JAMBELI CULTURE — ESTRADA, MEGGERS, EVANS 539 



detect temporal differences in the decorative motifs, but no distinctions 

 could be found that could not be attributed to deficiencies in the size 

 of the sample. 



Analysis of the rim forms and reconstructed vessel shapes showed 

 the same picture of cultural uniformity through time (fig. 39). Only 

 two forms have a restricted distribution: Form 5, a bowl with an 

 everted rim, and Form 12, a large compotera. Both are limited to 

 the early half of the sequence. Annular bases are present throughout, 

 and this is probably also true of hollow legs since their absence from 

 the earliest and latest levels can be accounted for by the small size of 

 the sherd sample. The only solid leg comes from an early site. 

 Spouts, although very rare, appear to occur throughout the sequence. 



Discussion of the temporal differences in the frequency of various 

 types of artifacts is hampered by the fact that these are relatively 

 rare in comparison to sherds. The majority of those artifacts included 

 in this report are from Site 0-6, where the constant erosion of the site 

 by the sea, and the existence of a resident population able to salvage 

 objects as they were revealed, provides a combination of collecting 

 circumstances not met with at any other Jambeli site. In view of the 

 relatively small amount of excavation undertaken, it consequently is 

 difficult to determine whether the absence of figurine fragments from 

 the lower third of the sequence reflects a real absence or the inadequacy 

 of collections from early sites. The two bark beater fragments are 

 from sites that occupy an early and a late position, suggesting that 

 this trait was present throughout the sequence. 



Of considerable interest is the question of whether the entire 

 area in which the Jambeli culture remains are found was occupied at 

 the same time, or whether a movement from north to south or south to 

 north can be inferred. The seriated position of the sites does not 

 provide a clear-cut conclusion, but the absence of any El Oro Province 

 sites in the lower third of the seriated sequence may reflect a more 

 recent movement into that part of the area from the north. In the 

 upper third of the sequence, the situation is reversed. The largest 

 sites are in El Oro or southeastern Puna Island, and those in the 

 Playas-Posorja region are typically small with sparse pottery refuse. 

 If this interpretation of a shift in population concentration from 

 north to south is correct, it may be correlated with the gradual 

 desiccation of the northern portion of the area, which caused the 

 extinction of the mangrove swamp on which the subsistence of the 

 Jambeli Phase was largely dependent. 



In summary, it can be said that all the evidence derived from an 

 examination of the chronological distribution of ceramic traits, 

 artifacts, and other cultural features of the Jambeli culture presents a 

 picture of unifonnity and consistency. There is no indication that 



