George Peele’s 1588 play, The Battle of Alcazar, is reflected in many surviving documents that might reveal much about the play’s performance history, the history of the Lord Admiral’s acting company, and Elizabethan stage practices. The documents may be variously interpreted, however, and in consequence the story they tell about the events which underlie them has remained elusive. This essay re-examines the documents and attempts to reconstruct the probable course of underlying events. Alcazar was revived by the Lord Admiral’s in 1594, again in 1598, and lastly in 1601. The play’s backstage plot, laying out a scheme for the performances, was prepared for the 1598 revival. Those conclusions entail the probable resolution of many subsidiary questions concerning the company’s history and Elizabethan stage practices. Philip Henslowe’s inventories of company costumes and properties, for example, were prepared in 1599, not 1598 as often thought, and concerned items kept in storage at the Rose theatre. Edward Alleyn, the company’s leading actor, continued to participate in the company’s affairs during his temporary retirement from acting between 1597 and 1600. Henslowe’s inventories suggest that severed heads presented onstage specifically resembled the heads of the characters who had been beheaded, and of the actors who had performed the roles. The inventories are in that respect supported by a range of other evidence.
Scholarly controversies surround George Peele’s play, The Battle of Alcazar, conflicts which, although perhaps less portentous than the historic battle itself, have been no less earnestly waged. A principal reason for those conflicts lies in the unusually rich number of documents that survive concerning the play: its printed text, a manuscript backstage plot laying out a scheme for the play’s performance, and many potentially relevant entries in Philip Henslowe’s records. Those documents lend themselves to varying interpretations, and a coherent resolution of the consequent difficulties has remained elusive. And yet the documents reflect an actual course of underlying events––those events should be capable of determination to a reasonable probability. Any such resolution would explain the controversial documents consistently inter se, and also locate the underlying events within the limits defined by other, uncontroverted, evidence.
Let us undertake that project here. If we can discover a consistent explanation for the documents, we would learn in the process a great deal about the performance practices of Elizabethan acting companies. How often, for example, did they revive their moderately successful plays? What objects represented the stage’s many severed heads? We would also learn a great deal about the history of a particular Elizabethan acting company, the Lord Admiral’s, with whose history Alcazar was intimately intertwined. Our inquiry will focus on four topics: 1) the date and auspices of five lists of Lord Admiral’s costumes and properties, inventory lists apparently prepared by Henslowe; 2) the likelihood that Alcazar appears in Henslowe’s Diary as ‘Mahomet’; 3) the date of Alcazar’s backstage plot; and 4) what Henslowe means when he records on the properties inventory list an ‘owld Mahametes head’. From those findings we shall extract in the conclusion a performance chronology for the play, a chronology unique for Elizabethan drama.
I. The Lord Admiral’s Inventory Lists
Five inventory lists were found by Edmond Malone among the manuscripts at Dulwich College. Malone printed the lists in 1790, but the originals are lost. Each list’s title specifies that the list concerns items belonging to the Lord Admiral’s company. Based on their provenance and spelling, the lists are generally attributed to Henslowe, an attribution which seems right, although the fifth list’s title was not set down by him.1 The titles are of some significance:
Although Henslowe evidently intended these inventories to be comprehensive in some respect, they obviously do not account for all the Lord Admiral’s properties and costumes.2 One reason for this incompleteness seems to be that the lists concern only properties and costumes kept in storage at the Rose.
The booke of the Inventary of the goods of my Lord Admeralles men, tacken the 10 of Marche in the yeare 1598. Gone and loste.
The Enventary of the Clownes Sewtes and Hermetes Sewtes, with dievers other sewtes, as followeth, 1598, the 10 of March.
The Enventary of all the apparell for my Lord Admiralles men, tacken the 10 of marche 1598.—Leaft above in the tier-house in the cheast.
The Enventary tacken of all the properties for my Lord Admeralles men, the 10 of Marche 1598.
The Enventorey of all the aparell of the Lord Admeralles men, taken the 13th of Marche 1598, as followeth.
