University of Virginia Library

The Printing of the Second Quarto of Romeo and Juliet (1599)
by
Paul L. Cantrell and George Walton Williams [*]

THE RECENT PUBLICATION OF TWO EDITIONS OF Romeo and Juliet and the appearance of supporting articles by the several editors have suggested the pertinence of an examination of the printing of the second quarto of the play.[1] This quarto, the 'good' quarto, was printed in 1599 by Thomas Creede and is generally thought to depend in some manner on a Shakespearean manuscript. It is therefore the text from which all subsequent editions derive. In this article the writers will investigate two aspects of the 1599 quarto: the composition and the copy. The first section will discuss the compositors and the press-work; the second will offer some observations on the nature of the copy.

I. The Composition

1. Compositor Determination. It has been assumed that the second quarto of Romeo and Juliet was set by one compositor, but this opinion is contraverted by several kinds of bibliographical evidence. Variant compositorial characteristics suggest the presence of two compositors, and the mechanical evidence of the press-work corroborates that suggestion.

The evidence from typographical practices is of four kinds. The first concerns the setting of proper names in stage directions. In this quarto proper names in stage directions appear generally in distinguishing type (roman) on 35 signatures; for example: A4: Enter old Capulet in his gowne, and his wife. B4: Enter Capulets Wife and Nurse.[2] F2v: Enter Mercutio, Benuolio, and men. I2: Enter Frier and Countie Paris.[3]

However, on five signatures — A3v, A4, K3v, L4, and L4v — proper names in non-distinguishing type (italic) occur as follows: A3v: Enter Benuolio. A4: Enter Tibalt. K3v: Enter Will Kemp. L4: Enter Romeos man. L4v: Enter Capels. / Enter Capulet and his wife. / Enter Mountague.

A second kind of evidence is found in the forms of certain speech prefixes. On pages L3v and M1 appears the prefix "Frier" in that fully spelled-out form, eight times on L3v and twice on M1. Only this form is found on these pages, but in the rest of the book "Frier" never occurs; instead, the prefix appears 43 times (including two catchwords[4]) in the short form "Fri."[5] In a similar manner the spelled-out form "Peter" occurs once on E3 and ten times on K3v. The abbreviated form "Pet." occurs on E4 (once), E4v (once), and L2 (twice). On K3v the speech prefixes "Minstrels" (1), "Minstrel" (1), and "Minst." (5) appear seven times (once as a catchword); on K4 the form is "Min." (1). Again, on K3v occurs the prefix "Fidler" (1); on K3 the form is "Fid." (1).[6] The short forms, "Fid." on K3 and "Min." on K4,

are in marked contrast to the long forms on K3v and bracket that page. Though other instances of fully spelled-out prefixes occur, no comparable concentration of these forms exists in the quarto.

The third kind of evidence is provided by the catchword when it happens to be a speech prefix. On 25 pages in the quarto the catchword indicates the beginning of a new speech on the top of the following page. On 22 of these pages the form of the catchword consists of the speech prefix and the first word of the speech (e.g., K2v Fri. Peace).[7] On two occasions the catchword consists of the prefix only: F2v Mer., K3v Minst. On one occasion the catchword consists of the first word of the speech only: B1v But.[8]

The fourth kind of evidence is provided by the catchword when it is juxtaposed with a stage direction directly above it. This situation occurs thrice in the quarto. On I4v and on K3 the stage direction and the catchword appear in contiguous lines; on K3v the direction "Exit" is separated from the catchword "Minst." by an intervening line of quads, though such an insertion results in a page of abnormal length (39 linear spaces).

It will be noticed that on several pages the first three categories of evidence are conflicting. Thus on A4 there is one stage direction with proper names in italic type, three with names in roman; on E3 there is the long prefix "Peter" at the first appearance of the speaker, but there is also a catchword composed of a prefix and the first word of a speech; on F2v there is a catchword composed of a prefix only, but there are also two stage directions containing proper names in roman type. And on A3v, though there is no conflict, it is evident that the stage direction "Enter Benuolio.", by virtue of its unique position on a line of dialogue — not set off, constitutes a special case in the quarto.

If we then confine ourselves here to those pages on which no conflicts occur, we may tabulate the clear evidence thus:

The four kinds of typographical evidence combine to distinguish signature K3v markedly and to connect that page with the seriatim group grouped and seem actually to indicate a second hand, we tentatively posit a second workman for these pages, Compositor B. To Compositor A we assign the greater portion of the work. Spelling tests will clarify the distinction further.

