tulate that the highly irregular formation of the leaves 
is caused by a critical concentration of a diffusable sub- 
stance, and that in the exceptional cases of very vigor- 
ously growing stem apices the critical level of concentra- 
tion is not reached. Such a substance might be viral or 
hormonal in nature. That the natives find the smallest, 
most malformed leaves the most toxic suggests that even 
the tropane alkaloids should not be ruled out of con- 
sideration as possible inhibitors of regular leaf develop- 
ment. As Robert Kahn has suggested to me, grafting 
experiments should help to clarify the development of 
the various leaf shapes. 
ae 
Seedlings 
Through thirteen months of constant observation of 
the Datura cultivars in the Valley of Sibundoy, no seed- 
ling tree was seen. It has been possible, however, to 
grow healthy seedlings of three clones, ‘Buyés’, ‘Ama- 
ron’ and ‘Quinde’, in the greenhouses of the Biological 
Laboratories of Harvard University. The seedlings 
within each cultivar are not uniform. Seeds collected 
trom D. candida ‘Quinde’ trees have produced seedlings 
the leaves of which resemble ‘Dientes’ (1596, 1611; 
ECON) as well as ‘Quinde’ (7607; ECON). Seedling 
variation from a single ‘Quinde’ fruit includes both 
‘Quinde’ and ‘Dientes’ types (7608; ECON), suggesting 
that a ‘Dientes’ tree may have been the pollen parent. 
‘Quinde’ seedlings exhibit the same failure to form 
‘Quinde’-type leaves on a very vigorously growing stem 
that was observed on mature trees in the Valley of Sibun- 
doy. As the growth rate of the seedlings slows some- 
what, these same stems that grew vigorously produce 
typical ‘Quinde’-type leaves. Assays of these seed grown 
‘Quinde’-type leaves for the presence of mechanically 
transmittable virus by Robert Kahn (pers. comm.) have 
[ 180 | 
