University of Virginia Library

The Composition of the Quarto of Much Ado About Nothing
by
John Hazel Smith

For many years critics of Much Ado spoke highly of Benedick and Beatrice, but found little else to praise. Particularly, many of them complained of disunity in the play. In the nineteenth century, theories of stratification began to be advanced in explanation of the supposed defects.[1] In 1923 J. Dover Wilson refined upon these theories and gave them apparent respectability by an elaborate analysis which concluded that the "old play" which was imperfectly blended into the new was "an early play by Shakespeare himself." He concentrated on the bibliographical problems in the quarto, finding in them evidence that the copy had been a theatre prompt copy in which Shakespeare had marked revisions so unclearly that the compositor had frequently become confused. For example, he pointed to sig. G1, which has a narrower tail margin than other pages and contains thirty-nine lines of text instead of the normal thirty-seven; in addition, a total of seven verses are compressed into five lines of prose. Obviously, the compositor was forced by some factor to make unexpected adjustments in the page.[2]

Only very recently have critics found a unified theme in Much Ado,[3] and as late as 1948 G. B. Harrison spoke favorably of the stratification theory in an edition which has been used by thousands of undergraduates.[4] Similarly, though Wilson's position has been convincingly

disputed by Sir Walter Greg, Sir Edmund Chambers, and others,[5] they have not offered positive refutation of all his bibliographical arguments. Only limited scientific work has been done on the quarto. Greg proved that the copy for the play had been Shakespeare's foul papers, not a theatre prompt copy (loc. cit.). And W. Craig Ferguson, arguing from the absence of stops after speech prefixes and from other typographical practices, has disproved an old theory of multiple compositors.[6] I have applied to this quarto the scientific bibliographical methods illustrated by the work of Charlton Hinman on the Shakespeare First Folio and of George W. Williams and Robert K. Turner, Jr., on other Renaissance quartos.[7] These scholars have repeatedly disproved the old belief that Renaissance compositors invariably set the pages of their copy seriatim. Using their methods I intend to prove that the Much Ado quarto must be added to the growing list of quartos known to have been cast off and composed by formes. This knowledge will let us answer with relative certainty a number of the bibliographical questions raised by Wilson and not specifically answered by others. Thus may we finally lay to rest the theory of stratification.[8]

The quarto of Much Ado was printed in 1600 by V. S. (probably Valentine Sims) as part of the same job with 2 Henry IV.[9] It collates

A-I4 [A1] is the title-page, and [A1v] is blank. Through sheet G two skeletons were used:

• Skeleton I imposed A(i), B(i), C(o), D(o), E(o), F(o), G(i);
• Skeleton II imposed A(o), B(o), C(i), D(i), E(i), F(i), G(o).

The precedence of A(i) is proved by three italic types (B 1, B 2, d 1) from that forme, which are then divided between the formes of B.[10] The running-titles would suggest that B (i) was also precedent; but two italic types (B 1, B 2) from B (o) which are divided between the formes of C prove the precedence of B (o). Though there is no indication of the cause for delay, or explanation of what the press was doing during the delay, there seems to be no alternative but to assume that the compositor stopped working after A long enough for both skeletons to be available when B (o) was imposed. By chance the skeleton which had imposed A (i) was used for B (o).

Such evidence as there is substantiates the running-title evidence for the precedence of formes in sheets C through G. In these sheets the following formes were precedent: C(i), D(i), E(i), F(i), and G(o). No type evidence (save the doubtful e6) confirms the order for C, but almost certain proof of the precedence of C(i) lies in the fact that a word is divided between C4 and C4v (aſſu-/rance). If the copy was cast off, it is inconceivable that allowance for a divided word would have been made, especially in a prose passage. Thus, the division of a word between formes is proof either that the pages were composed seriatim or that the forme containing the first part of the divided word was composed first, with the compositor re-marking his copy for the other forme. That Much Ado was composed by formes will be abundantly clear later. It is already indicated by the types (B 1, B 2, d 1) which appear in two adjacent sheets: under seriatim composition it would be very rare to find, as we find several times in this quarto, on the first or second page of a second sheet a type from either forme of a first sheet. Hence, the word divided between C4 and C4v proves that C(i) was precedent.

