Reproducing Linen Trousers from Roman Egypt

SEEKING CROWD-SOURCED HELP

I'm putting this post up now, but it's really not quite ready.  Below, I've laid out my best approximation of how to pattern some interesting trousers (possibly better described as underwear) from Roman Egypt.  

These proportions worked out very well for me.  However, I've not had any opportunity to try this out on people with different body shapes than mine.  I'm hoping that by putting this up here, I can get other re-enactors out there to help me out by trying this and seeing how well it works for a wider range of bodies.  So, if you're reading this and you think these trousers are kind of neat - please try this out and let me know how well these proportions work for you.  

You can find a link to email me, by clicking on "view my complete profile" in the right-hand sidebar.

_________________________________________________________________________________

Romano-Egyptian linen trousers

            Most investigations into the clothing of Egypt in the Roman era (sometimes called Coptic era) focus on the elaborately decorated tunics that have survived.  But, what did they wear underneath?  It turns out that there are 4 extant linen trousers from Roman (Byzantine) Egypt.  A marvelous article analyzing these trousers is available online (here).  Two of them were subjected to radiocarbon dating and both are from the late 6^th or early 7^th century CE.  I highly encourage anyone interested in these garments to read Kwaspen and De Moor’s article.  This brief tutorial will focus on the process of reproducing this style of trousers.

            The 4 trousers that Kwaspen and De Moor analyzed are all constructed in the same way, with some slight variations.  The internal proportions of the pieces are relatively similar among them as well.  This enabled me to reproduce a pair sized to fit me and assess the fit based on proportions of my measurements. 

            It must be pointed out that the construction of these trousers is completely unlike modern pants.  The fit is likewise unlike modern pants.  The use of a back panel and triangular gussets create a shape that seems much baggier in the rear than any sort of modern pants.  However, this room easily accommodates the movements of the body in a wide range of positions.  It’s unclear to me whether these trousers were worn by men, women, or both.  However, the cut of the gusset is particularly well-suited to accommodate male anatomy.  As an interesting side note, the construction of these Egyptian trousers has some similarities to northern European finds from earlier in the first millennium. 

            As Kwaspen and De Moor point out in the article, there is no evidence that any sort of physical pattern was used by the people of this time – and quite likely the cut was known well by those who made them.  Since the pieces that make up these trousers are all simple geometric shapes (rectangles and right triangles), it is easy enough to mark the necessary measurements directly on your fabric.  Be sure to double check your measurements – measure twice, cut once!  

            The original trousers are made from linen that (in most cases) seems to have been repurposed. In some, different pieces of the same garment are cut from different cloths.  They range from 13-35 threads per centimeter.  So, a lightweight or medium weight modern linen would be suitable for a reproduction. 

Measurements

Cutting the pieces for these trousers only requires 3 measurements:

Hip measurement (H) – is the circumference around the widest part of your hips and buttocks

Outseam length to the ankle (OA) – this is the full length of the pants from where the drawstring will sit to the ankles

Outseam length to the knee (OK) – This is the length from where the drawstring will sit to the knee

             One of the pairs of surviving trousers is knee length, and if you want to make a knee length pair, you will only need to measure the outseam to the knee (OK).  If you want full-length trousers, measure to the ankle also (OA). 

Pattern pieces

            The main body piece is a large rectangle.  The width will be 3/4 of the hip measurement (H) and the length will equal the full outseam (OA).  You will also need to mark the top of the slit.  Measure down from the midpoint of the top edge 1/8 and mark this point.  The main body piece will be slit from the bottom up to this point. 

            The back panel is also a rectangle.  The width will be 3/8 H and the length will be ½ H + 2 in/5cm.  The extra length is will be folded over to make a channel for the belt. 

            There will be two triangular panels.  The length will be 1/2 H.  The diagonal will be OK.  Mark a vertical line of length 1/2 H on your fabric and mark two lines perpendicular to this.  These will be the top and bottom.  Measure from the top of your vertical line and find the point where a measurement of OK meets the bottom line.  This will show you width of the rectangle.  Cut out the rectangle and then cut it in half along the diagonal line.  . 

