161/Ia. Senna multijuga (L. C. Richard) Irwin & Barneby subsp, multijuga var. multijuga. Cassia multijuga L. C. Richard, l.c., sens. str.—Typus supra sub subsp, multijuga indicavimus.—Peiranisia multijuga (L. C. Richard) Britton & Rose ex Britton & Wilson, Sci. Surv. Porto Rico and the Virgin Is. 6(3): 350. 1926.
Cassia calliantha G. F. W. Meyer, Prim. Fl. Essequib. 169. 1818.—"In sylvis plantationis Konigsberg [near Georgetown, Guyana]."—Collected by E. K. Rodschied; no typus found at GOETT (1977), but the detailed description, as recognized by Bentham (1870, p. 123), decisive.
Cassia richardiana Kunth, Mimoses 139, t. 42. 1823.—"Crescit in ripa fluminis Atabapo (Alto Orinoco) . . . bords du Rio Atabapo pres de son confluent avec l’Orenoque [near lat. 4°N in Amazonas, Venezuela]."—Holotypus, labelled 1180, Rio Atabapo," collected by Humboldt & Bonpland in April, 1800, P-HBK! isotypus, B (hb. Willd. 7990)!—Reduced by Bentham l.c. to C. multijuga.
Cassia fulgens Wallich ex Vogel, Syn. Gen. Cass. 44 & Linnaea 11(6): 685 (descr. ampliat.). 1837.—". . . 5310 . . . Hbt. in India orient."—Holotypus, B (hb. Kunth.), not seen; isotypi, Wallich Cat. 5310 ex hort. Calcutt., K, NY (misit Bentham)!—Equated by Bentham, l.c., with C. multijuga.
Cassia ampliflora Steudel, Flora 26: 760. 1843.—"[Hostmann & Kappler . . . im Innem von Surinam gesammelt haben . . .] nr. 829."—Holotypus, P (hb. Steud.), 3 sheets! isotypi, K, NY!—Equated by Bentham, l.c., with C. multijuga.
(?) Peiranisia aristulata Britton & Killip, Ann. N.Y. Acad. Sci. 35(3): 182. 1936.—"Villavicencio, Meta, Colombia, December, 1928, Apollinaire Marie, type (Herb. Instituto de la Salle, Bogota)."—Holotypus f 1948, survives as photograph, K! clastotypus (fragm), NY!—Certainly referable to C. multijuga sens, lat., but, lacking stipules, ambiguous between var. multijuga and var. lindleyana.
Cassia multijuga sensu Bentham, 1807, p. 123, ex parte, exclus. syn. C. selloi, C. magnifica et C. semifalcata cum locis pluribus austrobrasiliensibus, t. 37 (legumen male depictum); Schery, 1951, p. 72, fig. 127; Dimitri & Rial Alberti, 1954, fig. 2; De Wit, 1955, p. 253; Worthington, Ceylon Trees 190. 1959.
Trees attaining 25 m with trunk 3(-4) dm diam, flowering precociously as shrubs as low as 3(-2) m, most commonly seen as slender trees with rounded crown ±6-15 m, the annotinous growth sometimes quite glabrous but the branchlets with lf-stalks and lfts (on dorsal or both faces) and axes of inflorescence commonly strigulose, pilosulous or subtomentulose with straight appressed, spreading, or curly and entangled, pallid or lutescent hairs up to 0.1-0.7 mm, the inflorescence sometimes in addition charged with erect orange or discolored setules mixed with the filiform trichomes.
Stipules 4-12 mm, unilaterally dilated at base and ± falcately incurved, the base of the firm, herbaceous blade 0.8-2.5 mm wide, undulately crimped or replicate on itself.
