With those nice hexagonal beads, I wanted to frame them with a hexagonal knot. It didn’t need to be particularly stable as I’d be reinforcing the central area with thread (and more beads), so likely I’d be taking a hexagonal knot and reduce the amount of overlapping an favour of a bigger centre hole, the better to display the framed bead rather than hide it. Looking through Lydia Chen’s third book, I decided on the “Type 2 Compound Constellation Knot” which unlike your basic hexagonal stellar knot (which is uouu) has a structural loop path of uuouo. At this point I’ve tied it about 6-7 times and each time it looks… funny. As usual with a new polygonal knot, it’s the last interlacing pass on the back side that’s tricky.

So here’s the creeper knot. You’ll notice the free ends extending to either side versus the usual free ends making up an ear of the knot. That’s because the creeper is meant to wrap around an object, like the bow on a package. The description in the book makes me think that she maybe wanted “vine” versus “creeper” but…

So, if I already have a mat knot and a polygonal knot then do we need a mystic knot? Polygonal knots are such a distinct class (which conveniently intersects with the vast family of mystic knots) that I think it’s beneficial to show it in this context.

To better illustrate the midpoint I redid this knot with one of my fused cords. These cords that I often use to tie example knots are actually shoelaces, sold locally as “oval shoelaces”. They look like president’s cord with the edge parts in a contrasting colour. I would love to find some real president’s cord in these designs as (a) shoelaces are a very expensive way to get string and (b) shoelaces are meant for dragging through the muck and aren’t exactly woven in a very refined way, so (c) real president’s cord takes the curves better and would probably have made a fine double coin example.