University of Virginia Library

Jonson's 1616 Workes, a folio-in-sixes collating ¶6 A-4P6 4Q4, is a bibliographically complex book, and nowhere more so than in the final quires, 4F-4Q, which contain the text of the masques. Here we find frequent headline dislocations and, in 4M-4Q, extensive resetting, as well as the reimposition of a number of unreset pages.[1] In quires 4M-4Q and quire ¶, which was printed last, it is sometimes difficult to establish the sequence of some of

the printing variants. However, evidence from headline rules found in the Workes and in another book helps to resolve most of this difficulty. Two skeletons that first appear in quire 4P of the Jonson volume come from another folio-in-sixes produced by Stansby in 1616, The Surueyor in Foure Bookes by Aaron Rathborne (STC 20748), collating A-V6 X4. The headline rules show when these skeletons left the Rathborne volume for use in the Jonson Folio, and when they returned to The Surveyor, but changed in ways ascribable to their use in the Workes. The condition of the rules in The Surveyor helps to establish the temporal sequence of variant impressions in the last quires of the Workes.

Besides settling the problem of the order of variants, the evidence from the headline rules shared by the Workes and The Surveyor enables us to recognize an instance of printing-house negligence in quire 4P. What appears to be an obvious press-correction in the text of Mercury Vindicated from the Alchemists, on 4P5v, can now be identified as the introduction of a gross error by the printers; the correct state is the earlier one.

In addition, a dated preface in the Rathborne book provides a later date for Jonson's Workes than has yet been proposed.

The basis of this study is an examination of eight copies of the Jonson Folio at the Folger Shakespeare Library (including two large-paper copies), two at the Library of Congress, six copies at Harvard's Houghton Library, one complete and one incomplete copy at the Boston Public Library, one at the Newberry Library, the Scolar Press facsimile of a copy at the Bodleian Library, microfilm reproductions of two copies (one on large paper) at the British Library, and a microprint of a copy at the New York Public Library. The copy of Rathborne's Surveyor referred to is that in the Rare Book Room of the University of Wisconsin's Memorial Library, which I have checked against a microfilm copy.[2]

The extensive resetting in quires 4M to 4O of the Workes is documented by Simpson in the Oxford Ben Jonson (HS, IX, 30-40). The text is wholly reset in all the pages of quire 4M except 4M2r and the bottom half of 4M2v; and in formes 4N3r.4v, 4N1v.6r, 4O3v.4r, and 4O3r.4v. In addition to this resetting, as Gerritsen recognized, "with two exceptions [4P3v.4r and 4P1v.6r],

all those formes of 4M-4P which were not reset were reimposed."[3] That is, all of these formes are found to have been printed twice, each time with a different skeleton. We may add to this list 4Q1r.[4v] and ¶3v.4r, which were also twice impressed.[4] What seems to have happened is that a mistake was made in counting the number of sheets required for printing these formes, which were then printed short. Apparently the counting error was discovered about the time that quire 4P was being printed, since there are two invariant formes in this quire. Then "all the undistributed type of these quires was collected, the missing portions were reset, and further copies printed to make up the numbers" (Gerritsen, 1959, p. 55). The resetting did not involve correction of the text, and some trivial errors were introduced. In the unreset but reimposed pages only three variants appear in sheets between impressions; two of these are trivial and the third is the miscorrection of 4P5v mentioned above which, as we shall see, was probably caused by a disturbance of the standing type.

This theory would require that at the time of resetting and reimposition Stansby had a good deal of type standing—"37 folio pages" in Dr. Gerritsen's estimate (1959, p. 55). That Stansby had a large amount of type at his disposal is indicated in other ways. The headlines in the Workes reveal a large number of skeletons in use. In contrast to the Shakespeare Folio of 1623, printed with never more than two skeletons, and to the Beaumont and Fletcher Folio of 1647, whose various sections were printed with only one or two skeletons at a time, four to six skeletons at a time were normally employed in the Jonson Folio.[5] This difference is not due to the different style of page layout used in the Jonson Workes. The Stansby/Jonson Folio, using english-size type stretching the width of the page rather than pica crowded into double columns, required fewer types per page than did these two other folio collections; however, Stansby nonetheless seems to have had much more type at his disposal than did Jaggard. Although the circumstances involved in the final quires are exceptional, my examination of recurrent types in Every Man Out of His Humour, where printing began, found that Stansby usually had an entire quire of type—twelve folio pages—left standing at a time.

