University of Virginia Library

Recently George Walton Williams showed, by means of type-shortage evidence, that certain Elizabethan dramatic quartos were set by formes rather than seriatim as was commonly supposed and that this method of composition could be employed by two compositors working simultaneously as well as by a compositor working alone.[1] To the plays


Page 199
cited by Mr. Williams I shall add another, the second edition of Marston's The Insatiate Countess, largely a two-compositor book which seems to have been set by formes within the stints of the respective workmen.

The Insatiate Countess Q2 (which collates A-I4) was printed in the shop of Nicholas Okes in 1616. It was "apparently set up, more economically, from the 1613 edition and used a slightly wider type page — an advantage in the prose passages — and one line of print more to a page . . . . There is no lavish spacing; superfluous [lines] are avoided . . . and the stage directions are contracted and inset."[2] In many respects Q2 is a typical reprint; many of the textual inconsistencies and errors which abound in Q1 were transferred without change to the later edition. Among the typographical features adopted by Q2 from Q1 was the use of italics, indented, for speech prefixes; and, as I shall point out later, certain of these italics are the source of information about the order of composition.

Spelling evidence shows clearly that in sheets A through H two compositors were at work, setting the following type pages:

Compositor X:  A1-A2v   B3-C2v   D3v-E2v  
Compositor Y:  A3-B2v   C3-D3  E3-H4v  
This pattern is a variation of that found in Okes' The History of the Two Maids of More-clacke (1609) and The Duchess of Malfi (1623).[3] Sheet I, however, was evidently set by formes. I1v, I2, and I4 contain enough of Compositor Y's preferred spellings to make assignment to him almost certain, and on I2 the use of an italic capital in the signature links this page with D2 and F1 where Y, unlike X, used italic capitals for the same purpose. I cannot be sure of his presence on I3v, but, since he set the rest of the forme, I think it likely that he set that type page also.[4] Sigs. I1, I2v, and I3, on the other hand, contain a total of seven examples of a distinctive change of the copy spelling to Ile, a form which is characteristic of neither X nor Y. I believe, therefore, that these three type pages, and probably I4v (which contains few significant spellings in its nineteen lines), were composed by a third workman, Compositor Z.

The nature of the collaboration between Compositors X and Y enforces the inference that they were setting simultaneously: generally one composed one-half of the type pages in a sheet while the other composed the other half of the same sheet. The interesting point is that within their respective shares of a particular sheet they seem to have set by formes rather than


Page 200
seriatim. Although the evidence is not conclusive, the pattern of type shortages suggests that this method was chosen by both compositors; indeed, there would be no point in one compositor's setting by formes unless his collaborator adopted the same procedure.

The only letter that seems consistently to have run low during the composition of the book was the italic capital A, which was needed frequently to set the speech prefixes and the names of the characters Anna, Arsena, and Abigal, and both workmen made this deficiency good by substituting the roman capital A. That this substitution was more-or-less systematic is indicated by the clean breaks on certain pages between the appearance of the italic letter and the substituted roman letter,[5] and, as Mr. Williams has shown, where the depletion of one letter is remedied by the systematic substitution of a letter from a different font, the resultant pattern can be very useful as evidence of the order of the setting of the type pages. In The Insatiate Countess Q2 the clearest pattern can be seen in sheet C, where the probable order of setting seems to be as shown in the following chart, in which the numbers to the left of the slash represent A's and the numbers to the right are roman A's substituted for italic:

4v   1v   2v  
Compositor X  2/0  2/8  0/7  0/2  0/1 
2v   3v   4v  
Compositor Y  1/0  5/1  1/4  0/5  0/5 

Here what happened seems to be clear enough. Both compositors had enough italic letter to finish setting sheet B (I shall show later that B2v and B4v were probably the last type pages of the sheet to be composed). But both ran short in sheet C, began to substitute the roman letter (X on C1v and Y on C3v and C4), and finished out the sheet with roman letter exclusively. The implications of this arrangement are clear also; while X was setting two pages of the inner forme of C, Y was setting the other two pages, and thus the inner forme was completed and imposed before work was started on the outer forme.

At this point we can look to the running-title evidence to establish the order of composition of the remaining formes, working on the assumption that the first formes through the press were also the first composed. The book was machined with two skeleton formes: skeleton I imposed A(o), B(o), C(o), D(o), E(o), G(o), H(o), and I(i), and skeleton II imposed


Page 201
A(i), B(i), C(i), D(i), E(i), F(o), F(i), G(i), H(i), and I(o). As we know from type-shortage evidence that C(i) was composed before C(o), it follows that the order of composition of the formes of the first five sheets was probably A(i), A(o), B(i), B(o), C(i), C(o), D(i), D(o), E(i), and E(o). The type-shortage pattern in these formes is as follows, where once again the left-hand numbers represent italic types and the right-hand roman for italic:[6]                    
1v   2v   3v   4v  
Compositor X  1/0  1/1  6/0  3/0  2/0  2/0 
3v   4v   1v   2v  
Compositor Y  5/0  0/1  3/1  0/1  1/1  0/2  5/0  1/0 
3v   4v   1v   2v  
Compositor X  1/0  1/0  3/0 
1v   2v   3v   4v  
Compositor Y  5/0  3/3  0/5  2/3 