Another word divided between D2 and D2v (Bene-/dicke) confirms the precedence of D(i). A type (B 3) used in both C(o) and D(o) further substantiates this order: since C(i) was precedent, a type from C(o) would not have been available for the precedent forme of D. By the same token, a type (B 3) which is used in both D(o) and E(o) confirms the running-title evidence for the precedence of E(i). For F I have found no confirming type evidence. But if F(i) was

precedent, as the running-titles suggest, then there is confirming type evidence for the precedence of G(o): namely, a type (B 2) used in both F(o) and G(i).

After G there was another delay in the pressroom, as shown by the running-titles of H and I. For those two sheets new running-titles were composed:

• Skeleton III imposed H(i), I(i);
• Skeleton IV imposed H(o), I(o).

Some of the types in these two skeletons came from the old skeletons:          

SKELETON III	SKELETON IV
H1v used the running-title from G4v;	H1 used the running-title from G2;
H2 used the running-title from G4;	H2v apparently used all new types;
H3v used at least some types from G1v;	H3 apparently used all new types (though the N may be that from G1);
H4 used types from both G3 and G1 (some are questionable, but some types from both pages are clear).	H4v used some types from G4v (M, h); some others seem to be new.

Thus, each of the new skeletons used running-title types from both of the old skeletons. Both formes of G must have been wrought off before either forme of H was imposed. Type evidence shows that both formes of G were wrought off before either forme of H was composed: H(i) was composed before H(o), for types (I3, m1, n1) from H(i) are divided between the formes of I. But a type (B 2) from the non-precedent G(i) appears in H(i), on sig. H2, and this could not have happened if G(i) were in type when the compositor was working on H(i).

The delay after G was evidently not the fault of the pressmen. While they were working on G(o), the compositor was certainly setting G(i), but what he was doing after that can only be conjectured. The breakdown of the skeletons which necessitated the resetting of the running-titles indicates that the letterpress was removed from the loosened formes and that the running-titles then lay unused for a time. Removal of the letterpress implies distribution of the type, but I have found no types from G(o) in H(i). Distribution of the type implies an expected continuation of composition, but as we have seen the composition of H did not proceed. I strongly suspect that while the pressmen were working on G(i) the compositor was working on some extra job—probably a small, one-sheet job which was wanted quickly—on which he used the type which he had distributed from G(o). In anticipation of his return to the Much Ado quarto, he laid Skeleton II

aside for re-use, but during the ensuing interval it got jostled about and some of its type misplaced. If the extra job required two formes, perhaps the process was repeated with Skeleton I and the type from G(i). In any case, he sooner or later returned to H(i) of the Much Ado quarto, presumably while the pressmen were working on the (conjectured) extra job. After it was composed, he salvaged what he could from the intact running-titles of both of the old skeletons: Notice that he tended to transfer to Skeleton III, which imposed H(i), pairs of running-titles from the old skeletons (those from G1v-G4 going to H2-H3v, and those from G3-G4v going to H1v-H4), though one running-title in each pair had to be patched up with some new type. Notice also that it is only in the skeleton for the non-precedent H(o) that we get some running-titles which required all new types: the compositor had apparently salvaged all he could in the earlier forme and in the first page of H(o).

Some of this is conjectural, but there is a certain amount of circumstantial evidence to give it credence. If there was an emergency extra job, one wonders what it was. The answer may lie in the 2 Henry IV quarto. Briefly to review the well-known facts, this quarto survives in two issues of which the first (Qa) lacks a scene (III.1). At some time after the entire first issue had been printed off, the same compositor who had set it added the missing scene as part of a six-leaf gathering (E) in the second issue (Qb): cancelling E3-4 of Qa, he substituted four leaves, E3-6, newly composed, in several ways stretching his copy to fill the four leaves. The reason for the omission from Qa and the length of time between the completion of Qaand the composition of E3-6 (Qb) have been subjects of much controversy, and a thorough exploration of these questions is beyond the scope of the present study.[11] But it is quite possible that the omission of the scene was discovered while Sims' workers were still printing Much Ado; and the discovery might be just the sort of emergency which would cause them to stop work on what they were doing. According to the theory as described above, the types from G(o) of Much Ado were distributed after G(i) was composed. I have found one of these G(o) types in the added pages of 2 Henry IV Qb: I1 on E5v, l. 28 (Iohn). Presumably the types from G(i) were distributed when they became available. I have found two types from G(i) of Much Ado in Henry Qb: m2 on

