Once I had completed my analysis of stays from 1790-1820, it was then time to make a pair for myself. It's fair to say that I spent more time researching patterns from various sources than I did actually attempting to make anything. I found free patterns online, scaled patterns in books and I bought all the commercial patterns that I could find.

One of the free online scaled patterns looked very appealing so I decided to try it out. The stays are listed as being formerly in the collection of the Ohio Historic Society, and the pattern and photos are online at the 19th US Regiment of Infantry living history organization website.  I decided to first make a sample in its original size, which would help me assess how it might work for my body type without too much effort.

There are a couple of notable differences in the Society's description and the description/photos on the re-enactment group's website. First, the fabric is described as linen by the OHS and cotton by the 19thUS. The OHS describes the busk as 2" wide and doesn't mention the bone grommets that I would expect them to mention. To complicate things, the 19thUS stays photos are a mixup of two different garments. One has the shoulder strap sewn to both front and back, hand sewn eyelets and a wide busk pocket. The other garment (the one that is patterned) has a narrow busk, bone eyelets and the shoulder strap ties at the front. I could speculate whether the wrong Collection Number is written on the scale pattern and webpage and this is where it would have been so helpful if the OHS had photos of the garment they owned under this Collection Number.

The Original Size Sample

These stays are a winner from my perspective. They were fairly quick to sew together and even the cording didn't take as long as expected. I scaled up the pattern using the Paint program, and added seam allowances to all the pieces except at the top and hem edges. I double checked that the seamlines would match in length and the front piece needed adjusting at the hem edge on the underbust seam. I also misread the pattern and didn't print out the photos to refer to when cutting out the pattern pieces, so I didn't grasp that the busk width was not 2¼ inches (5.6cm) wide but instead 1 1/8th inch wide (2.8cm). Once that was together I did learn one thing ... make the opening for the busk first!

I didn't expect the resulting garment to be as curvy as it turned out to be: it's clear that the original wearer had an exaggerated hourglass figure. I'm no longer that slim, and I'm not quite as extreme an hourglass as the figure that this garment was made for, but I'm not as far from it as I'd prefer. I would love to see the clothing that this woman owned. It would be great to know what someone with this figure type really did wear back then.

The finished measurements of the stays made to the scale pattern (had I used a 2.8cm busk) are:

These are slightly larger than the pattern would indicate, as the scaling method I used resulted in the seamlines drawn 1/8" (3mm) wide. I added the seam allowance to the outside edge, and over the eight seams that's potentially the 2" difference.

Regardless of that, assuming the back had an even gap from top to bottom, there is a significant 16" (40cm) difference between bust and waist. The OHS website had these listed as c1830 and the website that provides the original garment photos and the scaled pattern has it as c1800. Without knowing what each site based their date on, one of the possible reasons for the OHS date is the depth of the bust gusset, as they did become deeper (not wider) in the 1830s. However, one thing that comes with larger breast volume is that the base diameter is larger too, and with that the lowest edge of the breast is closer to the waist than in those with less volume. With more than 15" difference, it's safe to say that the gusset was deeper simply because it had to be. With these stays, the overall design and the fact that they're designed for such an extreme figure, I don't have such a strong opinion on their date. I think 1800-1810 is possibly too early, but I could be convinced with specific information.

I photographed them on a store mannequin I own as I don't know anyone living near me that would fit the stays. With a 70cm waist measurement the mannequin is still too big, so there is a 12cm or 4.75" gap down the back and the shoulder strap isn't at the back, but near the underarm. I used almost all the shoulder pads from my sewing stash to pad out the bust & hip area. In the first photo, the left side isn't padded.

Personalising the pattern

The pattern sheet has four main pattern pieces and two minor. The pattern sheet only names two of the main pieces, the Centre Front (the busk pocket) and Front, so I've named the other two Side and Back. The two minor pieces are the Bust Gusset and the Shoulder Strap.

As well as grading up the garment to fit me, I wanted to make a couple of minor style changes:

• Firstly I want to use light boning instead of rows of cording at the centre back.
• Secondly, I also used a heavier cord – I just didn't think 1/16th inch piping would be the best for my more padded figure.
• Third, I also wanted to extend the horizontal band of cording around the body, to the side back seam, and not just further around the body but also deeper, right up to underneath my bust. 
• The fourth and last change was to make the shoulder straps a little narrower, and to bind them separately so they would be attached after the top edge had also been bound. I also reverted to the original size busk, just over 1" wide, so although it's different from how I'd made the sample, it's not really a change to the original pattern. From these changes, the only pattern pieces to alter were the shoulder strap and correcting the Centre Front.