Acting companies presumably kept in storage somewhere their properties, costumes, and play-books not currently in use, and their practices in that regard seem to have varied. The Lord Chamberlain’s/King’s company seem to have stored such material off-site. When the Globe burned down in 1613, ‘nothing did perish but wood and straw, and a few forsaken cloaks’. However, the Lord Admiral’s company and their successors apparently stored at least some such material inside the theatres in which they performed. When the Fortune burned down in 1621, ‘all their apparell & play-bookes [were] lost, wherby those poore companions are quite undone’.3
Henslowe’s lists here apparently concern material kept in storage at the Rose. The second list so indicates, as it specifies that it concerns items which have been ‘Leaft above in the tier-house in the cheast’. The tiring-house was where the company kept its costumes and properties currently in use, so the apparel ‘left above’ there in the chest were not currently in use but were intended for storage––by inference in a space below the tiring-house. The fourth inventory list, moreover, concerns stage properties and contains many items associated with specific plays, yet lacks properties that would be used in every play, such as chairs and tables. Those presumably were always kept in the tiring-house for current use.Henslowe divides the inventoried items into descriptive categories—suits, apparel, and properties. Such a division was possible for items kept in storage, but would have been difficult for items kept in the tiring-house. Costumes and properties there presumably were arranged not by descriptive category but to promote efficient manipulation during performances.
Henslowe’s probable purpose in taking these inventories also suggests that they concern items kept in storage at the Rose. The lists could have found their way to Dulwich College only if Henslowe had made them for his own use, not for the use of the actors. He was a pawnbroker, for whom records of collateral served important functions. He had by this time begun to loan the acting company money and their apparel and properties stored in his building effectively functioned as collateral for the debt. He could lock up the property pending issuance of a Chancery Court writ allowing him to take it. Henslowe had good reasons to record for this purpose only items kept in storage, for items in the tiring-house were essential to the acting company’s then current performances. If Henslowe were to seize them, the company could no longer perform. Henslowe would lose his own share of the gallery takings from the forgone performances, and the company would be deprived of any means to make payments to him on their debt. So Henslowe needed no inventory of items in the tiring-house; he knew that he would not seize those things in any event.
Henslowe’s five inventory lists must be distinguished from a sixth list found by Malone and printed together with Henslowe’s five. The sixth list is titled:
This list apparently was made late in August 1598. It includes many plays that, as we know from Henslowe’s Diary, were acquired by the Lord Admiral’s company from December 1597 through 10 August 1598. Those were among the ‘suche bookes as belong to the Stocke’. Other play-books on the list were purchased by the list’s author, ‘such as I have bought’. Two reasons tell us that the author was not Henslowe, but Edward Alleyn. First, the list includes the play of ‘Vayvode’. Neither Henslowe nor the Lord Admiral’s owned Vayvode at the time the list was made. Alleyn did. He sold ‘the playe of vayvod’ to the company 5 months later. Second, between the two plausible candidates whose papers ended up at Dulwich College, a telltale mannerism reveals Alleyn, not Henslowe. The writer of the title uses an ordinal for the day in his date, ‘the 3d of March 1598’. Perhaps more than 1,000 dates handwritten by Henslowe survive and among them, as far as I can tell, Henslowe never employs an ordinal for the day. But Alleyn does, for example, in a memorandum ‘this 8th of december 1597’.4
A Note of all suche bookes as belong to the Stocke, and such as I have bought since the 3d of March 1598.
