446 AKNTTAL REPORT SMITHSONIAN INSTITUTION, 1955 



The Brazilian agricultural scientists who have been studying this 

 problem are convinced that it is not economically possible to raise 

 crop plants on these sandy upland soils. Therefore, they are urgmg 

 the settlers to go down and dike the lowlands along the rivers and use 

 them for rice and pasture, and use the uplands only for building sites 

 for their farmsteads. The effectiveness of this system is being demon- 

 strated on their governmental experimental farms near Belem, and 

 it is promising. It should be remembered, however, that these river 

 lowlands are along branches of the local streams which discharge into 

 the Amazon estuary so that their levels are not greatly affected by 

 the changes in level of the Amazon River proper. 



THE AMAZON VALLEY 



The Amazon Valley has been more often described by popular writ- 

 ers than almost any other tropical region,^ and its imagined possi- 

 bilities for food production have been enlarged upon at great length.^ 

 Unfortimately, the potentialities seem to be very limited, even more 

 limited than the experiments near Bragan^a would suggest. As 

 Marbut and Manifold pointed out more than a quarter of a century 

 ago, the alluvial plain of the Amazon Valley is relatively very narrow, 

 often only a few miles in width. 



The alluvial soils constitute a sixth group. They are in general of two kinds. 

 (1) Well-drained loams and very fine sandy loams occupy the immediate banks 

 of the rivers in a narrow belt ranging from a few feet to a few hundred yards in 

 width. They lie on the natural levee and are moderately well drained, subject 

 to flooding for a short period each year, but highly productive. (2) Heavy, 

 imperfectly drained to poorly drained "back swamp" soils are often dark in 

 color and heavy throughout the whole section. They are subject to long periods 

 of inundation. Considerable areas are treeless. The belts in which they lie 

 contain many shallow lakes and swamps.* 



The reports are that during the high-water season of 1953 the 

 Amazon River levels were 15 to 20 feet above the street level at 

 Manaos. Higher up the Amazon, as at Iquitos, the usual variations 

 in the level of the river in different seasons of the year are from 50 

 to 60 feet or more. Obviously, it would be quite impossible to keep out 

 floodwaters by raising dikes high enough along the mid- Amazon which 

 would be needed in order to plant rice or other lowland crops in the 

 remaining portion of the alluvium. 



The largest-scale experiments in the development of the Amazon 

 Valley were certainly those of the Ford Motor Company. The engi- 

 neers planned first to exploit the forest timber and, by clear fellmg 



" Michael Scully, Amazing Amazon, The Rotarian, August 1952, pp. 35-36. 



'' Earl Parker Hanson, Mankind need not starve. The Nation, vol. 169, No, 20, 

 pp. 464-467, Nov. 12, 1954 ; New Worlds Emerging, 373 pp., 1949, New York. 



' O. F. Marbut and C. B. Manifold, The soils of the Amazon Basin in relation 

 to agricultural possibilities, Geogr. Rev., vol. 16, pp. 414-442, 1926. 



