2 Henry VI.—This play exists in two forms: one in the 1623 Folio, hereafter for convenience called F.; the other in Quarto, entered S. R. 12th March 1594, hereafter called Q. It was published in 1594 as The First part of the Contention betwixt the two Famous Houses of York and Lancaster. This Quarto version is a mangled and probably surreptitious copy of the original play, greatly abbreviated for acting. The play as first written will be hereafter called O. But F. and O. are not identical, although in many parts O. was more like F. than Q. It will be convenient to enter on the proof that O. was revised and altered before beginning the discussion of the authorship of either version, which is the most difficult, if not the most important, problem in Shakespearian criticism.
In the Folio of 1623 a list is given of the principal actors in Shakespeare's plays. The method in which this list is arranged has never been pointed out. It is chronological. The first ten names are those of the original men actors when the Chamberlain's company was instituted in 1594; the next five were added not later than 1603; the next five (excepting Field, who is inserted here from his early connection with Underwood and Ostler) c. 1610; the final six after 1617. By a comparison of this list with the names of the actors in The Seven Deadly Sins, originally acted before 1588, but the extant plot of which dates c. 1594, we shall get the evidence we want. The first seven names in the Folio list are (1.) W. Shakespeare, (2.) R. Burbadge, (3.) J. Hemmings, (4.) A. Phillips, (5.) W. Kempe, (6.) J. Pope, (7.) G. Bryan. The last five of these we know to have been members of Lord Strange's company in 1593. In the 7. D. S. we find neither Shakespeare nor Hemmings; but we do find (2.) R. Burbadge, (4.) Mr. Phillips, (5.) Will Foole, (6.) Mr. Pope, (7.) Mr. Bryan. It will be noticed that the prefix Mr. is confined to members of Lord Strange's company. Next in the Folio list come (8.) Henry Condell, (9.) William Sly, (10.) Richard Cowley. These appear in 7. D. S. as (8.) Harry, (9.) W. Sly, (10.) R. Cowley. At this point we are struck with the fact that Harry, Will, and Dick are names of three Cade conspirators in Q., and naturally try to see if the other names, Nick, Jack, Robin, Tom, and George, occur in 7. D. S. For it is certain that in very early plays up to the end of the sixteenth century it was frequently the case that the actors in plays are designated by their proper christian names. We do find (11.) Nick (i.e., Nicholas Tooley, a boy-actor in 1597, but a man c. 1610 in the Folio of 1623), (12.) John Duke, (13.) Robert Pallant, (14.) Thomas Goodall; but George, i.e., G. Peele, is not there discoverable. I may notice that Duke and Pallant, like Beeston, all three of whom left the Chamberlain's men for the Earl of Derby's in 1599, are excluded from the Folio list. On turning to another play, Sir Thomas More, c. 1596, the only other one that can give us similar information on the same scale, I find (8.) Harry, (13.) Robin, (14.) T. Goodall, (15.) Kit (i.e., Christopher Beeston), and two boys, (16.) Ned and (17.) a second Robin, i.e., Robert Gough, who occurs in the Folio list as a man c. 1617. In the 7. D. S. these latter correspond to (15.) Kitt, (16.) Ned, (17.) R. Go. In Sir T. More there are two other names of this kind, Giles and Rafe. Of Giles nothing more is known, but Rafe Raye is mentioned in Henslowe's Diary as a Chamberlain's man in 1594. A further examination of older plays leads to little additional information; but what is to be found all confirms the opinion that I had formed (as will be seen), on other grounds, that 2 Henry VI. was written for the Queen's men. Thus in plays known to have belonged to that company, I find in The Famous Victories, (12.) John, (13.) Robin, (14.) Tom, (16.) Ned and Lawrence; in Orlando, (14.) Tom and Rafe (Raye); in Friar Bacon, (10.) Dick, (14.) Tom; and in James IV., Andrew. There is no Andrew in our lists, but one occurs in Much Ado About Nothing, iv. 2, 1597-8, in place of Kempe: apparently a remnant of the older form of Love's Labour's Won before Kempe undertook the part. But our list of the 7. D. S. is not yet exhausted: (18.) Sander (a boy-player, but the same as Alexander Cooke, a man in 1603 in the Folio list), (19.) T. Belt, and (20.) Will (another boy), occur in The Taming of a Shrew, 1588. Of (21.) Vincent, nothing is known; but (22.) J. Sinkler acted with Gabriel (Spenser) and Humfrey (Jeffes) in 3 Henry VI., which belonged to Pembroke's company. Now as the last two, with Antony Jeffes and Robert Shaw, appear in Henslowe's Diary for the first time immediately after the partial breaking up of Pembroke's company and their juncture with the Admiral's in October 1597, it is morally certain that Sinkler had gone to the Chamberlain's, and Spenser Shaw and the two Jeffes to the Admiral's, at or before that date. I feel, therefore, justified in concluding that the 7. D. S. gives us a nearly complete list of the Chamberlain's actors, formed of Lord Strange's players as a nucleus; such of the Queen's men as joined them in 1591-2, when they obtained many Queen's plays (see p. 108), and such of Pembroke's as joined them in 1594, when they obtained Pembroke's plays (see p. 21). I have omitted only one name, and the absolute coincidence of nearly every one of the rest with the lists obtained from other sources is too remarkable to be the mere effect of accident: in fact, the chances are many millions to one against this being the case. The one name omitted is (23.) John Holland. This name occurs nowhere else to my knowledge, but in the 7. D. S. plot and 2 Henry VI., Act iv. in the Folio, where he replaces Nick of the Quarto. There can be no doubt of this being an actor's name; and its occurrence shows at once that the Cade part of the play was revised, and that the revision was probably made after 1594. Had it been earlier, there would have been two Johns in the company, Duke and Holland, and Duke would not have been called simply Jack.
