If that isn’t a clickbait title, I don’t know what is. Seriously, though, this was a little project I did to try to understand not just *how* the garment was put together, but how the maker might have patterned the garment for a specific recipient – that is, how they determined the proportions and dimensions of each of the pattern pieces. I started with an excellent article by Anne Kwaspen that analyzed several pairs of 6th-7th century CE linen trousers from the eastern end of the Mediterranean, where conditions were favorable and preserved the garments almost perfectly. From there, I tried to figure out how I might have gone about making a pair sized for myself without the benefit of a modern tape measure or paper pattern. What follows is my trial-and-error process for figuring out how to come up with a pair of pants that use body measurements and result in a finished garment that generally matches the proportions of the original.
Note: The images of the original pair are from Kwaspen’s article – the article is free to read online at the link above, but I do not hold the copyright to the images so I will take them down upon request.
Background
According to Paulus Diaconus’ Historia Langobardorum, the Langobards had adopted the Roman/Byzantine custom of wearing underpants known as femoralia, a pair of short linen trousers or drawers.[1] My 7th-century CE Langobard persona needed underwear, so I went looking for something appropriate. In the absence of an extant find in a Langobard context, I looked at the similarities between northern European finds of wool trousers and a group of well-preserved linen examples from Egypt and Syria (areas that were nominally part of the Byzantine sphere until the middle of the 7th century). Although the crotch construction is slightly different, the northern European finds, including the Datgen, Damendorf, and Thorsberg trousers, all use the same construction method found in the Egyptian and Syrian examples – namely, each leg is a tube, seamed up the back of the leg, and connected by means of an arrangement of square and triangular pieces to form the seat and crotch.
In 2013, textile conservator and researcher Anne Kwaspen conducted an analysis of four well-preserved pairs of linen trousers from the Egypt and Syria, dated from the mid-6th through mid-7th centuries, in which she noted the parallels in patterning and construction with these European examples.[2] Kwaspen hypothesized that the patterning of all these trousers (both those found in Northern European and Eastern Mediterranean contexts) derives from a method that began with a tube of fabric wrapped around the wearer. The four examples from Egypt and Syria are different enough in dimension that it was clearly not a “one-size-fits-all approach,” but while Kwaspen’s work included a detailed examination and faithful replica of each pair based on the original dimensions (including trial on a variety of models of different ages and heights), she stopped short of speculating on exactly how such a garment would be patterned to fit an individual. My objective with this project was twofold: to make a pair of femoralia using the example now in the museum collection in Dusseldorf while simultaneously attempting to pattern the garment in this manner – that is, beginning with a draped tube and deriving a pattern from body measurements without the use of a paper pattern, math, or other later methods.
One caveat: I’m fairly diminutive by modern western standards, but just about exactly the right size to fit into the original Thorsberg trousers (based on published dimensions), so the premise of the whole project is that I am proportioned more or less like an average man in late antiquity or the early medieval period and could use myself as a dressmaker’s dummy while experimenting. Others’ mileage may vary, but my suspicion is that this method would still produce a wearable garment, even if the proportions of the pattern pieces are not the same as the original example. I would love to expand the information here to cover a wider range of body types, so feel free to get in touch if you are interested in making some very comfortable but relatively unflattering 7th-century trousers.
The Originals
These knee-length breeches have a 101.5 cm waist circumference (39”), 37 cm inseam (14”), and 53 cm outseam (20”), consisting of a front/leg panel, rear panel, triangular front gore, and triangular inserts at the back of the legs (while these are generally the same as the three other pairs in Kwaspen’s study, the maker cut one of the triangular inserts along with the back panel). Two belt loops are sewn to the front, while the upper back edge is folded and stitched down to form a belt tunnel. Likely made from recycled textiles (note red decorative stripes), the individual pieces are not all cut from the same fabric but are generally similar – a linen tabby with S-spun warp and weft. The main leg/front panel has a warp/weft count of 18/17cm, while other pieces vary from 17 to 30 warp/cm and 14-18 weft/cm. The garment was sewn together with linen S2Z thread. The upper edge and leg openings are finished with a rolled seam, while all other seams are flat felled.
Patterning steps:
In trying to determine body measurements, I looked at the ratio of the widths of the front/legs and back panels to see if I could find some uniform characteristics that could guide me and found that in all four examples, the width of the front panel comprised between 56 and 64% of the total waist circumference of the finished garment. I found that the two largest, rectangular panels could be derived from body measurements, while all the dimensions of the triangular panels were derived either from a side of a rectangular panel or were non-critical “guesstimates” that did not affect the patterning. Once I established what I thought were workable body measurements, I back-checked by comparing the ratios of each pattern piece’s dimensions to those found on the original garment. The diagram below is what I came up with, based on Kwaspen’s description of the originals, and is the reference for the explanations that follow:
- FRONT/LEG and BACK PANELS – DRAPED FROM A TUBE
Test 1: Wrap fabric to natural waist. Make this the width of the front panel. With this panel draped around the widest part of hips/buttocks, cut back panel to easily accommodate the difference.
Result: this could work but does not produce a ratio that reflects the originals, all of which have a front panel that constitutes an average of 60% of the total circumference. My measurements produced a front panel closer to 75% of the total circumference.
Test 2: Start with fabric wrapped tightly to natural waist. Half of this is the width of the back panel. (E to F). Mark (if cutting from continuous piece) or pin/pinch together with fabric intended for the front panel. Wrap the whole thing loosely around widest part of hips/back to determine maximum width needed to be able to pull the trousers on, allowing some ease to be taken up by seam allowances. This additional width of fabric becomes the front panel.
