Hoyle's "Sixth" Edition and Progressive Ornament Damage

(updated 9/29/2012 with link to online edition of Maslen on Richardson)

Last month I was the successful auction bidder for an early Hoyle I'm quite excited about. The book is the "sixth" edition of the treatise on whist, published by Thomas Osborne, which I referred to as Whist.6 when I discussed it here. I have other copies of this work in Osborne collections (discussed here), but this copy was separately published.

Whist.6 title page
Levy [1761]

Whist 6 title page verso
with Hoyle autograph

What I like about the book is that it is in its original binding. Previously, I had written:

What did these Osborne's individual treatises actually look like as published? I've seen a very few of them in original bindings. They were sold in drab blue unprinted paper wrappers. They were stab sewn through four holes in the paper, rather than sewn through the folds. In nearly all cases, surviving individual treatises have been rebound some time in the past quarter millennium and the original appearance is lost.

My new copy, despite some minor paper repairs, is a rare survival of the original binding and confirms my earlier statement: the unprinted blue wrapper and the four-hole stab sewing. The sewing adds weight to my contention here, that one of my Osborne collections consists of separately sold treatises bound by the customer.

front wrapper

detail of stab sewing

The full imprint of the book is "Printed for T. Osborne, at Gray's Inn; J. Hildyard, at Hork; M. Bryson, at Newcaste [sic]; and J. Leake, at Bath. M DCC XLVI." The imprint does not reveal who printed it, nor is there a colophon identifying the printer. I was able to identify the printer through his use of woodblock ornaments such as the one appearing in Whist.6 on the first page of text below.

ornament from Whist 1746

The ornament provides a clue as to the printer. For a very small number of printers, there have been published studies their ornament usage:

• Maslen, Keith., The Bowyer Ornament Stock. Oxford, Oxford Bibliographical Society, Bodleian Library, 1973
• Maslen, Keith., Samuel Richardson of London Printer. A Study of his Printing Based on Ornament Use and Business Accounts. Dunedin: University of Otago, 2001. As of 9/28/2012, the book is freely available online.
• Goulden, R. J., The Ornament Stock of Henry Woodfall, 1719-1747 : a preliminary inventory London : The Bibliographical Society, 1988.

This ornament appeared in none of the books, so a lot of tedious work was required. I looked at reproductions of lots (and lots) of 18th century books where the printer had been identified. After some time, I was able to find the same ornament in a contemporary work, Life's Progress Through the Passions: or, the Adventures of Natura. London: Printed by T. Gardner, and sold at his Printing-Office. 1748. The author has been identified as novelist Eliza Haywood. The form of imprint "printed by" rather than "printed for" identifies Thomas Gardner as the printer of the Haywood work and thus as the printer of Hoyle's Whist.6. 

Osborne used Gardner to print many of the early Hoyle's and the same headpiece often appeared at the beginning of the text. What is striking is that we can see damage to the ornament over time. See the examples below, perhaps enlarging them in a separate window. In 1745 and 1746, it appears to be in good condition. In Whist.7, the ornament is fine at the beginning of the book, but when it was reused on page 72, the beginnings of a vertical crack are visible to the left of center. By March 1748, the crack is more pronounced in the "eighth" edition of Hoyle's Games. It is difficult to tell from the reproduction from Haywood's Life's Progress if there is any worsening by April. In 1750 the crack appears to be slightly larger in the "tenth" edition of Hoyle's Games.

Backgammon 1745

Whist "seventh" edition page 1

Whist "seventh" edition page 72

Hoyle's Games "eighth" edition March 1748

Haywood, Life's Progress, April 1748

Hoyle's Games "tenth" edition 1750

One could expand this study by identifying and looking at all of Gardner's output during this time period. I can only imagine how much work Maslen and Goulden must have done to produce ornament inventories for the printers they studied.

I noted earlier that the "tenth" edition was a reissue of the "eighth" with the same setting of type. There appears to be more damage apparent in the 1750 "tenth" edition as compared with the 1748 "eighth."  Obviously this suggest that the pictured copy of the "tenth" was printed later than the copy of the "eight." Was there a single continue printing operation or multiple impressions from the same set of type?

A Final Look at the "Twelfth" Edition

In the past several essays, I have looked in detail at the "twelfth" edition of Hoyle's Games. The first described what must be the authorized edition (A), as it is always autographed by Hoyle and Osborne. A second essay noted common elements among the three variants, (B), (C), and (D), and concludes that absent the autographs they are piracies. This essay will focus on the differences among the four books, concluding that they are four separate settings of type and therefore, four distinct editions.