If the above conclusions are well founded, 2 Henry VI. was originally written for the Queen's men as a continuation of 1 Henry VI., and, like the latter-mentioned play, passed into the hands of Lord Strange's men in 1591-2, but was not, like it, then revised; or it may, like George a Greene, have passed to Sussex' men; from them, like Titus Andronicus, to Pembroke's; and thence to the Chamberlain's. It is noticeable that although published in Quarto by the same person, Millington, who published 3 Henry VI. as the True Tragedy of Richard Duke of York in 1595, he put no name of acting company on the former play, as he did that of Pembroke's on the latter. This distinctly shows that the original companies for whom these plays were written were not identical, and that that of 2 Henry VI. was probably unknown to Millington. As to the authorship of 2 Henry VI., it will be well to make F. the basis of investigation, always having in mind the possibility of passages having been inserted by the ultimate reviser. The corruption and omission in Q. caused by the shortening for stage purposes have been so great, that the usual plan of beginning with Q. becomes altogether misleading. The example of 1 Henry VI. induces me to attach great weight to the chronological arrangement of the historical facts. Henry's marriage in 1445 forms the subject of i. 1, evidently written by Greene originally. The word "alderliefest" in 1. 28 should specially be noted: it is used by Greene in his Mourning Garment, and "aldertruest" in his James IV. Such words are not found in Marlowe, Peele, Lodge, or Shakespeare; yet here one occurs in a passage found in F. but not in Q., plainly indicating omission in Q., not addition in F. The next portion, i. 2-ii. 4, is concerned with the banishment of the Duchess of Gloster, 1441, and the story of Saunder Simcox, 1441, with which is incorporated the accusation of the armourer for high treason, 1446. This part (except i. 3. 45-103) is mainly by George Peele, but much altered in the F. revision. Peele his mark, "sandy plains," occurs in i. 4. 39. The Simcox anecdote, however, ii. 1. 59-153, which is quite unconnected with the rest of the play, is more like Kyd's work than Peele's, and may have been written by him. The exceptional bit, i. 3. 45-103, to the conversation in which no historical date can be assigned, is manifest Marlowe; a preparation for iii. 1-iv. 1, which is beyond question by him. The events in this section are (iii. 1a) the accusation and (iii. 2) murder of Gloster in 1447; (iii. 3) the banishment of Suffolk, 1447; (iii. 3) the death of Winchester in 1447; (iii. 1b) the Irish insurrection in 1449; and, finally, (iv. 1) the death of Suffolk in 1450. These scenes are the salt of the play. The opening lines of iv. 1, the description in iii. 2. 160, &c., the awful pathos of the death of Winchester, are from the same hand as the end of Doctor Faustus. The differences of Q. and F. in this portion are mostly due to omissions in Q.: iii. 3, for instance, could not have been left in the state in which Q. has it by the meanest of the authors of the play: it is cut down by some illiterate actor. That revision there has been is, however, plain from the singular circumstance that in iii. 2 Elianor is given for Margaret as the Queen's name. This is probably due to Marlowe's almost simultaneous work on the older John, in which Queen Elianor is a prominent character. It would seem that the revisor missed this scene, although correcting Margaret properly in the others. It is no printer's error; for in l. 26 we have "Nell," for which some modern editors euphoniously substitute "Meg." The rest of the play, iv. 2-v. 3, is by one hand, and that hand Lodge's. The notion that Greene wrote it arises from want of discriminating Greene's work from Lodge's in The Looking-glass for London, all the better part of which is by Lodge. I fear that those who underrate the powers of this elegant and (in his own line) powerful writer estimate him by his earliest dramatic effort, Marius and Sylla. He should be read in his Glaucus and Rosalynde; and his evident wish to avoid being known as a dramatic writer should be taken into account. That he did continue to write plays for many years, I have no doubt, but the evidence is too extensive to be given here. This part of the play includes Cade's insurrection, 1450, and the battle of St. Albans, 1455.