As we shall see, the large-paper copies of the Workes always contain the original impressions of all the reset and reimposed sheets in quires 4M-4P,

and possibly in quires 4Q and ¶ although the evidence is less conclusive there.[6] While some small-paper copies are confined to the later impressions, others have some sheets of the earlier printing bound up with reimposed sheets. Evidently the large-paper copies were run first during the printing of the final quires; then Stansby began the small-paper copies, where the counting error occurred.

Most pages in the Workes have headlines consisting of double-pica italic running titles and page numbers enclosed between two horizontal rules, although sometimes the rules alone appear, when the headlines have been removed for specific reasons, such as the presence of a head title on the page. The individual types of the running titles, especially those with kerns, are often battered or bent in distinctive ways that allow them to be readily identified. In most of the book, the orderly alternation of running titles paired as forme-mates shows clearly that the headlines were treated as an integral part of the skeleton. The rules, too, are usually distinctively bent or nicked, and regularly recur with the same headlines. Some pairs of headline rules continue relatively unchanged for long periods in the volume.

In the later quires of the Workes, the rules become the primary means of skeleton identification, because of the frequent dislocation of the running titles. In quires 4F-4Q the running titles on both the recto and verso pages read "Masques", although thirteen different "works" are included in this section. When a new masque began a page and was introduced by a head title, Stansby's men usually removed the headline, leaving only the headline rules. When this occurred in quires 4K-4N, several running titles moved from one skeleton to another. However, the rules continued to recur regularly in the same positions; thus we can still identify the skeletons used in this section as the same four that are found throughout quires 4F-4O, reset pages excepted. Even though the positions of the running titles change several times, we can recognize on the basis of the headline rules that the original impressions of quires 4F-4O are printed with the same four skeletons found in the last quires of the preceding section, Entertaynments. (In quire 4P two new skeletons appear.)

The headline configurations found in quires 4F-4K are identified in the chart below. First, however, a word of explanation is necessary. The succeeding charts describe the configurations of the running titles and headline rules, which are the usual basis for skeleton identification. For convenience I have labelled each configuration of paired headline elements as a "skeleton," although strictly speaking only four skeleton formes were used in the original

impressions of quires 4F-4O. The headline configurations are labelled with an arabic numeral and a capital letter. The former refers to the rules which, except for the change in quire 4G, remain constant; however, since the running titles change position so frequently, I assign a letter to the skeletons that changes with the successive changes in the headlines.

• Skeleton 1A:
• Running title I: 4F4r, 4F2r, 4G4r, 4G5r, 4H1r, 4I5r, 4I1r, 4K3r, 4K1r
• Running title VIII: 4F3v, 4F5v, 4G3v, 4G2v, 4H6v, 4I2v, 4I6v, 4K4v, 4K6v
• Skeleton 2A:
• Running title IIa: 4F6v, 4G4v, 4G6v; Running title IIb: 4H2v, 4H1v, 4I4v, 4K1v, 4K2v
• Running title VII: 4G3r, 4G1r, 4H5r, 4H6r, 4I3r, 4K6r, 4K5r
• Skeleton 3A:
• Running title VI: 4F4v
• Running title V: 4F3r, 4F6r
• Skeleton 4A:
• Running title III: 4F2v, 4G1v
• Running title IV: 4F5r, 4G6r

