In search of the Carolingian Cross

In my last blog posting, I noted an interesting shape based upon the triquetra I'd found in a wikipedia article and was pretty sure could be accomplished with poi--I've spent most of the past evening and day thinking through how and coming up with different approaches to it.

About mid-day today, I thought I'd had it: imagine a square, bisected through both corners. You start at the middle and do a single triquetra, but when you reach the point of origin, you perform another triquetra 90 degrees off of the first one and so on and so forth all the way around the square. I discovered I could do the pattern in under 3 seconds which, according to Abject, is what I should be shooting for to create a recognizable pattern. Below is a diagram I did in Illustrator to visualize the concept.



There's just one problem: this doesn't actually come out looking like the Carolingian Cross in the wikipedia entry. Conceptually they bear similarities, but there is no distinction between the triquetra nodes at each corner. Back to the drawing board.

Using the same method, I tried separating out the center nodes till it began to have the same appearance as the wikipedia figure:



As you can see, the hand path is no longer a simple square--in fact, it appears to be an antispin flower in its own right. Months ago I'd toyed with the idea of compound flowers in this way and this appears to be an excellent test of the idea. Back to practice--let's see if this one is doable.