Miklagard Garrison
Bottom of page

By Peter Beatson - NVG Mikligard


Woolen bands of a noticably uniform type occur amongst textile remnants excavated from early medieval urban sites across the breadth of Europe (Fig. 1). The bands are usually 75-100 mm broad (Fig. 2), and were deliberately woven to this narrow size rather than cut, as selvedges exist on both sides. The weave is almost invariably of 'herringbone' type, often of a very fine quality.

Technical analysis

Weave: A few exceptions aside (see Table 1), the bands are woven in 2/2 warp chevron twill with non-matching repeats (Fig. 3-a), in worsted woolen yarn. The warp is Z-spun, and the weft S-spun. Thread counts are variable, but usually high, indicative of high quality cloth (Fig. 4). The weave was usually 'unbalanced' to some degree, ie. there were more warp threads per cm. than weft. This was countered to some extent by making the weft thread slightly thicker than the warp. The warp:weft ratio used may reflect local weaving traditions, for example the three Hedeby examples, though of different qualities, have a similar unbalanced (2:1) warp:weft ratio, whereas those from the British Isles are all closer to 1:1 (Fig. 4).

In chevron twills the order of binding reverses frequently to give the characteristic 'zig-zag' appearance. The number of warp threads lying between each reversal seem to have been chosen purely for convenience, as it does not seem to vary with the thread count. The most commonly chosen interval was ten warp threads, though six was also popular, especially in Britain (Fig. 5). Mistakes in counting are frequently apparent, and in some examples there does not seem to be any attempt to produce a regular pattern at all Fig. 6). One Elisenhof example has only a single reversal, in the middle of the band.

Selvedges are characteristically of simple unreinforced type, ie. the weft thread simply turns back on itself to pass through the next shed at the edge of the fabric (Fig. 7-a). An exception are bands, up to 16 cm wide, from Finnish graves at Zalatovye (Khvoschchinskaia 1992), with tubular selvedges (Fig. 7-b). A band (S-26) from Hedeby had one selvedge reinforced by hemming, though in this case it seems likely that it was hemmed during reuse, probably as a clothing patch (Hägg 1991).

Colour: Although only a few fabrics have been analysed for dyes, lichen purple was found in the York and nearly all the London bands, and indigotin (blue from woad?) in two of the latter. Bands E-32 from Elisenhof, and S-26, H-28 and S-34 from Hedeby appeared to have been artificially coloured- the first two red, the others yellow. In some examples different-colored warp and weft were used to accentuate the weaving pattern- one of the Elisenhof bands (E-479) has dark (naturally pigmented?) warp and light (unpigmented?) weft; whereas Dublin band E172:12064 appears to have a light warp and a dark weft.


These pieces are usually interpreted by costume historians as 'puttees', spiral wrappings for the lower leg ( = German: Wickelbander), so often seen in manuscript illustrations from the period. The stretchiness of herringbone weave makes it particularly suited to this type of function. As the finds are casually discarded rags rather than clothing excavated in situ upon a corpse, this identification is hard to confirm. Similar bands sometimes accompany North German bog sacrifices from the earlier Iron Age, but even in these cases they are not worn by the body (Hägg 1984; Hundt 1981). In manuscripts, wickelbander are almost always worn by male figures, though occasionally they peep out beneath long female garments (Owen-Crocker 1986). In Latvia, similar bands appear to have been worn around the lower arm (Hägg 1984; Zarina 1994), and herringbone was used for women's headdresses as well as limb bindings (Zarina 1990).

Fastening must usually have been accomplished by simply tucking in the loose ends. Occasionally there is evidence of more elaborate systems. Ties below the knee with tassels or tags are seen on some high status figures in manuscripts, etc. (Fig. 8; & Fig. 9). Narrow tablet woven garters were used by Viking Age Finns to fasten their chevron twill puttees (Khvoschchinskaia 1992), the garters often had plaited terminals interlaced with coils of bronze wire (Fig. 10), a particularly Baltic embellishment.

A pair of hooks were found below the kneecaps of a male skeleton in Birka grave Bj903 (Fig. 11-a). On the reverse of the hooks were remains of woolen twill cloth (Wickelbander?). The hooks were fastened into small iron rings, to which linen tabby (from the trousers?) was rusted (Hägg 1986). Similar bird or mask-like hooks are known from other finds (Fig. 11-b). Small silver hooked tags were found in a similar position in a 9th cent. burial at Winchester, England (Fig. 12), here the sex of the deceased was indeterminate. Biddle (1990) concludes they may have been on the end of garter straps or ribbons, but suggests that a better use for such hooks are to hold together two edges of a garment. For instance, they could have been used to fasten a split in the upper edge of a short stocking, allowing it to fit snugly to the calf.

The Germanic aristocratic fashion for leather garters with elaborate metal fasteners, which reached its zenith in the Merovingian period, was almost extingushed by the Viking Age, but examples, in bronze (Fig. 13) and iron (Fig. 14), are known from three graves at Lejre, Denmark. A set, comprised of a buckle, usually with a strap retainer loop, and a strap end was found near each knee.


Apart from weaving bands yourself, a reasonable solution is to cut strips from whole cloth, using inconspicuous whip or blanket stiching in matching thread to stop the edges from fraying. 'Herringbone' wools are widely available, though modern quality is generally not as high, four measured samples varied between 8-13 warp by 8-10 weft per cm, reversing every 8 or 16 warp threads (compare Fig. 4). Modern weaves usually have different-coloured warp and weft, to accentuate the herringbone effect. It is unclear whether this was the rule in the early medieval period, but at least some wickelbander were made this way (see above). Because of the high thread counts and frequent reversals, the 'arrowhead' patterns were generally smaller and 'pointier' than available modern fabrics, but the known variation in patterns would certainly accommodate modern examples. Blue and red-purple seem to have been the popular colours.

How long should they be? No intact specimen exists, but three surviving Viking Age band fragments longer than one metre have been reported - Elisenhof E-479 at 124 cm, Hedeby H-55C at 118 cm and Hedeby H-43 at 102 cm (Hägg 1984; Hundt 1981). Complete bands are known from pre-Viking bog burials mentioned above, eg. Damendorf (2nd cent. AD) 105 cm; and Bernuthsfeld (7th cent. AD) 370 and 292 cm (Hägg 1984; Hundt 1981; van der Sanden 1996). Reconstructed shin bands, based on Hedeby band S-26, made at the Historical-Archaeological Research Centre at Lejre, Denmark were 360 cm long (Krag 1994). From my experience, a minimum of 2.5 to 3 metres per leg is needed to wrap the foot and shin - a pair can be made from a single metre of cloth, by joining up strips end to end (as Bernuthsfeld Man's are) with reversible seams.

An earlier edition of this article appeared in the 'Varangian Voice', issue 37 1995.

All rights reserved by the author.

Textiles Table - Published Viking Age textiles identified as 'Wickelbander'
Textile Picture's - Published Viking Age textiles identified as 'Wickelbander' ( Load's Textiles Picture's - 1-14 )
References for - Published Viking Age textiles identified as 'Wickelbander'

Information Supplied for this page was by Peter Beatson.

This Page was Created By Roy Pickton A Member of - NVG Mikligard

Top of page