The distinction between Henslowe’s five inventory lists and Alleyn’s list of books becomes important when we seek to determine when Henslowe prepared his five lists. Henslowe dates his lists March 1598, but that could be our year 1599 if Henslowe was using the official calendar. Scholars often have thought that the date was 1598 on the assumption that Henslowe’s lists were linked to Alleyn’s list. Without such a link, the date remains ambiguous. Henslowe’s choices between the modern and official calendars were as erratic as his spelling. His proclivities in these 2 years offer no help. In our year 1598 Henslowe dated all his March Diary entries ‘1598’. And again in our year 1599 he likewise dated all his March entries ‘1598’. He used the official calendar in that month even past the change of the year on 25 March. His account entries for the year begin on ‘the 4 of Jenewary 1598’, and continue through ‘the 31 of marche 1598’. Only in April does Henslowe switch years, his first such entry being for ‘the 7 of aprell 1599’.5
Thus, to determine whether the lists were made in March 1598 or March 1599 we must examine the inventories’ contents. They indicate that the year very probably was 1599. The property inventory contains an entry for ‘j helmet with a dragon’. That surely was a helmet worn by Arthur Pendragon—King Arthur—in relief of the Saxons’ siege at Bath. In Historia Regum Britanniae Geoffrey of Monmouth reports that Arthur put on a helmet ‘simulacro draconis insculptam’, on which was carved the likeness of a dragon. The helmet in storage thus presumably was worn in performances of a play reflected in Henslowe’s Diary as ‘the Lyfe of Artur king of England’. Hathaway finished writing that play in April and the Lord Admiral’s were preparing to perform it in May 1598. The helmet thus probably was fabricated for performances that began 2 months after 10 March 1598. It could only have been listed in a March inventory if the year were 1599.6
The property list also includes a ‘shelde, with iij lyones’. That shield no doubt was used in the play Henslowe records as ‘the funerall of Richard cordelion’. Richard died while laying siege to a castle in Limousin. He received an arrow wound near the neck, a wound which caused the onset of fatal gangrene. The play presumably dressed this end as a more heroic form of combat, one requiring a shield. In any event, Richard’s shield probably was used often in the play, during both the siege and the funeral. Battles and ceremonies were the principal occasions on stage for shields. Wilson, Chettle, Munday and Drayton were writing The Funeral of Richard Coeur de Lion in June 1598. If the 3-lion shield was acquired for that play, the inventory must have been made in March 1599. The shield probably was not acquired for Anthony Munday’s two Robin Hood plays, The Downfall and The Death, both of Robert Earl of Huntingdon. Richard appears once at the end of Downfall and twice at the beginning of Death. The occasions in those plays are informal gatherings, neither combat nor formal ceremony, and Munday’s stage directions in Downfall specify that Richard is ‘clad in green’.7
Munday’s Robin Hood plays and another play, Black Joan, do, however, lead again to a date of 1599 for the inventory lists. The company obtained licences to perform both Robin Hood plays on 28 March 1598. Because they could not perform the plays until they obtained the licences, they probably were not thinking about acquiring costumes and properties until sometime approaching that date. ‘Blacke Jonne’ is named on Alleyn’s list of play-books. Its play-book, like Vayvode’s, probably belonged to Alleyn, not the company. The play is not named in Henslowe’s records of plays the company performed daily from May 1594 to November 1597, nor is it included among the great many plays the acquisitions of which by the company were thereafter financed by loans from Henslowe. Alleyn acquired his play-books between 3 March, as his title indicates, and late August, 1598, when he made his list. Henslowe’s inventory lists contain many items for the Robin Hood plays and ‘j frame for the heading in Black Jone’. None of those items probably had been acquired by 10 March 1598, and certainly would not then have been placed in storage.
The consequent inference that the inventories were taken in 1599 is further supported by a grammatical choice Henslowe made. When Henslowe associates items in his lists with particular plays, he usually employs the preposition ‘for’, a preposition which could apply to past, current or future performances. But four times Henslowe employs the preposition ‘in’, a choice which implies performances which have already occurred. Thus the lists include a ‘dragon in fostes’, a ‘whell and frame in the Sege of London’, a ‘frame for the heading in Black Jone’, and ‘the fryers trusse in Roben Hoode’. Doctor Faustus and The Siege of London by any reckoning had long since been performed. We may reasonably conclude that the Robin Hood plays and Black Joan had similarly been performed when Henslowe took his inventories.8
II. Alcazar in Henslowe’s Diary
The circumstances strongly suggest that performances of The Battle of Alcazar by the Lord Admiral’s company are reflected in Henslowe’s Diary. Henslowe appears to record there every play the Lord Admiral’s performed at the Rose from May 1594 to November 1597. He also extended hundreds of loans to the company to meet their play-related expenses, and from October 1597 until it ends in March 1603, the Diary usually identifies the relevant plays. Loans from Henslowe surely were not necessary for every Lord Admiral’s play or performance run. Yet his lending activity was sufficiently extensive for a fair chance to exist that even a revival after 1597 is reflected in Henslowe’s Diary.
