On September 23, Peter Collingwood responded to a message on the TWIST mailing list and brought the expressions "twist-on-twist" and "twist-against-twist" to my attention. He refered to Page 79 in TTW which is relevant to the bands below. His full message was:
Sponsored by TWIST - Tablet Weavers International Studies & Techniques Gudrun. You could take 4 S-plied threads and 4 Z-plied threads and twist them both in the same direction, making two 4 strand cords. I think this will show the difference between a "twist-on-twist" cord (bumpy irregular surface) and the more normal "twist-against-twist" cord (smoother surface). Former will try and untwist much more than latter, which tends to be more balanced. So this is the visual difference you would get if using this idea as stripes in a TW band. I know that page 79 in my TTW looks very dull (no pictures or diagrams !) but it was the result of much experiment in this field. There is, for example, an unexpected effect that a turning tablet has on the actual twist in the threads it holds. This means that a warp thread on near side and on far side of the tablet pack can acquire a (sometimes quite visible) difference of twist. I suppose looking at any pre-1800 TW in a museum collection would give an idea about the type of hand-spinning originally used for warp yarn. Peter Collingwood Send private reply to "peter collingwood"
Those of you who saw me at Convergence 2008 or the Complex Weavers Conference in Florida may remember my necklace with the TWIST pin. Later I wove a similar piece and wrote an article for HANDWOVEN, Nov/Dec 2008, p. 62, 63.
Both pieces were woven the same way. Each band consists of 4 parts: flat beginning - tube - reverse tube - flat ending. The woven yarn is S-plied silk, the cards are Z-threaded. Up to the midpoint all cards were turned forward creating S-twist (twist-on-twist) and a lot of torque in the tubular section.
The torque developed rightaway while I was weaving. In the second half the cards were turned backwards creating Z-twsit (twist-against-twist). I expected to get some torque here but to my surprise there was almost none. However, the flat end piece spiralled a lot more than the beginning (after being taken off the loom). The shuttle always went from right to left through the shed - at least in the second piece. I don't remember which way I wove the first piece, it may make a difference. There would probably more torque after the midpoint if the shuttle had gone the other direction.
The tube has a core of several threads of crochet cotton.
As you can see there is a lot going on here and more experiments can be done. Read page 79 of TTW for a better understanding of what causes corkscrewing.