Introduction
Information Source I: Wayne G. Hammond
Information Source II: Rayner Unwin
Information Source III: The A&U Archive
Printing & Binding History
   - The Fellowship of the Ring
   - The Two Towers
   - The Return of the King
Summaries
   - By Impression
   - By Date
Postscript
References

Introduction
Although Tolkien had largely completed The Lord of the Rings by 1949, it was not until late in 1952 that Allen & Unwin were able to assess the manuscript. After obtaining composition and printing estimates from Unwin Brothers and Jarrold & Sons, Rayner Unwin informed his father that he thought it was a work of genius, but that, if published, it might lose the company a thousand pounds.  Fortunately the book buying public were in agreement with Rayner and the three volumes sold steadily, but not spectacularly, until the mid-1960s when interest in Middle-earth blossomed and sales soared.

In the absence of the present study, very little information regarding the number of books printed and sold is available.  The little that is known has been reported in books by Wayne G. Hammond and Rayner Unwin and relates, in the main, to the first impression of each book in the "trilogy".  The figures supplied are given below, together with more detailed numbers determined using correspondence and publishing ledgers held in the Allen & Unwin archive at Reading University.

Information Source I:  Wayne G. Hammond
The conventional view of the size of the first impression print runs is set out in J.R.R. Tolkien: A Descriptive Bibliography by Wayne G. Hammond.  It states:

* HM = Houghton Mifflin

Information Source II:  Rayner Unwin
According to Rayner Unwin’s account of the later history of A&U, given in George Allen & Unwin: A Remembrancer, selling part of the print run to an American publisher was common practice, and gaining such an order would underwrite most of A&U's own manufacturing costs.  The figures given above are contradicted by Rayner's account, which suggests that the figures quoted for A&U are in fact the total number printed, i.e. they include the sets of sheets exported to the United States for use by Houghton Mifflin.  The figures reported by Rayner breakdown as:

 
Information Source III:  The Allen & Unwin Archive
The A&U Archive is part of the British Publishing section of the Reading University Library Special Collections.  The bulk of the archive consists of around 1,300 boxes of correspondence, covering the period 1914-1968, containing letters to and from authors, agents, publishers, booksellers, individual book buyers, binders, paper suppliers, advertisers, printers etc.  This is complemented by manuscript books listing manuscripts received (1914-1972), twenty one publishing journals and ledgers (1917-1957) and a host of other records.  The correspondence has been catalogued and can be searched using Unicorn, the library's online information system.  The other materials are detailed in a typed handlist.  Further information can be found on the Reading University website and in Brian Ryder, 'The George Allen & Unwin Collection: Reading University Library', Publishing History v.47 (2000), pp.67-78.

Publishing Ledgers 109 and 110 include details of the production and promotion costs, and the sales and stock figures for The Fellowship of the Ring, The Two Towers and The Return of the King.  They give the following figures, which, in general terms, agree with those given by Wayne G. Hammond:

[It was common practice for the printer to produce more sets of sheets than is ordered by the publisher to compensate for spoilage during the production process.  This often means that more bound copies are produced than were originally ordered.  The figures above for Total Printed give the number ordered, followed by the sum of the exported and the bound copies in brackets.]

Printing and Binding History
The correspondence and ledgers in the A&U archive include a wealth of further information.  Analysis of this has made it possible to determine the dates when each impression was first available, the number of copies printed and bound, and the number of copies exported to the United States for publication by Houghton Mifflin.

Where there is some uncertainty about quantities or dates this is denoted by an asterisk (*).  Differences between the Stated Date and the Date First Available occur because the publisher had to estimate when the printing and binding would be completed.  Unforeseen problems sometimes meant that the completion of the binding slipped into the following month.

