University of Virginia Library

Search this document 


expand section 
expand section 
expand section 
The Printing of Hamlet, Q2 by Fredson Bowers
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
collapse section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 

expand section 


Page 41

The Printing of Hamlet, Q2
Fredson Bowers

THE SECOND quarto of Hamlet, printed in 1604/5 by James Roberts, offers an interesting problem in running-title analysis. Over and above this, our ability now to separate the pages set by the two compositors provides a check, not usually present in running-title problems, to test results by other evidence. Thus when, as in Hamlet, the spelling tests for compositors equate so precisely with what one may conjecture to have been their stints from the evidence of running-titles, we may be somewhat more confident in the future about roughing-out two-compositor work in books on this running-title evidence. Not, as will be seen, that the running-titles will always tell us the precise pages set by different compositors, for they would not have detected in Hamlet the unusual fact that Compositor X assisted Y with sheet L by setting sig. L4v, and L1r. On the other hand, bibliographical evidence about the presswork may assist to differentiate the work of compositors when spelling tests prove doubtful. For example, in Hamlet the spelling evidence in sheet N is not so clearcut as in other sheets and might even be interpreted as indicating the possibility, at least, that compositor Y set sigs. N1-2v, and X sigs. N3-4v. But the evidence of the running-titles practically enforces the hypothesis that sheet O + A was imposed before either M or N, and, moreover, that inner N was imposed before any forme of M. If this is so, and it seems to be true, then it is all but impossible to work out stints for the two compositors that will enable them to have shared sheet N; and hence it is necessary to weight the almost equal evidence and to award all of sheet N to compositor X, a view confirmed by typographical evidence. Running-title and spelling tests, therefore, offer valuable complementary evidence to decide difficult problems in the analysis of printing.

Some time ago a preliminary survey of the running-titles convinced me of the possibility, even the necessity, for the use of two presses in printing Hamlet Q2. But since I was then concerned only with discussing the nature of bibliographical evidence in criticism of Dover Wilson's non- bibliographical


Page 42
hypothesis that the quarto was set by only one compositor, I merely sketched in the general pattern which the running-titles in their skeleton-formes took, and did not list precise details. The purpose of this note is to present the complete evidence, and from this evidence to construct a hypothesis about the order of the sheets through the press.

Mr. John Russell Brown's article in this volume of Studies gives the result of spelling tests which show that compositor X set sheets B. C, D, F. I, sig. L1 and L4v, sheet N and half-sheet O (doubtless including the title-page on half-sheet A machined with 0). Compositor Y set sheets E, G, H, K, sigs. L1v-L4r, and sheet M. This assignment is confirmed by the running-titles, which show that sheets B-D, F. and I were imposed by one pair of skeleton-formes; and sheets E, G, H, K, and L by a different pair. Sheets M-O exhibit a curious mixture of titles from the two different sets, which will be analyzed below, but the main pattern is clear that compositor X served one press and imposed his formes for that press, whereas compositor Y served a second press and, correspondingly, imposed his own distinct formes for that press.

The evidence of the running-titles follows. I have assigned roman numerals to identify every title and have not indicated the few times when minor re- spacing somewhat alters the appearance of the typesetting. These running- titles are occasionally so similar in their typographical characteristics and in their spacing determined by measurement as to lead to considerable difficulty in precise identification. However, I have reasonable confidence that the table below presents the facts with some accuracy. The relation of the N1 running-title is the only point of serious doubt.

