Evans and é 
ok ae ARCHEOLOGY IN BRITISH GUIANA 301 
ethnographic evidence is that metal griddles were preferred but not 
always available in large enough supply to eliminate the necessity of 
manufacturing pottery ones occasionally. 
The strikingly short duration of Rupununi Phase villages, implied 
by the sparse accumulation of sherd refuse at all habitation sites, finds 
its explanation in the ethnographically documented practice of aban- 
donment of the village when a death occurred. The strictness with 
which the Indians adhered to this custom regardless of the newness or 
convenience of the settlement caused comment from several early vis- 
itors (e.g., Schomburgk, 1836, p. 238), who appear to have considered 
it wasteful. In view of the contrast between the depth and density 
of refuse at these sites and at those in other parts of British Guiana 
visited in this survey, it seems probable that in general thin deposits of 
the Rupununi Phase type can be interpreted as reflecting the aban- 
donment of the village for supernatural reasons rather than ecological 
ones. 
Although the existence of trade relations between the Taruma and 
the tribes of the Rupununi savannas is attested by contemporary 
observers (Im Thurn, 1883, p. 273; Farabee, 1924, p. 21), archeologi- 
cal verification is practically nonexistent. The only evidence is the 
presence of a few sherds of diagnostic Taruma Phase decorated pot- 
tery types at the Rupununi Phase site of R40, which dates from 
around 1850. Several sherds of Koriabo Phase decorated types came 
from the same site (pl. 36, f,7) and the significance of these is more 
dubious. Although the location of R-40 on the bank of the Rupununi 
River, a potential route of communication between the interior and 
the coast, makes trade relations between the two phases reasonable, 
other evidence suggests that the aboriginal cultures of the Northwest 
District were no longer in existence in the 19th century. Aside from 
the possibility that the Koriabo Phase sherds at R-40 are not related 
to the Rupununi Phase occupation of the site, the only apparent ex- 
planation is that the Koriabo Phase either lasted longer than has 
been assumed in the Northwest District, or retreated toward the 
interior where European contact was less intense and survived there. 
Perhaps future archeological investigation along the lower and mid- 
dle course of the Essequibo will provide the missing evidence and 
permit a choice between these alternatives. 
The ethnographic data disagree with the archeological record in 
two important points. One concerns the method of disposal of the 
dead, which is reported by all observers to be primary interment. 
The Wapisiana are said to have a tradition of an earlier practice of 
cremation, but there appears to be no contemporary record of urn 
burial. Considering the fact that missionaries were among the 
513186—60——_23 
