1953] 



BOTANY OF THE GUAYANA HIGHLAND 



123 



cies would contribute nothing constructive to the situation. In subsequent studies 

 it will most likely be considered only of sectional importance, but any such in- 

 novations must, for the moment, be postponed to such time that very careful, de- 

 tailed study of the entire subtribe Cuspariinae may be undertaken; such a study 

 is contemplated for the future by the author. 



Ravenia ruellioides Oliver, 



Cerro Sipapo, Maguire & Politi: Rare, small shrub, along banks of Lower 

 Cano Negro, January 1, 1949, 28095; rare, small shrub 5 dm. tall, fls. scarlet, 

 stream banks of Lower Carlo Negro, January 1, 1949, 28111; infrequent, shrub 

 1 m. high, corolla coral-colored, Cano Profundo, Cerro Sipapo, January 10, 1949, 

 28271; occasional, shrub 1 m. high, flowers tubular, coral-red, watercourse in 

 Upper East Basin, 1800 m., January 20, 1949, 28459; occasional, shrub 1 m. high, 

 fls. scarlet, South Savanna, South Basin, January 26-28, 1949, 28667. 



This very uniform species probably extends over the entire length of the 

 Pacaraima System. It was described originally from collections from Mount Ro- 

 raima where it has been since recollected; now it is known from Auyan-tepul, 

 Ptari-tepuf toward the northern end of the System, and the collections cited above 

 represent the southernmost extension of the range. 



MALPIGHIACEAE 



Diacidia Grisebach and Sipapoa Maguire. 



The first collection, a single specimen, of the new genus Sipapoa was made 

 by Schomburgk in the "mountains of British Guiana." It was assigned to the 

 genus Coieostachys as C. vestita by Bentham, 22 who remarked at the similarity of 

 the enlarged calyx with those of previously known species of Coieostachys. But 

 he also pointed out that the ovary of the Schomburgk specimen is "entirely un- 

 divided, while in Coieostachys it is three-lobed and the style ventrifixed. " Later 

 Bentham and Hooker 23 transferred the species to Diacidia as Diacidia vestita 

 (Benthc) Benth. & Hooker, to which genus it is indeed closely related. Niedenzu 24 

 did not admit this remarkable species to his monograph of the family. Recently a 

 number of new collections have come into my hands which have required a revalu- 

 ation of the several relationships within the subtribe Byrsoniminae to which they 

 all belong. 



The genus Diacidia is characterized essentially by evaginate but connate 

 petioles, pariculate-scorpioid inflorescence, sepals which apparently do not be- 

 come ampliate in maturity, 10 stamens, and anthers which are barbate at the base 

 and curiously bicornute by two inwardly recurved spinose awns. In Diacidia 

 vestita (Benth.) Benth. & Hook, and four additional recently collected species, 

 the petioles are conspicuously connate-vaginate, the inflorescence simply spi- 

 cate, the sepals conspicuously ampliate in maturity, the stamens 8 or 6, and 

 the anthers bicornute but not barbate. It is obvious that these latter species are 

 generically inconsistent with Diacidia. As a consequence I propose the new 

 genus Sipapoa to accommodate them. 



As the two genera now stand, Diacidia consists of two closely related species 

 and is so far as known confined to the rain-forest area of the upper Rio Negro. 

 Sipapoa with 5 remarkably distinct species seems to be restricted to sandstone 

 areas of the Guayana Highland. 



"Lond. Jour. Bot. 7: 124. 1848. 

 "Gen. PI. I 1 : 253. 1862. 

 24 Pflanzenreich 4: 141. 1928. 