I B2v-C1v-D1v-F1&v-I3v-N3v   IX E1v-G1v-H1v-K1v-L3v-M3& v  
II B2-C2-D2-F4-I4-02-N1(?)   X E2-G2-H2-K4-L4-M2  
III B1v-C3v-D3v-F3&v-I1v-N1v   XI E3v-G3v-H3v-K3v-L1&v-M1v  
IV B4-C4-D4-F2-I2-N2   XII E4-G4-H4-K2-L2-N4  
V B3-C1-Dl-Fl-Il-Ml-Ol   XIII E1-G3-H3-K3-L3-N3  
VI B4v-C2v-D4v-F2&v-I4v-O1v-N2v   XIV E2v-G4v-H2v-K4&v-L4v-N4v  
VII [Bl]-C3-D3-F3-I3-M4   XV E3-G1-H1-K1-L1-M3  
VIII B3v-C4v-D2v-F4&v-I2v-M4v   XVI E4v-G2v-H4v-K2v-L2&v-M2v  

Before we come to the major problem of sheets M-O, several minor irregularities call for discussion. The first concerns the transfer of running-titles from the formes of sheet B to those of sheet C whereby in an unusual manner two running-titles break their formes and are exchanged for each other. Such an event could not arise unless both formes of B were rinsed and lying together on the bench waiting to be stripped and their furniture, including the running-titles, to be transferred to the C formes. Presumably


Page 43
this situation would occur only if the conclusion of machining a sheet coincided with the end of a working day when, as here, the first forme of sheet C did not need to be ready imposed before the last forme of sheet B came off the press. On the evidence it would seem probable that both formes of sheet C were imposed the next morning at the start of the working day and the type of B thus released for distribution. The error, though that is too serious a word, could have been caused if we imagine the two formes of B to have been lying on the bench in the following relation, the outer forme having (a common practise) been turned end to end:      
B (i)   B (o)  
(4) IV VIII (3v)   (1) [ ] I (2v)  
(1v) III II (2)   (4v) VI V (3)  
Since sig. B1 contained the head-title, an additional running-title (VII) needed to be set above what, owing to the turn of the outer-B forme, was the C3 type-page. If, when this was done, the compositor had moved to the left instead of to the right to select the B furniture to place about the C4v type-page, he would have taken title VIII from inner-forme B3v instead of title I from outer-forme B4v.      
C(i)   C(o)  
(4) IV III (3v)   (3) VII VIII (4v)  
(1v) I II (2)   (2v) VI V (1)  

The transfer of the running-titles from C(i) to (Di) was accomplished normally, with each title being placed in its corresponding position in the newly imposed forme. An irregularity developed in the transfer of the outer forme, since C1 and C3 went properly to D1 and D3; but C2v to D4v and C4v to D2v Such an exchange diagonally across a forme is somewhat unusual but not unknown. I have observed it before and speculated whether it meant that some compositors stripped the form not horizontally or vertically, but instead on a diagonal. It will be noticed that precisely the same diagonal exchange takes place in both formes when sheet F is imposed from D. In the inner forme D2 and D4 go respectively to F4 and F2, and in the outer D2v and D4v respectively to F4v and F2v. And in the transfer of the running-titles from both formes of F to I, once more each forme exhibits one diagonal exchange: F1v and F3v to I3v and I1v and F2v and F4v to I4v and I2v respectively. This is an interesting compositorial habit, seemingly, although compositor Y indulges in it four times also: between H(i) and K(i) the titles of H2 and H4 are exchanged diagonally, and between K(i) and L(i) the titles of K1v and K3v; in the outer forme G2v and G4v titles exchange in the transfer to H, and H2v


Page 44
and H4v in the transfer to outer K. The transfer of titles between E(o) and G(o) is not irregular since it apparently results from the forme of E(o) having been set down on the bench in a turned position, although one cannot in this book, at least, rule out the slight possibility that a double diagonal exchange has been made.

At this point two statements may be made about the printing before sheet M which are pertinent in connection with what is to follow. First, except for the evidence that under special circumstances C(o) was first imposed although not necessarily first sent to press, we cannot distinguish, except inferentially, the order in which formes for any sheet were imposed and machined; but we do know that there was perfect regularity of procedure and that each compositor faithfully imposed inner from inner and outer from outer. Second, we know that whatever slight irregularity developed, as in the diagonal exchange of quarters common to both compositors in transferring furniture, each forme maintains its integrity, and except for the special circumstances of sheet C there is no transfer of running-titles and furniture from one forme to another. Hence when this transfer develops with sheet M, and running-titles indeed are exchanged even between compositors, the effect is not random but there is a cause to seek which must lie in the circumstances of composition or presswork and is therefore subject to analysis.

