Woody vines; stipules small, interpetiolar, distinct; leaves decussate; petiole eglandular or bearing 2 small glands near middle; lamina eglandular or bearing 2–6 glands on the abaxial surface (sect. Dicella) or 2–8 glands on the margin (sect. Macropterys). Inflorescence terminal or axillary, a decompound panicle, the flowers ultimately borne in short pseudoracemes with decussate bracts, each bract subtending a 1-flowered peduncle with apical bracteoles; bracts and bracteoles eglandular. Flowers bilaterally symmetrical in all whorls; sepals abaxially sericeous, adaxially glabrous, pressed in against the androecium in anthesis, the anterior sepal eglandular, the 4 lateral sepals biglandular, all much enlarged and becoming papery and winglike in fruit, shorter than or subequal to mature fruit in sect. Dicella, much longer than mature fruit in sect. Macropterys; petals yellow, turning red in age in some species, abaxially densely sericeous or lanate, adaxially glabrous or tomentose; stamens 10, all bearing fertile anthers, the filaments abaxially sparsely to densely sericeous and adaxially glabrous, the anthers ± alike (the pair opposite the anterior-lateral petals smaller than the others), all strongly reflexed in anthesis, the connectives large and reddish, not exceeding locules (sect. Dicella) or exceeding locules, especially those of anthers opposite sepals (sect. Macropterys); pollen globally symmetrical, polyporate with the pores non-equatorial, and bearing broad fissures; receptacle glabrous between androecium and gynoecium; gynoecium comprising 3 carpels completely connate in the ovary; ovary densely velutinous or sericeous, the 2 posterior locules full-sized and fertile, the anterior locule rudimentary and empty; 2 posterior styles distinct, stout, glabrous, obliquely truncate or short-hooked at the apex and with stigmas large and internal; anterior style much reduced or absent. Fruit a dry, hard, indehiscent, obconic or spherical, nutlike structure with a thick fibrous wall, containing 1–2 seeds, most often only 1, subtended by dry wings formed by enlargement of the sepals; seeds large, consisting of a large embryo and no perisperm (sect. Dicella) or a small embryo and abundant perisperm (sect. Macropterys). Chromosome number: n = 10 (D. bracteosa; W. R. Anderson, 1993a); photos.
Seven species in two sections. Sect. Dicella comprises two species, D. bracteosa (Adr. Juss.) Griseb. and D. nucifera Chodat, in southeastern Brazil and adjacent Paraguay and Argentina. Sect. Macropterys M. W. Chase comprises five species: D. aciculifera W. R. Anderson in Costa Rica and D. conwayi Rusby, D. julianii (J. F. Macbr.) W. R. Anderson, D. macroptera Adr. Juss., and D. oliveirae M. W. Chase in South America from Colombia south to about 19°S. [maps] — Regional key to genera: Central America.
Dicella is easily recognized by its yellow, abaxially hairy petals, ovary with two fertile posterior locules and sterile anterior locule, two well-developed posterior styles with the anterior style rudimentary or absent, dry indehiscent nutlike fruit, and sepals enlarged and winglike in fruit. It seems unlikely that the small fruit-wings aid dispersal in sect. Dicella, but in sect. Macropterys the fruit-wings are large enough that it does seem possible that they function in dispersal. The two sections are quite distinct (see leaf glands, androecium, fruit-wings, and seeds in the description above) and both are well-supported in the latest phylogeny of the family (Davis & Anderson, 2010 [pdf]), but surprisingly there is no support in that tree for treating the two sections as a single genus. Still, they do come out together, and they are united by the shared characters in the description above, so until there is a good reason to treat them as separate genera we prefer to take the more conservative course of continuing to recognize the single genus Dicella.
Etymology: Grisebach stated that the name Dicella came from the Greek word dikella, a two-pronged hoe, apparently in reference to the two short-hooked stigmas (see the drawing of D. nucifera), which he regarded as an important character distinguishing this genus from Bunchosia. The name has also been taken (e.g., by Niedenzu, 1928) to refer to the fact that the ovary in this genus has only two of the three locules developed and fertile (di- means two- in Greek and the Latin word cella means chamber). The latter explanation of the name is surely an error.