Professional Documents
Culture Documents
Chain-mail is known in Europe as far back relation to plate armor. Chain-mail is also
as around 300 BC. The method of lighter – the almost complete shirt from
manufacture is believed to have been Vimose, displayed in the Danish National
discovered by the Celts, but the oldest Museum weighs only 8 kg. But a shirt of
remains discovered (in Denmark) are from mail represents much more work than sheet
the Hjortespring find on the island of Als, armor, such as the well-known ‘Lorica
dated to around 350 BC (Rosenberg segmentata’. In the Roman army, only the
1937). This find was excavated by Gustav better-off legionnaires were able to afford a
Rosenberg in 1921-22 and today, there is shirt of mail (Robinson 1975:164). We do
almost nothing remaining of the material not know how long it took to
deemed to be chain-mail. Rosenberg
describes several square meters covered
with heavily corroded material. But there is
doubt as to whether this was chain-mail, or
a layer of natural iron separation formed
around plant roots, the occurrence of
which can often be in the form of rings.
The Romans adopted shirts of mail and
they were subsequently used throughout
the Middle Ages (Blair 1972:19ff,
Robinson 1974:154) most often in the 12th
and 13th centuries but small pieces of mail
were used right up until the 17th century
as a part of armor. Although shirts of
chain-mail were used for nearly 2000
Fig. 1
years, there has been surprisingly little
Typical chain-mail construction. Rows
interest in the manufacture of them. Since
consisting of jointless rings alternating
they are a very fine and without doubt a
with rows of riveted rings. Each riveted
highly specialized craft, carried out over a
ring connects four jointless rings.
very long period of time, they are an
excellent object for studying technological
make a shirt of mail. In medieval German
development from the Iron Age to the
guild books we can see that a mail-maker’s
Middle Ages.
masterpiece was a shirt that would take
him 6 months.
The construction of chain-mail Common to all the types studied is that
A shirt of mail is a form of armor, made up
they consist of two ring types (1). The first
of rings. As the rings can move in relation
is made of a wire that is bent, overlapped
to each other mail is far more supple in
and connected using a rivet. The other has The biggest overlap is found on the
no visible joint. The two types always Vimose shirt, where there is up to 7.0 – 8.0
alternate in rows allowing each riveted ring mm and on the same shirt, the overlap
to hold four jointless rings. Fig 1. consistently follows the same direction,
Fig. 2.
Ring size and appearance The rivet heads are nicely rounded in
In the four types of chain-mail studied, ring contrast to the rivets used in some cases
size varies greatly. The finest were found for repairs where hammer marks can be
at Hedegaard, the rings of which had a clearly seen on the rivet heads.
diameter of around 5 mm and a wire
thickness of around 0.95 mm. Somewhat Metallurgical study
poorer is the Brokbaer shirt, with a ring The jointless rings: A plane-section was
diameter of around 7.2 mm and a wire made across the jointless rings and a series
thickness of around 1.0 mm. A shirt with of cross-sections of the wire were taken
very similar rings was found at Vimose (2). The plane-sections showed that the
(Engelhardt 1869:12). The biggest mail rings are made of carbon-free iron. Only a
rings were discovered at Vimose and few rings from the Thorsbjerg shirt were
Thorsbjerg: 10.5 mm and 12.5 mm found to contain up to 0.3% carbon. The
respectively, with wire thicknesses of iron is generally quite homogenous and
approx. 1.5 and 1.7 mm (Engelhardt contains almost no slag – in other words, it
1863:59; 1869:12). is high quality. The rings of the Hedegaard
Corrosion to the shirts from Hedegaard shirt contain 0.2% phosphor. There is no
and Kastenskov has caused the rings to sign of weld joints, which may be expected
swell, making it difficult establish the if the rings were made of wire bent into a
original thickness. The Brokbaer shirt is ring shape and then welded.
also heavily corroded, but it is possible to
see that the jointless ring had a flat side in
towards the hole and a more rounded
exterior. The clearest cross-section of the
jointless ring can be seen on the shirts from
Vimose and Thorsbjerg. Once again, the
inside is flat and the exterior more rounded
and a groove can be seen all the way round
on one side of the rings. On the small
Vimose shirt, the edges of the groove
actually form burrs of 0.5 – 1.0 mm.
The cross-section of the riveted rings is in
every case more rounded, although the
wire thickness can vary. These rings are Fig. 2
not as uniform as the jointless rings, but The overlap of the riveted ends differs,
have a tendency to be oval and with much from 1.5 – 2.0 mm to 7.0 – 8.0 mm. The
greater variation. The overlap at the rivet overlap is always in the same direction
is always flattened, but least on the within the same mail-shirt, but can be both
Brokbaer shirt, where only a canted cut is ways.
apparent with an overlap of 1.5 – 2.0 mm.
Fig. 3
Plane-sections of jointless and riveted
rings show different slag structures. The Fig. 4
origin of the jointless ring structure is best A cross-section of jointless rings shows a
explained by theorizing that these rings typical deformed structure where the
were cut from a thin iron sheet. If they had material at the sides has been forced
been formed by drawn wire, the structure downwards (a). In some mail-shirts the
would be similar to the form found in outside of the rings are rounded, probably
riveted rings. by deliberate grinding in order to remove
sharp edges (b).
Some hint of the method of manufacture
can be gleaned from the characteristic slag The riveted rings have an entirely different
pattern, which runs in parallel bands slag pattern to the jointless rings: The slag
regardless of the rounded shape of the pattern runs parallel with the surface of the
ring. A slag pattern of this nature can only ring. Cross-sections reveal no depressed
occur if the ring is made of a thin, beaten structure.
sheet and not if the ring (as previously
assumed) was made of welded wire. Fig The manufacture of the rings
3a. The observations noted here can be
Cross-sections of the whole ring also show interpreted as follows:
a very characteristic slag pattern. The The jointless rings: The raw material used
structure is clearly deformed, as the for the rings was a 1.0 – 2.0 mm thick
material has been forcibly depressed at the sheet of almost carbon-free iron. The rings
sides. This explains the burrs that even were either produced using a single
with the naked eye can be seen on many of stamping process or in two steps, using
the rings. These are the remains of grooves hollow punches. Stamping would seem to
such as those that can still be found on the be the most logical process for the
small Vimose mail shirt. On the rings of the manufacture of very large numbers of
Thorsbjerg shirt, the large Vimose shirt uniform rings, but a stamping tool must
and possibly the Brokaer shirt, the groove have been difficult to produce given the
on the inside of the rings is very visible technology of the time. The individual
whilst the outside is more rounded. Only a components of the tool must have been
trace of an outer groove can be seen. exactly the same size and the upper part
The riveted rings: A horizontal section of must be placed accurately in relation to the
the rings shows that they are also made of lower in order to avoid damaging the tool
carbon-free iron. The phosphor content in itself. Additionally, the newly stamped ring
the Hedegaard shirt rings is 0.1 – 0.2%. could get stuck in the tool. Today, they are
Fig. 5
The rings can be formed either by a
stamping machine or by first punching the
outer rim and the hole with hollow
punches.
Fig. 9
Based on bloomery slag analyses,
Northern Europe may be divided into eight
main areas, each having its own unique
chemical characteristics.