First, some overall comments about the structure of the books, as  presented in the chart below. 

	(A)	(B)	(C)	(D)
chain lines	horizontal	vertical	horizontal	vertical
leaves per gathering	12	6	6	6
dimensions of largest copy seen	14.6 x 8.7	15.4 x 8.4	16.3 x 10.1	15.8 x 9.1

While all the books are duodecimos, only (A) is in twelves. One normally expects duodecimos to have horizontal chain lines, but there is a mixture here. Note that (C) and (D) are larger than (A), with (B) falling somewhere in the middle. Recall that Osborne advertised the book as "a small genteel pocket volume."

To determine if these books are four settings of type, we will first focus on pages 212 and 213, as those pages highlighted the differences that Marshall noted in Notes and Queries. As the earlier essays pointed out, page 212 was a logical end to the book, concluding the treatise on backgammon and having an erratum at the bottom. Page 213 is the late addition of "two new cases at whist, added since this book was printed off." You will need to click on the images to enlarge them (perhaps in a separate window) to follow the discussion below.

Variant (A) Games.3 Levy [1118]	Variant (B) Games.3.P.3 Levy [1634]
Variant (C) Games.3.P.2 Levy [1632]	Variant (D) Games.3.P.1 Levy [1119]

	(A)	(B)	(C)	(D)
ornaments	yes	yes	no	no
errata	yes	no	no	no
punctuation after law number	period	period	comma	period
"Man" in law 1 (212)	"M" is above "p" in "played" in line below	"M" is to the left of "p" in "played" in line below	"man" is not capitalized	"M" is above "p" in "played" in line below
"Parties" in last line of law 5 (212)	capitalized	capitalized	not capitalized	not capitalized
word on last three lines of first new case at whist (213)	nei- Cards win	neither in 6	If 3 to	If 3 to
second to last line (213)	"and that A has 4..."	"A has 4..."	"A has four..."	"A has 4..."

Looking only at the ornaments and the errata as Marshall did is not completely conclusive. It is possible that the type was left standing, but somebody needed the type ornaments for another printing job and they were removed. Similarly, the errata could have been corrected in a re-impression from the same type. However, we can look at differences in capitalization, line breaks, and spelling to conclude that we are looking at four distinct settings of type.

On page 212, only (A) and (D) have "Man" and "played" aligned similarly, but (A) capitalizes "Parties" in the last line, while (D) does not. So there are four distinct settings of page 212. We can reach the same conclusion about page 213 by looking at the ends of the lines of the first case at whist. Only (C) and (D) are the same, but (C) spells out "four" on the second to last line, while (D) uses a numeral.

I used these pages, because they are the ones noted by Marshall, but really the same analysis must be done for each gathering—in fact for each forme within a gathering. See "Pirates, Autographs, and A Bankruptcy" for an example (Whist.2.1) where some gatherings were reset and others not, and another (Whist.3 ve. rsus Whist.4) where some formes were reset and others not.

I have gone through the four variants gathering by gathering. The work would be much easier with a Hinman collator, but absent that I have convinced myself that no two are from the same setting of type.We are left with four editions, and with the exception of (A) which must have been first, I have not been able to determine priority. Indeed, it is not clear that (B), (C), and (D) were released in 1763 along with (A). Recall from the last essay that Brambles and others reprinted this text as late as 1807!

More on the "Twelfth" Edition

In the two essays I have discussed the authorized "twelfth" edition of Hoyle's Games (1760) and three pirated variations of the same book. I plan to publish an analysis of the three variants to determine whether they are three separate editions or re-impressions from a single setting of type. First, I want to raise a couple of topics connected with those two essays.

An earlier piracy?

I the last essay, I argued that the three variations are piracies because they lacked the autographs of Hoyle and Osborne and because they were not printed in the usual format of the Osborne Hoyles. I mentioned what may be a similar example of piracy at the bottom of this essay. A single copy of the "tenth" edition of the individual treatise on Whist is at the British Library The authorized "tenth" edition (actually a reissue of the "eighth" edition, as discussed here) was a collected edition of all of Hoyle's treatises. I believe the individual treatise to be a piracy because it was not autographed by Hoyle (Osborne was not yet signing any of Hoyle's work) and because it was an octavo, where all the contemporary genuine Hoyles were duodecimos.

