curves in the region of the anthers, exceptionally long, 
32.8-39.8 cm.; not known to fruit. 
Selecting seven characters which most frequently dis- 
tinguish D. candida from D. suaveolens, I find that in 
three, the leaf shape, calyx apex and corolla tooth length, 
‘Andrés’ is intermediate. In the length of its corolla and 
the conspicuous slender basal part of the tube, it is simi- 
lar to D. suaveolens, but the calyx width and free anthers 
are like those of D. candida. The S-curved style is unique 
among the tree Daturas. There seems little alternative 
for the present but to interpret cv. Andrés as a hybrid 
between D. candida and D. suaveolens. 
Common names: Andaqui borrachera. 
Andaqut is the name of a now almost extinct Indian 
tribe to the northeast at the head of the Magdalena Val- 
ley. The area lies many days’ journey away, mostly 
through the eastern lowlands. 
On the assumption that ‘Andrés’ is a hybrid derivative 
of D. suaveolens, a lowland species not known to occur 
in the Valley of Sibundoy, it is probable that it was 
brought to the Valley from the eastern lowlands, or pos- 
sibly directly from the Upper Magdalena area of the 
Andaqui people. In the immediate eastern lowland area, 
Datura is represented by several collections (Klug 1889, 
A, GH, NY, US: Cuatrecasas 10752, US; Schultes 
3472, GH) which resemble ‘Andrés’ but have connivent 
anthers and lack a curved style, and thus approximate 
D. swaveolens more. 
Reference to Datura dolichocarpa (Lagerh.) Safford 
in the Valley of Sibundoy (86) is probably to D. candida 
‘Andrés’. 
Cotombia: Comisaria del Putumayo, Valle de Sibundoy, alt. 2200 
m.—San Andrés. Shrub 2.5 m.; corolla white. Border, infreq. 18- 
IIl-1963, 650 (COL, ECON, PASTO, US).—San Andrés. Arbores- 
cent 3 m.; corolla white; style with S-curve in region of anthers. 
Roadside, infreq. 20-VIII-1963, 1314 (ECON).—2 km. SE Sibun- 
[ 223 ] 
