b BUREAU OF AMERICAN ETHNOLOGY [Bull. 193 



vague affiliations with the types as established. Thus, the Ortiga 

 variety, although placed with the Parita type, shares various elements 

 with the Macaracas type, and the Calabaza and Cerit6 varieties, 

 which were grouped as a separate Calabaza type, may prove to be 

 sufficiently related to one of the three major types at the site to be 

 combined with it. Others, for example the Higo and the Pica-pica, 

 possibly should be combined since their shapes coincide and their 

 design elements are so frequently shared. However, such difficulties 

 are inherent in any system of classification which deals with as many 

 possible variables and subtleties of impression as are included in the 

 Azuero or Code styles. It is worth noting, as well, that the typology 

 used in this report was necessarily established without benefit of ade- 

 quate stratigraphic checks. These varieties, then, are my impres- 

 sions of ceramic units, or clusters of attributes, which may have 

 chronological and areal significance. The classifications should be 

 used with the obvious caveat that since they are based on a fairly 

 limited sample of total aboriginal manufactm-e at a particular time and 

 place, they are subject to revision in the light of subsequent knowledge. 



In the same way that no one category of attributes was chosen as 

 decisive in distinguishing varieties from each other, I followed no 

 explicit rule in grouping varieties into types. In a sense, the types 

 obtruded themselves from the material at hand; in many cases the 

 larger grouping was apparent first, and varieties within it were dis- 

 tinguished later. The El Hatillo type is based primarily on a similarity 

 of design style and, to some extent, shapes which appeared to me to be 

 readily distinguishable from other vessels; the varieties within it 

 were separated out only after further examination. Although some 

 attributes are obviously shared between types, the final decision as to 

 whether a variety belonged in type A or type B was made on the basis 

 of the subjective impression that it was more similar to the varieties 

 of one type than to those of another. In some cases (e.g., the Calabaza 

 variety) this was difficult to determine and the variety, although in 

 this case placed within a type, is left with a vague status. 



The ceramic group as used here is simply a more inclusive grouping 

 at a higher level of generalization ; a device which is handy, for example, 

 in discussing the differences or similarities between the pottery 

 complexes of the Late Period Code graves at Sitio Conte and the 

 "classic" Chiriqui grave complex, or that found at He-4 and other 

 sites on the base of the Azuero Peninsula. It includes both character- 

 istic polychrome and other ceramics, bichromes, plain and plastic 

 decorated wares. As distinct from the pottery or ceramic "complex," 

 it presumably includes only pottery of local manufacture that is 

 characteristic of the area and time and excludes trade sherds or vessels. 



