Paper No. 8
Presentation Time: 10:45 AM


ROTH, Karen A.1, STERNBERG, Robert S.2, SHACKLEY, M. Steven3, FEINBERG, Joshua4 and RAHL, Jeffrey M.1, (1)Department of Geology, Washington and Lee University, Lexington, VA 24450, (2)Earth and Environment, Franklin & Marshall College, Lancaster, PA 17604, (3)Anthropology, University of California, Berkeley, 232 Kroeber Hall, University of California, Berkeley, CA 94720-3710, (4)Institute for Rock Magnetism, University of Minnesota, Department of Earth Sciences, 310 Pillsbury Drive SE, Minneapolis, MN 55455,

Mule Creek and Mt. Taylor, located in southwestern and western New Mexico respectively, are obsidian source regions for archaeological lithic artifacts. The volcanism that produced the obsidian-bearing rhyolite in these regions dates to 17 Ma for Mule Creek and to 3.2 – 3.5 Ma for Mt. Taylor. Within each region, we sampled non-artifact obsidian at three geographically-distinct sources, each about 0.25 km2 in area, for the purpose of geochemical characterization. Obsidian is exposed as black marekanite nodules ranging from 2 – 4 cm in diameter at Mule Creek sources and 3 – 12 cm in diameter at Mt. Taylor sources. Marekanite nodules occur in situ in hydrated perlitized outcrop and as float weathering out from the ground surface. Obsidian specimens at all sources often bear percussion marks from ancient knappers testing the obsidian quality. At some sources the marekanites contain sanidine phenocrysts, 2 – 5 mm long; at others the marekanites are aphyric. Prior laboratory-based X-ray fluorescence work has characterized the geochemistry of obsidian from these regions, with the aim of establishing their inter-source geochemical variability and facilitating the sourcing of archaeological artifacts. We build upon this work by employing portable XRF to characterize the geochemistry of a much larger population of samples than has previously been analyzed, using spatially referenced samples. This permits us to examine inter-region, inter-source, and intra-source variability, all with a large enough group of samples to provide statistically robust results.