Editor's Note: This excerpt appears in Issue #7, Margaret Mitchell of BookThink's Bookseller's Author Reports.
... inevitably, most books with universal appeal spawn bibliographical complexity. Later printings and editions arrive quickly, and
textual changes are usually ongoing. In the case of Gone With the Wind, though editorially sound at the outset, minor errors persisted through many
printings before all were addressed. These errors, however, as they often do in other books, have not evolved into determinant issue points -
for that matter, nobody has enumerated them. For years Macmillan was consistent in not only stating each printing as it was issued but also in
noting previous printings. The single statement of Set up and electrotyped. Published May, 1936 on the copyright page of the first printing soon
became a paragraph of subsequent printings, a sort of capsule publishing history that obviated citing of other points of edition identification.
So - it's these statements that can guide us, but there were two hiccups early on that threw all of it off, and consequently, GWTW has become perhaps
the most widely misrepresented book by booksellers of the 20th century.
This story begins with the Book-of-the-Month Club. On April 15, 1936 the club, having decided it wanted GWTW for a summer selection, offered Macmillan $10,000
for the right to publish 50,000 copies. There was only one catch: The scheduled May release would have to be pushed back to late June to accommodate BOMC's schedule.
This was common practice with book clubs; it guaranteed that members would receive their copies within a week of their appearance in bookstores. However,
Macmillan had already fired up the presses in early April and printed 10,245 copies that carried the statement Published May, 1936. In Depression times, Macmillan
could hardly afford to toss them and start over, and they had already exerted considerable promotional effort toward a May release. But nor could they afford to
pass up either BOMC's $10,000 or the massive spike in publicity that would most certainly ensue with their participation. It was decided, finally, to cancel the
May release of the US and Canadian editions and distribute the May copies in June, guessing that an incorrect date would not create copyright issues as long
as two June control copies were sent to the Library of Congress - that's right; the Library of Congress received two second printings.
By early June so many advance orders had come in that the publisher ordered that very second printing of 25,000 copies, these dated June, 1936. A first issue
dust jacket back panel that had featured a Macmillan list of 17 "Spring Novels" in which GWTW appeared 10th was changed to a list of 13 "New Macmillan Books" with
GWTW at the top. Actual shipment of the books began in the latter half of June, and copies of the first two printings were commingled so that all stores would
have at least some first printings available to them.
But here's where things start to get somewhat confusing. The June printing makes no mention of the May printing on the copyright page, stating only
this: Set up and electrotyped. Published June, 1936. Nor is there mention of it in any subsequent printing. And yet many collectors had contacted
Macmillan prior to publication, inquired about First Edition issue points and were assured that the first printing carried the May statement. It seemed as though the
May printing had "disappeared" for the purposes of protecting copyright and yet remained in place from a collector's standpoint.
And there's more confusion. In late June a third printing of 15,000 copies was ordered as well as the aforementioned 50,000-copy printing for the BOMC - all produced
at the Ferris Printing Company and all 65,000 copies identical, that is, trade editions could not be distinguished from book club editions. This practice was uncommon but
not unprecedented, especially in the early years of the BOMC. In any case all 65,000 copies carried this statement, despite the fact that two separate printings were done in late June:
Set up and electrotyped. Published June, 1936.
Reprinted June, 1936.
One would assume that Macmillan simply (and reasonably) regarded the BOMC printing separately from its trade printing totals, and this would explain the omission.
But wait. The indication of two June printings persisted well into multiple August printings, when, lo and behold, a third June printing suddenly pops up in the following statement:
Set up and electrotyped. Published June, 1936.
Reprinted June (twice), July (three times),
August (five times), 1936.
To compare, here is the previous printing statement:
Set up and electrotyped. Published June, 1936.
Reprinted June, 1936. July, 1936 (Three times.)
August, 1936 (Four Times.)
Both of these hidden factors wreak havoc on an attempt to establish printing priority. If one didn't know better - and most booksellers do not - the fourth August
printing would look very much like a 9th printing. Or, even if one knew of and acknowledged the existence of the May printing, a 10th printing at best. In fact, however,
it is an 11th printing, given the hidden BOMC printing that surfaced again on the copyright page of the 12th (the 5th August) printing.
The bibliography that appears at the end of this report will lay everything out from start to finish, but for now, booksellers and collectors should note two factors:
One, for the purposes of buying and selling this book, there were effectively no book club editions of GWTW issued in 1936, and it was not until the early 1950's that a
very conspicuous BCE was issued with a Roman Numeral copyright date, a double-column text format, and a different binding, this despite the fact that huge numbers of "book
club editions" were sold from the get-go to BOMC members. The BOMC did issue a laid-in Book-of-the-Month Club News in 1936, and if one was encountered in a copy, it would be
at least somewhat persuasive that one had a book club edition, perhaps more persuasive yet if there was evidence, via offsetting, that this literature had been with the book
for many decades and left a footprint, but there simply is no definitive means of identification. For purists, one can state with certainty that all copies carrying the May
or first June statement only are trade editions in every sense of the word.
Two, though Macmillan waited until the 12th printing to acknowledge the existence of the June BOMC printing on the copyright page, it's clear in their correction that
they regarded it as a separate printing and in fact had noted it much earlier on an inside-the-front-cover full-page ad in the July 11, 1936 The Saturday Review of Literature which
stated "Over 100,000 copies before publication!" This total would have been only 50,245 at the date of publication had the BOMC printing had not been counted.
Given the widespread ignorance of these two factors, a careful explanation of why a given printing number is asserted should ideally accompany any bookseller's description
of a later printing. This would not be necessary if the later printings were not enthusiastically collected.
Click here to purchase the complete Margaret Mitchell Author Report.
