University of Virginia Library

JAMES SHIRLEY'S TRIUMPH OF PEACE:
ANALYZING GREG'S NIGHTMARE
by
STEPHEN TABOR

THE surviving exemplars of James Shirley's Triumph of Peace (London, 1633/4)
have defeated the attempts of generations of bibliographers to classify
them. Any two copies are likely to show some pages with identical typesettings,
some with partly different settings, and some entirely different. The title page
comes in four variants of authorship and edition statements (fig. 1):

• By James Shirley, of Grayes Inne, Gent.
• By James Shirley, Gent.
• By James Shirley, of Grayes-Inne, The third Impression.
• By James Shirley, of Grayes-Inne, Gent. The third Impression.

These states have an obscure connection with the text pages that follow, whose
dozens of permutations require one to specify the state of each page in or-
der to fully characterize a copy. The normal hand-press taxonomy of edition,
impression, issue, and state is inadequate to sort the copies into well-behaved
families.

As convoluted as The Triumph of Peace (henceforth "ToP") may be, part of the
reason for bibliographers' difficulty in coming to grips with it was the unavail-
ability, until the 1960s, of a technology that could supply quick and cheap repro-
ductions of complete copies for thorough comparison. W. W. Greg, who came
closer than anyone to understanding the book, died in 1959, the year the Xerox
machine was introduced. (He would have had to wait another half-century for
the silent invasion of library reading rooms by digital cameras and smartphones.)
In 1946, he published a paper entitled "The Triumph of Peace: A Bibliographer's
Nightmare". 1 His chief purpose there was to point out inconsistencies in the
account of the work by William Jackson and Emma Unger in the Pforzheimer
catalogue published several years earlier. 2 However, he concluded that "a final
solution will have to wait till a bibliographer equipped with the necessary skill
and patience finds the leisure and the opportunity to make a thorough scrutiny of
all available copies." 3 The descriptions of ToP that Greg published five years later
in his Bibliography of the English Printed Drama to the Restoration showed no further

[Description: FIGURE 1(A). Title state 1: Huntington Library, San Marino, California. RB 69433.]

[Description: FIGURE 1(B). Title state 2: Houghton Library, Harvard University. STC 22459a.5.]

[Description: FIGURE 1(C). Title state 3: Folger Shakespeare Library STC 22459b.2.]

[Description: FIGURE 1(D). Title state 4: Folger STC 22459b c.1 (bd. w/4629).]

progress. 4 He did agree with Jackson and Unger that two initial "duplicate"
text settings were printed more or less simultaneously in expectation of heavy
demand, and that the resulting quires were indiscriminately combined in copies
with title pages 1 and 2. 5 We shall see that both of these conclusions were wrong,
but my purpose here is not to analyze the achievements and errors of previous
investigators. 6 To get efficiently at the many problems that ToP poses, we need
to start with a clean slate.

The Triumph of Peace was a masque produced by the Inns of Court for the
court of Charles I. It was conceived, fittingly, as a peace offering. In 1633, William
Prynne, a member of Lincoln's Inn, had published Histrio-mastix, a puritanical
indictment of the theatre that, among a multitude of objections, criticized women
for appearing on stage. Archbishop William Laud, looking for an excuse to si-
lence Prynne, contrived to frame this as a personal attack on the Queen Consort Henrietta Maria, who was known occasionally to breach this barrier in private
performances. Prynne was famously hauled into Star Chamber for sedition and
sentenced to have his ears cropped, among other penalties. But he had dedicated his book to the Inns of Court, thus implicating his own professional organization
in an act of lése majesté that, as both sides realized, called for a display of contrition.
The barristers loosened their purse-strings, and the masque they offered is chiefly
famous today—its bibliographical complexities apart—for being the most lavish
production staged in England at least until the Restoration. Performed on 3 Fe-
bruary 1634, 7 it commenced with a very public torchlight procession from the Inns
of Court down Chancery Lane and the Strand to Whitehall. The Venetian ambas-
sador reported home, "Their display at the palace with a numerous, stately and

glittering cavalcade, by their dresses, liveries and devices, attracted a great crowd,
exciting the curiosity and applause of all the people ...". 8 The whole spectacle was
funneled into the Banqueting House for the private performance featuring sets by
Inigo Jones and music by William Lawes and Simon Ives. The result so pleased
the Queen that Charles requested a second performance, at further expense to the lawyers (or perhaps this was his real motive), ten days later at Merchant Taylors'
Hall. 9 This shift of venue entailed some hasty demolition and construction to clear
an approach for the royal couple, and residents along the parade route were re-
quired to mount torches on the front of their houses and provide security for the
whole night. The combined cost of the two performances came to very roughly
£20,000–at least $4 million today, by a conservative estimate. 10 Whatever the
literary merits of Shirley's text, the extravagance and unusually public aspect of
the productions created a heavy demand for the printed record, both within and
without the court. 11 The book of the masque bears witness to this demand;
the bibliographical evidence I present here will reveal a printing house running at full
tilt, distracted by other jobs and grasping for short-cuts.

The surviving financial accounts for the production of the masque record
payments to a "Mr. Wakelye" (variously spelled), whom they call "the printer". 12
This was the publisher Thomas Walkley, who until at least 1632 had a shop in
Britain's Burse, quite close to the Inns of Court. By early 1634 he had moved
nearer to Whitehall, but he seems to have retained his connections to the legal
community. Walkley sub-contracted the actual printing of ToP to John Norton,
who had printed works by Jonson and Massinger for Walkley in the previous few
years. And he entrusted the distribution of copies to William Cooke, who pub-
lished all six of James Shirley's plays issued in 1632–35 and many more besides. 13
Cooke's shop at Furnival's Inn Gate was a nexus of literary activity associated
with the Inns. Shirley himself lived near Gray's Inn, and on 17 January 1634–no
doubt as part of his compensation for writing The Triumph of Peace—he was ad-
mitted to honorary membership in the Inn as "one of the valets of the chamber
of Queen Henrietta Maria, 'absque fine'". 14 Three of the four title-page states of

ToP record this honor. So, the production and distribution of the printed book
was very much an inside job, carried out by men with close ties to the Inns of
Court and their associated literary coteries.

The book that Norton produced is a quarto that collates a2 A-D4, with pagi-
nation [i-iv], 1–8, 21–24. The first half-sheet bears the title (verso blank) and a
leaf of dedication by Shirley. Quire A describes the procession before the per-
formance, and the decoration of the Banqueting Hall; quires B-D, with their
own pagination sequence, contain the text of the masque. Four surviving cop-
ies contain an additional singleton leaf of verse headed "A speech to the King
and Queenes Maiesties, when they were pleased to honour the City with their
presence, and gave a gracious command, the former tryumph should attend
them." Its eighteen lines, delivered by a character called "Gen[ius].", allude to
the masque's second performance on 13 February.

THREE PUZZLES

§1. Main Settings

Table 1 shows, in shades of gray, the distribution in surviving copies of the
completely different settings of each page. The rows of the table correspond to
the copies or reproductions I have examined, grouped by the four title-page vari-
ants which are separated by the horizontal white bands. The columns represent
the pages of each copy grouped by quire and forme. One can immediately see
what perplexed the earlier bibliographers. Some settings persist throughout the
print history: the outer formes of quires A and C, for instance, and even the
title page (a1r), remain substantially the same with relatively minor variants.
Other pages go through as many as three settings. Most of the new settings do
not line up with changes on the title page. The concept of "edition" as applied
to the whole book, and the normal role of the title page as the primary taxo-
nomic organizer, go out the window here. So does any hope of dealing with the
book quirewise, or even formewise. Only a page-by-page analysis will allow
us to specify the relationships among copies.

As it happens, this patchwork of settings (taken in conjunction with the vari-
ants they contain) works in our favor when we set out to put all the states, from
major to minor, into chronological order. Let us take the example of quire A
inner forme, where page A2r goes through two partial resettings, A1v gets one
complete resetting, and the other two pages exhibit only stop-press corrections.
In table 2, the heavy horizontal line separates the completely different settings
of A1v, while the lighter lines separate partial resettings or more minor altera-
tions. The relative positions of the lines between columns show the points of
introduction of new states, either simultaneously with changes on other pages or
independently from them. We have no immediate basis for deciding the order
in which some of these occurred. However, since the forme is the basic unit of
printing, these four pages must all live their lives in the same direction: a prov-
able sequence of variants on one page imposes an obligation on the sequences
of the others.

A4r, in its single setting, shows progressive improvement in spelling and
punctuation. In the first wave of correction, "sharpe sited" becomes "sharpe
sighted" and "Caduseus" becomes "Caduceus". In the next wave, the addition of
commas to a listing of architectural features helps to improve the sense. On A3v,
the correction of "Arber-worke" to "Arbor-worke", occurring part-way through
the second state of A4r, follows the same order. If we accept this sequence of
states for these two pages, the other two (A1v and A2r) must follow suit. There-
fore, setting 1 of both A1v and 2r does indeed come first, a fact that we would
otherwise have difficulty proving. Setting 2 of A1v then goes through two waves
of correction that lock partial resettings 2 and 3 of A2r into order. When, some
twenty-five years ago, I began to sort through all of the states of ToP, I was pleas-
antly surprised to find most of them clicking into place pretty readily in this same
way. I have recorded the rationales for the order of states in appendix 1, though
I would not encourage the reader to venture into that thicket quite yet.

The important thing to note in table 1 (the table of editions of the whole
book) is that there appears to be an orderly progression of settings of the constitu-
ent sheets. The various settings of one quire do not mix randomly with settings
of other quires. As the quire-breaks always fall at the same points in the text in
all the states, copies could in theory have been assembled from sheets of what-
ever combinations of settings were available at the gathering stage. The lack of
random mixing implies discrete stages of production and collation, and provides
one argument against the theory of simultaneous setting that has prevailed until
now. Only the late states of quire D show a more promiscuous behavior; I
will explain this in §8B.

The title-page variants, though seeming to dance to their own drummer, at
least dance in the same direction as the variants in quires A-D. In the earliest
copies of the title (fig. 1[A]), the "O" in "OF" in line 3 is sound. A small accident
soon dents the top of the O, and this damaged type persists through all of the
subsequent states. Reassuringly, the "third impression" title pages are found only
in copies having the latest states of the text, and we can see that the unique copy
of title 3, with its dangling comma after "Grayes-Inne", is just an uncorrected
version of title 4. Quire "a", therefore, shows a rational order of variants that also
tracks the evolution of the other quires, even if the title-page changes are out of
phase with those of the text pages.

Although, after we become familiar with table 1, we discern some comforting
regularities, we are left with the question of why some pages were reset and others
were not. The mystery deepens when we look more closely at the pages which
retain at least 50% of their original settings. TOP shows the normal pattern of
stop-press corrections found in most books of the hand-press period. But we also
encounter two phenomena which are unusual: (1) pages on which only the page
number is reset, where no correction was required, and (2) pages on which blocks
of text in a particular type style (usually italic) are reset—again with no substan-
tive changes to the reading—while the other portions remain untouched.

The seemingly capricious page-number resettings turn out to be a powerful
tool for reconstructing the sequence of events in John Norton's shop during the
printing of ToP. Once we understand what motivated them, we will be in a better

TABLE 1. Distribution of completely different settings

 

The three shades of gray indicate, from light to dark, one to three settings.	Clear cells indicate missing leaves. No attempt is made here to show partial resettings.

TABLE 2. Complete and partial resettings in forme A(i)

position to interpret the equally puzzling replacement of blocks of text on some
pages.

§2. Headlines

ToP contains no running titles. Except for the words "The Epistle" on a2v,
the headlines consist of simple page numbers within parentheses—three or four
types plus spacers. Figure 2 shows two states of a page (C1r) in which the headline
has been reset. This is not just a spacing shift—all three printing characters are
different types. But the setting below is unchanged, and even the direction lines
are unaltered (both contain the catchword "An_" with a dropped hyphen). On
some other pages, both the numbers and the direction lines show resetting—
for example B3r in Houghton STC 22459 as compared with the same page in
Houghton STC 22459a.5. Still others have page-number resettings in conjunc-
tion with stop-press corrections. On A1v, Bodleian Mal. 160(3) has three errors
that are corrected in National Art Library 25.c.79. 15 The priority seems clear.
However, along with the corrections comes a reset page number: originally well
printed, the "2" in the apparently later state is badly damaged and the left pa-
renthesis prints much thinner than the right. The nature of the damage to the
numeral makes it unlikely that we are looking at the same unhappy type, and
the parentheses are also clear substitutions. In the act of improving the text, the
printer seems for no discernible reason to have swapped out a good page number and substituted a bad one.

Finally, a new headline always appears in those cases (mentioned in the pre-
vious section and discussed more fully in the next) where selected blocks of text
are reset seemingly without need. Headlines are reset even when they are not
adjacent to the reset text. Huntington copies 69434 and 69428, for instance,
show this phenomenon on C2v.

Most of these headline replacements also entail a reset direction line, though
the latter can be difficult to detect. Of the two features, I have found the page
numbers to be a clearer first indicator that something unusual is going on, for

[Description: FIGURE 2. Headlines on page C1r of Huntington RB 69428 (top) and Huntington RB 69432.]

three reasons: unlike the catchwords, they are letter-spaced, and seldom the same
way twice; their horizontal alignment with the first line of text is variable; and
the numbers and parentheses in Norton's type-cases contain a variety of forms
and damaged sorts that make substitution easy to detect.

The page-number changes in ToP are so clearly without purpose that they
must have been unintended consequences of other operations. We can explore
their patterns by modifying table 1 to show where the substitutions occur. In table
3, the double horizontal lines indicate page-number changes. 16 These obviously
occur also at the points where whole pages are reset (still shown by changing
shades of gray), but others divide the main settings into smaller groups. A reset
number on one page usually accompanies a similar operation on its forme-mates,
and the table shows corresponding fault-lines that often break out of individual
formes into whole quires and even to the other sheets. The initial impression of
chaos in table 1 begins to yield to a pattern that is more complex, but also more orderly.

The page numbers in ToP belong to an element common to most books
of the hand-press period called the skeleton forme. This consisted of non-printing
material such as the furniture that held the pages in a set layout on the press
bed, but it also included the headlines and direction lines. The latter contained

TABLE 3. Correlation of page number resettings with text resettings

 

Heavy horizontal lines show points of text resetting; horizontal double lines show points of page number resetting. Blank cells signify missing leaves.

type-matter that did not vary much from forme to forme; most of it was spacers.
So, the skeletons tended to get re-used, with changes to the visible type as needed,
in successive formes as an edition went through the press. When a forme was
finished printing, the units of the skeleton were normally separated from the type-
pages and used in another forme, while the stripped pages went back to the type
cases for distribution. 17 After the edition was finished, the type of the skeletons
would go down the same path.

It was entirely possible to do multiple printings of a short book using the same
setting of type. In such a scenario, the type-pages, after their first printing, would
be tied up with string, with or without their skeletons, for an arbitrary period,
then reimposed and reprinted. This cycle could repeat itself as many times as
needed. Printers commonly stripped the headlines and direction lines when pages
were finished printing, but if the skeletons happened to remain with their pages, a
reprint could easily escape detection even by a modern bibliographer. Numerous
examples of reprinting from standing type, at least for stretches of a few pages,
are documented from the early seventeenth century and before, 18 and the prac-
tice became rather common by the late eighteenth century. The reasons it was
not more widespread in the 1630s were both practical and regulatory. Between
1587 and 1637, an order of the Stationers' Company set a limit of 1500 copies
on print runs from a single setting. 19 If more copies were wanted, a scrupulous
printer would have distributed the type and reset a new edition from scratch.
This rule was intended to provide sufficient work for journeyman compositors,
but it also made economic sense. Because hand-cast type was expensive, a printer
would keep only limited amounts and would not want large chunks of it tied up
and unusable for other jobs. Nevertheless, printers interpreted the regulation
flexibly, and when faced with a sudden high demand for a particular title they
were understandably tempted to exceed the limit. 20

If we encounter a book whose copies show only minor changes to the text
but two entirely different settings of the running titles, we have a strong case
that the edition went through two impressions. In the interval between them,
the skeletons got distributed and had to be reset for the second run. This is
what I believe happened repeatedly with The Triumph of Peace. I would contend

that nearly every instance of arbitrary page-number substitution resulted from
an interruption of work during which the whole forme was removed from the
press. The type-pages were set aside for a time and then received at least reset
page numbers, and usually new direction lines as well, in preparation for re-
impression. Some pages underwent a complete resetting during the interval, of
course including the skeleton. This is the explanation of the horizontal fractures
in table 3. It is even possible that some pages were stored with their skeletons
intact and went undetectably into later impressions. This could have occurred
with formes D(o) and A(i) at certain points in the printing history; it is tempting
to posit "silent" reimpositions that would bridge the discontinuous fault-lines in
table 3. We will explore this possibility when we go into the life-histories of the
individual quires.

§3. Italic Raids

Figure 3 shows a sort of mutation frequently found in ToP: the entire song
(composed in italic type) has been reset, but not the passages in roman above.
Note that short stretches of italic embedded in roman text remain unchanged.
Eleven pages in ToP show this localized resetting of italic at some point in the
printing history, and one of them (D2r) has part of the italic reset yet again. Two
pages (including one that previously had the italic changed) also show an arbi-
trary change-out of some of the roman type.

Table 4 shows where these resettings occur. (Solid arrows indicate replace-
ment of italic type, dashed arrows the replacement of roman.) All of them, except
for the roman replacement on C1r, occur at points in the production cycle where
the numbers in the headlines were also reset—in other words, during the pauses
between impressions that we have already identified. Table 5 presents the history
in another way. (It contains codes for production phases and variant states which
I will discuss later.) This table shows all the pages of ToP that contain concentra-
tions of italic. A great deal of that type is used for the songs that cluster toward
the end of the masque.

The local resettings do not always draw on the same font. In the earliest states
of ToP, all the text—both roman and italic—is set in "english" size (20 lines =
96 mm). In the first wave of changes, the italic Song 5 on C4V reappears in pica
size (82 mm). During the same transition, four lines of italic stage direction on
A2v-3r switch to a 120 mm great primer. Pages C3v and D3r are a more subtle
case: the reset songs are still in 96 mm italic, but the font is of a different design,
which I will call i-2 as opposed to the more prevalent i-1. The differences are
hard to describe, but we can compare the two in figure 3. The page on the left
features Norton's regular english italic i-1, while that on the right shows the song
reset in i-2—a less attractive face, smaller on the body and also (in its ToP out-
ing) badly worn. The most diagnostic feature of i-2 is the h, whose legs nearly
close at the bottom, so that the letter resembles a b. In its i-1 cutting, the right
leg only begins to close.

None of these partial resettings result in substantive textual changes. Like the
page-number resettings, they bear the marks not of perfectionism but rather of

[Description: FIGURE 3. Block of italic text reset: page D3r in Huntington RB 69428 (left) and Huntington RB 69432.]

accidental contingencies of unrelated operations. Their timing, arbitrary nature,
and recourse to alternative, wrongly-sized, and sometimes inferior fonts strongly
suggest that they were a response to a shortage of type, that Norton was juggling
other jobs, and that ToP had to stand aside—even sacrifice some of its stand-
ing type—to let those jobs "play through". I therefore refer to these incidents
of localized distribution as "raids", which created holes in the composition that
Norton refilled once the more pressing need was past.

In §9, I will show that some of the italic distributed in the first raid ended up
in ToP itself, in the two pages of dedication (leaf a2). As part of the preliminaries,
these pages would have been printed after the main text if Norton was following
the usual practice. Once the dedication was set, on the other hand, it remained
substantially unchanged through the remainder of ToP s printing history. Norton
was able to muster enough italic to restore the raided pages, though he some-
times patched the holes with a mis-matched font. The first wave of raids liberated
about 2130 ens of italic, but the preliminary quire of ToP contains only 42% of
that, about 900 ens. Distributing more than twice the amount of needed type
might have been a prudent action to ensure an adequate supply of all the sorts.
If there was another job or jobs that prompted the italic raids, I have not been
able to identify them. The candidates would obviously contain large amounts of
Norton's english italic (96 mm). This was a slightly unusual size for standard text;
the smaller pica served as the workhorse. In his books dated 1633–34, Norton
used english-sized italic mostly for matter subsidiary to the main texts—prelim-
inaries and indexes. The most extended stretch of it comes in the twelve-page
index of Richard Verstegan's Restitution of Decayed Intelligence (STC 21363), dated
1634. It contains double columns of short lines of italic with a heavy admixture of
roman. Its requirements hardly seem enough to stretch the resources of a print-
ing house responsible for nearly thirty known jobs, or parts thereof, during the
two-year span with ToP at its center. It may be safest to hypothesize that Norton
was doing some job work now vanished, maybe longer than a single sheet, greedy
for a larger size of italic than usual, and urgent. The fact that the raids took place
at more than one point in the ToP production suggests either several ephemeral
jobs, or one larger project which for some reason has not survived.

In this section I have focused on localized resettings of italic passages in re-
sponse to a need for type, but we need to bear in mind that any distribution of
type served to make it available for other jobs. We will see instances throughout
the printing history of ToP in which entire pages or formes were reset, with no
discernible pattern in which pages were chosen. At the end of each printing
phase, Norton had three choices of what to do with each type-page: (a) distribute
it, (b) keep the text standing and strip the skeleton, or (c) keep the text stand-
ing with the skeleton intact. It is quite possible that many of the resettings were
motivated by a need for more roman type, and that the selection of pages was
driven by which type-pages were accessible when the need arose.

It is now time to look at the variants in The Triumph of Peace in detail. The
following sections, §4–8, are the densest portion of this article and some readers
may find in them too much of a good thing. Once we have dealt with the book
at the level of quire, forme, and individual page, we will pull back to see how all

TABLE 4. Correlation of italic raids with page number resettings and text resettings

 

Solid arrows indicate replacement of italic; dashed arrows indicate replacement of roman. Heavy horizontal lines show points of text resetting; horizontal double lines show points of page number resetting. Blank cells signify missing leaves.

