If you don’t weave on a mechanical dobby loom then you may never have given much thought to how they work. But there are a lot of these looms about, and here in the UK they quite often pop up secondhand, so it can actually be an affordable way to increase the number of shafts at your disposal compared with going full-on computer loom.

I’ve been weaving on a 16-shaft mechanical dobby loom for over a decade, and its strengths and limitations have really worked their way into my thinking about weave structure and design.

But what has this to do with liftplans?

When you treadle a mechanical dobby loom, you advance a moving chain of ‘lags’. This lag chain is like a physical embodiment of the liftplan.

Each lag is a little like a mini-treadle, in that it is set up to raise a certain combination of shafts. The lag has one hole corresponding to each shaft, and if a peg is inserted into the hole then that shaft will be raised when the lag is activated.

There are two distinct differences from treadles, however.

• You are likely to have many more lags than you have shafts, so you can use more unique shaft combinations.
• The lags are chained together, so they have a default ordering, and there is a cost - in both time and wear - associated with activating them out of sequence.

How do these things play out in practice?

Last week I shared the draft above: a wave design in three-end huck lace, which has a repeat of 54 picks. 

Psssst… if you're new to this series and want to catch up, links to previous episodes are at the bottom of the page.

The effect of the curve in this design is created by two vertical areas of lace connected by two diagonal areas of lace. In each of the straight vertical areas we are directly repeating one pair of half-units; and the two diagonal areas we have the same sequence but in opposite directions.

If I have enough lags then I could potentially peg up the whole liftplan exactly as written. However, besides the possibility that I don’t have that many available, the loom itself may operate more smoothly with a shorter chain. 

On the other hand, if I reduce the liftplan to the ten individual shaft combinations we identified last time, then I am going to need to be skipping about the chain after every pick looking for the next lag that I need. 

I am seeking the sweet spot between 10 and 54 lags that will give me an efficient weaving process which is comfortable for both the weaver and the loom.

Although we want to join things up as much as possible, it’s worth noting that we can skip about the chain, and also that we can reverse its direction. I can weave it from the first lag to the last, but also from the last lag to the first, or I can dive right into the middle and start there.

Let's take it section by section. The first vertical alternates two sets of three picks A-B-A and C-D-C. It starts and ends with the same set of three. I'm going to assign nine lags to this section and peg A-B-A-C-D-C-A-B-A. 

The second vertical alternates a different two sets of three picks: I am going to treat these in the same way as the first set. That makes 18 lags in total.

For the diagonal, there are four sets of three picks. I am going to peg it in full using 12 lags, but I am just going to peg it once. Altogether, then, I propose using 30 lags.

This is not the most economical pegplan I could create, but it strikes a happy balance between economy and ease of use. By way of an experiment, I have made a very simple video below, showing how it can be applied.

In case you are unable to view the video, the weaving sequence is very simple. I weave from one end of the chain to the other, then I reverse it and - starting from the fourth lag - weave it back the other way. 

Each time I reverse it, I will advance to lag 4 and start from there in order to avoid duplicating the end section. You may observe that I could just drop the first and last lag shown here and pivot on the mid point of the final set of three. That would be a good option too, and maybe even a better one. It is simply my personal preference to peg in complete groups of three, so that when I am looking at the chain, I can easily identify the rhythm and the structure.

A dobby loom weaver can never have too many pegs.

I need to take a break from this series next week, but in the subsequent instalment I plan to get my skeletons out of the closet…