Source provenience data for artifacts from Willamette Valley sites point to predominant local procurement of archaeological obsidian from local secondary sources of glass found in the Valley. The Obsidian Cliffs and Inman Creek geochemical source groups are represented by the great majority of characterized artifacts. Other minor western Oregon sources (e.g., Devil Point and Clackamas River) are occasionally identified but the use of sources outside of western Oregon is very rare. Willamette Valley tools are typically rather small, often no longer than a few centimeters, and debitage frequently retains surface cortex. In most cases, this surface cortex shows evidence of mechanical abrasion (small surface pits). Small split rounded pebbles of raw material are also often found and it is likely that the majority of artifacts produced in the Willamette Valley were manufactured from small nodules of obsidian found in river gravels and in alluvial deposits. Small nodules of glass correlated with the Obsidian Cliffs and Inman Creek sources are found today at many locations along the Willamette River between Eugene and Oregon City.

On the other hand, characterized artifacts from Western Oregon sites located to the east and northeast of the study area usually originate from a much wider variety of geologic sources, some of them situated in central Oregon. While Obsidian Cliffs is often the major source identified at sites in the in Mt. Hood and Willamette national forests, other western Oregon sources are commonly present (Devil Point, Inman Creek, Clackamas River), as are locally available sources (Jordan Creek, Rhododendron Ridge, Lemiti Creek, Carver Flow). Glass from a few central Oregon sources, most often those located in the Newberry Volcano region, are also occasionally identified in small proportions. Cortex associated with artifacts from Western Cascades sites is most often fresh and angular, evidence that the material has been collected at or near their primary sources.

The objectives of the research presented here are straightforward:

1. Using XRF trace element analysis methods, to determine the geologic sources of artifacts found at the Orr Site.
2. To provide obsidian provenience and procurement information for an area in which no extant characterization data are currently available.
3. To further demonstrate the feasibility and importance of trace element methods in recovering information from severely disturbed archaeological sites.

The primary source location for one of these source groups, the Butte Creek source, remains unknown at the present time. In the current investigation, 30 of the 47 artifacts correlated with the Butte Creek source retained at least a trace of original surface cortex. The cortex on all of these specimens was angular and mechanically unmodified by transport processes, suggesting procurement at or very near the primary source of the material. This evidence, when combined with the large proportion of material from the Butte Creek source found at the Orr Site, suggests that the parent source lies within a relatively short distance of the site.

A comparison of the trace element composition of the new Butte Creek source with previously analyzed western Oregon artifacts indicates that only a single specimen (originally identified as an unknown source) from Site 35-MA-141 in the eastern Willamette Valley has been found in previous investigations.

An Additional Note: Since the completion of the Orr Site artifacts study, the Butte Creek source has also been identified among characterized artifacts from 35-MA-183 (N=3) near Salem and at 35-CL-293 (N=1) in Estacada.