During the printing of quire 4G, the headline elements of Skeletons 3A and 4A were recombined; the recto headline of Skeleton 3A (i.e., Running title V) was combined with the verso headline (i.e., Running title III) of Skeleton 4A to form 'Skeleton 3B'. The verso headline (i.e., Running title VI) of Skeleton 3A was combined with a new running title (Running title IX) within the recto headline rules of Skeleton 4 to form 'Skeleton 4B'. The reconstituted skeletons appear on the following pages:

• Skeleton 3B:
• Running title III: 4G5v, 4H4v, 4H5v, 4I5v, 4I1v, 4K5v
• Running title V: 4G2r, 4H3r, 4H2r, 4I2r, 4I6r, 4K2r
• Skeleton 4B:
• Running title VI: 4H3v, 4I3v, 4K3v
• Running title IX: 4H4r, 4I4r, 4K4r

After quire 4G there were no further complications until quire 4K. In quires 4K and 4L and in the original impression of quires 4M and 4N, running titles were again shifted from one skeleton to another. These dislocations were caused by the removal of the running titles from pages that begin with head titles. Thus the following headline changes occurred in quires 4K and 4L. After forme 4K1v.6r was printed, with Skeleton 2A intact, forme 4K2v.5r was printed with the same skeleton; however, Running title VII was removed from between the recto rules of Skeleton 2A because 4K5r contains the head title for The Masque of Queenes. (4K5r was printed with no headline but with the recto headline rules of Skeleton 2A intact.) The removal of Running title VII led to a number of consequent changes in the headlines of quire 4L. When Skeleton 2 next appears, in formes 4L2r.5v and 4L1v.6r, Running title VII has been replaced by Running title I, removed from the recto headline of Skeleton 1A. The new Skeleton 2 configuration is

now designated Skeleton 2B. It consists of the same headline rules as Skeleton 2A, with Running title IIb still in the verso position and Running title I now in the recto position.

In addition, these same dislocations caused the changes that transformed Skeleton 1A into Skeleton 1B. We have seen that Running title I went to Skeleton 2B. Replacing Running title I in the recto headline of Skeleton 1 is Running title VI (which had been in the verso headline of Skeleton 4B as late as the printing of forme 4L3v.4r, printed with Skeleton 4B intact). Running title VII, removed from Skeleton 2A when forme 4K2v.5r was printed with a head title, next appears in the verso headline of Skeleton 1, displacing Running title VIII.

Skeleton 4B is in turn changed into the configuration I have called Skeleton 4C. Running title VIII, displaced from the verso headline of Skeleton 1A, reappears in the verso headline of Skeleton 4, which was left vacant by the relocation of Running title VI to Skeleton 2B cited above. The recto headline of Skeleton 4 remains unaffected by these changes; Skeleton 3B, too, is unchanged.

The headlines remained this way until another, similar dislocation occurred in quires 4M and 4N. Thus the skeletons from quire 4L into 4N are identified and designated as follows:

• Skeleton 1B (same rules as Skeleton 1A):
• Running title VI (replacing I): 4L3r, 4L5r, 4M5r, 4N5r
• Running title VII (replacing VIII): 4L4v, 4L2v, 4M2v, 4N2v
• Skeleton 2B (same rules as Skeleton 2A):
• Running title I (replacing VII): 4L2r, 4L6r
• Running title IIb: 4L5v, 4L1v, 4N5v
• Skeleton 4C:
• Running title VIII (replacing VI): 4L6v, 4M3v, 4M1v, 4N4v
• Running title IX: 4L1r, 4M4r, 4M6r, 4N3r
• Skeleton 3B (unchanged):
• Running title III: 4M5v, 4M6v, 4M4v
• Running title V: 4M2r, 4M1r

In the original impressions of quires 4M and 4N, the three pages 4M3r, 4N2r, and 4N6v contain no running titles because they contain head titles. The removal of the headlines on 4M3r and 4N2r brought about changes which transformed Skeleton 3B into Skeleton 3C; this occurred in the following way. After formes 4M1r.6v and 4M2r.5v were printed with Skeleton 3B intact, 4M3r.4v was printed without a recto running title, because 4M3r begins with the head title to Prince Henry's Barriers. When Skeleton 3 next appeared, in quire 4N (formes 4N3v.4r and 4N1v.6r), Running title I occupied the recto headline (filling the space left empty on 4M3r); this is the configuration labelled Skeleton 3C.

