[ 97 ] 



IV. On several instances of the Anomalous Development of the Raphe in Seeds, and the 

 probable causes of such deviations from the usual course of structure, especially in 

 reference to Stemonurus (Urandra of Tliwaites), tcith some Trefatory Hemarks on 

 that Genus. By John Miebs, Esq., F.B.S., F.Ii.S. 8fc. 



Read April loth, 185C. 



In a recent Number of Hooker's Journal of Botany (vol. vii. p. 211), Mr. Thwaites has 

 recorded a new genus, of which he gives the characters under the name of Urandra. 

 This he refers to " OlacacecB, tribe Icacinece.'" I have read over his description with the 

 utmost care, and am obliged to say that I perceive no difference whatever in the cha- 

 racters of Urandra, and those I have detailed of Stemonurus, to which genus he confesses 

 it is closely allied, differing only in all its flowers being fertile, in its small, not pulvinate, 

 stigma, and in the structure of its fruit. The character founded on the constancy of the 

 hermaphi'odite flowers in the plant which he describes, cannot be considered of the 

 smallest generic value, because this circumstance has been shown to exist not only in 

 some species of Stemonurus, but in the contiguotis genus Flatea, in which, althougli 

 some of the plants are imisexual, others are frequently hermaphrodite. I have also 

 shown that the stigma in Stemonurus {Gomphandra, Wall.) is small, and not large and 

 pulvinate as it had been described, but that it becomes subsequently immersed in the 

 epigynous gland which crowns the ovary, a cu'cumstance evidently not observed by 

 Mr. Thwaites ; and that it is tliis gland wliich assumes a pulvinate form on the summit of 

 the fruit, and not the stigma, which may always be seen hidden in a small central depres- 

 sion of the cushion. There remains therefore to be considered only the structure of the 

 fruit ; and Mr. Thwaites's details of the ovary, fruit and seed in Urandra, closely agree 

 with what I have observed in Stemonurus. In this genus, as in all others of the 

 Icacinacece, where the ovule is usually unilocular, the cell is always excentrically placed 

 on one side of the pistil, the point of suspension of the ovules not being from the summit 

 of the cell, but constantly inclined against the side on the line of the displaced axis of the 

 ovary : the lobes of the stigma are always two or fom-, thus showing prima facie that 

 the ovary is normally bilocular, and that the two ovides observed in the cell are reaUy 

 attached, near its summit, to the dissepiment, which, owing to the abortion of the other 

 cell, appears to form the waU of the ovary. I was fortunate enough to meet with the proof 

 of this conclusion in a ripe fruit of the closely allied genus Fennantia, where the ovary 

 is usually unilocular as in Stemonurus; but in the instance alluded to, the fruit was 

 regularly two-celled, and only one seed was perfected in each cell, the remaining abortive 

 ovule being stiU \isible on the dissepiment at the point of attachment of each seed ; this 

 partition was of thin textiire, and the nomishing vessels proceeding from the base formed 

 a longitudinal nerAOU'e in the line of its axis, extending thence to the point of attachment 

 VOL. XXII. O 