In most cases the A&U copies were bound in batches, usually two or three per impression.  There were a number of reasons for this:

• If all of the copies were bound at the same time then the costs of binding all had to be paid at the same time.  By spreading the binding over a period, say a year, the costs were also spread.
• Should demand for the book decline then there was relatively little value tied up in the unbound flat sheets (essentially just the cost of the paper), whereas if the books had been bound then costs involved were much more substantial.  For instance, for the 1st Impression of The Fellowship of the Ring, the costs were £152 for printing, £273 for paper and £237 for binding of the A&U copies.
• Bound copes, especially the dustwrappers, are more susceptible to damage than unbound flat sheets.

The method used to produce the Houghton Mifflin copies varied depending on circumstances prevailing at the time.  For earlier impressions, when Houghton Mifflin were able to indicate their needs in advance, they were printed with the Houghton Mifflin imprint on the title page, while for some of the later impressions a joint A&U/HM imprint was used.  When Houghton Mifflin were unable to place an order prior to printing, A&U copies were 'converted' to Houghton Mifflin copies by use of a cancel title leaf or fold.  This is indicated below by a double asterisk (**).  For further information see the notes following the tables.

Readers may notice small differences between the total printed and the sum of the A&U and Houghton Mifflin copies.  This is one of the many problems encountered when working with the A&U records.  The differences are due, in the main, to the printer compensating for spoilage during binding by printing more sets of sheets than were ordered.  While most of the printing work was completed by Jarrold & Sons, five impressions of The Fellowship of the Ring were printed by Purnell & Sons as indicated below.  [N.B. K&W = Key & Whiting Ltd (Bookbinders)]

Notes - The Fellowship of the Ring
Galley Proofs - Four sets of sheets were sent out by Jarrolds in a piecemeal fashion commencing on or about 30 June 1953, presumably in an unbound state.  The final galleys were sent out no later than 19 July.

Page Proofs - The quoted figure of 35 copies is not certain.  27 copies were returned to Jarrolds for perfection of a binding error on 28 September 1953.  On the following day, one of the copies without the fault was sent to Tolkien.  The figure of 35 has been assumed because that was the number produced for The Return of the King - publishers are creatures of habit!

Travellers' Samples - Printed (and presumably bound) at the same time as the page proofs, these samples consist of the first two signatures of a book bound up for use by the publisher's sales representatives.  While there is no direct evidence that any were produced for The Fellowship of the Ring, it does appear to have been common practice to do so.  The figure of 14 was chosen as this was the quantity produced for both Beowulf and the Finnesburg Fragment and Farmer Giles of Ham.

Advance Copies - Sent out by Jarrolds on 11 June 1954.  The endpapers were white rather than cream-coloured, as is usual with most copies of this book.  It is unclear whether any of the bulk supplies also had white endpapers, but as Jarrolds would not have known that the endpapers must be cream-coloured until 14 June 1954, and by 8 July the 2,500 copies had "been bound for a few weeks", it seems possible that they had begun binding before they received the instruction.  The advance copies were probably wrapped in a proof version of the dustwrapper, which included two errors in the text on the lower cover.

Paperbound Copies - These were identical to the First Impression, but bound in paper wrappers rather than the usual red cloth covered binding case.  Jarrolds despatched them to A&U on 22 June 1954.

1st Impression - Twelve of the Houghton Mifflin copies were despatched by air in June 1954 for use as advance copies, while the remaining 1,538 copies were shipped by sea in July 1954.

3rd Impression - The January 1955 batch of Houghton Mifflin copies included a Houghton Mifflin title page as printed, and lacked any A&U device or colophon.  The remaining 3,000 sets of sheets all bore the A&U imprint.  In April 1955, 500 copies were converted to Houghton Mifflin copies by use of a cancel title.  The correspondence indicates that this was a 4-page affair to replace the title page, title verso and pages 29-30, and was printed on Art Laid paper.  However, Steven M. Frisby's study of the Houghton Mifflin Edition of The Fellowship of the Ring (see References) suggests that a cancel title leaf was used - attached to the stub of the original title page.  Analysis of the quantity of paper used for the printing supports the use of the 4-page cancel indicated by the correspondence, but it may be that only the title page was inserted when Unwin Brothers bound together the sheets.  In June 1955, a further 2,000 copies were converted using the cancel titles, this time printed on Drury Antique Wove paper.  A figure of 1,000 copies is recorded in Publishing Ledger 109, but this appears to be an error caused by the official order for 1,000 copies being amended to 2,000 copies in a subsequent letter.