            The last piece is a crotch gusset.  The crotch gussets on the extant garments are not perfectly square.  But, for simplicity we will be cutting our crotch gusset as a square.  Two of the surviving trousers have a gusset that sets into slits on both the main panel and the back panel.  One has a triangular crotch gusset that sets into only the slit in the main panel.  The last has no crotch gusset at all, but this would create a very high-stress point where a straight edge from that back panel sets into the slit of the main panel.  I would not recommend this construction.  I actually got a better fit using a triangular gusset seamed to the bottom of the back panel than I did by setting in into a slit. 

            The crotch gusset is based on a square, with each side 1/8 H in length.  For these instructions, you can cut that in half along the diagonal. 

            Lastly, you will need 4 belt loops that will be attached to the top of the main panel.  I made mine by using a strip of selvedge from my fabric.  Kwaspen and DeMoor don’t explicitly list the size of the belt loops on the originals.  I used a strip 2 inches wide, folded in thirds with the selvedge on the outside, then cut into 3 inch long pieces. 

Please excuse my poor skills with computer graphics.  I hope these diagrams are clear.

I have used ½ inch/ 12mm seam allowance throughout (except for the belt channel, which is described below).

Order of construction

            The original garments all have seams that are stitched and then flat-felled.  This arrangement allows the order of operations to be deduced.  It can be done either by hand-stitching, or by machine.  To reduce bulk, I recommend trimming down one seam allowance on each seam before folding them over to make the flat-fell seam. 

            First, sew each triangular panel to the sides of the back panel.  The long leg of the triangle (the side which you measured at 1/2 * H) is sewn to the back panel.  Match the bottom (short leg) of the triangles to the bottom of the back panel.  Press these seams toward the back panel to flat-fell the seam allowances.  When pressing the seam allowances to make the flat-fell, you will also press a double-fold hem for sides of the back panel above the point of the triangular panels. 

You can then fold down the extra 2 inches at the top of the back panel.  Fold these 2 inches to the outside.  Fold the raw edge under and stitch this down to create a channel for the belt. 

Match the middle of the bias edge of the crotch gusset to the middle of the bottom edge of the back panel.  Stitch the gusset to the back panel.  Press these seam allowances toward the gusset, but don’t flat-fell them yet.

Stitch a double-fold hem along the top edge of the main panel.  Cut a slit up the middle of the main panel from the bottom edge to the marked point (1/8 H from the top edge).  If you will be making full length trousers, you should mark the knee level on both sides of the main panel and both sides of the slit.  This will be at OK from the top edge.  Set the corner of the crotch gusset into the top of the slit.  Continue stitching the sides of the gusset, then the back panel and triangular panels to either side of the slit.  Stop stitching at the point that marks the knee level.  The point of the triangular panels may go past this.  That’s fine and if it does, you will trim it off later. 

To flat-fell all these seams, first fold under and stitch the seam allowances for the seams joining the crotch gusset to the slit.  These will be pressed away from the gusset.  As you approach the edges of the gusset, trim down the seam allowances.  You can now flat-fell the remaining seam allowances here as one long seam, pressing the seam allowances toward the main panel (toward the gusset in the middle of the seam). 

On each side, stitch the side of the main panel to the hypotenuse of the triangular panel.  Again, stop stitching at the mark for knee level.  If the point of the triangular panel extends past this point (or if it extends past the knee marking on the slit) trim it off.  Fold seam allowances toward the main panel and flat-fell those seams.  The folding from the flat-felling will continue naturally into a double-fold hem on both sides of the main panel and both sides of the slit below the knee level.  (see the diagram below)  Stitch these hems.

 If you had to trim off the points of the triangular panels, make a narrow hem along that cut edge.  Stitch a hem along the bottom edge as well.  Lastly, attach your 4 belt loops to the top of the main panel. 

I'm including a line diagram of one of the surviving trousers from the Kwaspen and DeMoor article here.  The solid lines indicate seams and the dotted lines indicate the direction the flat-felling is folded.  If this is unclear, please follow the link at the top and read the paper itself for more details. 