It is impossible to be sure that these pages were composed in just the order indicated in the chart, and the patterns are not clear enough to permit much more than a guess at the procedure that was adopted. Nevertheless, I suspect that some such circumstances as the following attended the composition of the early formes. The fact that A's were in short supply suggests that some type was standing when the setting of The Insatiate Countess Q2 was begun; hence, it is not surprising to find that both compositors were forced to substitute roman for italic in order to finish out sheet A. Some of this standing type seems to have been distributed by Compositor Y after the setting of A3v, since he secured from some source three A's for use on A4v. Compositor X also seems to have obtained enough of this letter to last through the setting of his part of sheet B and the first few lines of C1v. That certain type pages were set out of order (e.g., A4 before A3v) seems quite possible since the book was set up from printed copy which could have been cast off with great accuracy.

A(i) and A(o) probably came from the press and were distributed during the setting of sheet B (possibly the six italic pieces obtained by Compositor Y for use on B1 and B2v came from A[i]), but the evidence is by no means as clear as one could wish it. It is fairly sure, however, that no distribution was made during the composition of sheet C, and I suspect


Page 202
that D3 was set by Y rather than X because the latter had stopped composing in order to distribute the type which had been used to set B(o) and B(i). C(i), distributed by Y after the composition of C4v and before the composition of D1, may have been the source of the eight italic types consumed in the setting of D1 and the first part of E3v, but it is impossible even to speculate about the time of distribution of C(o) since that forme contained no pieces of the italic letter.

The spelling evidence indicates that after he had set his part of sheet E Compositor X was called off the job, and the running-title evidence suggests that his removal caused Compositor Y to fall behind the press. The use of skeleton II to impose both formes of sheet F probably means that at the time of the completion of the last forme of that sheet to be set both I and II were off the press and available for use. In order to effect the minimum press delay, the formes of sheet E and F would have had to go through the press in the following order:

Skeleton:  II  II  II 
Forme:  E(i)  E(o)  F(i)  F(o) 
If E(o) had preceded E(i) through the press, the compositor doubtless would have used skeleton I rather than skeleton II to impose one of the formes of sheet F. Since he did not, I think it likely that this order is correct and that Compositor Y continued to set by formes after Compositor X left the job. Had Y set sheet F seriatim, three type-pages of F(o) would have been on the stone at the time F(i) was imposed. It seems improbable, then, that he could not have set F4v, thus making F(o) ready for the press, before F(i) was machined and skeleton II released.

In sheet G, however, the typographical evidence indicates that the compositor began to set seriatim. The following pattern is found:

1v   2v   3v   4v  
Compositor Y  1/2  1/0  7/0  1/2  2/0  5/1 
On both G1 and G2v the substituted roman types appear after the italic types, but on G4v the one roman type is preceded by one italic and followed by four, probably an accidental rather than a deliberate substitution. I judge that distribution took place between the setting of G1 and G1v and between G2v and G4, although in neither case is it possible to be certain of the forme that was distributed. The workman's reason for changing his method of composition is not apparent. It is true that he was no longer setting in collaboration with Compositor X, but, since he was unable to keep pace with the press during the composition of sheet F, it is strange that he would have chosen to set sheet G in such a way that he had to compose three additional type-pages before a forme was ready for the press


Page 203
Nevertheless, he seems to have done so, and we can infer that another press delay occurred between the machining of F(o) and G(i). There is no type shortage in sheet H.

In sheets G, H, and I there is some rearrangement of the running-titles within their skeleton formes, for which haste may have been responsible. It is interesting to note that skeleton II retains in I(o) the same form it had in G(i), which leads one to think that I(o) was machined before H(i). Apparently Compositor Z started work on I(o) while Compositor Y was still engaged in the setting of sheet H, and, one judges, he probably came on the job to prevent another press delay. Thus it appears that the order through the press of the last few formes was as follows:

Selection:  II  IIa  Ia  IIa  Ib  IIb  Ic 
Forme:  F(o)  G(i)  G(o)  I(o)  H(o)  H(i)  I(i) 

From this analysis it is possible to draw two conclusions. First, with regard to the play itself, it may be inferred that the edition was smaller than usual since two compositors were needed to match the pace of the press and since a delay appears to have occurred when one of them withdrew from the job. Secondly, this analysis tends to support Mr. Williams' thesis that setting by formes in quarto printing was a more common practice than it has been thought to be, and it further indicates the value of type-shortage evidence in bibliographical investigation. Moreover, by comparing Q2 with Q1 it should be possible to make some estimate of the accuracy of three of Okes' compositors when they were setting from good copy. The Insatiate Countess Q2 should be a particularly valuable control on the work of the man I have called Compositor Y, since it is possible to see him operating under different conditions of composition — at a fairly moderate pace in the first part of the book and in considerable haste in some sections of the last part.