A CHEMICAL AND PETROGRAPHIC STUDY OF THE ORIGINS OF PREHISTORIC POTTERY IN THE NORTH CAROLINA SANDHILLS

Archaeologists have long suspected that much of the prehistoric Native American pottery found on sites in the North Carolina Sandhills was not made from local resources. Limited replication tests have shown that the plasticity of some Sandhills clays is inadequate for making pottery. This paper reports on the results of instrumental neutron activation and optical mineralogical analyses designed as a first step in testing this hypothesis. Thirty archeological potsherds from three sites (10 samples each) in three different river basins (two Piedmont and one Coastal Plain) outside of the Sandhills are compared to 20 pottery samples from Sandhills sites. About 77% of the 50 potsherds comprise five chemical-composition groups that sort into two broad categories describing high-calcium/sodium sources in the Piedmont and low-calcium/sodium sources in the Coastal Plain. Petrographic analysis further indicates the source of the calcium-sodium difference is igneous (diabase and granitic) rock fragments included as temper in the ceramic matrix. Potsherds from Sandhills sites reflect a mixture of these types suggesting that pottery was transported into the Sandhills from source areas in the Piedmont and Coastal Plain.