REPORT ON THE BOTANY OF THE ISLANDS OF THE SOUTHERN OCEAN. 265 



Although our plauts do not agree in every particular with Reichardt's description 

 of Plantago stauntoni, we have no doubt of their being that species. Reichardt describes 

 the capsule as 4-seeded ; but he probably overlooked the solitary apical seed, which we 

 found in every one of the capsules opened, and also in the ovular state in our Plantago 

 pentasperma. 



With regard to dimorphism in the flowers of this species, Reichardt makes no allusion 

 to it, and it is possible that what we have taken for a form having the tube of the corolla 

 much shorter than the lobes is only a young state of the form represented in Plate XLI. 



Plantago pentasperma, Hemsl. (Plate XLII. — B. and C.) 



Plantago pentasperma, Hemsl., n. sp. 



jfferba annua ? acaulis, villosa, glabrescens. Folia lineari-oblonga vel subspathulata, 

 semi- ad sesquipollicaria, obtusa, deorsum attenuata, crassiuscula, enervia. Scapi erecti, 

 graciles, supra medium floriferi, bracteis quam flores longioribus vel brevioribus. Flores 

 dioici ? dimorphi, aliis corollae tubo quam lobi breviore et staminibus inclusis, aliis corolla? 

 tubo quam lobi longiore et staminibus exsertis. Capsula pseudo-trilocularis, ssepissime 

 5-sperma ; semina dimorpha, 2 collateralia oblonga in utroque latere septi, et 1 subhemi- 

 sphericum in apice septi. 



Amsterdam Island. — Not found in the lower parts of the island : commencing at an 

 altitude of about 500 feet, and ascending nearly to the summit — De Vlsle. 



The flowers of Plantago pentasperma are apparently dimorphic ; whether they be 

 functionally unisexual, or the one form unisexual and the other hermaphrodite, the material 

 is insufficient to determine. In the one form the tube of the corolla is much longer than in 

 the lobes, and the stamens apparently exserted, though we have not actually seen anthers ; 

 in the other, the tube is much shorter than the lobes, and the stamens apparently included. 

 We say apparently included, because the two or three flowers we have been able to examine 

 were young, and it is possible that in an older state the filaments would be much longer and 

 the anthers exserted. It is more likely, however, that the corolla of this form, as in analo- 

 gous forms of some northern species, never fully opens, and the stamens remain enclosed. 

 The pistil of this form seemed to be perfect, but the one ovary examined was very small, 

 and no ovules were detected. The specimens of the form having long-tubed corollas bear 

 nearly ripe capsules, and the form of the septum or placenta — for it is more like an irregu- 

 larly shaped placenta than a true septum — is very singular. It was not possible to deter- 

 mine with certainty whether the body bearing the seeds was attached anywhere except at 

 the base. When the seeds are ripe the placenta detaches with the operculum, with which 

 it may, or may not, have organic union. This placenta or septum is so formed that it 

 divides the lower two-thirds of the capsule into two cells, and leaves the upper third 

 one-celled. In each of the lower cells there are two oblong seeds covering the whole face 



(bot. chall. exp. — part ii. — 1884.) B 34 