For a start, we have one useful piece of information derived from the spelling tests: compositor X assisted his companion by setting type-page L4v and very likely L1r of a sheet not assigned as part of his stint. Another piece of evidence derives from the running-titles. Although the types for the letterpress of the running-title I4v-O1v-N2v are identical, an adjustment is found in its spacing which indicates that the title was imposed above O1v directly from I4v, and thereupon the spacing differs in the same title on N2v It would appear, therefore, that sheet O+A was imposed and machined before sheet N. Even without this important evidence, we should be forced to infer this order of printing as I-O-N for the work of compositor X, since the only non-random explanation for the evidence of the running-titles in sheets M, No, and O requires some unusual employment of only certain titles in a forme leading consequentially to a necessary breaking-up of the integrity of the skeleton-formes. This requirement can be satisfied only by the imposition and machining of half-sheet O (by twin half- sheet imposition with half-sheet A) before either sheet M or N. Moreover, the evidence of the running-titles requires inner N to have been imposed before either forme of M. This latter requirement cannot be met if the two compositors split sheet N between them, for workman Y would be free to tackle N only after he had finished M. We are able, as a result, to lay out an approximate time schedule for the relation of the two compositors'


Page 45
typesetting. On the evidence that X helped Y by setting sigs. L1 and L4v it would seem fair to conjecture that X was the faster workman, assisting a slower compositor who was falling behind his press; and some evidence in support of this conjecture may be found in Dr. Walker's analysis of the work of the two compositors in her "Collateral Substantive Texts" in the present volume of these Studies. How much faster he was, however, is difficult to assess. Moreover, since we may reasonably assume that the time taken to machine a sheet at each of the two presses would be relatively constant, faster composition would not mean faster printing unless the slower compositor delayed his press by failing to have formes ready when required. The fact that the hypothetically slower Y used two skeleton-formes for each sheet, however, would appear to indicate that at the start, at least, he was able to keep up with presswork. By the time he came to sheet L, however, he may have begun to experience difficulty, and that may explain the intervention of X with this sheet.

The assistance that X gave to Y by setting L1 and L4v is rather curious. From typographical evidence Dr. Gerritsen has discovered—that the italic in each compositor's font was cast on a different size of body so that a similar speech-prefix will vary in measurement by about one millimetre according to the fonts-it is clear that when X helped Y. he abandoned his own cases and used Y's. This is most curious, although it could perhaps be explained as a device not to mix the types when the formes of L were distributed by Y and thus to prevent the depletion of X's cases if the pages set by him in L were not segregated for his own distribution. However, it raises the question whether the two workmen were composing simultaneously, or whether X took over composition while Y was imposing one or other formes of K and very likely engaging in some distribution Whichever it was, both men could not be using the same cases for their separate purposes. It may be, however, since compositors seemed to have worked with pairs of cases, that X composed from one set while Y was either composing from or distributing to the other set.[1]

Still another problem is raised by the particular pages L1 and L4v both from the outer forme, chosen by X. One of three explanations may conjecturally account for the apparent anomaly that X did not proceed to


Page 46
L1v after setting L1. According to the first, he may have been casting off copy and setting by the forme so that when Y. returned Y. would have only two type-pages to compose to have outer L ready for the press. According to the second, when X first started on L1 he may have swung over to help his companion because he thought his own stint was finished, and only a short time later was it decided that he should continue by preparing the title-page.[2] Since further cast-off copy would indicate that some text type-pages would need to be imposed with the title-page, it may be that in next selecting L4v compositor X was killing two birds with one stone. That is, he would not only provide Y with the opportunity to set the rest of sheet L most expeditiously by formes, but he would also provide himself in L4v with the farthest point in sheet L from which to cast off the manuscript to insure accuracy when he came to set the pages of half-sheet 0. Thirdly, we cannot be sure, of course, that L1 and L4v were set seriatim, as has been assumed in the above. Hence it may be that X assisted Y by setting L1 that he then proceeded to the pages of O + A, and finally returned to assist once more with L4v before continuing with N1. But except for the very dubious evidence of capital W's on L4v following on two VV's at the foot of L1 there is no evidence, in fact, of the order of setting L1 and L4v.

