538 PROCEEDINGS OF THE NATIONAL MUSEUM 



explained as intrusions resulting from later Mantefio occupation 

 of the area around the Gulf of Guayaquil. 



Unclassified Decorated 



Most of the decorated sherds left unclassified were too small or too 

 badly eroded to identify. The remainder consisted principally of 

 applique nubbins. 



Applique Nubbins. — One rim of Form 9 from Site 0-5 has two 

 rounded pellets attached side by side under the everted rim. Traces 

 of red wash are visible on the upper surface of the rim, which has a 

 diameter of 14 cm. Another rim from the same provenience has a 

 nubbin with a punctate in the center in the same position on the 

 rim exterior. 



Several sherds from tiie surface of Site 0-3 are ornamented with a 

 horizontal row of low applique nubbins 2-5 mm. high, applied either 

 to the surface or to a low ridge. 



Implications of the Seriated Sequence 



The seriated sequence of the Jambeli culture is based on trends in 

 change of frequency in the plain types derived from analysis of the 

 sherds from stratigraphic excavations at sites G-84, G-86 and 0-5. 

 Into this framework, surface collections from 10 additional sites were 

 seriated. Finally, the attempt was made to include sites represented 

 by a sample of less than 100 sherds, and 7 of these fit the pattern of 

 change sufficiently well to be incorporated into the graph (fig. 38). 



The two principal plain types exhibit a rather smooth pattern of 

 change, in which Jambeli Plain, tempered with coarse micaceous 

 sand, increases in popularity while Ayalan Plain, tempered with fine 

 micaceous sand, shows a corresponding decline in frequency. Some 

 of the samples exhibiting a larger amount of both types than is 

 consistent with the pattern of the curve reflect an unusually high 

 degree of surface erosion. This loss of surface results in the classi- 

 fication of some sherds as plain that might originally have had a 

 slipped or painted surface. Sites P-2 and G-L-3 are examples of 

 this kind of distortion. 



The only real time marker, in terms of presence or absence, is 

 Posorja Polished Plain, which is restricted to the early half of the 

 sequence. It perhaps represents a survival of the polished plain 

 ware tradition characteristic of the preceding Formative Period. 



None of the decorated types exhibits any consistent pattern of 

 change. Each type maintains a relatively uniform frequency from the 

 earliest to the latest site, with fluctuations that can be attributed 

 either to differential amounts of surface erosion or to errors resulting 

 from the small size of the sherd sample. Attempts were made to 



