I found the Moire effects of the radial coloring of the root 1001 and 3001 configurations (page 3) especially beautiful -- *and*, it reminds me of a computer graphics experiment I tried about 25 years ago: 1. Create a color table, from 0 to N-1, by using N (N large) consecutive colors equally spaced around a circular rainbow (e.g., follow the circle inscribed in the hexagon that bisects the RGB cube halfway between black (0 0 0) and white (1 1 1)). 2. Now on a preferably high-ish resolution screen whose pixels are indexed by (K,L) for 1 <= K <= xmax, 1 <= L <= ymax, color (K,L) with the color whose number is (K^2 + L^2) mod N. I'm sure many other people have also tried this experiment. I don't think I ever learned exactly why I got the Moire pattern that resulted, not dissimilar from Bourke's 1001 and 3001 configurations with radial coloring. --Dan