Since it is on signature K3v that the typographical evidence most strongly indicates the presence of a second compositor, we may look to that page first in an attempt to discover spelling differences that distinguish it from the other pages of the quarto. Two words, "dry" and "cry", appear at once to set K3v off from the general pattern of the book and at the same time connect it with pages L4 and L4v, already tentatively assigned to Compositor B on typographical evidence.

         

dry-beate	K3v	drie beate	F3v
		drie	D4v, G3, H3v, K3
cry	K3v	crie	D1, E3, L4v
descry	L4
outcry	L4v

The use of the termination -y for these short forms is in marked contrast to the spellings in the rest of the quarto of these particular forms (as indicated in the list above) and of similar or related forms (as indicated in the list below):

• alie F4
• defie L2v
• denie C3, D2, D3, I2v, K4
• flie C1, G1, G4(2), L2v
• prie L2
• replie I1
• spie F2v, I3
• trie I4(2)

The instances of these endings point to a decided preference on the part of Compositor B for the -y spelling and an exclusive preference on the part of Compositor A for the -ie spelling of these short forms.[9]

Since the evidence of this -y /-ie spelling difference confirms the evidence of the typographical practices on pages K3v, L4, and L4v, we feel confident in examining Compositor B's five pages together, including pages L3v and M1 that share the same practices, for additional -y / -ie variations. A pattern similar to that for the short forms is found in the termination of nouns, verbs, and adjectives:

• K3v merie, cary, mary, mercy, speedy
• L3v vnthriftie, stony, goarie, guiltie

---

111




• L4 happy, bloudie, Countie, safetie
• L4v County, emptie
• M1 Citie, Countie

The proportion in these is ten -ie forms to seven -y forms. When this ration is applied to spelling on the pages of the rest of the book, it is seen that one page, L3, maintains the same proportion: L3: merie, bewtie, bloudie, enemie, Appothecarie, vnsauoury, weary, Appothecary. We therefore assign this page also to Compositor B on the ground of this spelling distinction. In these six pages there are then 15 -ie forms and 10 -y forms of the endings of nouns, verbs, and adjectives.

On the other hand, with the exception of such words as "any", "many", "every", "ready", and "already", and such special cases as "Lady" and "body", Compositor A in the larger portion of the book sets nouns, verbs, and adjectives with terminal -ie almost exclusively. Of 270 instances of relevant forms on the 83 pages that we think were set by Compositor A, only twelve variants from the -ie spelling are noted:[10]

• A1 sundry, Burby
• C2 watry
• D2 liuery (liuerie, F3)
• E4 penny
• G4 vnworthy
• G4v sympathy
• H4v vnworthy, worthy
• I4 harlottry
• K4v penury
• L1 hurry

The distinction is even clearer in the use of adverbs and adjectives ending in -ly or -lie. Of the ten instances of these modifiers in the work of Compositor B, two adverbs on signature L4 end with -lie: "Happlie" and "newlie". Compositor A employs the -ly termination almost exclusively and, in contrast to the Compositor B forms, sets for example: "happily" H4, I2v; and "newly" A1, G1. In a passage in the second quarto demonstrably set from the first quarto (see below) Compositor A changes Q1 "louelie" and "liuelie" to his preferred forms "louely" and "liuely" B3v. Of 136 instances of suitable forms on the pages of Compositor A, all occur with the -ly ending with three exceptions: "earlie" C2v, C4v (cf. "early" passim); "solie" E3 (cf. "soly" E2v, immediately preceding).[11]

The evidence may be summed up thus:

       

	Compositor A	Compositor B
Short Forms	Exclusive -ie	Prefers -y
Nouns, verbs, adjectives	Much prefers -ie	Indifferent
Adverbs, adjectives in -ly	Exclusive -ly	Indifferent

Though this single spelling variant serves generally to distinguish the work of the two compositors in this play, there are particular spellings on each of the six pages of Compositor B that contribute further to this distinction.

                                             

Compositor B	Compositor A
K3v	merie (L3)	merrie B3v, B4v, I3v, (merie I4), K1v
	cary	carrie A3
	yron (2)	iron L1v, L2
	musique (4)	musick (e) C3, D4, F1, F2, K1v(2)
	mary	marrie B4v(2), C4, E1v, H4v, I2, I3 (2), I3v, I4
L3	point	poynt F4v(2), G1, K1, K1v
	merie (K3v)	[see above]
	staie	stay [throughout]
	dym	dim I1v
L3v	Whoes	whose G4v
	Master	maister [throughout]
	entents	intents L2 (2), L4
	yong	young [throughout][12]
	heere (L4, M1)	here [throughout]
	hower (M1)	houre [throughout]
	nest	neast F1v
L4	poison	poyson B3, E1, G2, G4, H4, K1, K4v (2) L1 (3), L4, M1
	Pittifull	pitifull K3
	heere (L3v, M1)	[see above]
L4v	maners	manner (s, ly) C2v, C4, I3v
M1	foorth	forth A2, A4v, D4v, G3v (2), H1v, I4v, K2, L1, M1v
	heere (L3v, L4)	[see above]