Running title I had been removed from Skeleton 2B before forme 4N2r.5v was printed because 4N2r contains the head title for Oberon the Faery Prince. Thus Skeleton 2B appeared on forme 4N2r.5v with a complete verso headline

but without a recto running title. However, when forme 4N1r.6v was being imposed in this skeleton, the verso running title had to be removed because 4N6v contains the head title for Love Freed from Ignorance and Folly. Whereas the recto headline of Skeleton 2B lacked a running title and now needed one on 4N1r, the verso running title (IIb) was transferred to the recto headline for the printing of forme 4N1r.6v. The skeleton now lacked a verso running title. A new one was composed, using letters from Running title V (removed from Skeleton 3B on page 4M3r); I have labelled it Running title X. This new configuration, "Skeleton 2C," can be seen in formes 4O3r.4v and 4O2r.5v. Skeletons 1B and 4C are unaffected by these changes. Thus the skeletons in the original impressions of formes 4N3v.4r, 4N1v.6r, and 4N1r.6v, and of quire 4O are as follows:

• Skeleton 1B (same rules as former Skeleton 1A):
• Running title VI: 4O4r
• Running title VII: 4O3v
• Skeleton 2C (same rules as former Skeletons 2A and 2B):
• Running title IIb: 4N1r, 4O3r, 4O2r
• Running title X: 4O4v, 4O5v
• Skeleton 3C (same rules as former Skeleton 3B):
• Running title I: 4N4r, 4N6r, 4O5r
• Running title III: 4N3v, 4N1v, 4O2v
• Skeleton 4C (same rules as former Skeleton 4B):
• Running title VIII: 4O6v, 4O1v
• Running title IX: 4O1r, 4O6r

All of the preceding discussion has concerned the original impressions only. The headlines show that the resetting in quires 4M-4O took place after the original imposition and printing of all the quires in the volume. All of the reset pages were printed with the same skeleton, which we will call "Skeleton o", whose headlines were composed of various elements taken from the headlines of Skeletons 3C and 4C.[7] Because Skeletons 3C and 4C remain intact to the end of the volume in the original impressions and are also found intact in some of the reimposed but unreset formes, Skeleton o cannot have been constructed before the miscounted sheets of the final quires were all printed and before some unreset formes were reimposed to correct the counting error. In addition to the reset formes, several of the unreset formes were reimposed in Skeleton o, specifically 4N3v.4r, 4N1r.6v, and 4N2v.5r. Thus these sheets were reimposed at about the same time as the reset pages, and there is no difficulty distinguishing the earlier impressions of these formes from the reimpressions.[8]

The remaining unreset but reimposed formes were not reimposed with Skeleton o; thus the earlier states are not distinguished from the later ones so readily. These are 4N2r.5v; both formes of sheets 4O1.6, 4O2.5, and 4P2.5; and formes 4P3r.4v, 4P1r.6v, 4Q1r.[4v], and ¶3v.4r. Forme 4N2r.5v varies between some copies (including the large-paper copies) printed with Skeleton 2B and others printed with Skeleton 3C. The latter state is found in sheets perfected with Skeleton o, which, as we noted above, could not have been constructed until after the completion of the original impressions and some reimpositions. The use of Skeleton o in the perfecting forme suggests that the impression of 4N2r.5v in Skeleton 3C is the later one, and the impression in Skeleton 2B the original. This conclusion is supported by the fact that Skeleton 3C was used in other reimpositions (as we will see below), while Skeleton 2 is last seen intact in the original imposition of quire 4O. Finally we may note that the original state of 4N2r.5v, printed with Skeleton 2B but lacking a recto headline, helps to explain the progression of the headlines from the configuration of Skeleton 2B to that of 2C, as well as the change from Skeleton 3B to Skeleton 3C, as described above. Since Skeleton 3C came into being as a result of the impression of 4N2r.5v with Skeleton 2B, the impression of this forme with Skeleton 3C must be the later one.

