I really like double running stitch. The more complex the pattern, the better. Best of all are the amazingly detailed ones from the late 1500s/early to mid 1600s that are an explosion of vegetal forms. Some are inhabited by natural or mythical creatures. Here’s an example:

StitchPuppy, a stitcher new to double running stitch (aka Spanish Stitch, Holbein Stitch) asked me about the logic and method of working double running. She’s familiar with the working method – that the final effect of a solid line is achieved by two passes of the needle. On the first trip every other stitch unit is made, and on the return trip, the “in betweeners” are filled in:

(Pix above are from TNCM). She understands that with careful stitching, pieces in double running stitch can be made to look exactly the same on the back and the front – a plus for cuffs, collars, napkins and other applications where both sides of the work are likely to be seen. Where StitchPuppy has problems is on understanding how the method can be applied to complex patterns. She wants to know where I start when I tackle a complicated double running pattern, and whether or not I use a logic that helps keep me from running into dead ends, or that helps ensure that I do end up with a front/back reversible end product.

I’ll try to answer.

First – not everything that’s graphed out for double running can be done easily totally two-sided. Any design with an isolated bit of stitching that’s not connected to the main pattern presents a problem. The small dolphin just below the mermaid in the stitched panel above is not connected to the rest of the design. It’s a stand-alone element. To work this particular piece double sided, one would need to both begin and end off a separate strand of thread, just for that dolphin, or connect it on one or both sides to the main motif by one or more bridging stitches. Either way, the dolphin presents only a small problem. A larger one is posed by the mermaids’ facial features. The eyes, nose and mouth are isolated from the main stitching areas, and are too small to be worked double sided and have enough area to finish off the ends.

The rest of the mermaid pattern can be worked double sided. There are no other logical impediments to completion. But how to work a complex design? Not hard. Any design without a discontinuity (like the orphan dolphin) can be envisioned as a single baseline, with detours to fill out the details or as a series of areas. Let’s look at the phoenix I posted here a couple months ago:

(By the way – see that border? The octagonal interlaces are not connected to the little “Vs” filling out the border north and south. Lots of discontinuities there, and if you saw the back of that work you’d notice the bridging stitches I used to connect the design elements).

Back to the phoenix. It’s pretty easy to identify a baseline around the phoenix’s perimeter:

Sometimes I stitch this way – working a long every-other-stitch outline around the entire motif, then going back and doing the “detours” from that line. The advantage of working this way is that it’s quick to block in the major design elements and to make sure they’re properly aligned to each other before investing time and thread in filling in the rest of the design. The primary disadvantage is that it’s hard to keep count during long straight runs. This is the working logic described in most blackwork books. This piece shows another example of the conventional baseline-first attack method:

You can see that I’ve outlined the blossom’s main elements, and am now following along to work the individual petals.

However, I’m far more likely though to work my pattern in a more compartmentalized manner, either identifying the baseline but instead of following it and filling in detail later, starting on the baseline and taking every detour that presents itself. I’m using the baseline identified above, but instead of following around the bird, I immediately zip down to do that first little feather slice, returning to the baseline when that’s done.

Worked this way, the design gets filled in early on, moving down the baseline and accomplishing the detours, and returning to the baseline after each one. All that’s filled in on the second pass is the every-other-stitch segment of the baseline. .I find this method much easier to use for complex charts. It’s quite easy to count little completed feather units in the bird’s wingtips as I finish them. The flower strip above also shows the second method. I used it for the acorn sprigs. I stitched along the baseline, but every time I got to a branch, I finished the branch before returning to the baseline. The second pass is a straight run along the baseline itself.

Where to start? It depends on your work, the style of frame you are using, and your own preferences. In general it’s better to minimize handling of the stitched area. Working from the center out is an accepted practice because it tends to keep sweaty hands away from finished stitching. But there are times when working that way isn’t logical. I began the phoenix with its head, having matched the center of the pattern with the center of my to-be-embroidered area. The phoenix was also at the rough center of my finished project and was one of the early elements I completed on it. The strip below though was done bottom up. And the patterns I’m working on my current piece were begun at the cloth’s center. It’s all situational.

Where is the baseline in an all-over pattern? Wherever it’s convenient. Here you can see that I’m using two baselines for the twisted frame element, and not worrying about completing the entire interlace in one gulp:

Is there any way to determine which method was used on historical pieces? Scholars may have made figured it out but I haven’t run across word of it in popular stitching literature. The most reliable way to figure out historical stitching logic would be to pick apart an artifact. NOT something anyone sane would do.

One word of caution to those who want to work something two-sided. Resist the temptation to use veeerrryyyy loooonnnnnggggg strands of thread to minimize the number of ends. They WILL tangle and abrade as they are stitched. You will curse the day you started the project. (Trust me on this.) I do have a trick to share, though. If I use a very long strand I start from the middle of it. I pull the thread half-way through my work, then in an inconspicuous spot, I wind the excess thread around a straight pin. I stitch away with the free end until it’s ready to be terminated. Then I go back and free the other end of the thread from the pin, and use that. Since I am stitching with a sane length each time I avoid tangles and thread wear, but I minimize total ends. Of course this presents its own logic problem – how do you know where to start the next mega-thread, but that’s a conundrum for another day.

I hope that this is helpful to StitchPuppy and with luck others, too. If anyone has questions about identifying baselines or stitching logic in double running, please feel free to post them here.

Technorati : counted-thread, embroidery