The Scottish Hoyle

In two essays (here and here) I have discussed an edition of Hoyle printed in Edinburgh by Mundell and Sons. Here I want to highlight that the Scottish Hoyle is textually identical to the three variants we have been discussing. The verso of the title page contains only the "Advertisement" and not the "To the Reader." It is not autographed by Hoyle or Osborne. It lacks the errata at the end of the book, but contains the "Two New Cases at Whist" as the last leaf of the book. While the Scottish Hoyle has all the characteristics of the piracies, I believe it was authorized. Why would a printer put a colophon on a piracy? The lack of an autograph is more likely to do geography than piracy.

A nineteenth century reprint

The "twelfth" edition was followed by a "thirteenth" with additional text in 1763. As Hoyle came off copyright in 1774 (discussed here), new editions of Hoyle appeared with the text of other writers treating other games (discussed here). Thus it is puzzling that a reprint of the "twelfth" edition appeared in 1807, 44 years later, with none of the improvements of the intermediate editions.

Games.3.R.2
Levy [0085]

The book is printed for J. Brambles, A. Meggitt, and J. Waters, by H. Mozley, Gainsborough. 1807. It appears to be quite rare with only one copy listed in WorldCat in the John G. White chess collection at the Cleveland Public Library and a second listed in Copac at Oxford. It seems surprising that the book is listed in the three major Hoyle bibliographies (Horr, Jessel, and Rather, discussed here), but I suppose its inclusion is due to a coincidence of geography. 

Cleveland attorney and gaming bibliographer Norton Horr must have seen White's copy (or White's catalogue which must have listed the book) and was able to include it into his 1892 Bibliography of Card-Games and of the History of Playing-Cards as item 681. Frederic Jessel, who's collection is now at the Bodleian Library at Oxford, listed the book as item 826 in his 1905 Bibliography of works in English on Playing Cards and Gaming, but made it clear that he had not seen the book and was relying on the description in Horr. The Oxford copy is not catalogued as part of the Jessel collection and must have been acquired after Jessel wrote his bibliography. Rather, of course, would have had access to the White collection and the Jessel bibliography.

Two new cases at whist

Brambles reprinted the "twelfth" edition completely, even including the two new cases at whist on the final two pages of the section on backgammon. They removed the phrase "added since this book was printed off" and corrected the errata in the text. Note the colophon at the bottom of page 195, shown at right.

How did Brambles, Meggitt and Waters come to reprint a work almost half a century old, particularly one that had been superseded by so many editions of Hoyle's Games Improved? The Brambles Hoyle is a small book, 11.3 x 7.8 cm, with horizontal chain lines and no visible watermarks. It is likely an 18^o or 24^o, a cheap reprint. I looked through WorldCat and found about four dozen titles issued by the three publishers in the first decade of the 19th century. They all seem to be reprints of rather stale 18th century titles: Aesop's Fables translated by Croxall, Dodsley's Economy of Human Life, Gay's Fables, Elizabeth Moxon's English Housewifery, Defoe, and...Hoyle. We can infer that their business model was to reprint cheaply books in the public domain.

Still, shouldn't they have used a more recent text?

Other "Twelfth" Editions

Last essay, we looked at the autographed "twelfth" edition of Hoyle's Games. In this essay we look at other printings, two of which were identified by Hoyle bibliographer Julian Marshall:

[The twelfth] edition was reprinted (B) without the signatures (autogr.) of author and publisher on the face and back of the title; but they are printed on the verso, as before, at the end of the "Advertisement." "To the Reader" is omitted, and the first page of the "Contents" is numbered iii. The "Two New Cases" appear on pp. 213 and 214, but not the list of errata, for the errata have been revised in the book. This shows it to have been printed later than the former (A) issue...In another variation (C), on larger paper, the author's and publisher's (autogr.) signatures and the errata are again wanting, and the lines of ornament at the top of pp. 212, 213, are omitted. (Notes and Queries, 7th ser. ix, 25, January 11, 1890)

Marshall's analysis is based on his examination of his own copies of the autographed edition and variant (C) and copies of (B) at belonging to the British Library (then British Museum) and Henry Jones ("Cavendish", discussed here).

I have seen a half dozen copies of (A), three each of (B) and (C), and five copies of another variant (which I shall call (D)), not described by Marshall. Before focusing on the minute differences among the variants, let's examine the commonalities.

Marshall has touched on most of the important:

• The variants are not signed by Osborne or Hoyle, though they both contain the Advertisement noting that "no copies of this book are genuine, but what are signed by us. Edmond Hoyle, and Thomas Osborne."
• The variants lack the note "To the Reader" (pictured in the last essay).
• The errata have been corrected in the book.