The reimpressed but unreset formes in quires 4O and 4P can be considered as a group; they are listed below. The large-paper states are the originals, the others the reimpressions.

                 

Forme	Skeleton found in large-paper copies [9]	Skeleton found in other copies
4O2v.5r	Skeleton 3C	Skeleton 1Bv.6r
4O2r.5v	Skeleton 2C	Skeleton 4C
4O1v.6r	Skeleton 4C	Skeleton 3C
4O1r.6v	Skeleton 4C	Skeleton 1Br.2Cv
4P3r.4v	Skeleton 6	Skeleton 4C
4P2v.5r	Skeleton 5	Skeleton 1Bv.6r
4P2r.5v	Skeleton 6	Skeleton 1Br.2Cv
4P1r.6v	Skeleton 1B	Skeleton 1Bv.6r

A few words of explanation are in order. The skeletons in the right hand column above labelled 1 Br.2Cv and 1Bv.6r are so designated to describe their composite character. Skeleton 1Br.2Cv is a combination of the recto headline and rules of Skeleton 1B (now placed in the verso position) with the verso headline and rules of Skeleton 2C (now placed in the recto position). Skeleton

1Bv.6r is composed of the verso headline[10] of Skeleton 1B paired with the recto headline rules and verso running title of Skeleton 6 as seen in the large-paper copies of 4P3r.4v and 4P2r.5v. The hybrid character of Skeletons 1Bv.6r and 1Br.2Cv is itself an indication that the formes printed with them are reimpressions, not originals. Skeleton 1B, which was split apart to create these two hybrid skeletons, is found intact in the large-paper copies of formes 4P1r.6v, and in all copies of 4Q2v.3r and ¶3v.4r. This suggests that the reimpression of the variant sheets in quires 4O and 4P took place after the original imposition of all of the final quires; at any rate there can be little doubt that Skeletons 1Bv.6r and 1Br.2Cv are found on reimposed sheets, and that Skeletons 1B, 2C and 6, when found intact, precede them.

The second salient feature in the list above is the appearance of two new skeletons in the printing of the original copies. Skeletons 5 and 6 first appear in the large-paper copies of quire 4P; Skeleton 5 can be seen in all copies of the Workes on 4P3v.4r, an invariant forme. Both the rules and the running titles of these two skeletons are wholly new to the Jonson volume. Fortunately, the rules are quite distinctive: they are the same as those found in quires N-R of The Surveyor. The rules of Skeleton 6 are particularly recognizable. Here are the verso headlines as seen on R3v of The Surveyor:

The headline rules of Skeleton 5 are less striking, but can also be readily identified. Skeletons 5 and 6 can be traced back to the early quires of the Rathborne book, although they are at first hardly recognizable as the same rules.[11] Gradually, through wear and through changes in the positions of the rules, they assume the configurations seen in the Jonson Folio. Quires N-R of The Surveyor are printed with Skeletons 5 and 6 alone. In quire R the headline rules look virtually the same as those in certain copies of the Workes, quire 4P. In quire S of The Surveyor to its end, however, not only do these rules enclose running titles in different types, but the position of the rules changes and the two pairs of headline rules in Skeleton 6 are now separated and appear in two different skeletons.

This separation of the two sets of headline rules from Skeleton 6 is explained by the composition of Skeleton 1Bv.6r in the Workes. The recto headline was separated from its forme-mate and paired with the verso headline from Skeleton 1B. This fact indicates that Skeletons 5 and 6 were transferred to the Workes between the printing of quires R and S of The Surveyor.