E4v, l. 4 (mountaines), and T1 on E3, l. 15 (The).[12] I have found many other types which appear in both quartos, but it is very interesting that I would find in this one added forme three types from sheet G of Much Ado—and that none of the three reappear in the precedent forme of H: I1 reappears in H(o), T1 in I(i), and m2 in I(o). All these reappearances are possible if we assume that the types from 2 Henry IV Qb were distributed immediately after impression, as they probably would be, for use in Much Ado. Of additional interest is the fact that when the compositor came to reconstruct some of the running-titles for the Much Ado quarto, he used at least one of the types which had printed the running-titles in Henry Qb: u in Henry Qb E4, and Much Ado H4v. This conjectured sequence of events is based on the assumption that 2 Henry IV Qa was printed before Much Ado, and that is the order which Greg assumed from the appearance of the two title-pages (Bibliography, loc. cit.).

Now, the fact that the two issues of the Henry quarto survive in approximately equal numbers made Shaaber (and others) assume logically that about half of the run was issued as Qa, the other half as Qb: one might ask, then, why so many copies of an imperfect book would have been distributed if the imperfection was discovered even before its companion volume was completed. Several answers are possible. Though these two quartos are companion volumes, it is very likely that the sheets of the earlier volume were delivered to the publishers (Wise and Aspley) as soon as it was completed: after all, there might be some buyers for that first volume. If it took Sims' workmen a week or more to print the first seven sheets of Much Ado, the copies of the Henry quarto could have been in the booksellers' stalls for at least a few days (depending on how long was required for its sheets to be made into books), and quite a number of copies could have been sold. These could hardly be recalled when the error was discovered—even if the publishers wanted to recall them. An alternative answer implies some cynicism in the publishers (or in Sims, depending on who was responsible for the original omission), who could save the cost of several hundred sheets of paper by not making perfect the whole run of the first issue, but perhaps rectifying only those copies not yet

stabbed, where cancellation would be the easier. At least the question does not present an insurmountable hurdle to the theory.

Theory it remains, however, and I do not insist on it. But I do insist—to return to the analysis of the Much Ado quarto—that a delay after G has been proved and, more important for our purposes, that the order of composition through H is as I have outlined. Finally, if H(i) was precedent, the running-titles indicate the precedence of I(i), and I have no type evidence to confirm or dispute the indication.

Two factors make very difficult the determination of the precise times of distribution. Although I have identified almost half a hundred types used almost 140 times in this book, and although several italic types occur in adjacent sheets, through sheet G I have found no roman type which reappears earlier than two sheets after a given use. There is one exception: e6 occurs in B(i), C(o), and D(o), but the identification is questionable. Apparently the compositor held more roman types relative to need than italics: no roman type occurs more than four times in the quarto, and most occur less often; on the other hand, two italic B's occur in six of the nine sheets and another in five. Even so, if the roman letter were distributed at the same time as the italic, we could expect at least several times to find a roman letter used in two adjacent sheets. That we do not suggests that the formes (through G) were at least sometimes distributed in two stages: the italic letter as soon as they became available, the roman later. Perhaps another eye might identify more roman types than I have found. But in any case the following patterns of reappearances of roman types (excluding e6) demonstrate the difficulty of using them to determine times of distribution:

• A types are not found in B, but both A(i) and A(o) types are found in the precedent C(i): e3, h1, I2, ſh1, ſt1 (see the type chart);
• B types are not found in C, but both B(i) and B(o) types are found in both formes of D: e1, e4, m2, n2, q1, ſh2;
• C types are not found in D, but both C(i) and C(o) types are found in both formes of E: b1, e3, h2, I2, n1, s1.
• D(o) types are not found in E, but are found in both formes of F: e4, n2, ſh2, w2. But I have no D(i) types before G: e1, f1, m2, ſ1;
• E types are not found in F. E(o) types are found in G(o) but not in G(i): I1, ſſ1. E(i) type is found in G(i) but not in G(o): T1.

I omit mention of the later sheets because the delay after G renders conclusions about the distribution of F and G roman types uncertain; and we have already seen that H(i) was distributed before composition of the next sheet.

The other factor which complicates our problem is the pattern of substitution of one kind of type for another. One of these involved the use of VV in place of W.[13] The following chart shows the number of W's (to the left of the slash) and the number of substituted VV's (to the right of the slash) in the text on each page in G:

   

G1	G2v	G3	G4v	Total	G1v	G2	G3v	G4	Total
8/0	2/0	3/2	1/1	14/3	3/0	5/0	1/0	3/3	12/3.

