E d r 
pyane an ARCHEOLOGY IN BRITISH GUIANA 321 
Types of ornamentation that seem indigenous to the Rupununi 
Phase, but occur only occasionally, are conical applique nubbins on 
large jars (pl. 65, 0) and nicks or finger impressions on the rim edge 
(pl. 63, 7-4). Sherds with a thin red film on one surface occur 
sporadically but are too frequent to have temporal significance. 
Analysis of the vessel shapes of Kanuku Plain and Rupununi Plain 
shows that the same forms are represented in both wares (Appendix, 
tables 88-39). However, a considerable difference in relative popu- 
larity is discernible. In Kanuku Plain, Form 1 is dominant, being 
almost twice as common as the two next most frequent shapes, Form 
2and Form 4. In Rupununi Piain, by contrast, Forms 1, 2, and 3 are 
all equally represented and Form 4 is relatively rare. Form 6 is rare 
in both wares. ‘The’ total number of rims from each site is too small 
for reliable percentage analysis, but several tentative conclusions can 
be drawn from the range of occurrence and differential frequency of 
forms. It seems significant, for example, that Form 4 (griddles) is 
most abundant in Kanuku Plain; which is the earlier type. This 
vessel, indispensable to the baking of cassava bread, ceases to be made 
in quantity in the late part of the Rupununi Phase sequence although 
the type of food with which it is associated is still a major item of 
diet today among Rupununi savanna Indians. The dying out of this 
form is undoubtedly correlated with the introduction by Europeans of 
a metal substitute, and therefore reflects increasing Kuropean contact 
and consequent greater ease of obtaining the more durable type of 
griddle. 
Of the three kinds of base found on Rupununi Plain and Kanuku 
Plain vessels, the flat vase (Form A) is by far the most common 
throughout the sequence. Form B (pedestal) is next and appears to 
be largely restricted to the early half ofthe Phase. Form C (rounded) 
is rare and sporadic (Appendix, tables 38-39). 
Most types of stone tools are too infrequent to permit the drawing 
of any conclusions about changes in popularity. However, it may 
be noted that hoes occur only in the latter half of the sequence and 
only at sites in the northern savanna. They are sufficiently abundant 
here to suggest that they ought to be represented at earlier sites if 
they had been in use. Possibly these hoes are stone copies of Euro- 
pean meta] hoes received by trade and considered more efficient than 
the aboriginal digging stick. Axes, by contrast, are found throughout 
the Phase, except at the latest sites where they can be presumed to 
have been superseded by metal counterparts. 
The 'seriated site sequence (fig. 125) derived foun ceramic fre- 
quencies has an interesting geographical aspect. All of the sites in 
the lower half of the sequence are in the southern savanna except R-2 
and R-8, which are on the north savanna near its southern boundary. 
