Wednesday, August 15, 2018

It's All in the Details: The Mystery of the 101st Row...

Hey all! Back from my hiatus. Saber Seminar was a lot of fun, and my talk--now posted on YouTube--was well-received. (Can't ask for more than that!) Now back to the Ty Cobb Sweater...

You may recall that, a couple posts back, I described the sweater's construction as "elegant." The sleeves are another example of that. They're also one of the better (and more subtle) demonstrations of the fact that this sweater was factory-made.

Picked-up stitches at top of sleeve
Unlike modern sweaters, one of the goals appears to be to minimize the number of seams (something I discussed when I talked about the shoulders). In fact, so important is that goal that gauge is occasionally sacrificed; put another way, the makers preferred to pull at the stitches and stretch the fabric rather than allowing it to lie flat.

Take a look at the sleeves. Rather than attaching them afterwards, every other stitch at the end of each row along the sleeve opening--yes, we're now knitting perpendicular to the body--is picked up and becomes a new stitch.
"Stretched" sleeve sts at shoulder (HOF Sweater)
Since the number of rows is less than twice the number of stitches for our gauge (15 sts and 22 rows/4", remember?), that means the stitches "spread out" at the shoulder and only gradually return to their original gauge.

However, the vertical body stitches don't make up the entire sleeve. You may have noticed in some of the body pictures that there were ten loose stitches at the bottom of each sleeve opening. This is an unusual design element (I'd certainly never seen it before), and all I can figure is that the manufacturer meant to simulate a "gusset." Until we started making armholes that curved in the same shape as sewn clothing, knitted gussets were common. A diamond would be inserted at the underarm, minimizing bunching and allowing the sweater to hang more naturally. While those ten loose stitches aren't an inserted diamond, they appear to serve the same purpose as a gusset, making for a more comfortable sweater (and incidentally, an easier knit).

Example of a typical diamond gusset
Sleeve opening (note loose sts at the bottom)


















When working out the construction of the sleeve, I noticed something interesting that, once again, points toward mass production. The sleeves are exactly 100 rows long (actually, 101 rows long, but I'll explain why shortly), each of the decreases is five rows apart, the central portions are fifty stitches around, and the cuffs have twenty-five rows each. Needless to say, this struck me as odd, especially when you consider that the decreases are done in two sets--one near the shoulder and one near the cuff--rather than smoothly along the length of the sleeve. (Note: That splitting up of decreases was also new to me.)

Sleeve (reproduction) - note the two sets of decreases
All I can figure is that working in multiples of five and ten it made it easier for the workers to keep track. It also might explain why the sleeves were so long that they needed to be rolled up.

And now we come to one of my bigger mistakes. Having reached the end of the sleeves, I proceeded gamely on to the ribbed cuffs. No, they didn't really look like the ones on the original HOF Sweater, but I knew from our only picture that Ty Cobb had rolled up his sleeves, and even that there was a run in one of the cuffs. I figured that maybe wear-and-tear had stretched out the cuffs, leaving them shorter and with a tighter gauge.

Cuff - HOF Sweater (Note the unraveling...)
When will I listen to my own explanations?!?! After all of my earlier concerns with gauge, I of all people should know that wool simply doesn't do that. Sure, if you wash wool in hot water, it will shrink (and mostly vertically), but there was no sign of that here, and the rest of the sweater was pristine. Not only that, but when I looked at pictures of some of the other Hall-of-Fame sweaters, the cuffs were identical. Clearly I had screwed up. The question was: HOW?

So, back to the pictures I went...where I discovered something very strange. [insert Twilight Zone music] Looking very carefully (and this is nit-picky enough that I won't bore you with visual examples), I could tell that the ribbed cuffs (a) had a much tighter gauge and (b) had clearly been worked from the bottom up. That made absolutely no sense. Clearly the sleeves had been worked from the shoulders down, meaning the only way to make those cuffs  would be to start them separately (on much smaller needles) and then attach them using a Kitchener Stitch (such as we saw with the shoulders). Who the %&*! would do that?!?!

And then I remembered that 101st row... And the light dawned. [insert angelic choir]

Cuffs (reproduction) - Left cuff is correct; Right cuff is WRONG
Remember that this sweater was made in a factory, meaning that one garment was probably made by multiple people. It's unlikely that whoever was knitting the sleeve would switch to smaller needles for the cuffs. Rather, it was probably quicker and more cost-effective to have somebody else make cuffs and only cuffs, which would then be attached to the finished sleeve. Since the attachment clearly used a Kitchener Stitch, that would explain the 101st row. It would also account for the placement of the ladder on the original sweater, since the obvious place to start/finish grafting would be at the center back of the cuff (where the decreases are). If the Kitchener Stitch row hadn't been finished properly, one end might have come loose, which would explain why the ladder occurred there (and also why the edge of the cuff itself was still intact).

So, what to do? If I was going to be truly historically accurate, the obvious thing would be to make the cuffs separately and attach them. Here, however, I took a liberty. After all, part of the point of this project is to create a pattern that can be made at home by an average knitter. Asking somebody else to make cuffs in this way--rather than simply switching needle sizes and continuing to the end--seemed to be design ad absurdum. Therefore, for my own sanity and everybody else's, I chose to do the logical thing, rather than the historically-accurate one, and work the cuffs as a seamless part of the sleeve. Was it the right decision? I don't know. However, if it took me that long to figure out the small oddity in cuff construction, hopefully nobody else will even notice. (And yes, if enough people really care, I'm happy to include two versions of the sleeve construction in the final pattern.)