Archaeologia Islandica - 01.01.2007, Blaðsíða 59
Ancient ironmaking in Iceland, Greenland and Newfoundland
their approach becomes descriptive. The
prescriptive aspect of process metallurgy,
expressed by this author, is largely miss-
ing (e.g. Espelund 2004b).
Out of 51 iron artefacts only 5
contained metallic iron and were studied
by SEM analysis. Carbon and sulphur
were determined by combustion and
absorption. The two nails were as expect-
ed high in phosphorus - low in carbon,
while the rivet contained 0.12 and a steel
sheet fragment 0.36 % C, both low in %
P. No less than 15 trace elements were
searched for, giving values around 0.01%,
which can be regarded as the lower detec-
tion limit and therefore disregarded. Only
two values for molybdenum 0.061 (nail)
(together with 0.60% P) and 0.13 (rivet)
(together with 0.076 P) might indicate a
special source of ore (:65). The values
for the elements Mn and Si were copied
from Rosenqvist (1977/85). The lump of
cast iron measuring 2.5x2.2x2 cm (Ung-
lik and Stewart 1999:20) can hardly be an
intended product as no such production
took place before the end of the Middle
Ages (1300 in Sweden, 1600 in Norway,
never in Denmark, Iceland or Greenland).
It is more likely that it represents unsuc-
cessful smelting at L’Anse aux Meadows,
without proper slag control or, alterna-
tively, a newer and foreign sample. The
value for Si (0.73%) can represent strong
reducing conditions and/or entrained
slag. Pieces of cast iron have also been
encountered by the present author in
slag heaps at the Norwegian Roman Age
smelting site Myggvollen and at an Even-
stad type of furnace in Kása, community
ofTolga. The authors also conclude that
the cast iron was unintentional (Unglik
and Stewart 1999:116).
This author selected slag No.
2 (4A505B1-1) for a closer study. If the
values for the two iron oxides are added
as FeO, the value for the “fayalite ratio”
R becomes 2.35, which is a normal figure
after extraction of iron. The sum for this
sample amounts to 97.4%. An interesting
observation from these tables is the com-
plete absence of molybdenum in the ore
and in slag pieces, indicating that the nail
and the rivet mentioned above appear to
be of foreign origin. The two nails con-
taining 0.60 and 0.56% P can hardly
be the result of smelting at L’Anse aux
Meadows, as two of the ores presented
are so to speak phosphorus-free. Phos-
phorus ensured stiffness, useful for a nail,
to be driven into wood when a new ship
was built, and not to be exposed to any
impact later. This is a good expression of
professional shipbuilding with access to
different materials, characteristic for con-
temporary Norway.
Among the trace elements the
values for molybdenum 0.061% (large
nail) and 0.13% (large rivet) deserve
special attention. As Mo is fully unex-
pected, it has so far not been searched
for in my own analyses of Norwegian
ores and slags. Geochemical mapping
of Scandinavia might reveal the source.
According to the Norwegian Geologi-
cal Survey there are two areas in South
Norway where this element occurs in
significant amounts (Korneliussen pers.
comm.): Knaben in Setesdal, and Nor-
dli near Eidsvoll. Bloomery ironmaking
during the Viking Age - Medieval period
cannot be excluded in the two areas. This
information might reveal the provenance
of the iron for the rivets. An analysis by
means of microprobe of a sample from
a bloom from Skeibrok, community of
Lista, County of Vest-Agder UO C 22 615
at the south coast of Norway has shown
similar values for molybdenum. By sheer
57