Further evidence that the final quires of the Workes were printed between quires R and S of The Surveyor is seen in the fact that on S3r of The Surveyor we find the recto headline rules of Skeleton 2 as they appear on 4Q4r of the Workes, their last appearance in the book. When Skeletons 5 and 6 returned to The Surveyor from the Workes, these headline rules went with them.

Recognition of the prior integrity of Skeleton 6 establishes the originality of the large-paper impressions of formes 4O2v.5r, 4O1r.6v, 4P3r.4v, 4P2v.5r, 4P2r.5v, and 4P1r.6v. There is good reason for supposing that the large-paper copies of formes 4O2r.5v, 4O1v.6r, and 4P3r.4v are also original impressions, as they perfect original impressions. The large-paper copies of 4O2r.5v, showing the last appearance of Skeleton 2C intact, had to have been printed before forme 4Q1v.4r, where we find the recto headline of Skeleton 2C paired with a new forme-mate in all copies. As we have noted, the reimpressions

were apparently not done until after the printing of all the original and invariant formes in the final quires and preliminaries.

We are now in a position to describe the original printing of quires 4O an 4P. 4O3v.4r was originally printed with Skeleton 1B, 4O3r.4v and 4O2r.5v with Skeleton 2C, 4O2v.5r with Skeleton 3C, and both formes of sheet 4O1.6 with Skeleton 4C.

In quire 4P Skeletons 5 and 6 were introduced. Formes 4P3v.4r and 4P2v.5r were printed with Skeleton 5 and their perfecting formes with Skeleton 6. 4P1v.6r was printed with Skeleton 3C and 4P1r.6v with Skeleton 1B.

The sequence of the variant impressions has some textual significance. On 4O6v, line 35, the original setting gives the word "ripned"; in the later impression this reads "rip'ned." This would appear to be a straightforward case of press-correction.[12]

A more interesting case is that of the variants in 4P5v. In the large-paper copies and others, we find at lines 9-10 "a thing of nothing . . . a | toy, a lease"; the other copies give "nothing of nothing . . . a | any lease". The two states of this page are recorded by Simpson as a press-correction, which is what they would normally appear to be; they seem also to have been accepted as such by Jackson.[13] However, the headlines show that the correct state, the one printed with Skeleton 6 intact, is the earlier one, while the garbled version is the reimpression. The variants are confined to the left-hand margin of two consecutive lines. Apparently the standing type was disturbed between the two impressions, and Stansby's staff haphazardly reconstructed the gap in the text by guesswork, rather than referring to their copy. This example underscores the Oxford editors' recognition that "the text of the entertainments and masques is often carelessly printed" (a recognition that, unfortunately, weakens the authority of the Oxford edition for the texts of several masques).[14]

We have been able to identify with confidence the large-paper copies of the masques' quires through 4P as the originals. Such confidence diminishes when we examine the headline evidence in 4Q and ¶. Quire 4Q is a gathering of two sheets. The text of the masques ends with the conclusion of The Golden Age Restored on 4Q4r.4Q4v is a blank page. A major textual variant occurs on 4Q3v and 4Q4r with the repositioning of two speeches. In most copies, Astraea's final speech and the "Galliards and Coranto's" (ll. 221-240 in HS) precede Pallas's final speech and the Chorus's couplet of praise to Jove (ll. 199-220); in the large-paper copies (and some small-paper copies) they are transposed as they appear in the Oxford edition. The positions of the speeches were changed without any disturbance to the surrounding text, and the headlines of these two pages and of their forme-mates are invariant.