The next step was to grade the pattern to fit me. As evenly grading the original size wouldn't result in a match with any two of my bust, waist or hip measurements I had to choose which single measurement to grade to, and then alter the others (known as an "uneven grade"). At this point I was less concerned with a precise bust measurement, as I could alter the gusset size to fit if needed, so I graded to the waist measurement and altered the bust and hip.

These are the original pattern pieces (below left) and my altered pattern pieces (below right). The areas with purple cross hatching were removed.

Fabrics used

The outer layer is firm white cotton, known as calico in Australia and the UK, but I think as muslin in the States, and for the lining I picked a cotton print from my stash. This was partly so that I didn't muddle the pieces up as I assembled them, as the front piece is easy to turn upside down or back to front. As I'm machine sewing every step it's clear that I'm not aiming for complete authenticity, and to be honest I don't really like white underwear even when it's period.

Construction Steps

Decide where the busk opening is going to be. If you need to make a buttonhole opening on the lining layer, do it now before any front seams are sewn.
Place Centre Front and Centre Front lining with the wrong sides together and baste near the edges.
Sandwich the Centre Front pieces between the Front & Front lining pieces (careful to make sure they were facing the right directions).
Press the seams flat. This method means that there are no seam allowances inside the busk pocket, which I felt was an improvement on what I'd done with the sample size.
Measure where the gusset will need to begin on the underbust seam and leave the upper part of that seam open when attaching the side panel pieces. Repeat on the lining. Press the seams open.
Set the assembled pieces aside and turn to the back & back lining. Sew the centre back seams, turned so that the wrong sides face each other, press the centre back flat and edgestitch the centre back.
Before I went any further I wanted to see how this fitted on me, so I basted the front/side section and back together at the side back seam to compare the stays to myself. As a side point, I often use my digital camera as a fitting assistant. For me, it's much easier to be objective about how something fits if I look at a photo on a computer screen, compared to looking at myself in a mirror. So I've learned to take photos and look at them when I'm fitting a pattern I've never used before. It's saved me from making a few mistakes. As I don't have lacing strips that I could attach and remove, I did the much less scientific method of wrapping the stays around me and holding it at the back while taking a timer-delay photo. It was obvious that the gap was too wide before I looked at the photo. As I live alone there isn't the option of asking someone else to measure the gap for me.
So to work out how wide it was, I turned the stays back to front around my body and measured the gap at the front waist. I was shocked to see that it was almost 6" (15cm). Now I want a 3" (7.5cm) gap as I've been slowly losing weight and that slightly wide gap means it would fit me a little longer. Before I continued, I wanted to figure out why there was such a difference. It wasn't going to change the result in my hands but I wanted to know where things had gone wrong. This is the point when I realised the narrower busk hadn't been compensated for when I graded the pattern up. That took into account 1" (2.5cm) of the problem. The seam allowances I'd used had been a smidge on the generous side compared to those I drafted, and over the 8 seams that was approximately another 1" (2.5cm). What I should have done is measured myself to see if there was a difference between my measurements that day and the ones I graded to. I do sometimes get rapid fluid retention (3kg or 6.5lb) over a few of days and at other times fairly bad bloat that affects how waistbands fit me. With hindsight one of these was a contributing factor.
So how could I fix my perceived 6" gap? My three options were to: (1) Let out all the seams. However, the underbust seam was perfectly positioned and letting out the seams would move it 0.5" (1.2cm) closer to the underarm. (2) Re-cut wider side panels. (3) Insert another panel. Something I considered is that one of my body's non-standard features is that my torso is deeper front to back than average, and this helped me decide what to do. I unpicked the basted side back seams and inserted a strip at each the side back seam (finished width of 1.5" each). This would provide the extra fabric where my body needs it and piecing is period, so I wasn't too distressed at taking this option.
Sew the side back seams and press them open, and turn stays right side out. Pin the front to the lining along the matching seamlines.
Stitch in the ditch on the side back seams, and from below the waist on the underbust seam (so it doesn't interfere with inserting the gusset).
The next fitting step was the bust, and already I thought with the shaping on the front and side pieces the gusset would be too big. Again my fitting method was less than scientific and ultimately flawed. I pinned the stays in place onto a tank top and took a photo to see how much of a gap there was. It wasn't a large gap (at least not compared to the volume underneath), so I decided to leave out the gusset entirely and sew a narrower seam instead - an example of hubris as I should have measured the gap. Just because it wasn't large compared to total volume didn't mean that I should have skipped such a basic step.
If you don't have a large volume and still want to use the gusset, you can take out some of the volume from the front/side pieces and put a gusset back into the seam. If I were to insert a gusset I would have basted the gusset piece and lining together, then sew the gusset into the lining seam. I would then place the outer layer into place on top and edge stitch around the gusset. Here is a photo of the original size with a single layer gusset. For myself I would prefer to use a double layer.
Now that I had the stays as one piece I moved on to the cording, boning and binding. I settled at the machine to sew all the channel seams for the cording. Again I simply marked the start and end points with pins & just aimed for the finishing point. There wasn't any fussing with marking lines or basting. I didn't do that on the sample set and they turned out perfectly fine. After sewing what felt like endless rows at the waist I did wonder why I'd added to the amount of cording work by extending the waist rows around & up the body. I'm really hoping that it act as I think it will, as a supportive countermeasure to the bust above it. I could have sewn the channels and only inserted the extra cording after I'd finished and worn the stays to see if there is a difference, however this didn't occur to me until a couple of days after the cording had been done.
The stores near me were out of the piping cord size I'd used in the sample pair and the next size up had a coarser different texture after I'd pre-shrunk it with boiling water. This would make it harder to 'sew' through the fabric and not suitable. Not easily defeated, I went looking for knitting cotton to use as a substitute. The yarn I bought is a bamboo & cotton blend, and its 18 strands aren't twisted together and so easily separate. This would be irritating if I was knitting with it, but as cord filler it wasn't an issue. I bought a couple of different needle types to test (tapestry and bodkins) and the first one I tried & continued to use is a size 17 bodkin needle.

