Self stripers, multi-ply color twist yarns, and hand-dyed yarns are lots of fun. There’s a ton of color there to enjoy. But sometimes a vast and unbroken expanse of color play can be boring or can produce some oddly pooled or cluttered results that look nowhere as nice as the yarn did in the skein. That’s why I like to fool around with multicolor yarns, trying to find a way that they play best, or are shown to better advantage. Sometimes it’s not easy (regular readers here know of at least two of my multicolor yarn experiments have landed in the limbo of my Chest of Knitting Horrorstm over the past couple years. But sometimes it is easy.
I finished up a pair of socks as a “chaser” after the vest was done. I used a stash-aged 50g ball of Stahl Wolle Socka Color, in a multistrand twist of maroon, pine, blue, and marigold – #9140, plus another well-aged 50g ball of Patons Kroy Socks in hunter green – #409. While the colors aren’t exact matches, they are close enough to complement each other. Both of these were found in last-ball sales, but several years and many miles apart. Since I need around 80-90g of most fingering weight sock yarns to make a pair, between the two bargain basement balls I had enough to finish and still have leftovers – provided I used more or less equal amounts of each.
My solution was to work heels, toes, and ribbing, plus about a little under a third of the sock’s body in my solid green, plus the remainder in the multicolor. I used very simple seven-row striping repeat, working five rows of multi, and then two rows of solid green (2/7 = about 28% of my sock’s body). I like how the multi is visually broken by the bands of solid green. The end result has at once more contrast and more subtlety than working the whole sock from multi alone, even if I still did contrasting color toes, heels and ribbing.
Of course the other advantage of working simple stripes on socks is to idiot-proof achieving two socks of identical size. It’s very easy to count five row units and two row units. When I had completed ten muticolored stripes, it was time to start the heel, which is much easier than having to count every row or trust in doing a measured or eyeballed estimation of foot length.
So. If you find yourself with odd lots of sock yarn, don’t despair. 100g of fingering weight sock yarn knitting is ample for most socks up to around men’s US shoe size 10.5 or so (slightly smaller if yarn-eating textures are used). You can either work color block style, using up one leftover and then another, or you can stripe. But how wide to make the stripes?
Heels, toes, and ribbing in my standard short-rowed heel sock consume about 25g (a conservative estimate). I have large and wide feet for a fem, so if you are knitting for yourself chances are that you use roughly what I do or less. Weigh it out and set it aside. Then weigh the rest of your leftovers. If you have (for example) 40g of blue, 20g of yellow and 20g of green, you’ve got a ratio of 2:1:1. If you worked a stripe repeat conforming to that ratio (let’s say two rows of blue, two rows of yellow, two rows of blue, two rows of green), you should have enough of each color to complete the pair.
Obviously, I had enough and did complete my pair. And I did have leftovers. As expected, I had a bit more of the multi left than I did the green, because my heels, toes, and narrow green stripes added up to about 60% of total yarn consumption.
Techknitting is posting an interesting series on stranding, and as part of it, mentioned the use of Strickfingerhuts (knitting yarn guides/knitting thimbles), linking back to my original post on the subject.
For those who are unfamiliar with them, they are those gizmos that sit on the end of the left hand index finger, that are used by Continental style knitters (pickers) to hold and separate two or more yarns while doing stranded colorwork.
Adding some more detail on the subject, I’d like to address a problem TK points out as being common among those who hold two yarns in one hand while stranding – differential feed.
If a row has more or less equal numbers of stitches of both colors, both yarn strands are consumed at the same rate. But if a row has lots of Color A, but very little Color B, A will be eaten at a much greater rate, eventually causing the knitter to readjust his or her grasp of the yarn to even things out.
Those of us who do use Strickfingerhuts find that the differential feed rate problem is greatly minimized compared to trying to hold both yarns in the left hand unassisted. Yes, eventually the difference in yarn consumption catches up with us and we have to yank the strands even, but no where near as often.
We do however find that over time we prefer to put the dominant color (the color most represented on a row) in either the left or right eyelet to minimize the feed problem. There’s no hard and fast rule to this, it’s a matter of personal preference.
In stockinette in the round, I prefer to have the dominant color in the right eyelet, and the less represented color in the left. This helps when I lock in my floats:
Although I usually work stranding in the round, occasionally I have to do it in the flat. If I’m knitting stockinette in the flat using a Strickfingerhut, and I’m on the purl side, I prefer to have the dominant color in the left hand eyelet.