7th Impression - The date given for the second binding (April 1958) is an estimate based on the number bound in 1957. The date for the third binding (August 1958) is based on an invoice for binding issued in that month.

8th Impression - The number of copies produced by the first and second bindings is unclear, but they totalled 3,000 copies.  A quantity of 1,500 copies has been indicated for the first binding to tie in with the corresponding binding for The Return of the King.  The quantity for the second binding was then derived by calculation.

9th Impression - The first binding order was actually for 1,500 copies, but for some unknown reason only 770 copies appear to have been produced.  A subsequent binding order was placed on 9 June 1960 for 2,194 copies.  When this was completed is uncertain, but by 25 August it had "been ready for sometime" and more than one delivery had been taken from it.

10th Impression - On 30 August 1960 A&U asked Houghton Mifflin to indicate their requirements for the forthcoming reprint.  There is no record of the response but copies of the 10th Impression have been seen in a Houghton Mifflin binding.  2,000 copies are not accounted for in the records, so an even split between A&U and Houghton Mifflin has been assumed.

12th Impression - A total print quantity of 10,000 copies has been assumed, matching that of the 1962 Impression of The Return of the King.  The figure of 3,500 copies for the first binding was calculated by subtracting the 2,500 copies for Houghton Mifflin and the second binding order of 4,368 copies (including overs and waste) from the 10,000 total print run - assuming that the 368 copies were the overs produced.

13th Impression - There is no indication in the records of when the second binding order was completed, but the first deliveries were taken from it in October 1964.  It is unlikely that the books were bound before March 1964, because at this time A&U were planning to use some of the existing sheets for the Deluxe Edition.

14th Impression - The quantity of 4,000 copies produced by the second binding order is an assumption based on the fact that the first binding orders for both The Fellowship of the Ring and The Two Towers were for 2,500 copies and the second binding order for The Two Towers was for 4,000 copies.  The figure for the fourth binding order was then derived by calculation.

15th Impression - Only one binding order is recorded for the 15th Impression, so presumably the entire print run of 4,000 copies was bound at this time.

Notes - The Two Towers
Galley Proofs - Four sets of sheets were sent out by Jarrold & Sons on 26 November 1953, presumably in an unbound state.

Page Proofs - The exact quantity is unknown.  The figure of 35 has been assumed because that was the number produced for The Return of the King.  The exact date these were available is also unclear, but A&U sent a copy to Tolkien on 4 March 1954.

Travellers' Samples - Printed (and presumably bound) at the same time as the page proofs, these samples consist of the first two signatures of a book bound up for use by the publisher's sales representatives.  While there is no direct evidence that any were produced for The Two Towers, it does appear to have been common practice to do so.  The figure of 14 was chosen as this was the quantity produced for both Beowulf and the Finnesburg Fragment and Farmer Giles of Ham.

Advance Copies - Twelve copies were sent out by Jarrolds in two batches of six.  The second batch arrived with A&U on 23 September 1954.

2nd Impression - Houghton Mifflin did not order any copies of the 2nd Impression when it was printed, presumably because their own edition had not yet been issued.  In April 1955 they placed an order for 1,000 copies, and in May increased this to 2,000 copies.  Only 1,200 sets of sheets remained from the 2nd Impression and these were converted to Houghton Mifflin copies by use of a cancel fold to replace the title page, title verso and pages 27-28.  The remaining 800 copies were taken from the 3rd Impression.

3rd Impression - The July 1955 batch of Houghton Mifflin copies included a Houghton Mifflin title page as printed, and lacked any A&U device or colophon.  The remaining sets of sheets all bore the A&U imprint.  In September 1955, 1,000 copies were converted to Houghton Mifflin copies by use of a cancel title.