Diagram of surviving trousers by Kwaspen and DeMoor

To wear these trousers, a belt of some sort must be threaded through the channel at the top of the back panel and through the belt loops.  There is no evidence as to what sort of belt was used with the originals.  I have used a length of twill tape, tied at the front.  However, a cloth or leather belt with a buckle would also be a justifiable choice.  The legs are left open at the back below the knee.  The originals have ties attached at the corners of the hem.  Thus, it is assumed that these sides were tightly wrapped around the lower legs.  Again, Kwaspen and DeMoor have some photos of their reconstructions in the article and I encourage everyone to read it.  

Finding the Intersection of Authenticity and Safety

 Clothing vs. armor – What to wear for fencing?

            I wrote this up with the intent to present it at a display at an event this spring which was cancelled.  I came across the mostly finished write-up, going through some files on my computer and figured I may as well put it up here for those who may be interested.  

 The concept:

            The rules for rapier fencing in the SCA presume that fencers are wearing “common civil attire” for the purposes of acknowledging blows.  The rapier was, indeed, a civilian weapon – used for self-defense, affairs of honor, and occasionally brawling, but not typically on the battlefield.  So, this convention makes perfect sense. 

             However, for reasons of safety there are certain requirements for the clothes worn by fencers in the list.  These requirements regulate the type of fabric and number of layers in order to minimize the risk of injury.  These requirements don’t always match up with typical clothing construction from the pre-1700 era.  Most fencers implement the protective requirements by building additional layers into a garment that would not normally have them, or by using modern materials underneath a more period-appropriate item. 

             Since I practice a Spanish style of fencing (first described in writing by Jeronimo Carranza in his book which was published in 1582), I have been focusing on Spanish clothing of the 1570s-1580s.  Attire for a gentleman of this era would typically consist of several layers.  (Gnagy)  So, my goal with this project was to see if I could create an outfit for fencing with as few deviations from historical construction methods as possible. 

Fencing suit, version 1 (before the Unfortunate Incident)

Planning:

            The SCA rules for fencing include several requirements that I did not try to adjust.  This mainly had to do with “rigid material”.  There is not really any part of normal clothing that would use these sort of materials.  So, I accepted these as the necessary intrusion of modern considerations into my fencing attire.  The “rigid materials” include groin protection (worn under the clothing), a fencing mask (definitely NOT optional), and neck protection (in my case, a leather gorget worn over the clothing). 

            The primary considerations for this project were the construction of the main body garments.  These are the doublet, breeches (calçones), and the sayo (the outer garment, which could be called a jerkin in English sources).  These will each be considered separately.  The rules also require complete coverage of the back of the head, with no bare skin showing.  This requires a sort of garment that is essentially foreign to late 16^th century fashion, but I did create one that was ultimately rooted in it.  All these garments count as “puncture resistant” – able to withstand 1.5J of force over a 5/32” diameter circle in accordance with the rules’ procedures for punch testing. 

            In addition to the items constructed by me for this project, there are some additional accessories that require no modification to comply with the SCA fencing rules. 

The shirt, as underwear, is essential to dressing in 1580s style.  However, as underwear, I did not consider it in my reckoning of the number of layers required.  It is essential for wicking of sweat, especially in hot weather.

The legs must be covered by “abrasion resistant” material.  Period-appropriate knitted stockings meet this criteria.  Stockings in the late 16^th century would have been knitted from wool or sewn from woven linen.  (Arnold)   Knitted silk stockings were also known, but would be inappropriate for anything short of fancy occasions.  While I may eventually sew some linen hose, I decided not to expand the scope of the current project too much.  Knitted wool stockings are readily available.  However, while these are fine in cooler weather, they are not usually thin enough to be comfortable for me in hot North American summers.  At Pennsic, I compromised with knitted linen stockings.  There is no evidence of knitting with linen in the 16^th century, however. 

The fencing rules require shoes that completely enclose the foot.  I have a pair of leather reproduction shoes that work well.  However, under very muddy conditions at Pennsic I did resort to some period-ish wooden clogs (photo).  

Lastly, the rules require gloves of at least “abrasion resistant” material.  I am not a glover, so I use commercially produced gloves.  They are made with leather and the cut is not too different from surviving late 16^th century gloves. 

