The second and fourth ad pages are inverted in two “original” issues.
In the first issue:
The second ad page is of McClure, Phillips & Co.
The fourth ad page is The Macmillan Company
(this is the same format as shown in the 1964 reprints)
The second original issue has these two ad pages inverted with every other page (to include ads) being exactly the same.
My March 1902 issue is an original and the same as your 1st reference regarding the ads.
Apparently this was the format copied for the 1964 reprints.
It "seems" obvious that the NGS had different print runs, i.e. the different positioning of the two ads.
My reasoning for this is:
I have "red brick" originals that have uncut pages, i.e. the tops and/or bottoms did not get trimmed sufficiently to "open" the pages. This shows the printing technique was to print one large sheet (with many pages front and back), fold the sheet into page size; bind them with covers then trim/cut the three edges to open the pages. The production back then "often" got the trim wrong leaving pages connected and unable to open.
Using this technique means all the pages must fall in the same sequence for the entire production run. Only a "different" production run using different page placement on the large sheet (in this case the back of two consecutive pages) could produce a different page sequence.
I'm open to suggestions as to any other method for producing two originals with different page sequences.
Note 1: I'm not trying to equivocate different production runs to reprints (though this could still be the case). A different production run could simply be resetting the presses due to scheduled maintenance or a number of other reasons.
Note 2: On-the-other-hand, it could also be a reprint. Buxbaum identified NGS reprint orders for Volumes II and VI, but could not identify the date of these orders. He surmised the reprints were ordered prior to 1896 as the NGS was advertising "that year" for the offer to purchase these numbers in good condition from older members due to their absence from inventory for new members. He also believed these "reprint orders" continued beyond the red brick issues but could not prove it. He stated the differences between the reprints and the originals were indistinguishable (except, perhaps, in my opinion, if the page sequences were different?).
Don't know what I was thinking - it was Volumes IV and V that Buxbaum discovered an NGS reprint order for.
I had to read it a second time to understand. My problem not yours. Every time I see the word reprint I automatically think 1964. You were talking about "down the road production runs" a la the April 1905 red cover run which was in the Oct.-Dec. 1905 time frame judging from the ads. This reprinting practice seems to have been common.
Interesting production technique. Thanks for the info.