       

Compositor B	Compositor A
mariage	marriage D3v, E1v, F1, H1v, I1v, I2, K1, M1, M1v
stayed	staid H1v, L1
hower (L3v)	[see above]

One other characteristic may be mentioned, though deduced from negative evidence only. In Compositor A's pages the single -e form of the personal pronoun and the verb occur regularly, but the double -ee form appears twenty-one times; Compositor B's pages in this quarto contain only the single -e form: hee, shee, mee, wee, yee: B2(2), B2v(2), B3, B4(2), C4v, E2(2), E3v(2), F2v, F3v, H2v(2), I2v, I4, L1v / bee: B3, D3v.

It seems clear that the spelling evidence corroborates the typographical evidence in distinguishing the work of two compositors in Q2 Romeo and Juliet.[13] Signatures K3v, L3v, L4, L4v, and M1, tentatively assigned to a second compositor on typographical evidence, may be confirmed by spelling evidence. To them signature L3 may be added on spelling evidence alone since the nature of the text here admits none of the typographical evidence (no stage directions, no speech prefixes, no prefix catchword). The three pages noted earlier on which there were conflicts of typographical evidence are confirmed by spelling evidence as not belonging to the second compositor, Compositor B, and the three conflicts in question may be attributed to chance or to the oversight of the principal compositor, Compositor A.

2. The Printing Process. The quarto is printed with two skeleton formes recurring in regular sequence throughout sheets B through K.[14] In these sheets one skeleton is used for the inner forme, the other skeleton for the outer; they alternate on the press seriatim. The two running titles for the inner forme of sheet A (I and II) are used in the outer forme in sheets B-K, and one of the titles (II) is also used in the outer forme of sheet A.

The sequence of two alternating skeletons is consistent through sheet K, with the following exceptions: a turn of the inner skeleton at the imposition of sheet E and a turn back at sheet K; similarly a turn of the outer skeleton at sheet E and a turn back at the imposition of sheet G. The outer-forme running title I is replaced or reset in F (I'), and titles II and III seem to have been readjusted after loosening.

After sheet K this sequence is observed:

It will be noted that the outer forme of sheet L breaks the previously established pattern by utilizing the skeleton (here turned) that has been regularly working the inner forme and simultaneously discards the usual running title (VII) that has been used in the inner skeleton. The rejection of running title VII and the use of title II in the respective quarter of the outer forme of sheet L suggest that signature L4v had been set complete with running title II and its furniture before the other three quarters of L outer were ready for imposition; and further, that title II was available for such imposition because it had previously been discarded from sheet M, which must therefore have been imposed before L outer.

This evidence from running titles can be explained only with great difficulty as accompanying the work of one compositor; but a reasonable explanation may be offered by resorting to the hypothesis of a second press, and thus of a second compositor. The presence of the second compositor in sheet L, strongly suggested by the evidence already advanced, is then confirmed at the same point in the quarto by the evidence of the press-work. The compositorial and the mechanical

evidence in conjunction leave little doubt as to the presence of a second compositor or the extent of his work.

A conjectural time table may now be set up for sheets K, L, and M, indicating the pages set by the two compositors and the passage of the formes through the press.

From this information we may visualize something of the situation in the shop. As Compositor A approached the end of page K3, Compositor B concluded his work on another job and became available to assist him. Compositor B accordingly set page K3v while Compositor A set K4, in that manner making the inner forme of sheet K ready for the press at the earliest possible moment. (The inference is that Compositor A was near the end of K3, for had he been near the beginning of that page, Compositor B could have set K4v and thus made ready first the outer rather than the inner forme. But perhaps this hypothesis is too nice.) After completing K3v Compositor B proceeded to set the second half of sheet L and, while K inner was printing, took the skeleton generally used for the outer forme and imposed it around K outer for his press. His press then perfected sheet K. When Compositor A had finished setting signature L2, he imposed and sent to his press the pages for L inner (two set by each compositor) in the regular inner forme skeleton which had just been used for K inner. Meanwhile Compositor B set type through M1, but he imposed half-sheet M for his press to print in the skeleton from K outer that his press had just released. In making this imposition he removed running title II and the furniture for that page, not needed for sheet M, and fixed them around L4v, which he had recently set and placed on the imposing stone. When Compositor A came to impose L outer in the skeleton that his press had been using for L inner, he found L4v with a running title already in position. He therefore removed from his skeleton the now unneeded running title VII.