An approximate composition schedule for the work of the two men might resemble something like the following. In it, I have included arbitrarily two extra type-pages in X's stint as against Y's, on the evidence for his faster composition. I have also assumed in this schedule that X set both L1 and L4v.

An alternate ending might have been:

Before discussing this curious alignment, we may survey the evidence that leads to the reconstruction of such a schedule. Since there are only four pages of letterpress to combine with four blanks in sheet O + A, there would have been time for compositor X to turn to the assistance of Y by setting O + A and L1 and L4' before the first forme of O + A was required by the press. It may be that I(o) was first off the press and that O(o), requiring only the single type-page O1, was immediately imposed from it,


Page 47
though it is clear that the I(o) type-pages with their running-titles were not used to furnish-out the blanks.      
Bench I(o)   Forme O(o)  
VII VI   bl bl or bl bl  
VIII V   V bl or bl V  
If the last forme of sheet I to come off the press coincided with the end of a working day, perhaps as with C the two formes of O were simultaneously imposed at the start of the next day.      
Bench I(o,i)  Formes O(o,i)  
VII VI II I   bl bl II tp  
VIII V III IV   bl V VI bl  
Whichever it was, we seem to have a roughly similar occurrence to that found in the imposition of sheet C with the selection of II and VI on a line, though from different formes, to impose O(i).

If we arbitrarily assume that O(o) was first sent to press, and that N(i) was imposed while O(i) was printing, the quarters available on the bench, and the imposition, would be: The prime difficulty here lies in the reason why title XII was taken from Y's bench to fill up the fourth quarter when VII and VIII and probably V were available. The regular transfer to the same respective positions of titles I, III, and IV shows that II must have been previously removed, and this evidence joins to the clearcut case of respacing VI between O1v and N2v to indicate the high probability that the imposition of O preceded that of N. As remarked above, this order is also indicated by the necessity to explain the irregularity of the running-titles in M and N as consequent upon the use of only some titles in imposing sheet O + A. If, as seems possible, the wrought-off formes for both presses were placed on the same bench, in spite of the fact that previously there had been no mixing of the running-titles, it would appear that in this case, partly perhaps because no complete X forme was present, the compositor, once more on a line, abstracted the furniture and running-title from another forme. This time he was forced to do so by the presence of only three quarters in I(i). The only


Page 48
odd matter is that he used a quarter from his fellow compositor's forme, which is highly unusual in two-press printing.

When Y came to impose M(i), he had on his part of the bench only the three remaining running-titles from L(i): IX X, and XI. These he imposed in their regular positions except for a diagonal exchange of IX and XI, and he necessarily abstracted his fourth quarter from X's share, choosing VII from I(o).

M(i)   imposition  
VII   IX  
XI   X  

When N(o) came to be imposed by X he had these miscellaneous quarters available, and also L(o) on Y's part of the bench:

I(o)   O(o)   O(i)  L(o)  
VIII   V   VI   XVI  XV  
He seems to have ignored V and VIII, and started to impose N(o) from the last available forme O (i), inserting the furniture of quarters II (title respaced) and VI, and finished by borrowing XIII and XIV from Y.      
N(o)  imposition  
VI   II?  
Thus when Y came to impose M(o), only four quarters were left on either bench, VIII from I(o), V from O(o), and XV and XVI from L(o). These four produced the final imposition of the quarto:      
M(o)  imposition  
XVI  V  

The above reconstruction has been somewhat arbitrarily presented without an attempt made at showing possible alternatives, in large part because no alternative that I have been able to conceive of has in any way fitted the evidence of the running-titles. It is necessary to repeat that on the evidence of sheets B-L of this quarto, as well as the evidence of countless other books, there is every indication that printing would have continued normally to the end, each skeleton-forme as a unit imposing the next forme, if some unusual circumstance had not intervened. So far as I can see, this could only have been the printing of sheet O + A after sheet I. Thereafter, no compositor could have an integral skeleton-forme to transfer as a unit, once compositor X had abstracted XII from Y's materials to


Page 49
impose inner N. On the evidence it would seem that after formes were broken up, the compositors (or at least compositor X) were indifferent where they secured their quarters for imposition.