VV's do not occur in the other formes, which use W's in the following numbers: A(i) 4, A(o) 3; B(o) 14, B(i) 12; C(i) 6, C(o) 3; D(i) 6, D(o) 8; E(i) 10, E(o) 11; F(i) 10, F(o) 10; H(i) 8, H(o) 11; I(i) 10, I(o) 9. Thus, the VV's are used where the demand for "W's" is greatest, no doubt where the W's ran low. Yet the substituted types do not occur in a group unmixed with normal ones. The last two "W's" on G3 are VV's, as is the first on G4v; from somewhere the compositor got another W for the last page of G(o). He could, of course, have found it lying on the floor. He could have got it by distributing an earlier forme: clearly the supply was replenished before G(i) was composed. (The only type evidence of distribution of the roman letter is that E[o] was distributed before G1 was composed [see I1 and ſſ1 in the type chart] and E[i] was distributed before G2 was composed [see T1].) But I suspect that the distribution occurred between formes and that the mixed pattern of W's and VV's is due to a curious practice of the compositor. Note that on G4 the first, second, and fourth "W's" are W's; the third, fifth, and sixth VV's. Apparently the compositor did not wait until complete exhaustion of his supply of a letter to begin using substitutes.

This is a curious kind of aesthetics, but no other explanation suffices for the pattern of the far more extensive substitutions of roman B's in speech prefixes and stage directions, which were normally set in italic. The play has an exceptionally large number of characters whose speechprefix and stage-direction identifications begin with B: Beatrice, Benedick, Balthaser, Bastard (John), Brother (Antonio), Boy, and Borachio—and of course Benedick and Beatrice are identified often. In view of the previously demonstrated shortage of italic B's, it is not surprising that we find frequent substitutions of roman B's in speech prefixes and stage directions. They occur in five of the nine sheets, on seventeen of

seventy pages. Fourteen pages contain both roman and italic "B's"; on nine of them one or more italic B's appear after the first roman B. Foul cases cannot explain the mixtures, for the roman letter appear with too much system to be explained thus. G3 is an extreme example of the interspersing: on that page, which contains more "B's" altogether and more substituted roman B's than any other page, twenty-three "B's" are arranged two roman, three italic, four roman, two italic, six roman, six italic. Although intra-page distribution might explain the last six B's (see below), it does not explain the earlier mixture. While extreme, this page differs from several others only in degree: for example, C1v has five italic B's, one roman B, then two more italic; C2 continues with three italic, but then there are two roman; the only roman B out of six "B's" on I3v is the second one; etc. I think the compositor's desire (however half-hearted) for a balanced use of the substitute fount whenever he saw that substitution would be necessary has been demonstrated.

The first roman substitutions occur in B(o). Twenty-one italic B's were used in A(i), twenty more in A(o). We know that the compositor replenished his supply of italic letter from A(i) before composing B1 (see B 1 in the type chart). Even so, he ran short. The italic and roman "B's" in speech prefixes and stage directions of B(o) occur in the following pattern:[14]

   

B1	B2v	B3	B4v	Total
6/0	6/0	6/1	10/6	28/7.

The sixteen italic B's on B3 and B4v are consecutive, and the lone roman B on B3 cannot be certainly explained. Perhaps the compositor had distributed only part of the italic types from A(i) before composing B1 (when he knew that he would need some B's for the new forme); then, after running out of B's on B3 he finished the distribution. At least we know that a distribution had occurred before B4v was composed (see B 2): thus, we know that the substitutions were not dictated by composition in the order B4v, B3, followed by distribution. In any case, I infer from these figures that the compositor began the quarto with something like fifty B's in his tray.[15] Whenever in succeeding formes we find nearly fifty B's in type, we can expect to find roman B's in numbers roughly sufficient to fulfill requirements in those formes.

Before composing B(i), the compositor must have distributed the twenty B's from A(o), though I have no type evidence of such distribution. In B(i) he used twelve italic B's. In C(i) he again substituted roman B's. If he had distributed the twenty-eight B's from B(o) before beginning C(i), he would have had no need to substitute; and yet he did distribute B(o) before finishing C(i): see B 1 in the type chart. Thus, he must have distributed the italic letter from B(o) after composing C3v; hence, we find the following pattern of substitution in C(i):

   

C1v	C2	C3v	Distribute	C4	Total
7/1	3/2	0/2	B(o) italic	6/0	16/5.

