The piecing of the Charles de Blois pourpoint
The silk fabric of the pourpoint of Charles de Blois was cut into an intimidating 27 pieces before assembly. Upon consideration, its tailoring is not 27 pieces-worth of complicated, however, because many of those original pieces could have been combined — with no change to the shaping of the garment — had the fabric been wider. To be exact, the pieces of the garment could have been reduced to 18 with no change to its shape at all. I believe the overly-pieced cut is proof that the fabric was both narrow and scant in length. To this date, a detailed analysis explaining why the extra piecing was placed where it was has not been published. This article shows how the original pieces fit neatly into a layout on narrow, folded fabric while also revealing the original fabric’s width, which I extrapolated using the armhole measurement and a basic understanding of tailoring.
The pattern diagrams used as the basis of this study were published by Adrien Harmand in 1929 in Jeanne d’Arc: Ses costumes, son armure: Essai de reconstitution. Of the various published pattern diagrams for this garment, I believe these to be the most accurate. I compared them with decent photos of the original and so far have not found any noteworthy discrepancies.
The diagrams show two portions of the garment, each one on its own page:
Before starting, I first needed to have confidence that the two separated diagrams were proportional to each other. I measured the armhole on the left and compared it to the measurement of the sleeve cap on the right. They matched, confirming proportionality:
Laying out the Body Pieces
How does one begin a task like figuring out such a garment’s cutting layout? I began with a hint I found in the text accompanying the diagrams. Harmand states that a sliver of fabric was added to each side seam and that the bottom back piece was filled out with two small slices of fabric because of the “insufficiency” of the width, presumably that of the fabric itself. (116) In the figure below I have highlighted the side sliver in red and the lower back slices in yellow to make clear which pieces Harmand referred to:
The reason we can assume the fabric was too narrow is because the lower back slices would have been cut as one with the larger pieces, had the fabric been wide enough. Some basic tailoring theory explains why. The seams that join the slices to the main lower back piece are created from two straight edges sewn together, both on-grain (for the most part; there appears to be a very slight slant in the figure above). When fabric is joined in this way, it does not create 3-D shaping. It merely extends a flat shape. There is no benefit to sewing three pieces together when one would do. The obvious exception would be noticeably-different fabrics joined to create a visual pattern, but that is not the case here.
But which part of the fabric was narrow — the width or length? When woven fabric is limited size-wise in one direction or another, it is usually selvedge to selvedge, also known as “width”. This is because long lengths of fabric can be produced on a loom, but the fabric’s width is always constrained by the loom’s width. It seems likely that the lower back piece’s width is artificially limited by the available fabric’s width. I am extending this concept to include almost all of the pattern pieces because examination of photographs shows the brocade medallions facing the same direction as they face on the lower back piece. In the figure below I demonstrate the orientation of the lower back piece on the fabric:
Is it possible that the pattern was laid out transversely instead? It is within tailoring propriety to rotate pattern pieces 90 degrees and cut them consistently in that direction. This type of cutting is certainly possible in the middle of the 14th century, when the pourpoint was made. Agnes Geijer asserts that a common width of Italian silks in the time period was between 115-120 centimeters, which translates roughly to 45-47 inches. (102) While some silk available in Western Europe was wide, not all of it was, nor was all of it made in Italy, with its impressively wide looms. Imports from Persia and further East were still common. If the fabric were narrow, however, the long front piece would be impossible to accommodate:
But let us imagine that the original tailor had generously wide fabric and chose to lay the pattern in a transverse orientation. Would he still have needed to use 26 pieces when 18 would achieve the same tailoring effect? The only reason to include extra straight-edge piecing in a pattern made of uniform fabric is to waste as little of the fabric as possible. I performed a little experiment where I laid the 26-piece version of the pattern out on a set size of folded fabric and did the same with the 18-piece version. Both fit on the same sized fabric, which proves the tailor did not use a transverse layout. The layout on the left shows the 26-piece version while the layout on the right shows the 18-piece version:
It is worth noting that the upper and lower back pieces are symmetrical. This means they can be cut on folded fabric. The advantage of this is the use of less scissor strokes, which saves time and energy in the cutting. The top back piece is nearly the same width as the lower back piece, and each front piece (side sliver excluded) is exactly half the width of the lower back piece. By limiting the shape to half the width of the fabric, the tailor uses the available fabric very efficiently. If the fabric were folded and then the front, top-back, and bottom-back pieces were laid on top, this is how they might appear:
Finding the Fabric’s Width
Before going much further, I found myself wanting to know exactly how wide the original narrow fabric really was. Harmand states that the grande assiette armhole measures 1.03 meters. With this one measurement, I used basic math to find the original fabric’s width. As mentioned previously, the widest of the body pieces is the lower back piece. If its widest measurement were found, the fabric width could be guessed, probably within an inch of accuracy. For practical purposes, rather than enlarge the diagram to match the life-size measurement, I measured the armhole circumference on the page-sized pattern, in inches. This number came to 4.5 inches. Then I measured the widest portion of the lower back piece, which came to 2.125 inches, bringing the lower back piece’s width to approximately 47% of the armhole circumference.
