C. bicolor Bentham, J. Bot. (Hooker) 2: 139. 1840. — "Uruguay, Tweedie." — Holotypus, Tweedie s.n. from rio Uruguay, K (hb. Hook.) = NY Neg. 1981 (2 plants at bottom of sheet)!.

C. microphylla Bentham, J. Bot. (Hooker) 2: 139. 1840. — "Minas Gerais. R Claussen." — Holotypus, Claussen 24 e Cachoeira do Campo, K (hb. Benth.)! = NY Neg. 1979. — Feuilleea multifoliolata O. Kuntze, Revis. Gen. Pl. 1: 190. 1891.

C. myriophylla Bentham, London J. Bot. 3: 111. 1844. — "Brazil (Minas Geraes?), Sello, Pohl, Miers!' — Lectoholotypus, Sello s.n., K! = NY Neg. 1980. — Feuilleea myriophylla O. Kuntze, Revis. Gen. Pl. 1: 188. 1891.

C. peckoltii Bentham, Trans. Linn. Soc. London 30: 555. 1875; & in Martius, Fl. Bras. 15(2): 425. 1876. — "Habitat ad Canta Gallo [= Cantagallo, near 22°S, 42°30'W], prov. Rio de Janeiro, Peckolt n. 399." — Holotypus not found in hb. Mart. (BR). — Feuilleea peckoltii O. Kuntze, Revis. Gen. Pl. 1: 188. 1891.

C. microcalyx Harms, Feddes Repert. Sp. Nov. Regni Veg. 17: 91. 1921. — "Brasilien: Minas Geraes, Caraca (Glaziou no. 14648)." — Holotypus, †B = F Neg. 1246; isotypi, K! = NY Neg. 1987, P!.

C. falcifera Ducke, Arch. Jard. Bot. Rio de Janeiro 3: 71. 1922— "[BRAZIL. Pará:] . ad stationem Arumateua viae ferreae alcobacensis prope cataractas inferiores fluvii Tocantins, l[egit] A. Ducke 4-1-1915 n. 15.650, fruct, 15-7-1916 n. 16.256." — Syntypi, RB n.v.; isosyntypus, Ducke 15650, MG = F 602743, photo + fragm!.

C. parvifolia sensu Burkart, 1952: 111; Cabrera, Man. Fl. Aired. Buenos Aires 246, fig. 82(A-C). 1953; Burkart in Parodi, Encicl. Argent. Agri. Jard. 455, fig. 126(B). 1959; Fl. Buenos Aires 4(3): 421, fig. 133. 1967; Jozami & Muñoz, Arbust. Prov. Entre Ríos 118, 120. 1982; Bemardi 1984: 175; Hoc, 1992: 212, fig. 4 + map 2; Stannard, 1995: 382.

C. bicolor sensu Bentham, 1844: 107; 1875: 555, exclus. syn.; 1876: 425.

C. microphylla sensu Bentham, 1844: 110; 1875: 555; 1876: 424.

C. myriophylla sensu Bentham, 1844: 111; 1875: 555; 1876: 425; Glaziou, 1906: 188 (n. 14648, P! exclus. n. 10620, P!). C. microcalyx sensu Renvoize, 1981: 67; G. P. Lewis, 1987: 175.

Slender arborescent shrubs or subshrubs (0.5-)l-4 m with fuscous, stiffly virgate long-shoots and narrow multifoliolate primary lvs, the young growth and especially the dorsal face of lf-axes finely pilosulous with spreading or forwardly subappressed, white hairs to 0.25-0.9 mm, the tiny, closely imbricate, either glabrous, or microscopically cioliolate, or loosely ciliate (cilia to 0.4—1.1 mm) lfts subconcolorous, the umbelliform capitula arising singly from brachyblasts, subtended either by a developed coeval If or by a pair of annotinous stipules, the older brachyblasts thatched with imbricate stipules; phyllotaxy distichous. Stipules lanceolate or triangular-lanceolate 1.5-6(-7) mm, at first soft green, early stiff, dorsally 1-9-nerved, persistent but becoming dry fragile, often blanched in age. Lf-formula variable, even on one plant, between primary lvs and those of brachyblasts, (v-)viii-xvii (-xx)/(20-)24-52; lf-stks of primary lvs 2-8(-9.5) cm, the petiole 3-12 mm, the longer interpinnal segments 2.5-6 mm, the stalk little swollen at insertion of pinnae, the ventral groove continuous or weakly bridged; rachis of longer pinnae (8-) 12-22 mm, the interfoliolar segments 0.15-0.5 mm; pulvinules 0.1-0.2 mm; lfts linear from obtusely auriculate base, obtuse or apiculate, straight, the longer ones (2-)2.2- 5.2(-6) x 0.5-0.9(-l) mm, 4—6.5(-6.9) times as long as wide; midrib simple or very faintly pinnate, subcentric, immersed on upper face. Peduncles slender or subfiliform 1-3.4 cm, either bracteate or not, the bract when present ±1 mm, inserted either above or below mid-peduncle; capitula 6—16-fld, the fls ordinarily heteromorphic, the peripheral ones subsessile to slenderly pedicellate, the terminal one (sub)sessile, not or scarcely longer than the rest but broader, the floral receptacle 1-1.5 mm, sometimes produced as a short terminal pedestal; bracts of outer fls submembranous, caducous, those of further fls wanting; perianth either reddish or greenish-yellow with rubescent teeth and lobes, thinly puberulent distally or almost glabrous; PERIPHERAL FLS: longer pedicels (0.4-)0.7-4 (-4.5) mm; calyx vase-shaped 1.8—2.6(—3) x 1.4—2.2 mm, weakly 5-nerved, the obtuse teeth 0.25-1 mm; corolla vase-shaped 4.2-7 mm, the ovate lobes 1.1-1.6 mm; androecium 12-32-merous, (19-)21-50 (-54) mm, the stemonozone 0.45-0.8 mm, the tube 1.5-3.1 mm, the tassel of filaments bicolored, white or pallid in lower half, carmine distally, the color deepening with age; ovary subsessile, at anthesis either glabrous or distally puberulent; intrastaminal nectary 0; TERMINAL FL (in some capitula abortive): broadly campanulate, with 5-lobed intrastaminal nectary 0.6-0.8 mm tall, the stamens sometimes more numerous than those of outer fls. Pods in profile (4-) 5-10.5 x 0.6-1.2 cm, straight or gently retro-arcuate, the dilated sutural ribs ±2-3.5 mm wide in dorsal view, the plane, recessed valves stiffly coriaceous or lignescent, transversely sinuously venulose, the whole pod either pilosulous overall, or glabrous except for puberulent margin, or glabrous except for red-granular faces; seed-funicles basally dilated; seeds (few seen) ±6.5-7.5 x 5 mm, the U-shaped pleurogram 6x3 mm.