Even with this fresh supply of italic B's, the previous record of substitution indicates that the compositor would not have had enough to carry him through C(o), which required twenty-nine B's. He must have obtained some additional B's during the composition of C(o), probably from B(i). The only type evidence of such distribution is the questionable e6; otherwise I have found no B(i) types before D(i): see e 2 and m2. The issue here is confused by the additional substitution of two roman B's in D(i):

   

D1v	D2	D3v	D4	Total
5/0	1/0	6/1	0/1	12/2.

The roman B on D3v is followed by six italic B's. According to my calculations, the compositor should have had some half-dozen italic B's left after composing C(o)—assuming that B(i) was distributed during that forme. This is precisely the number of italic B's which he used in D(i) before having to use the roman B on D3v. Immediately after using that roman B, he must have distributed some italic letter, probably that from C(i), enabling him to finish the page with italic B's. Again there is no type evidence of such distribution: I have found no re-use of C(i) types before E2v. And even this sequence does not explain the lone roman type on D4. It could have been caused by an accident or been part of some press correction. But this single anomalous type does not disturb the findings thus far for the book as a whole.

After D(i) the compositor ran into no more difficulty until G(o). C(o) was distributed before the composition of D(o): see B 3 in the type chart. Those twenty-nine italic B's were more than enough, with further distributions, for the next few formes. D(o) required only six B's, E(i) sixteen, E(o) twenty, F(i) seventeen (including one for a Latin word in the text), and F(o) ten. Distributions of italic type had occurred as follows: D(i) before or during E(i) (probably, though there is no type evidence of it); D(o) before or during E(o) (see B 3); E(i) and E(o) before or during the corresponding formes of F (probably, though I have found no types from these formes before G). F(i) was distributed before or during the composition of G(o): see d 2. Despite this distribution, the compositor had to use many roman B's in G(o):

   

G1	G2v	G3	G4v	Total
3/1	9/3	11/12	0/1	23/17.

I suppose the distribution occurred during composition of G3, after all the roman B's and just before the concluding six italic B's (see p. oo). The roman B on G4v is, under this circumstance, as mysterious as that on D4.

No roman substitutions were made in G(i), which required seventeen B's. F(o) was distributed either before composition of G1v or, more likely, before composition of G3v: see B 2.

In H(i) I have found a type (B 2) from G(i) but none from G(o); in H(o) I have found types (B 1, I1, W2) from G(o) but none from G(i).[16] If my conclusion is correct that 2 Henry IV E3-6 (Qb) or some other extra job was composed after G, and if it was in two formes, then types from G(o) went into the precedent forme of the extra job. After that forme was composed, the types from G(i) presumably were distributed and used in the other forme of the extra job: they certainly were if the extra job was Henry Qb. The compositor then returned to Much Ado H(i), but he apparently made no distribution before

composing it; else we would expect to find in that forme types from the (original) G(o). The G(i) type which appears in H(i) is no doubt one that was not used on the extra job. Before composing H(o), he distributed the precedent forme of the extra job and used several of the types from the original G(o): see B 1, I1, W2 in the type chart. Notice that both roman and italic types were distributed. There was no shortage of B's in H: H(i) required eleven, H(o) thirteen. (Notice that Henry Qb requires only a few B's.) The non-precedent forme of the (conjectured) extra job must have been distributed by the time the compositor worked on I(i), and H(i) was distributed by the time he set I3v (see I3). Yet there were not enough B's for either forme of I:    

I1v	I2	I3v	I4	Total	I1	I2v	I3	I4v	Total
1/9	14/0	6/1	1/10	22/20	1/5	1/0	4/1	1/2	7/8.

If the compositor composed the pages in this order and if he still had his full complement of B's, then I do not know why the substitutions occur in such numbers or in such a pattern. It would seem that he ran low on B's on I1v, replenished his supply (from H[i]?) for I2, but ran low again for the last two pages of I(i). He was still short of B's in I(o), but may have got some from somewhere before composing I3. I have no evidence that H(o) was ever distributed before the quarto was completed. An alternate explanation is that the compositor set the pages in a different order, perhaps    

I2	I3v	I4	I1v	I2v	I3	I1	I4v
14/0	6/1	1/10	1/9	1/0	4/1	1/5	1/2.

Such an order makes sense for I(o), but I do not know why it would have been adopted for I(i).

If there is some doubt about some of the details of distribution and substitution, there is no doubt whatever that the copy was cast off and the quarto composed by formes. As we have seen, this fact is established by the overwhelming evidence of the types which occur in adjacent sheets and of the many substituted roman B's which, except for sheet I, occur in only one forme per sheet. Seriatim composition cannot explain these phenomena so well as the sequence of composition by formes which I have described—minor discrepancies notwithstanding.

