Securidaca complicata H.B.K., Nov. Gen. Sp. 5: 423. 1823.
Securidaca orinocensis Rusby, Descr. New Sp. S. Am. PI. 40. 1920.
The early synonymy of S. pendula parallels that of S. coriacea (vide supra). Although the carinal crest in S. complicata was described by Kunth and can be seen on one flower in the Macbride type photograph (13089), Bennett in Flora Brasiliensis keyed the species as without a crest. Characterized by the rather elongate and externally strigulose alae and essentially evenly bialate fruit, S. pendula is common throughout the Orinoco drainage at lowland river margins and ranges also to Colombia (Vaupés basin and eastern lowland Boyacá).
Both S. longifolia Poepp. & Endl, and S. bialata Benth. are closely related to S. pendula, but are distinguishable by the relatively broader (length/width ratio 1.1-1.2 rather than 1.4-1.9, excluding the claw) alae and very asymmetrically bilobed fruit. The Venezuelan population of S. longifolia is certainly conspecific with that of northeastern Peru and upper Amazonian Brazil. Generally S. longifolia is distinguishable from S. bialata by the moderately dense and persistent (rather than sparse and early caducous) appressed short hairs of the lower leaf surfaces, the alar pubescence covering most of the external surface (rather than being restricted to the costal zone), and differently shaped fruit (very broad stylar sinus and widely diverging wings, rather than narrow sinus with parallel or incurving margins) with the larger wing 14-16 mm long (rather than 10-11 mm). The illustration in Flora Brasiliensis is not typical of S. bialata, and is probably from a very immature fruit. However, in the lower Amazon, much material has the leaf size and shape of S. longifolia, but the foliar pubescence and fruit of S. bialata; relying primarily on the fruit shape, I have referred collections from Belem to Amapa to S. bialata.
In 1952, Sandwith wrote me that Oort was quite wrong in reducing S. pubiflora Benth. to S. diversifolia; I agree with this, the relationship of S. pubiflora certainly being (perhaps, as Bennett indicated, synonymous) with 5. longifolia. In the same letter Sandwith also stated that the type number of S. bialata is Spruce 302, with alae strigose externally “mostly along the middle,” despite both Bentham’s and Bennett’s “glabrous” statements; also the Flora Brasiliensis citation of Spruce 2284 is erroneous, the collection being 2234 (later described by Chodat as S. warmingiana). Two other extra-Venezuelan binomials also probably involved in the S. bialata-S. longifolia complex are S. rosea Barb. Rodr. and S. engleriana Chodat.