TABLE 5. Italic raids

           

Phase

A outer

B outer

C outer

C inner

D outer

D inner

A1r

A2v

A3r

A4v

B1r

B2v

B3r

B4v

C1r

C2v

C3r

C4v

C1v

C2r

C3v

C4r

D1r

D2v

D3r

D4v

D1v

D2r

D3v

D4r

I

1

(1)

(1)

1

1

1

1

(1)

1a

(1)

(1)

(1)

(1)

1a

(1a)

(1a)

1

1

(1)

1a2

(1)

(1)

(1)

(1)

II

2a

[2a]

[2a]

2a

2a

2a

2a

(2a)

1b

[2a]

[2a]

[2a]

2

1b

(1b)

(1b)

1

1

(1)

1a2

[2]

[2]

[2]

{2}

III

2b

2b

2b

2b

2b

2b

2b

(2b,c)

1c

2b

2b

2b

3

2

{2}

1c

2

2

[2]

1b

3

[3]

3

3

IV

[3]

() = eligible
[] = only italic reset
{} = entirely reset
of the changes fit into the production sequence of multiple printings. In §9–10,
I will demonstrate how recurring types give evidence of the order of printing
the formes; in §11–14, I will show how copies were assembled from the stock
of printed sheets on hand. Section 15 presents the entire workflow as a series
of specific actions, and §16 addresses basic issues like edition size and the time
required to print all the copies. The metamorphosis of the printed pages entails
a textual evolution in which deliberate revisions (mostly to the stage directions)
appear amid a chatter of accidentals generated by the resettings. I have reserved
a detailed account of these for the textual introduction to ToP in the upcoming
Oxford University Press edition of Shirley's works.

My method of organizing the variants is, I hope, as simple as the situa-
tion allows, but it still needs a bit of explanation. The comprehensive catalogue
of manifestations of all the pages, from complete resettings down to stop-
press variants and even the occasional random shift, will be found as appendix 1. It
is modeled on a biologist's taxonomic key, and as with all such keys the identi-
fication marks are the minimum required to distinguish the states; seldom are
all the variants on a page mentioned. During the first stages of organizing this
list, I saw the need for some kind of shorthand to refer to each state of a page.
Ideally the structure of this shorthand should express the hierarchical relation-
ships of the states of each page, with the main settings at the top level followed
by impositions at the next and stop-press variants at the bottom. However, as
the families of variants continued to ramify—sometimes manifesting states that
would not fit neatly into any hierarchy—practice sometimes overruled theory,
and the codes ceased to reflect any rigid correspondence with production phases
or the relative importance of the changes they denote. Attempting at this point
to revise the codes to make them more consistent—to the extent that that would
even be possible—would render precarious their connection with notes taken
over two decades, which embody the current system in a well-documented pro-
cess of evolution. For specification of the exact relationships between the states,
the reader should refer to appendix 1.

DEVELOPMENT BY QUIRES

§4. Quire "a"

A. a1R (TITLE PAGE)

The four main states of the title page are shown in figure 1. A quick compari-
son reveals that these states are largely in the same setting. Note, for instance, the
damaged "M" in "Majesties", line 9, and the wrong-font final "3" in the imprint
date, which is common to all states. It is relatively easy to demonstrate, with a
purely typographic proof, that the order of states shown in figure 1 is also the
order of printing. As mentioned in §1, the "O" in line 3 is sound (as illustrated)
in some copies of state 1, but shows damage in others. Therefore, state 1 must
be the earliest. State 3 (a unique survival at the Folger), with its dangling comma
after "Grayes-Inne", must precede state 4. One would assume that the "third
impression" copies came last, and this is borne out by the association of this state
with later states of quires A-D, as we shall see. If we play devil's advocate and
posit an order of 1,3,4,2—moving the "third impression" forwards—we would
not be able to explain why line 11 of the title ("February the third, 1633.") starts
in one setting in state 1, changes to another for states 3 and 4, then reverts to the
original one in state 2. (The presence of a comma or a period after "third" is the most obvious marker.) So, we can safely accept the order as illustrated.

State 1 is further complicated by a deliberate realignment (state 1a2, not
illustrated). Figure 1(A) shows the first word of the title sitting too far left with
respect to the following line. During the printing of state 1, but after the damage
to the "O" in line 3, the alignment was altered by shifting "THE" rightwards so
that the "T" lay closer to the "V" below than to the "I". As a result, the word
came to sit too far right on the page considered as a whole, but with respect to
the following line, it is an improvement. No version of the title ever gets each
line quite centered.

Table 6 shows the hierarchy of changes that we have just worked out. The
numbering of the states refers to appendix 1. For the title, I have called the four
states 1a (with three sub-variants), 1b, 1c1, and 1c2. The notation "1" recognizes
that all are in the same basic setting and that the fourth is a stop-press variant
of the third. The taxonomic key ignores other adjustments, even the resetting
of line 11, that accompany the revisions signaled by the chosen markers. (A full
account of the evolution of the ToP title page would provide enough material for
a separate article.)

TABLE 6. Evolution of the title page

The reason for the deletion of "of Grayes Inne" in state 1b, and its restoration in
1c, is unclear. Shirley was admitted to honorary membership in the Inn on 17Janu-
ary 1634. This was a common enough practice for individuals who had rendered a
significant service, and it should not have been controversial. Quite probably, Shirley
was led to expect it as part of his compensation for writing the masque. He may have
allowed the words to creep into the printer's copy in advance of the actual bestowal.
When someone called attention to the error, the words were deleted, only to be re-
stored (in a new setting, of course) in state 1c, which was evidently printed after the
honor was official. Since there is no evidence that anyone tried to suppress them in
earlier copies, the deletion could represent a minor adjustment for a minor embar-
rassment. It could even have happened as a stop-press correction. This theory would
have a corollary that the printing of at least state 1a was well underway or completed
by the middle of January, and this would support Greg's assumption that the pub-
lisher aimed to have copies ready for sale by the first performance of the masque
on 3 February. 21 However, the archival evidence and other facts I will adduce in
§16 argue against the book's having been on sale that early. The disappearance and
reappearance of "of Grayes Inne" is the first of many mysteries surrounding ToP that
I will have to leave without a complete solution.

B. a2R

As with a1r, there was only one setting of this page. The following table shows
its states along with those of the title page.

TABLE 7. States of the title page and a2r

At the head of a2r is a decorative band composed of pairs of acorn orna-
ments. The early state has 28 pairs of acorns, the later two have 27. It is nearly
impossible to find individuality in acorn ornaments, but tiny differences show
that the change results from resetting, not a loss of one pair on the end. The
reduction happens simultaneously with a clear resetting of the direction line—
signature mark "a2" and the catchword—without a change of reading. There
being no reason to reset the acorns and direction line in mid-run, it appears that
the printing of a2r underwent a pause at this point during which the skeleton
(including the acorns in this case) was replaced. Another arbitrary change oc-
curred later, when the equal leading above and below line 5 of the heading (the
single word "THE") was disturbed and that line drew closer to the one below.
The same operation apparently dislodged a space two lines above, following the
words "Equall and", so that copies printed thereafter show a gradual rightward
drift of those two words.

C.a2v

Like the others in quire "a", this page has only one major setting (see table 8).

The first change to a2v is the stop-press correction of "ond" to "and". (The
error was a simple one for either the typesetter or the previous distributor to
make—the "o" box in the typecase lay just above the "a".) All copies of the un-
corrected state also have the undamaged "O" on the title page (a1r:1a1a). After
the correction to "and", further copies were printed before the "O" on the title
became damaged (initiating state a1r:1a1b). The addition of the comma after
"Honourers" on a2v occurs at the same point as the re-alignment of the first line
of the title page (producing a1r:1a2). The copies with the "Third impression"
title page (starting at a1r:1c) also show a resetting of the closing two lines of the
dedication on a2v, with no difference in reading. This sort of arbitrary change
should remind us of the headline resettings that, as I have argued, signal pauses

TABLE 8. States of the title page, a2r, and a2v

TABLE 9. States of quire "a"

in the printing. The loss of the rule below the dedication's signature is another
such indicator; it coincides with the disappearance of "of Grayes Inne" on the
title (a1r:1b). Shirley's signature to the dedication and the rule below it seem to
operate as part of a quasi-skeleton that is less stable than the text on which it borders.

D. QUIRE "a" ENCOMPASSED

Let us now combine in one sweep of the eye the states of the three type-pages
comprising quire "a". Table 9 is a "plain English" version; table 10 shows the
variants using the shorthand keyed to appendix 1. In both tables I have provi-
sionally placed double lines at the points of possible reimposition that I have
discussed above, which seem to indicate breaks in the printing.

TABLE 10. States of quire "a"

Note that the changes on all three pages tend to occur in lockstep. Quire
"a" is a half-sheet, which would most economically be produced by half-sheet
imposition run off by work-and-turn. 22 In this method, all the pages of the quire
are printed at once, and the sheets are flipped end-to-end and perfected with the
same pages on the back. When cut in half, each sheet yields two copies of the
quire. If this were the case with ToP, and if one examined the type impression
in many copies, about half will show that a2 recto was printed before the verso,
while the other half will show the opposite sequence. I have attempted this exami-
nation on 23 copies of ToP—a difficult process because the outer leaves normally
show most of the soiling and were usually washed and pressed at some point.
Only leaf a2 can be analyzed in this way because the title leaf verso is blank.
But the test did show a mixture of sequences of printing, so we can be confident that
quire "a" was printed by work-and-turn.

Because Norton used half-sheet imposition, all three type-pages of quire "a"
were accessible to alteration at the same time. So, it is not remarkable to see
changes taking place across all three pages simultaneously. The thing to note
at these transitional points is the mixture of arbitrary revisions with apparently
rational ones. When the number of acorn pairs drops from 28 to 27, the title
page gets a small realignment and the comma is added on a2v. Later, when "of
Grayes-Inne" is removed from the title (state a1r:1b), the rule disappears from
under the dedication on a2v and the line spacing changes on a2r; the latter is
an apparently random disturbance that entails the accidental loosening of a line.
The addition of "The third impression" to the title (state 1c) brings with it some

needless resetting on a2v. These bursts of purposeless activity strengthen our ear-
lier suspicion that the printing of the preliminaries proceeded in several discrete
spurts of work, of which the double-rule boundaries I drew in the tables above
mark only the most obvious transitions. I believe we have good evidence of four
printings, and that the double-rule segments in tables 9–10, indicating pauses in
the work, should run straight across both formes. I will refer to these printings,
or discrete periods of work on the formes, as "phases", and table 11 incorporates
them into a comprehensive scheme for quire "a":

TABLE 11. Production phases of quire "a" and their variants

My accounts of quires A-D will be less detailed than the foregoing and will
ask the reader to refer to the taxonomic key of appendix 1.

§5. Quire A

Table 12 results from mentally removing quire A from every copy of ToP
and arranging its pages in order from early to late states. Heavy horizontal cell
borders indicate complete resettings; double-rule borders show reset headlines
indicating reimpositions of standing type. Again the printing seems to divide into
several phases. The table shows that the formes evolve in the same direction,
which indicates that the printer was executing an orderly flip of the stack after
printing the first forme: the first-printed sheets of one forme became the first-
printed of the other. This pattern generally holds throughout ToP, though we will
find some aberrations in the other quires.

The table reveals that, as with the preliminary quire, Norton turned out sheet
A in several discrete phases of work. The most common marker of a shift to a
new phase is the resetting of the skeleton, or extensive text, of multiple pages
at the same points in the production history, though quire A does not show this
consistently. As mentioned in the previous section, it is often impossible to say
what motivated Norton to distribute some pages and not others. In the case of
quire A, all but one of the pages (A1v) retain at least half of their original settings
throughout the printing history. A1v was reset after Phase I, and the bottom half
of A2r was reset twice. Work on Phase II of quire A—the most clear-cut transi-
tion point in the quire's evolution—began by reimposing all of the outer-forme
type-pages set for the first phase. (Analysis of the type bite shows that quires A–D

TABLE 12. Production phases of quire A and their variants

of ToP always went through the press outer-forme first.) For the perfecting run,
Norton reimposed the inner forme using two pages with minor changes from
Phase I (A3v and A4r, which lay head-to-head), but the other two pages of the
forme were either completely or partly reset (A1v and A2r respectively).

Other phase shifts were more subtle, and the transition from IIb to III was
barely a whisper: only two pages got stripped and a couple of raised quads were
pushed down on a third. The backing forme shows no sign of a corresponding
interruption. This lack of correlation requires some fancy explaining. My feeling
is that the reimpositions on A1r and 2v, though they affect only two pages out of
eight in the quire, were still associated with a work stoppage. The lowering of the
quads on A4v, which happens at the same time, was probably an unintentional
result of removing the type-page at the end of Phase IIb and replacing it for
Phase III. Furthermore, on the evidence of the surviving copies of ToP, the number
of quire-A sheets in Phases IIa and IIb combined (11) is approximately equal to the number of sheets of the other text quires in their own undivided Phase II
(quire B, 10; C, 10; D, 11). Copies of the book that contain Phase III
of sheets B-D also contain Phase III of sheet A. (One exceptional copy from Phase III (Folger STC 22459.2) has an A-sheet from IIb, which indicates a possible slight
overrun of the A sheets that found their way into the following phase. I discuss
the overrun in more detail in §13.) These are the reasons I have designated the
two sub-phases within the second impression of quire A.

It is likely that Norton's original intent was to print Phase II as a continuous
job, but that something interrupted the machining of the outer forme and the
type was removed from the press. This marked the end of Phase IIa. When he
was able to resume work, he had two options. The simpler was to put the outer

forme back on the press, complete the intended second impression of that, then
perfect the sheets from Phases IIa-b of the outer forme together. However, a
skeleton resetting on A1v and a page-number shift on A3v in the inner forme
suggests that the perfecting of IIa-b also proceeded in two phases of work. Norton
perfected all the IIa sheets before doing another printing of both formes to
complete Phase II. Although it required more shifting of formes, this tactic would
prioritize the completion of some number of finished A-sheets. This would allow
a more speedy assembly of at least some copies of the whole book.

The inner forme of Phase IIb is indistinguishable from that of Phase III except
for a stop-press correction which appears part-way through the latter phase
(A1v:2b2). On the evidence of table 12, Norton could even have printed the inner
forme of Phase IIb and continued straight on to print the inner forme of Phase
III in the same run, perfecting the sheets of the latter phase with the outer forme.
But the type-impressions show that Norton always printed his outer formes first. So, we have to conclude that Norton began Phase III of quire A with only minor
changes to the outer forme (new skeletons for A1r and A2v) and none at all
to the inner. The boundary line between Phases IIb and III continues invisibly
right across the inner forme. We will later see similar arbitrary patterns in the
production of the other sheets.

A few other observations on quire A are worth making:

• In the outer forme, the four pairs of variants A1r:2c1–2,
  A2v:2a1–2, A3r:2b1–2, and A4v:2b1–2 all involve printing accidents. The last two pairs
  result from specific, one-time events. On A3r, the "3" in the signature mark
  drops out in the second state. (This may have occurred during the act of repairing
  some type damage and mis-alignment at the upper left, shown uniquely in copy
  bL 644.0.44. 23 ) A4v:2b1 has the quads printing between the two paragraphs, as
  mentioned above; these rose during the reimposition preceding Phase IIb and
  disappeared again at the inception of Phase III. (The rule below the text shows
  various bends during these changes.) In contrast, the "variant" on A2v involves
  a progressive downward creep of the last letter of the catchword, so these are not
  two distinct states. Similarly, the A1r variants involve a page number that prints
  more or less faintly—evidently a function of variable depths of impression—and there is no clear progression one way or the other over the course of printing. All
  these changes present a random picture in appendix 2, but if one plucks out only
  Phases II and III of quire A and puts the rows in order by state, the progression
  becomes clear.
• In A1v:2b1, the jockey in the torchlight parade holds a "bride" in his
  hand; state 2b2 corrects this to "bridle". Previous attempts at ordering ToP (most
  recently in the STC) have tried to use this difference to characterize major groupings
  of whole copies, but we see here that it is simply a stop-press variant within
  Phase III.
• Copies of the book with sheets from Phase IV of quire A come only with a
  "third impression" title page, and vice versa. In this final phase, a paragraph of description
  added to A2r affects the page breaks for the rest of the quire. Sometimes

  ---

  135

  
  
  
  the transferred text was moved in type, sometimes it was reset. The method used
  depended largely on whether the lines travelled to a facing page (i.e. in the same
  forme) or to the verso (printed in the next forme, in which case the transferred
  text was usually reset). Section 15A gives the specifics of this process.
• Between Phases I and IIa, the four lines of italic split between A2v-3r were
  raided and reset in great primer size (120 mm).
• In Phase IIb, the catchword on A2v, which heretofore correctly read
  "The", changes incorrectly to "All". This marries with the first word in roman
  type on the next page, but ignores the two lines of italic that precede it. We would
  expect this sort of error to arise during the italic raid after Phase I, but that had
  already occurred. Instead, it appears that the workman doing the reimposition
  for Phase IIb did not notice the italic lines and took his cue from the first paragraph
  in roman. The error persists through Phase III.

§6. Quire B

A. PHASE I

The early phase of quire B presents one significant difficulty:

TABLE 13. Variants in quire B, Phase I

Only three pages show variant states in this phase. Those on B1r are:

1a1 heading wholly in italic

1a2 heading reset with characters' names in roman

• type relatively undamaged
• first letters of ll. 7 up and 9 up are damaged

We can be reasonably certain that this is the correct order because the type
damage does not occur during state 1a1, and the reset heading follows the normal
convention by giving proper names in a contrasting type style.

The variants in B3v-4r, in the backing forme, involve commas. 24 Neither
state is obviously the earlier, but the direction of the change is less interesting at
this point than its confused relationship with the outer forme: the early and late

states of the outer forme each occur with both states of the inner forme. The
distribution of surviving copies is as shown below:

TABLE 14. Distribution of selected variants of quire B in Phase I

If, after printing the outer forme, the printer followed a normal turn of the
heap of sheets that contained B1r, we would expect the changes on B3v-4r, in the
backing forme, also to follow an orderly progression from early to late. We saw
this pattern in quire A. In contrast, the distribution of states in Phase I of quire
B shows a sort of "countercurrent eddy", indicated by the shaded cells above.
It appears that a disorderly stack-turn took place. To understand this concept,
first recall that a normal print run of a forme might produce several hundred
sheets, yielding a thick pile of damp paper with the last-printed sheets on top.
To preserve the same order and orientation of sheets going through the press
on the perfecting run, the pile had to be inverted end-to-end. Since the pile was
too bulky to turn as a unit, the workman had to transfer it in several handfuls,
inverting each one onto the new stack.

Suppose instead that the workman carelessly transferred one handful of sheets
straight across without turning it. This would produce a stratum of sheets with
the printing facing the wrong way. We would expect the pressman, when working
down the pile and finding this mistake, to utter a curse, turn the affected sheets
to align with the others, and go on working. But he might also utter a curse and
start picking up the sheets from the opposite end, using a slightly different motion
to place them on the tympan. The pinholes—normally positioned slightly
asymmetrically—would help prevent mishaps in perfecting, and the pressman
might enjoy the little bit of variety until reaching the underlying layer of correctly
oriented sheets. This seems to be what happened in Phase I of quire B. Let us
use our sample of the nineteen copies of that phase to stand for the entire first
printing of this gathering. The machining of the outer forme (containing B1r)
produced a heap like the one on the left below, with the later states on top.

If we now back up these partially garbled sheets with the inner forme containing
B3v-4r, which begins with state 1a1 and switches to state 1a2 about half-way
through, we get the observed distribution of variants (table 15).

A stack turn with handful B transferred without flipping is the only scenario
that yields the observed results if we assume the printer started the perfecting
run at the top of the stack and worked his way straight through. Our thought
experiment also shows that state iai of B3v-4r most likely came first. If state 1a2
had been earlier, we would have to hypothesize a second turning of the garbled
stack—this time correctly done—yielding the inverse of the first column above.
The reader may verify that this does not yield the observed distribution of states.
We cannot change the boundaries of group B without creating a conflict with the
observed combinations, so that cohort must represent the survivors of a manageable
number of sheets that the worker was comfortable picking up with two
hands. It is tempting to extrapolate from our nineteen-copy sample to estimate
the total number that were printed in Phase I. If group B represents one "grab",
the total first printing of the quire might have contained four roughly equal
handfuls of damp sheets, each of which in ToP measured about 41 by 31 cm. 25 A
practical trial might provide an estimation of the edition size. However, it proves
difficult to locate a source of poor-quality, lightweight, handmade paper with

TABLE 15. Distribution of copies resulting from anomalous transfer of handful B

similar characteristics to that used in ToP; and I have been no more successful in
getting usable data from present-day fine printers regarding their paper-turning
habits. It turns out that there are a host of variables that make it impossible even
to approximate an upper limit on the edition size by this means.

It should make us slightly uncomfortable that B1r state 1a2b is absolutely
coextensive with group A. The last handful (or more likely two) appears to have
contained only that state, which occurs nowhere else; and state 1a2a is found
only in handful B. If my analysis above is correct, and if further copies of ToP
were to come to light, one of them might contain B1r:1a2a combined with a
later state of B3v-4r.

B. QUIRE B, PHASES II-IV, OUTER FORME

The next phases of quire B are so complex that it is necessary to analyze the
two formes separately.

After Phase I, three of the pages of B(o) were reset and one (B3r) was reprinted
with a new skeleton. Phase II's only variants occur on B4v (in states 2a1 –3) in the
passage "ayre, ... with an artificiall bellowes coold". The last word also appears

TABLE 16. Variants in quire B outer forme, Phases II-IV

as "cool" and "coo". Neither reading makes as much sense as "coold" (2a1), and
the variants probably result from a two-stage accidental loss of letters from the
end of the line. Frisket slip is not a suspect here because there are no intermediate
states—the letters are either completely present or completely lacking.

For Phase IIIa, all of the pages have new skeletons. Page B4v state 2b2,
found uniquely in a copy at St. Catharine's College, Cambridge (Z.59), presents
a problem because the page number has shifted horizontally. However, the type
appears to be the same and the page is unchanged in other respects, so reimposition
is not likely to be the cause. If looseness in the line caused the shift, we
would also expect slightly different spacing among the three characters of the
page number, but we do not see this either. It remains an unexplained problem,
but a minor one.

In the discussion of quire A, I identified a point within Phase I at which press-
work was interrupted, but the other formes of the book from that phase did not
show a similar hiatus. Sub-Phases IIa and IIb were in effect separate printings of
quire A, with both formes showing changes. Quire B is a slightly different case.
Again, it appears that a more urgent job interrupted the printing, this time during
Phase III of the outer forme. The delay was long enough that the type-pages
were stripped and probably tied up. After the other job played through, printing
of B(o) resumed with a new skeleton, including a new arrangement of acorns on
B1r. (I have not been able to determine whether the direction lines were reset as
well—they are closely similar.) These re-imposed copies constitute Phase IIIb.
However, both of these impositions of the outer forme occur with identical states
of the inner forme. The two sub-phases of the outer forme seem to have gone
through a single perfecting run. 26 Apart from this one forme—and a couple of
other small variants which I will discuss in §11–14—these five copies resemble
the twelve others from Phase III.