The application of this result to the critical problem of whether Much Ado is a stratified play is simply that the bibliographical problem on G1 now seems likely to have been caused immediately by composition-room practices and ultimately by the foul papers which were the copy. Some miscalculations in casting-off are to be expected. All the more is this true of Much Ado, which contains a high percentage of prose: verse is simple to cast off, but accurate estimation of the space

which manuscript prose will occupy in print is not so simple, especially if the manuscript has revisions.

It has escaped notice, apparently, that G1 is not the only page in the quarto which has more or fewer than the normal thirty-seven lines. A3, B3, C4v, and E3 contain each only thirty-six lines (not counting, for these or any other pages, the lines which contain non-textual matter: running-titles, signatures, and catchwords): in each case the blank line occurs before a stage direction, whereas stage directions are not normally separated from the context by blank lines. A3v, B4v, C2v, F4, and I2v have each thirty-eight lines of text, the last one sharing a line with the catchword.[17] Half of the anomalous pages occur in non-precedent formes and suggest that the compositor was adjusting for miscalculations. It might be argued that the compositor did not care whether his pages came out to a consistent length; but the many adjustments in apparently normal pages which I shall discuss below reveal that he went to great pains to make his pages consistent. Aesthetic considerations may at times have dictated a too-long page: the thirty-eighth line on half of the too-long pages is of one word, and nowhere in the quarto did the compositor carry a single-word line onto a new page. But he often carried a single line of a speech onto a new page, and a few of these overruns have only two or three words. In any case, there is little question about three of the long pages with thirty-eight lines: the last line of each is full and argues for a motivation out of necessity rather than aesthetics. Finally, some of the anomalies could have been accidental; but the number of adjustments discussed in succeeding paragraphs argues for deliberation by the compositor.

The compositor saved one line each on A4, C1, F3, and I2v by crowding an overrun into the margin above or below the end of a line; scores of other overruns (including several short ones on C1) were printed as separate lines of type. The crowded overrun on C1 involves an entrance; by contrast, many other entrances were printed on separate lines even when there was room for them at the ends of verses. (Some other possibly crowded entrances occur on F2, l. 35, and H2v, ll. 1-2, both discussed below; and on I2, l. 29.) The crowded overrun on

F3 involves an exit. Exits were usually printed in the margins at the ends of lines, but in at least one instance where the exit was too long for the space (C3, l. 29) the entire exit was printed on a separate line. Either C3 was being stretched to fill the page[18] or F3 was being compressed, or both. The crowded overruns on A4 and I2v involve the text. The crowding on the former page (ll. 4-5) may have been necessary because of a misjudgment in the handling of the stage direction at l. 3: here "Exeunt. Manent Benedicke & Claudio" occupies one line; by contrast, on C1, l. 24 (a crowded page), "exeunt: manet Clau." is placed in the margin after a line of text. Whatever the case, the different handling of the similar directions on two pages is probably significant. The crowded overrun on I2v is unquestionably significant, for the page is crowded in other ways: It contains thirty-eight lines. The identifications "Epitaph" (l. 3) and "Song" (l. 13) do not occupy separate lines as the identification "The Song" does on D1. And the epitaph and song are not spaced stanzaically as the earlier song is on D1v.

Compression of other sorts occurs on other pages. On A2v, the final speech-prefix is not indented as most are; moroever, in the final speech the spacing between words and especially after commas is very tight, and there are almost no final -e's. On G4v, l. 6, two brief speeches are set up on the same line, for the only time in Much Ado; contrast, for example, C2v, ll. 20-22, and C4v, ll. 9-10. H2v, ll. 1-2, presents a striking handling of stage directions. An entrance is marked in the margin rather than in a separate line; the compositor's method elsewhere indicates that the motive was to save space. As it happened, however, an exit was to be marked at the same place, and the marginal entrance occupied the space normally assigned to exits. Faced with this dilemma, the compositor used Dogberry's rule ("an two men ride of a horse, one must ride behind") and placed the exit above the entrance. This order was normal: in every other place in the quarto, exits appear before entrances, whether on the same line or not (e.g., A4, l. 3; B2, ll. 4-5, 27-28). In the present instance, the placing of the exit was unfortunate,

for it left Antonio with one line to speak after he had exited. There is no reason to suppose, however, that the compositor shared our concern for precision in such matters: his concern was simply to get his material onto the page, and his material was apparently longer than he had anticipated. Possibly he did not notice that "Bro." (Antonio), identified in the speech prefix, was one of the unnamed persons who were leaving ("Exeunt amb."), or he deliberately decided that proper placement was more important for an entrance than for an exit.[19]

