For years, I made buttonholes in costumes the same way I made them for everyday clothing – using the buttonhole feature on my sewing machine. My mother and Home Economics teacher had both taught me how to do them by hand, but since I never made anything haute couture enough to warrant such attention to detail, I merrily got by with machine-made buttonholes.
Until I made a gardecorpes for my husband. Suddenly, I had a desire to make buttons authentic to the garment, but had little to go on. So, I made the kind I’d learned how to as a girl – mark the slit, surround it with double running stitch, cut the slit, and put down retentively perfect stitches, laid neatly side by side.
Later, I acquired Arnold’s Patterns of Fashion and Crowfoot, Pritchard and Staniland’s Medieval Finds From Excavations in London: 4 Textiles and Clothing, c.1150-1450 and the scales fell off my eyes. To begin with, prior to the 16th century, buttonholes were very simple, being just a slash with a few buttonhole stitches worked down either side, with nothing done to strengthen the ends of the slits (Crowfoot, p.171). The assumption is that 16th century clothing, being more closely fitted and tailored, resulted in more stress upon the button closures, and required reinforcement. My own theory is that the seam allowances turned under along the opening of a lined garment, or the fold-over hemmed along the edge, served as a stop-cut, making an end bar unnecessary, and that bars in the 16th century are largely decorative.
Even in the 16th century, reinforcing stitches across the ends of the slits was not universal. In Arnold, there are examples of:
Bars only across end of the slit facing the garment’s opening, (Arnold, p.23, fig. 148)
Bars across both ends of slit (ibid., p.20, fig.120),
Bars across neither end of slit (ibid., p.27, fig. 178)
In several instances, a decorative braid running down either or both sides of the buttonhole placket might be covering bars, or acting in lieu of them.
Tailors didn’t appear to have knocked themselves out making the buttonholes a work of art. Crowfoot, describing medieval London finds, notes that buttonholes cut from sleeve openings and hood fastenings are made up of stitches set about 0.5mm apart, rather than being “packed closely together to form a solid band as is normal today” (Crowfoot, p.171). Arnold gives an example of a red doublet, c.1560, with buttonholes worked in a packed satin stitch (Arnold, p.20, fig.120); but she also shows us a German doublet, c.1605, on which the buttonhole stitches gap about 1 thread’s thickness apart (ibid., p.20, fig. 123).
Buttonholes varied in length, generally depending upon the size of the button. Medieval examples range from 7-10mm (5/16”-3/8”) on wrists to 12-14mm (7/16”-9/16”) along body closures (Crowfoot, p.171) Sixteenth and seventeenth century doublets might have buttonholes as long as ¾” (Arnold, p.20). Buttons, however, might be disproportionate to the size of the buttonhole. Arnold gives an example (ibid., p.20, fig. 123) in which the button diameter is only half the width of the buttonhole, which might indicate the buttons were made up at a different time or by a different tradesman.
The depth of stitching wasn’t terribly deep. They could be as short as 1mm deep (Crowfoot, p.171), though there are also examples of longer stitches. Arnold gives little detail on buttonholes save the length of the slit (and that only in some cases); from this, however, I extrapolated that the stitches range in length from 1/8” (Arnold, fig. 84) to 3/16” (ibid., fig. 120).
Medieval and renaissance tailors set buttonholes much closer to the edge of the garment than we do today. A slash ending as little as 1/8” from the edge of a garment may look alarming to us, but as noted earlier, 14th-15th century buttonholes did not take a lot of stress, and the layers of turned seam allowances of the 16th century acted as a stop-cut for the button shank. In the 14th century, a length of card-woven braid often was worked directly onto the edge of the fabric on sleeve cuffs, and the buttonholes cut right up to the edge of the braid. Since worn-out sleeve cuffs pitched into rubbish heaps have the braid still intact, this method worked satisfactorily.
On medieval garments, button plackets on unlined garments generally had a facing, usually a band of fine silk, linen, or possibly a cotton fabric called “bokeram” (Crowfoot, pp. 160-1). Today, manufacturers of better-quality cardigan sweaters use the same technique by stitching a strip of grosgrain down the back side of the opening.
For the most part men’s clothing, particularly in the 16th century, lapped the left side over the right, as is done today, but this was not universal. There are too few examples of women’s buttoned clothing to draw any solid conclusions as to how feminine garments lapped.
No one has as yet - that I’m aware of - picked apart a period buttonhole to find out if reinforcing stitches were used, but Crowfoot specifically mentions that there is “No evidence of a circuit of running-stitches around the hole to hold the two layers together and to strengthen the vulnerable slits” on 14th-15th century garments (Crowfoot, p. 170).
Buttonholes were generally stitched in a heavy, tightly-twisted silk thread, much like the silk buttonhole twist tailors use today. Silk sewing thread was widely available throughout most of Europe from at least 1300 on, though until the latter part of the 14th century linen seems to have been used for construction seams, and silk reserved for decorative and other visible stitching. Linen thread was sometimes used for buttonholes, as woolen archeological finds with only stitch holes remaining were probably worked in linen, which does not survive in acidic conditions found in most European soils. Hemp thread may have been used, but linen and hemp threads look virtually identical, and only recently have the experts started to differentiate between the two fibers. Cotton was rarely used, even the latter part of our period.
Both ordinary buttonhole stitch and tailor’s buttonhole stitch (sometimes called “twisted” buttonhole stitch) were used in period. Tailor’s buttonhole stitch, made by inserting the needle tip through the slit and bringing out through the fabric, creates a tiny knot at the top of the stitch, making a more abrasion-resistant edge. There does not seem to have been a preference for one over the other. Also, there is no standard whether the stitching is worked right-to-left or left-to-right.
I haven’t yet come across evidence for welted buttonholes up through the 16th century, but will gladly update this article to include them if any come to light.
Making up the buttonhole:
Period buttons were cut open before they were stitched. Substances such as wax and fish glue were known to have been used to treat the cut edges of fabrics to prevent fraying (Arnold, p.17, fig. 100). The technique is called “cereing”, from “cere” meaning to “wax.” Think of it as historical Fray-check.
Although Crowfoot notes a specific lack of reinforcing stitches around the slit, I frame the slit with a row of tiny double-running stitches. This fastens the various layers together and gives me a stitching guide.
The examples in the instructions are stitched with a doubled length of unwaxed linen sewing thread on a fulled, skirt-weight woolen twill.
(Click thumbnails for larger images)
Tailor’s buttonhole stitch
The top buttonhole has bars worked across the short ends, based on examples from Arnold.
Grosgrain ribbon backing of buttonhole placket. Ribbon is permanently stitched to the hem and basted down the other side. The basting is removed after all buttonholes are completed.
Do not worry if the stitches don’t cover the back of the buttonhole as neatly as they do the front. Period examples have messy back sides to them.
Below are pictures of some buttonholes I have made for various garments. As you can see, some came out tidier than others, but you find the same variation of neatness in examples from the period.
Front opening of a man’s 16th c. jerkin, worked in waxed bookbinder’s linen thread. I was deliberately fussy about stitch placement, as I made the garment to be reversible.
Front opening of a man’s cotte, (Herjolfsnes No. 63), worked in dark blue worsted with a linen lining. I worked a few spoke stitches around the outside end of the slit because I was worried about how the wool might ravel. It turned out to be unnecessary.
Doublet based on pattern from Arnold, taken from garments worn by Cosimo diMedici. After several years of wear, the linen twill lining has frayed a bit along the cut edges. So have similar examples depicted in Arnold’s book. I did not make reinforcing stitches around the slit. Structurally, they were not necessary, though they would have helped me make all my stitches the same length.