Forme 4Q2r.3v was printed with Skeleton 4C. 4Q1v.4r is anomalous, as it was printed with a new hybrid skeleton, composed of the recto headline of Skeleton 2C paired with another headline, one that appears nowhere else in the volume. We have already noted that 4Q4r gives us the last appearance of the recto headline of Skeleton 2C in the Workes (the recto rules then go to S3r of The Surveyor). The verso headline of Skeleton 2C (i.e. Running title X within Skeleton 2C's verso rules) is found in the more common state of 4Q1r—the only variant headline in quire 4Q—and in most copies of ¶6r; this formerly verso headline now occupies a recto position. The forme-mates of both 4Q1r and ¶6r are blank pages. (This configuration would seem to be preliminary to the creation of Skeleton 1Br.2Cv, which pairs this headline in this position with the recto headline from Skeleton 1B, now occupying the verso position.) The large-paper copies of 4Q1r were printed with a headline found nowhere else in the volume except for the variant copies of ¶6r. This headline will be discussed below at greater length.

Quire ¶ contains the preliminaries to the Workes, and the headlines have no running titles, but rules and page numbers. The first leaf of this quire is a blank. ¶2r contains the title page to the Workes and ¶2v is a blank page. Thus only ¶3r-¶6v have headlines. ¶[2v].5r bears the recto headline rules from Skeleton 3C, and ¶[2r].5v the verso headline rules from Skeleton 4C. Besides ¶[1v].6r, only one other forme, ¶3r.4v, was reimposed. The large-paper copies etc. were printed with Skeleton 4C, the others with Skeleton 1Br.2Cv. The large-paper copies would seem to be the originals, as ¶3v.4r shows Skeleton 1B still intact in all copies.

Three wholly new headlines appear in quires 4Q and ¶. One of these occurs in all copies of ¶6v, whose forme-mate is a blank page; the only other place in the volume where these rules appear is the headline of 3S2v, an anomalous headline over the text of Catiline, whose forme-mate, 3S5r, the title-page of the Epigrammes, lacks a headline; Dr. Gerritsen believes that "forgetfulness in printing . . . or perhaps a reluctance to decide on its form

[i.e. the typography of the title-page to Epigrammes], deferred the printing of [3S2v.5r] till the end of the volume."[15] Another new headline appears in all copies of 4Q1v, paired with the recto headline of Skeleton 2C as a forme-mate (on 4Q4r). This is the only appearance of this headline in the Workes.

The third new headline is particularly intriguing. It is found in the large-paper copies of 4Q1r and a few variant copies of ¶6r. In both of its appearances it has no verso forme-mate; thus it displays a distinct skeleton configuration and can be labelled "Skeleton 7". In both of its appearances Skeleton 7 is related to textual variants. Although there are no variants on 4Q1r, 4Q1r.[4v] perfects 4Q1v.4r, where a major variant, the change in the ending to The Golden Age Restored, occurs. The so-called revised ending is always found in copies (including all the large-paper copies) that show Skeleton 7 on 4Q1r.[4v]; Skeleton 7 appears only in copies with the "revised" ending. The other appearance of Skeleton 7 is even more remarkable in that it is directly related to a textual change and in the fact that it presents the only variation among large-paper copies that I have seen in these later quires. On some copies of ¶6r, including some though not all of the large-paper copies, side-notes are added to the commendatory verses found on this page. The copies containing the side-notes were imposed with Skeleton 7, the others, including some large-paper copies, were imposed with the verso headline of Skeleton 2C in the recto position, i.e. with the recto headline of Skeleton 1Br.2Cv, perhaps before it was paired with the recto headline from Skeleton 1B. The introduction of the side-notes indicates that some of the large-paper copies of this sheet were printed after most of the copies. It is only logical to suppose that the side-notes on the verses were added late rather than deleted early during the printing of forme ¶[1v].6r. There is no good reason for deleting them, whereas their addition serves a good purpose; "Without them it would not be clear that Donne's verses referred to Volpone" (HS IX, 52). Likewise the large-paper version of the ending to The Golden Age Restored has always been regarded as the revised state by literary critics.[16] It is quite likely that we have in the outermost sheets of quires 4Q and ¶ a reversal of the pattern we have seen in quires 4M-4P, where the large-paper copies were demonstrably the first printed.

