13, 1854. — Professor Balfour, President, in the Chair. 



The following papers were read : — 



1. "Notice of a new species of Cavlerpa" by R. K. Greville, 

 LL.D. &c. This paper will be found in the ' Annals ' for September, 

 and in the Society's Transactions. 



Professor Edward Forbes remarked, that the plant resembled 



