The first thing to check with a Gid rubbing, is the scale. The rubbings are not all at a 100% scale, and there is no indication of scale. In this case we have a perfectly accurate scan of example three with which we can test the scale. This is what I have done in Comparative Diagram 1. I have cut and pasted a strip of AE a 1 next to the rubbing and found that the rubbing was smaller than scale. I then resized the rubbing 115% at a rough guess and tested it again, it was now just a fraction out and I changed it to 116%. By running the green lines across the imprints I can see that they are now the same size. Now we can test the DC a 1 roulette using both Gids type model and a strip of actual imprint that has been transformed into grayscale and inverted so as to resemble a rubbing. We can never be too careful in testing imprints they may look very similar and yet not be the same roulette, I was not really convinced until I added the blue lines, the chances of these details, that are all at different angles, being the same are very slim, therefore we can be say with a certain measure of confidence that the 254 binding was decorated with a DC a 1 roulette, the same roulette found on example 3. The last roulette was really tricky, even with all the little circles to help us, but eventually we have to concede that this is a FR 6 roulette. The roulette is also found on Gid 101, we absolutely have to test this because if this is right it opens up a whole new can of worms.