Choice of materials:

            The SCA fencing rules do list standard materials that are known to meet the standard of “puncture resistant”  as well as testing procedures to assess puncture resistance.  Other than leather, the listed examples are not typical of fabrics used for common everyday clothing in the late 16^th century. 

Surviving examples of clothing from the 16^th century are often silk. (Arnold, Braun et al.)  However, this is clearly survival bias and internal structural layers of these fancy clothes are of linen and wool.  Everyday clothing would almost universally be made from wool as the primary material, with wool and/or linen interior layers and linings.  (Gnagy, Johnson)  Linen, in addition to use for undergarments, was occasionally used for doublets and jackets as well.  This use is attested in Spanish colonies in America and the Caribbean.  (LaPorta) 

Since the clothing I made to fence in would be subjected to significant wear, it needs to be sturdy.  It also needs to be worn during occasionally vigorous exercise, sometimes in very hot weather.  Since it is difficult to find densely woven, sturdy wool in a weight other than those intended for winter coats, I have chosen to use linen for all parts of these items.  Linen canvas doublets are known from this period.  Outer garments (jerkins, sayos, ropillas, etc.) would probably not have used linen as the outer layer.  However, heavier weight linens were known and used for the internal layers of tailored garments.  So, in the absence of suitable wool fabrics I have made the sayo and breeches out of a heavy weight linen as well. 

Since the complete outfit would consist of a doublet and breeches under a sayo, my estimation was that the combination of the layers in the doublet and the sayo would meet the required puncture resistance.  Not counting the linings, the doublet has one layer of heavy linen canvas and the sayo has two.  There is additional linen in certain areas, as required by the tailoring methods, but puncture resistance must be measured in the most vulnerable part of the torso clothing (the fewest number of layers).

The Doublet:

            The doublet is the foundation of men’s clothing from the late 16^th century.  Aside from the use of machine stitching for many of the construction seams, this doublet was made without any deviation from typical historical construction methods.  The pattern was drafted using Mathew Gnagy’s reconstructed proportional system and construction mostly follows his recommendations based on analysis of extant garments. (Gnagy)  It is a single layer of linen canvas, with additional reinforcement of the same canvas used in the buttonhole areas and the collar.  A lightweight linen was used for lining.  The buttons are ball-shaped metal buttons.  Cast metal buttons were known in this period, and would have been a cheaper alternative to expensive jeweled buttons, or buttons wrapped in silk thread (which are seen on higher class surviving garments).  There is a strip with eyelets inside the doublet waist to use for attaching the breeches. 

The pieces of this linen canvas doublet are shaped as Gnagy describes (using steam to stretch certain areas) even though there is no interlining layer for the body.  However, a padstitched piece of canvas in the collar provides essential support for it to hold its shape.  Seen here before the buttons and buttonholes were made.  

The Breeches:

            The breeches pattern was also drafted from Gnagy’s method.  The SCA rules only require abrasion resistant covering on the legs, but since the breeches do cover the lower part of the torso, I have to consider them in the overall outfit.  I used the same heavy linen as the first version of the sayo for the outer fabric.  The same canvas used for the doublet was used as a lining.  Aside from the use of machine stitching for some construction seams, the construction mostly followed historical methods.  The single deviation is discussed below.  One huge advantage of the historical construction is the use of lacing holes on the waistband the tie the breeches to the doublet.  This essentially prevents any possibility of gapping between the breeches and doublet which would expose non-puncture resistant underlayers.  I am currently using cords made by loop braiding linen thread.  As a utilitarian garment, I tried to just make wrapped thread ends, but these are proving less than satisfactory.  I intend to put plain metal aglets on the ends of the points. 

            A significant safety consideration in the construction of the breeches was the form of the button opening.  This style of breeches, which appears in the 1570s-1580s (photos), lost the ornamental codpiece that was typical of men’s clothing of the mid-16^th century.  The necessary front opening was closed with buttons.  However, in portraits these are often not seen.  Gnagy reports that surviving examples use a deep pleat at the center front to cover the buttons.  This configuration is surprisingly practical for safety purposes.  A strip of fabric with buttonholes extends from the left front of the breeches.  The buttons are sewn to the right front.  In order for a broken blade to slip under this fly, it would have to come from the right.  But, the right edge of the fly is never exposed due to the deep pleat covering it and the fullness of the breeches. 