6th Impression - The number of copies produced by the first and third bindings is unclear, but they totalled 3,000 copies.  A quantity of 1,500 copies has been indicated for the first binding to tie in with the corresponding binding for The Return of the King.  The quantity for the third binding was then derived by calculation.

8th Impression - Details for this period are very sketchy, so it is possible that the total print quantity was higher than is noted above.  The figure stated is the sum of the quantities produced by the two known binding orders plus the Houghton Mifflin copies.

9th Impression - A total print quantity of 10,000 copies has been assumed, matching that of the 1962 Impression of The Return of the King.  The figure of 1,000 copies for the first binding is dependent on a total print quantity of 10,000 copies - 5,000 sets of sheets remained after the second binding of 3,000 copies.  Houghton Mifflin took a second batch of 1,000 sets of sheets in February 1963, leaving 4,000 for the third and final binding in May 1963.

10th Impression - The first binding was due to be completed in August 1963, but may have slipped over into September.  The figure of 4,000 copies for the first binding was selected as this was the quantity produced for the other two volumes.  The quantity produced by the second binding is dependant on the assumed figure for the first binding.  There is no indication in the records of when the second binding order was completed, so the date given is the same as that for the 1963 impression of The Fellowship of the Ring.  It is unlikely that the books were bound before March 1964, because at this time A&U decided to use some of the existing sheets for the Deluxe Edition.

12th Impression - Only one binding order is recorded for the 12th Impression, so presumably the entire print run of 4,000 copies was bound at this time.

Notes - The Return of the King
Galley Proofs (Main Text) - Four sets of sheets were received by A&U on 5 July 1954, presumably in an unbound state.  A further 10 sets were ordered the following day, with a request to cut them to a depth of eight inches.

Page Proofs (Main Text) - There appears to be some correspondence missing from the A&U Archive for the period 28/10/1954 - 23/11/1954.  Due to this the exact date when the the 35 copies were ready is not certain.  Jarrolds started the make-up around 18 October and the margin pages were approved on 29 November.  Typically this process took five to six weeks with the proofs being sent out a few days before the margins were approved.  If this was the case with The Return of the King then the page proofs were ready in the later part of November.  There is a mention of proofs in a letter from A&U to Tolkien dated 18 November 1954, but it is unclear which proofs are being discussed.  The J.R.R. Tolkien Companion: Chronology records that on 26 January 1955 Rayner Unwin may have delivered a set of page proofs to Tolkien in Oxford.  These were not 'proper' page proofs, but cut up galley proofs with printed headlines and page numbers.  How these unbound proofs fit in with the galley and page proofs noted above is unclear, but similar proofs also exist for the Appendices.  A complete set, including the main text and Appendices, was included in a Sotheby's sale in New York in 2002 (see References).

Travellers' Samples - Printed (and presumably bound) at the same time as the page proofs, these samples consist of the first two signatures of a book bound up for use by the publisher's sales representatives.  A standing order was placed in April 1954 for an extra 19 sets of the first two signatures of the page proofs to be printed and bound as travellers samples.  These samples were produced for The Return of the King, but there is no direct evidence of the number of sets supplied.

Galley Proofs (Appendices) - The date given is for the issue of complete proofs for all of the Appendices.  The issue of galley proofs for the Appendices of The Return of the King was not a simple matter due to the piecemeal fashion in which Tolkien submitted the copy.  The first section to be proofed was Appendix C (Family Trees).  This was sent to Tolkien on 18 November 1954.  Galleys of Appendix F and the Shire Calendar from Appendix D were issued on 2 March 1955, and these were followed on 4 April by a proof of one of the alphabets.  Complete galley proofs of all the Appendices were finally sent out on 29 June when one set was sent to Tolkien and another to Allen & Unwin.