Larger mature lvs 13-31(-35) cm; gland at proximal pair (1.2-)1.5-4.5 x 0.3-0.8 mm, sometimes similar glands at a few distal pairs, between all other pairs commonly a tuft of spicules, sometimes a minute subulate gland; pulvinules (0.8-)1-2(-2.4) mm; lfts 16-37 pairs, inserted up to (4-)5-13(-16) mm apart along rachis, in outline oblong, oblong-elliptic, lance- or exceptionally linear-oblong, mostly (20-)23-43(-53) x 5-12(-13) mm and (2.7-)3-5.2(-5.8) times as long as wide, exceptionally (in upper Amazonian Brazil and adjacent Peru), up to 15-20 x (2.6-)3-4 mm, the secondary camptodrome venation sometimes fully immersed on both faces but the (5-)6-11 pairs of major secondary veins usually prominulous beneath, sometimes on both faces, tertiary venules then erratic.
Long inner sepals 4-6.5(-7) mm; longest petal 16-26 mm; androecium and pod of the species.—Collections: 296.
Riverine forest, both on varzea and terra firme, river- and creek-banks in rain forest, entering savanna and cerrado by way of forest-islands and gallery-forest, becoming locally abundant in disturbed forest or capoeira, from near sea level in the Guianas up to 850 m on the Guay ana Highland, to 1100 m in Ecuador, 750 m in Bolivia and 1000 m on the Planalto of s.-centr. Goias, widespread around the periphery and along rivers within the Amazonian Hylaea, the Guayana Highland and the Guianas, from s.-e. Colombia and Venezuela s. of the Orinoco through the foothills of the Ecuadorian and Peruvian Andes to the Yungas in Bolivia, e. in Brazil to coastal Amapa, the Amazon Delta in Para, and coastal Maranhao, s. in Brazil to the Amazon-Paraguay divide near 17°S in s.-centr. Goias; widely dispersed in tropical gardens both within and outside its native range (Trinidad, Martinique, Puerto Rico and Virgin Is.; s. Florida; tropical Africa; s. India, Java, Sumatra), reported from Central America and Mexico but the only specimens seen (from Honduras and the Canal Zone in Panama) are from cities and probably either street- or weed-trees.—Fl. irregularly throughout the year.—Marimari or San Francisco Marimari (Guayana); riariadan, kashadan (Guyana); palaka (Surinam), kabana-fau (Taki-Taki in French Guiana); arrepillo, una de gato, marupa (Amazonas, Venezuela); jarupari (Terr, do Roraima); sabueira, camunze (Para); mocairey (Huitoto of Loreto, Peru); quillocaspi, quillosisa (Loreto and San Martin, Peru).
The commonest and most widespread form of S. multijuga is characterized by the syndrome of basally dilated stipules and large membranous leaflets well spaced along the rachis. The populations differ much in pubescence and in flower- size, but erratically so. Trees with setulose panicle and sometimes a few setules along the leafstalks are known from the segment of the whole range that extends from Sierra Imataca and Mount Roraima in southeastern Venezuela south westward to the upper Rio Negro and (interruptedly) to Peruvian Amazonia, but does not replace the ordinary, simply pilosulous or glabrate form. Locally around Iquitos in Loreto, Peru and on the sources of Rio Jurua in Acre, Brazil, the species is represented by a notable variant with 20-36 pairs of exceptionally narrow leaflets up to (13-) 15-22 x 2.6-4(-4.3) mm relatively crowded along the rachis at points (2-)3-4.5 mm apart. This might suggest passage into distantly allopatric var. verrucosa, but has substantially fewer and (even though narrow) shorter leaflets; we interpret it as a minor, but visually striking variant.
All but the last of the synonyms listed above were recognized as such by Bentham and require no further comment. We refer here Peiranisia aristulata with diffidence, for it is known only from fragments of the holotypus lost with the collection of Instituto de la Salle in Bogota. These fragments seem certainly to represent S. multijuga sens, lat., but lacking stipules we cannot be sure that they are not var. lindleyana. The species, except for var. peregrinatrix introduced into Bolivar, has been collected in Colombia only on the Putumayo River and can hardly be expected as a native element of the llano flora in Meta. Perhaps a topotypus from Villavicencio will one day settle the question.
It seems likely that the earliest name applicable to S. multijuga is Cassia marimari Aublet (1775), for which there is a voucher in the Aublet herbarium at BM (=MO Neg. 2271). As published, however, C. marimari is merely a technical synonym of C. galegifolia Linn.