Change view: Change style to alternate Change style to default

Silk-Wrapped Wire Circlet

"... a deceptively simple thing ..."



I'd like to say that I set out to research and make a 14th century Lady's circlet and veil after a style depicted in the Luttrell Psalter (c. 1300-1350), using extant articles as a technical guide. But it isn't true. Ultimately that's exactly what I did, but when I started, all I was curious about were the silk covered wire frames described in Dress Accessories. They seemed so simple, and yet I couldn't see how they were made. How did they wrap such thin wire with silk and then twist it into little coils and other shapes? I spent almost two years considering this, occasionally trying ideas as they came to me. I did finally figure out how it could have been done, and in the process learned quite a bit about silk threads, dying, veil shapes and fabrics, as well as how those ladies kept such a sheer veil on the back of their head without it slipping out from underneath the fillet! To see if I was correct on any of it, I had to actually do it.

The materials

To understand how the item was made, I needed to use the same materials to produce my own - thin wire and unthrown silk. Unthrown silk is made of the filaments from a number of cocoons unwound together. One cocoon provides one very long silk filament - about a mile long. Thrown silk has gone through a twisting process, which strengthens the thread and keeps the filaments in order. Unthrown silk skips these steps. In either case, the thread is then boiled in an alkaline solution (usually soap) to remove the sericin, the glue that held the silk together as a hard cocoon. So how could I get this unthrown, filament silk? Unfortunately, embroidery or sewing silk is usually made of short bits of silk twisted together and so will not fan out in a fine layer as it wraps around the wire - and trying to untwist a sufficient quantity just isn't practical.

I tried several methods of producing my own unthrown silk, including twirling a bamboo skewer (substituted in a pinch for wire) above a pot of 20-25 softened and ready-to-reel cocoons. The effect was lovely and very durable, because the sericin stuck everything together, but the method is almost certainly wrong. The wires would have had to be produced right in the filatures where the cocoons were reeled and there's no indication of this.

Ultimately, I found a commercial source for a heavy, 2-ply weaving yarn of filament silk. Being so thick, it was easy to unply, and any remaining twist could be gently unrolled while wrapping it around the wire.

The wire was the easiest part to match. Dress Accessories describes the silk covered-wires as "made from drawn wire, usually copper-alloy but occasionally iron." The finer wires ranged in diameter from .2mm to .8mm, and were often coiled into spirals and sewn or further coiled around somewhat thicker wires (1 - 2mm). I used 2mm copper wire and .5mm brass wire (28 gauge), consistent with the extant circlet that was my guide.

Dying the silk

Dress Accessories does not say whether the silk-covered wires had any traces of color. I decided to dye some of the silk, taking my cue from the book's mention of "14th and 15th-century reliquaries embellished with multi-coloured silk wire flowers, often with seed pearls and beads threaded onto them" and from the floral circlets seen in some illuminations. I used Brazilwood for the deep rich reds it produces. Although rarer to find on excavated textiles than madder or kermes, according to Clothing and Textiles, it was imported from the late 12th century and is "encountered in larger numbers in textiles surviving above ground."

I had to dye several batches before finally getting it all right. I discovered that the yarn didn't dye evenly unless it was unplied first, but unplied yarn is difficult to keep from tangling. For another batch, I mistakenly used tap water and produced a vile shade of light purplish brown. When I finally got the dye bath made correctly with good spring water, I didn't strain it completely and tiny bits of bark permanently embedded themselves between the silk strands. At last I produced the color I wanted, using approximately equal weights of brazilwood chips to silk with an alum pre-mordant. I was pleased to discover that medieval dyers also spoke of dyeing by weight of goods.

Covering wire with silk

I still had the problem of how to get the silk onto the wire efficiently and in a way logical to the time period. Many of the silk-covered wires recovered from digs were very thin and then coiled into spirals after being covered with silk. This required a lot of wire to be covered.

For my first attempt, I began with the thickest copper wire - 2mm. I cut it to length and found I could simply twirl it between my fingers while guiding the silk around it. This was only possible because of its limited length and its thickness - enough to hold and control easily. (Trying this with the .5mm wire was an exercise in frustration. ) I steamed the silk after wrapping it around the thicker wire, but still had a problem with the ends slipping. I tied them tightly, and tucked the tail under the silk as much as possible, but the last few twists of silk were still free to rotate around the wire, ultimately loosening the silk along the length of the wire and causing gaps.

I really didn't want to use my hot glue gun on it, I wanted to understand how medieval craftsmen made silk covered wire. Perhaps they did use an adhesive of some type, but I found an alternative that felt like a very natural solution - both available and logical. I drew the wire along a cake of beeswax, just as I would to strengthen a sewing thread. It made the wire just tacky enough to keep the silk from slipping.