On river banks, rocky shores, and in gallery-mar- gins, n.-ward around outcrops in campo cerrado, from near sea level on dunes of coastal n.-e. Bahia and on the Plata estuary in Argentina to 950-1150 m on the Brazilian Planalto and 1250 m on Pico de Almas in Bahia, discontinuously dispersed over e. Brazil from s. Maranhão (Grajaú) and Ceará to interior Rio de Janeiro, w. to s. Goiás and w. Paraná, thence w. into s.- e. Paraguay and along river banks s. to the w. Uruguay and the Plata estuary in Buenos Aires, Argentina; disjunct, in Amazonian campína, on lower rio Tocantins in Pará (near Camet Arumateua) Brazil. — Map 12.

-Fl. (VIII—)X-I(-II). — Angiquinho; plumerillo; borbas de obispo', niho azote; flor de seda; chicote de niho (Argentina, Hoc).

As might be foreseen from the synonymy cited above, C. parvifolia varies considerably in features of which the importance has been exaggerated: leaf- formula, pubescence (especially of the fruit), length of peripheral pedicels, flower-size, and number of stamens. Bentham acknowledged close similarity between C. bicolor (an admitted synonym, now untenable, of C. parvifolia), C. myriophylla, C. peckoltii, and C. microphylla, but noted these particularities (see his conspectus of Calliandra, Bentham, 1876: 408) of each: in C. bicolor only 3-6 pairs of pinnae; in the rest 10-20 pinna pairs coupled in C. myriophylla with densely pubescent young foliage and scarcely pedicelled flowers about 6 mm long, in C. microphylla with glabrescent leaves and sessile flowers about 4 mm long, and in C. peckoltii with glabrous leaves and pedicellate flowers 4 mm long. The fruit of C. bicolor was pubescent, that of C. myriophylla glabrous, and that of C. microphylla facially pubescent but marginally glabrous. In material now available for comparison, the leaf-formula, leaf-pubescence, and flower-size are independently and continuously variable, and I have found no credible geographical or morphological patterns in them. Vesture of the fruits needs further study, for few have been collected, and the ovary at anthesis is always glabrous, acquiring vesture only after fertilization. There are trends toward higher leaf-formula and longer pedicels in Brazil northward from state of Rio de Janeiro, and these are accompanied by slightly shorter and sometimes narrower calyx; whereas from central Goiás to Paraguay and Argentina the pedicels are often (but not in fact always) shorter and the corolla a trifle longer, but not in concert with pinna- or leaflet-number. I evaluate C. microcalyx, a name taken up by Renvoize (1981) for the Bahian forms of C. parvifolia, as essentially equivalent to C. peckoltii though the leaflets are a little fewer. Leaflet-number is, however, quite variable between leaves of primary long-shoots and those of short-shoots, which may be the only leaves present at flowering time. I have found, in addition, that the androecium in plants that approximate the C. peck- oltii-microcalyx form is commonly less than 20-mer- ous as contrasted with 20-32-merous elsewhere in the species. It might in consequence be possible to recognize in C. parviflora two varieties divided approximately by the 45th meridian (see Map 12), but certainly not more than one independent species. The altitudinal range of C. parvifolia is notable. It is assumed that the riparian populations along the lower Uruguai and Paraguai rivers are colonists from further north, where precisely similar forms occur at elevations up to 650 m. In Bahia C. parvifolia is well known from elevations of 900 m upward in Chapada Diamantina and recurs, somewhat surprisingly, on dunes of the Atlantic coast north of Salvador. The pedicels in this coastal form are exceptionally long (to 4.5 mm, Queiroz 2534, HUEFS).