(B) Title Page
Games.3.P.3
Levy [1634]

(B) Title Page verso

Marshall omits the most salient point that the three variant printings share a physical structure, though one that differs from the authorized edition. Last essay, I noted that the autographed version collates 12^o: π1 A⁶ B-I¹² K⁸ L² χ1 = 114 leaves and the pagination is pp. [2] 1 ii-x [2] [1] 2-86 [87-9] 90-115 [116-9] 120-152 [153] 154-168 [169-171] 172-214 = 228 pages. In discussing the contents, I noted that both structurally and textually, the final leaf χ1 was added after the bulk of the book had been printed. The structural argument is that the leaf is an insertion outside of a gathering. The textual argument is that an erratum appeared on L2v (p212), suggesting it was originally the final leaf. More importantly, χ1 contains new cases at whist "added since this book was printed off" with more errata on the verso (p214). The new whist cases appear with the section on backgammon, not the section on whist.
 
The variants all collate 12^o: A-S⁶ T⁶ (T6 missing, blank?), so they are duodecimos in sixes rather than twelves. In all the copies I have seen, the final gathering is five leaves only. The sixth, which must have been present during printing, is perhaps blank or contained a leaf for another book. There are 18 gatherings of 6 leaves ('J' is not used as a signature) plus 5 leaves in gathering T for a total of 113 leaves, one fewer than the authorized edition. The pagination of the variants is [i-ii] iii-xii [1] 2-86 [87-89] 90-115 [116-119] 120-152 [153] 154-168 [169-171] 172-214 for a total of 12 + 214 = 226 pages, two fewer than the authorized edition.

The text of the variants is almost a line-for-line resetting of the autographed edition. What accounts for the missing leaf? The variants lack a section title for whist, which appeared on the last of the seven leaves of the preliminaries. The other six leaves of preliminaries are in the A gathering. In all books, the autographed edition and the variants, B1r begins the text of the whist treatise.

Marshall is correct that the variants were all printed after the autographed edition. The corrected errata are conclusive, as is the structure with the new cases on pages 213-4 integral to the final gathering, rather than an insertion. What Marshall doesn't address is whether the books are piracies. I believe they are, because:

• The books are not autographed by Hoyle and Osborne. Hoyle was under contract to autograph all of his books and Osborne would have insisted upon it—it was a major component of his marketing, as evident in the advertisement I quoted last essay. 
• After 1747, all of the Osborne Hoyles were duodecimos in twelves, not in sixes. 

The next question, and one that will wait for another essay is whether these three variants are three different settings of type and hence three separate editions, or whether they are different impressions from the same setting of type. A Hinman collator would be the best tool to answer that question, but I will attempt manual comparison.

The "Twelfth" Edition

Julian Marshall, the first Hoyle bibliographer (discussed here), noted that there appeared to be multiple printings of Hoyle's "twelfth" edition. In Notes and Queries, (7th ser. ix, 24-5, January 11, 1890), he describes what he calls variant (A) in full detail, transcribing the title page, and noting Hoyle's autograph at the foot of the title page and Osborne's on the verso. He provides a collation, describes the contents, and notes changes in the text since the "eleventh" edition. He identifies two other printings which he calls (B) and (C), and I have discovered a fourth.

In this essay and the next, I will look at these four books in some detail, trying to uncover how they came to be. This essay will focus on (A), the only one of the four which is autographed by Hoyle and Osborne, and thus the one that is clearly authorized.

The story of the "twelfth" edition begins with an advertisement appearing in The London Evening Post of December 23, 1760:

"This day were publish'd, beautifully printed on a fine Paper, in a small genteel pocket volume, (price only 3s. neatly bound) the twelfth edition, to which are added, two cases in whist, never printed before; and the new laws of the game, as play'd at White's and Saunders's Chocolate House: Mr. Hoyle's Games Compleat...printed for T. Osborne...S. Crowder...and R. Baldwin.

"Be pleased to observe what you buy are signed by Edmund Hoyle and Tho. Osborne, all others being pirated, or a bad edition, and for which a reward is given to any who will inform of the sale of them."

The advertisement notes three things that we will keep in mind as we examine the physical books. The size ("small genteel pocket volume") is relevant because, as Marshall notes, (C) is printed on larger paper. Second, the advertisement notes changes in text ("new cases" and "new laws"). Finally it cautions that unsigned editions are piracies.