The following corrections demonstrate that Phase IIIb indeed followed
Phase IIIa:

TABLE 17. Variants in quire B, Phases IIIa,b

In Phase IV of this forme, a few things are of interest:

• On B1r, the page number wanders a bit horizontally over the 15th pair
  of acorns.
• Likewise, on B2v, the stem of page number "4" can point to the first letter
  of "the" below, to the "o" in the preceding word, or anywhere in-between.
• In this phase, B3r picks up the first three lines of text (in a new setting)
  from the following page, relieving some crowding on the latter.
• On B4V, the four flush-right italic lines have been raided and are reset in
  smaller (pica) size. The two centered italic lines on the same page are unaffected,
  perhaps because it was not worth the trouble to get at them. On this page also
  the page number is horizontally unstable. This example of wandering, along with
  items 1 and 2 above, are among many in ToP that demonstrate that lockups on
  the press bed must have been much less rigid than Joseph Moxon's descriptions
  would lead us to expect.

C. QUIRE B, PHASES II-III, INNER FORME

TABLE 18. Variants in forme B(i), Phases II-III

The inner forme of quire B is even more perplexing than the outer one.
Phase II begins with a complete resetting of all four pages. Shortly afterward,

B4r alone acquires a reset skeleton (state 2b), which persists into Phase III. The
direction of the change from 2a to 2b is confirmed by a textual correction and
progressive damage to line 7 during state 2b. For only one page of a forme to
change its skeleton is puzzling enough. But it also occurs against a background
of changes to its forme-mates B1v-2r which do not develop in consistent directions.
These variants are:

B1v: The "V" beginning line 10 is more or less displaced downward, or is
entirely absent. Where it is absent, the "D" from two lines down drifts up into
the vacant space.

• B2r: 1. St. Catharine's College (Cambridge) Z.55 shows a unique setting of
  the page number. The number "3" comes from a different font than that found
  elsewhere in the book, being rounded and with the upper loop noticeably smaller
  than the lower. It is evenly spaced between undamaged parentheses. All the other
  copies have the usual narrow "3" with loops about the same size, and one or
  both of the page-number parentheses are either damaged or fouled with ink. This
  number was apparently not locked up tight, and it can fall anywhere from close
  to the left parenthesis to centered between the two. The direction line appears
  to contain the same type as does the St. Catharine's copy.
• In some exemplars, the right ends of lines 1 and/or 4 creep upwards by
  varying degrees.
• The comma ending line 6 prints either normally or as a blob.
• A stray bit of metal resembling a hyphen may show in the empty space
  between lines 6 and 7 and below the "na" of "nature" in line 4.

A tabulation of these features shows some correlation among certain manifestations
of variant 1 above (a bit of dirt on the right parenthesis at about 2 o'clock),
variant 2 (both lines displaced), and the blobby comma in variant 3. But other
characteristics do not line up, either with each other or with the amount of type
damage on B4r. Most of the features are of a transient nature and could be
reversible: for instance, the wayward "V" on B1v could drift, fall out, and be
replaced, so late copies might look like early ones. The loose page number with
its parentheses could have fallen apart near the end of the run, and all three types
could have been reset to produce the unique manifestation in the St. Catharine's
copy. The stray metal in variant 5 could have been slightly raised in the forme
throughout this phase, but would print only when sufficient pressure

Excessive attention to these details might not be either productive or healthy,
but it is not so easy to dismiss the skeleton change on B4r early in Phase II (the
transition from state 2a to 2b), mentioned above. Because there is no evidence
of changes on the other three pages of the forme at that point, it presents a challenge
to my theory that arbitrary page-number changes are signs of work stop-
pages. It is another occurrence I will have to leave unexplained.

D. QUIRE B, PHASE IV, INNER FORME; QUIRE B RECAPITULATION

The final phase (IV) of forme B(i) shows a resetting of all the pages. For no
obvious reason, the four-line roman-face stage direction on B3v beginning "The
Gentleman ..." is now set in pica, one step smaller than the surrounding text.

TABLE 19. Forme B(i), Phase IV

Combining all that we have discussed about both formes of quire B generates
table 20, which is puzzling in its fragmentation and necessarily evasive about
correlations among certain states below a certain level of detail.

TABLE 20. Synopsis of variants in quire B

Nevertheless, it is possible to discern two events, initiating Phases II and IV,
that clearly interrupted work on the entire quire. These resulted in the resetting
or reimposition of almost all the pages. The outer forme underwent two more
skeleton changes on its own, introducing Phases IIIa and IIIb.

§7. Quire C

Table 21 shows four relatively clear-cut phases of production for all of quire
C. Phases II and III are signalled by reimpositions or resettings that run straight
across both formes. Phase IV is more subtle, having only two pages reimposed.
But the remaining pages (except for C3r, which comes over completely unchanged
from Phase III) all show revisions, some of them major, at the inception

TABLE 21. Variants in quire C

of Phase IV, and the timing of these coincides with the major discontinuities that
we have already shown initiating the final printing phase of the preceding quires.
(Appendix 3 shows this correlation most clearly.) Phase IV of this quire is found
only with the "third impression" title pages and vice versa. So, despite the lack of
extensive resetting at this point, we can still assert
the likelihood that the alterations to both formes preceding this final phase took place during a significant
pause in the press run.

On C2v, the action of the masque enters its final phase, which features numerous
songs set in italic. These provide the main source for the italic raids
discussed in §3.

The production of the quire went as follows:

A. PHASE I

C outer forme shows a normal pattern of variants whose sequence is all but
certain. The inner forme, however, presents an insoluble problem in the sorting
of the two variants of C1v, which at line 23 has either (1a1) "We change
the Seene" or (1a2) "We change the Scene". These are found in the combinations
shown in table 22. (The remaining three pages of C(i) are invariant, so
are not shown.) About half the sheets (the eight copies shown in the last row of
the table) have the latest states of both the outer and inner formes, as we would
expect, but the other ten sort illogically. Unlike the retrograde pattern of quire
B Phase I (§6A), the existence of all these permutations cannot be explained
by any plausible irregularity in turning the stack. Only dropping some portion
of the sheets on the floor and picking them up more or less at random would
cause the first state of the inner forme to be printed on the back of three of the
four variants of the outer forme. "Accidents at the press" is the last resort of the
baffled bibliographer, and I will not waste space here in trying to develop such an
explanation.

TABLE 22. Distribution of C1v variants in Phase I

On C4V the bottom lines of text are bent upwards at the right. The greatest
amount of deflection is a bit more than 2 mm—so, rather less than half the
height of one line—but bad register with the text on the back of the leaf makes
it quite noticeable. This much distortion would be impossible if the non-printing
space on the page were packed solid; there must have been a gap in the lockup,
and I would guess that one of the lines of Song 5 somehow got into the forme
without being justified. Such an oversight would also explain the slight misalignment
in lines 7–9 on the same page. This precarious situation might have made
it impossible for a workman to pick up and transport the locked-up forme in
the manner described by Moxon, 27 but the interaction of pressures along both
dimensions of the page was apparently sufficient to keep types from being pulled
during presswork. This is the most graphic demonstration in ToP that printing
could proceed with adequate stability even if regions of the forme lacked support
on all four sides. We have already seen instances in §6B of wandering alignment
between adjacent lines that probably arose from similar causes. 27a

B. PHASE II

Phase II shows raiding of italic on pages C2v, C3r, and C4v. These pages contained
all of the main caches of italic in the outer forme. C4v was particularly
heavy-hit, being totally reset except for the roman lines 7–9 and the last line. After
the raid, the italic Song 5 and its roman title on that page reappeared in pica size.

The inner forme of quire C offered an even richer source of italic than the
outer, but it was never raided at this point, or ever during the history of ToP.
The reason is probably that C(i) was on the press and inaccessible while C(o) was
being raided. C1v does show resetting of the italic in lines 8–10 and 25–27, but
this page was not a plentiful source for the font, and the resettings could have
arisen from some other cause. (For instance, the earlier state showed a comma

out of place in line 7, and it is possible that lines 8–10 were pied in the act of
repositioning it.)

C1r in this phase shows some horizontal instability in the page-number
position.

C. PHASE III

In this phase, pp. C2r and C3v—which lie side by side in the forme—undergo
the only complete resettings in the history of quire C. Head to head with
these, but later in the phase, C1v and C4r both show page-number shifts. These
are probably simultaneous but random drifts, as the types in the page numbers
appear to be the same before and after. However, the five copies which show this
deviation happen to be the same ones which contain quire B from Phase IIIb. I
will discuss this uncomfortable coincidence in §13.

Also in this phase, the text of C1v shows further instability around the same
points which changed in Phase II: the italic in lines 1 –10 and the last four lines
is reset, the fourth line up ("A Landschape the Scene.") going missing in the first of
two steps to a smooth transition between antimasques. When lines 8–10 were
changed in the previous phase, the period after "carryed off" in line 9 ended up
below the line (which shows another unstable lockup). In Phase III, it is properly
aligned, and the resetting of the nearby italic may have had some connection
with that repair-work.

D. PHASE IV

Phase IV, as mentioned above, is found only in "third impression" copies of
ToP, and, like the final phase of quire A, it introduces important textual revisions
that alter the page breaks within the quire. Edits on C1v and 2r shorten the word
count significantly, causing two-line chunks of the text to move backwards until
the original breaks resume at the end of C2v.

This phase gives us further examples of text that is reset either larger or
smaller than the prevailing 96 mm "english" size:

C1r: At the top of the page, a brace separates the list of three characters from
their stage directions on the right. In the original setting, the directions crowd
right up to the brace. In Phase IV, the direction is reset in 82 mm pica; the line
breaks are the same, but the page gains some needed white space.

C2r: The antimasque of the "phantastique Adventurer" has the same layout
as the previous example, with a brace separating the names of the characters
from their stage directions. In this last phase, the directions are rewritten and
lengthened, so the compositor reset them in pica type apparently to maintain
visual balance.

§8. Quire D and "A Speech"

Readers who have followed me to this point will, I trust, agree that the thirty-six
pages of The Triumph of Peace comprise a Wunderkammer of bibliographical
problems, any one of which could support a separate study. The last quire will
not disappoint.

A. QUIRE D PHASES I-II

Quire D begins life with an outer forme showing an orderly progression of
variants:

TABLE 23. Variants in forme D(o), Phases I-II

The second state of D1r supplies a missing word; at the same time the first
three lines lose their horizontal alignment. This state gives way to a third one
in which the same three lines of type are damaged. D4v shows a more subtle
change, in which the right parenthesis of the page number suffers damage near
the bottom; this occurs while D1r is in its second state.

The inner forme on this set of sheets also shows a regular progression, but
with a clear work stoppage indicating two phases:

TABLE 24. Variants in forme D(i), Phases I-II

The variants are

D1v: State 1a1 has a comma that has slipped into the line below, and a damaged
parentheses to the right of the page number. In state 1a2, the comma is
correctly positioned and the parenthesis is sound. This sequence, which could
have gone in the opposite direction, is confirmed by

D2r: The second state (1a2) corrects a typo, improves sense by adding a
comma, and pushes down one of two quads that were printing inappropriately
in the first state.

The conspicuously damaged parenthesis in D1v:1a1 is the same one found on
C2v, also in Phase I, in the same position to the right of the page number and in
the same quarter of the forme. This recurrence gives us a clue about the order
of printing these quires in Phase I. There are two possibilities:

• Quire D came first. The damaged parenthesis was removed part-way
  through Phase I as one of five stop-press improvements to the inner forme, but
  instead of being discarded it was put back in the type-case and later set—coincidentally
  in the same position—in the headline of C2v.
• Quire C came first. After Phase I of that quire the damaged parenthesis
  either persisted on D(i) as part of the skeleton, or returned to
  the type-case to be set—coincidentally in the same position—on D1v. In either case, part way
  through Phase I of quire D, its damaged state caught the eye of the corrector
  and it was discarded.

The second possibility—that the parenthesis remained as part of the skeleton
and was later replaced by a sound type—is much more likely; and indeed this
damaged parenthesis never reappears in later states of ToP.

The independent typographic evidence that Phase II of the inner forme followed
Phase I is rather slight: the correction of one error in line 8 of D3v, which
kept the same setting of the first twenty-one lines from Phase I. However, as most
copies of ToP containing Phase II also contain later states of quires A-C, we can
be confident of this direction of change.

When we put together Phases I and II of the two formes of D, their relationship
becomes complicated:

TABLE 25. Variants in quire D, Phases I-II

This table shows two oddities beyond the fault-lines indicating the pauses in
production, which by now we have become accustomed to:

• There were two production phases of the inner forme, marked by skeleton
  changes on three pages (D1v, D2r, and D3v) and a complete resetting of D4r.

  ---

  148

  
  
  
  However, the outer forme of the same sheets appears to be the product of one
  continuous printing, with states 1a2 and 1a3 each appearing within Phases I or II,
  not during a pause between them.
• Phase I of this quire shows an illogical mixing of states between the two
  formes. If we consider only the states in this phase and ignore the invariant pages,
  the scheme looks like this:

TABLE 26. Distribution of states of quire D, Phase I

This is very nearly a random shuffling; like the permutations of quire C in
Phase I, it cannot be attributed to any plausible method of flipping a single stack.
The three combinations that do not occur are early/early, middle/middle, and
late/middle; but there is no guarantee that these would not emerge if new copies
of the book were to come to light. As with quire C, I am unable to explain the
distribution of variants here.

Point 1 above—the fact that the inner forme goes through two production
phases while the outer forme appears to have only one—is easier to explain if
we suppose that the outer forme went through two printings with no alteration
during the pause between them. Phase II of D(i) associates mostly with the third
state of the outer forme (D1r:1a3), in which the first three lines of type on D1r
show damage. The accident that caused this may in fact have come in the transition
to Phase II. The distribution of surviving copies is as follows:

TABLE 27. Correlation of D1r variants with phases of D(i)

Only two copies (McMaster University, and British Library Ashley 1697,
shaded above) show sheet D with its inner forme from Phase II but the outer
forme in the undamaged state typical of Phase I. 28 If we take the hypothesis that

the outer forme had two separate printings corresponding to Phases I and II of
the inner forme, there are at least two possible explanations for the observed
distribution:

• The damage occurred early in the second printing of D(o), or
• McMaster and Ashley 1697 represents a small residue of sheets which for
  some reason were not perfected during Phase I. The damage occurred during
  the replacement of the outer forme in preparation for Phase II. The Phase II
  sheets with the damaged type were combined with the Phase I leftovers and
  perfected with Phase II of the inner forme.

It is more likely for damage to occur at points of transition than during
periods of repetitive work, so I lean toward the second explanation. Either one
requires some special pleading to fit the hybrid copies into a plausible narrative.
There is no doubt, though, that they are closely associated with the skeleton
change in D(i), which marks the end of Phase I of that forme, and the same phase
also produced about equal numbers of quires A-C.

Before Phase II came the pause during which italic was raided in quires A
and C. Norton raided quire D as well, but only the inner forme. Three of its
constituent pages show the italic in a new setting in Phase II, while the fourth
page—consisting almost entirely of italic—was entirely reset. D3r, in the outer
forme, could have supplied a good amount of the font but was not touched.

For Phase II of quire D, Norton printed the outer forme, as I have argued,
unchanged from its Phase I state except for the damage to D1r (which could also
have happened early in Phase II). Following Phase II, another small italic raid
occurred on both formes, carrying off the songs on D2r and D3r. The former leaf
had already been raided, even more extensively, following Phase I. The second
raid shows a continuing need for the font beyond its earlier use to set the dedication
for ToP; some other job was interrupting at this point. Concurrently with
these replacements, and quite bafflingly, the roman portion of D3v—about two
thirds of the page—was reset. I cannot begin to suggest a reason for this.

B.QUIRE D PHASES III-IV

Uniquely for ToP, Phases III and IV of quire D contain two different settings
that mix randomly in copies containing Phases III and IV of the other quires. I
will call these settings A and B:

TABLE 28. Quire D settings, Phases III-IV

Setting A, in six out of the eight pages of the quire, is simply a reimposition of
the standing type from Phase II; the first two pages are reset. Setting B contains
typographic features not found elsewhere in ToP, nor in any of Norton's printing
from the years 1633–34 that I have seen. The characteristics are:

           

Quire D, setting B		Rest of ToP (Norton)
k	The diagonal arms are stunted, the upper-right serif failing to reach the x-height and sometimes tilted upward.	k	The upper diagonal arm consistently reaches the x-height.
w	In some examples, the left V meets the right one a bit downward from the middle serif; the bottom of the left V falls below the baseline of the right one, so that the letter usually appears to tilt to the left.	W	The left and right V's meet the middle serif at the same height, and the letter does not tilt.
h	The curvature of the legs is midway between that of Norton's i-1 and i-2; the right leg ends pointing left.	h	In i-1 the legs curve slightly toward each other but do not approach closure; the right leg ends pointing 45 degrees below horizontal. In i-2, they close almost completely; the right leg ends pointing straight left.
3	Full-bodied.	3	Abnormally narrow.
o	Ranges with the other lowercase letters.	o	Usually too large.

[Description: Setting B, showing stunted "k" (Houghton STC 22459a.5)]

[Description: Setting A (Norton), showing oversized "o" (Huntington 69432)]

These are strong tendencies, not invariable markers, but the differences are
pronounced enough to indicate that a second printer's type was used in setting
B. The immediate suspicion falls on Nicholas Okes, with whom Norton at this
time had an informal and increasingly fractious partnership. 29 Okes had already

left traces in ToP, though probably involuntarily. The factotum on a2r came from
his 1609 stock as catalogued by Peter Blayney 30 —Norton is known to have filched
other woodcut material during their association. To confirm my hunch about
Okes, I surveyed English books printed in 1633–34, searching for the odd "k"
and "W" in particular. Unsurprisingly, this did not yield an unequivocal answer.
Early-modern printers' typographic stocks were not proprietary, homogeneous,
or static, and there are no woodcut elements in quire D to help finger the collaborator.
I found versions of the stunted "k" in the repertoire of numerous
printers, including at the university towns, and "W" showed great variability in
most samples. Many texts I looked at did not even contain english-size type—
the smaller pica is much more common—and a good proportion were Latin, in
which letters "W" and "k" are rare. My search for definite fingerprints became
an exercise in probabilities, trying to estimate frequencies of the characteristic
sorts in scattered pools of the target font. In the end, though, my survey supported
Nicholas Okes as the most likely owner of the type used in quire D setting
B. 31 A hardier investigator than I might be able to clinch the identity with
a survey of broken types in Okes' other productions, but for narrative simplicity
in this account I will treat it as a certainty.

Though Okes' type may appear in setting B, we cannot assume that he had
personal responsibility for the printing. A series of undated petitions beginning in
late 1635 shows a deteriorating relationship with his partner. Okes' most interesting
complaint is that Norton, "wanting Capacitie himself for the Government
of a presse hath enterteined ayded & assisted a Company of disorderlie & factious
persons for the erecting of an unlawfull presse in a secrett place, and hath
without [my] knowledge & Consent ... secretly conveyed out of [my] house [my]
formes and lettres which afterwardes were discovered and seized by the master
& wardens of the Company of Stacioners & at their hall melted and defaced to
[my] great disgrace and losse." 32 The chronology of these events is vague, but
another document, dated January 1635, reinforces the impression that both men
were accustomed to playing fast and loose in their business. This list, endorsed
by Sir John Lambe of the High Commission, claims that the two printers jointly
employed, in addition to five authorized apprentices, six individuals "brought
upp to Printing against order". 33 The latter group, which included a butcher,
a gingerbread maker, and a garbler, may have constituted the "disorderly and
factious" team that Norton recruited to run his clandestine press. There is no
evidence that the "Okes" ToP quire D was printed off-site, or even that Norton

had such an operation while TOP was being printed. But the partnership clearly
operated in the shadows, and we cannot take for granted the ostensible paternity
of any of their productions.

Okes' setting of quire D occurs in seven of the seventeen surviving copies of
Phase III and in four of the nine copies of Phase IV Possibly Norton and Okes
each printed about equal numbers of sheets to supply the last two phases, or Okes
printed rather fewer. I have not found any variants in Okes' sheets. Although the
typesetting is not error-free, one gets the impression of a more assured compositorial
hand than Norton's crew could muster. It is not clear why Norton brought in
a second printer at this late stage to run off a "duplicate" setting of the quire. In
§16, I will show that the time gained by running the same eight pages of text on
two presses was nearly offset by the time taken in setting them anew. The complications
of using a clandestine off-site press would further reduce or even negate
the advantage. This consideration, along with the greater professionalism of
Okes' quire D setting, argues against Norton's direct agency in its production.

We know that Phases III and IV of sheets A-C contain multiple impressions,
so we would expect sheet D to show evidence of this as well during Okes' involvement.
However, as mentioned, Okes' sheets contain no variants that I have
identified. Norton's sheets for Phases III-IV are also virtually identical, but D1r
shows some interesting small changes:

2a1: l. 10 has "Masqncrs"; signature mark under the "hi" of "white,"

2a2: l. 10 has "Masquers"; signature mark under the comma of "white,"

2a3: l. 10 has "Masquers"; signature mark under the "w" of "white,"

From the spelling correction it appears that states 2a2 and 2a3 follow 2a1. The
shifts in the direction line raise the spectre of reimposition. However, the signature
mark—anomalously, an italic "D"—appears to be the same type in all three states.
(Transient bits of foreign matter appear either as specks within the letter [mostly
in 2a2] or falling across one or the other serif.) The page number also appears
invariant in composition and position. It would be tempting to attribute the states
of the direction line to random shifts due to loose lockup, except that there are n
ointermediate states, and the variants tend to sort in a complex but fairly orderly
fashion with sheets of the other quires. Norton's states of sheet D show the following
distribution:

TABLE 29. Correlation of Norton's D1r states with B(o) and C(i) in Phases III-IV

The table shows that state D1r:2a1 is found only with Phase-III copies of the
other quires, while 2a3 associates (with a single exception) with Phase-IV copies.
State 2a2 (shaded) is transitional between 2a1 and 2a3, and it occurs in copies
having B(o) from Phase IIIb. Those same copies have unique states of pages in
the inner forme of sheet C. In §11–14, I will return to these links among the
three quires. For now, it appears that the D sheets containing Norton's D1r:2a3
could represent a discrete Phase IV, his final one for that quire. The "leakage" of
2a3 into Phase III, represented by the Newberry copy, could have resulted from
Norton's having printed sheet D a bit short in phase III. Thus, a small group of
Phase-III copies of the book may have been made up with Phase-IV sheets of
quire D. Those hybrids probably went to the warehouse along with the purebred
Phase-IV copies.

