238 BUREAU OF AMERICAN ETHNOLOGY [Bull. 177 
sites in which only one cut was made. E-30 embraces in five levels 
almost the entire span of the Taruma Phase, while E-1, Cut 2, covers 
more than two-thirds of the sequence. In E-30, Cut 1, the first and 
second levels are at the top and middle of the sequence respectively, 
and the third level is almost as far below the second as the second 
level is below the first. Levels 4 and 5 are very close to level 3, and 
obviously represent the same occupation. The spacing of the levels 
in Cut 2 at E-1 is similar, and here the reoccupation theory is 
bolstered by the presence of Cut 1, which correlates with the second 
level of Cut 2.. This indicates that the location of the second village 
was predominantly in the vicinity of Cut 1, which the earlier and 
later settlements did not overlap. Several of the other cuts show 
considerable separation between successive levels, but this seems partly 
a function of the large number of contemporary sites, whose inter- 
digitation pushes apart what would otherwise be seen as a continuous 
series of Jevels. This appears to be true of H-8, Cut 1, since the two 
levels involve no greater amount of ceramic change than those of 
E-3, Cut 2, although the presence of a considerable number of con- 
temporary sites and levels spaces them much farther apart (fig. 101). 
The situation at E-30, Cut 1, and E-1, Cut 2, however, is much too 
extreme to allow for any such explanation, and the reoccupation theory 
seems to be the best answer. 
The existence of a general dating based on historical records for 
both the Taruma and the Wai Wai occupations of the upper Esse- 
quibo region makes it worth while to reexamine our attempts to arrive 
at a formula for using the rate of pottery refuse accumulation to 
estimate village and phase duration in Tropical Forest sites (see 
Meggers and Evans, 1957, pp. 245-257). As applied to sites in the 
region around the mouth of the Amazon the formula was: 2,600 sherds 
per 1.5- by 1.5-meter cut=100 years. Since all the stratigraphic ex- 
cavations in the upper Essequibo area were 1 by 1 meter, the equation 
was reduced to conform with this dimension, giving: 1,156 sherds per 
1- by 1-meter cut=100 years. Durations were calculated on this basis 
for all the Taruma Phase sites with stratigraphic excavations and 
controlled tests, and these are presented on table I, arranged in 
descending order. The probable number of occupations represented 
at: each site is also shown. Comparison of the estimated total years 
of occupation represented by each stratigraphic cut or test (table TI) 
with the number of occupations indicated by the separation of the 
levels shows that, with a single exception of 66.1 years, all cuts and 
tests with only one occupation have durations of less than 50 years. 
Eight cuts from various sites with two or three occupations have total 
estimated durations ranging from 24.8 to 128.8 years. 
