The Patent and Initial Observations

Henry T. Marsh filed this patent in January of 1878, claiming two “improvements” to corsets. His first improvement is the pattern on Sheet 2 of his patent. The second is the use of a non-traditional material to replace the fabric in parts of the corset, namely kid leather.

Right off the bat, I'm somewhat sceptical of Marsh's first claim. He states that by using his pattern, “the resulting structure shall fit and adapt itself to the body of the wearer” but gives absolutely no further explanation of what makes these pattern pieces different or how they work. The pieces do not seem particularly novel or unique; many corsets of the era and much earlier boast extremely similar bust gussets and hip gores (see this example from Symington's), and even the curvy front pieces seem common enough (see this example from the Met). I don't know that his pattern is very original, but I'm game to give it a try!

As for his second claim, Marsh tells us “it is well settled by long experience that cloth (cotton, linen, or silk of stout texture) is the best material for the body of a corset. It is also well settled that there are certain places in corsets which wear out long before the remainder of the corset is substantially worn.” His remedy for this is to use kid leather or “other fine leather dressed to the same condition” in place of fabric at these points of wear.

What kind of wear is sending corsets to an early retirement is not specified, but we might gather some clues from the fact that he has placed the leather around the busk, the grommet area, and at the sides. I'll hazard a guess that the wear issues he's attempting to prevent are grommets pulling out, seams splitting at the point of greatest pressure, and abrasion of the fabric from laces or waistbands. I'm intrigued by his idea to use leather at these points, but not sure if it will actually help. I guess I'll have to test it out!

What he doesn't specify:

• He is unconcerned with the type of boning or busk material.
• Number of bones/placement is not mentioned.
• Single layer/double layer is not specified.
• It's left up to the maker to decide how to join the pieces.
• Anything regarding the strange strips at the top of the bust in Fig 1 of his drawings. They aren't shown in the pattern, nor are they mentioned in the description.

 

The Plan

I'll scale up the pattern and make a paper mockup to see how the pieces fit and go together. From there, I'll adapt the pattern to my measurements and make a mockup (or two, who knows!) to check fit. Along the way I will test various types of seams to find the best way to join the fabric and leather pieces. I'll also stress test leather vs. fabric in a variety of scenarios to find out if there's any merit to Marsh's claims. Finally, I'll get around to making the actual corset.

 

The Pattern

First off, I've got to get this pattern scaled up and in wearable condition. I've chosen to import the image of the pieces into CorelDRAW and trace over them so that I can make a digital pattern. I'm not going to go into a great amount of detail about how to trace the drawing, as Mark Garbarczyk has already written a fabulous article on how to tackle this task.

 

After tracing everything, I've scaled the pattern up a bit so that it's larger and easier to work with. I'm not concerned with getting it to proper human size yet; I just want pieces that aren't super tiny. I also put everything in order, since some parts are drawn out of place (looking at you, busk piece).

Right off the bat I've run into some issues. The patent tells us that pieces 5 and 6 are meant to be doubled over (folded in half), thus they should be symmetrical on either side. Unfortunately they're anything but!

 

It's a simple fix at least. Divide the pieces down the center, figure out which side that looks most correct and cut off the bad side. Copy & paste the remaining side, then mirror it (flip it to face the opposite direction) and snap it into place.

 

In addition to the wonky bust and back, some of the pieces are drawn tilted (look at 4 and 5, their edges should be parallel relative to each other) so I use a little guesswork to orient them more vertically.

 

Now I've got to decide whether the pattern already includes seam allowance. Luckily, it's actually really obvious this time. If there were no seam allowance, the bottom edges of the pieces would flow smoothly from one piece to the next; instead there are weird little upturned bits at the edges of several pieces. This is evidence that seam allowance is already included. I'll need to remove the seam allowance to accurately measure the pieces.

 

The upturned bits are all slightly different sizes; not really surprising given the inaccuracies so far. At the pattern's current scale (remember, not human sized yet) those little bits measure out to an average allowance of ¼”, so I need to remove that much to find the correct stitching line.

I often use the Contour tool to add seam allowance when making patterns in CorelDRAW (also covered in Mark's article); luckily the same tool can be used to remove it as well by placing the contour inside the shape rather than outside. The downside is that it also takes the seam allowance off the top and bottom, which is unnecessary in corsetry. For now I'm just going to leave it there, but eventually I'll have to correct it and extend the stitch line all the way to the top and bottom edges.

 

I've guessed that the waistline hits right where the hip gores meet pieces 2 and 4. For now, I've extended that line straight across the rest of the pieces and hope that it's correct.

 

At this point the pattern has not been altered other than to scale it up, fix the asymmetrical back and busk pieces, slightly tilt some pieces, and add the stitching line and waistline. Now's a good time to make a paper mockup to see how it all goes together!

Right away it's apparent that this is going to be fairly curvy up top. It's also obvious that the pattern pieces aren't trued up; in some places the top and bottom edges match, while in others they're way off. I can see that the hip gores are both too short.

