Archaeologia Islandica - 01.01.2007, Qupperneq 67

Ancient ironmaking in Iceland, Greenland and Newfoundland
Fe FeO Fe203 SÍ02 MnO AI203 P205 Sum* R
1 Slag V (54.5) 70.1 18.6 1.75 4.48 0.35 99.0 3,22
2 Slag L (49.6) 63.8 21.85 1.4 9.6 0.24 101 2.49
3 Slag B1 (48.8) 62.7 24.9 2.0 5.9 0.35 100.1 2.15
4 Slag B2 (49.6) 66.8 22.25 2.03 5.01 0.37 ? 2.58
5 Ore B (59.4) 84.82 9.89 0.56 2.90 0.133 99.13 6.63
E.v.B1/Ore B 2.52 3.57 2.03
E.v.B2/Ore B 2.25 3.62 2.46
‘The sums include minor amounts of CaO, MgO, BaO, Ti02 and K20.
Table 6. Analyses of samples from Viðivellir (V), Lundur (L) and Belgsá (B). E.v. enrichment value
(% in slag/%> in ore).
the waste product slag, while the ore
found at Lundur contained even less iron
oxide. These values are omitted.
During smelting the oxides of
iron and phosphorus are partially reduced.
The other oxides MnO, Si02 and A1203
are only transferred from the ore to liq-
uid slag. As metallic iron and also some
oxygen (as the gases CO and C02) are
removed, the percentage of such oxides
must increase from ore to slag. If the raw
material is representative, the ratio (%
X in the slag/ %X in the ore) therefore
should be identical for these three oxides,
normally present in large amounts. Minor
corrections for wood ash and dissolved
clay are not considered here. This ratio is
given the name “enrichment value” (E.v.
in the table).
After the ochre was excluded
as a raw material, the author returned to
Belgsá in the year 2001 with the hope that
a “valid” type of ore could be found. A
close examination of the bare slag heap
revealed some brownish small lumps
measuring 2-5 mm, reminiscent of raw
sugar. In the hope that these pieces might
represent ore lost by the smelters, about
5 g of this material was taken to Norway
for analysis, with results presented in the
table above as Ore B. This time the analy-
sis shows a good ore, with R = 4.69. The
“enrichment values” presented are also
much more coherent than those related to
fresh ochre. Only MnO, present in small
amounts, diverges from values between 2
and 2.7. This shows that the small pieces
seem to represent the ore used and there-
fore can give a sound basis for an evalua-
tion of the iron production (see later).
The small pieces were non-mag-
netic. As shown in figure 6, the X-ray
diffraction diagram shows only the pres-
ence of the mineral goetite FeOOH. The
lack of diffraction lines for Si02 can only
be explained by a finely divided, X-ray-
amorphous substance. The origin might
be diatoms, which seem to thrive in fresh
precipitates such as ochre. E.g. the dia-
tomaceous earth created in lake Myvatn
some 60 km from Fnjóskadalur and sold
as an excellent high-temperature isola-
tion material consists of such algae. Inci-
dentally, the character of this ore is so to
speak identical to ore found in Modum,
Norway. In most Norwegian bog ores
particles of sand by XRD overshadow the
possible presence of other types of silica.
The presence of iron as a hydroxide seems
to prove that the Icelanders did not roast
their ore prior to smelting, which is in
agreement with the opinion of those who
65