angular, with a greatly thickened, suberose testa. n=12 
(Bristol 1193, 1255; GH). 
Both the calyx and corolla are variable in length, but 
the corolla rarely measures less than 20 cm. or more than 
30 cm. long. Fruiting is very infrequent, and this sup- 
ports the contentions of Joshi (8) and Barclay (3) that 
the tree Daturas, the populations of which often consist 
of asingle clone, are self incompatible. Despite the illus- 
tration by Ruiz and Pavoén (reproduced by DeWolf, 6) 
upon which Persoon based his description, a persistent 
calyx is rarely seen either in the field or in the herbarium. 
D. candida has been spread through much of Latin 
America and elsewhere from the northern Andes, where 
it thrives in the warm days and cool nights between 1500 
and 2500 meters elevation, 
D. affinis Saff. is based on the description of Brug- 
mansia arborea (1..) Lagerh. of which D. arborea L. is 
the basionym. Lagerheim noted its good agreement with 
Pavon’s specimens labelled **D. arborea Fl. Per.”’ (Ruiz 
and Pavén’s D. arborea L.), e.g... D. candida (Pers. ) 
Saff. In proposing 2. affinis, Safford relied upon the 
glabrous peduncle and the 2—5-toothed calyx to separate 
the concept from 2. candida. The white corolla is short, 
as in J. aurea, and Safford observed that D. affinis is 
‘very closely allied”’ to this species. 
DD. aurea Lagerh. differs from 2. candida only in its 
longer calyx and shorter corolla which remains yellow 
after emerging. As with D. affinis, the fruit and seeds 
are those of D. candida. 
D. Pittier’ Saff. is easily referable to D. candida in 
spite of the short corolla and anthers which chiefly dis- 
tinguish it. 
SPECIMENS EXAMINED: 
Mexico: Bottert 1090; Greenman 131; Muller 265; Pringle 6321, 
13126; Rose 3243; Seaton 94; Seler 56; Schultes & Reko 455, Guarr- 
[ 238 ] 