As regards the date, &c., of revision, see under the next play.
3 Henry VI. is of very different character from the two preceding plays. If read in the F. version, no change of authorship is perceptible; all is consistent; and if the Q. version had not come down to us, no one would have suspected a second author. It is plainly by Marlowe, but the Marlowe of Edward II., not of Faustus, later in date than 2 Henry VI. F. is nearly if not quite identical with the original play. Q. is not, as in the case of the preceding play, an abridgment for the stage made by the actors, but one made for the same purpose, carefully and accurately, apparently by the author himself. The reason for this difference in the treatment of the plays is manifest. 3 Henry VI. was, as we know from the title-page, acted by Pembroke's men, and F. is printed from a prompter's copy, in which the names of Gabriel [Spenser], Humphrey [Jeffes], and [John] Sinkler appear in the stage directions; and they were actors for that company. There is not a particle of evidence that this stage copy was ever altered in any way after the Chamberlain's company acquired it. A careful examination of such passages as ii. 5, the stronghold of the revision theory, shows too much coincidence between Q. and F. for any likelihood of rewriting having taken place, except by way of abridgment in Q. But in 2 Henry VI. things are quite different: the Greene and Marlowe parts are merely abridged in Q., and the Peele a good deal revised in F. as well as abridged in Q.; but the Lodge part at the end is absolutely rewritten in the St. Alban's battle, and the very names of the actors are changed in the Cade insurrection. Who could have done this but Shakespeare? Here, and here only, can we find an explanation of the inclusion of these plays in the Folio edition of his works in 1623. In my opinion the history of the plays is this: About 1588-9, Marlowe plotted, and, in conjunction with Kyd (or Greene), Peele, and Lodge, wrote 1 Henry VI. for the Queen's men. About 1589 the same authors wrote 2 Henry VI.; in that year I have ascertained that Marlowe left the Queen's men, and in 1590 joined Pembroke's, for whom he alone wrote 3 Henry VI. In 1591-2 the Queen's men were in distress, and sold, among other plays, 1 Henry VI. to Lord Strange's men, who produced it in 1592 with Shakespeare's Talbot additions as a new play. In the autumn of that year or in 1593-4, when the companies travelled on account of the plague, they cut down their plays for country representation; among others, 2 Henry VI. (altered by some illiterate) and 3 Henry VI. (abridged by Marlowe himself). On this point compare the parallel instances of abridged plays, Hamlet, Orlando, and The Guise. In May 1593 2 Henry VI. passed to the Sussex' men with Leir, &c., when the Queen's men broke up; in February 1594 with Andronicus to Pembroke's; in April, when Pembroke's company partly dissolved, all three plays were reunited in the hands of the Chamberlain's men; and for them 2 Henry VI. was, c. 1600, after Lodge had retired, remodelled by Shakespeare, and 3 Henry VI. corrected—the other authors, Peele, Marlowe, (Kyd?), and Greene, having died before 1598. Meanwhile Millington published 2 Henry VI. Q. as York and Lancaster, and 3 Henry VI. Q. as Richard Duke of York, these abridged copies having become useless to Pembroke's men on the ceasing of the plague and of their travels.
I have not noticed here the many parallel passages from the works of Marlowe and others which confirm the assignment of authorship now advocated. It would be out of all proportion to give them here unless imperfectly: the reader will find some in Dyce's Marlowe, and more in my edition of Edward II. Nor have I noticed the schoolboy interpretation that explains "their" in Henry V., Epil. l. 13, as referring to 2 and 3 Henry VI.: "their," more Shakespeariano, like "they" in the previous line, refers in form to the "many" of l. 12, but in meaning to the actors of 1 Henry VI., in which play, and not in 3 Henry VI., the loss of France is treated of. It is also most unlikely that the 1600 edition of The Duke of York should have been issued as played by Pembroke's servants if the play had been previously acted by the Chamberlain's. Compare the parallel case of Andronicus. Miss Lee's statement, "Greene wrote, Nash tells us," more than four others "for Lord Pembroke's company," is absolutely without foundation. Nash says "the company" (Apology, 1593), and evidently alludes to the Queen's men, for whom Orlando, Bacon, Selimus, and The Looking-glass were written. In fact, Greene's only known connection with any other company was his fraudulent selling of Orlando a second time to the Admiral's. Marlowe, and he alone, is known as a writer for Pembroke's: Kyd may have been, however, and in my opinion was, a contributor to their stage.
