560 ANNUAL REPORT SMITHSONIAN INSTITUTION, 196 



Along the Pacific coast of North America there was another food- 

 collecting pattern which paralleled in many ways that of the Eastern 

 Woodlands. ITere, by 2000 B.C. if not earlier, semisedentary societies 

 based upon fishing and acorn gathering were established all along the 

 coast from southern Alaska to southern California [2, pp. 133-137]. 

 In South America there were also ancient fishing societies along the 

 coasts. The Quiani phase [30] of northern Chile displays this adjust- 

 ment. On the Brazilian coast are the huge samhaquis^ piles of shell 

 refuse containing the skeletons and artif actual remains of food-collect- 

 ing peoples who lived along these shores probably as much as two 

 millennia before the beginning of the Christian Era [31]. Coastal 

 shell-mound dwellers are also Imown from Venezuela at about this 

 same period [32, 33]. 



I have mentioned that in both the North American and the South 

 American plains there were retentions of big-game-hunting patterns 

 into later times ; even these cultures, however, show the result of con- 

 tact with the neighboring food collectors in their possession of an in- 

 creasing number of food-grinding implements. This is exemplified in 

 the later North American Plains phases, such as the Signal Butte I 

 [34], and by the later phases in the Strait of Magellan sequence and 

 on the Argentine pampas [35]. 



INCIPIENT CULTIVATION 



The change from food collecting to a subsistence based upon plant 

 cultivation was one of the great turning points in human prehistory. 

 This is true of the New World as well as the Old, and there are in- 

 dications in both hemispheres that this switch-over was not a rapid 

 one, but that it was effected only over a period of experimentation. 

 It is this era of experimental or incipient cultivation in the New 

 World that I now wish to examine [36] . 



In the Americas it would appear that there may be at least four 

 distinct and semi-independent traditions of incipient farming. Two 

 of these are Nuclear American. The northern one, the probable 

 propagator of maize, was located in IMiddle America and in the adja- 

 cent deserts of northern Mexico and the southwestern United States ; 

 the southern one had its focus on the Peruvian coast. A third in- 

 cipient-cultivation tradition cent ered somewhere in the tropical forests 

 of the Amazon or Orinoco. Its existence is difficult to demonstrate 

 archeologically, but such a tradition is needed to explain the domesti- 

 cation of manioc and other root crops. A fourth, and distinctly 

 lesser, tradition rose in eastern North America in the Mississippi 

 Valley system. 



Tlie earliest evidence for incipient cultivation in any of these tradi- 

 tions comes from northern Nuclear America. The region is the north- 

 eastern periphery of Middle America, in the semiarid hill country of 



