454 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
7. Cobaea villosa Standley, sp. nov. 
Stems rather stout, striate or subangulate, densely villous about the nodes 
and sparingly so elsewhere; petioles and tendrils villous to puberulent; leaflets 
subequal, similar, obovate to oblong-obovate or oblong, 40 to 65 mm. long, 16 to 
40 mm. wide, obtuse, mucronate, unequal and rounded or truncate at the base, 
thin, glaucescent, usually glabrous on the upper surface, sparingly puberulent 
beneath or glabrate, on petiolules 4 to 12 mm. long; peduncles solitary or 2 to- 
gether, 13 to 19 cm. long, slender or stout, frequently flattened and usually 
coiled in age, puberulent or glabrous; calyx lobes united only at the base, 18 to 
28 mm. long, 3 to 8 mm. wide, linear-oblong to lance-oblong, rather abruptly 
acuminate, conspicuously nerved, villous-ciliate; corolla yellowish green, 4 cm. 
long or slightly less, the tube campanulate, glabrous, the lobes about as long as 
the tube, ovate-triangular, acuminate, densely short-villous outside, glabrous 
within; filaments 8 cm. long or less, slender, villous at the base, the anthers 
yellow, 1 cm. Jong; capsule elliptic-oval, 4 cm. long, acute, glabrous, the cells 
8 or 4-seeded; seeds irregularly oval or oblong, 18 to 21 mm. long, the wings 
very broad, finely reticulate-veined. 
Type in the U. S. National Herbarium, no. 575607, collected in Salvador by 
Carlos Renson (no. 218). 
ADDITIONAL SPECIMENS EXAMINED: 
SaLvapor: San Salvador, Velasco (J. D. Smit, no. 8882). 
From Cobaeca lutea the present species differs in about the same respects as 
does C. viorna, besides having villous-ciliate instead of usually glabrous calyx 
lobes. From the latter species it differs in its villous stems, broader, obtuse, 
glaucescent leaflets, and villous-ciliate calyx segments. 
8. Cobaea lutea Don, Edinburg Phil. Journ. 10: 112. 1824. 
Cobaca macrostema Pay.; Don, loe. cit., as synonym; Hook. Curtis’s Bot. Mag. 
66: pl. 3780. 1840. 
Cobaeca acuminata DC.; Hook. loe. cit. 
Cobaea macrostoma DC. Prodr. 9: 322. 1845. 
Rosenbergia macrostoma House, Muhlenbergia 4: 24. 1908. 
TYPE LOCALITY: Originally given as “Ad Portum Guayaquil in Regno Quitensi 
Peruvianorum,” but Hooker states’? that this locality was probably incorrect 
and that the type came perhaps from Mexico. If, however, the plant that has 
usually been given this name is correctly determined, the type probably came 
from Guatemala. 
RANGE: Guatemala. Brand’ also reports a specimen from Salvador, and 
eredits the species to Costa Rica. It may have this range, but possibly these 
reports are the result of incorrect identifications. 
SPECIMENS EXAMINED: 
GUATEMALA: Laguna de Ayarza, Department of Jalapa, alt. 2,400 meters, 
Heyde & Lur (J. D. Smith, no. 8987). San Lucas, Department of 
Antigua, C. ¢ EH. Seler 2452. Between Guatemala City and Chiquimula, 
August 18, 1860, Hayes. Without locality, Heyde 240. 
Well distinguished from the related species by the large corolla. The speci- 
men figured by Hooker in the Botanical Magazine was grown at Kew from 
seeds sent from Guatemala by Skinner. 
It is impossible to justify the use of the name macrostema (or any of its 
variations) for this species. Don plainly publishes the plant as lutea, citing 
Pavon’s manuscript name macrostema as a synonym. Yet lutea has never been 
used by any other author to designate this species. 
‘Toe. eit. 
?In Engl. Pflanzenreich 27: 28. 1907. 