Within the time frames covered by the Diary, the Lord Admiral’s almost certainly performed Alcazar. The backstage plot anticipated their performance of the play, and was prepared sometime between 1597 and 1601. Jonson in Poetaster elaborately parodies Edward Alleyn’s performances in the play’s role of Mahamet.9Poetaster was written in 1601, and because Jonson expected his audiences to recognize the parodied performances, the play probably had been performed since Alleyn joined the reorganized Lord Admiral’s in 1594. Most significantly, the title page of the play’s 1594 quarto print advises viewers that the books offer The Battle of Alcazar ‘As it was sundrie times plaid by the Lord high Admirall his servants’. This statement was intended to induce viewers to come to the bookseller’s shop to buy the books. Offprints of the title page evidently were distributed or posted around London as advertisements. The page informs viewers that the books are ‘Imprinted at London […] for Richard Bankworth, and are to be solde at his shoppe in Pouls Churchyard at the signe of the Sunne’.
Few patrons would have been induced to appear at Bankworth’s shop unless the performances to which the title page refers had been given since the Lord Admiral’s were reorganized in May 1594. Alleyn had left the company in 1591. They are not known to have performed again in London until the reorganisation. In the intervening years, two remnant companies apparently toured, one in the English provinces, the other on the Continent. Prospective London book buyers may not even have known about any performances of the play by the touring companies, and, as the company had recently begun performing in London for the first time since 1591, Bankworth presumably was seeking to capitalize on those recent performances. With an imprint date of 1594, Bankworth’s quarto could have been printed as late as 24 March 1595, the last day of 1594 on the official calendar. The official calendar remained at this time the operating calendar of the Stationers’ Company, as reference to their records will show. The performances to which the title page refers, then, probably were given sometime between May 1594 and March 1595. At that time Henslowe was recording every play the Lord Admiral’s performed in London.
If Henslowe’s Diary reflects Alcazar, however, Henslowe must have used a pseudonym, as the play is never recorded there under its formal name. Among plays the Diary does record, the dates alone will draw our attention to one Henslowe calls, in various spellings, ‘Mahomet’. The company performed ‘Mahomet’ eight times from August 1594 to February 1595. Those dates fall within the range suggested by the 1594 quarto’s title page. The run apparently was a revival, moreover, as a run of Alcazar would have been. Henslowe does not mark the first performance as ‘ne’, as he customarily did for new or substantially revised plays. The Diary reflects another performance run of ‘Mahomet’ beginning in August 1601. Henslowe made several loans to meet expenses for a new production.10 That run began soon after Edward Alleyn returned to the company, following his temporary retirement begun late in 1597. Several other plays in which we know or might plausibly conjecture that Alleyn starred were revived at about the same time.
‘Mahomet’ aptly substitutes for Alcazar’s formal name. Mahamet is the villain-hero of Alcazar, the leading character, the actuator of all the play’s events, the only character larger than life. No other Lord Admiral’s play named by Henslowe may plausibly be identified with Alcazar. Few commentators deny that ‘Mahomet’ might be Alcazar, but some express doubt based on the variance from the play’s formal title. Those doubts are too punctilious. Henslowe need not have concerned himself with a play’s formal title. His Diary is a business records book kept for his own use. It was enough there that he knew a play by the name he called it. Many plays of the period, moreover, are formally titled by their leading characters’ names. Others were given descriptive titles and alternately called by the names of their leading characters. Henslowe and nearly everyone else, for example, called The Spanish Tragedy by a version of Hieronimo’s name, ‘Jeronimo’. Henslowe invariably calls Munday’s Robert Earl of Huntingdon plays ‘Robin’ or “Robert Hood’. He often identifies The Massacre at Paris by a short form for the Duke of Guise, ‘the Guise’, and The Spanish Comedy as ‘Don Horatio’.
