- lronworking in the Celtic World -
earliest furnaces used for smelting iron were probably based on the furnaces used
for smelting copper. The superstructure of such furnaces is rarely well preserved,
unless, as was sometimes the case, they were sunk into the ground, probably to
reduce heat loss. Recent work has confirmed that many furnaces which appear to be
bowl furnaces were almost certainly the bases of low shaft furnaces, 30 to 40 cm in
diameter, with a cylindrical superstructure half a metre or so in height. Experimental
reconstructions have shown such furnaces to be relatively efficient in producing iron,
whereas there are considerable problems with bowl furnaces. Such furnaces lacked
arrangements for tapping the slag, and this will have accumulated in the furnace; the
bloom could be removed through the top.
The taller shaft furnaces, standing some 2 metres high and often with an adjacent
pit into which the slag was tapped, may have been introduced in the later Iron Age,
although the best evidence for their use is in the Greek and Roman world, and they
do not appear in any numbers in Celtic areas until after the Roman conquest.
If the low shaft furnace dominated the western Celtic tradition, two alternative
types were in common use in the eastern part of the Celtic world. The first was a
short shaft furnace placed above a deep pit in which the slag was collected. In North
Africa, where similar furnaces have remained in use until modern times, it was
possible to move the furnace shaft to another pit when the first was full, but the sheer
weight of most of the east European examples would probably have prevented
this. Such evidence as we have suggests that this type of furnace did not appear
until relatively late in the Iron Age. Another form, commonly found in Bavaria
and Austria, had a dome rather than a shaft. With a diameter of a metre or more and
standing about 80 cm high, it was larger than most of the other types. Although
essentially a central European form, isolated examples have been claimed as far west
as northern England.
By modern standards such furnaces were highly inefficient; not only did a large
part of the iron end up in the slag, but great effort was expended to produce a rela-
tively small piece of iron. Experimental work by Mr Peter Crew based on late iron
age iron production sites in Snowdonia, suggests that a low shaft furnace could be
used for up to twenty smelts, each producing a bloom weighing about 2 kilograms.
Thus the entire production of the ironworking settlement at Crawcwellt, where
a total of fifteen furnaces were used, was probably no more than 600 kilograms of
raw iron, which, after lengthy processing to expel the slag, would have left about
300 kilograms of workable iron. Experiments using replica furnaces suggest that the
production of this amount of metal would have required 30,000 kilograms of
charcoal and taken some 7,500 man-days, up to four years' work for a team of six
men if we allow for occasional holidays. By comparison the time taken to make a
simple iron stool would have been quite short, not more than a few hours.
Even if we allow for the uncertainties in such calculations, the effort required to
produce iron artefacts assumes its true proportions. Such objects were prized because
they represented weeks or months of work, and the great quantities of ironwork
found at cult sites throughout the Celtic world can be seen as offerings of real
munificence. The fact that for much of the Iron Age even the smallest scrap of iron
was collected and reprocessed is understandable when the effort of producing the
raw material is appreciated.
313