342 BUREAU OF AMERICAN ETHNOLOGY [Bull. 177 
tivated plants, then it would seem likely that they should be found 
in earlier sites when the economy was indisputably nonagricultural. 
In view of the small scale of Alaka Phase excavations, it is possible 
that the milling stones have a longer time span than existing data sug- 
gest. However, the fact that similar artifacts continue throughout the 
British Guiana sequence, in association with economies known to be 
based on agriculture, seems significant. In this connection, it should 
be noted that Rouse and Cruxent (1959, pp. 266, 229) also report a few 
grinding tools from both Barrancas and Los Barrancos styles at the 
mouth of the Orinoco. Thus, while it is obvious from archeological 
and ethnographic evidence that bitter manioc was a major staple in 
the Guiana area, it cannot be argued with equal validity that maize 
was absent here. The best that can be said is that the existing data 
are inconclusive so that either interpretation is possible. 
The correlation of pottery griddles with the use of bitter manioc 
is so widely attested ethnographically in lowland South America 
that the same correlation can be assumed when griddles are encoun- 
tered in the same area archeologically. It is more questionable, how- 
ever, to infer the absence of bitter manioc from the absence of grid- 
dles, since other methods of manioc preparation are still in use in the 
Tropical Forest area (e.g., Levi-Strauss, 1948, p. 363; Lipkind, 1948, 
pp. 181-182), and probably were more common in the earlier stages 
of domestication of the plant. In a discussion of the distribution and 
antiquity of griddles, it must therefore be kept in mind that these 
data reflect the spread of a particular technique of bitter manioc 
preparation and not necessarily the first spread of the use of the plant 
as food. 
In view of the ethnographic association of bitter manioc with the 
Tropical Forest area, it is interesting to note that the greatest antiq- 
uity of griddles appears to be marginal to that area. In British Guiana 
they appear earliest in the Northwest District, and diffusion along 
the coast toward the east was very slow. Cruxent and Rouse (1959, 
p. 244) report griddles from the earliest pottery making cultures in 
eastern Venezuela, which they date in the vicinity of 1000 B.C. By 
contrast, griddles do not appear at the mouth of the Amazon until the 
Arua Phase, which came from the north shortly before the first Euro- 
pean contact (Meggers and Evans, 1957, pp. 602-603). In the inte- 
rior of British Guiana they are even later, since the first pottery mak- 
ing groups here did not arrive before the end of the 17th century. 
This dating rules out the Guianas as the place of origin of griddles. 
At present, the earliest occurrences are Saladero at the mouth of the 
Orinoco and possibly Momil I on the north coast of Colombia. Rela- 
tively early dates, in the vicinity of A.D. 500, are attested from the 
Upper Orinoco (Evans, Meggers, and Cruxent, 1960, pp. 359-369). 
