Oregon samples were represented by 6,644 items from 83 archaeological sites in the eight counties intersected by the Pipeline. Sample sizes ranged by county from four artifacts in Umatilla County to 2,602 items in Deschutes County. Forty eight sites from eight California counties were also selected for characterization studies and a total of 2,864 items were submitted for source identification. Two items of obsidian debitage from 10-BY-444, the only two Idaho artifacts of adequate size for XRF analysis, were also included in the trace element studies.

This combined corpus of 9,510 specimens still constitutes, to our knowledge, the largest sample of characterized obsidian artifacts associated with any single archaeological project in the world to date.

Because of this, we now feel that the time is right for a reevalulation of the original Oregon trace element data that resulted from the PEP obsidian studies. Although most of the original source assignments were valid, new source information should permit us to identify the geologic sources of many of the artifacts that were assigned to unknown sources. In addition, recent extensive trace element investigations of obsidian sources at Glass Buttes and the Ochoco and Malheur national forests should allow us to clear up many provisional source assignments that resulted from our incomplete knowledge of these areas prior to the time of initial project completion. These provisional assignments were largely associated with the Whitewater Ridge, Little Bear Creek, and Juniper Spring sources. We now know that the Juniper Spring sources originated from the nearby Glass Buttes Source Complex and that the Whitewater Ridge and Little Bear Creek sources belong to a single highly variable geochemical source group. After several years of searching, however, we have still failed to locate the so-called Unknown X source that probably lies somewhere on the lower western flanks of Newberry Volcano (now the Newberry National Volcanic Monument).