38 BULLETIN 1422, U. S. DEPARTMENT OF AGRICULTURE 
Plantations started in the Amazon Valley now would have the 
benefit of the planting knowledge which has been developed by the 
Eastern planters. Many mistakes which were made in the beginning 
of rubber planting could be avoided. For example, no on£ now would 
think of planting rubber in swampy water-logged soil, but such soil 
was originally considered a requisite for the successful development 
of the trees. Friable well-drained soil with an ample water supply, 
the sort best suited to rubber trees, is abundant and should be care- 
fully chosen. 
Eastern experience has shown the importance of planting only 
stock of the highest quality. Formerly seed of any sort was planted, 
emphasis being laid only on the percentage of germination. Some 
planters insisted on procuring seeds from old trees, a practice which 
seems to have no scientific basis, but until recently none of them 
insisted that seeds for planting should be taken from trees of high 
yield. 
Plantations grown from miscellaneous seeds show great variations 
in the yielding power of the trees, and it is not always possible to 
save all the good trees and eliminate the poor ones in thinning oper- 
ations, because the distribution of good and bad trees is not uniform. 
Selection of seeds of high-yielding trees in plantations offers some 
advantage, but on account of the large proportion of cross-pollination 
it can not be depended upon to give progeny of much greater uni- 
formity than common seeds (19). Seeds from special seed gardens 
containing only high-yielding trees should be used. 
Budding young stock with buds from selected trees offers a means of 
obtaining a much higher yield than propagation by ordinary seeds. 
DonkersToot (7) describes the practical methods of budding on a 
plantation scale which were developed in Sumatra under the direction 
of the writer or in collaboration with him. The preservation of wood 
for budding, so that it may always be available for use and may be 
transported to considerable distances from its source, is absolutely 
essential to success in budding on a practical plantation scale. Meth- 
ods for doing this were developed by the writer, and they make 
possible the transportation of bud wood even from one continent to 
another. These methods also are given by Donkersloot in the paper 
cited. 
If budding were practiced in these hypothetical plantations some 
difficulty would be met in selecting trees from which to take buds. 
The actual yielding power of the jungle trees is difficult to determine, 
because the trees are of varying ages and some have been subject to 
much more severe competition in the jungle than others more fortu- 
nate. The experience of Eastern workers will need to be supple- 
mented by careful studies before suitable methods can be developed 
for the selection of trees, either for seeds or for bud wood. 
Recent trials of budded trees indicate that the character of the 
stock exercises a considerable influence on the scion. This is more 
pronounced in some cases than in others. That this might have been 
expected was shown by some preliminary experiments on the flow of 
latex which were performed several years ago by the writer on the 
Sumatran plantations of the United States Rubber Co. 
Records of individual yields of a group of trees were kept for some 
time, to determine the normal range oi variation from day to day. 
Some of these trees were kept in tapping as a check plat. The others 
