That's easy... Taking Graham Esler's article into consideration from I believe the September 2006 CPMS Newsletter, along with a bit of take-my-word-for-it estimation based on prefixes, notes were printed 40/on at the time and skip-numbered by 1,000 throughout the series.
Normally, serial number errors occur as a result of one numbering head being mis-set or otherwise not on the proper number. For example, a wheel has jumped ahead a couple spots. This will stay out of sync until reset by the press operator, and the result is a vertical stack of serial number errors in only a single position.
However, it is clear from the information available that it was in fact two whole sheets that were numbered identically. How it would come to be that two sheets were numbered identically, I don't know. The 001 sheet is the second sheet in the stack, assuming the changeover from 001-1000 to 000-999 numbering occurred at A/X as the catalogue suggests.
So that would mean there should be 40 pairs of these errors. Four are already accounted for. I would estimate, again based on those seen so far, that the numbers will be 2040001 through 2079001 with only the bolded digits changing.
Logged
BWJM, F.O.N.A.
Life Member of CPMS, RCNA, ONA, ANA, IBNS, WCS.
President, IBNS Ontario Chapter.
Treasurer, Waterloo Coin Society.
Show Chair, Cambridge Coin Show.
Fellow of the Ontario Numismatic Association.
