In the die-link diagram in the original paper there seem to be 4 links leading from O7, which go to R9, R10, R18 and (apparently) R19, although it is hard to be sure of this last link. Coins illustrating the first three links are shown, but I can't find an example of O7-R19.
Ross G.
Indeed in the diagram the
obverse die 7 and the
reverse die 19 are connected but I really think that it is a mistake because in the article an O7-R19 coin is never mentioned. The line
ran away ... Excuse me
[/quote]
OK, dropping the supposed O7-R19 link means that we can now re-order the die-link diagram into a strictly linear order, with no crossed links.
There are several crossed links groups in the original die-link diagram. The first, starting with coin 2, can be re-ordered as coins 2, 6, 5 and 4, preceded by coin 1 and followed by coins 7 & 8. Coin 3, with no links, needs to move to somewhere else, although where it might fit (given its supposed numbers) is unclear.
The second crossed links group can be reordered as coins 20, 21, 22 , 23, 10, 11, 12, 13/14 (much as Mark Fox proposed), preceded by coin 19 as before and followed by 24.
Finally the third group can be reordered as coins 16, 17, 25/26, 27, 28 29.
This all seems quite neat in terms of the (known) die-links, but with certain coins (e.g. 3 and 18) the resolved numbers now no longer fit into the proposed overall number sequence scheme, and coins 10 through 14 now follow 19 through 23 when the numbers require the
reverse of this. In other words this revised linear die-link sequence doesn’t fully match the number sequence.
So, if the linear die-link sequence is valid then the number theory isn’t, but alternatively it could mean that the actual die sequence wasn’t strictly linear, i.e, that there were periods when more than two
obverse (and
reverse?) dies were in use at the same time.
Ross G.