Richard III. is closely connected with 3 Henry VI., and written with direct reference to it. In i. 2. 158, iv. 2. 98, iv. 4. 275, scenes in that play are plainly alluded to. Nor is it possible, if the two plays be read in immediate sequence, to avoid the feeling that they have a common authorship. On the other hand, a closer analysis shows that in Richard the Latin quotations, classical allusions, and peculiar animal similes which are characteristic of Henry have entirely disappeared. There are also discrepancies, such as Gray's fighting for the Lancastrians, i. 3. 130, whereas in 3 Henry VI., iii. 2. 2, he is represented as a Yorkist, which shows a different hand in the two plays. Richard III. has always been regarded as entirely Shakespeare's, and its likeness to 3 Henry VI. has more than anything else kept alive the untenable belief that this last-named play was also, in part or wholly, written by our greatest dramatist. Yet the unlikeness of Richard III. to the other historical plays of Shakespeare, and the impracticability of finding a definite position for it, metrically or æsthetically, in any chronological arrangement, have made themselves felt. Even cautious Mr. Halliwell says, "There are slight traces of an older play to be observed, passages which belong to an inferior hand;" and again, "To the circumstance of an anterior work having been used do we owe some of its weakness and excessively turbulent character" (Outlines, 94). A careful examination of the editions will be found to confirm and extend this conclusion. The 1597 Quarto Q1, which is evidently an abridged version made for the stage, and which no doubt was the version acted during nearly all Elizabeth's reign, differs from the Folio in a way not to be paralleled in any other Shakespearian play. Minute alterations have been made in almost every speech, in a fashion which could not have been customary with him who uttered his thoughts so easily as scarcely to make a blot (i.e. alteration) in his papers. The question of anteriority of the Q. and F. versions has been hotly debated on æsthetic grounds; but the mere expurgation of oaths and metrical emendations in F. are enough to show that it is the later version, probably made c. 1602; while the fact that it was preferred by the editors of the 1623 Folio shows that they considered it the authentic copy of Shakespeare's work. In other instances, Macbeth, The Tempest, &c., they have indeed given us abridged editions; but there is neither proof nor likelihood that any other were accessible. We do not know what original copies were destroyed in the Globe fire of 1613, and should be thankful for such versions as we have, which were probably the acting versions used at Blackfriars. But in this case the editors had at hand the Quartos, and unless they thought the Folio more authentic, I cannot see why they preferred it. Furthermore, the F. version appears to have been defective in some places; for v. 3. 50, end of play, and iii. 1. 17-165, are certainly printed from Q3 (1602). This has been controverted, but on very insufficient grounds. Now directly we compare the Folio and Quarto versions, we meet with evidence that alteration and correction have been largely used in both of them. For instance, Derby is found as a character in the play in i. 1, ii. 1, 2, iv. 5, v. 5, in both versions; in iii. 1. 2, iv. 1, v. 2, he is called Stanley. This shows correction by a second hand. In iv. 1, while Stanley has been inserted in the text, Derby remains in the prefixes; v. 3 is only partially corrected, and both names occur. The names were not used indifferently, for in iv. 2, 4, we find Stanley in F. but Derby in Q. This shows a progressive correction in which Q. precedes F. It may be noticed that Darby is the original author's spelling. In like manner, Gloster, the original prefix, has in i. 1, 2, 3, ii. 1. 2, iii. 4, 5, 7, been replaced in F. by Richard, but in iii. 1, in the part printed from Q3, and there only, Gloster remains. So again Margaret is indicated in the older version by Qu. Mar., Qu. M., &c., but never Mar., as in F. iv. 4. In F. i. 3 we find by side of Mar. a remainder of the older form in Q. M. This is not an exhaustive statement, but sufficient I think to show that alterations were made, as I suggest. There can be little doubt that in this, as in John, Shakespeare derived his plot and part of his text from an anterior play, the difference in the two cases being that in Richard III. he adopted much more of his predecessor's text. I believe that the anterior play was Marlowe's, partly written for Lord Strange's company in 1593, but left unfinished at Marlowe's death, and completed and altered by Shakespeare in 1594. It was no doubt on the stage when, on 19th June 1594, the older play on Richard III., "with the conjunction of the two Houses of Lancaster and York," was entered S. R. That was acted by the Queen's players. The unhistorical but grandly classical conception