On the verge of giving up on the thinner wire, I picked up a small brass rod with a bend at the end that I use to teach spinning. I twisted the wire around the hook, secured the remaining supply of wire, and found that now I could twist the rod between my fingers instead of trying to handle the fine wire on its own. It went so well, that soon I was at my arm's limit. For any good spinner, the next logical thing to do was to wind the excess length around the rod, pass it under the hook and continue "spinning". That was when I discovered that if carefully done, this produces tidy coils of silk covered wire. There's no need for a second step to form the wire into coils - it's a side effect of wrapping the wire with the silk! I could slide it off the end and use it coiled, as many wires were, or unwind it if straight pieces were needed - as with the circlet I was trying to make. As before, drawing the thin wire though beeswax helped immensely.

Unfortunately, I don't yet have any information on the diameter of the medieval coils to know how thick of a rod they were wrapped around. With a little bit of work, I can tighten my coils to a smaller diameter if I need to.

Circlet Circlet

Assembling the circlet

I prepared two bundles of silk-covered wire in red to knot around the thicker wire, which I covered in white silk. Dress Accessories had a diagram of the knots shown in two colors, but I think this was intended to clarify how the knots were made, and not to suggest that the original item used two different colors. (No indication of color is given at all.) It's possible that the knots and base frame were all the same color, or even undyed. I liked the look of red knots because they reminded me of tiny rosebuds. I kept the base frame white for contrast.

As I manipulated the wires to form the knots, I tried to keep the alternating pattern shown in the diagram. I discovered that I didn't have to keep track of a "right over left, left over right, reverse for the next knot" pattern as long as I had one bundle I could identify (it was smaller) always dominant - it always crossed over the other wire, keeping the pattern on track without any additional effort.

The original item is described as having the knots "bound in place by silk thread." This is the only part I didn't reproduce. I just couldn't see that it was needed. I can always do it later if I discover why the original maker thought it was necessary.

The Veil

For the veil, I originally made the oval shaped veil I'm used to seeing. Then I noticed that the Luttrell Psalter has several illuminations of ladies with their veils flying out behind them in two long streamers. A long, narrow rectangle could do this, but not an oval. Supporting a rectangular shape, relic veils described in Textiles and Clothing are all quite narrow, from 235mm (9 1/4 in.) to 460mm (18 in.) - which was also noted as being the complete loomwidth. Only one length was mentioned: 720mm (28 1/4 ins) on the narrowest veil. I tried out a similar sized sample, and it's quite small. Positioned with the front edge of the veil at the crown of my head, the long edges came only a few inches down my upper arm, and the back barely reached the base of my head. The illuminations (see below) show a much more generous veil. In the dining room scene, it comes to the tip of the elbow, and in the arming scene to the small of the back. If the flying veil in the garden scene were brought down, it would also reach to the small of her back.


Aware that illuminations are not photographs, I still opted for the more generous look. I was interested to notice that the front corners of my veil reach to my elbow while the back edges reach the small of my back. The width after hemming is about 16 3/4 in. - wider than some, but still within the known range. When worn, it extends a few inches below the back neckline of my dress.

The fabric used for the extant veils is described as being in a balanced tabby weave (a simple over-one under-one weave with a similar number of threads per inch in the warp as in the weft), lightweight with a very open, transparent look. The threads are heavily twisted in the same direction for both warp and weft (Z-twist or clockwise), giving a light crepe (crinkled) effect. There were between 48 and 60 threads per cm in the warp, and 42 to 52 per cm in the weft - although I discovered that this information isn't of much help if the thread diameter isn't known. I could find silk fabrics that each had a few of these characteristics, but none with all of them, so I compromised with a very sheer, balanced tabby weave silk for my veil.

The selvedges (or at least one selvedge) should have extended the length of the veil, but this would have required me to buy a far longer piece of silk than my budget would allow, so I settled for turning it on its side, using the width of the fabric for the length of the veil, and hemming the long edges. As appropriate for a veil of this type, I used a narrow rolled hem. I found my hem to be within the measurements indicated in Clothing and Textiles - less than 1mm deep and with 5-6 stitches per cm. For fun, I used a silk thread that I reeled and threw myself. It slid through the fabric beautifully and was a pleasure to use.

Illuminations in the Luttrell Psalter show that the veil was worn under a fillet on the back of the head. The front hem starts at the very top of the head. So why didn't it slip out from under the circlet without even a barbette (chin band) to pin it to? Probably because it was sewn to the circlet. A roughly contemporary silk-covered wire frame described in Dress Accessories had fragments of a sheer veil stitched to it. It works quite well, and can be removed if needed.



It was very satisfying to produce an item that matched both an artifact and contemporary illustrations, not just in the way it looked, but also in the way it was made. It was very important to me to test my theories about construction methods by duplicating an existing item. If mine didn't measure up (literally!) to the original, then I didn't have something right - be that materials, technique or skill.

Now that I have more experience and confidence handling unthrown silk, I think I'll make my own next time. I had some difficulty with the commercial silk thread I used for the finer wire. Even using just one ply, it was too thick and it kept bunching up in places until I learned how to control it. If it works, my own silk would be finer than the commercial silk and I think that will solve the bunching problem. I also would like to research the wire flowers mentioned in Dress Accessories and try making them to learn more about medieval dying and working with silk wires.