> Okay, I haven't much time, so this ain't gonna be clear, but I'm going to
> try my best. I've found a way of counting which has yielded partial
> results, but those may be coincedental. Here's what to do. You start from
> any arbitrary number within a given glyph, say, G1. Starting from, say,
> 24, you start counting CLOCKWISE. 24 (w), 25 (nw) 26 (n) and so on. If the
> increment you're counting equals one of the numbers of the glyp, then that
> number is "validated," and the accompanying letter may be used.
> So... for an example of "validation," start counting from 19....
> 19 (19), 20 (50), 21 (62), 22(75), 23 (20), 24 (24) -- bingo!

Nice idea! However, G6 has too many coincidences 'clockwise':

10 is validated (thru 14 & 15)
4 is validated (thru 66)
15 is validated (thru 80)
14 is validated (thru 15)

BTW a number x is validated by y if the following holds

x n1 n2 n3 n4 ... y <- numbers in the glyphs
x+1 x+2 x+3 x+4 ... x+n <- GH counting

and y-(x+n) is divisible by 8. For example

4 15 80 14 15 20 66

5 6 7 8 9 10

and 66-10 = 56 = 7*8.