of Margaret, the Cassandra prophetess, the Helen-Ate of the House of Lancaster, which binds the whole tetralogy into one work, is evidently due to Marlowe, and the consummate skill with which he has fused the heterogeneous contributions of his coadjutors in the two earlier Henry VI. plays is no less worthy of admiration. I do not think it possible to separate Marlowe's work from Shakespeare's in this play—it is worked in with too cunning a hand; but wherever we find Darby, Qu. M., Glo., &c., we may be sure that some of his handiwork is left. Could any critic, if the older John were destroyed, tell us which lines had been adopted in the later play? Nor can I enter, unless in a special monograph, on the relations of the Quartos to each other. The question is of no importance, and I need only say that the usual corruptions take place from Q1 to Q5, and that in Q6 (1622) many readings are found agreeing with F. which are not in the other Quartos. The same phenomenon is observed in the 1619 edition of The Whole Contention, and far too much has been made of it. It merely indicates correction by attendance at the theatre and picking up a few words during the action. The only Quartos deserving special notice are Q1, as containing Shakespeare's first "additions," and Q3, as having been used in printing part of F. I do not think the allusion in Weever's Epigrams, written 1595-6, is to this play. It may be so.
Titus Andronicus.—That this play is not by Shakespeare is pretty certain from internal evidence. The Latin quotations, classical allusions, use of pour as prefix in iv. 1, manner of versification, and above all the introduction of rape as a subject for the stage, would be sufficient to disprove his authorship. Fortunately we know that it was produced by the Earl of Sussex' men, 23d January 1594, and Shakespeare belonged then to Derby's (Lord Strange's). It was afterwards, on the breaking up of that company, acted by Pembroke's and Derby's before 16th April, when Lord Derby died. Enlargement in the Folio or abridgment in the Quarto, 1600 (we have no copy extant of the first edition, entered S. R. February 1594), appears in iii. 2, found in F., not in Q., and there is a distinct continuity between Acts i. and ii.; at the end of Act i. we have "manet Moore," not Exeunt simply. Whether this play got into the Folio by some confusion with Titus and Vespasian, played by Lord Strange's men 11th April 1592, which was, as we know from a German version extant, written on the same subject, and in which Shakespeare may have had some share, we cannot tell; but it was certainly played and revised (there was another edition in 1611), while the other play has perished. That it was written by Marlowe I incline to think. What other mind but the author of The Jew of Malta could have conceived Aaron the Moor? Mr. Dyce has warned us against attributing too many plays to the short career of Marlowe, but he did not consider that Marlowe probably wrote two plays a year from 1587-1593, and that we have only at present seven acknowledged as his. Those now attributed to him, in whole or part, by me will raise the number to a baker's dozen; but in some of these, as the older John and 1 and 2 Henry VI., his share was comparatively slight. Nevertheless, I think the opinion that Kyd wrote this play of Andronicus worth the examination, although, with such evidence as has yet been adduced, Marlowe has certainly the better claim. Shakespeare probably never touched this play unless by inserting iii. 2, which is possible.
Edward III. The Shakespearian part of this play, i. 3, ii. 1. 2 (beginning at "What, are the stealing foxes"), which contains lines from the then unpublished Sonnets, ii. 1. 10, 450, and an allusion to the recently published Lucreece, ii. 2. 194, was clearly acted in 1594, after 9th May, when Lucreece was entered on S. R. Edward III. was entered 1st December 1595. This love-story part is from Painter's Palace of Pleasure. The original play is by Marlowe, and was acted in 1590 and is thus alluded to in Greene's Never too Late, c. December in that year: "Why, Roscius, art thou proud with Æsop's crow, being prankt with the glory of others' feathers? Of thyself thou canst say nothing; and if the Cobler hath taught thee to say Ave Cæsar, disdain not thy tutor because thou pratest in a king's chamber." Ave Cæsar occurs in i. 1. 164, but not in any other play of this date have I been able to find it. There are many similarities between the Marlowe part of this play and Henry VI. As the Roscius in Greene's pamphlet was the player who had interpreted the puppets for seven years, who induced Greene to write for the stage, and had himself written The Moral of Man's Wit and The Dialogue of Dives, there can be no doubt that Robert Wilson is Roscius, and that he was an actor in Edward III. in 1590. It was acted by Pembroke's company, and must have been acquired by Lord Strange's men with the other Pembroke plays in 1594.