Page Proofs (Appendices) - A&U received three sets of page proofs on 18 July, and sent one set to Tolkien and another to Houghton Mifflin.  An additional 35 sets were ordered, to be combined with the page proofs of the main text.

Advance Copies - The exact date when the advance copies were sent out is unclear.  They were due on 21 September, but on this date Jarrolds reported problems with the binding.  There is no sign of any complaint from A&U, so it seems that the advance copies appeared only a few days later.

1st Impression - Due to a misunderstanding between A&U and the printers, the text on the verso of the title page of the American Edition was not in accordance with Houghton Mifflin's requirements.  A&U offered to print a four page cancel fold, but Houghton Mifflin decided to print their own cancel title leaf as they did not wish to delay the publication of their edition.

4th Impression - The quantity printed is not noted in the records, but the quantity of paper used suggests 3,200 copies.  This is supported by a request to print a minimum of 3,000 dustwappers.  The quantities of A&U copies produced by the two binding orders are little better than guesses, the only certain figure being the 1,000 Houghton Mifflin copies. As printed, all sheets bore the A&U imprint.  In December 1958, 1,000 copies were converted to Houghton Mifflin copies by use of a cancel title.

5th Impression - As printed, all sheets bore the A&U imprint. In February 1960, 1,000 copies were converted to Houghton Mifflin copies by use of a cancel title.

6th Impression - The total quantity printed and the number of copies produced by the first A&U binding are unknown.  The total of 3,000 A&U copies was selected to match the 3,000 copies of The Fellowship of the Ring rather than the 4,000 copies of The Two Towers, because both FR and RK were reprinted before TT.  The quantity for the first binding was then derived by subtraction.

7th Impression - Details for this period are very sketchy, so it is possible that the total print quantity was higher than is noted above.  The number of A&U copies is unknown, but was undoubtedly small because the book was reprinted three months later.

8th Impression - As with the 7th Impression, details for this period are very sketchy, so it is possible that the total print quantity was higher than is noted above.  The number of A&U copies is unknown, but was probably small because the book had been reprinted three months earlier.

9th Impression - The figure of 3,500 copies for the first binding was calculated by subtracting the 1,500 copies for Houghton Mifflin and the second binding order of 5,280 copies (including overs and waste) from the 10,000 total print run - assuming that the 280 copies were the overs produced.

10th Impression - There is no indication in the records of when the second binding order was completed, so the date given is the same as that for the 1963 impression of The Fellowship of the Ring.  It is unlikely that the books were bound before March 1964, because at this time A&U decided to use some of the existing sheets for the Deluxe Edition.

11th Impression - The total print quantity is unknown, but was significantly higher than 10,000 copies.  A&U asked Jarrolds to print 10,000 copies of The Fellowship of the Ring and The Two Towers, and to use all of the remaining paper for The Return of the King.  RK was not reprinted with FR and TT in 1966 because sufficient stock remained.  The figure of 14,000 was chosen as this was the number of copies of FR and TT produced over the 1965 and 1966 impressions.  Quantities of 2,500 and 4,000 copies for the first and second bindings were chosen to tie in with FR and TT.  The quantity given for the third binding was then determined by calculation.

Summaries
The summary tables below list the approximate quantity printed for each impression - Total, A&U and Houghton Mifflin.  Where there is some uncertainty about the quantity this is denoted by a question mark (?).  Further details regarding these uncertainties can be found in the notes for each volume - FR - TT - RK.

3,000 copies of each volume were printed for Readers Union with the 1960 impressions, and 1,000 copies of each of the 1963 impressions were used for the A&U Deluxe Edition.  These have been excluded from the figures given in the summaries below.

In 1964 Houghton Mifflin decided to print all three volumes in the United States, therefore they stopped importing folded and collated copies from England.  The last copies exported to the United States were from the 13th Impression of The Fellowship of the Ring, and the 10th Impressions of The Two Towers and The Return of the King.

 
Summary - By Date
Houghton Mifflin imported more than one batch of sheets from some impressions.  These are recorded separately.