Reverting back to the metric system, I multiplied the original metric armhole circumference (103 cm) by .47 to find the widest width of the lower back piece. It came to 48.41 cm. In inches this converts to slightly more than 19 inches, which means the fabric was probably about 20-21 inches wide, if we assume at least a half inch of wiggle room near the selvedges. This allowed me to visualize the fabric folded lengthwise to a 10-inch width — seriously skinny fabric!
Laying Out the Sleeve Pieces
Now how about that sleeve? As you can see from the marked-up pattern diagram below, both the upper sleeve and the lower sleeve are vertically split in two, with approximately 25% of the width in one piece, and 75% in the other:
We already know this serves no tailoring purpose on its own due to the straight edge to straight edge non-tailoring rule.* However, if you take the pieces outlined in red in the diagram above and place them on our 10-inch wide folded fabric, they fit perfectly:
What remains are the four sleeve gores (one front gore, one underarm gore, and two back gores), the sleeve cuff, the remaining vertical pieces of the sleeves, the lower back slices, and the sliver intended to complete the shaping of the front pieces. How might they have been laid on the remaining fabric? Here is one speculative take on fitting all the remaining pieces except for the back gores — for which the grain direction required more analysis — and the side sliver, which was probably cut last and which will be discussed in more detail below:
As for the back gores, you can tell at what angle they were laid on the fabric by examining the direction of the lion and eagle medallions on the fabric. The photo below is marked with arrows to indicate the grain direction of the gores:
To further illustrate how this translates on the pattern diagram, I added yellow lines showing where the back gores got sewn together and became part of the seam that runs down the back of the entire length of the sleeve:
Transferring this information to the fabric looks something like this:
But that is undoubtedly incorrect. The tiny triangle completing one of the back gores is apparently only present on one side of the original, if we are to trust the redrawing of the garment in Janet Arnold’s article on the Black Prince, Edward of Woodstock’s pourpoint (18). This lack of symmetry is a mystery, considering that everything was cut on folded cloth, which produces two mirror versions of each piece cut. My best guess is that a mishap befell one of the gores after it was cut out but before it was sewn. Perhaps the tailor accidentally snipped into one piece at some point in his process. Or, perhaps the tip of one piece got soiled and needed to be cut off and replaced. I doubt we will ever know for certain. If the more likely scenario is that the tiny triangle was a later correction, I think the layout would have looked more like this originally:
And Finally, the Side Sliver
The only piece yet missing from my layout is the side sliver, which, if you will recall, is oriented in the same direction the front piece is. According to the original diagrams, the sliver was cut perpendicular to the rest of the pieces’ direction, which means it was rotated to point at the selvedge. It measures longer than 10 inches, which is the proposed width of the fabric folded in half, however. It must have been cut when the fabric was folded end to end, not selvedge to selvedge. The diagram below shows how this works:
It most likely was cut after everything else. It has been speculated that it was added after the fact purely as a correction for better fit, but as already explained, the front pieces were perfectly limited to the 10-inch width to take maximum advantage of fabric conservation. I believe the side slivers were intended as a planned part of the front pieces, not to correct a fitting mistake by the tailor.
The total cut pieces in this garment numbered 27 by the end of the garment’s construction, but likely started as a set of 26 pieces.
So there you have it — a full garment-worth of pieces crammed into less than three yards of 20-inch-wide fabric. To do this, the tailor broke the garment down into 26 pieces from a starting total of 18. Somewhere in the process of construction, however, one gore was pieced further, bringing the total number of pieces to 27 on the finished garment.
What I have not done is proven every way it could have been cut. Proceeding from the premise that the fabric was narrow and then folded in half lengthwise, there were probably a number of possible cutting layouts. Variation would have likely centered around the order of all the smaller pieces placed in the spaces left by the main body pieces. I performed this exercise to understand exactly why some of the non-tailoring-related piecing existed on the original. In seeking that answer, my respect for the original tailor and his clever approach to cutting precious fabric grew greater than it already was.
* In all fairness, there is a very slight curve added to one side of the smaller upper sleeve piece. This was probably a fit adjustment added after the garment was assembled and fit on the body. Whatever it was, its effect is subtle and not enough reason for cutting both the top and bottom portions of the sleeve in multiple pieces.
Arnold, Janet. “The Jupon or Coat-Armour of the Black Prince in Canterbury Cathedral.” The Journal of the Church Monuments Society VIII (1993): 12-24.
Geijer, Agnes. A History of Textile Art. London and Totowa NJ: Pasold Research Fund Ltd, 1979.
Harmand, Adrien. Jeanne d’Arc: Ses costumes, son armure: Essai de reconstitution. Paris: Librairie Ernest Leroux, 1929.