Analysis of Alaskan Archeological Obsidian Artifacts Presented at the Archaeometry Research Graduate Group Annual Symposium February 1994, University at Buffalo Anthropology Department Henry J. Chaya 146 North Walnut Street Beacon, New York 12508 Submitted March 29, 1994 Note: Figures 1 to 3 are graphics files available separately. Background A number of obsidian artifacts were analyzed by X-Ray Fluoresence from Archeological Site XSI-040 on Chernabura sland in the Schumagin Islands west of Kodiak Island in Alaska. Radiocarbon dates extend to 1500 B.P. or approximately 500 A.D. Experimental The obsidian artifacts were analyzed in terms of relative compositions of trace elements iron, rubidium, strontium, yttrium, zirconium, and barium. We did not have any geological source samples of obsidian to compare the artifact obsidians with. The X-Ray Fluoresence setup consisted of an Americium- 241 source with a tin target to irradiate the samples. The time of exposure for each obsidian sample was one hour. Elemental peaks appear as shown in Fig. 1. An AXIL software program was used to measure areas under the elemental peaks above the baseline. The program automatically subtracts the background below the baseline. Barium is outside the spectral graph in Fig. 1 and is due to unshielded Americium- 241. The areas of the elemental peaks in terms of counts are shown in Table 1 for 16 obsidian artifacts analyzed. The Yellowstone obsidian sample shown in the table is of known composition in terms of the trace elements being analyzed. Table 2 shows the elemental peak areas summed up for each sample and normalized to 100% and the relative percent of each element is its elemental area over the total sum. Fig. 2 shows a bar graph representing the relative amount of each element in a sample, and it can be seen that the bar graph of each obsidian sample is almost similar within the precision of the analytical procedure. Fig. 3, shows a plot of percent of zirconium and percent yttrium and cluster formation of the samples in terms of these elements. Fig. 4 shows the same thing in terms of the elements strontium and rubidium. The clustering of the obsidan samples in a graphical way shows the variability or precision of the analytical procedure. In order to obtain absolute concentrations of the elements in parts per million, one has to compare the spectrum of the obsidian artifact with that of a reference obsidian sample of known composition. These comparison samples must be of the same size and analyzed under exactly the same instrument conditions. Conclusions It may be concluded that the obsidian artifacts analyzed for this archeological site are similar in composition and are all from the same geological source. Further work may be the comparison of obsidians from various geological sources with these artifacts to determine their geographical source. Acknowledgements Appreciation and thanks are expressed to Professor Willian Lanford and students Badle Assallami, Armin Knoll and Janice E. Lau of the Physics Department, State University of New York at Albany for obtaining the data used in this work. Appreciation and thanks are also extended to Professor Lucy Johnson of the Anthropology Department of Vassar College for providing the obsidian artifacts. Table 1: Name of Sample Peak area / Counts Fe Rb Sr Y Zr Ba #1 19/20d * 24867 3087 3159 3963 28637 5317 #2 B12 D11 E2 19090 2177 2519 2947 20787 3262 #3 B12 P13 D3 22016 2729 2817 3478 23950 2404 #4 (S)12 P20 E1 21974 2665 2810 3428 23794 3648 #5 B12 Dl5C 2 18650 2500 2246 2661 20428 2119 #6 B12 P13D 2 16222 1892 2128 2242 16971 1917 #7 B12 P15 Dl 15309 1665 1586 1788 14902 1338 #8 B12 P13 C1 20992 2632 2731 3436 22946 2443 #9 B12 P12 ** 17491 2214 2066 2435 17565 1509 #10 B12 P23 D3 16284 1711 1808 1920 14875 1102 #11 B12 P10 D1 16583 1884 2129 2371 17214 1497 #12 B12 P18 D1 19578 2387 2328 2556 20582 1987 #13 B12 P20 *** 15253 1479 1953 1961 14176 1201 #14 B#13 21054 2758 2777 3227 22751 2437 #15 B#8 22323 2678 2727 3125 23754 2886 #16 B12 P23 C 21613 2843 2815 3467 24723 3164 Averages 19331.19 2331.31 2412.44 2837.81 20503.44 2389.44 Std dev's 2921.25 483.91 444.32 659.16 4190.20 1094.63 Normalisation Sample: Yellowstone 12946 9126 67 4704 12442 45 * from firepit at base of layer ** layer above floor *** O/F contact Table2: Name of Sample Normalised amount of elements Fe Rb Sr Y Zr Ba #1 19/20d * 36.02% 4.47% 4.58% 5.74% 41.48% 7.70% #2 B12 D11 E2 37.59% 4.29% 4.96% 5.80% 40.93% 6.42% #3 B12 P13 D3 38.36% 4.75% 4.91% 6.06% 41.73% 4.19% #4 (S) 12 P20 E1 37.68% 4.57% 4.82% 5.88% 40.80% 6.26% #5 B12 DISC 2 38.37% 5.14% 4.62% 5.47% 42.03% 4.36% #6 B12 P13D 2 39.21% 4.57% 5.14% 5.42% 41.02% 4.63% #7 B12 P15 D1 41.84% 4.55% 4.33% 4.89% 40.73% 3.66% #8 B12 P13 C1 38.04% 4.77% 4.95% 6.23% 41.58% 4.43% #9 R12 P12 ** 40.41% 5.12% 4.77% 5.63% 40.58% 3.49% #10 B12 P23 D3 3.19% 4.54% 4.80% 5.09% 39.46% 2.92% #11 B12 P10 D1 39.79% 4.52% 5.11% 5.69% 41.30% 3.59% #12 B12 P18 D1 39.30% 4.79% 4.67% 5.93% 41.31% 3.99% #13 B12 P20 *** 42.34% 4.11% 5.42% 5.44% 39.35% 3.33% #14 B# 13 38.28% 5.01% 5.05% 5.87% 41.36% 4.43% #15 B# 8 38.83% 4.66% 4.74% 5.44% 41.32% 5.02% #16 B12 P23 C 36.87% 4.85% 4.80% 5.91% 42.17% 5.40% Averages 39.13% 4.67% 4.85% 5.66% 41.07% 4.61% Std dev's 1.98% 0.28% 0.26% 0.35% 0.79% 1.28% Normalisation Sample: Yellowstone 32.92% 23.20% 0.17% 11.96% 31.63% 0.11% * firepit at base of layer ** layer above floor *** O/F contact