If we accept that Norton printed Phases III and IV of quire D as distinct
impressions, what about Okes? Might he have printed all his sheets for both
phases in one go? If we allow this possibility, we have to explain the fact that, in
the surviving copies, Okes' contribution makes up about an equal proportion in
each phase: respectively 41 and 44%. When, after Phase III, a workman began
collating complete copies, he would probably start picking D-sheets from a homogeneous
batch of one or the other man's production, whichever came to hand
first. This process would almost certainly produce different Norton/Okes ratios
in Phases III and IV So, either the D sheets after Phase III already contained
about 40% of Okes' work, or Norton deliberately manipulated the ratio to equal
that in the upcoming Phase IV. Such an intervention supposes obscure motives
and more fastidiousness than I can credit Norton with. The same objection applies
to a hypothesis that both men printed their sheets for Phases III and IV
in a continuous run. The more plausible explanation is that Norton brought in
Okes to print a half-share of Phase III sheets, but asked him to keep the eight
pages in type, skeletons intact, against a probable fourth printing. When that
need came about, Okes was able quickly to reimpose the pages without alteration
for the new run. This scenario asks a high degree of precision from Okes, and
requires that he tie up eight pages of type without knowing for certain whether
they would be needed—otherwise, why not go ahead and print Phases III and
IV together? I cannot give a detailed and definitive account of Okes' work cycle.
It is intertwined with the mystery of why he got involved at all; if we knew the
answer to one, the other would probably follow. However, I favor the theory
that Norton's and Okes' teams worked in tandem on Phase III of quire D, and
that—on the slim evidence of surviving copies—each produced an approximately
equal number of sheets, of which Norton's contained states D1r:2a1 and
the much less common 2a2. Once they reached the end of Phase III, the pages
of type from both printers' pairs of formes were tied up and there was a pause,
during which completed copies from Phase III were assembled and dispatched.
After an indeterminate time came Phase IV, for which both printers again machined
approximately equal runs of their respective settings of quire D. Norton's
sheets of these final copies differ from those of Phase III only in the position of
the signature mark on D1r, which moved about three millimeters, probably during
the reimposition process, producing state 2a3 of that page.

TABLE 30. Phases in the printing

Norton and Okes seemed to feel that a further impression would be called for
after Phase III, but something prevented them from going on to print a double
run of quire D. The publisher Cooke may have been uncertain about demand.
But Phase IV also ushers in the bulk of the substantive textual variants in ToP
These occur in the description of the procession on A2r and A3r, in the tavern
direction on B3r, and in antimasques on C1v and C2r. None of the variants are
stylistic; they seem rather to document actual changes in the performance. By the
end of Phase III, Norton and Okes may have known that revisions were coming.
Although, in the event, none of these fell in quire D, the printers may not have
known this, and therefore put the type-pages for the last quire aside until the
matter was settled.

The singleton epilogue leaf "A speech to the King and Queenes Maiesties",
found in only four copies (and Greg's own copy, now untraced), 34 exists in a single
state, printed by Norton using mostly his workhorse italic i-1. The page is headed
by the now-familiar motif of a double row of acorns (28 pairs in this case). I will
consider the history and status of this addendum in succeeding sections.

§9. Order of Printing the Quires

In the preceding sections, we have been exploring ToP "vertically", as it were
—working out the histories of individual pages, studying their chemistry as they
combine into formes and quires, and watching each quire evolve in isolation
from the others. It is now time for a "horizontal" approach as we begin to reconstruct
the progress of printing the whole text, phase by phase. Table 30, a
modified version of appendix 3, shows how these phases align across the quires
of the surviving copies. It shows a fairly orderly stratification: for a given copy
of the book, the phase of any one of quires A-C serves as a good predictor of
the phases of the others. The few exceptions to this among the survivors result,
I believe, from the occasional necessity of making up copies with sheets from
neighboring phases. (I will discuss this further in §11–14.) But the phases do not
leapfrog one another: II and IV never mix in a copy, nor III and I or IV and I.
If sheets from more than one phase were available simultaneously for gathering
complete copies, there was nothing to stop Norton from mixing them indiscriminately,
because the text-breaks across quires never change. Except for the "third
impression" copies he would have no motivation for keeping the products of
one phase segregated from another. Therefore, I think we can safely hold to the
theory of four discrete printings of the book as a whole.

The recurrence of types can give us clues about the order of printing the
quires. (Lacking evidence to the contrary, I will assume for the ensuing discussion
that the order of printing in ToP is the same as the order of composition.) We would
expect Norton to work through quires A-D in sequence. But here, as so often, ToP
hands us a surprise. The recurring types in the first phase of printing that I could
confidently identify are distributed as follows:

A. PHASE I

TABLE 31. Recurring type material in Phase I

On first glance, the table appears to show nothing out the ordinary. Quire B
inner forme could have been set with type distributed from the outer forme of A;
Norton could even have begun at the title page and proceeded straight through
the book. But we need to take into account what happened to the individual
pages at the end of Phase I. In fact, the outer forme of quire A remained in substantially
the same setting through the printing history of ToP, whereas quire B
was distributed after Phase I. So, types from A(o) were not available to set B(i)—
rather the opposite. Therefore, A(o) was set after B(i) went through the press and
got broken down. Similarly, quire "a" continued to hang onto most of its type
(predominantly 96 mm italic) while D2r was raided for its italic after Phase I. (I
have indicated the point of the raid with a wavy border in table 31, distinguishing
the fate of that page from that of D4r, which was entirely distributed.) The
preliminaries could not have been set until printing was finished on quire D inner
forme. Furthermore, two letters on D2v show that that page could not have
been set until after the distribution of at least part of each forme of quire B; and
a distinctive bracket on C2r shows that C(i) cannot have been set before B(i)
was off the press. So, quires C and D both followed B. In §8A, I showed, on the
evidence of one broken parenthesis in the headline, that quire D was probably
printed after quire C. No other types from C show up elsewhere; it was probably
set while quire B was still being worked off, after which it held onto most of its

roman type through the succeeding phases. Italic raids on C after Phase I must
have supplied some other printing job than ToP. If, then, we may consider entire
quires as production units, it appears that the printing of Phase I followed the
sequence B-C-D-A-a. Nothing in the data shows that the preliminary quire "a"
had to come last, as long as it followed D from which it drew some of its type. But
preliminaries were normally printed at the end of the job, and if "a" followed D
directly, there would have had to be a pause for D(i) to be at least partly distributed.
(Remember that in ToP the inner formes were printed after the outer ones,
and it is only D(i) that provided types for quire "a".)

Something caused Norton to place quire A late in the schedule. A probable
reason emerges when we recall that quires B-D contain the text of the masque,
while quire A sets the stage, as it were, with a description of the costumes, scenery,
and the public procession that opened the performance. Most of this is written
in the past tense, as if reporting the finished event, while the stage directions
for the masque itself are mostly (though not exclusively) in the conventional
present tense. This division between reportage and direction again raises the
question of whether copies of the book were in fact available on the day of the
first performance—a matter which I will explore further in §16. The important
point here is that the details of the procession were more likely to be in flux until
the last minute, whereas a relatively stable text of the masque would have been
needed earlier so the performers could learn their parts. Though both sections
show light revision through the printing history, it would have made sense to
start work on the part least likely to change. Quite possibly, Norton began typesetting
the main text before Shirley had even finished writing the description of
the procession.

If typesetting and printing began with quire B, as the evidence above indicates,
this may explain why this quire alone was mostly distributed after Phase I—
seven of its eight pages broken up, as opposed to only one page each in quires A
and D, and none in C. I have already mentioned the often arbitrary pattern of
which pages got distributed and which survived into later impressions. However,
this first quire printed could represent an initial intent to play by the rules and
distribute type after running an impression—a routine that Norton could discard
if he learned that he would have to achieve faster production than planned.

Two other observations are worth mentioning here:

• Quires A, C, and D have a type-page width of 89–91 mm, but quires "a" and
  B measure 97 mm—three ens of english type wider. 35 These different widths
  imply at least two differently adjusted composing sticks, suggesting more than
  one compositor.
• 2.The last page of quire B has some overly generous leading, indicating some
  distress in copy-fitting. This could reflect a compensation for the unexpected
  use of the wider measure just mentioned.

B. PHASE II

TABLE 32. Recurring type material in Phase II

After Phase II, only four pages were distributed, plus italic portions of D2r
and D3r; recurring types therefore provide less evidence for the order of printing
in this phase than for Phase I. (The wavy border below D2r in this table shows
that only the italic portion was reset; one of its types turns up on B4v.) Here
again, it appears that quire A was printed after quire D: D1v was distributed at
the end of this phase and two of its types show up on A2r. On the other hand,
quire A held onto most of its type from this phase onward and could not have
supplied it to D or any other quire. It also looks (on the evidence of only two
types, but both are quite distinctive) as if quires D and C were finished print-
ing before quire B could be set—again, because quire B carries its types largely
unaltered into the next phase, while the donor pages in quires C and D give up
type after Phase II. These observations show that quires C and D were printed
before quires A and B. The fact that Phase II of quire A actually consisted of two
separate printings complicates the picture. In §5, I suggested that, after an inter-
ruption in the Phase-II printing of A(o), Norton chose to perfect the sheets he
had already printed rather than resuming the planned run of the outer forme and
then perfecting both sub-impressions in a single operation. The motive would be
to prioritize the completion of at least part of the planned impression of sheet A
in order to get some complete sets of sheets to the publisher for sale. This strat-
egy would make more sense if quire A came later, even last, in the sequence of
quires as printed. However, if quire A came last and if therefore B followed D
directly, the composition of B's outer forme (the first one printed in B) would have
had to wait for the italic from D2r (in the second forme printed in D) to become
available. This would leave the press idle, at least for purposes of machining ToP.
If, instead, A followed D directly, its outer forme would still be standing from
Phase I and printing could have proceeded on that while D was broken down
and the types incorporated into A(i). So, the slightly more likely sequence for this
second phase of printing was C-D-A(Phase IIa)-B. Probably Phase IIb of quire A
came last, but it cannot have been delayed long because two of the three surviv-
ing books that contain it also have Phase II of sheets B-D.

The textual changes in quire A are minor for Phase II and would have had
no bearing on a decision to leave the production of that quire for later. Possibly it

came later in the sequence because that was the order followed in Phase I. Quire
B, on the other hand, came into this phase with only one type-page intact; all the
others had to be reset. So, it would have made sense to commence printing with
quires C, D, and A, leaving time for the resetting of B.

I have left quire "a" out of the discussion for this phase because it holds onto
most of its original types throughout the production. We would normally expect
it to come late in the series, probably after quires B and/or A(Phase IIa).

C. PHASE III

TABLE 33. Recurring type material in Phase III

This phase began with the resetting of four pages, italic raids on two others,
and some localized resettings. Only the inner forme of quire B was distributed
before the next phase. Its donation of a distinctive bracket to C1v shows that C(i)
was composed after B(i) had finished printing. This is the only evidence for work
sequence that recurring type-matter affords in Phase III. (Of course Norton's and
Okes' typesettings of quire D did not share any material.)

Italic raids before this phase point to a continuing need for english italic for
another job; the distributed material does not reappear in reset portions of ToP.

D. PHASE IV

TABLE 34. Recurring type material in Phase IV

Both Norton's and Okes' quires D continue intact from the preceding phase,
and most of sheets A-C also enter Phase IV unchanged except for localized reset-
tings to accommodate revisions and resulting page-break changes. The exception
is quire B inner forme, which was entirely reset. As usual, there is no obvious
reason why Norton reset this forme and not others. But table 34 shows that some
of the types from the first page of text (A1r) turn up again in the two last pages—
B3V and 4r—of the newly set B(i). So, A(o) was already printed and being dis-
tributed before B(i) was ready for the press. A single letter "a" shows that Norton
printed sheet D and distributed some of its type before resetting three lines on
B3r. Therefore—if we again take sheets to be the units of production—both A
and D preceded B through the press. Quire C contains no distinctive types that

allow us to place it in a sequence, and again the few changes to quire "a" are
insufficient to establish links with any other. The sequence, then, was A-D-B,
with the positions of C and "a" unknown, though as usual we would expect the
preliminaries to come last.

Though the evidence of recurring types is very incomplete, it shows that Nor-
ton distributed at least portions of quires A and D while other quires remained to
be composed. It looks as if Norton knew that he had reached the end of the pro-
gram and there was no point in leaving type standing against further reprints.

Some type-matter distributed from Phase III found its way into reset passages
in Phase IV:

TABLE 35. Recurring type material between Phase III and IV

It appears that the types freed up by the distribution of B(i) after Phase III still
lay near the top of their boxes, available for picking up for Phase IV alterations
in quire A. This implies a rather short delay, or none at all, between Phases III
and IV.

A different case is the singleton leaf bearing "A speech to the King and
Queenes Maiesties", surviving in four copies of the final phase. It is set mostly
in Norton's italic i-1. I could identify none of its types with those used in the rest
of the book in any phase. This lack of recurrence is not too surprising, given the
small sample size, but it would also be consistent with a significant passage of
time between the distribution of ToP's main text and the setting of "A Speech".
During this period, the ToP types might get deployed on other jobs or covered in
the boxes by other types more recently distributed, making them less likely to be
picked for setting the new epilogue.

§10. Skeleton Formes

The patterns of transfer of the visible parts of the skeleton can sometimes
yield information about the progress of formes through the press. In ToP, the only
identifiable recurring components of the headlines are the page numbers and
their surrounding parentheses. The pagination of the book is [i-iv], 1–8, 2i -24;
as quires A and B have duplicate pagination, whole headlines could potentially
have been transferred between them without change. The digits themselves provide
scanty evidence for the re-use of skeletons: though the workman imposing
the forme might have recycled digits from previous pages,
the types would have
tended to shuffle positions randomly, and it is often difficult to tell whether one
of them is damaged or is just another design variation in Norton's heterogeneous
stock. Pairs of parentheses, on the other hand, occur on every page and could

very easily be re-used along with the surrounding spacing material that held the
number more or less centered. Being long, thin lines, parentheses are prone to
damage, and particularly in the case of a battered stock like Norton's, they ac-
quire an individuality that bibliographers can take advantage of. In my catalogu-
ing of those in ToP, I relied on visibly injured individuals, but made no attempt to
classify the infinite variations among the sound examples that could result from
different cuttings, variations in inking and impression, or even inversions.

A. PHASE I

In §9A, I showed that the sheets in Phase I were probably printed in the
order B-C-D-A-a. The preliminaries do not participate in the sharing of visible
skeleton formes, and in the following table I have arranged the other formes in
my proposed order of printing. In this and the subsequent tables, the codes fol-
lowing the page references stand for pieces of type. The abbreviations L, R, and
# indicate left and right parentheses and the digit(s) they enclose. A code "L#R"
signifies that the headline was transferred unaltered from a page earlier in the
printing. An "x" indicates an individual type that is undoubtedly different from
the one in the previous appearance of the headline; a question mark means that
the type is not distinctive enough to tell whether it is the same or different.
For example, in the table below, the left parenthesis from B1v recurs on C1v; the
page number changes from 2 to 10, so neither of the two digits can be the same
types; and the right parenthesis might be the same one from B1v, but I could
not verify this. The same left parenthesis then migrates to D2v, where again the
the following digit of the page number changes from 1 to 2 (hence "x"), and the
next zero and right parenthesis could be the same types from C1v but are not
distinctive enough to tell. So, all we can say definitely about these two transfers
is that the left parenthesis is conserved. We cannot say even that the transfer was
direct from one forme to the other; the headline could have been distributed
and the type coincidentally picked off the top of the pile in the case for setting
the new headline. In the first scenario, the action is deliberate and systematic, in
the second it is unconscious and random. There is perhaps a tendency for types
to remain in the same position in the new forme; codes printed in boldface
in the following tables signify pages on which the type material recurs in the same
quarter as in its immediately previous use.

TABLE 36. Recurrence of headline types in Phase Ia—b

The appearance of Phase I is not one of particular regularity. Every identifi-
able parenthesis switches at least once between outer and inner formes during
the printing history. A slightly puzzling feature is the rapid re-use of certain
types, at the transitions marked by the dashed vertical lines. For instance, the left
parenthesis on C1v (inner forme) recurs immediately in D2v (outer). Since outer
formes of ToP sheets were always printed before the inner ones, this proximal
re-use means that the imposition of D(o) had to wait for the machining and strip-
ping of C(i) to receive some of the latter's headline. If presswork were proceeding
smoothly, it would have avoided a bit of delay to take D outer's skeletons from
C(o), or from somewhere else further back in the production. The same sharing
occurs with two types between D(i) and the next following forme, A(o). These
"pinch points" suggest small delays in the workflow after quires C and D, pos-
sibly to accommodate other jobs or even some of the composition of ToP itself.
We can eliminate some of these points if we drop the assumption that the order
of printing is the same as the order of composition and rearrange the quires in
the table above; but no arrangement gets rid of them entirely.

The parentheses that start their journey on B4v travel through all four quires
of the main text, remarkably ending up on A4v with the same damaged "8". (My
notation system cannot show that particular coincidence.) Because quires C and
D intervened, this distinctive type could not have travelled with the parentheses,
but was evidently put aside, or back into the case, after B(o) was printed, and
randomly retrieved for use in A(o).

B. PHASE II

TABLE 37. Recurrence of headline types in Phase II

Type-recurrence analysis (§9B) shows that the order of printing the quires in
this phase was probably C-D-A(Phase IIa)-B-A(Phase IIb). Table 37 follows this
order. In comparison with Phase I, Phase II seems to have proceeded in a much
more orderly fashion. Recurring characters leapfrog adjacent formes, so that any
one piece of type will appear only in the outer or the inner, and always in the
same quarter of the chase (as shown by the bold typeface in the table). All three
characters of the A2v headline, including the page number, recur on B2v. It is
odd not to find any headline material from quire C participating in a transfer, as
all of its pages were either reimposed or completely reset after it was printed.

Quire A was printed in two distinct stretches of work, Phases IIa and IIb.
After the first of these, the skeletons were stripped from five of its eight pages.
Table 37 shows that A(o) in Phase IIa shares much of its headline material with

the corresponding pages of B(o). When A(o) went back on the press for Phase IIb,
all of its headline material was new. By the time I realized this, the three copies of
A(IIb) were not easily accessible and I was unwilling to exasperate my colleagues
with yet more requests for images. Therefore, I have no specific information on
the headlines of A(IIb).

C. PHASE III

In §9C, I was unable to find evidence for the sequence of printing the quires
in this phase, so I tabulate them here in alphabetical order:

TABLE 38. Recurrence of headline types in Phase III

This phase shows regular patterns of movement of headline material between
quires A and C, and B and D. Once again the printer either did not have or did
not take the opportunity to transfer the page numbers intact between sheets A
and B. The re-use of material between B(i) and C(o) would create a "pinch point"
if Norton followed the order shown, because the imposing of C(o) would have
had to wait until B(i) was off the press, cleaned, and stripped. There are several
ways of reordering the quire sequence to place comfortable intervals between the
re-uses shown in the table: repeating Phase II's order of C-D-A-B would be one
of them. But as we lack evidence of recurring types in the text, the most we can
say is that it is slightly more likely that Norton printed the quires out of sequential
order, but we do not know what that order might have been.

The five surviving copies from Phase IIIb are characterized by unique page-
number settings in the outer forme of quire B. Their typographic material shows
no distinctive forms that occur elsewhere in Phase III.

Okes could conceivably have transferred some headine material between the
two formes of his quire D. Visual inspection shows no evidence of this, partly
because because his type was more uniform and in much better condition than
Norton's and the parentheses are too regular to show useful differences.

D. PHASE IV

From the evidence presented in §9D we know that the printing of sheet A
in Phase IV preceded that of quire B. This phase is unique in having all the
headlines from A(i) transferred intact to the same-numbered pages in B(i).
We would expect a similar transfer between the outer formes, but this did not oc-

TABLE 39. Recurrence of headline types in Phase IV

cur. Norton's printing of quire D carries over its headlines from Phase III, and
quire C has only two new headlines in eight pages, with none of their material
used elsewhere. Okes' headlines remain with their type-pages, unchanged from
Phase III.

In conclusion, the patterns of headline re-use during the ToP printing history
show that Norton's men took only partial advantage of this potentially labor-saving
practice. This lack of regularity might be interpreted as evidence of a chaotic
work-flow, but in fact, it is not that common to find any book of the period to be
completely regular in this respect, and many show no headline re-use at all. The
limited evidence of the practice in ToP gives the appearance that Norton settled
into a slightly more systematic routine as he worked through the impressions.

GATHERING THE SHEETS

Now that we have laid out the evolution of each page of ToP and considered
how the pages shared typographic material in their progress through the several
phases of printing, we will examine how the various states group themselves
together in formes and sheets, and how the finished sheets were gathered into
copies ready for folding and binding. Earlier investigators (notably Unger, Jack-son, and Greg) were perplexed by the range of permutations of states in copies
of ToP that they examined. They attempted to explain some of the complexity as
resulting from duplicate settings run off concurrently to satisfy the expected high
demand for the text. While this was true diachronically—Norton did, in fact,
set many of the pages more than once through the various phases—it is not true
synchronically: different settings were not in production simultaneously except
for Norton's and Okes' quire D. Instead, surviving copies show a degree of ho-
mogeneity consistent with separate phases of printing the whole book, each phase
yielding a discrete batch of complete copies that left the establishment before the
products of the next phase could mix with them. The uncommon exemplars that
show mixing between phases can be attributed to slightly unequal supplies of in-
dividual sheets at the time of collation, requiring some copies of a given phase to
be completed with products of the previous or following one. 36 The ensuing dis-
cussion will refer to appendix 3, and it assumes that the proportions of surviving

copies of the various states shown in the table bear some statistical relationship to
the numbers that were originally printed. I have found no evidence that Norton
treated the products of any one phase with particular favor, nor do I have any
reason to suppose that different versions had different rates of survival.

§11. Phase I

Appendix 3 shows that eighteen of the nineteen copies surviving from Phase
I (about 95%) are homogeneous throughout (ignoring stop-press corrections).
In other words, they contain the earliest settings or impositions of all the sheets
including the preliminaries. The one exception is Huntington 69434, in which
quire D comes from the next printing. This is an example of the situation just
described in which one quire was evidently printed a bit short in one phase and
had to be completed from the next. There is nothing special about quire D that
might have led to a short printing; recall from our type analysis that it was prob-
ably printed third of the five quires in Phase I.