Several pages show evidence of the stretching of copy to fill a page. B3v C3 (already discussed in another context), and E1v run over onto their last lines two to four letters plus a point which would not have needed to be run over. Especially on E1v there seems to have been deliberate stretching. Spaces between the words of l. 36 are slightly wider than in most similar prose passages (e.g., E3, ll. 28-30); the tightening of these spaces and the elimination of two final -e's (in "foole" and "appeare") and of one of the p's in "appeare" would have left room for "is." on l. 36. On C3 the one-word last line ("sute.") could easily have been fitted onto l. 36 by the abbreviation of the speech prefix and the elimination of two superfluous -e's; the same speech prefix was abbreviated in a speech just above. On the same page l. 26 is spaced out (by an exceptionally unabbreviated speech prefix) so that the speech would run over onto l. 27. On B1v, l. 1 was spaced out so that "good." would have to go onto l. 2: the speech prefix was not abbreviated as it was elsewhere on the page; four words in l. 1 have unnecessary final -e's; and the spacing of l. 1 is very wide. D2, ll. 17-18; F1v ll. 13-14; F2, ll. 10-11; D2v, ll. 33-34; and E2, ll. 12-13, show similar wide spacing and long spellings apparently designed to allow the overrun of a single word onto a separate line.[20]

The fact of compositor adjustment beyond the normal needs of "justification" of lines seems irrefutable, and at different times the adjustments reflect a varying need, either to stretch or to compress the copy. The likeliest explanation is that the compositor wanted to make each page end at a predetermined point and that he could not rely on succeeding pages to compensate for miscalculations. If, say, I (i) was already in type when a miscalculation was discovered on I2v, the compositor had to adjust for the miscalculation by the end of I3. To be sure, only about half of the adjusted pages are in non-precedent formes; moreover, the adjustments on sig. 4v of several sheets seem surprising, for presumably there the compositor would have some room for error. In either case, however, it was probably easier to adjust for miscalculations immediately than to face the distasteful alternative of re-marking (or allowing for the mis-marking on) whatever succeeding pages had already been cast off. Some of the adjustments within the same formes seem to offset each other: for example, the shortage on A3 may have been a compensation for (accidental?) compression on A2v. But even this compensation is a significant adjustment showing carefulness about the amount of copy to go on each page.

Thus, the pages of anomalous length and those showing significant adjustments amount to more than a third of the total pages of the quarto. And most of the adjustments were no doubt caused by miscalculations in the casting off of the copy. The causes of the miscalculations were the high percentage of prose and also, no doubt, partly the nature of the copy. Greg's proof that the copy for Much Ado was the author's foul papers makes very likely the conclusion that, if revisions in the copy caused the compositor to miscalculate, they were currente calamo revisions. I would, then, answer Wilson's question "how can the compositor have miscalculated, if the 'copy' was in order?" (p. 98) by saying that the copy was not in perfect order, but for reasons other than Wilson thought. This applies to the most drastic miscalculation on G1 as well as to the others. The bibliographical problems in Much Ado are just that—problems for the bibliographer, not for the critic of Shakespeare's art.

Type Chart[21]

