To interpret the three phases taken by this species in Guiana, it has been necessary to review the many tropical American collections of it together with the relevant literature. The taxonomic confusion surrounding C. bicapsularis is partly biological and partly bibliographic. In Central America, West Indies, and coastal northern South America, C. bicapsularis is a glabrous shrub with membranaceous leaves, pale yellow flowers 1.5-2.0 cm in diameter and the longer filaments 6-9 mm long. This combination of characters is present in material collected in the Paria Peninsula, (Broadway 165, 588, 705) and coastal British Guiana (Jenman 5625; Irwin 503; Forest Dept. of British Guiana 7082).
A second form, essentially similar but for the pubescent branchlets, leaflet undersurfaces, sepals, and ovary; longer rachises; and larger flowers to 2.5 cm in diameter with the longer filaments to 15 mm long, found quite commonly in interior Venezuela and Colombia, extends as far east as the lower Orinoco (Rio Manimo, Buelte Triste, Bond, Gillin & Brown 148; May., Sacupana, Rusby & Squires 189). The cited Venezuelan collection is referred here.
A third variant, to which the cited Brazilian collections belong, differs from the second by the localization of leaflet pubescence on the midrib and margin, the nearly glabrous sepals, and the scanty ovary pubescence. This group extends from Colombia and Amazonia south to Paraguay.
Bibliographic confusion stems largely from the disparity between the application of C. bicapsularis by Linnaeus and Bentham. Linnaeus8 gave “India” as the province for the species but no doubt intended “India occ.9” Prevailingly the West Indian plant is the glabrous small-flowered type. However, in Flora Brasi-liensis Bentham10 11 cited Martius 129 Rio de Janeiro), 585, and 718 (both from Mato Grosso) as representing the typical C. bicapsularis. These materials are quite pubescent on the stems, ovary, and at the base of the leaflets. Bentham’s11 treatment of the West Indian plant is as a variant, with “leaflets rounder and thinner” than in more southerly populations. Nomenclatural adjustments, if any, will follow a more complete study of the species in the Planalto do Brasil.