Result: this was closer to the original ratio, with about 64% of the total circumference in the front panel, but the front panel would be narrow enough that when split to form the legs, the cuff opening would be too narrow to get my knee through. For a pair of above-knee trousers, this wouldn’t work either. For this reason, Kwaspen concluded that the Düsseldorf trousers were made to fit an adolescent or smaller teen. I realized that to follow the ratio of the examples, I would have to calculate backwards from the thigh dimensions. This meant that if I also used the 50%-natural waist = back panel method, the front panel would in fact have to be the circumference of my natural waist, and that would have meant a nearly 50” circumference for the waist. Of the four pairs, the largest had a 48” waist, (although the outer tips of the triangular side panels were truncated and formed part of the width of the knee circumference, reducing the needed width of the front panel to achieve a well-fitting leg hem).
Test 3: I realized I was sizing the whole thing with a modern eye and aiming for a more fitted garment than was likely typical and that I should be aiming for a larger waist circumference than the absolute minimum required to pass over the hips. I returned to the first method, beginning with a snug natural waist measurement (below, left), moving that down to the widest portion of the hips and this time doubling the gap (below, right).
Result: Success!! This produced the appropriate 60-40 front-back ratio as well as an adequate thigh circumference. I double checked against the Düsseldorf trousers to see what would happen if I scaled up those measurements to fit me (based on thigh length and leg opening diameter, since the waist measurement was obviously not reliable on its own) and found that when increased by a factor of 10% all around, the Düsseldorf measurements were almost exactly the same as the ones I had arrived at.
- FIND HEIGHT OF BACK PANEL.
To find height of back panel and thus E-K and G-J, I tried some possible body-based measurements:
Test 1: With one leg bent 90 degrees (foot on chair), wrap back panel from center back through crotch to the approximate level of the top of thigh
Result: this won’t work – the back panel would be far too tall. E-K must necessarily be less than G-H. This dimension should be closer to the length of the front slit
Test 2: While standing normally, wrap back panel from center back at natural waist loosely through crotch so that it extends just beyond the front of the thigh.
Result: Success! This produced a ratio much closer to the measurements Kwaspen recorded for the original pair. Some excess should be included to form the belt tunnel as well – I eyeballed this.
- FIND TRIANGULAR PANEL DIMENSIONS
The diagonal of thigh gusset needs to be reasonably close to match edge of leg tube, thus G to H is less than or equal to B to D. Two of the four examples show that this hypotenuse was stretched to reach the needed length, so this was clearly not an exact science.
- Lay out back panel and fabric from which one of the triangular panels will be cut so that they are side by side.
- Hold one upper corner of the front panel to the upper left corner of back panel (slightly below upper edge, allowing for the folded down portion for the belt tunnel).
- Holding this pivot point, swing lower corner of front panel away until it matches the lower edge of the triangular panel. This will form a right triangle that should be the right size – mark and cut a rectangle of this width and the height from bottom of the back panel to the top edge of the belt tunnel, then cut along the diagonal to create two triangles.
- DETERMINE DIMENSIONS OF FRONT (CROTCH) GUSSET
Make a triangle large enough that the three points are above and to the side of any sensitive areas – this is also not a super precise measurement because the front slit will be cut to accommodate it as needed, but a generous hands’ breadth plus seam allowance should be fine.
Sewing up:
Based on the seams on the extant garments, Kwaspen was able to determine the order in which the pieces were put together
- Crotch gore is sewn to center of bottom edge of back panel with the seam allowance felled against back panel
- Triangular thigh inserts are sewn along the sides of the back panel to form the back trapezoid (flat felled seams on the back-panel side)
- Upper edge of back panel is folded down with a 1cm seam allowance and whip-stitched to the right side to form a tunnel for a belt
- Upper edge of front/leg panel is finished with rolled hem
- Sew B-D to G-H and same on opposite side, felled against the front panel
- Use H-J-K-L-M to calculate length of front slit and cut up middle front – C-N
- Sew the bottom edge of the back piece to the slit so that the point of crotch gusset is centered at the top, then fell seams against front panel.
- Finish leg openings with a rolled seam.
Lessons learned:
Step 6 is the magic trick that simplifies the task – cutting the slit to match the other pieces, rather than attempting to dimension two panels with a combined length to match an existing slit, eliminated what might have been the most challenging aspect (having to reverse-engineer the pieces comprising the combined H-J-K-L-M measurements based on a preordained C-N measurement). I actually was able to start at the bottom edge/leg opening and work my way up one side of the inseam so that I could clip the last inch or so of the front slit to the exact length needed.
The other hard-to-calculate factor is the width of the triangular side pieces (and the corresponding acute angle at the upper and outer corners). The indications of stretching along the hypotenuse in two of the four examples in the study group shows that this also did not have to be exact, as the bias stretch of the fabric provided several centimeters of leeway.
Summary
In short, two body measurements from a draped tube were sufficient to establish a garment that fit well and yielded the same proportions of individual pattern pieces as the original example. The other measurements were either based on the first two or did not require precision prior to sewing up. Given the close similarity to the proportions of the original garment pieces, this method certainly seems like a plausible way of deriving the pattern from the body with ease.
[1] Paulus Diaconus, Pauli Historia Langobardorum: In Usum Scholarum Ex Monumentis Germaniae Historicis Recusa (impensis bibliopolii Hahiani, 1878), V:38.
[2] Anne Kwaspen, “The Pattern-Cutting of Linen Trousers in Late Antiquity,” In: Drawing the Threads Together, Textiles and Footwear of the 1st Millennium AD from Egypt, accessed February 28, 2020, https://www.academia.edu/5383533/The_pattern-cutting_of_linen_trousers_in_Late_Antiquity.
farolfvs me fecit 