• b1: C4v, l. 27 (be); E2; l. 30 (hobby); H2v, l. 11 (beene)
• b2: A4v, l. 33 (humble); F3, l. 20 (bid); I1v, l. 21 (borne)
• b3: B2, l. 18 (breake); H4v, l. 8 (bid)
• B 1: A3v, l. 27 (Beat.); B1, l. 30 (Bened.); C4, l. 21 (Bor.); E2v, l. 6 (Bast.); G2v, l. 31 (Beat.); H2v, l. 12 (Bened.)
• B 2: A3v, l. 2 (Be.); B4v, l. 21 (Beat.); C1, l. 12 (Borachio); F4v, l. 35 (Beatrice); G3v, l. 5 (Beat.); H2, l. 15 (Brother)
• B 3: A3, l. 12 (Beat.); C2v, catchword (Beatr.); D1, l. 17 (Balth.); E4v, l. 21 (Bor.); G3v, l. 8 (Beat.)
• B 4: G4, l. 17 (Bor.); I2, l. 35 (Beat.)
• d 1: A2, l. 2 (daughter); B4, l. 26 (Pedro); E2, l. 12 (Claud.); G3, l. 17 (Bened.)
• d 2: F3v, l. 32 (Claudio); G3, l. 32 (Bened.)
• e1: B3, l. 12 (to the); D4, l. 16 (Beatrice); G4, l. 10 (alreadie)
• e2: B1v, l. 36 (her); G1v, l. 35 (heare); I2, l. 4 (mingle)
• e3: A4, l. 10 (ſpeake); C2, l. 20 (her); E4, l. 22 (the)
• e4: B2, l. 35 (prefent); D3, l. 4 (peace); F4v, l. 15 (mifgouernement); I1, l. 26 (fweete)
• e5: B3v, l. 15 (Berrord); F3v, l. 13 (me); H1, l. 18 (before)
• e6: B1v, l. 7 (Hero); D3, l. 6 (doe); H4v, l. 25 (henceforth); perhaps also C4v, l. 23 (new)
• e 1: D4, l. 7 (Hero); F2v, l. 12 (Leonato)
• e 2: B4, l. 25 (Hero); D2, l. 17 (Leonato)
• f1: A2, l. 15 (of); D1v, l. 18 (foole); G4, l. 22 (efteſt)
• h1: A2, l. 21 (haue); C1v, l. 9 (Why); F3v, l. 10 (ha); I3, l. 2 (that)
• h2: C4v, l. 20 (had); E3v, l. 35 (child)
• h3: H3v, l. 4 (hath); I2v, l. 24 (the gentle)
• H1: B2v, l. 35 (Hero); E3, l. 35 (How)
• I1: B2v, l. 22 (I make); E2v, l. 35 (I ſee); G1, l. 7 (I thy); H2v, l. 37 (I will)
• I2: A3v, l. 12 (I had); C2, l. 17 (I was); E4v, l. 32 (I haue); I4, l. 21 (I would)
• I3: C4, l. 8 (I told); H3v, l. 24 (I aske); I3v, l. 6 (Ioue)
• m1: C2, l. 28 (may); F4v, l. 25 (mans); H2, l. 11 (me [second]); I2v, l. 2 (monument)
• m2: B3v, l. 11 (my); D2, l. 6 (amaze); G1v, l. 7 (nor my); I1, l. 32 (man)
• M 1: A4, l. 3 (Manent); F2, l. 2 (Mar.)

---

26




• n1: C1, l. 33 (Againſt); E2v, l. 11 (manifeſt); H1v, l. 17 (not); I2v, l. 29 (and)
• n2: B4v, l. 2 (dauncer); D4v, l. 2 (fortunate); F1v, l. 9 (fine)
• q1: B2, l. 8 (muſique); D3, l. 4 (quarrel)
• s1: C4v, l. 16 (is); E2, l. 18 (his); H2, l. 28 (this)
• ſ1: D2, l. 20 (ſo); G1, l. 26 (ſo); see n. 16
• ſh1: A4v, l. 26 (ſhould); C4, l. 5 (diſho-); F3v, l. 27 (ſhe); H1, l. 11 (ſhould)
• ſh2: B3, l. 13 (ſhall); D4v, l. 12 (ſhape); F1v, l. 3 (faſhion)
• ſſ1: E3, l. 19 (defartleſſe); G2v, l. 15 (inwardneſſe)
• ſt1: A2, l. 17 (muſt); C1v, l. 13 (ſtrike); G4, l. 29 (Conſtable)
• T1: B4, l. 3 (The); E3v, l. 27 (Truely); G2, l. 6 (To); I3v, l. 27 (The)
• T2: A4v, l. 32 (That); H1v, l. 31 (Thy)
• w1: C1v, l. 30 (willow); F3v, l. 37 (what); I2, l. 19 (why)
• w2: D2v, l. 15 (wiſedome); F2v, l. 23 (were)
• w3: A3v, l. 11 (would); I1v, l. 9 (who)
• w4: B1, l. 3 (with ſickneſſe); E2v, l. 36 (wed); I1v, l. 29 (betweene)
• W1: C1v, l. 2 (Whither); F3v, l. 12 (Will); H1v, l. 14 (Why)
• W2: G3, l. 13 (Why); H1, l. 4 (Whoſe)
• y1: A4v, l. 9 (your); H3, l. 12 (body)

Notes