For the record, I notice no difference in the appearance of the finished product if I mix eyelets – sometimes putting the dominant color in one, and sometimes in the other. I do however note that some other Strickfingerhut users do, and advocate always keeping the background color in the same eyelet regardless of its relative dominance on any one row. Again, experimentation is your friend.
I’ve played around some with methods of producing and applying the edge finish to the khaki vest. First I tried the separately knit/sewn on band method, using a couple of different approaches to the seaming (fold band longitudinally, sew the band up, then apply it; sew on both sides in one pass; sew on the display side, then do a separate seam to affix the facing side). Of all of them, the last method worked best, but it was the most effort intensive of them all.
So I looked further. Plain I-Cord (knit on or applied) was too narrow to stabilize the edge, and two courses of it would have been too bulky. I didn’t like the way that picking up along the edge then knitting out looked – especially along the curve of the armhole.
Even more experiments ensued. Finally I landed on knitting-on a strip parallel to the edge, then going back and seaming down the free side on the inside of the piece. Doing that I could produce an edge of any desired width, go around curves and even plan on mitering the vest point corner. Here’s a swatch with a mitered corner. Note that I haven’t sewn down the facing on the inside yet, but natural stockinette curl is keeping it nice and neat. (For some, the inside seaming might be optional, but I plan on doing it on my finished piece).
To miter the corner of this 8-stitch strip, I used short rows. Here’s how I did it:
Applied 8-stitch Strip Facing with Mitered Corner
Start with the public side of the work facing you, holding it with the bulk of the piece on the left, so that you’re working up the right side of the thing (upside down from the picture above). Using straight needles, cast on 9 stitches, then pick up one stitch in the edge of the piece being finished. While the strip is 9 stitches wide, one is consumed during joining, so the part that protrudes is really only 8 stitches wide.
Row 1 (wrong side): P8, k1.
Row 2: S1, k6, ssk, pick up one stitch in edge of swatch
Row 3: S1, p7, k1
Repeat Rows 2 and 3 until you reach the corner, having just completed an odd number (wrong side row)
Row 4: S1, k6, wrap and turn.
Row 5: Slip the wrapped stitch, p6, wrap and turn
Row 6: Ignoring any previously wrapped stitches, S1, k5, wrap and turn
Row 7: Ignoring any previously wrapped stitches, Slip the wrapped stitch, p4, wrap and turn
Row 8: Ignoring any previously wrapped stitches, S1, k3, wrap and turn
Row 9: Ignoring any previously wrapped stitches, S1, p2, wrap and turn
Row 10: Ignoring any previously wrapped stitches, S1, k2, knit the next stitch along with the loop around its base, turn
Row 11: Ignoring any previously wrapped stitches, S the stitch you just knit, p2, purl the next stitch along with the loop around its base, turn
Row 12: Ignoring any previously wrapped stitches, S1, k3, knit the next stitch along with the loop around its base, turn
Row 13: Ignoring any previously wrapped stitches, Slip the stitch you just knit, p4, purl the next stitch along with the loop around its base, turn
Row 14: Ignoring any previously wrapped stitches, S1, k5, knit the next stitch along with the loop around its base, turn
Row 15: Ignoring any previously wrapped stitches, Slip the stitch you just knit, p6, purl the next stitch along with the loop around its base, turn
The corner is complete, return to repeating Rows 2 and 3. Optional finish – seam down the inside edge of this facing.
I’ve stated applying this same edging to the armholes of my vest (having previously seamed the shoulders).
I plan to do the bottom edge next, incorporating the mitered corner on the vest points. But I haven’t played with the buttonhole band treatments yet. Sadly, I have misplaced my copy of InkKNitters. It’s here. Somewhere… Weekend plans include tossing my knitting library to find it.
Oh. Unless a monsoon is upon us, weekend plans also include attending the annual Gore Place Sheepshearing Festival in Waltham, Massachusetts. Not a big festival as fiber fairs go, but very local and lots of fun. Look for me with both Elder and Smaller Daughter in tow.