If quire D was printed short, it seems that quire "a" was printed long. In
addition to the nineteen complete copies from Phase I, four copies from the
second printing have Phase I preliminaries. This extrapolates to a 21% surplus
of first-printing quire "a"—an odd amount, being too little to represent an in-
tended second impression but too large to have resulted from a trifling accidental
overrun. The discrepancy persists in Phase II, in which each quire survives in
approximately equal numbers, but a residue of preliminaries is pushed over into
Phase III copies. In Phase III, Norton apparently shorted the run of quire "a"
by a number about equal to the initial overrun, so Phase IV copies are finally
homogeneous (if we disregard Okes' contribution)–all "third impression" title
pages come with text sheets from the last printing. The phase relationship of
quire "a" with the main text looks like this:

TABLE 40. Staggered distribution of production phases of quire "a"

The first two impressions of the preliminaries may occur with either of two
impressions of quires A-D, and the second and third printings of the main text
can each come with two possible variants of the title page. It was this ladder-like
asynchrony that led earlier bibliographers to believe that Norton had printed
simultaneous duplicate settings of the text.

In a given copy, we should not necessarily expect early stop-press variants of
any one sheet to sort with early states of the others, or later with later, though
there may be a tendency for this to happen. 37 Appendix 3 indeed shows an es-
sentially random pattern in this regard.

§12. Phase II

Phase II represents a print run that was probably smaller than the first; there
are ten survivors, as opposed to nineteen for the first printing. The phase is com-
plicated by the two sub-phases of sheet A, the products of which consort with the
other sheets of the book in contrasting ways. Sub-phase IIa is found only with
Phase II of sheets B-D, and it sometimes occurs with the residue of Phase I prelimi-
naries. Sub-phase IIb, on the other hand, comes only with Phase II prelimi-
naries, and one copy (Folger STC 22459.2) even has quires B-D from the third
phase. These groupings support my earlier assertion that Phase II consists of two
batches of copies that left the shop at different times, or were at least sequestered
so they could not intermix. The A sheets may have been a bit over-printed, as
evidenced by the Folger copy that is hybridized with Phase III.

TABLE 41. Schematic contents of copies with quire A from Phase II

§13. Phase III

This is the phase where the quire D sheets printed by Nicholas Okes first
make their appearance, in 41% of the surviving copies from that phase. Most,
though not all, copies come with the title page omitting the words "of Grayes
Inne". During the course of Phase III we see the skeleton changes to the outer
forme of quire B that define Phase IIIb of that sheet, as discussed in §6.

The following table shows the composition of Phase-III copies. The patterns
of association, based on a very small sample size, are at particular hazard of be-
ing broken by new copies of ToP that may happen to emerge, and the following
discussion need to be read with that caveat in mind.

TABLE 42. Schematic contents of copies with quires B—D from Phase III

Three copies with quires C and D from Phase III (shaded above) contain
quire "a" from the previous phase, and one of these also has quire A from the
previous phase. Because of their residual Phase-II content, they are likely to have
been among the first exemplars from Phase III to be gathered into complete cop-
ies. Two of these three copies contain Norton's quire D and one has Okes'. This
statistically equal representation is consistent with my earlier conclusion that the
two men's sheets were effectively printed "simultaneously", so were about equally
available to use in any copy of Phase III.

In §6B, I discussed an imposition of B(o) that I called Phase IIIb. The five
copies containing it show some puzzling patterns in their makeup. In most re-
spects they behave like normal copies from Phase III. However, none of them has
quire "a" from the earlier phase, and all five have quire D in Norton's printing.
The first point is not especially surprising, given the rarity of Phase III copies
with quire "a" from Phase II. On the other hand, if Norton's and Okes' printings
of quire D were equally available to make up those five copies, the probability
of their ending up only with Norton's is 1/32, or about 3%; so perhaps this does
call for some explanation. In §6B, I argued that Phases IIIa and IIIb of B(o) were
perfected in one run with the same inner forme. In the resulting stack(s) of sheets,
the two sub-phases would probably have remained separate strata. Hanging the
sheets up to dry could have led to some intermixing of the two states, but by the
time Phase III copies were assembled, the two sub-phases of sheet B might still
have had enough spatial segregation to create dissimilar patterns of association.
If, for example, the IIIb's ended up in a remote stack and were retrieved late in
the collation process, and if the supply of Okes' quire D had arrived somewhat
before Norton's and was exhausted first, there could be a tendency for Phase IIIb

of sheet B to end up with the remainders of Norton's quire D. But we cannot draw
any conclusions about what happened, and with the discovery of more copies the
pattern shown in table 42 could easily fall apart and we would have something
different to explain.

Phase IIIb of forme B(o) also associates with unique states of forme C(i) and
page D1r:

• C1v has the page number and parentheses displaced about 2 mm to the right
  with respect to other copies of Phase III.
• On C4r, in the same forme, the page-number types have moved 4 mm to
  the right.
• In both these cases the page-number materials are unchanged except in position,
  and there is nothing to indicate whether the states are earlier or later than their
  Phase III cohorts.
• On D1r, the italic signature mark "D" shows three possible positions in the
  Phase IIIb copies:

• 2a1: Under the "hi" of "white," in the last line. This is the normal position
  in Norton's Phase IIIa copies, but it is also found in the Phase IIIb copy at the
  Bristol Central Library. Line 10 has the misprint "Masqncrs".
• 2a2: Under the comma. "Masquers" is now correctly printed, so this state is
  later than the preceding, and the handling required to make the correction may
  be sufficient to explain the signature-mark shift. The state is uniquely associated
  with Phase IIIb of B(o), occurring in three of the five copies.
• 2a3: Under the "w". Line 10 also has "Masquers". This state is found in
  the Newberry copy of Phase IIIb, but also in all of Norton's Phase IV sheets, so
  it is evidently the latest of the three. Its presence in the Newberry copy prob-
  ably results from a slightly short run of quire D which had to be made up from
  Phase IV.

The materials and alignments of the headlines are invariant in all of these
states.

The five copies containing Phase IIIb of quire B are already unusual in their
B(o) skeleton, but the association of that skeleton with unique states in two other
quires is unsettling. I cannot imagine an event happening during the printing of
one sheet that would manifest itself at about the same point during the printing
of another. The arbitrary realignments in quires C and D are reminiscent of St.
Catharine's College Z59, with its unique page-number placement on B4v (§6B);
but here they coincide with the unique skeleton settings of B(o). Three of the
seven states that appear during this sub-phase improve the readings found in
their Phase III cohorts, 38 though another (B2v:2c2) actually introduces a textual
corruption that persists through Phase IV. One is tempted to suggest that the
copies containing Phase IIIb of quire B are witnesses to a separate production
phase of the entire book, with no changes except the reset skeleton of forme B(o),
corrections to D1r, and incidental shifts on C(i). (There were later deliberate
changes and attendant shifts on B2v, B4r, and D1r.) How the five IIIb copies

came to combine all these unusual features is, to my mind, ToP's greatest remain-
ing mystery.

§14. Phase IV

The latest copies of ToP—those which the title page calls "the third impres-
sion"—comprise the most homogeneous group in the whole publication history.
Of the 27 pages in quires "a" and A-C, 24 have settings, impositions, or vari-ant states that are unique to this printing. 39 The sheets of quire D printed by
Norton for this phase are characterized by state D1r:2a3, whereas his output
for Phase III contains D1r:2a1–2. (The latest state of that page also occurs in
the Newberry copy, one of the five from Phase IIIb discussed in the preceding
section; but we can probably regard this as another instance of an earlier-phase
copy that had to be completed with a quire from the next phase.) No unsophisti-
cated copy of Phase IV contains sheets from earlier printings, if we except Okes'
quire D, which I have argued was likely the product of two separate print runs
that are now indistinguishable.

The single-leaf "Speech to the King and Queenes Maiesties," alluding to
the second performance of 13 February, is bound into four of the nine traceable
copies of Phase IV. Its placement among the sheets would probably have been at
individual binders' discretion. As a sort of dedication, it could logically belong at
the end of the preliminaries after Shirley's original dedication; or, as it referred to
an event later than the first performance documented by the book, it could come
last. The copies I have seen show two in the first location and two in the other
(and Greg reported that his copy had it bound after quire a.) Being a singleton, it
would have been prone to loss, and one does not know whether it was originally
included with the sheets of all Phase-IV copies or whether it was available only
for those sold later. I could not find any of its damaged types elsewhere in the
book, so there is no obvious evidence that it was set soon after the last phase of
printing. In §16, I will give archival evidence that might have a bearing on the
addendum, but from the physical evidence it is not possible to say anything with
greater certainty.

§15. A Digest of the Workflow

In this section I attempt a narrative of the entire production cycle of ToP. It
follows the order of printing the quires in each phase as far as I have been able
to determine. "Distributed" means the entire page was broken down and the type
put back in the cases. "Stripped" means that the skeleton (always the headlines,
and generally the direction lines as far as I can tell) was removed—either for
distribution or for transfer to another forme—leaving the bare type-pages to be
tied up for later impression. "Replaced" means that the tied-up pages, complete
with their previous skeletons, were newly imposed for reimpression. When all the
pages of a forme were replaced unaltered, I would not rule out the possibility that
they were still locked in their chase when set aside.

As it is difficult to assign a significance to transfers of individual types in skel-
etons, I have mentioned only cases where two or more types in a cluster seem to
have been moved together.

A. PHASE I

As I demonstrated in §9A, the quires were likely printed in the order B-C-
D-A-a. Quires B and C have different measures and could have been composed
simultaneously. Quire "a" has the same measure as quire B and was possibly
composed with the same stick. English-sized type (20 lines = 96 mm) is used for
the text throughout.

B outer forme

Pauses for correction: one, affecting B1r only: heading reset with characters'
names in roman. The same page later incurs damage to ll.7 up and 9 up.

After printing: the stack of sheets is flipped in a non-standard but explainable
way (§6A), scrambling the states during the perfecting run. All pages are distrib-
uted except B3r, which is stripped and kept standing. One piece of type from
B4V reappears in D(o), and parentheses from the B4V headline are transferred
to the headline of C4r.

B inner forme

Pauses for correction: one: one comma added to B3v, two to B4r.

After printing: all pages distributed. One identified type reappears in D(o).

C outer forme

C4v has an egregiously cock-eyed lockup, the last lines of text bending up-
ward into a hole left by the omission of some spacers; page number 16 is mis-
printed "61".

Pauses for correction: three, eventually affecting all of the pages; one change
deletes a stage direction on C3r

After printing: all pages stripped. The italic songs on C2v, C3r, and C4v are
removed and distributed, presumably for use in another job but certainly not for
ToP; the remainder of the text is kept standing. The fact that C(i), an even richer
source of italic, remains untouched in this wave of raids shows that the songs in
C(o) may have been taken while C(i) was on the press and inaccessible.

The unperfected sheets are shuffled in an unexplained manner that results in
an anomalous distribution of the two states of the inner forme.

C inner forme

C4r headline gets its parentheses from B4V

Pauses for correction: one, correcting a spelling on C1v.

After printing: distributed: C3v, and C2r except most of ll.5–9. Other two
pages stripped and kept standing. At least one type from the C1v headline is
transferred to D2v in that quire's outer forme. That is the next one on the press,

so printing on D(o) probably follows a short delay. Headline parentheses are also
transferred from C4r to D3v (both on inner formes).

Some time before the next phase, C4r l.10 and C1v italic and brackets in
ll.8–10 and last three lines are removed and distributed.

D outer forme

Pauses for correction: one: a word added to D1r, making better sense but dis-
turbing the first three lines in the process. A bit of type damage takes place later
on D4V, and still later on the previously disturbed lines of D1r.

After printing: all pages kept standing with their skeletons.

The unperfected sheets are shuffled in an unexplained manner that results in
an anomalous distribution of the three states of the inner forme.

This quire is printed a little short.

D inner forme

D3v gets its parentheses from C4r.

Pauses for correction: one, fixing a typo and a misalignment.

It is possible that a small number of sheets printed with D(o) are not perfected
in this phase, but are perfected with the inner forme characteristic of Phase II.
(See the discussion of B.L. Ashley 1697 in §8A.)

After printing: D4r distributed, the rest stripped. Some headline material is
transferred to A(o). That is the next forme of ToP to go onto the press, so the
transfer probably causes at least a short delay in presswork.

Some time before the next phase, sections of italic are removed from D1v,
D2r, and D3v, to be used at least partly for setting quire "a". Remainder of text
kept standing.

A outer forme

A4v gets its parentheses from D3v (same position in forme).

Pauses for correction: none.

After printing: distributed: A1r ll.1–9 and the last; A4v ll. 1–2. Remainder of
text stripped and kept standing.

Some time before the next phase, four lines of italic on A2v-3r are also
distributed.

A inner forme

Pauses for correction: three, affecting spelling and commas.

After printing: A1v and bottom half of A2r distributed (from "These moving
forward ..." to end); other pages stripped and kept standing.

"a"

The dedication (pages a2r-v) contains italic raided from D2r and 4r. All
three pages of preliminaries are imposed together and printed by work-and-turn.
There is no evidence of the use of bearer types on blank a1v. Most likely, each
sheet is perfected immediately.

Pauses for correction: one, correcting a typo on a2v. Part-way through the cor-
rected state, the "O" in l.3 of the title page suffers damage.

After printing: acorns and direction line stripped on a2r, otherwise all type-
pages kept standing.

The number of sheets printed, on the evidence of surviving copies, is about
20% greater than those of the other quires.

At this point, the products of Phase I either leave Norton's shop or are pack-aged and remain unmixed with later printings. A few sets of sheets apparently
linger without quire D, which was printed a bit short.

Meanwhile, composition is proceeding on another unidentified job or jobs
which cause italic passages to be removed from the standing type in quires A,
C, and D.

B. PHASE II

The sequence of quires in this phase is less certain than for Phase I (see §9B).
The history is complicated by the two sub-phases of quire A, and there is no
evidence for the position of quire "a" in the sequence because after Phase I it
has no typographic interaction with the other quires. I have placed it last below,
as it was normal to print the preliminaries after finishing the text.

C outer forme

Preparation for printing: italic songs on C2v, 3r, 4v reset. On the latter page,
Song 5 and its title are set in pica (20 lines = 82 mm, with the title in roman;
the title "Song 3" on C3r is also in pica). All new skeletons. The distortion of the
last line on C4v is remedied.

The use of pica type on C4v probably indicates a shortage of the english size,
which is the default for ToP. The use of pica roman for the title "Song 5." may
be deliberate, to match the type of the song itself; but it also creeps into the title
on C3r, where the song itself is reset in english.

Pauses for correction: none. The page number on C1r shows some looseness.

After printing: all pages stripped and kept standing. None of the headline mate-
rial shows up in the other quires.

C inner forme

Preparation for printing: C1v (italic and brackets, ll. 8–10, and last three lines)
and C4r l.10 reset. Other minor adjustments to 3v and 4r. All pages reimposed
with new skeletons.

Pauses for correction: none.

After printing: C1v: italic and brackets in ll. 1–10 and last four lines distrib-
uted. Otherwise, all pages stripped and kept standing. None of the headline
material shows up in the other quires.

D outer forme

Preparation for printing: pages replaced along with their original skeletons. The
damage to the first three lines of type on D1r, already disturbed, probably occurs

at this point, producing state 1a3. If not, the damage happens early in the print-
ing (§8A).

Pauses for correction: none.

After printing: D1r distributed; all other pages stripped and kept standing.
Some headline material from D1r, 2v, and 4v is transferred to the same posi-
tions in A(o).

Some time before the next phase, two stretches of type are distributed: D2v
ll.18 through the end of the page, and the italic song on D3r.

D inner forme

Preparation for printing: D4r reset; five-line stage direction at the top of D1v
reset in roman; raided italic portions of D2r and 3v reset in the same italic font.
All new skeletons.

Pauses for correction: none. It appears that a residuum of sheets printed with
the outer forme from Phase I are perfected with this batch.

After printing: all pages stripped and kept standing except for the following,
distributed: D1v (all); D2r italic song ll. 1–8 only; D3V roman section. Headline
parentheses from D4r are transferred to A4r (inner forme).

A outer forme Phase IIa

Preparation for printing: the following are reset: A1r ll. 1–9 and last; A2v-3r,
four lines of italic (set in great primer size, 20 lines = 120 mm.); A4v ll. 1–2.
Textual revision to A3r.

All new skeletons, including the rule on A4v, but omitting its catchword;
the lines of acorns on A1r may be new. A2v headline gets its parentheses from
D2v.

Pauses for correction: none. The catchword on A2v goes gradually out of align-
ment. Printing is halted before all sheets of this forme are machined.

After printing: all pages stripped and kept standing. Some headline material
from A1r, 2v, and 4v is transferred to the same positions in B(o).

After a delay, apparently to allow other work to play through, it appears that
Norton decides to perfect this partial run of sheet A rather than complete the
Phase-II impression of outer-forme sheets.

A inner forme Phase Ila

Preparation for printing: A1v, and bottom half of A2r (from "These moving
forward ...") reset. Commas added on A4r. All new skeletons, including the row
of ornaments on A3v. A4r headline gets its parentheses from D4r.

Pauses for correction: none. A4r shows looseness in l. 7.

After printing: all pages kept standing: A1v stripped, the others with their
skeletons.

If Norton's strategy at this point is to rush out as many copies as he can with
the partial run of sheet A, the logical next step is to machine the two remaining
quires before returning to complete the impression of A:

B outer forme

Preparation for printing: B1r, 2v, 4V reset; B1r has 28 pairs of acorns, with two
inverted. On B3r (reimposed), rewriting of the tavern direction from one to three
lines forces the last four lines of type on this page onto the verso. B2V gets its
entire headline from A2v.

Pauses for correction: none. On B4v l. 2, the last two letters of "coold" are lost
in two stages.

After printing: all pages stripped and kept standing. B1r loses its acorns along with
the page number. The four overflow lines of type from B3r are respaced to three
lines and transferred to B3v, which is therefore probably being reset at this point.

B inner forme

Preparation for printing: all pages reset. B3v starts with the three overflow lines
of type from B3r, but the page break at the bottom is kept unaltered by reducing
leading and running a stage direction into a dialogue line.

Pauses for correction: one: page number on B4r is reset part-way through. There
is a confused series of seemingly accidental variants (§6C). One copy also has
page number on B2r set in a font not seen elsewhere.

After printing: all pages kept standing with their skeletons.

At this point Norton would be able to assemble a certain number of Phase-II
copies using the remainder of quire "a" from Phase I. He might now return to
print further copies of the preliminaries with some revisions:

"a"

Preparation for printing: title page realigned; direction line and acorn pairs
(27) on a2r newly set; comma added on a2v. The forme is printed by work-and-
turn.

Pauses for correction: none.

After printing: all pages kept standing.

At this point Norton can assemble further complete copies, exhausting the
supply of sheet A from Phase IIa. Unlike the previous batch, these copies all
contain sheet "a" from Phase II. Now Norton returns to complete the Phase-II
impression of sheet A:

A outer forme Phase Ilb

Preparation for printing: all new skeletons. A2v catchword marries with first line
of roman on next page. (That is the third line of text, so the workman evidently
overlooked the two short lines of italic that begin the page). Some quads left up
on A4v.

Pauses for correction: none.

After printing: all pages kept standing: A3r and 4v with their skeletons, A1r
and 2v without.

A inner forme Phase IIb

Preparation for printing: A1v gets new skeleton; other pages replaced with their
skeletons, but A3v page number shifts.

Pauses for correction: none.

After printing: all pages kept standing with their skeletons.

This brings the print run of quire A equal to those of the other text quires,
and the last copies from Phase II can be sent to the warehouse, uniquely contain-
ing quire A from Phase IIb. There may actually be a few copies of sheet A left
over: one surviving copy from Phase III contains quire A from IIb.

C. PHASE III

The sequence of printing the quires in this phase is indeterminate, though
rather more likely to have been in scrambled order. I give it here in the order
A-B-C-D-a.

A outer forme

Preparation for printing: A1r and 2v get new headlines; A3r and 4v are replaced
unaltered, but in the process the quads on A4V get lowered.

Pauses for correction: one? On A3r, one copy shows displacement and damage
at upper left, and slight displacement of the skeleton elements. Possibly during
the act of repairing the damage, the "3" goes missing from the signature mark.
The page number on A1r prints indistinctly or not at all.

After printing: all pages kept standing: A4V probably keeps its headline but
loses the rule; the others are stripped.

A inner forme

Preparation for printing: all pages replaced unaltered.

Pauses for correction: one: the famous "bride/bridle" correction on A1v.

After printing: all pages stripped and kept standing. Some headline material
appears in the same positions on C(i).

B outer forme Phase IIIa

Preparation for printing: all pages get new skeletons; B1r receiVes 25 pairs of
acorns, with endpieces. Headline parentheses on B4V are the same as on D4v. In
the process of reimposition, some disturbance and damage occur on B1r and 2v,
and the catchword is left off of 4v.

Pauses for correction: none. One copy shows a page-number shift on B4V.

After printing: all pages stripped and kept standing. B1r loses its acorns along
with the page number.

This printing of B(o) is somewhat short with respect to the other text quires
in Phase III—something prevents Norton from machining the full white-paper
stack. He will complete it when he has dealt with whatever has caused the inter-
ruption. Meanwhile, he strips the outer forme of B and puts it aside. The situ-
ation is similar to the interruption in the run of quire A in Phase IIa, but here
only the outer forme is affected and both sub-phases are backed up with the same
state of the inner forme. Once again it is illogical for Norton to strip the skeletons
if he expects to resume presswork soon. Nevertheless, after a presumably short

delay, it appears that B(o) gets reimposed and put back on the press for a short
completing run, which constitutes Phase IIIb of the outer forme:

B outer forme Phase IIIb

Preparation for printing: all pages get new skeletons; B1r gets 28 pairs of acorns,
all properly oriented. Typo on B1r gets corrected. The tavern direction on B3r is
rewritten from three lines to one again and positioned flush right.

Pauses for correction: one: on B4V the word "stndy" in l.13 is corrected, after
a fashion, by turning the "n" upside-down. Perhaps as an accident during this
operation, the comma after "Laughter' on B2V, l.23, goes missing.

After printing: all pages kept standing; B2v keeps its skeleton, the others are
stripped. B1r loses its acorns along with the page number.

Phases IIIa and IIIb of B(o) are now perfected together:

B inner forme

Preparation for printing: all pages reimposed along with their skeletons.

Pauses for correction: none.

After printing: all pages distributed. Headline parentheses on B3V are the same
as on C2v, which indicates that either C(o) does not follow B(i) directly, or there
is a short delay between the stripping of B(i) and the imposition of C(o).

The five surviving copies containing B(o) from Phase IIIb all contain Norton's printing of quire D, and most show displacements of skeleton types on C(i) and D(o). I am not able to explain the coincidence of these particular variants in this cohort.