 

The most important discovery was that I had placed piece 3 too low in relation to the others, so when I extended the waistline across it ended up in the wrong place. I moved piece 3 up and its waistline down to correct for this and the piece fit much better, though not perfectly.

 

I also can now draw the bust and hip lines directly on the paper model. I find it's harder to guess where these go on a flat drawing; having a three-dimensional model makes it a lot easier to find out where these lines go.

 

Now that I've got more info, it's back to the digital pattern. Based on the paper model, I've moved the waistline on piece 3. I've also copied the bust and hip lines onto the digital pattern. I've also removed the seam allowances for now; I want to measure the actual pieces, and the SA just gets in the way of that.

 

There are several methods to scale a pattern up to your relative size. I'm going to simply enlarge it until the waist is as big as my half my desired waist - 1 inch (for back gap).

To do this, I snap all the pieces together edge to edge at the waist. I measure the size of the waist by using the Horizontal Dimension Tool to snap a measuring line along the waistline from the center of the busk piece to the center of the back piece. I almost made the mistake of measuring to the edges of those pieces, but since they're intended to be folded in half that would have been wrong! I then scale the whole pattern up until the dimension line reads my desired result of 12”.

 

Now that the waist is the correct size, I can measure the bust and the hip and compare those measurements to my own. This is a pretty rough method, and often gives comical results. However, I'm surprised to find that this time the measurements of the bust and hip end up very close to mine! Rather than adjust the pattern further, I decide that I should make a mock up. Since I've done very minimal alterations to it at this point, this is a chance to see how Marsh's pattern fits a real live person. I make a new copy of the current pattern, add seam allowance back in, and I'm ready to cut.

 

Cutting

I've never been a fan of cutting out corset pieces. There are always so many, and accuracy is super important. Luckily, I have a laser cutter at work. What would take me two hours to cut by hand takes less than five minutes. You can see how fast it is in the video below!

 

I cut the fabric pieces out of a cheap coutil, and the leather pieces out of blue canvas. I'll use leather in the final corset, but for now I just want to be able to get a sense of what the corset will look like with two different materials.

 

I was so focused on measuring the waist, bust and hips that I didn't pay attention to the vertical measurements. As I was watching the laser cut, I realized that these pieces are super long! The center front alone is more than 15” in length. I'm very long waisted, but I hazard a guess that I'll need to shorten everything on the next go around.

 

Making The Mockup

Because piece 6 is meant to be folded in half, I have to cut slits for the busk hooks to fit through. This results in fraying fabric, but shouldn't be an issue in the final corset as this piece is supposed to be leather.

 

I usually use detachable lacing strips on my mockups, but my current pair is worn out and in need of replacing. Not wanting to waste time and materials grommeting a mockup, it occurred to me that since this is just meant to be worn briefly I could simply punch holes in the fabric and forgo the grommets entirely. The bones along either side keep the holes stable and prevent them from ripping through the edge of the fabric. Sure enough, I was able to lace into the mockup multiple times and wear it with no issues whatsoever.

 

Marsh offhandedly mentions any type of boning can be used. He doesn't mention where to apply it. I experiment with a mix of spiral and flat white steels, with the flats mostly at CB and CF. On one side, I place boning channels running vertically down the length of the corset. On the other side I let the boning channels follow the curves of the pattern pieces.

 

The Result

Color me surprised, it fits and it's pretty comfortable!

 

The high back is great and contains any overspill. The fit in the torso is snug and just a little bit curvy, allowing my ribs some space.

 

The bust is nice and curvy on my left side, but somehow ended up much flatter on my right. I could use a little more practice inserting gussets!

 

As expected, it is a bit too long in front, but only from the waist down. I don't feel I need to shorten it all the way around; instead I think I can just take the extra straight off the bottom front. Unexpectedly, above the waist it's actually too short in the torso! In the above images, the waist is sitting just a little too high for perfect comfort. Pulling the whole corset down an inch solved that issue, but left the bust too low. I'll have to add some length up top for the final corset.

 

The hips fit well enough, but I was expecting a little more curve. I'm also betting that I can get close the waist just a little more if I add bit more hip spring. To get an idea of how much I might be able to add, I slit the seams on either side of piece 3 up to the waistline. The resulting gap showed me that I could probably get away with adding a little less than an inch to each gore.

 

Fixing the Pattern

To add that extra length into the top, I cut across all the pieces a little bit above the waist and move the upper parts up one inch. I then redrew the lines to smoothly connect the two pieces.

I also added width to each hip gore, and just for the heck of turned the straight sides of each gore into a bit of a curve. Finally, I cut nearly two inches off of the center front but left the side and back the same length. The image below shows the mockup pattern in green, with my new pattern in black. I'm confident enough with these alterations that I'm not going to make a second mockup.

 

Next up: I'll test Marsh's claims about leather, plus make the final corset!

 

 

add