After yesterday’s post on my Galaga Hat, I’ve gotten a couple of questions about the method for working back and forth seamlessly to make a tube. In specific, some people wanted to see illustrated how I make the wrap and turn join. I try to oblige them (click on pix to see them larger):
I’ve shown just the knit-side round. The purl side round works in exactly the same way. Work to the marker, making sure to work the last stitch before the marker along with the loop around its base, shift the marker over, wrap the stitch after the marker, flip the piece over, return the marker to the right-hand needle, and continue with the rest of the round.
Why go through all this trouble?
I don’t have enough yarn to strand around the entire piece. Nor do my motifs span the entire circumference of my hat. I am in effect working spot Intarsia motifs (actually I’m stranding between them, but limiting that stranding to the spot motifs). Rather than cut the yarn at the end of each motif, or stretch it back to the beginning of the spot design on each row, I am working the equivalent of flat knitting – going back and forth, alternating rows of knit and rows of purl. When I purl or knit back to my spot motif, my contrasting color ends are on the correct side of the motif for the next round. But I hate sewing up, and want to make a hat without seams. Rather than knit this totally flat (a valid option), I’m using wrap and turn to make the join at the end of each round.
I had started this post back when I was up to the shoulders of my ribbed leaf pullover, but life intervened and it languished. Also, the diagrams ended up taking more time than I thought they would. For the record, I write these entries mostly in the half-hour I steal in the morning after breakfast, while my kids are getting dressed for school. Some of the longer and more illustrated ones can take a couple of days to pull together. Yet another reason why my blogging rate has fallen back since leaving the world of consulting for full-time employment.
For the record, I’m now just a couple of rows away from completing the sleeves of the ribbed leaf pullover. I’ll use the piece to do some assembly and finishing posts later this week and next.
Some deadlines have passed, others remain. I did have an hour or so of knitting time last night, which I used to excellent effect, both for some much needed relaxation, and to advance my leaf pullover. I am now finishing up the front, at the point where the centermost stitches are set aside and the shoulders are completed.
Now this stage of production is one that has inspired a huge number of wiseNeedle advice board questions. The directions to join in a second ball of yarn and knit both shoulders at the same time tend to confuse people who are new to knitting. Here’s the basic concept. My postulated directions say something like
Work across 25 in pattern, place center 20 on holder, attach second ball of yarn and work remaining stitches; continue in pattern and commencing on the next wrong-side row, working both sides at the same time, decreasing 3 stitches at each neck edge 2 times, then 1 stitch at neck edge three times. Continue until piece measures 20 inches from bottom and bind off.
Here you see a basic sweater front (or back), knit in green yarn bearing a big R in the center so we can keep track of the right (read public) side. You see all 70 stitches on one needle, ready to commence a right side row.
At this point, I’ve followed the direction to “Work across 25 in pattern, place center 20 on holder”. Note that the stitches on my right hand needle have been completed.
Now I’m beginning the part that confuses many beginners, “attach second ball of yarn and work remaining stitches.” It’s not difficult. We’re going to do the left and right shoulders simultaneously, mirroring all shaping so that they are symmetrical. The stitches on the holder form the bottom of the neck opening. Sometimes the pattern specifies that they be bound off, other times it asks that they be placed on a holder so that they can be used “live” to form the collar. In either case, they are now parked and won’t be touched again until the pattern revisits collar production and finishing.
Take another ball of the same yarn and starting with the stitches on the far side of the stitch holder, finish out the row. Leave enough tail at the neck edge for easy finishing later. This next diagram shows the work after I’ve completed the “work remaining stitches” bit. I’ve finished my right side row.
The diagram below shows the work flipped over to work back across the wrong side (the non-public side). I’ve got my two balls of yarn set up, one for each shoulder area, and I’ve indicated the spots where the decreases should happen.
We’re up to continue in pattern, working both sides at the same time, decreasing 3 stitches at each neck edge 2 times.” The pattern is now directing the shaping of the neckline. When a pattern calls for decreasing more than one stitch at an edge I usually bind off at the beginning of a row. Yes, that makes a stepwise decrease, but as you’ll see I minimize the jaggedness a bit. The only exception to this is if I’m working in a giant superbulky (3 stitches to the inch or fewer). In a yarn that big, the steps can be quite noticeable. But back to 99.99% of all knitting.