C outer forme

Preparation for printing: all pages reimposed with new skeletons. C2v shares
some headline material with B3v.

Pauses for correction: none.

After printing: all pages kept standing with their skeletons.

C inner forme

Preparation for printing: C1v: italic and brackets in ll. 1–10 and last four lines
reset, dropping the third line from the bottom; the italic is face i-i; first letter of
"Invention" lowercased, l.6 up. C2r reset except for most of ll. 5–9. C3V reset,
with italic sections in alternative face i-2. Some respacing on C4r.

Pauses for correction: none. For reasons still unclear, the copies bound with
B(o) in Phase IIIb show page-number shifts on C1v and C4r (side-by-side in the
forme), and the comma at the end of C4r l.6 shifts right.

After printing: all pages kept standing, C1v and 4r with their skeletons, C2r
and 3v without.

By this point in the production, Norton eVidently has made an arrangement
with his partner Nicholas Okes to divide the job of printing quire D. (For the
possibility that Okes was an unwitting collaborator, see §8B.) Okes composes
his own setting of those eight pages using his own type. Each man then prints
roughly equal numbers of the quire for Phase III. Some of each man's sheets are

gathered with Phase II of the preliminary quire, in which "of Grayes Inne" still
appears on the title page after Shirley's name; one of these copies, with Okes'
quire D, also contains Phase IIb of quire A. From these groupings, it appears that
each man's production was equally available for assembling the first copies from
Phase III, so their work on this quire was effectively simultaneous.

D outer forme (Norton)

Preparation for printing: D1r reset. D2v ll. 6–7 respaced to one line, and bot-
tom half of page reset beginning with l.17. Italic portion of D3r reset in alterna-
tive italic i-2. On D4v, l.4 respaced. New skeletons all around; D4v shares some
headline material with B4v.

Pauses for correction: one, correcting spelling and punctuation on D1r and ap
parently causing a signature-mark shift.

After printing: all pages kept standing with their skeletons.

This printing of quire D is probably a bit short, as one surviving copy from
Phase III was apparently completed with quire D from Norton's final state
(D1r:2a3).

D inner forme (Norton)

Preparation for printing: D1v reset. D2r ll. 1–8 reset in the default italic (i-1).
D3v roman section reset, and the word "Proclayming", l. 4 up, gets a lowercase
first letter (incorrectly, as it follows a period).

D2r, 3v, and 4r are reimposed with headlines set in pica size; D4r lacks
catchword, and word-spacing of l. 3 and last three lines shifts a bit. Because the
headline on D1v is a different size from those of the other three pages, it was
evidently set in a different operation, probably during the resetting of the text.
The use of pica for the other headlines could indicate a shortage of english-sized
numerals and parentheses, but more likely the workman simply dipped into the
wrong case. The direction lines of all four pages are set in the correct size.

Pauses for correction: none.

After printing: all pages kept standing with their skeletons.

Quire D (Okes)

Preparation for printing: all pages newly set.

Pauses for correction: none.

After printing: all pages kept standing with their skeletons.

Some complete copies of the book are probably made up at this point, using
up the leftover preliminaries from Phase II and containing Okes' and Norton's quire D indiscriminately.

"a"

Preparation for printing: "of Grayes Inne ," is removed from the title; motto re-
spaced to two lines to compensate; spacing of imprint slightly disturbed. a2v loses
the rule below the signature, which may therefore have been treated as part of
the skeleton. The headline, however, is unchanged. On a2r, leading is adjusted

around l.5 of the heading, moving the line closer to the succeeding one. In the
process, the loss of a spacer after "and" in l.3 causes a gradual rightward drift
of the first two words in the line.

The forme is printed off by work-and-turn.

Pauses for correction: none.

After printing: on a2v, "The humblest ... Shirley." stripped; all pages otherwise
kept standing unaltered.

On the evidence of surviving copies, quire "a" has a print run sufficient to
supply all copies from Phase III. The remainder of the third printing of ToP can
now be collated and shipped.

D. PHASE IV

Products of this last phase of work are the most uniform. All copies have "The
third impression" on the title page. Although most of the formes underwent ex-
tensive alteration, or even resetting, before printing began on each, there were no
stop-press alterations thereafter. The only complication is the existence, carried
over from Phase III, of the two settings of quire D printed by Norton and Okes.
These occur in eight and six surviving copies respectively.

In light of the evidence that there were four main work phases, and B(o) and
all of quire A seem to have had five printings each, we need to take Norton's final
claim of three "impressions" with a grain of salt. We cannot know whether Nor-
ton actually believed this, whether the various starts and stops in the work cycle
caused him to lose count, or even what he understood as an "impression". In the
period 1600–40, the word was used on title pages essentially synonymously with
"edition", but only one-fifth as often. One finds it most commonly on title pages
of devotional works that went into many settings, like Arthur Dent's Plaine Mans
Path-Way to Heaven; but the title pages of Shakespeare's second and third folios
also identify them as later "impressions". Norton himself used the term in 1624
(STC 25090a, in a collaboration with Augustine Mathewes), 1633 (23503), 1634
(twice, 3129a and 22459b [ToP]), and 1637 (20274); all of these except Triumph
of Peace were completely new settings. It is likely that, in the case of ToP, Norton
was basing his count of impressions on the changes of title-page wordings which
would have been noticeable to contemporary readers: states 1a1 and 2, which
contained "of Grayes Inne"; 1b, lacking those words; and ici and 2, the avowed
third impression.

Phase IV seems to have followed Phase III rather closely, as some types from
distributed portions of III turn up in IV. These types evidently remained near
the top of their compartments in the type case without moving off to other jobs.
The recurring types within this phase (see §9D) show that quire B inner forme
(which had been distributed after Phase III) was reset after the last copies of quire
A outer forme came off the press; and D(o) contributed one type to B(o). For the
account below I have followed the order A-C-D-B-a.

In this phase, a new paragraph added to A2r causes a change of page breaks
through the rest of the quire.

A outer forme

Preparation for printing: all pages reimposed; A4V probably retains its headline
but gets a new rule and regains its missing catchword. A1r: heading slightly re-
spaced and the leading removed between it and the text. A2V receives the last
three lines of text from the preceding page, reset; its last two lines of type are
moved to the top of A3r. A3r: one textual correction; last three lines removed
and transferred, probably immediately, to A3v in the inner forme. A4v: last four
lines from previous page reset and added at top; l.2 up, "variety" changed to
"uariety".

Pauses for correction: none.

The text of ToP, especially in the later phases, shows struggles with the letters
u and v. These resulted from orthographic changes taking place in the printing
trade around 1630–35. Prior to this, the two graphics were variants of the same
letter, which was used to denote the sounds u and v indiscriminately. The letter
took the v-form at the beginning of words and the u-form elsewhere. Hence the
words "beauty" and "variety" used the graphics as we do today, while "every"
and "unless" were rendered as "euery" and "vnlesse". Within the space of a few
years, the English printing trade converted nearly universally to the modern us-
age, but for some compositors this must have been a period of confusion. Thus on
A4v of ToP in Phase IV we find, in the same line, the aberrant spellings "beavty"
and "uariety".

A inner forme

Preparation for printing: on A2r, the addition of a new paragraph of three lines
after the first one leads to the following dislocations:

A2r: last three lines of text removed and distributed. (The text has already
been incorporated, reset, on A2v in the outer forme.) For no apparent reason,
the bottom half of this page is again reset from the line beginning "These mov-
ing forward". (This same portion of text received an arbitrary resetting between
Phases I and II.)

A3v: receives the last three lines of type from A3r in the outer forme. These
force the last four lines of type to the top of A4r.

A4r: receives the four lines of type from A3v. These force out its last four lines,
which have already been incorporated, reset, in A4v in the outer forme.

There is one spelling change on A3v. New skeletons all around.

Pauses for correction: none.

After printing: headlines removed and transferred to B(i); the rest distributed.

C outer forme

Rewriting of the text on C1v and 2r leads to some altered page breaks.

Preparation for printing: all pages restored with their skeletons from Phase III.
C1r: four lines of roman at top reset in pica, perhaps to relieve crowding. C2v:
first two lines removed and probably distributed; blank line added above direc-
tion line to compensate. C4v: spelling correction.

Pauses for correction: none.

C inner forme

Preparation for printing: text revised and shortened on C1v; to compensate, the
setting of the first two lines on the next page is transferred in type to the bottom of
this one. More revision to C2r, and two lines of text brought over from the verso
in the outer forme (probably reset). C3v: improvements to spelling and leading.
C4r: spelling correction. C1v and 4r retain their previous skeletons, though 4r
shows a slight page-number shift; new skeletons for C2r and 3v.

Pauses for correction: none.

As in Phase III, both Norton and Okes contribute sheets of quire D to this
final printing. The sequence below assumes that Okes machines his Phase-IV
sheets in a separate printing, though his two impressions are indistinguishable.

D (Norton)

Preparation for printing: all pages replaced with their skeletons, but the signa-
ture mark on D1r shifts about 5 mm to the left. This is the only feature distin-
guishing this phase of Norton's quire D.

Pauses for correction: none.

D (Okes)

Preparation for printing: all pages replaced with their skeletons.

Pauses for correction: none.

Apparently, some D sheets from this impression are used to complete copies
from Phase III.

B outer forme

Preparation for printing: B2v replaced unaltered from the previous phase; the
others get new skeletons. B1r gets a heading of 26 pairs of acorns with an end-
piece on the right only. The bottom of B3r receives the first three lines of text
from B3v, reset, apparently to relieve an imbalance; the tavern direction revised.
B4v: last two lines of italic reset in pica.

Pauses for correction: none. At various points during printing, the page numbers
shift on B1r, B2v, and B4v.

B inner forme

Preparation for printing: all pages reset, with pica roman used for the directions
for gentleman and beggars on B3v. Headlines transferred from A(i) with numbers
unchanged.

Pauses for correction: none.

"a"

Preparation for printing: title page: "The third impression." added above motto;
probably as a consequence, leading adjusted elsewhere, and ll. 1, 4, and 7 re-
spaced; the line "February the third, 1633." reset, with a period in place of the
comma; "Gent." in authorship statement replaced with "of Grayes-Inne,", with
accidental omission of the "Gent.". On a2v: "The humblest ... Shirley." newly
set. The forme is printed by work-and-turn.

Pauses for correction: one, restoring "Gent." to the end of the authorship
statement.

"A Speech to the King and Queenes Maiesties"

Printed an undetermined period after the main book, with no variant states
found in five examples. It is not found in surviving copies from before Phase IV;
I cannot say whether all copies from the last impression originally contained it.

Phase IV of The Triumph of Peace brings us to the end of a process that left
behind artifacts of a bibliographical complexity unsurpassed, in my experience,
for a book of its modest length. Once the last copies were sold, the masque stayed
out of print for the next two centuries. 40

§16. When, How Many, and How Fast?

I have argued that the complexity of ToP shows the printer's response to a
high but short-lived demand for a text, so it behooves us to ask: were they first put
on sale, how many copies were printed, and how long did it take to print them?
Although ToP is the most thoroughly documented stage production before 1640,
the archival record fails to give straightforward answers to these questions. The
relevant documents are as follows: 41

• 17 October 1633: a letter from Thomas Coke to his father Sir John mentions
  that the King had requested the masque "about a fortnight agoe". "Who is the
  poet or who makes the maske dance I doe not yet understand." 42 We do not know
  whether Thomas was simply not privy to the decision process, or whether the
  organizers had not yet picked their poet. But if his chronology can be trusted, the
  members of the Inns had only four months to find a writer, commission music,
  construct sets and costumes, learn the parts, and pull it all together. No wonder
  that Coke, in the same letter, reports "no law studied in the Ins of Court now all
  turnd dancing scools".
• Bulstrode Whitelocke, in his Memorials of the English Affairs, gives an even
  shorter timeline, dating the original idea ofthe masque to "about Allholantide"—
  so, around the beginning of November, with three months to prepare. 43
• 17 January 1634: Shirley is admitted to membership in Gray's Inn. This honor
  is recorded on the earliest states of the title page of ToP, so it would seem to
  provide a terminus post quem for the first printing. However, I have suggested in
  my discussion of the title-page variants in §4A that Shirley's claim to the honor
  might have anticipated its actual granting, which created an embarrassment that
  led to the subsequent deletion of "of Grayes Inne" from the title pages of Phase
  II of the preliminaries. So, this timepoint could be less significant than it seems.
  It is possible, however, that his actual admittance was contingent on delivery of
  the completed manuscript.
• 24 January: the publisher William Cooke brings a copy of ToP to Stationers
  Hall for license and registry. The entry, signed by Cooke, reads "Entred for his
  copy under the hands of Master Attorney Sir John ffinch and Master weaver
  warden The maske of the four Inns of Court with the Sceane as it is to be presented
  before his Maiesty at Whitehall the third of ffebruary next." (Finch was one of the
  eight members of the main committee overseeing the masque, and one of two
  from Gray's Inn. Edmund Weaver was the Stationers' Company official who
  recorded the entry.) The wording of the entry shows that the clerk, as we would
  expect, was looking at a manuscript rather than the printed book. It is entirely
  possible that Norton was already hard at work on ToP, but a scrupulous printer
  (which Norton was not) would have waited until entry was completed.
• 3 February (first performance): a warrant is issued for payment of £15 to
  "mr Sherley poett for the Masque in full of his gratuity". 44 Fifteen pounds re-
  presents the Middle Temple's quarter-share of the total amount: expenses for the
  masque were borne equally by the four Inns of Court. So, on the day of first
  performance, Shirley is authorized to be paid £60 in full for the text. However,
  a further warrant of 21 November 1634 requests £5 for Shirley on behalf of the
  Middle Temple, its share of a total additional payment of £20. 45 No reason is
  given for this warrant; either the the Inns decided on a bonus to the handsome
  original fee, or it involves compensation for writing the epilogue "A Speech to the
  King and Queenes Maiesties", which was added to copies of the last printing of
  ToP. In either case, we are still lacking pieces of the puzzle, for Shirley's eighteen-
  line addendum was certainly not worth one-third of the whole masque text. A
  general accounting of all expenses for the masque complicates matters further by
  recording a total of £100 paid to Shirley. 46 The accounts often present this sort of
  confusion. Even the second warrant's November date offers no evidence bearing
  on the printing of the epilogue, for the warrant belongs to a group of late settle-
  ments for services that were mostly rendered in February 1634 or earlier.
• John Finet, in his undated memoirs, 47 describes both performances of ToP
  and mentions "the description of the Maske since being printed". This seems to
  mean that the book was not finished even by the time of the second performance
  on 13 February. But drawing a definite conclusion from Finet's wording runs the
  risk of being over-literal.
• Between 3 and 13 February: Justinian Pagitt writes to his cousin Tremyll, 48 "I
  have sent you a booke of our Masque which was presented on munday last." So,
  the second performance had apparently not yet taken place. At the time of writ-
  ing, Pagitt was a member of the Middle Temple in his early twenties, and he rode
  in the procession introducing the first performance. We cannot say for certain
  that the "booke" he sent was a printed copy, but it would seem extravagant to
  prepare a 7500–word manuscript to send away to a cousin. All of the surviving
  eyewitness accounts focus on the visual and social aspects of the performance;
  none mentions the qualities of Shirley's text.
• 19 February: William Gawdy writes to his father Framlingham, sending him
  a copy of "the booke". 49 William, aged about 22, had used family connections
  to get a ticket to the second performance. Again it is not certain that he was
  sending a printed copy, but the combination with Pagitt's letter adds to the evi-
  dence that the book had come out at least by mid-February. If manuscripts were
  circulating, Pagitt and Gawdy would more likely refer to them as "copies" than
  as "books". 50 Furthermore, neither man is known to have had direct access to
  Shirley, and it was only from Shirley or the very few organizers of the event that
  complete transcriptions of the masque could legitimately have been had. (Even
  the actors would have received copies only of their own parts, with cues.) The
  circumstances of the masque made it ideal for publication, and it was in nobody's
  interest to cut into potential profit by encouraging scribal circulation.
• 21 November 1634: John Herne, disburser of funds for the masque, issues
  a warrant for 25s. payable to "mr Wakelye—printer". 51 This figure represents
  the Middle Temple's quarter-share of a £5 payment to the book wholesaler (not
  printer) Thomas Walkley, who was evidently acting as middleman between the
  Inns and the publisher William Cooke. As with Shirley's fees, we do not know
  precisely what this payment was meant to cover, and the date of the warrant
  probably has little relation to the dates of services rendered.

From a marketing standpoint, it would have been ideal to have the printed
book available the day of the first performance, if not a few days before. Milit-
ating against that were the short timeline from the conception of the masque to
its performance and the inevitability of many last-minute changes, particularly
in the procession. Whitelocke's memoirs tell of adjustments made to head off
squabbles over matters like precedence and color schemes. 52 Such frictions must
have been constant, especially in the staging of a masque that dealt so overtly
with class conflict. 53 In §9A, I showed that the first impression of quire A, which

contains the description of the procession and the decoration of the Banqueting
Hall, was printed after the text of the masque itself, comprising quires B-D. This
sequence probably reflects the order in which the sections arrived at Norton's
shop. A professional like Shirley would first compose the more laborious liter-
ary portion of his commission, then turn to the reportage. Small changes to the
latter could be incorporated up to, and even during, the printing of the quire
containing it. 54

Printed masques conventionally preface the main text with a description of
the set design in the past tense, and the title page almost always gives an exact
date of performance. These features create a framing device that presents the
performance—usually the only one—as a fait accompli. This convention might
have no more value in dating the publications than a condemned criminal's "last
words", which might be already printed and available at the public execution. 55
But there is no more reason to suppose that masques were printed in advance of
their performance than there is for any other type of dramatic text. Specifically,
we have no evidence that the first printed ToP was available before or at the first
performance, but we have good reason to believe that it came out shortly after-
wards, even before the second performance.

As to how many were printed, there is no archival record. In fact, we have
almost no figures for the edition size of any early Stuart dramatic text. The only
report for a play that I am aware of is 1500 copies for the second edition of Beau-
mont's Philaster (1622, STC 1682), but Peter Blayney cited this as an optimistic
response to an unusually successful first edition, and therefore atypical. 56 David
Bergeron found that the print runs for the Lord Mayors' shows in the 1630s gen-
erally ran 300–500, 57 but neither the genre nor the intended readership of these
brief accounts compare closely with those of a play or masque. Edition sizes of
non-dramatic works that emerge in archival sources vary widely, and since the

motive for recording the numbers is often a disciplinary action for irregular print-
ing, we need to exercise caution in treating the figures as normative.

In the absence of relevant archival records we can draw our evidence only
from the number of surviving exemplars. I have been able to locate 55 reason-
ably complete copies of ToP (which I define as lacking no more than one quire).
This represents an average of 13.75 copies per printing. I performed a survey of
39 editions of other individually printed masques catalogued in ESTC. These show
an average survival rate of 10.64 copies per edition (σ = 5.84; in contrast to ToP,
none of them shows evidence of reprinting from standing type). One is tempted,
then, to suppose that ToP's average print run was somewhat higher than average
for a masque. It was also unique in running to four impressions; only Chapman's
Memorable Maske, Campion's Discription of a Maske, Heywood's Loves Maistresse, and
Daniel's True Discription of a Royal Masque went into a second edition before 1640.
We could extrapolate from our reasonably hefty sample to conjecture that the four
printings of ToP yielded a total of about 1.3 ˙ 4n copies, where n is the average
print run for a masque and 1.3 is the factor by which the average print run of ToP appears to exceed that of an average masque. However, these figures rest on shaky
foundations. For several reasons, ESTC often fails to list all known copies of an
edition. This incomplete reporting could magnify the apparent disparity between
ToP's numbers and those of other masques which have not attracted as much scru-
tiny. Even if the factor 1.3 is valid, the higher-than-average survival of ToP could be
partly the result of higher retention rates due to the masque's contemporary fame.
And because we have no good evidence for the normal print run of a masque, as-
signing a value to n in the equation above is a very subjective process.

The relative sizes of ToP's four printings are also hard to fix. If we use quire C
as the most stable marker of which phase a particular copy of ToP comes from,
the numbers of survivors are 19, 10, 17, and 9 respectively. By the time I was
reaching the 50-copy mark, I assumed that the relative distributions had stabi-
lized. But I was surprised when, late in my investigations, several more copies of
ToP emerged and most of them clustered in Phase III. This skewing of the data
scuttled my preliminary conclusion that a large first printing was followed by
three approximately equal but smaller ones.

Is the spread in the survival rates of the four printings of ToP statistically significant? Suppose we assume that each production phase yielded the same
number of copies. A random sample from the entire printing history should yield
approximately equal numbers from each printing—again, 13.75 in our case. Out of a sample of 55, the chances of drawing 19, 10, 17, or 9 copies of any one print-
ing would only be about 3, 6.6, 7.1, and 4% respectively. 58 So, the dispersion ap-

pears significant, and we have reason to believe that the first and third printings
were larger than the second and fourth, perhaps by a factor of 11/2 to 2.

The temptation to estimate the print runs from these admittedly wobbly
figures is a powerful one, as some of my respected predecessors have demon-
strated with recourse to even skimpier data. I hope that the disclaimers above
will safeguard any speculations I make from hardening into claims if they are
quoted later. I will guess that the anticipation generated by the run-up to ToP's performance led Walkley or Cooke to set the initial print order optimistically
at 1200 copies. Perhaps they would have gone higher had it not been for the
Stationers' Company restriction. The three remaining printings might have con-
tained 750, 1000, and 750 copies respectively, yielding a total of 3700 copies, or 16,650 sheets. 59

The printed masque, to my thinking, was a nine-days' wonder that the text
alone could not sustain once the impact of the public spectacle faded. How
long did the interest last, and how long did it take Norton to produce the four
printings? We can start by estimating the time required for the first printing.
The initial setting contained about 24,800 ens of visible text type, not including
skeletons and display text. Donald McKenzie, in his The Cambridge University Press
1696–1712 (Cambridge: University Press, 1966), cites an exceptional typesetter
who averaged about 10,600 ens per day over a five-week period in 1702, but a
more normal rate was 6,300 ens for that compositor, and usually lower for other
individuals. 60 In January 1634, John Norton had two apprentices, the senior one
having close to seven years' experience. 61 We have no evidence that he also
employed a journeyman, but recall from �8B that a document dated one year
later charged Norton and Okes with hiring illegal workmen unaffiliated with the
Stationers' Company. This allegation makes it impossible to estimate the number and skills of workers thatNorton
could potentially put onto a job. But let us as-
sume that either Norton or his senior apprentice, working flat-out, were capable
of setting 10,000 ens per day. Based on this rate, the undistracted composing of
ToP should have occupied two and a half man-days. So, typesetting should not
have been a bottleneck in getting ToP onto the market.

