BE d 
apenean ARCHEOLOGY IN BRITISH GUIANA 343 
and from eastern Ecuador (Meggers and Evans, MS.). More evi- 
dence from other parts of the northern and western lowlands is 
needed, however, before an attempt to pinpoint the origin of the use 
of griddles for bitter manioc preparation can be considered anything 
more than wild speculation. 
The fact that several of the Phases distinguished by archeological 
evidence in British Guiana survived not only into historic times but 
to the present day provides an unusual opportunity for the checking 
of archeological techniques of classification and analysis and the in- 
terpretations derived from them. Although we are less concerned 
with the reverse, this situation also permits verification of ethno- 
graphic reporting. The only instance that we will cite here is the 
conflict between archeological evidence that the practice of secondary 
urn burial existed on the Rupununi savanna well into the 19th cen- 
tury and contemporary reports that the only method of disposal of 
the dead known to the Indians was direct interment (see p. 325). 
However, in:view of the frequently encountered tendency of Indians 
to conceal customs that. are frowned upon by white people, especially 
missionaries or priests, this contradiction is easy to explain. 
The existence of ethnographic and historical evidence for the Ru- 
pununi and Taruma Phases provides an independent means of check- 
ing the seriated site sequences, and consequently the validity of the 
classification of pottery into types upon which the principle of seria- 
tion is based. For the Rupununi Phase, the absence of deposits deep 
enough for stratigraphic excavation made it impossible to identify 
the direction of the trends of pottery type change by archeological 
means. Fortunately European trade goods existed in a number of 
surface collections and these could be dated. The dates that were as- 
signed coincided very closely with the chronological order derived 
from the seriation of the percentage occurrence of pottery types at 
the habitation sites (table N). This striking correlation is rather 
strong demonstration that the method of ceramic seriation and the 
premises on which it is based are sound. Agreement between the 
seriated site sequence and the historical records was equally good for 
the Taruma Phase. Here it was possible to identify the archeological 
remains of two villages visited by Farabee and their position at the 
end of the Taruma Phase sequence is exactly as it should be. These 
results do not of course imply that the method of ceramic seriation 
cannot be misused or the results misinterpreted, but they do suggest 
that the method itself is a sound approach to the problem of building 
a detailed, relative chronology for a phase. 
Ears few techniques for arriving at ar cheological in- 
ferences are as objective and systematic as pottery type classification 
and seriation. Consequently, it is easy to go too far in making inter- 
