52 BULLETIN 1422, U. S. DEPARTMENT OF AGRICULTURE 
In the neighborhood of Sena, Bolivia, some seeds were collected 
which were only slightly larger than those from the Igarape do 
Soldanho, Matto Grosso, whereas others were near the average size of 
seeds from Sumatra and Java. All these seeds came from large 
trees, as may be seen by referring to Table 10, where measurements 
of these trees are given. 
From the Acre Territory a larger number of seeds was obtained. 
These are all normal in size. It may be noted that tree No. 2, Cobija, 
and the six young trees at Cobija bore the largest seeds. These are 
all young trees. 
The seeds from the Museu Goeldi are very like similar seeds from 
a plantation in Sumatra. They are all from planted trees, most of 
them probably grown from seeds from the lower Amazon. For the 
sake of comparison, measurements of some seeds of H. spruceana, H. 
guianensis, H. randiana, and H. sp. (itaubaf) are included in Table 12. 
These show how distinct some of the species are in respect to size of 
seeds. Such distinction is not to be found between any of the varie- 
ties of H. brasiliensis so far as could be determined. 
None of the seeds of Hevea brasiliensis collected in South America 
have shapes which can not be duplicated in a random collection from 
a plantation in the Orient, and no shape could be found characteristic 
of a special type of tree. 
The seeds from the Acre region and those from the Museu Goeldi 
at Para were collected soon after they fell, and the color patterns 
could be studied. These were entirely typical of the species as grown 
on the plantations and were of no aid in making varietal distinctions. 
To summarize, it may be said that none of the characters of the 
seeds collected were of any value in classifying the trees which they 
represented. 
FRUIT CHARACTERS 
All of the fruits examined resemble very closely those of plantation 
trees generally, and none were found characteristic of special tree 
types. Fruits with a number of locules other than three appeared 
to be fewer than in plantation trees as studied by the writer in 
Sumatra (17) ; but this difference may be only apparent and not real, 
since the number of fruits seen in South America was very small 
compared with those studied in Sumatra. 
Table 13, compared with Table 12, shows that the relation of seed- 
length to carpel length is about the same as that shown in the pub- 
lication cited (17). 
