In the summer of 2009, I was asked by a friend of mine, Develon, if I could make an authentic Viking-Age Gambeson for him to use in live steel combat. As I have done many sewing projects in the past, and have a specialization in historical clothing, I agreed.
After researching various styles of gambesons and arming coats in use throughout the middle ages, we decided that a quilted layered linen gambeson would be the best choice for the late Viking age, as it would consist entirely of materials easily available in Viking Scandinavia, and as layered linen gambesons were commonly used by the poorer classes as armor by themselves, unlike later quilted and stuffed gambesons that were primarily worn under other armor, whether chain or plate. Henry Skodall, a German historian, in his report on armor during the Norman invasion, states that “medieval Steppwämse [gambesons] were not padded, but by countless layers of linen were to hold the arrow fire.” Images in the Bayeux tapestry reflect this, with some armor appearing to be chain mail, some appearing to be ring mail, some appearing to be scale mail, and some appearing to be various forms of quilted gambesons.
However, there are few extant layered gambesons which could be looked at for construction methods and sewing techniques, and very few historical accounts of such gambesons aside from illuminated manuscripts and tapestries, making it more difficult to construct one correctly. So, the only option, after researching and finding next to nothing concrete on early, or pre-14th century Medieval padded armor, was to create our own from what is known, and then test its effectiveness. As Elizabeth Wayland Barber said in Women’s Work, textiles and textile products are most difficult to learn about, as few survive. Thus, it is often necessary to recreate textiles and clothing in order to learn how they were made and why.
It was decided to cut the gambeson in the ‘T’ tunic style, which would eliminate bulky shoulder seams and create a stronger cloth as a whole, and to cut every fourth layer on the bias, that is, with the weave of the cloth at a forty-five degree angle to straight, which is a feature found in some extant layered linen gambesons, and would cause the quilted cloth to be stronger and resist being pierced by arrows and other such points better than so many layers of straight cloth. The layers on the bias would also create more ‘stretch’ in the layers, which would again make the layering and quilting stronger. Unlike many reproductions of quilted gambesons which are available to purchase or see today, and are usually stuffed, it was decided to make the gambeson at least twenty-eight layers thick. This seems like a lot, but there are references to gambesons that had up to thirty layers. Indeed, the only way a gambeson of plain quilted linen could provide much padding on its own or much protection would be if it were incredibly thick.
After all was decided, Develon ordered the linen—forty yards of plain, tabby-weave, unbleached canvas-weight linen. Once it arrived, I cut it out according to measure: seven layers with sleeves, seven on the bias, and the other fourteen layers straight with no sleeves. (It was decided early on to only cut a fraction of the layers with sleeves to allow for greater mobility of the arms.)
The layers were then laid out together, alternating straight cut with bias cut pieces in order to make the fabric stronger. As the material was too thick to pin the layers together, the layers were then tied together in a grid with heavy string, much in the manner that a comforter is tied, to stabilize the layers and keep them from shifting while being quilted.
The next step was the quilting of the gambeson, which took approximately 34 hours. The lines of quilting were made three inches, or 7.62 centimeters, apart. The gambeson was quilted entirely by hand with a steel needle and rough, unbleached linen thread. To make stitching easier, the thread was waxed with natural beeswax.
Once the gambeson was entirely quilted, the edges had to be bound, gussets inserted under the arms, and lacing panels put on the sides to facilitate lacing eyelets to fasten the gambeson. Unlike some stuffed gambesons, which were tied up the front, we decided early on to lace this one up the sides, which would allow for a closer fit and better protection, less weight, and the ability to wear this gambeson easily under a coat of mail. So, the edges were bound, using the same running stitch that had been used for the quilting. Gussets for the underarms were cut out, much the same way that gussets would have been cut for a standard ‘T’ tunic, but instead of being sewn in one piece like a standard tunic, they were folded over and stitched as separate panels, allowing lacing eyes to adjust the fit on the arms as needed. Leather strips were sewn into the edges that would come under the most strain, those of the side lacing panels, and eyelets were sewn in—six in each side lacing panel, and seven in each underarm gusset, again with waxed linen thread.
Finally, a collar was cut out in the form of three strips on the bias, which allowed the fabric to curve smoothly around the round edge of the neckline and gave the collar a thickness of six layers, and binding was applied to the edges of the neck slit. The collar was top-stitched on the edge to attach it to the gambeson itself, then stitched through all layers to make it stand up. Leather lacing was cut out to lace up the sides, and the gambeson was completed!
However, our work was not done yet. To simply make a correct gambeson was not enough if we did not test it. So, a 15 by 18 inch (38.1 by 45.72 cm) ‘test piece’ was made, of exactly the same materials, layers, stitches, and quilting technique as the gambeson itself. Viking period correct arrowheads were ordered, both bodkins and broadheads, and placed on cedar shafts with hand-tied goosefeather fletching.
Develon already owned a 45-pound wooden bow and a Viking-age reproduction spear, sharp, as well as a couple of hand-axes and a sharp sword, which were also used in the test.
The primary goal of the test was to ascertain the effectiveness of the gambeson as stand-alone armor and when coupled with chain mail.