            This configuration would probably have been enough for safety.  But I added another buttonhole stand extending in the opposite direction on the inside.  I will admit that a partial consideration here was that I couldn’t decide whether to cut the breeches from the same cloth as the doublet or the sayo (both ways are shown in the cutting layouts of Freyle’s tailor’s manual).  With a little planning, and the addition of the second button stand, I have effectively made the breeches reversible.  This is in no way supported by any historical practices.  But, I like the idea of effectively having two pairs of breeches for the effort of one. 

The Sayo:

            The sayo was originally cut from the same heavy linen as the breeches, with a linen canvas interlining and a loftier linen for the tailored padding layer.  The lining is a lightweight linen.  Again, the pattern was drafted using Gnagy’s system with construction in line with his methods.  Again, machine stitching was used for many of the construction seams.  I have made the sayo to button right-over-left, while the doublet buttons left-over-right.  The direction of buttoning was not standardized in the 16^th century and examples of both can be found.  While I find left-over-right much easier, as a right-handed person, right-over-left has two advantages in this instance.  Firstly, it has the opening facing to my left, which is usually the side away from my opponent.  This reduces the risk of a broken blade sliding in between the buttons.  Secondly, by using opposite buttoning for the two layers, it further reduces the chances of any broken blades penetrating fully. 

Machine pad-stitching on the interlining layer of the sayo. The 'padding' is actually the same canvas as the interlining itself.  This provides no real contribution to the protective aspect of this garment, but is essential to the construction process that gives the garment its shape.

            My original plan for the sayo had been to use a short-sleeved style.  The SCA fencing rules require puncture resistant material in the armpit, covering the upper part of the inner, upper arm.  The remainder of the arms only require abrasion resistant material.  Short sleeves are rare, but there are several portraits from the 1560s showing men with short-sleeved overgarments (jerkin, sayo, etc.).  The original plan was to make sleeves only as long as the portion of the upper arm that needs to be protected with puncture resistant material.  However, after two attempts, I could not find a configuration of short sleeves that I felt would not allow a blade to slide up under the sayo sleeve.  So, I changed plans.

            The final version of the sayo has a sleeve shown in tailor’s manuals from just after 1600.  This is called a “manga de armar” in Spanish and usually translated as “arming sleeve”.  I don’t know of any images of this sleeve style, but there are many reasons why it might not be readily identifiable.  It is a one-piece sleeve, cut in a very unusual shape that folds up to create the same shape as a more common curved, two-piece sleeve.  If fully constructed in this way, the arming sleeve would have a seam up the back of the forearm, across the elbow, and then up the front of the upper arm.  The more typical two-piece sleeve has two seams, running the full length of the sleeve at the front and also the back of the arm. 

Since the arming sleeve shape is much less efficient in use of fabric than the two-piece sleeve, I find it hard to believe that 16^th century tailors would have constructed the arming sleeve this way.  I suspect that it may have been constructed without sewing any seams at all – leaving the “seam” lines as openings (possibly with buttons or ties) through which the sleeve underneath would be visible, or even puff out.  Gnagy suggests that the seam on the upper arm would be sewn closed with slashes only across the outside of the elbow and down the back of the forearm.  This arrangement suited my purposes even better, as a large opening on the front of the upper arm would be a spot where a blade could potentially get caught.  However, with the upper arm sewn closed I can still put my forearm through the elbow slash, leaving the lower sleeve hanging.  Since the opening here is at the elbow, any blade that slipped up under the sayo sleeve would still not get up past the mid-bicep.  I could put my forearm into the sayo sleeve as well, but having the option to wear it as a hanging sleeve helps reduce layering on parts of the body where it is not necessary – something useful in hot weather. 