Of the half dozen copies I have seen of (A), the largest is 15.5 x 8.7 cm., which seems to be the same size as the "eleventh" edition published in 1756. It appears to be smaller than the "eighth" edition and its many reissues (as discussed here, the reissues are stated "ninth" and "tenth" editions), which can be as large as 16.6 x 9.7 cm.

Title page
Levy [1118]

Verso of title page

The title page is for the collected edition, containing separate treatises on whist, quadrille, piquet (including chess), and backgammon. As we shall see, each treatise also has its own title page, even though none was published separately. It is odd that the book is noted as a "twelfth" edition. Hoyles are numbered, more or less, by the edition of the whist treatise. For example what is called the "eighth" edition of Hoyle's Games lacks an overall title page, but the whist section title is noted a stated "eighth" edition (Osborne.1748 discussed here). It the same essay, I note that the "ninth" edition (Reeve.1748, discussed in the same essay) has a collected title The Accurate Gamester's Companion, "ninth" edition, though Whist is still the "eighth." With Osborne's "eleventh" edition, the statement of edition is on the collected title only; the section title for whist has no statement of edition, although the other treatises do. This "twelfth" edition continues that pattern: a stated edition on the collected title, with no stated edition on the section title for Whist (pictured below).

Unlike Marshall's copy, the autographs of Hoyle and Osborne are both on the verso of the collected title. From the copies I have seen, this is the more typical arrangement.

The book collates 12^o: π1 A⁶ B-I¹² K⁸ L² χ1 for a total of 1 + 6 + 12 * 8 + 2 + 1 = 114 leaves. π1 refers to a single preliminary unsigned leaf. χ1 refers to a single leaf later in the book.

The pagination is pp. [2] 1 ii-x [2] [1] 2-86 [87-9] 90-115 [116-9] 120-152 [153] 154-168 [169-171] 172-214 for a total of 2 + 10 + 2 + 214 = 228 pages. The bracketed numbers refer to unnumbered pages. Where we can infer page numbers, they are shown in roman number in brackets; where we cannot, the page count is shown in italicized numbers in brackets.

Let's match the contents of the book with page references and page numbers. It will be easy to see why pages were unnumbered and we will see how the physical structure of the book relates to its contents and reveals something about how it was written.

Signature Reference	Page Reference	Contents
π1r	[1]	Combined title page (pictured above)
π1r	[1]	Advertisement and To The Reader, autographed by Hoyle and Osborne (pictured above)
A1r-A5v	1 ii-x	The Contents
A6	[2]	Section title for Whist (with no statement of edition); verso blank
B1r-E7v	[1] 2-86	Whist Treatise
E8	[87-88]	Section title for Quadrille; verso blank
E9r-F10r	[89] 90-115	Quadrille Treatise
F10v	[116]	blank
F11	[117-8]	Section title for Piquet; verso blank
F12r-H4v	[119] 120-152	Piquet Treatise
H5r-H12v	[153] 154-68	Some Rules and Observations for Playing Well at Chess (no section title as Chess is included in the section on piquet)
I1r	[169-70]	Section title for Backgammon; verso blank. While the section title says "fifth" edition, that designation is meaningless.
I2r-L2v	[171] 172-212	Backgammon Treatise with an erratum at the bottom of page 212
χ1	213-4	Two New Cases at Whist, added since this book was printed off, with 11 errata at the bottom of page 214

The pattern of the page numbering is clear: section titles, their blank versos, and the first page of each section are unnumbered. The rules and observation on chess were always a part of the treatise on piquet and lack a separate title page, but still the first page of text is not numbered.

Interestingly, the book seems entirely complete without the final leaf χ1, pages 213-4. Physically, it is a single leaf outside the final gathering. The text supports that view as well. The erratum at the bottom of page 212 suggests the book is complete and an error was noted during printing. The advertisement and collected title page promise two new cases at whist which appear in chapter 17 on pages 64-5 (discussed here). χ1 contains two further cases at whist ("since this book was printed off"), not noted in the advertisement or on the collected title, awkwardly placed after the section on backgammon, rather than with the other material on whist. The errata on the bottom of χ1v suggest a second round of proofreading. Everything suggests that this leaf was a late addition to the book.

We have noted, beginning with the "thirteenth" edition, the material on χ1 migrated to the section on whist. As we shall see in the next essay, the oddities of the "twelfth" appear in other printings, although the physical structure is different.