(In the following, piercing refers to an arrow or other weapon breaking some layers of the gambeson, but not all layers. Penetration refers to a weapon breaking or cutting through all layers, and amount of penetration refers to the part of the blade that came out the other side.)
The test piece was hung vertically from a wire flush against a stack of standard square hay bales.
A test shot was fired at just the haybales without the test piece, and the arrow nearly went completely through the bale.
A succession of arrows were then fired at the test piece from 30 feet with the following results:
1. Bodkin-tipped arrow, 16mm penetration.
2. Broadhead-tipped arrow, cut all the way through (However, this was on a seam between two haybales. May have been a lucky shot or due to absolutely no resistance behind test piece.)
3. Broadhead-tipped arrow, 13mm penetration.
4. Bodkin-tipped arrow, 36 mm penetration.
5. Bodkin-tipped arrow, 15 mm penetration.
6. Broadhead-tipped arrow, 7mm penetration.
7. Broadhead-tipped arrow, 37mm penetration.
8. Broadhead-tipped arrow, 9mm penetration—broke arrow.
9. Bodkin-tipped arrow, 7mm penetration.
10. Bodkin-tipped arrow, Bounced off, minimal piercing.
After ten shots were fired, a piece of 16-guage zinc-plated riveted steel chain mail was added on top of the gambeson, to compare with damage inflicted on the gambeson only.
Shots were as follows, again from 30 feet:
11. Broadhead-tipped arrow, bounced off, dented one link in mail.
12. Bodkin-tipped arrow, 7mm penetration, broke one link.
13. Broadhead-tipped arrow, small pierce, did not penetrate. Broke two rings in mail.
14. Bodkin-tipped arrow, 2mm penetration, broke two rings.
After the shooting was concluded, we decided to try a spear against the gambeson, to see how it would hold up to the significantly greater force of a spear. Develon threw the spear from 20 feet with chain mail over the test piece.
15. Sharp throwing/sparring spear, thrown. Small pierce, damaged chain mail, bent tang of spearhead seventy degrees.
We then removed the chain mail from the test piece, and repaired the spear. Erik, who was also helping to test, then threw the spear again from 20 feet.
16. Sharp throwing/sparring spear, thrown. Small pierce, no penetration.
Develon then thrust the spear at the armor in order to see if the spear could pierce the gambeson if there was steady pressure behind the spear.
17. Sharp throwing/sparring spear thrust, small pierce, no penetration.
Next, we attempted to damage the armor with both sharp hand-axes and a sharpened sword, and failed to cause any damage to the test piece. To dispel doubts about the sharpness of the sword, with one easy swing it was embedded into a block of wood, and stuck there on its own accord.
In all, there were 17 piercings or penetrations on the front of the test piece, and 12 on the back of the test piece.
From these results, it is possible to see that the gambeson is more effective against bodkin arrows than against broadheads. While the arrows would still pierce one’s skin to some extent were you shot while wearing it, a gambeson would be vastly superior to one’s tunic or even chain mail on its own, as the arrows tended to break links in the chain mail and then proceed to pass right through. However, broadhead arrows, which were more easily stopped by the chain mail, tended to slice right through the layers of the gambeson. The multiple bias-cut layers of the gambeson worked excellently against the bodkin arrows, as they pierced more than cut, and the bias layers caused an elasticity that resisted the piercing points of the bodkin-type arrows. As for the other weapons used, they appeared to be futile against the gambeson, but would have caused a great deal of bruising and broken limbs, ribs, etc., a condition that would have been easily dealt with in the Middle Ages, as cuts and infection were more to be feared.
The one disadvantage to quilted layered linen over a later stuffed gambeson is the cost and labor. While a stuffed gambeson can even be made mostly from scraps and waste fabric for the stuffing, with minimal ‘good’ fabric for the shell, a quilted layered gambeson would require about forty yards of good linen fabric, a good deal more than the four or five required for a stuffed gambeson. It would also require a great deal of labor—this gambeson required nearly sixty hours of skilled labor to construct, and that was from the fabric. (My finger turned greeny-blue from my brass thimble. I call it the badge of a seamstress.)
In an early medieval society, the linen would not only have to be quilted by hand, but the linen would have to be spun and woven into fabric first, before the gambeson could be constructed. To give one an idea of how labor intensive that would be, one must realize that it takes about ten times as long to weave a piece of cloth as it does to make something from it, and about ten times as long to spin the thread for a piece of cloth as to weave it. However, as the linen gambeson is made from a lower grade linen, we can surmise that it would only take 5 times as long on each count. Still, this would be a phenomenal amount to labor to construct a gambeson. However, when compared to the cost of shirt of chain mail, which would cost as much as a small farm, the cost of a gambeson would significantly less. For the amount of protection a layered linen gambeson provides, it would be a viable piece of armor for a lower-class fighter.
Skodall, Henry. “Schutzausrüstung Zur Zeit Der Schlacht Von Hastings” http://www.reenactment.de/reenactment_start/reenactment_startseite/diverses/kitguide/kitguide.html Accessed March 2010.
 Barber, Elizabeth Wayland. Women’s Work: The First 20,000 Years. 26.
UPDATE July 2011: This style of gambeson, custom-made, is now available for sale here from Spindle, Shuttle, and Needle.