Árbók Hins íslenzka fornleifafélags - 01.01.1982, Blaðsíða 56

60
ÁRBÓK FORNLEIFAFÉLAGSINS
nails etc., and the waste products, the residues or slags that are the by-product of the ironworking
process. The technological ability of the early ironworkers can be judged from the examination of
the artefacts and to some extent of the (smelting) slags. The re-use of artefacts and their high sus-
ceptibility to corrosion usually deny the archaeometallurgist the opportunity of examining them
to determine their mode of manufacture. It is therefore the slags that may provide the most suit-
able opportunity for technological study.
The working iron can be subdivided into two processes. First the smelting process, which is the
reduction of the ore to malleable iron (the bloom) in the furnace. Secondly the smithing process
which is the working of the bloom into artifacts or the subsequent re-working of repairing of arti-
facts. Both these processes produce slag as a by-product. It is therefore of the greatest importance
to be able to distinguish smelting slag from smithing slag, in order that the correct process or pro-
cesses can be ascribed to a given site. It should be noted that the smelting process requires con-
trolled furnace conditions, ie. high temperature, reducing atmosphere, whereas the smithing
process can be carried out in virtually any available herath. Therefore smelting slags tend to be
deposited in the vicinity of the furnace, while smithing slags are commonly found widely distri-
buted across archaeological sites. Thus stray slag finds are more likely to derive from smithing
rather than smelting. Both slag types, unfortunately, have a similar chemical and mineral com-
position, principally iron oxide (normally in the form of Wustite, FeO, and iron silicate (Fayalite,
2FeO. SÍO2) in a glassy (Anorthitic, CaO. AI2O3 2SÍO2) matrix. It is therefore very difficult to
differentiate the two slags by chemical analysis alone.
This problem has been investigated over the past three years at Aston University,
(Birmingham), and in many cases the smelting and smithing slags can now be distinguished, but in
others the identification is at best tentative. The criteria for indentification will be published at a
later date.
Two samples in the form of thin sections were submitted for identification, with a description
of their finds location.
The specimen from Steinfinnsstaðir showed a densely packed microstructure of rounded iron
oxide dendrites (probably wustite) with very little fayalite or glassy phase present. This structure is
typical of smithing slags, the iron deriving from the hammer scale flaking from the artefact dur-
ing heating and fusing together in the smithing hearth to form a hearth bun or hearth bottom. The
latter is typically planoconvex in section, and varies from (approximately) 2—15 centimetres in
diameter, and 2—8 centimetres in depth. The find location, which was in the area of abandoned
settlement would be in keeping with the specimen being smithing slag.
The sample from Þuriðarstaðir has a more varied microstructure comprising of a high proport-
ion (about 60%) of fayalite laths with dispersed iron oxide dendrites (probably wustite) in a glassy
matrix. A structure typical of many smelting slags. The fayalite laths are indicative of a rapid
cooling rate for the slag (commonly smelting slags are tapped from and cool outside the furnace).
The iron oxide dendrites are more skeletal in structure than those from Steinfinnsstaðir and
comprise about 20—30% of the total mineral composition, which can be regarded as large for a
smelting slag and suggests that the sample derives from a relatively inefficient technology, ie. that
some of the iron that could have been extracted from the ore has been retained in the slag.
The position of the find location of the Þuríðarstaðir slag, on the summit of a hill does support
the theory that it is a smelting site. This exposed position should not necessarily be interpreted as
having been selected for the prevailing winds etc. to supply a natural draught. Since induced
draught (chimney effect) and bellows are dependable and more efficient.
There are three major raw materials required for iron smelting. Firstly, turves and stone for the
furnace(s) structure which are normally easily available. Secondly fuel (normally wood charcoal