To see how long it actually took, I timed sewing the cording into all the channels: 2.5 hours. I only used 60cm lengths of yarn at a time so that the yarn wouldn't fray from being forced though the fabric too many times.

I clipped the yarn as close to the fabric as I could and then tugged the fabric along the length of the channel (without putting any pressure on the yarn ends) and with that little stretch the yarn disappeared into the stays. I repeated with every channel and did all the cording in one sitting. I'm not sure if it was the TV distraction or the desire to finish, but it wasn't until I'd finished that I noticed that my right thumb was red and a little swollen. It took another 24-36 hours before I could easily use it again and another couple of days after that to not be tender so my recommendation would be to break it up in to three or four sessions, a day or so apart.

All the cording was now complete apart from the rows I would need below the busk. I should have sewn the boning channels at this point. As I see this as a washable garment I didn't want to use metal bones in the centre back and the plastic boning I could find was either useless for this purpose (rigilene) or wider than I wanted to use. I ended up using narrow cable ties, about 5mm wide, or just under 0.25".
I bound the top, bottom edge and shoulder strap with commercial poly-cotton bias binding. I did find some cotton bias tape but it was a bit too stiff and bulky for my liking. Possibly I'm just being lazy about cutting bias binding strips, and yes I did machine sew the binding on ... tennis elbow is my excuse to get out of hand sewing as that aggravates it.
The original stays had 12 bone eyelets on each side of the back, but I did 11 hand sewn eyelets. Not the first hand sewn eyelets I've ever done but close to it. I no longer hate doing them with a passion, but they did take just as long to sew as inserting the cording took. The pattern sheet doesn't indicate if this was spiral or cross laced, I'm assuming spiral although I placed mine for cross lacing. I'm planning to work out how to arrange fan lacing in the same way as the 1830s stays in the book Fashioning Fashion (pages 86 & 87). I'm not sure how much experimentation it's going to take for fan lacing to be workable, but until then I'll need a lady's maid or friend to dress me.
Shoulder straps: baste the outer and lining piece together and bind all edges. Repeat for the other strap. Sew eyelet where marked on pattern.
Hand sew them to the back shoulder with a small close overcast stitch on the lining side.
As for the busk I used a 1" or 2.5cm wide wooden ruler. As it was a couple of inches shorter than the centre front pocket I filled the bottom section of the busk pocket with rows of vertical cording as per the stays in c1835 stays on Abiti Antichi. I ought to have sewn the channels before the binding was done and inserted the cording once the binding was done.

Conclusion

A friend to laced me into my completed stays. I'm very happy with them, although I plan to remake them fairly soon with a couple of changes.

I leapt to conclusions about the stays being 3" (7.5cm) too small, when in fact they were only about 1.5" (4cm) too small. However, given my body shape I need to take the extra width from the centre back edge and leave the extra width in the side panel. You can see in the photo of my back that the shoulder strap is as close to my underarm as is comfortable. The bones buckle at centre waist so either I need to remove a small horizontal wedge (about 1.5cm at centre back tapering to zero at the side) as am a little swaybacked and this is an alternation I do when making all my modern clothing or I get the same wrinkle. However I've seen bucking in garments that have cording and not boning.

I should definitely have inserted a gusset as I'm quite flattened without it. Perhaps only 1-1.5" (2.-4cm) at the top edge rather than the original 2.5" (6.3cm).

The shoulder straps are too short by around 3" (8cm) though I can switch these out due to my construction method.

Despite planning changes, these stays work well enough that I don't need to remake them immediately. I'm looking forward to using them at a re-enactment event later in the year.