Archaeologia Islandica - 01.01.2009, Blaðsíða 50

Richard E. Hughes and Gavin M. Lucas
140
120
100
£ 3 80 ■
cn 60-
40-
20-
k|h Katla
Krafla
\ / \ i \ \ / /
Kerlingarfjöll Hágöngur
Torfajökull
z'* '’v
í } \
20 40 60 80 100 120 140 160 180 200
Y (ppm)
Figure 4. Sr vs. Y Composition of Obsidian Artifacts from Hofstaðir. Dashed lines demarcate the
range of composition measured in geological samples. Open circles represent measurement for
single obsidian source standards. Filled triangles represent measurements for Hofstaðir artifacts
from Table 2. Error bars are 2-sigma (95% confidence) error estimates for artifacts and single
obsidian source standards. The numbers of artifact plots do not correspond exactly to the number
of specimens in Table 2 because of data convergence at this scale.
excitation of the elements selected for
analysis. In this case analyses were con-
ducted for rubidium (Rb), strontium (Sr),
yttrium (Y), zirconium (Zr), niobium
(Nb) and barium (Ba) using the Ka emis-
sion line for each element. All artifacts
also were analyzed to determine iron vs.
manganese (Fe/Mn) ratios. The analyses
were conducted at 120-600 deadtime-
corrected seconds, with tube current
scaled to the physical size of the speci-
men. Other details involving laboratory
analysis protocol appear elsewhere
(Hughes & Pavesic, 2005); the interested
reader should consult this paper, and
Hughes (1988, 1994,2005, pp. 249-250),
for additional information on calibration
and element-specific measurement reso-
lution.
Using these same instrument condi-
tions, analyses were first conducted on a
series of geological obsidian samples
from fíve areas in Iceland to serve as a
comparative baseline for artifact analysis
(see Table 3; Figure 4). Table 3 com-
pares trace element data published by
48