Archaeologia Islandica - 01.01.2009, Page 51

Geochemical Identification of Obsidian Artifacts
Trace Element Concentrations Ratio
Cat. Obsidian Source
Number Rb Sr Y Zr Nb Ba li Mn ES2Ö3^ Fe/Mn (Chemical TvDel
97-28 63 97 99 512 55 593 nm nm nm 43 Krafla
±4 ±3 ±3 ±4 ±3 ±15
98-111 66 101 102 522 57 626 nm nm nm 43 Krafla
±4 ±3 ±3 ±4 ±3 ±15
99-404 64 97 100 504 55 626 nm nm nm 42 Krafla
±4 ±3 ±3 ±4 ±3 ±15
00-102 66 101 102 522 58 593 nm nm nm 43 Krafla
±4 ±3 ±3 ±4 ±3 ±14
U.S. Geological Survey Reference Standard
RGM-1 152 109 25 220 10 812 nm nm nm 64 Glass Mtn., CA
(measured) ±4 ±3 ±3 ±4 ±3 ±14
RGM-1 149 108 25 219 9 807 1600 279 1.86 nr Glass Mtn., CA
(recommended)
Values in parts per million (ppm) except total iron [in weight %] and Fe/Mn intensity ratios; ± =
expression of x-ray counting uncertainty and regression fitting error at 120-360 seconds livetime.
nm= not measured; nr = not reported. Recommended values for RGM-1 from Govindaraiu
(1994).
Table 4. Quantitative Composition Estimates for Obsidian Artifacts from Hofstaðir
Jónasson (1994, 2007) with data gener-
ated here on the same samples. The
measurements are in very good overall
agreement, engendering confídence that
the non-destructive quantitative compo-
sition estimates generated here can be
compared directly with published litera-
ture (see also Macdonald et al. 1992:
Appendix I, pp. 158-160).
With these comparative data in hand,
trace element analyses were conducted
on the four obsidian artifacts recovered
from Hofstaðir (see Table 2 for proven-
ience information). Table 4 presents trace
element data for each of these four arti-
facts. Although other trace elements
could be used to illustrate the differences
between obsidian chemical types, Sr vs.
Y (Figure 4) and Rb vs. Nb (Figure 5)
contrasts were employed here. The
dashed lines in both fígures represent the
range in variation measured in geologic
source samples, while the filled triangles
are plots for obsidian artifacts from
Hofstaðir. Trace element values for
Torfajökull obsidian samples show a
slightly greater range than other obsidi-
ans (perhaps due to magma mixing [Blake
1984; McGarvie et al. 1990]) but its
overall elemental “fíngerprint” is none-
49