On the other hand, the large-paper state of the only other reimpressed forme in these two quires, ¶3r.4v, printed with Skeleton 4C, precedes the variant state, which was printed with Skeleton 1Br.2Cv; the latter skeleton cannot have been composed until the two halves of Skeleton 1B were split apart, but Skeleton 1B is found intact on all copies of formes 4Q2v.3r and ¶3v.4r.

The evidence suggests that the sheets printed with Skeleton 7 were reimpressions. Thus, the large-paper copies of sheet 4Q1.4 and some large-paper copies of sheet ¶1.6 were reimpressed after the printing of ¶3r.4v, a forme which seems to conform to the pattern of quires 4M-4P in that the large-paper version is manifestly the original. The alternative supposition, that the sidenotes on ¶6r were deleted rather than added and that the revised ending of The Golden Age Restored (i.e., the large-paper version) is in fact the original, and the version ending with the speech of Pallas and the praise of Jove the revision, is much less likely.

Why so many new skeletons should have appeared in these final quires is not wholly clear. We now know that Stansby interrupted the printing of The Surveyor to transfer Skeletons 5 and 6 to the Workes. It seems reasonable to suppose that the counting error which necessitated the resetting and reimposition in the final quires was discovered in the course of printing quire 4P, where we find two formes (4P3v.4r and 4P1v.6r) which were not reimposed; we might then suppose that the discovery of the counting error caused Stansby to devote all of his resources to the completion of the Jonson Folio. The problem with this explanation is the fact that three of the four formes impressed with Skeletons 5 and 6—4P3r.4v, 4P2v.5r, and 4P2r.5v—were also printed short, as was ¶3r.4v.

Because we can see when Skeletons 5 and 6 left and returned to The Surveyor, we have new evidence for the date of the completion of the Workes. Dr. Gerritsen concluded that the Jonson Folio was completed "not long after the beginning of summer" in 1616 (1959, p. 55). However, Rathborne's preface to the reader on A6v of The Surveyor ends with the following: 'From my Lodging, at the house of M. ROGER BURGIS, against Salisburie-house-gate in the Strand, this sixt of Nouember, 1616.' Quire A contains the preliminaries of The Surveyor; as one woúld expect, it was not the first quire in the volume to be printed. However, neither was it the last. The condition of the headline rules shows that quire A was printed some time between quires K and M.[17] The dated preface may have been delivered to Stansby with the rest of the book, in which case 16 quires or 47 sheets of the book were printed between the first week of November and the time of the original printing of quire 4P of the Workes. However, even if the dedication were written just before quire A was printed, at least seven or eight quires (A plus L or M through R), or 20-23 sheets of The Surveyor were printed between that time and the printing of 4P in the Workes. (We have already seen that Skeletons 5 and 6 entered Workes quire 4P between the printing of The Surveyor quires R and S.) At this point, eight sheets still remained to be printed in the Jonson Folio, not counting resetting and reimposition. Thus at the very least 25 edition sheets, and possibly 55 edition sheets, were printed between November 6 and the completion of the Jonson Workes. Stansby printed over 695 (and possibly over 700) edition sheets in 1616, which gives a

monthly average of over 57 sheets.[18] No doubt the rates of production varied greatly, but this figure at least provides a point of reference: that about two weeks to a month passed between November 6 and the completion of the Workes seems a plausible estimate. Thus late November or early December, 1616 seems a likely date for the completion of the Jonson 1616 Folio.

There is one more issue raised by the evidence of Skeletons 5 and 6 that I would like to mention, and that is the possibility of concurrent printing in Stansby's shop. When Skeletons 5 and 6 entered the Workes, 17 quires (A-R), or 51 sheets, of The Surveyor had already been printed, as well as 83 quires, or 249 sheets, of the Workes. I see no reason to suppose that the printing of the Workes was interrupted when the 17 Surveyor quires were printed; the headline rules in the Workes are too regular in their continuation from earlier quires. The Surveyor quires may well have been printed concurrently with the latter part of the Workes.