To accomplish my first set of bind offs I have to remember to work my rows in pairs beginning on a wrong-side row- two rows each with stitches bound off at the beginning yields symmetrical decreases at the right and left edge of the work. In the diagram above, I am poised to begin my initial shoulder decrease. I have worked back across the first bunch of shoulder stitches, ending at the neck edge. No bind-offs yet. But as I begin the second set of shoulder stitches I bind off the first three, then continue across the row. Then I flip the work over to begin my right-side row, work across the shoulder side I just decreased, and perform a similar decrease on the other shoulder
At the end of my second decrease row (in this case, a public side row) I finally have symmetrical decreases on either side of my neck edge, formed by binding off stitches at the commencement of two successive rows. My bind offs are a bit jagged and step like, but that can be diminished somewhat by slipping rather than working the first stitch bound off prior to ending it off.
I am ready to go on to the next direction in my instructions. It says to decrease “1 stitch at neck edge three times”. It doesn’t say to do this by binding off. I could do it that way, and many patterns say so. But I don’t like the jaggies formed by binding off. If I’ve got only one stitch to get rid of, I’ll use plain old K2tog and SSK decreases. Depending on the pattern, I might work them in the edgemost stitches, or in the next-to-edgemost stitches, allowing them to form some sort of decorative detail. Also unlike the bind-off style decrease, there’s no logical reason to separate these between two successive rows. I generally work them on the same row. Most of the time that’s a right-side (public side) row, but in my current project – a piece with heavy texture patterning – it’s easier to do them on the plain purl worked wrong side, using P2tog and P2tog through the back of the loop so as to produce the same effect on the public side as K2tog and SSK. In any case, I place them on either side of the neck edge, creating the curve that is the foundation for whatever collar treatment is specified by the pattern in hand.
An aside: It’s interesting to note that older patterns more commonly suggested completion of one shoulder and then the other rather than knitting them in parallel. Most often those pattens gave directions for the first shoulder, and then said something like “repeat for second shoulder, reversing shaping as necessary.” (A direction that caused me to blink in wide-eyed terror while knitting my very first sweater.) There’s no reason why patterns written in that style can’t be worked in the “at the same time” method. I prefer the two-together method because it’s how I idiot-proof my own knitting to ensure that my shoulders end up being exact cognates of each other. But not everyone likes working this way.
Reasons to stick with the older method might include the unavailability of a second ball (if for example you are working off one immense cone of yarn and don’t want to break it to create a second ball); or the need to concentrate on one set of shaping directions at a time. So long as the you take care to make sure that row counts are the same and that placement of the decreases is a parallel as possible, working one shoulder at a time is a perfectly legitimate way to go. There’s no shame in working the one at a time method, it’s just a matter of mental wiring and personal comfort.
June over at Twosheep recently wrote about a tubular cast on. That sent me off looking up various ones. June recommends the one from Montse Stanley’s Readers’ Digest Knitters’ Handbook, although she notes that doing it in stockinette is not as stretchy as doing a ribbed tubular cast-on. She gives links to a couple of nicely photographed instructions at My Fashionable Life, and Little Purl of the Orient. I don’t have that particular Stanley book on my shelf, but I use an entirely different tubular cast on than the one described at those sites and in the book.
I learned an at once more fiddly and simpler method for a ribbed tubular cast-on during the second sweater I ever knit – Penny Straker’s Eye of the Partridge unisex raglan. Straker’s pattern format included a side bar with helpful advice or bonus illustrations of techniques and tricks. This one included instructions for the cast-on I did Partridge as a gift for one of my sisters. I knit it in Germantown worsted (very much like Cascade 220), in an dusty antique rose and a deeper, almost blood rose for the darker complementing color. It’s long gone now otherwise I’d put a photo here instead of the sample photo from the pattern, shamelessly lifted from a web-based retailer (the pattern itself is still available, and also comes in a kids’ version).
Straker’s method calls for using a provisional cast-on, and casting on half of the stitches called for in the pattern. If for example, the pattern asks for 100 stitches, I cast on 50. Then I knit in plain stockinette for four to ten rows (usually 6). At the completion of the last row, I unzip my provisional cast-on, and place all the newly freed stitches along the bottom edge onto a second needle. I often use a needle one size smaller than I used for the stockinette piece to make this easier.
