139g.  Dalea carthagenensis (Jacquin) Macbride var. brevis (Macbride) Barneby

(Plate CXXVII)

Either glabrous to the spikes, or the young twigs puberulent, or the stems, leafstalk and -rachis, and lower face of leaflets thinly and loosely pilosulous; primary cauline leaves up to (1.5) 2-5.5 cm long, with 11-15 (17) leaflets up to (4) 5-15 mm long; peduncles terminal to main branches developed, up to 2.5 cm long, the later ones obsolete or nearly so, the whole inflorescence irregularly corymbose; spikes short and dense, the villosulous axis 3-11 mm long; bracts 2-4.5 mm long, gland- tuberculate dorsally, the lowest sometimes glabrous, the rest (or all) pilosulous; calyx 3.8-6.3 mm long, densely silky-pilose throughout, the tube 2-2.6 mm long, the intercostal glands small, 3-5, not conspicuous externally, the teeth triangular at base with a subulate or setiform tip, the whole 1.5-3.7 (4) mm long; banner ± 3.8 mm, the blade 2.4-2.8 mm long; keel 4.9-5.9 mm long, the claws 1.8-2.2 mm, the blades 3-4 (4.4) mm long, 1.7-2.6 mm wide; androecium 5.5-6.5 mm long. — Collections: 19 (o).

Dry hills and valley floors, in arid tropical brush or thorn-forest, 20-2000 m, locally abundant, Pacific slope and coast of s. Ecuador (Guayas, Azuay, Loja) and adjoining Peru (Piura, Tumbes), and upper Maranon valley in n. Peru (Cajamarca); apparently isolated (? native) in the upper Mantaro valley, at ± 2400 m, in Ayacucho, centr. Peru. — Flowering April to October, perhaps through the year.—Representative: ECUADOR. Azuay: Hirsh, s. n. (NY). Loja: Espinosa 592 (IAN, NY, US); Rose, Pachano & Rose 23,350 (NY, UC). Guayas: A. S. Hitchcock 20,120 (NY, US). PERU. Tumbes: Tumbes: Ferreyra 10,726 (US), 12,289 (REGINA). Piura: Tabara: Haught 170 (F, NY, US). Huancabamba: typus of D. microphylla. Ayabaca: typus of D. vicina. Cajamarca: Jaen: typus of var. brevis; Weberbauer 7119 (F, US). Celendin: Ferreyra 15,060 (US). Chota: Ferreyra 785 (US). Ayacucho: Ayacucho: Soukup 4043 (F, US).

Dalea carthagenensis (Jacqu.) Macbr. var. brevis (Macbr.) Barneby, comb. nov., based on Parosela microphylla var. brevis (short, of the calyx-teeth) Macbr., Field Mus., Bot. 4: 103. 1927.— "Peru: between Shumba and Taen [= Jaen], Dept, of Cajamarca, April, 1912, Weberbauer 6191..." — Holotypus, F! = Field Neg. 50204, NY! isotypus, US!— Dalea microphylla var. brevis (Macbr.) Macbr., Field Mus., Bot. 13: 370. 1943.

Dalea microphylla (small-leaved) H. B. K., Nov. Gen. & Sp. (folio) 6: 482. 1823.— "Crescit locis montosis Peruviae, inter Loxam et San Felipe, prope Guancabambam, alt. 1028 hex." — Holotypus, labelled "Guancabamba", Humboldt & Bonpland 3536 (P, herb. H. B. K.)! — Parosela microphylla (H. B. K.) Rose, Contrib. U. S. Nat. Herb. 10: 106. 1906.

Parosela vicina (close, to the preceding) Macbr., Field Mus., Bot. 4: 103. 1927.— "Peru : between Olleros and the valley of the Quiros, Province of Ayavaca, Dept, of Piura, May, 1912, Weberbauer 6350..." — Holotypus, F! = Field Neg. 50209, NY! isotypus, US!—Dalea microphylla var. vicina (Macbr.) Macbr., Field Mus., Bot. 13: 370. 1943.

The var. brevis varies in amount and distribution of pubescence, the types of var. brevis itself and of D. microphylla having glabrous leaves, whereas those of the type of P. vicina are thinly pilosulous beneath. The hairy form occurs also in Ecuador, and there is no reason to suspect that it is ecologically or geographically separated from the glabrous ones. Variation in length of the calyx-teeth is more striking: the dorsal tooth varies from 0.8 mm shorter up to 1.7 mm longer than the tube. The material from the Pacific slope in Ecuador and neighboring Peru furnishes examples of a dorsal tooth 2, 2.3, 2.6, 3, 3.7 and 4 mm long, forming a random series, not to be sorted into a significant pattern of dispersal. It seems best, therefore, to treat D. microphylla and Macbride’s two varieties of it as minor variants. It will be noted that the three proposals were based on collections from the same general area in Piura and northern Cajamarca. So far as I can see, the material of var. brevis from the Maranon valley does not differ significantly from that of the Pacific slope. The one collection from Ayacucho is also inseparable morphologically. If native there, it forms a link between var. brevis of northern Peru and var. pilocarpa on the east edge of the Andean plateau in Bolivia.

The var. brevis often closely resembles the Mexican var. capitulata, apparently differing chiefly in the organization of the inflorescence, which remains more or less corymbose even in the drought-inhibited, nearly leafless stage. The calyx-teeth of var. brevis are on the average both shorter and less densely plumose-ciliate. In habit var. brevis resembles also var. pilocarpa, which has fewer leaflets and ultimately looser flower-spikes; and the geographically nearer var. trichocalyx, discussed under the following.