IN NORTHERN HILLSBOROUGH COUNTY, FLORIDA.

ABSTRACT: During the summer of 1981, the University of South Florida conducted Phase III mitigative excavations at three sites located within the boundaries of the Tampa Palms Development Area in northern Hillsborough County. This thesis presents the results of archaeological investigations at two of these sites, the Tampa Palms site (8-Hi-557) and the Priscilla site (8-Hi-559).

The primary objectives of the project were

1) the determination of site function at the Tampa Palms and Priscilla sites;

2) the delineation of site boundaries at the Priscilla site;

3) the identification of site activities and the isolation of specific activity areas at the Tampa Palms site and, if possible, the Priscilla site.

An activity set model was formulated and tested at each of the sites. The results indicate that the Tampa Palms site was utilized on a relatively continuous basis throughout the Archaic stage. Site function varied as a result of changing hunter/gatherer subsistence strategies. These changes were in response to concomitant changes in climate and in the location and availability of exploitable resources. A shift from primarily a foraging existence to one of central based collecting is suggested.

The Priscilla site was occupied primarily during the Late Archaic and functioned as an extractive camp for specialized task groups exploiting locally available resources. These groups operated within a logistically organized system of central based collecting.

An additional goal of the project was to provide data pertinent to the problem of functionally identifying the large number of sites in the region which have been subsumed under the ambiguous label "lithic scatters." To this end, a debitage attribute analysis was conducted on a sample of flakes recovered from the third site excavated during the summer, the Rock Hammock site (8-Hi-556); a spatially and temporally isolated area of lithic reduction activity.

The attribute analysis was conducted in an effort to isolate specific flake types representative of different stone working activities. A model of stone tool reduction was formulated which predicts the co-occurrence of certain flake attributes within each level of the flaking sequence. A multivariate analysis of the Rock Hammock sample was conducted utilizing these attributes. The results revealed the presence of ten flake types at the site. These are interpreted as representing "micro-levels" of flaking activity within the general reduction continuum. The analysis suggests that stone working activities at this and other sites can be isolated with a greater degree of specificity than has previously been attempted.