I now have a long, skinny snake of knitting, suspended like a hammock between two active needles. I hold the needles so that the strip is folded in half, with the purl side on the inside. Then I take a third needle and alternately knit one stitch off the needle closer to me:
and purl one stitch from the needle that’s further away:
(Pictures courtesy of Younger Daughter, already at 8 as good a photographer as her mother will ever be)
When I’m done, I have a nice, neat, stretchy tubular edge in K1, P1 rib that can be made wide enough to accommodate a drawstring. I use this routinely for almost all of my hem edges – even for circular knitting. I’ve made the small divot at the join into a design feature on some pieces where I’ve started my cast-on at the neckline. On others, I’ve used the dangling tail to snick it up and make the starting point invisible.
In this case, I broke into twisted rib the row immediately following the cast-on row. How’s the leaf sweater coming? The front of it is starring in the cast-on photos, above. Here’s the back – blurry and hard to see, but proof that I’m done with it. Also proof that yes – a texture pattern that’s mostly stockinette will also curl.
For the record, here’s a bit of detail in which you can barely make out the texture pattern and the armhole decrease area (click on this for a close-up):
And because I’m still sniffing around for a small project to run in parallel with my leaf sweater, plus I’m having fun with my ancient Unger Britania– I’ll take another lead from June. It’s mittens next. The shape of a traditional Norwegian mitten looks pretty simple, yet with ample scope for fun. Hello Yarn offers a PDF of a blank mitten graph. I think I’ll take that idea and run with it – redrafting the template for a smaller gauge, and using some of the historical graphed charts from my book on embroidery. If nothing else, I’ll enjoy the doodle time.
A quickie today.
There have been a few times when I’ve wanted to work I-cord (or a knit edging) onto the perimeter of something, completely encircling it, and ending up by grafting the final live stitches onto the original cast-on row with the hope of creating as near seamless a join as possible. Here’s an example:
To date when I’ve needed to do this, I’ve either knit several rows extra of the I-cord “free” prior to beginning to apply it to the edging, or I’ve used a provisional cast-on with waste yarn for wider knit trims. Working several rows of extra I-cord gives me a snip zone I can cut and then ravel back to produce the cast-on edge live loops I need for grafting. I suppose for narrow trims, I could do a similar thing – knitting several rows of plain garter or stockinette prior to beginning simultaneous application to the thing being trimmed and commencement of my trim pattern. A judicious snip and ravel back will reveal those live loops just as nicely as working sacrificial to-be-cut I-cord does.
But I had a “doh!” moment last night. Why not just cast those few first stitches directly onto a large safety pin or small stitch holder? Unclasp, transfer stitches onto a live needle, and go! To do this, I’d use the simplest of provisional cast-ons, starting out by holding my strand behind my stitch holder and picking one stitch up knitwise, then I’d shunt the yarn to the front of the holder and with my needle tip in back of it, pick up one stitch purlwise, and so on.
Here are seven stitches picked up on a stitch holder:
It looks kind of like the figure-8 cast-on I favor for toe-up socks:
EXCEPT that by picking up the stitches instead of winding the yarn around the needles I’ve managed to mount every other stitch with the leading leg in back. Not a problem. I’d work one corrective row of purls back before beginning my edging, and on that row, I’d purl into that back leading leg to eliminate any inadvertently twisted stitches. Or I could reverse the direction of the stitch holder and wind the yarn on exactly the same way as I do for my fig-8 cast-on, eliminating the problem entirely.
I got a note yesterday from someone who commiserated at the slow going doing a piece so full of left twist and right twist 1×1 cables, and who wanted to know if there were other ways to do them.
There are several ways to go about it. Some are documented in B. Walker’s stitch treasuries, others elsewhere. The first and most obvious is to do a plain old 1×1 cable, slipping the stitch that needs to go in back onto a cable needle or spare DPN, working the one that needs to land on top, then returning the slipped stitch to the active needle and working it, too. Nice and neat, but time consuming.
Some people have a knack for working these small cable crossings without using a cable needle or other aid to hold any stitches. This works best in a nice, cooperative and slightly sticky wool, but with practice can be employed in most other materials, too. Famous Wendy is especially good at it, and has a nice tutorial on no-needle cables on her website. Although it is employed there for a 3×3 cable, the same principle holds for a simple 1×1 twist. Grumperina also has an illustrated no-cable-needle tutorial. Her method is slightly different and works well, too.
