Examination of a bud of an isosyntype (Spruce 2045, K) showed a 5-locular sparsely setulose ovary. Both this collection and Spruce 2184 (Macbride photograph 36230) have immature inflorescences, with the latter collection apparently having slightly denser lower leaf surface pubescence. Schultes & Cabrera 13917 (Mitu, Vaupes, Colombia) and Froes 28211 (Rio Uaupes, Amazonas, Brazil) agree well with Spruce 2184 in leaf pubescence and have mature expanded inflorescences (the primary lateral fruiting branches up to 6-10 cm long). Cardona 1329 (VEN), from the headwaters of the Rio Siapa on the Venezuela-Brazil frontier, accords well with Sinuce 2015 in inflorescence development. Three recent Venezuelan collections have mature foliage pubescence slightly less dense and inflorescence more mature and expanded than the Cardona material, but are certainly conspecific with it; these numbers (Maguire & Wurdack 34956, from Piedra Nunca, Rio Negro; Maguire, Wurdack, d- Bunting 36700, Maguire, Wurdack, & Maguire 11943, both from the uppermost Rio Yatua) and Cardona 1329 all have 5-celled ovaries. I concur with Macbride 's implied exclusion of this species from the Peruvian flora; all Loreto and Junin material purporting to be L. francavilllma has glandular-strigose hypanthia.
As compared with L. francavillana, L. divaricala (Naud.) Cogn. has an inflorescence as wide as or wider than long, a glabrous 3-celled ovary, and smaller (usually) leaf blades which are strigulose or short-strigose beneath only on the primary and secondary veins. Gleason's notes indicate that he included here L. rhamnifolia (Naud.) Tr. var. macrodon (Naud.) Cogn.; certainly the Sellow and Riedel material (K, US) of L. rhamnifolia var. rhamnifolia does not seem specifically distinct from L. divaricata; many recent British Guiana collections ascribable to this complex (Gleason 665, Hitchcock 17324, Tutin 313, Maguire & Fanshawe 22918, de la Cruz 1149, 1525, 1.985, 2323, 3869) show varying degrees of hypanthial glandulosity. The rather similar L. secunda (Don) Cogn. (of Amazonian Peru, Ecuador, and Colombia), also with gland-tipped hypanthial hairs, has however a 5-celled sparsely glandular-setulose ovary. L. secundiflora (DC.) Cogn. was characterized as with 4-merous flowers by Cogniaux; however, at least one Cogniaux-cited collection (Traill 257, K ) has 5-merous fiowers (with a glabrous 3-celled ovary). If the Martius holotype (which I have not seen) is conspecific with the Traill specimen, L. divaricata would revert to synonymy under L. secundiflora.