Such a schedule should not be construed as exact, but it explains with some margin for error the time sequence of the order of the formes through the press.[15] It also confirms the presence of a second compositor in certain specific pages at the close of the quarto.

One other curiosity of this quarto may be mentioned. It appears

that Compositor A made a preliminary estimate of the length of his copy on the basis of 37 linear spaces to the page.[16] He was consistent in setting this number of lines from A3v to H4v with these exceptions: 36 spaces on B4, C4, and G2; 38 spaces on E2v.

At the beginning of Sheet I, Compositor A readjusted his practice so as to set 38 lines to the page, saving thereby an amount of space equal to 27 lines of text at the maximum. In sheets I and K he was faithful to his new count, varying only once by setting 37 spaces on page K2. Compositor B, as has been noted, set 39 lines on K3v in order to separate the stage direction from the catchword. Copy was cast off after sheet K with some care so that Compositor B could set the second half of sheet L beginning with L3. Compositor A set 38 lines for L1 and L2 and 37 lines for L1v and L2v; Compositor B set 39 lines for L3 (in order to begin L3v with the stage direction) and 38 lines for L3v, L4, and L4v. The two full pages of sheet M are set 37 lines.

II. The Nature of the Copy

Perhaps the most important question to be solved in any textual study of Romeo and Juliet is that of the contamination of the second quarto by the first quarto. Though the first quarto is an inferior text — generally thought to be a memorial reconstruction — in at least one passage it served as copy for the second quarto. Critics differ on the precise limits of this reprinted passage, but roughly it extends from I.ii.52 to I.iii.35, some ninety lines.[17] There are also indications of further contamination elsewhere in the quarto, but these sections are in even greater dispute. The precise textual inter-relationship of the two quartos and of the authoritative manuscript behind Q2 has never been explained thoroughly and convincingly. One sort of evidence that has received only cursory attention is that derived from speech prefixes, yet such evidence offers additional information of some value that bears on the question.[18]

We are fortunate in having in Romeo and Juliet the reprinted

passage already mentioned in which all authorities agree that Q2 was set from Q1. From the 33 prefixes in these reprinted lines that we may regard as forming a "control passage," we may draw some conclusions as to the practice of Compositor A (who set this section) when he is following a known printed copy.[19] The prefixes are as follows:                                    

Q1	Q2 (variants only)	Q1	Q2
1. Romeo		18. Ser
2. Ben		19. Ben
3.Romeo		20. Ro
4. Ben		21. Ben
5. Rom		22. Rom	Ro
6. Ser		23. VVife
7. Rom		24. Nurce	Nurse
8. Ser		25. Iuliet
9. Rom		26. Nurce	Nur
10. Seru	Ser	27. Iul	Iuli
11.Rom		28. VV	Wife
12. Ser		29. Nurce	Nurse
13. Ro		30. VVife
14. Ser		31. Nnrce	Nurse
15. Ro		32. VVife
16. Ser		33. Nurce	Nurse
17. Ro

In this group of 33 prefixes, nine in Q2 have been altered from their originals in Q1. (The purely typographical differences between Q1 VV/Q2 W and Q1 R/Q2 R are of no consequence.) Of these nine alterations, four are spelling changes, unimportant in this connection, "Nurce" to "Nurse" (#24, 29, 31, and 33; #31 involves a double change). The remaining five (#10, 22, 26, 27, and 28) constitute changes in length. But in spite of these five changes in length, the Q2 prefixes give the clear impression that the compositor is faithful to his copy. A profitable insight into the compositor's general practice is provided by an examination of these five variants; it will be found that all of the changes have been made to secure uniformity and consistency.

Prefix #10. In the reprinted passage the prefix for the Servant appears seven times. On six occasions the Q2 compositor follows his copy in the spelling "Ser." On the seventh occasion (#10) when Q1 has "Seru", the Q2 compositor changes the spelling to "Ser", thus following his copy faithfully and consistently but not slavishly. The

change tells us only that "Ser" is the form preferred by the compositor in the interests of consistency, a fact which is readily apparent. It is therefore for our purposes a non-significant change. It represents only a normalization to a consistent practice.

We may now apply the principle of consistency in the speech-prefix pattern for the Servant to the rest of the quarto. Outside of the reprinted passage the prefix for the Servant occurs eight times. On seven occasions the Q2 spelling is "Ser"; the single variant occurs on page B3 immediately before the reprinted passage and is spelled "Seru." It thus appears that Compositor A's consistent spelling "Ser" in the reprinted passage is also his preferred spelling. To secure this spelling he has even altered his copy spelling on one occasion (#10). Fourteen times in the play out of fifteen "Ser" is used; the single aberrant form "Seru" may then reasonably require some explanation.