But being a klutz and prone to dropping stitches, I prefer some of the other less adventurous methods. My irrational preference here is sort of like people who prefer to keep their fingers on the keyboard while using a word processing program, disdaining use of the mouse in favor of key command sequences.
Here are a couple of other ways to make 1×1 twists. B. Walker advocates the second method described below for each (the ones I attempted to illustrate). As with most cases in which there are several ways to accomplish the same thing, experimentation is always a good idea. Different methods will give different gauges and depending on the materials used, may have an effect on fabric drape and loft. If you’ve got a pattern that’s heavily dependent on LT and RT, take a moment to play with the various ways to accomplish them when you are swatching. You may find that one of the many ways to produce them works best for your project in hand.
Left Twist (LT) Methods – Rightmost stitch ends up on top
- Identify your two-stitch unit. Skip the first stitch and knit into the back of the second, then knit the skipped stitch through the back of the loop and slide the entire unit off your needle.
- Identify your two-stitch unit. Skip the first stitch and knit into the back of the second, then knit BOTH stitches together through the back of the loop and slide the entire unit off your needle
Right Twist Methods – Leftmost stitch ends up on top
- Identify your two-stitch unit. Skip the first stitch and knit into the front of the second, then knit the skipped stitch and slide the entire unit off your needle.
- Identify your two-stitch unit. Knit both stitches together, but do not remove them from the left needle. Knit the first stitch again, and slide the entire unit off your needle.
A person posting on one of the historical knitting lists asked a question yesterday about this 18th century Spanish knitted cap. I’ve poked around the Victoria and Albert Museum’s on line photo collection, but I hadn’t taken the time to zoom in and look closely at this particular item.
At first glance the cap appears to be covered with knit-purl texture patterning, but if you zoom in (and especially if you have the ability to get an even closer look at the image) you’ll see that the texture isn’t formed by knits and purls. Instead, the design is made up of some sort of stranding that floats over a stockinette background. The question was about how this might have been done. Unfortunately, we can’t see the back of the work. So I got to thinking…
The most obvious way would be for someone to work up a plain stockinette cap, then hand-stitch the floats over counted stitches, to produce a diapered or pattern darned effect. This would certainly work, but lacks elegance. If I were making a hat like this, I’d much rather do the decoration at the same time as the base knitting, rather than going back later.
This leaves two methods – some sort of in-row wrapping, or slipping stitches with the yarn in front of the work.
Let’s look at slipping first. If you knit a row, then holding yarn in front, slip several stitches, and then resume knitting, you make a fabric that has a base row of normal height, then a distended area where stitches were slipped. If you continue to do this on subsequent rows without rows of intervening plain knit, you pull those stretched stitches up even further, creating a vertical column with a grossly distorted base structure. It doesn’t look like the knitter of this cap made the floats by slipping with yarn in front because if you zoom in and examine the long vertical bars of the ornamentation, a float seems to happens on every row, and there is no evidence of vertical distortion.
This leaves the wrap method. Wrapping stitches for ornamental effect isn’t widely practiced any more although it still survives almost as a curiosity in some cotton knitting. You can see an example of wrapped stitches in the cover pattern on the Lewis Knitting Counterpanes book published by Taunton Press. In this case the wrapping is pulled very tightly to magnify the gathered effect of the pattern. The wraps are peeking out beneath the bellies of the scallops:
I’ve also seen texture designs in European pattern collections that use wrapped stitches. There are a couple of the tight-wraps-as-gathers type at the end of Omas Strickgeheimnisse, a German-language knitting texture pattern dictionary. I thought there was at least one in the Bauerliches Stricken series (another 3-volume German stitch dictionary), but thumbing through, I can’t find it now. Some of the on-line Russian language stitch collections also show wrapped stitches I found these by searching for which may mean pattern or stitch in Russian. It also seems to transliterate to the letters “uzori or uzor” in Western alphabets, which are also good starting points for searches. (No I don’t speak or read Russian, I’ve stumbled across this bit of trivia while web-walking.) I don’t have time this morning to fish up the citations for these dimly remembered Russian texture patterns. I’ll have to leave that for tomorrow.
However, none of the contemporary sources for these wrapped stitches employ them in the way I envision that the Red Cap Knitter did.
