Archaeologia Islandica - 01.01.2007, Blaðsíða 57
Ancient ironmaking in Iceland, Greenland and Newfoundland
MnO FeO Fe203 Si02 CaO Na20 K20 P205 Sum R
wt% 0.68 15.8 79.5 1.9 0.68 0.12 0.14 0.38 99.2 42.9
Table 2. Analysis ofbog iron ore from L’Anse aux Meadows by Rosenqvist (1985:414).
Fe203 MnO Si02 AI203 CaO MgO BaO K20 Ti02 P205 Sum R
1 32.18 31.93 22.62 4.97 0.81 0.66 0.17 0.86 0.22 0.12 94.5 2.3
2 89.50 5.33 1.24 2.45 0.47 0.05 0.01 0.04 0.01 0.08 99.18 58.0
Table 3. Analysis of bog iron ore from L’Anse aux Meadows by the author.
the Christian era, when deceased persons
had to be buried in consecrated cemeter-
ies. The nearest church was in Greenland.
Both Iceland and Greenland were bisho-
prics. The general picture therefore is
that of a temporary settlement. The dat-
ing, mainly 14C on samples of charcoal,
wood, turf/peat and whalebone indicate
that the place was inhabited during the
winter over a period of some 10 years
around year 1000 (Wallace 2006:73). In
full agreement with this assessment the
middens found were very small.
Ironmaking at L’Anse aux Meadows?
Material which seemed to be connected
with ironmaking was studied in the labora-
tories of Oldsaksamlingen in Oslo by A.M.
Rosenqvist, and later also in laboratories in
Ottawa by Unglik and Stewart (1999). One
should be aware of the immense develop-
ment in analytical methods after 1970.
A.M. Rosenqvist concluded that a minor
amount of iron was produced at L’Anse
aux Meadows (Rosenqvist 1977/1985).
Bog iron ore at the site was analyzed
chemically after oxidizing roasting, with
the results as shown in Table 2 (Rosenqvist
1985:414). If all iron is counted as Fe203,
normal for roasted ore, the amount of iron
oxide will be changed to 17.5 + 79.5 =
97%, giving a total of 100.9. Unfortunate-
ly A1203 is missing. - The value for R - the
fayalite ratio - (% FeO + % Mn0)/% Si02
is added by the present author.1 In short,
this is a very rich ore, comparable to the
best ores found in Norway and in Iceland.
The content of Si02 is probably too low
for adequate slag formation, as expressed
by the value for R. This effect could be
remedied by the addition of a little sand
to the charge of ore, thereby increasing the
amount of slag and securing refining, with
a corresponding change for the value of R
to about 10.
Two more ore samples were
obtained by the present author during my
visit. Analysis in the year 1992 gave the
results as shown in Table 3 (notice the pres-
ence of A1203 in the table). Sample No.
1 was fresh ore taken from a vertical cut
of Black Duck Creek. Sample No. 2 was
provided by Dr. Wallace and represents
archaeological material. Both samples
1 Because of similarity of FeO and MnO in liquid slag, they can be grouped together. This author has introduced the value R for
the molar ratio (FeO+MnO)/Si02 for an evaluation of slags and also the raw material. The value 2 corresponds to the main
constituent fayalite or iron silicate in solidified slag. Successful smelting should result in R = 2 while a good ore should have
a large excess of iron oxide relative to fayalite and therefore be represented by R from 5 to 10. A slag control of the metal
quality is necessary. In order to create enough slag, there is therefore probably also a lower limit for the silica content.
55