[toppcat]

Originally Posted by G1911

Just to add on, this structure is present in other sets that were part of the 1909-1912 project popularly credited to the ATC and AL. It is rare that cards were corrected, but when they were they typically had multiple backs. For example, the 3 spelling errors corrected in T218-1 all exist on 2 of the 4 backs, even though they represent a small percentage of surviving copies and the majority of those same 2 backs are the corrected version.

This ‘printing in waves’ appears to be a significant factor. The ATC ledger gives some evidence that some series were issued in waves; like the discordant dates for different sport subjects of T218-3.

Some 50 card sets seem to have had 25 unique cards to a sheet, and sometimes a back gets only half the subjects, like we see in T42. It seems to suggest wave printing again, not just a 2 sheet construction but those 2 sheets being done at a time gap during which decisions were made. I think our evidence suggests this happened with T220 also. The gap in between sheets saw multiple decisions made, to expand the back distribution, to cheapen the borders, to modify a couple cards, and to change the entire art style between at least four production runs over ~6 months.

While advertised and thought of as series, the traditional idea that all cards of a series were basically printed and issued together like Topps cards does not seem to be the case.

Topps cards were printed backs first then shipped to another plant for the fronts to be printed for a period of many years. It seems possible tobacco cards were handled the same way, especially as the reverses were monotone and would require less skill than the 6 color process used on the fronts. You would have a pile of back printed sheets to get through before the next batch came in, theoretically at least.