I don’t think it would be difficult to do this, just a bit fiddly. I like fiddly. Remember that this is a thought experiment. I haven’t tried the method out yet. Perhaps over the weekend I’ll have time to do so. Here goes.
Let’s say you want to lay a ladder across four stitches. You knit the four as usual. Then you take your yarn and move it to the back of the work. You transfer four stitches from your right hand needle back to the left hand needle, then you move the yarn strand to the front of the work, laying it in the “ditch” between the first stitch to be wrapped and the ones that came before it. Then you slip those four stitches back to the right hand needle. You draw the yarn strand across the front of the work over the four, then return it to the back. You have now “lassoed” your four stitches. Give the thing a slight tug to maintain tension, and knit the next stitch as usual.
Yesterday I described two of the more usual methods for attaching edgings to project bodies – plain old mattress stitch seaming, and knitting onto a live stitch or finished edge. Today I continue with a third method. I’m not quite sure what it’s called, so I’ll call it the drawn loop method for now.
I learned this while doing the Forest Path Shawl (Interweave Knits, Summer 2003). If you plan on finding a copy of IK and working the Path, be aware that it appears to be sold out from the back issue collection on the IK website, and that there’s a correction posted in the magazine’s errata pages.
Drawn loop is intended for working a knitted trim onto a finished edge, and seems to be the least bulky of the three methods when used for that purpose. Like plain knitting on, the attachment is worked row by row as the knitting proceeds. Many patterns that use plain knitting on include directions to pick up an endless number of stitches prior to working the attached edging so that the trim is applied to live stitches rather than the original finished edge. The drawn loop method avoids that annoying exercise in endless counting although it does work best when done on a slip-stitch selvage. Unlike knitting on, there’s no column of double stitch thickness decreases formed where the edging meets the main body. As such, it’s particularly airy. It is however a bit fiddly to do, and works more easily with a smooth finish yarn than with a hairy mohairy type lace yarn.
To use drawn loop, you cast on much as for knitting on. If you use half-hitch, knitting on or a cable cast on to add your edging’s worth of stitches, you do so to your right hand needle, but instead of making a slip-knot for the first, you establish that first stitch by picking up a stitch in the edge-most loop of your main body, then work back a wrong side row to return the yarn strand to the rightmost side of your edging (and point of attachment). If you had the foresight to have incorporated a slip stitch edge in your main piece this will be easy. Otherwise you’ll have to eyeball where to pick up. Difficult (which is why many patterns want you to pick up stitches along finished sides) but with practice this is do-able.
If you use a provisional cast-on like a crocheted chain, you’ll put those new stitches onto the left hand needle, then work a wrong side row using your good yarn to return the rightmost side of the edging (and point of attachment).
Once you’re back at that rightmost edge, you use your needle tip to draw another stitch up through the next selvage loop of the main body. Here’s where it gets tricky. Enlarge that “stitch” until it’s a loop of about 18 inches or so diameter (how big to make it will become clear after you’ve done a couple of iterations). That loop will have two “ends” – one firmly attached to the knitting, the other sliding free trailing back to your yarn ball. Grab the fixed end and give it a gentle tug to make sure there’s no extra slack, then using the loop, work across the right side row of your edging. Flip the work over as usual, and work the wrong side row back, again using the giant loop. When you get back to the point of attachment, give another slight tug to the strand coming from the yarn ball to pull out any excess left over from your giant loop. Then repeat the process, drawing up another giant “stitch” in the next selvage loop and using it to knit a pair of edging rows. All of this sliding of the yarn back and forth as the large loops are made is the reason why this method works better with smooth rather than hairy finish yarns.
Again, like any attachment method that involves butting two pieces of perpendicular knitting, some adjustment of the ratio between rows worked to selvage stitches may be necessary. If the newly done edging is beginning to get too ruffly and fluttery, you may need to rip back a row or two and skip a selvage stitch. If the edging is drawing up and the body is beginning to gather, you may need to work an additional pair of edging rows without attachment.
Like I said, it’s fiddly but effective, creating the lightest possible line of attachment between an edging and a live-stitch-lacking main piece, and avoiding pain-in-the-neck sessions picking up a zillion stitches around a piece’s perimeter. I used drawn loop to good advantage on my Spring Lighting Lacy Scarf pattern, and plan on using it again on future designs.