48 BULLETIN 1422, U. S. DEPARTMENT OF AGRICULTURE 
well on the plantations of the East where within a short time one can 
see thousands of trees of the same age which have a similar habitat 
and nurture. In the forests of Brazil and Bolivia it is very hard to 
collect leaf specimens from the mature trees on account of their 
great height. It is rarely possible to find a man who can or will 
climb these trees. In many cases leaves could be secured only by 
cutting them off with a shotgun or rifle. Leaves or twigs cut off in 
this manner often lodge in branches out of reach, and specimens are 
finally obtained only at a great cost of time, patience, and ammuni- 
tion. 
Leaves of a considerable number of trees of various types and from 
widely separated regions were studied, but they were of little use in 
tracing the relationships of the various strains. In general, the leaves 
of these trees showed less variability than those of the trees in the 
Orient. This may be due to the hybrid character of the Eastern 
trees; it may equally well be due to the smaller number of South 
American trees on which it was possible to make observations. 
FLOWER CHARACTERISTICS 
What has been said about the difficulty of collecting leaves applies 
even more forcibly to floral characteristics, since the collector can 
not visit the various regions during the flowering season of a single 
year. Some trees are found in flower at almost all times of the year, 
and one has to depend on these if he visits a region at other than 
the regular time of flowering. When such a tree is found, floral 
specimens are harder to obtain than leaf specimens because it is much 
easier to shoot down leaflets than flower clusters. 
The inflorescence of Hevea brasiliensis in Sumatra and Java has 
been well described by Arens (3), Sprecher (39), Maas (26), and 
Heusser (8) . Those of the Amazonian representatives of the species 
are in all respects like those described. Variations in South America, 
as in the East, occur principally in the size of the flowers. Certain 
trees are found which bear flowers considerably larger than the average, 
but the writer has never found that these were indications of a 
special type of tree. 
SEED CHARACTERS 
It is a difficult matter to collect representative seeds on a journey 
such as this. When the Madeira region was visited the seeds were 
not yet ripe, and the seeds of the past season had either germinated 
or decayed. It was found that the shells of many of the germinated 
seeds could be found at the bases of the young seedlings, and some- 
times a considerable number could be picked up under a given tree. 
These still retained their original shape and size, but were usually so 
discolored as to show nothing of the color patterns. The loss of the 
color patterns is probably not serious, as they seem to vary with 
every tree. In Sumatra, the writer (17) studied collections from 
hundreds of trees, but never found two collections which showed 
the same color pattern. Yet the pattern for a given tree is so definite 
that it may be used in determining which of the seeds picked up under 
a certain tree belong to that tree. The range of variation in the seeds 
of a given tree will include some seeds which resemble certain rather 
wide variations from the type of another tree, but they can still 
be distinguished by plantation laborers with sufficient certainty 
for routine selection of seeds for planting. 