Presswork went slower than composing, for a comparable amount of text. In
the case of the Cambridge Press, McKenzie found that "an output of well over
1000 perfected sheets a day from a full-press [employing two men] was quite
regularly achieved and sustained for lengthy periods." 62 ToP comprised four and

a half sheets. If we apply McKenzie's finding to my hypothetical figure of 1200
copies for the first printing of ToP, a full-press dedicated to the one job should
have taken about five and a half days—just short of a regular work-week—to
machine the nine formes of ToP, not counting the time to change the formes. If
Norton delegated two men to the press and one man to composing (and distrib-
uting, for the first italic raid), he should have been able to complete Phase I with a hard six-day week's work plus an extra day or so. Of course, preparing for
the job—making design decisions, casting off copy, readying the paper, cutting
frisket sheets, and the like—would have taken some additional time. But one can
imagine Norton receiving copy as late as the day of registry—24 January—and,
if necessary, having finished books from Phase I available very shortly after the
first performance nine days later.

The later printings, since they re-used standing type to varying degrees, de-
manded much less composition time. Norton's shop set 24,800 ens for Phase I;
Phase II involved the resetting of about 9680 ens, Phase III (including its sub-
phases) 3960, and Phase IV 4110. Nicholas Okes had to set 6220 ens for his
Phase III quire D; then his Phase IV was either an unaltered reimpression or was
printed as part of Phase III. I have figured his total press time as the same either
way. Using my thoroughly unreliable edition sizes, and the rates of work given
above—setting 10,000 ens a day and printing 1000 sheets a day—and rounding up to the half-day, we arrive at the following approximate figures:

TABLE 43. Estimate of minimum production time for all copies

So, the entire production cycle of ToP required a little less than one man-
week of composing, and somewhat less than three weeks at full-press. Of course
these calculations cannot give an actual total. We assume, for instance, that com-
posing and presswork often went on simultaneously—a sensible apportionment
of labor in Norton's small shop would be two men printing and one composing.
His partner Okes could have been working concurrently on his share, so his

hours (though only less than two days) should not figure in the total time. And
my estimates of the print runs could be too high. All of these factors would com-
press the computed time required to print all the copies. It is worth noting that
the imprint date remains 1633 in all four impressions. For this publication Nor-
ton was using the legal calendar, in which the year date changed on Lady Day
(25 March). If the fourth printing of the ToP preliminaries began after this day,
he would have had some motivation for advancing the year date to emphasize
the book's currency. But this would have been a small improvement, and we have
seen that Norton's attention to detail was fitful at best. Still, we can say with some
confidence that Norton could not have turned out all four impressions of ToP in
less than three or four weeks.

This is a best-case estimate, and I have not factored in some contingencies
that must have dilated the history. For instance, besides the preliminary jobs al-
ready mentioned (design, frisket cutting, etc.), I have omitted the time required
to compose the title page, other display lines, and skeletons. Every setting or
resetting implies a distribution; because that job can be put off indefinitely, or
run concurrently with printing, I have ignored it in my calculations. Most impor-
tantly, the later italic raids show that Norton had one or more other jobs in train
along with ToP, as was normal in printing shops. Since I have not been able to
identify these jobs, I cannot gauge their impact on the ToP history.

The paper evidence, on the other hand, serves to constrain the timeline. I
have assembled watermark information, in more or less detail, for three-fifths of
the copies of ToP as well as (mostly single copies of) fourteen of the twenty-seven
publications Norton put his name to in 1633 and 1634. Among this group, ToP is almost unique in having one type of mark that I have not found named or
described anywhere. In one version, it is a fat spindle with pointed or slightly
rounded terminations; in the other, one end of the spindle is either a full semi-
circle or softly flattened into a broad but still curved base, the other end being
pointed. (Figure 4 gives slightly idealized renderings.) All varieties have a sinuous
line inside, running roughly down the long axis, usually connected to the spindle
at the sharper end and sometimes terminating in a loop inside. Most of the forms
have an additional short line coming out of each end, often with its own termi-
nating loop. I have found eight spindle variations in ToP, and in Phases I-III of
quires A-D it is the dominant type of mark. Second most common is a hand, usu-
ally containing the initials "PD" (four varieties). Another has initials "GM", and
in a sixth variety the initials are lacking or illegible. The hands all have a flower
or a grape cluster at the fingertips. The third type of mark includes six varieties
of pot, two containing the initials RP or RF. One of these may have a handle and
might therefore be called a jug. Finally, there is one example of a broad mark
(over four centimeters) of uncertain form, but apparently bracketed by two short
convex columns. I have called this "posts". In all the copies examined, only one
sheet (quire C of B.L. Ashley 1696) seemed to lack a mark. The frequencies of
the marks in quires A-D are as follows: 63

[Description: FIGURE 4. Spindle watermarks in Triumph of Peace. ]

TABLE 44. Distribution of watermarks by phases

         

Phase	Spindle	Hand	Pot	Posts	No mark
I	29 (71%)	9 (22%)	2 (5%)	0	1 (2%)
II	22 (92%)	1 (4%)	0	1 (4%)	0
III	18 (90%)	2 (10%)	0	0	0
IV	9 (36%)	12 (48%)	4 (16%)	0	0

The spindles, I believe, are significant for two reasons. First, their frequency
drops at Phase IV, with hands coming into the majority and pots making a surge.
Although the sample size is small, these shifts could indicate a transition into a
different mixture of paper stocks and hence a rather longer pause after Phase III
than after I or II. Secondly, of the other books printed by Norton in 1633–34
that I was able to examine, only one contains a spindle: that is the Huntington
copy of the seventh quarto of Richard II , dated 1634. It has two specific versions
found in ToP, as well as two ToP hands, along with two marks which I found in
Richard II only. The received wisdom about printers' paper stocks is that they
bought only what they needed for specific jobs, and this makes economic sense.
Indeed, the stock found in ToP Phases I-III is consistent with a single batch of
paper assembled by a vendor from the products of two or more makers. The
odd post or pot could either be odds slipped into the ream by the vendor or
leftovers from a previous lot remaining in Norton's shop. In practice, the end of
a job never precisely coincides with the end of the paper supply, and Norton was
a busy man, printing about 250 edition-sheets that we know of in 1633 (though
that number drops to around 150 the next year). A more sensible response to a
continuous need for the middling-quality pot-sized paper found in virtually all
of Norton's books of the period would be to periodically estimate the projected
demand for incoming work and top off an in-shop supply that all the jobs could
draw upon. Thus, we would expect frequently to find the same marks (by which

I mean paper from the same specific moulds) in publications printed around the
same time. This is the sort of evidence we find in ToP and Richard II , 64 and I pre-
dict that if more copies of the latter are examined, they will contain a proportion
of spindles to hands comparable to that of Phase IV of ToP. A finding of similar
paper stocks would indicate that the two books were printed in quick succession,
ToP first, or even with some overlap.

The distinctive mixture of marks in the four phases of ToP tightens their
temporal bond. It reinforces the evidence that Norton produced the first three
impressions in as short a time as he could. It also shows that there is little chance
that he paused for long between Phases III and IV. Standing paper, like stand-
ing type, is vulnerable to appropriation by more urgent projects. The fact that
the paper in Phase IV shows significant overlap with the distinctive paper stock
used in the previous printings is consistent with my finding of broken types from
Phase III recurring in Phase IV (�9D). Both support a conclusion that Norton
did not wait long before embarking on the last impression. I would guess that the
four print runs were all over within two months.

CONCLUSIONS AND APPLICATIONS

I have presented evidence showing that John Norton printed The Triumph
of Peace in four impressions, with hiatuses within the second and third of these.
Varying amounts of text—but always much less than half of the book—were
distributed and reset before each of the later impressions, at least partly to sup-
ply other jobs running concurrently. The preliminaries, including the title page,
also went through four impressions, but in different quantities from the text
quires, so that the first two states of the title may each occur with two possible
impressions of the text. Other, smaller disparities in print runs among the text
quires produced occasional copies containing quires from adjacent impressions,
further complicating the problems of identification. However, in general the four
printings of the text come together in a few consistent combinations with little
intermixing. The state of quire C is the clearest indicator of which impression a
copy belongs to, but fully characterizing exemplars of the first three printings still
requires the specification of many variables. About half of the copies of the third
and fourth impressions contain type-material from Norton's partner Nicholas
Okes and are presumed to have been printed by him.

The difficulties that earlier bibliographers had with ToP arose largely from
their point of view. Like astronomers of the Ptolemaic paradigm, they put the
most familiar element—the title page, in this case—at the center and saw the
rest of the system behaving with a bewildering complexity. Moving the main part
of the book—the text—to the center and relegating the title page to its own odd
but explainable orbit eliminates much of the need to invoke mysterious causes.
Even so, the book offers seemingly unlimited surprises—even unto the fifty-fifth
copy—and mysteries whose unravelling I will now leave, as did Greg, to those

with more "leisure and opportunity". Those happy investigators might be able
to explain the unique variant in St. Catharine's (Cambridge) Z59; the garbling
of states in some phases of quires B and C; the weird quire interactions in the
Phase IIIb copies; the rationale of enlisting a second printer for help on quire D;
and the identity and dynamics of the competing jobs in Norton's shop. Mechani-
cal collation (a luxury that I did not allow myself) might uncover tiny differences
in the pairs of "identical" reprints that I have hypothesized above.

The complexities of ToP may tempt us to put it in a class by itself, 65 but it
contains lessons that might point the way to a fresh approach to certain diffi-
cult cases. For example, both Greg and STC treat Philip Massinger's The Picture
(1630) as one edition with a confusing array of variants. 66 A brief comparison of
copies, with attention to the skeletons, reveals a similar situation to that of ToP:
there were at least two impressions, though in this case the copies contain sheets
drawn indiscriminately from both. Perhaps not coincidentally, the printer was
again John Norton, and one immediately recognizes the characteristic slovenli-
ness that he inflicted on his lower-end work. Looking at The Picture in this light
may ease the work of the next editor, but it also at least doubles our estimate of
the play's sales during the author's lifetime. I cannot believe that The Triumph
of Peace and The Picture form the tip of a particularly large iceberg, but they
do underline the need to pay close attention to skeletons, especially when the
headlines contain nothing more than page numbers. We are still exploring the
range of ingenious—if not devious—shifts that early modern printers could use
to make their lives easier.

ACKNOWLEDGMENTS

This study had its origin in a cataloguing impasse at the Huntington Library,
to which I had been posted by the American office of the recently renamed En-
glish Short-Title Catalogue under the inspired directorship of Henry Snyder. It is
to Dr. Snyder and the Huntington that I owe the initial opportunity to be baffled
by ToP. Randall McLeod spent entirely too much of his busy retirement in re-
peated reading and marking up of the manuscript, and Joseph Gwara also made invaluable suggestions. The project could not have been completed without the
aid of staff at the libraries holding copies of the book. In particular I would like to
thank Matthew Baalham, Erin Blake, Tad Boehmer, Emily Dourish, Lynne Far-
rington, Stephen Ferguson, Jill Gage, Paul Gehl, Myron Groover, Colin Higgins,
Kathryn James, Kathleen Lesko, Ann Martin, Jenna Moore, Jason Moschella,
Richard Oram, Tim Pye, Joanna Parker, Catherine Uecker, Abbie Weinberg,
and Georgianna Ziegler. Barbara Ravelhofer and Eugene Giddens supplied valu-

able images of pages and watermarks; James Knowles helped with interpreting
archival records; and Roger Gaskell and Eva Griffith sent further information
about copies I could not easily get to. Finally, I would like to thank David Vander
Meulen and Elizabeth Lynch for their tenacious work on the manuscript through
multiple iterations.

APPENDIX 1

KEY TO VARIANTS

The leftmost numbers indicate major break-points in the production cycle. Suc-
cessive indentions show increasingly minor variants. The variants are identified
by combining the sigla at the successive indentions. For example: on page a2v,
variant 1b1 has a comma after "Honourers" in the dedication's closing; the clos-
est distance between "S" in "IAMES" and "S" in "SHIRLEY" is 3 mm; and
there is a rule at the bottom. Asterisks within the descriptions mark the resetting
of page numbers.

a1r (title page; p. [i])

• "of Grayes Inne"
  "T" in line 1 nearer to "I" below
  • "O" in "OF" intact, line 3
  • (Variant): "O" battered
  • "T" in line 1 nearer to "V" below
• No mention of Grayes Inne
"The third Impression"
• "of Grayes-Inne,", no "Gent."
• "of Grayes-Inne, Gent."

a1v (p. [ii]) blank

a2r (p. [iii])

• 28 pairs of acorns
(Reimposed): 27 pairs of acorns. Direction line reset
• Heading line 5 "THE" equally spaced between surrounding lines
• (Variant): "THE" closer to following line

a2v (p. [iv])

No comma after "Honourers" in closing
• Line 1 has "ond"
• Line 1 has "and"
• (Probably reimposed): comma after "Honourers" in closing; closest dis-
  tance between "S" in "IAMES" and "S" in "SHIRLEY" 3 mm
  [Probably later than 1 a because of broken "a" in "that", line 8.]
  • Rule at bottom
  • No rule at bottom
• (Probably reimposed): comma after "Honourers" in closing; closest dis-
  tance between "S" in "IAMES" and "S" in "SHIRLEY" 6 mm

A1r (p. 11)

• Caption title has "Jnnes (i.e. swash "I")
• (Reimposed*, at least first five lines reset): caption title has "Innes"
  • Line 5 of text has "Hornepipe"; page "1" over acorn pairs 14–15; page
    number parentheses 9 mm apart at center, right parenthesis lower and
    damaged
  • (Reimposed*): line 5 of text has "Hornepipe"; page "1" over acorn pair
    14; page number parentheses 7 mm apart at center, on same level, and
    undamaged
  • (Reimposed*): line 5 of text has "Hornepipe"; page number, or the space
    where it would be, over acorn pair 13
    • Page number printed
    • (Variant): page number not printed
    • [This variant probably depends on the presswork of individual copies.]
  • (Reimposed*): line 5 of text has "Horne-pipe".
  • [A few differences in line spacing and settings.]

A1v (p. 12)

Last line is one word, "fist.", plus catchword on same line
• Line 11–12 has "colours, laughing"
• (Variant): line 11–12 has "colours laughing,"
(Reset): last line begins "a bunch ..."
• Left parenthesis of page number over and to right of "D"
• [Line 19 has "bride".]
(Reimposed*): lines 11–12 have "colous laughing"
• Line 19 has "bride"
• (Variant): line 19 has "bridle"
• (Reimposed*): lines 11–12 have "colours laughing"

A2r (p. 13)

• Last line begins "ges"
• (Partly reset): last line begins "richly"
  [Reset from line beginning "These moving forward ..." to end.]
• (Partly reset): last line begins "After"
  [First paragraph is unchanged, then a paragraph of new text is added, forcing
  the last three lines of the page onto A2v. Thereafter, unchanged until the line
  beginning "These moving forward ...".]

A2v (p. 14)

• Last line has "Silver & Watchet."
• (Reimposed*, partly reset): last line has "silver & watchet."
  [Italic lines reset from english to great primer size. Later because bottom serif
  of "T" in "Triumphals", line 22, is damaged.]
  • Catchword "The"
    • Catchword properly aligned
    • The "e" displaced downward to varying degrees
  • (Reimposed*): catchword "All"; page "4" centered or somewhat to left
    within parentheses; right parenthesis badly damaged at 4:00
  • (Reimposed*) catchword "All"; page number "4" somewhat to right within
    parentheses; right parenthesis slightly nicked at 3:00
• (Reimposed*, partly reset): last line reads "the top."
  [Three lines of text are brought down from the previous page, in a new setting;
  last two lines move to A3r.]

A3r (p. 15)

• Line 15 reads "to the linings of the Chariots."
• (Reimposed*, partly reset): line 15 or 17 reads "to the linings of the
  Chariots."
  [First two lines and original lines 27–28 and 31 reset, the italic lines in a larger
  size than text below.]
  • Last line reads "prest in his place."; page number "5" over "so" of
    "crimson"
  • (Reimposed*): last line reads "prest in his place."; page number "5" over
    "i" of "crimson"
    • Signed "A3"
      [Damage to "White", line 2, occurs during this state. bL 644.c.44
      (now bound in nTxU Wrenn) shows damage and displacement at the
      left end of lines 1–2, and slightly different positions of the skeleton
      elements.]
    • (Variant): signed "A"
      [Lines 1 and 2 repaired, but all copies have damage to "White",
      line 2.]
  • (Reimposed*): last line begins "riding two a breast, ..."
    [First two lines of type brought down from A2v, getting slightly damaged
    in the process; last three lines of text move to A3v.]

A3v (p. 16)

• Page begins with a line of ornaments that contains no colon or inverted
  comma
  • Line 13 has "Arber-worke" with broken b
  • (Variant): line 13 has "Arbor-worke" with sound b
• (Reimposed*): page begins with a line of ornaments that contains a colon
  and inverted comma; page number "6" over and between pair of orna-
  ments 4 and 5
• (Possibly reimposed): page begins with a line of ornaments that contains
  a colon and inverted comma; page number "6" over and between pair of
  ornaments 5 and 6
  [Catchword appears identical; the page number may be reset or may sim-
  ply be shifted.]
• (Reimposed*): page begins with text, "The Habit of the Masquers ..."
  [First three lines of text brought down from A3r, reset; last four lines move
  to A4r.]

A4r (p.17)

Line 1 begins "in one hand,"
• Line 21 has "Caduseus"
• (Variant): line 21 has "Caduceus"; all commas listed in variant 3 except
  "drawne up the" and "Pallaces Lodges"
• (Variant, probably reimposed): line 21 has "Caduceus"; the following
  commas are present:
  • "Oliue-branch, all", lines 21–22
  • "drawne up, the", line 24
  • Pallaces, Lodges", line 26
  • "Architecture, with", line 27
  • "eye, opens", lines 28–29
• (Reimposed*): line 1 begins "pedestall fained".
  [First four lines of type brought down from A3v; last four lines of text
  move to A4v.]

A4v (p. 18)

• Line 1 begins "cleare Sky"
(Reimposed*, first two lines reset): line 1 begins "cleare skie"
• Page number "8" centered within parentheses
• (Reimposed*): page number "8" towards left within parentheses
  ["T" is damaged, line 3. The curvature of the rule varies during this
  imposition.]
  • Raised quads between the paragraphs
  • (Variant): quads down
• Line 1 begins "Trees and grounds, ..."
  [First four lines of text brought down from A4r, reset. Right parenthesis of
  page number now damaged, but headline not definitely different.]

B1r (p. 21)

Line 3 up has "A Bird!"
• Line 2 of heading: "they salute."
• (Variant): heading reset: line 2: "him, they salute."
  • "A" (line 9 up) and "B" (line 7 up) are sound.
  • (Variant): these letters are damaged.
• (Reset): line 3 up has "A Bird!"
  • 2 acorns are inverted
  • (Reimposed*): 25 pairs of acorns, all properly oriented
    [Later than 2a because "An" corrected, last line.]
  • (Reimposed*): 28 pairs of acorns, all properly oriented
    [Later than 2b because "Hce'l" corrected, line 12.]
  • (Reimposed*): 26 pairs of acorns, all properly oriented
    [The position of the page number varies a bit.]

B1v (p. 22)

• Question mark in line 8 under "er"; line 9 begins "Have"
• (Reset): question mark in line 8 under "nt"
  • Line 10 begins "Vs"
    [The "V" shows varying degrees of displacement.]
  • (Variant): line 10 begins "s"
  • (Reset): question mark in line 8 under "r" ; line 9 begins "Haue"

B2r (p. 23)

• Line 4 ends "nature?"; line 6 has "inuentions"
• (Reset): line 4 ends "nature"
  [This setting contains several features that do not reduce to a regular order;
  see essay, �6C.]
  • Page number is prevailing thin form (same as on A2r, C3r, D4r)
    [Position between parentheses varies.]
  • (Variant): page number is round form, and centered between
    parentheses
  • (Reset): line 4 ends "nature?"; line 6 has "inventions"

B2v (p. 24)

• Last line has "Twas"
(Reset): last line has "twas"
• Greatest distance between parentheses 11 mm; line 16 has "the
  Novelties."
• (Reimposed*): line 16 has "theN ovelties"
• (Reimposed*): greatest distance between parentheses 9 mm; line 16 has
  "the Novelties."
  [Within this imposition the page number shows slight variations in damage
  and position.)
  • Line 23 has "Laughter," with comma
  • (Variant): line 23 has "Laughter" without comma

B3r (p. 25)

• Tavern direction in one line, italic, centered
• (Reimposed*, partly reset): tavern direction revised, three lines, roman
  [Last four lines of type moved to B3v, respaced as three lines. "Except" is cor-
  rected to "accept", line 22.)
  • Page number "5" over the "f" of "foole" in line 2
  • (Reimposed*): page number "5" over the space between "the foole" in
    line 2
(Reimposed*): tavern direction revised to one line, flush right
• "The Scene a Taverne."
• (Reimposed*): "The Scene a Taverne."
  [First three lines of text transferred from B3v, reset.]

B3v (p. 26)

• Page begins "Another Antimasque ..." (regular italic "A")
  • Line 10 has "off throw"
  • Line 10 has "off, throw"
    [Almost certainly later; see essay.]
• (Reset): page begins "I may supply ..."
• (Reset): page begins "Another Antimasque ..." (swash italic "A")

B4r (p. 27)

Last line has "vnder water" (narrow word-space)
• Lines 16 and 17 have "beard ... yeares found"
• (Variant): lines 16 and 17 have "beard, ... yeares, found"
  [Almost certainly later; see essay.]
• (Reset): last line has "vnder Water"
  [This setting contains features that do not reduce to a regular order—see essay, �6C.]
  • Line 3 begins "a vapour"
  • (Reimposed*): line 3 begins "A vapour"
    [Line 3 now correctly begins with a capital, and type is battered in line 7 in some copies.]
• (Reset): last line has "under Water"

B4v (p. 28)

• Line 1 has "vast"
• (Reset): line 1 has "uast"; line 16: "The fift proiector dances."
  • Page number "8" over the space between "of ayre"; catchword "Ma-"
    • Line 2 has "coold"
    • (Variant): line 2 has "cool"
    • (Variant): line 2 has "coo"
  • (Reimposed*): no catchword; last word on page: "Th"
    • Page "8" over "e" of "ayre"
    • (Variant): page "8" over the "f" of "of" and centered within
      parentheses
      [The type-matter of the page number has shifted and possibly suf-
      fered damage but seems to be the same metal.]
  • (Reimposed*): page number "8" over "of" and to the left within the pa-
    rentheses; no catchword (last word on page: "The")
    • Line 13 has "stndy"
    • (Variant): line 13 has "study"
• (Reimposed*; flush-right italic lines reset in pica size): line 1 has "uast"; line 16:
  "The fift Proiector dances·"

C1r (p. 9)

First four lines of roman in same size type as the rest
Right parenthesis of page number heavier and nicked at 3:00
• Line 1 has "Returne as from the Taverne"
• (Variant): line 1 has "Returne, as from the Taverne,"
• (Reimposed*): right parenthesis sound and the same weight as the left
  [Retains the corrected reading of 1 a2.]
• (Reimposed*): page number "9" slightly to the right within parentheses
• (Partly reset): first four lines of roman in type smaller than the rest (pica, 83 mm)
  [The skeleton is not obviously changed.]

