tion for nearly the entire Amazonian watershed. The 
only major drainage-area for which, to the best of my 
knowledge, absolutely nothing is known is the Inirida 
River, north of the Vaupés. Among the minor drain- 
age areas which we still must explore are the great sand- 
stone mesas whose rivers pour their waters into the Yari 
River, emptying into the Caqueta below Araracuara; the 
Cahuinari, also flowing into the Caqueté; and the small 
but geographically strategical Cotuhé, entering the Put- 
umayo River near the Brazilian border. 
It is not always easy to define certain species of Hevea. 
The identifications of some collections must be made 
with reservation because of incomplete material. For the 
most part, however, the broad specific lines have been 
established. It seems to me that in eastern Colombia 
there may exist a number of subspecific variations which 
cannot easily be accommodated to concepts already de- 
scribed and established. There are perhaps even some 
undescribed species to be found in the more remote parts 
of Colombia’s Amazonian watershed, especially those 
areas which we now know to be rich in endemics. 
In the present short summary, I have purposely re- 
frained from consideration of any but the most easily 
recognized and well-established subspecific concepts. The 
complete and critical treatment of the subspecific varia- 
tions in Hevea in Colombia must await much more study 
and exploration. For one thing, we still do not have 
enough representative flowering and fruiting specimens 
to attempt to present even an outline of all the varieties 
and forms of the principal species. Furthermore, we are 
learning that bark characters, hitherto unstudied, except 
in isolated cases, are of unexpected importance in dis- 
tinguishing between subspecific variants in Hevea; we 
must retrace some of our steps and examine the barks of 
thousands of trees in numerous areas before making a 
[ 3 ] 