There are four possible explanations for this single variant spelling on B3. The first is that this is a fortuitous spelling; yet we have not found that this compositor indulges in random setting of this prefix; on the contrary, he sets it in a strictly consistent form. The second is that this "Seru" is so set because it is the first appearance of this speaker; but this theory seems weak considering the commonness of the character and the fact that "Seru" is little better than "Ser" in representing "Seruant" (Compositor A is by no means consistent in spelling out speakers' names on their first entrances). The third possibility is that "Seru" is taken from the stage direction; but here the direction has "Clowne", and this possibility is ruled out. The fourth possibility is that "Seru", "Serua", or "Seruant" stood in the copy and that the compositor is following such a spelling. This possibility we believe most readily explains the variant. But the form in Q1 at this point is "Ser", and it does not seem likely from the compositor's observed practice that he would change his preferred spelling when he finds it in his copy to an unpreferred spelling. We conclude then that the copy at this point was not Q1. This hypothesis is confirmed by the fact that the speech of the Servant is too much altered to have been taken from Q1 and must have been set from some sort of manuscript source. We infer that the prefix was set from the same copy (I.ii.38).

The example of the prefix for the Servant is a slight one, but it illustrates the principles of the investigation. To repeat: the change Q1 Seru / Q2 Ser (#10) in the reprinted passage is a non-significant change because it represents only the normalization to the preferred practice of Compositor A; the usage "Seru" on B3 is a significant usage because it represents a variant from the discovered practice. The

hypothesis that "Seru" is present in the copy at this point and influences the compositor is confirmed by the general admission of critics that this speech is taken from manuscript and not from Q1.

Prefix #28. The case of the prefixes for the Wife is similar. The compositor's preferred practice in the reprinted passage is the spelling "Wife." Accordingly he normalizes the Q1 "VV" to this consistent form. In the rest of the play the prefix occurs four times (we exclude for the moment all other titles to this part): twice set by Compositor A, and twice set by Compositor B on L4v (the latter settings do not concern us here). Compositor A's settings are on signatures A4 where the form is "Wife" and on I1 where the form is "Wi." The question is then raised, of six settings by Compositor A why is one irregular? We conclude as before that the Q2 compositor is following his copy, since we have not found that this compositor introduces innovations in speech-prefix form. As the first quarto has the prefix "Mo" at this point, we believe that the copy was not Q1 (III.v.176), for it seems unlikely that an editor or annotator of that quarto would have introduced the aberrant form into that text (which has some form of Mother in this scene) when preparing copy for the press. Furthermore, at this point in the text the two quartos are sufficiently near to have made the insertion of a transcribed slip unnecessary.

Prefix #26. Of the five prefixes for the Nurse in the reprinted passage, this one is uniquely abbreviated to "Nur." It is therefore in the passage an aberrant, but in the entire book this form of the prefix ("Nur" or "Nur") is overwhelmingly favored by the compositor (in 78 out of 93 uses).[20] This change is therefore non-significant; but variations from this norm should then be distinctive. Of the fifteen aberrant forms in the 93 occurrences of the prefix, ten are "Nurse" or "Nurse" and five are "Nurs." Of the ten prefixes spelled out "Nurse", four are in the reprinted passage (#24, 29, 31, and 33) and six in the section immediately following (sigs. B4v and C1). Since there are as many as ten of these spelled-out forms we can not positively say that they constitute an aberrant spelling, but we believe that they do represent an unpreferred spelling found in the original.[21] They seem to be, if not preferred, at least acceptable to the compositor. The five other spellings, "Nurs", that occur on signatures C4v and D1, do constitute

significant variants. These variants occur in the following pattern:                  

Q1	Q2
a. Nurse	Nur
b. Nurse	Nurs
c. Nur	Nurs
d. Nur	Nur
e. Nur	Nur
f. Nur	Nurs
g. Nurse	Nurs
h. Nurse	Nurs

In this sequence of prefixes the five Q2 "Nurs" variants are represented in Q1 by three spelled-out forms "Nurse" (#b, g, h) and two short forms "Nur" (#c, f). Since both the Q1 forms are used elsewhere by the Q2 compositor there seems but one reasonable explanation why Q2 should have the innovation "Nurs" five times: the compositor is reacting in his characteristic manner to his copy, following it. We conclude therefore that in these lines the first quarto was not the copy for Q2 (I.v.114, 131, 138, 144, 145).[22]

