450 Plants collected by Schomburgk in British Chiiana. 



Chamisso and Schlechtendal have already remarked upon the di- 

 oical nature of this genus. My specimens of each of the above three 

 species have all exserted stamens and short included styles, and are 

 consequently probably all male and sterile, although the ovules 

 exist in the ovarium. 



83. Clei-odendron fragrans. Vent. Malm. p. et t. 70. — " Sandy soils by 

 the sides of hills, British Guiana." Schomburgk, n, 475. 



There is, however, probably some mistake in the locality. The 

 flowers are all perfectly double and sterile, and it is only a cultivated 

 plant in America. 



Tribe Avicennie^. 

 In this tribe the inflorescence is generally irregularly racemose, 

 the upper pedicels being usually simple, and the lower ones cy- 

 mosely trifid. The ovules are attached very near the apex ; and as 

 the radicle is always as in other Verbenacecc pointing to the base of 

 the carpels, the ovules are amphitropous, or nearly orthotropous. 

 Besides Amasonia and Avicennia, I have observed this structure 

 in Gmelina. 



84. Amasonia erecla (Linn. Fil. Suppl. p. 294) perennis v. suftruticosa, 

 erecta v. basi decumbens, pubescenti-scabra v. glabriuscula, foliis elliptico- 

 oblongis grosse dentatis basi angustatis, bracteis ovatis acutis basi in petio- 

 lum brevem angustatis, calycibus corollisque breviter pubescenti-birtis. Ta- 

 ligalea campestris, Auhl. PI. Guian. ii. p. 625, t. 252. Amasonia punicea, 

 Vahl, Eel. ii. p. 51. — Savannahs of the Rupunoony Schomburgk, n. 228. 



Folia inflorescentia et calyx ut in descr. VahUana. Corolla 9-10 lin. 

 longa, tubo incurvo, fauce parum ampliata, limbi quadrifidi laciniis reflexis 

 ciliato-fimbriatis, superiors breviore latiore emarginata, lateralibus latis ob- 

 tusis, infima acutiuscula. Stamina 4, infra medium tub! inserta. Fila- 

 menta basi dilatata puberula, caeterum glabra, adscendentia, apice declinata, 

 exserta. Antherae et stylus ut in descr. Aubletiana. Ovarium quadrilocu- 

 lare? quadri-ovulatum, ovulis ex apice axeos amphitrope pendulis. Bac- 

 cara maturam non vidi ; junior dipyrena videtur globosa, pyrenis dispermis. 



Notwithstanding the discrepancies in the descriptions given by 

 the authors above quoted, it seems most probable, as Vahl suspected, 

 that the three are but one species. The pubescence and the brac- 

 tese are very variable. The corolla is said by Schomburgk and Vahl 

 to be red, by Aublet to be yellow ; but as the bracts are in all cases 

 red, the more or less yellow in the corolla maybe of no importance. 

 The flowers are so much pressed in my specimens, that 1 could not 

 satisfactorily ascertain whether the ovarium was two- or four-celled. 



The following new and very distinct species has been gathered in 

 Brasil by Pohl and by Langsdorff\. 



