Archaeologia Islandica - 01.01.2007, Blaðsíða 52
Arne Espelund
everyday life. Slaves from places such
as the British Isles were taken to Iceland.
However, it is very unlikely that the slaves
initiated ironmaking because it required
next to knowledge a certain organization.
In Greenland and at L’Anse aux Meadows
real slag heaps have not yet been found.
The main part of this paper is therefore
devoted to Iceland, where the process of
ironmaking at the site Belgsá in Fnjóska-
dalur can be followed from the raw mate-
rials bog iron ore and wood to metallic
iron and slag. Samples from the deserted
farm of Sandartunga in Þjórsárdalur have
also been studied. Small finds of iron in
Greenland and at L’Anse aux Meadows by
many are claimed to represent a local pro-
duction of the metal. Before presenting
the Icelandic material, a critical assess-
ment of such claims will be undertaken.
Greenland
Archaeology in Greenland has above all
focused on the physical remains of hous-
es, churches, costumes, tools and uten-
sils, and the contact with the Inuit culture.
The sudden collapse of the society in the
15th century has created much atten-
tion. Vestribygd disappeared suddenly in
1342, followed by a rapid decline also
in Eystribygd. The sagas and the dating
of physical remains correspond well. At
most there were 200 farms in Eystribygd
and about the half in Vestribygd (Vebæk
1963, Petersen 1967-76, 1984).
The well-known Danish archaeo-
metallurgist Niels Nielsen studied finds of
slag and metal in the 1930s and claimed
that the settlers in Greenland made some
iron (Nielsen 1930, 1936). In response
Rolf Falck-Muus told the same “iron story”
(Falck-Muus 1930). Nielsen refers to his
own studies of thin sections by microscopy
and mentions the phases magnetite, oliv-
ine, hematite and magnetic iron. Success-
ful iromnaking, leading to malleable iron,
required a slag control of the carbon con-
tent in the metal. After solidification such
slag contains mainly the silicate fayalite
Fe2Si04, in addition the oxide FeO and nor-
mally a glass, which included other oxides
such as A1203. Under no circumstance can
the oxide hematite Fe203 and metallic iron
occur together. The samples therefore
cannot represent the waste material after
iron production. He also mentioned crude
bog ore (Nielson 1930:202). He gives the
dimensions of individual pieces of slag,
sometimes 2 cm thick, 4.5 cm across,
with adhering clay. However, no weight
nor total chemical analyses are presented,
and these samples cannot have come from
a normal bloomery site.
The Danish archaeometallurgist
VF. Buchwald has spent much time in
Greenland and says that he has not come
across bog iron ore. In a book printed
some 140 years ago it was said that such
ore is found “here and there, always
scanty” (Rink 1877: 391). Probably the
lack of suitable wood is more indicative.
At present only small bushes of e.g. be-
tula nana dvergbj0rk are found (Buchwald
pers.comm.). Buchwald’s investigation of
iron objects represents a new approach:
the introduction of scanning electron
microscopy has created completely new
possibilities. By focusing a beam of
electrons on a cut and polished sample
each element will present a characteristic
spectrum, which can be analyzed by two
different approaches: by energy or wave
length dispersion. Buchwald has applied
the first technique on both samples of
slag and metal. Of special interest are
slag inclusions in metallic samples. A full
account of the method and its application
was presented recently (Buchwald 2005).
50