Prefix #22. Of the 170 prefixes for the part of Romeo, the form "Ro" is used 130 times; the form "Rom" is used 29 times[23] and the spelled-out "Romeo" 11 times. Since the two short forms are both common, we hesitate to draw any conclusions from them, but we must observe that there appears to be little correlation between the quartos in their use.[24] The use of the spelled-out form, however, seems to be

significant. Of the eleven instances of the prefix "Romeo", two occur in the reprinted passage, two occur on B1 at the first entrance of the speaker where Q1 has the same form (as does the manuscript [see below, footnote 27]), and the remaining seven occur on pages B1v (2), B2, C1, C2v, C3, and H3. Since at these seven points Q1 uses one of the Q2 compositor's preferred short forms, we conclude that the presence of the unpreferred form in Q2 is due to the influence of a copy other than Q1 (I.i.177, 191, 214; I.iv. 1, 95; I.v.42 [catchword]; III.v. 36).

Prefix #27. The case for Juliet is very similar. There are 116 prefixes for her part in the section of the quarto set by Compositor A (three more — "Iuli" — in the stint of Compositor B). Of these 116 the preferred form "Iu" is used 95 times; the form "Iuli" is used 19 times and the spelled-out form "Iuliet" is used twice. The change (#27) from "Iul" to "Iuli" is non-significant; the prefix "Iul" is never used in Q2. The spelled-out form is a significant usage because it is an aberrant form. It appears first in the reprinted passage (#25) and again on B4v. Since it is copied in its first appearance we conclude that it is copied also in its second, but as Q1 has here "Iul" the copy at this point cannot have been the first quarto (I.iii. 66).[25]

From this examination of the practice of Compositor A in the reprinted passage the following observations seem valid:

• 1. Whereas Compositor A has his personal preferences in the spelling of speech prefixes, he is greatly influenced by his copy and follows it exactly 28 times out of 33 (discounting the minor c/s shift in the prefixes for the Nurse).
• 2. When Compositor A does vary from his copy he does so in order to normalize the Q1 form to his own preference.
• 3. Compositor A in these lines never introduces an innovation in the speech prefixes. And on the basis of these observations the following extensions to the rest of the play seem valid also:
• 4. When a speech-prefix spelling preferred by Compositor A

  ---

  122

  
  
  
  occurs in Q1 and also in Q2, no conclusions can be drawn as to the nature of the copy, since there is not necessarily a relationship.
• 5. When an unpreferred speech-prefix spelling occurs in Q2 and the preferred or acceptable spelling is found in Q1, then the Q2 spelling must reflect a source other than Q1, since it cannot be accepted that Compositor A would change a preferred Q1 prefix to an unpreferred innovation.
• 6. When an unpreferred spelling occurs in Q2 and that spelling is found also in Q1, the presumption is that Q2 was following Q1 copy. This occurs so rarely, however, that it appears to be purely coincidental, and instead of confirming the assumption of Q1 copy rather weakens it.

Since the speech-prefix patterns of Q1 and Q2 vary independently (except in the reprinted passage) throughout the whole of the quarto to a comparable degree, we are constrained to accept the theory of an independent authoritative manuscript as copy for the second quarto with, as Dr. Hosley has termed it, "compositor's consultation" of the first quarto. Our confidence in this theory is strengthened when we consider that nowhere in Q2 save in the reprinted passage is there a sequence of speech prefixes of any considerable length that has any significant correlation with the corresponding passage in Q1. Furthermore, we believe that Compositor A by his intelligent fidelity and consistency reveals himself as a reasonable craftsman by no means so stupid or slovenly as some of his critics would have him.

These principles may be applied to the prefixes for the other characters of the play. Those for Benvolio offer interesting evidence. Of the 62 occurrences of this speaker, the form "Ben" is used 54 times; the aberrant forms "Benu", "Benuo", and "Benuol" are used respectively four, three, and one times. As most of these aberrants occur in the early speeches of the part, it may be argued that these variants indicate the compositor's attempts to find the form that suits him. But "Ben" is still used five times before signature B1v, as many times as the other forms put together and is therefore from the beginning the preferred spelling. The sequence of Benvolio's first prefixes is as follows:

                       

Q1	Q2
a. [Omit]	Benuo	A3v
b. [Omit]	Benuo	A4
c. Benuo	Ben	A4v
d. Ben	Benuo
e. Ben	Ben	B1
Q1	Q2
f. [Omit]	Ben	B1
g. Ben	Benu
h. Benuo	Benuol
i.Ben	Ben
j. Ben	Ben