It is unfortunate that the spelling evidence is not more clearcut for L1 since if X set only L4v the situation would be much more transparent. The collateral spelling evidence does seem to tend towards the hypothesis that X set L1, although something could be said for the appearance of the two VV's as favoring Y. The question has no intimate relation with the breakup of the skeletons consequent on the imposition of O + A and what this tells us of the order of the sheets. And the conjectural schedule is not vitally affected, though it would be much neater if only L4v were set by X. Since there is some small latent doubt about the assignment of L1 I add a proposed schedule in case X set only L4v This is an adjusted printing composition schedule:


One troublesome remaining question is why, if X started setting only shortly after Y, the two men did not alternate sheets regularly as they had previously done in the first Merchant of Venice quarto. It is distinctly odd that copy was cast off to start Y with sheet E when X was either just finishing sheet B or beginning sheet C. Then, although Y skips sheet F, he sets two adjacent sheets G and H, and after skipping I sets two more sheets in a row, K and L, followed after L4 recto by sheet M. It may be that this plan was adopted to enable X to draw more even with him. Why X set A + O after sheet I, and only then returned to assist Y by setting sheet N is explicable, apparently, only by supposing that the original plan was to put X to work on another book, leaving Y to finish the quarto at his own rate. This would help to explain the three-sheet sequence K, L, M given to Y; and we must suppose that the plan was cancelled when A+O was set, and L1 and L4v had been helpfully typeset to assist Y, once it was seen that the schedule was such that both compositors could finish at approximately the same time if X set N while Y worked on M. This hypothesis may appear to hang together, the more especially since only by some such reconstruction of motive can one explain why X helped Y by setting two pages of L instead of continuing with his own stint. However, if we take it that X's stint was planned to end with I, or just possibly with something after I, like A + O, then the distribution of the sheets for setting is fairly regular, given the fact that Y started with sheet E, for each compositor did in fact alternate sheets with his fellow (except for the G-H sequence) up to the start of K, from which point it would appear Y was at first expected to carry on alone to finish the book.


Page 50

If so, there remains only the question why so late a sheet as E was assigned as the beginning of Y's stint. This question has no easy answer, and it is more than possible that the question of copy had something to do with the case, for it must be remembered that the first act, in which contamination from Q1 has been detected, concludes seven lines down on E1. The precise manner in which Q1 influenced Q2 has not yet been established with certainty, nor have its limits been definitely determined. It may prove to have significance that the major, though not the only, pages in which contamination seems apparent appear in sheets B-D assigned as the stint of compositor X, even at the price of starting Y so late as sheet E. But it would be idle to speculate about this matter further until a more precise study is made of the influence of Q1 on Q2, a study which I hope to publish in the near future.



For example, it will be noticed that four regular roman capital W's appear in the upper part of L1 recto, set by X, but when next the letter is required he twice sets VV for W at the foot, perhaps indicating that the sort was exhausted. Yet on L1v, Y regularly sets W. Capital W appears on L4v. On the other hand, the worth of this evidence is doubtful. In X's stint the only other time he uses VV for W is on B4v once, though W appears on C1 which should have been composed before any forme of sheet B was distributed. Compositor Y sets VV once on G3v, line 7, but W precedes it in line 4, and various W's follow on the same page, four to be exact. Y also uses a VV on H1v along with W. and again on H3, K2 (thrice), L2v, and M2. Very likely Y liked to justify a line by substituting VV for W.


That X may have thought he was finished with his assignment could be indicated by the fact that O + A was set by X before N. Sheet N. therefore, would appear to have been added to his stint still later.