Unfortunately, after wearing this suit to fence at Pennsic 48 (2019), I washed it.  Being constructed entirely of linen, it should have been safe to wash in the machine.  However, it seems that I neglected to pre-wash the linen I used for the inter-lining in the sayo.  This shrank during washing, creating a hot mess of a garment that is totally unwearable.  I got myself some more heavy linen and made a second sayo, duplicating pretty much everything I did on the first one.  This is the currently displayed sayo.

Aside from the use of machine stitching, the main deviation from historical construction practices is the use of linen for the outer fabric.  While linen canvas doublets are documented, colored linen was exceedingly rare (LaPorta) and its use for a sayo is questionable. In the absence of sturdy, lightweight wool cloth, however, I felt that this was the better choice for a garment meant to be used during vigorous physical activity.

Version 2.  Replacement sayo with the same doublet and breeches.

Other options/future explorations:

            Initially, I thought that three layers of linen canvas would be required to meet the SCA requirement of puncture resistance.  However, I suspect that the sayo, as made, could pass punch testing on its own.  With proper choice of materials, it might be possible to construct a doublet that functions on its own.  This could include making a doublet with heavy linen canvas for the outer shell and the interlining.  It could also involve use of leather, a very historically appropriate material for clothing and one which would meet the SCA standard of puncture resistance as well.

            There are a few surviving examples which are presumed to be fencing doublets.  These doublets have quilted padding.  While not necessary for complying with SCA fencing rules, there are advantages to having a padded garment for fencing.  This would be more along the lines of recreating a period “fencing armor” than “common civil attire” that complies with SCA rules. 

Historical practice, however, is to stitch buttons to the edge of the garment and place buttonholes very close to the edge as well.  With a single garment, an additional placket of some sort would have to be pinned or stitched behind the buttonholes in order to comply with SCA rules.  My use of two layered garments with opposite-facing button openings eliminates the need for this ahistorical modification. 

Another deviation from historical practice to consider would be to only interline the armpit area of the sleeves.  It seems that doublet sleeves were either fully interlined, or not interlined at all.  However, to eliminate bulk in the sleeves and still comply with SCA fencing rules, interlining could be placed only in the areas where puncture resistance is required.  This would be completely unnoticeable in the finished garment.  Another modern “cheat” that could be combined with this technique would be to cut the layer of padding used in the tailoring to cover the entire body of the doublet, rather than only the chest and shoulder area.  This would add a third layer of cloth to the one garment.  

Sources:

Johnson, Caroline.  The King’s Servants: Men’s dress at the accession of Henry VIII, ed. Malcolm-Davies & Mikhaila, Fat Goose Press, 2009.

Gnagy, Mathew.  The Modern Maker Vol. 1: Men’s Doublets,  self-published, 2014.

Gnagy, Mathew.  The Modern Maker Vol. 2: Pattern Manual 1580-1640, self-published, 2018

              Includes LaPorta, A. “Cloth and Clothing in 16^th Century Spanish Florida”

Greco, El.  Portrait of a man, 1570-75.  https://commons.wikimedia.org/wiki/File:El_Greco_-_Portrait_of_a_Man_-_WGA10428.jpg

Mikhaila, Ninya and Jane Malcolm-Davies.  The Tudor Tailor, Costume and Fashion Press, 2006.  

Mor, Anthonis. Portrait of a gentleman  https://commons.wikimedia.org/wiki/File:Anthonis_Mor_Portrait_of_a_Gentleman.jpg 

Moroni, Giovanni Battista.  Portrait of Don Gabriel de la Cueva, Duke of Albuquerque, 1560.  https://commons.wikimedia.org/wiki/File:Giovanni_Battista_Moroni_-_Portrait_of_Don_Gabriel_de_la_Cueva,_later_Duke_of_Alburquerque_-_WGA16256.jpg

Sanchez Coello, Alonso.  Circle of?  Undated. https://blouinartsalesindex.com/auctions/Alonso-Sanchez-Coello-6125171/Portrait-of-a-Man,-three-quarter-length,-with-a-double-layered-lace-cartwheel-Ruff;-Portrait-of-a-Woman,-three-quarter-length,-with-a-double-layered-lace-cartwheel-Ruff