The first two Q2 prefixes, "Benuo" (#a, b), must derive from the manuscript since the first quarto lacks the lines.[26] So indeed must #d "Benuo", for here Q1 has "Ben", the preferred practice of the compositor of Q2 and already used by him once. In the light of these assignments, we submit that #c "Ben" is also from manuscript, since if the compositor had been following Q1 he would probably have set the then normal "Benuo" at this point. These four lines we therefore ascribe to manuscript copy (I.i.71, 75, 113, 125). "Benu" and "Benuol" (#g, h) are innovations; they do not occur in Q1 here which has "Ben" and "Benuo", forms already set thrice by the Q2 compositor.[27] We conclude that these also derive from manuscript (I.i.162, 166). Two more uses of "Benu" on B1v would not seem to have been set from the Q1 "Ben" which is the preferred spelling (I.i.189, 190), and the presence of "Benu" once again on C1v (I.iv.33) where Q1 omits the speech indicates that it was a common spelling in the manuscript.

Mercutio speaks 64 times. Of these occurrences the form "Mer" is used 59 times;[28] the aberrant form "Mercu" is used three times, and the aberrants "Horatio" and "M" once apiece. The three aberrant forms "Mercu" do not appear in the first quarto which has at these points the preferred spelling "Mer"; we conclude therefore that these three lines were set from manuscript (I.iv.13; II.iv.13; III.i.45). "Horatio" and "M" would seem to be special cases.[29]

Paris speaks 23 times. Of the prefixes, thirteen occur in the preferred form "Par"; aberrant forms also used are "Pa" (7 times), "Paris" (2), and "Pari" (1). None of these variants appears at the corresponding point in Q1. We conclude therefore that these forms were found in a manuscript (I.ii.12; III.iv.8, 19, 29; IV.i.19, 21, 24, 32, 35; V.iii.49). The five settings of "Pa" on signature I2v (IV.i) are singularly impressive, because Q1 has consistently here the preferred spelling "Par."

The name of the Friar in Q1 is spelled "Laurence" throughout the text, and in the brief scene with Friar John the prefix is abbreviated "Laur." In Q2 the name is spelled "Lawrence" in the text, and the

prefix is abbreviated "Law" in that scene. We believe that the spelling difference of the four prefixes is traceable to a manuscript source for the scene (V.ii.2, 13, 17, 23).

Of Compositor A's five settings for Peter, four are abbreviated "Pet"; one is spelled out. The aberrant spelled-out form occurs on the first entrance of the speaker. It is omitted in Q1 and therefore must derive from manuscript (II.iv.111) (cf. note 33).

Prefixes as seldom used as those for Mountague, Mountague's Wife, and for other less important characters can tell us nothing, unless as in the case of Sampson and Gregorie they indicate of necessity a source other than the first quarto. The prefixes for the Prince have no distinctive variants.

Capulet's Wife has 44 speeches assigned to her in Compositor A's share of the quarto by this variety of prefixes: Capu. Wi (1), Ca. Wi (1), Wife (5), Wi (1), Old La (6), La (14), Mo (12), M (4). In the face of such diversity it would be rash to choose one spelling as a preferred spelling. What does appear from this haphazard diversity is that the compositor is following his copy with the same faithfulness that has been observed in the reprinted passage, for judging from his observed practices it is impossible to conceive of him as introducing these inconsistencies on his own initiative. Of the 32 times the various Q2 prefixes have possible sources in Q1, only four times is there demonstrable influence (in the reprinted passage) and twelve times is there only arguable influence where Q1 has "Moth" and Q2 "M" (H4 (2), H4v, K2) or "Mo" (H4, I1v, I4v (4), K2 (2)). but sixteen times the possibility of influence is eliminated by a generic change in the prefix:

           

Q1	Q2
Wife	Old La	B4v(3), C1
Moth	La	H3v(5), H4, H4v, K1v(2)
Mo	Wi	I1
Mo	Ca. Wi	G1
M	Capu. Wi	F4v

We conclude that these sixteen instances indicate a manuscript source at these points (I.iii.63, 69, 77, 96; III.i.151, 181; III.v.65, 69, 70, 79, 81, 88, 140, 176; IV.iv.1, 11).