C1v (p. 10)

• Last line: "Three Dotterell-catchers." (all italic); line 9 has "off. Then"
  [The comma in line 7 is displaced below the line.]
  • Line 5 up has "We change the Seene."
  • (Variant): line 5 up has "We change the Scene."
• (Reimposed*): last line: "Three Dotterell-catchers." (all italic); line 9 has "off.
  then" (period below the line)
  [Page number, italic and bracket in lines 8–10, and last three lines in a differ-
  ent setting. Resetting of lines 8–10 may have resulted from an accident while
  re-aligning the comma.]
• (Reimposed*): last line: "Three Dotterell-catchers."
  [Italic and brackets in lines 1–10 and last four lines are reset, dropping fourth
  line from the bottom.]
  • "1" of page number over "d"
  • (Variant): "1" of page number over "a"
    [The page number types are the same and the shift is probably due to
    faulty lock-up.]
• (Partly reset): last line: "Ph. Dotterells, be patient, and expect."
  [The exchange between Opinion and Phansie is replaced by a new stage
  direction, and the remainder leaded out. Two lines of text are brought up,
  reset, from C2r.]

C2r (p. 11)

• Line 1: "Op. What are these?"; line 10 has "assavlted"
  • Left parenthesis of page number beyond the right end of the first line
  • (Reimposed*): left parenthesis of page number over "?" of the first line
• (Reset except for parts of lines 5–9): line 1: "Op. What are these?"; line 10
  has "assaulted"
• (Reimposed*, partly reset): line 1: "After the Dotterells be caught, by"
  [Probably the latest version because the bowlers now have something to do.
  The directions for the Quixote/Sancho figures are revised and reset in type
  smaller than the surrounding text, and the altercation with the gentleman
  and his servant is deleted.]

C2v (p. 12)

• Song line 5 has "skyes"
  • Line 6 begins "The Antimasques being past,"; line 11 has "Olives of
    her head"
  • (Variant): line 6 begins "The Antimasques being past,"; line 11 has
    "Olives on her head"
    ["On her head" is more correct.]
  • (Variant): line 6 begins "The Antimasquers being gone" (no comma,
    but spacer up); line 11 as in state 2
    [Further corrected.]
• (Partly reset): song line 5 has "skies"
  [Italic song reset, some other minor differences in setting or spacing.]
  • (Reimposed*): line 1 "Iol. I am as light now."; left parenthesis of page
    number over "now"
  • (Reimposed*): line 1 "Iol. I am as light now."; left parenthesis of page
    number over the period
  • (Variant): line 1 "Op. What new change"
    [Two lines of text have moved to C2r, but the skeleton is unchanged.]

C3r (p. 13)

• Line 5 has "calls;"
  Line 20 has "Starres, mantle"
  • Line 5 up reads "This Chariot attended as the former."
  • (Variant): this direction deleted
    [The attendants of the former chariot were never described, so this
    state is more correct.]
  • (Variant): line 20 corrected to "Starres, a mantle"
• (Reimposed*, partly reset): line 5 has "calls,"
  [Italic songs in a different setting.]
  • Right parenthesis of page number printed heavier than left and comes
    within about 1 mm of the bottom of the "3"
  • (Reimposed*): right parenthesis of page number only comes within about
    2 mm of the bottom of the "3"

C3v (p. 14)

• Line 13 begins "Thy voyce,"
  • Next line ends "neyther" (no period)
  • (Reimposed*): next line ends "neyther."
    [This version more correct.]
• (Reset): line 13 begins "The voyce,"
  • Line 5 up has "Diche"
  • (Reimposed*): line 5 up has "Dice"
    [This version more correct.]

C4r (p. 15)

• Line 2 begins "Dich. Swifly, [sic]"
• (Reimposed*): line 2 begins "Dich. Swiftly,"
  [The typo is corrected.]
• (Reimposed*, partly reset): line 2 begins "Diche" or "Dice"
  [Lines 1, 10, and 20 reset; some differences in word and line spacing (e.g.,
  lines 4, 6, 15–17).]
  • Line 2 begins "Diche"; left parenthesis of page number over "g" and
    the following space
  • (Variant): line 2 begins "Diche"; left parenthesis of page number
    over "4"
    [The comma at the end of line 6 is too far right. The page number
    looks to be the same types as c1, but is shifted.]
  • (Variant): line 2 begins "Dice."
    [This version is the most correct. Comma in line 6 as in 1c2. The
    page number again looks to be the same types, but is shifted so that
    the left parenthesis is over "ng"—possibly a result of correcting
    line 2.]

C4v (p. 16)

Page number "61"
• Line 6 up begins "Thus warme"; line 4 up has "Didche"
• (Variant): line 6 up begins "Thus warme"; line 4 up has "Diche"
  [The typo is corrected.]
• (Variant): line 6 up begins "Thus warm'd"
  [Line 4 up has "Diche", and further corrected.]
• (Reimposed*, mostly reset): page number "16"
  [Only the roman paragraph, lines 7–9, and the last line are retained from
  state 1. "Song 5" title and text are reset in pica size.]
  • Maximum distance between parentheses in page number (at midline)11 mm
  • [Line 3 has "Dich."; line 4 up has "Diche,"]
  (Reimposed*): maximum distance between parentheses 14 mm
  • Line 3 has "Dich."; line 4 up has "Diche,"
  • (Variant): line 3 has "Dic."; line 4 up has "Dice,"
    [This version is the most correct.]

D1r (p. 17)

Line 9 has "the Sky beyond"; line 5 has "ioyning"
• Line 9 has "the Sky beyond seene."
  [The last letter of line 1 is displaced upward in some copies.]
• (Variant): line 9 has "the Sky beyond was seene."; type mostly sound
  in first three lines, though disturbed
  [This version more correct.]
• (Variant): line 9 has "the Sky beyond was seene."; type damaged in
  first three lines
  [Later than 1a2 because of damage.]
(Reset): line 9 has "the sky beyond"
• Signature mark "D" under the "hi" of "white,"
  [Line 10 has "Masqncrs".]
• (Variant, later than 2a1): signature mark "D" under ","
  [Line 10 now has "Masquers". The page number is the same as 2a1,
  and the significance of the signature mark shift is unclear.]
• (Variant): signature mark "D" under the "w" of "white,"
  [Line 10 has "Masquers". The page number is unchanged, and the
  significance of the signature mark shift is unclear.]
• (Okes' setting): line 9 has "the Sky beyond"; line 5 has "joyning"

D1v (p. 18)

• First line has "Beneath ... Wings,"
  [First five lines set in italic, line 5 has "descended", both incorrectly.]
  • Comma after "Wings" displaced downward; right parenthesis of page
    number damaged
  • (Variant): comma after "Wings" in place; right parenthesis of page
    number sound
    [A sound parenthesis has been supplied and the comma replaced.]
• (Reimposed*, partly reset): first line has "Beneath ... with"
  [First five lines reset in roman, line 5 has "descending", both correctly.]
• (Reset): first line has "Beneath ... wings"
• (Okes' setting): first line has "beneath ... wings,"

D2r (p. 19)

• Line 7: "Live Royall paire, & when Your sands are spent"
  • Line 5 has "on ages"
  • (Variant): line 5 has "no ages"
    [This version correct.]
• (Reimposed*, partly reset): line 7: "Live royall pare, awd ..."
  [The song and the italic lines at bottom are in a different setting.]
• (Reimposed*, partly reset): line 7: "Live Royall paire,and ..."
  [Song lines 1–8 are in a different setting.]
• (Okes' setting): line 7: "Live Royall paire, & when your sands are spent"

D2v (p. 20)

• Line 6 ends "per-"
• (Reimposed*, partly reset): line 6 ends "perrill."
  [Line 6 is respaced to take up one line, and former line 18 (current line 17) to
  end is reset.]
• (Okes' setting): line 6 ends "their"

D3r (p. 21)

• Line 8 up ends "Spies,"; line 6 begins "uenge,"
• (Reimposed*, partly reset): line 8 up ends "spies,"
  [The song is in a different setting.]
• (Okes' setting): line 8 up ends "Spies,"; line 6 begins "venge,"

D3v (p. 22)

• Line 4 up reads "Proclaiming warres"
  [Line 8 has "a bout".]
• (Reimposed*, partly reset): line 4 up reads "Proclayming warres"
  [The song is in a different setting. Line 8 has "about" correctly.]
• (Reimposed*, partly reset): line 4 up reads "proclayming warres"
• [The roman section is in a different setting; the italic is the same as state 2,
  except that "proclayming" is unaccountably lowercased.]
• (Okes' setting): line 4 up reads "Proclaiming Warres"

D4r (p. 23)

• Line 3 up ends "prayer,"; no catchword
• (Reset): line 3 up ends "Prayer,"; catchword "Am-"
• (Reimposed*): line 3 up ends "Prayer,"; no catchword
  [Some variations in spacing at top and bottom.]
• (Okes' setting): line 3 up ends "prayer,"; catchword "Amphiluche" on last line
  of text

D4v (p. 24)

• "FINIS." in much larger type than text; swash italic "N"
  Right parenthesis of page number over "en"
  • Right parenthesis of page number intact
  • (Variant): right parenthesis damaged at bottom
  • (Reimposed*): right parenthesis of page number over "v"
• (Okes' setting): "FINIS." only slightly larger than text; no swash

APPENDIX 2

COPIES EXAMINED

I have examined all copies in person except as indicated. References are given
to reel numbers in the University Microfilms International (UMI) film set Early
English Books 1475–1640 (Ann Arbor, MI, [1937]-), subsequently digitized in
Early English Books Online (EEBO) (eebo.chadwyck.com). Thomas J. Wise tam-
pered with certain copies of ToP, now at the British Library and the Humanities
Research Center, University of Texas. For details beyond those given here, see
D.F. Foxon, Thomas J. Wise and the Pre-Restoration Drama: A Study in Theft and
Sophistication (London: Bibliographical Society, 1959).

bBRp: Bristol Central Library

SR 74. Reported by Roger Gaskell, who also sent digital copies. In a pam-
phlet volume, old maroon library buckram.

bC: Cambridge University Library

• Syn.7.61.1998
• Syn.7.63.3791. "Speech to the King" bound after sig. a. Sophisticated: D4
  supplied from a copy of the first printing (states D4r:1, D4v:1a2).

bCsc: St. Catharine's College Library, University of Cambridge

• Z.55. Examined in digital copy only. In a C19 to early-C20 collector's
  binding.
• Z.59. Examined in digital copy only. In a pamphlet volume, old leather.

bE: National Library of Scotland, Edinburgh

H.3.d.39(1). Examined in photocopy only. Bound with STC 22456 in later
19th-century quarter leather. Title page mended at inner margin.

bET: Eton College Library

S.170.Plays 27(04). Examined in digital copy only. Bound with six other C17
English plays in C18 gold-tooled mottled calf. Storer bookplate; given to the
library by Anthony Morris Storer, 1799. UMI reel 1585:6.

bL: British Library, London

• 644.c.44. David Garrick copy. Tampered with by Thomas Wise: A3, B3,
  C2–4 stolen, leaving stubs, as recorded by a pencil note in the book dated
  Jan. 1903. B3 and C2–3 were probably transferred to Ashley 1696. A3 was
  used to make up the Wrenn copy now at nTxU; a photocopy of that leaf
  is now bound into 644.c.44. C4 was found laid into Ashley 1697, and is
  now re-inserted in 644.c.44.
• Ashley 1695. Thomas J. Wise copy. Evidently the source of the t.p. illustra-
  tion in the Ashley Catalogue, vol. 5, recto of plate leaf preceding p. 161.
• Ashley 1696. Thomas J. Wise copy. Tampered with by Thomas Wise:
  B3 and the bifolium C2.3 were used to make up the Wrenn copy now at
  nTxU. Their lack was probably supplied from copy 644.c.44. Either Ash-
  ley 1696 or Ashley 1698 was probably the source of the t.p. illustration in
  the Ashley Catalogue, vol. 5, verso of plate leaf preceding p. 161.
• Ashley 1697. Thomas J. Wise copy.

---

206




• Ashley 1698. Thomas J. Wise copy. Disbound; paper flaw on B4. Either
  Ashley 1698 or Ashley 1696 was probably the source of the t.p. illustration
  in the Ashley Catalogue, vol. 5, verso of plate leaf preceding p. 161.
• Ashley 1699. Thomas J. Wise copy.
• Ashley 5388. Thomas J. Wise copy.
• C.12.f.15(7). Some cropping at head.

bLg: Guildhall Library, London
Closed access SR.4 Pageants.

bLv: Victoria and Albert Museum National Art Library, London

• 25.c.78. Alexander Dyce copy (Dyce Collection, 9113). Mid-C19 half calf
  with marbled boards (uniform with 25.c.79).
• 25.c.79. Alexander Dyce copy (Dyce Collection, 9114). Mid-C19 half calf
  with marbled boards (uniform with 25.c.78). "Speech to the King" bound
  at end. Washed: quires are all different shades.

bLONG: Longleat House, Warminster, Wilts.

Reported by Matthew Baalham. Half calf for Thomas Thynne, 2nd Marquis
of Bath (1862–1946).

bO: Bodleian Library, Oxford University

• Douce S 186. Francis Douce copy. UMI reel 1559:12.
• Mal. 160(3). Edmond Malone copy. UMI reel 1464:7. Misidentified as
  Huntington copy on EEBO.
• Mal. 254(5). Edmond Malone copy. Title page repaired along inner mar-gin; D4 repaired at top and bottom.

bOwo: Worcester College Library, Oxford University

• Plays 4.1022. In a pamphlet volume. -D4.
• Plays 4.1031. In a pamphlet volume.
• Plays 4.1032. In a pamphlet volume. Title page backed.

nCLU-C: University of California, William Andrews Clark Memorial Library, Los Angeles, CA

*PR3I44.T81. Bound by Riviere & Son. Some small mends to inner margin
of sig. a.

nC-S: Sutro Library, San Francisco, CA

822.S558t. Marbled boards, calf spine, ca. 1900. All page numbers more or
less cropped; state of D4v indeterminate because of this.

nCSmH: Huntington Library, San Marino, CA

• 69428. C19 sprinkled calf. Bookplates of Thomas Jefferson McKee and
  Henry William Poor (the latter's sale, Anderson Galleries, 14 Jan. 1909:942,
  $13; re-sold four months later as from "the library of Mrs. Constance E.
  Poor", Anderson Galleries, 14 May 1909:545, $9.50 to George Smith for Henry Huntington). Small paper fill, inside top of D4; the few missing let-
  ters supplied in pencil. UMI reel 1188:7.
• 69432. 19th-century calf. Bridgewater copy (en bloc purchase by Hunting-
  ton, 1917).
• 69433. Never bound, uncut, four stitching holes, now re-stitched with three
  holes, bolts cut, 20 x 15 cm. Badly printed line endings touched up in early
  MS. on B2v:20–22. Sotheby, Wilkinson & Hodge sale, 23 April 1904 (to

  ---

  207

  
  
  
  Joseph Sabin); folding morocco case by Riviere & Son added that year,
  according to dated cost code; Frederick R. Halsey copy (en bloc purchase by Huntington, 1915).
• 69434. Inlaid, some headlines cropped. MS. initials of John Philip Kemble, "Collated & perfect", 1798; Devonshire copy (pamphlet volume 81, item 7). Acquired by Huntington with the Kemble - Devonshire collection in
  1914; removed and bound separately in black morocco by Macdonald.

nCaOHM: McMaster University, Hamilton, Ontario

Mills Lower Level B 13846. Reported by Myron C. Groover. Bookplate of
Irving Kent Hall, 1901. Small paper fill in gutter of A2.

nCtY
EC: Yale University, Elizabethan Club, New Haven, CT

Eliz 218. Reported by Kathryn James. Red morocco signed by Riviere & Son.
Bridgewater duplicate stamp on title verso. Gift of Alexander S. Cochran to
the Elizabethan Club, Dec. 1911.

nDFo: Folger Shakespeare Library, Washington, DC

• STC 22459b c.2 (Bd.w. STC 1578). In a 1963 binding replacing "calf of ca.
  1780, one board off". Quire A only, following t.p. of STC 1578.
• STC 22459. Binding signed by Riviere & Son. Cropped at head and re-
  margined; state of D4v indeterminate for this reason. a1.2 not conjugate.
• STC 22459a c.1. Red morocco by F. Bedford. Griswold family crest
  stamped on cover; bookplate of Almon W. Griswold.
• STC 22459a c.2. Blue morocco by Riviere & Son; some deckles retained.
  -a2, some repairs to sig. D.
• STC 22459b c.1 (Bd.w. STC 4619). In a pamphlet volume bound in (prob-
  ably later) C17 calf. "Speech to the King" bound after sig. a.
• STC 22459b.2. C19 half green morocco, marbled boards. "Speech to the
  King" bound at end. Pencil notes: "From Mr: Kershaw's Lib. No. 1317–
  Bot: by him from Lacy (?)." (parentheses in source). Collated by Quaritch,
  25 Jan. 1924.
• STC 22459.2. Later wrappers. Title page repaired at inner margin.

nDLC: Library of Congress, Washington, DC

PR1241.L6 #188:6. Bound with five other plays in C18 calf. Title page mu-
tilated and backed; -a2; sig. D repaired at inner margin.

nICN: Newberry Library, Chicago, IL

Case Y 135.S62885. Reported by Jill Gage. Interleaved. Front cover with
initials "IZ[L?]G"; back cover with armorial of a lion rampant holding a fishing seine.

nICU: University of Chicago, IL

PR3144.T84 1633. Reported by Catherine Uecker. Acquired by the library
early C20. Heavily cropped at head (state of C1r-v and C4v indeterminate-for this reason), interleaved; t.p. and D4 mutilated.

nIU: University of Illinois, Urbana-Champaign, IL

IUA11400. Reported by Tad Boehmer. Bound with nine other plays by Shir-
ley. Bought 22 July 1942 from Dawson's Book Shop, Los Angeles, CA (Cata-
logue 161:36).

nMB: Boston Public Library, MA

G.3976.37. Reported by Jason Moschella. Thomas Pennant Barton copy.
With two other items in a late C19 library binding; some pages untrimmed.

nMH-H: Harvard University, Houghton Library, Cambridge, MA

• STC 22459. C18–19 sprinkled calf. T.p. repaired at inner margin; A1
  conjugacy not certain.
• STC 22459a.5. Early C19 calf. T.p. on a stub.

nMWiW-C: Williams College, Chapin Library, Williamstown, MA

STC 22459a.5. Henry Huth copy (his catalogue, vol. 4, p. 1349); his sale,
Sotheby, Wilkinson & Hodge, 8 July 1918:6825 ("new vellum"), �23. 10s. to
B.F. Stevens and Brown; subsequently sold by James Drake to Alfred Chapin
for the same amount; donated by Chapin to Williams College, 1 Feb. 1923.

nNhD: Dartmouth College, Hanover, NH

Rauner Ticknor VA Sh6t. Examined in digital copy only. George Ticknor
copy; bequeathed to Dartmouth by William Dexter, 1943; afterwards re-
bound in brown library buckram. Ticknor notes of prices at the Roxburghe
and Rhodes sales. -Sig. A; sig. a cropped at head; top half of last leaf torn
away.

nNjP: Princeton University, Firestone Library, Princeton, NJ

3930.3.394. Examined in digital copy only. C19 quarter leather, marbled
boards. Cost code of Dawson's Book Shop, Los Angeles, CA, with price
$22.50; acquired by Princeton in early 1940s. Hole in a2 affecting two lines.
Digitized at http://libweb5.princeton.edu/visual_materials/Misc/Bib

2458362_Shirley_Triumph_of_Peace_1633_NjP.pdf.

nPU: University of Pennsylvania, Kislak Center, Philadelphia, PA

PR3144.T75 1634. Reported by Lynne Farrington. Signature of Courtney
Kenny and date 1719. Acquired 1968 and probably rebound at that time
(marbled boards, leather spine, vellum tips). -a2.

nTxU: University of Texas, Harry Ransom Humanities Research Center, Austin,
TX

• Pforz 935 PFZ. Examined in digital copy only. Brown morocco by Riviere.
  Robert Hoe copy (initials, trace of bookplate); W. Carew Hazlitt? (Shake-
  speare Library Pt. 2, Anderson Galleries, 28–30 Oct. 1918:312); Carl H.
  Pforzheimer (Unger & Jackson 935). Evidently not item 3029 in the Hoe
  sale (Perkins copy with armorial bookplate, modern wrappers).
• Wh Sh66.633ta WRE. Examined in digital copy only. Binding signed by
  Riviere & Son. Bought from a Mr. Butcher (possibly the father of Robert
  W. Butcher of the Lancashire Fusiliers) 67 by Thomas Wise for John Henry
  Wrenn, 1902. Subsequently tampered with by Wise: A3 supplied from bL
  644.c.44; B3 and C2–3 supplied from bL Ashley 1696.

Meisei University, Tokyo

• MR 1063. Examined in digital copy only. Bridgewater copy ("3." in box
  on A1r). -sig. a.
• MR 3092. Examined in digital copy only. -sig. a.
• MR 3093. Examined in digital copy only. -sig. a.

W.W. Greg's copy

"Third impression", a2 in photographic facsimile. Half calf. With "Speech
to the King" bound after sig. a ("fore-margin closely shaved"), illustrated
in his 1946 article (see main article, footnote 1). Sold at Sotheby's, London,
28 March 1960:60; present owner untraced.

APPENDIX 3

Master Table of Printing Variants

The table on the following pages appears in color on the website of the Biblio-
graphical Society of the University of Virginia: http//www.bsuva.org.

APPENDIX 3: Master Table of Printing Variants