On the basis of this evidence we can now assign a page of some importance, B4v, immediately following the reprinted passage, as having been set from manuscript throughout. On this page there are five prefixes for Capulet's Wife, five for the Nurse, and two for Juliet. Of the five prefixes for the Wife, all spelled "Old La", three diverge from

the Q1 copy and two are not in Q1; all concur in pointing to a manuscript. These "Old La" prefixes occur on B4v; in the reprinted passage on B4 the character is regularly "Wife". The coincidence of this change of prefix form with the end of the reprinted passage, a bibliographical not a textual point in the quarto, is almost mandatory evidence for a manuscript source resuming at this point.[30] The presence of this manuscript is demonstrated yet more convincingly by the evidence of the unique "Iuliet" in the center of the page, as already discussed. This manuscript seems still to be in use on page C1 at I.iii.96 at the prefix "Old La" and at line 100 at the speech prefix for the Servant, for Q2 here has "Ser" which cannot be a copy of Q1 "Clowne." The manuscript continues to the end of the scene, supplying the final couplet omitted in Q1.[31]

The character of Capulet has the same sort of variety in speech prefixes: Capu (11), Cap (5), Ca (16), 1. Capu (3), Capel (1), Fat (1), Fa (11). Again it is risky to pick a preferred spelling, but as before certain generic changes are conspicuous:

     

Q1	Q2
Cap	Fa	I1 (4), K2 (2), K2v
Capo	Fa	I4v(2)

Amid a confusion of different spellings in both quartos (complicated by the presence of at least three compositors in Q1 who spell "Capulet" differently)[32] these generic variants indicate distinctly a manuscript source at these points (III.v. 161, 171, 174, 177; IV.ii.37, 39; IV.v.22, 25, 34). It is, further, at least surprising if Q1 had been copy for Q2 that the prefix "Capo", which occurs fifteen times in Q1, and the spelling "Capolet", are not inadvertently copied once in Q2.

Other equally valid though less clearly defined generic changes exist in the prefixes for the attendants in the play. There is no certain policy in setting these prefixes; in addition to "Ser" there are "Man", "Boy", "Pa[ge]", "Clowne", "Peter," and "Balt[hasar]" in unpredictable disorder. No system is apparent, but the following generic changes seem

to be of the same type as those for Capulet and Capulet's wife:            

	Q1	Q2
	Boy	Pa	L1v
	Clowne	Ser	C1
	Balt	Man	K4(2)
	Ser	Fel	K1v(2)
and	Apo	Poti	L1(2)

Such changes do not appear to us to be the results of any compositorial action, and we attribute them to the influence of a manuscript at these points (I.iii.100; IV.iv.15, 18; V.i.17, 27, 75, 77; V.iii.10). The change of Q1 "Balt" to Q2 "Pet" (L2 (2)) is a substantive variant and erroneous, but we conceive the error to be the author's, and not an editor's, and assign the variants to manuscript influence (V.iii.40, 43).[33]

Thus far in this paper we have examined the speech prefixes in the second quarto in order to determine the preferred spelling of Compositor A and to explain the variants from that preferred usage. No variant can be attributed to a source in the first quarto, and no positive links between the two quartos in speech prefixes outside of the reprinted passage have been found. The only possible exception to this situation exists in the case of Tybalt. The prefixes for Tybalt occur seventeen times in seven different forms. It seems to us unwise to attempt to draw any conclusions from such disparate evidence, yet if our method is applied to this character, we find the following variants from the preferred "Tib" (7 times) that do have corresponding forms in Q1:

     

Q1	Q2
Ti	Ti	C4
Tyb	Tyb	F3(4)

No preferred usage for the names of attendants has been noted, but these prefixes may represent copy influence:      

Q1	Q2
Balth	Balth	M1v
Boy	Boy	M1v

Two spelled-out "Romeo" prefixes on B1 are found in both quartos. It is conceivable to be sure that in these instances there is Q1 influence on Q2, but such a theory would be more acceptable if there were more than nine instances of the type. We prefer to regard these as coincidental; at best they can but serve to confirm the theory of compositor's consultation.

Though there are doubtless exceptions to the theory of speechprefix practice here advanced, the principle involved seems to be sound. It seems furthermore to point quite clearly to an independent manuscript as copy for the second quarto. At least the validity of this proposition is sustained by the overwhelming number of variants that do not originate in Q1 when compared with that almost negligible number that may derive from it. On the basis of the evidence already presented, and on the observation of Compositor A's fidelity to known copy, we argue for the theory of the independent manuscript, for we doubt that an editor or an annotator of the first quarto engaged with any reasonable or consistent attention to his editing would have made such unimportant changes in the speech prefixes. It is not credible to us that an annotator would make the useless change of one form of a printed speech prefix to another form of the same prefix, or, in the case of the generic changes, that he would alter the Q1 consistent use of some form of Capulet to the Q2 Father, the Q1 consistent use of some form of Mother or Wife to the Q2 Old Lady, Lady, or Capulet's Wife.[34] Furthermore, it has been demonstrated that outside of the reprinted

passage on no occasion do aberrants from the compositor's established and preferred practice positively have their origins in Q1 copy. We must therefore conclude that in the lines indicated above, Compositor A of the second quarto was following a manuscript of independent authority for his copy and following it faithfully.

Notes