Henslowe also had a particular reason to call Alcazar ‘Mahomet’. Edward Alleyn, who performed the role, was Henslowe’s son-in-law and business partner. They probably spoke often about the acting company’s business affairs. Those discussions would naturally lead Henslowe to think of the company’s plays in terms of the roles Alleyn was performing in them. Thus the Duke of Guise also was performed by Alleyn, and we may reasonably believe that Hieronimo was as well.11 Henslowe did not especially prize consistency, but he was in many respects a creature of habit, and a pattern may readily be perceived here:
|Play||Leading Role||Performed by||Henslowe name|
|The Massacre at Paris||Duke of Guise||Alleyn||‘Guise’|
|The Spanish Tragedy||Hieronimo||Alleyn (probably)||‘Jeronimo’|
|The Battle of Alcazar||Mahamet||Alleyn||( )|
|Play||Leading Role||Performed by||Henslowe name|
|The Massacre at Paris||Duke of Guise||Alleyn||‘Guise’|
|The Spanish Tragedy||Hieronimo||Alleyn (probably)||‘Jeronimo’|
|The Battle of Alcazar||Mahamet||Alleyn||( )|
Many other plays are similarly titled in the Diary by the roles that Alleyn is known to have performed, although in most, perhaps all, such cases, the roles also were incorporated into the formal names of the plays. Those plays include Orlando (presumably Orlando Furioso), The Jew (presumably of Malta), Tamar Cham, Cutlack, Tamburlaine and Doctor Faustus.
One more fact, pointed out by Martin Wiggins, provides additional support. Henslowe’s entry for a loan on 4 August 1601 records ‘the mackynge of crownes […] for mahewmett’. Alcazar rather unusually features multiple crowns. It concerns a battle which famously resulted, as its quarto’s first page reminds us, in ‘the death of three Kings’. In its final dumb show, just before the climactic battle, three crowns are hung ‘upon a tree’, apparently a tree that revolves over a fire. Each crown in turn falls off as the Presenter comments: ‘Fire, fire about the axiltree of heaven,/Whoorles round, and […] In fatall houre consumes these fatall crownes. […] Downe fals the diademe of Portugall, […] The crownes of Barbary and kindomes fall’. So the manufacture or remanufacture of multiple crowns for ‘mahewmett’ strongly suggests again that ‘mahewmett’ was Alcazar.12
III. The Date of the Backstage Plot
If we have interpreted the evidence correctly, the Lord Admiral’s began a performance run of Alcazar, starring Edward Alleyn as Mahamet, around the end of August 1601. Alleyn is named as the performer of Mahamet in Alcazar’s backstage plot, so it would be natural to date the plot to 1601. Wiggins makes a very reasonable case for precisely that date. Yet nagging difficulties remain. When we consider those difficulties, and re-examine evidence adduced by Greg, we should arrive at a different date, sometime between March 1598 and July 1600. Further exploration of the evidence will suggest a fairly precise time, around late November 1598.13
The foremost nagging difficulty is simple. If the backstage plot was prepared around August 1601, the company surely knew whether Alleyn would in fact perform. Yet the plot seems uncertain as to that very point. It assigns the role of Mahamet to Alleyn well enough, but it fails to assign any actor for the almost equally important role of King Sebastian. It also fails to assign two of the company’s principal actors, William Birde and John Singer, any role. That puzzle may be explained if the company were in doubt over whether Alleyn would in fact perform Mahamet. If, as matters turned out, Alleyn did perform the role, either Birde or Singer could perform Sebastian. But if Alleyn did not in fact perform Mahamet, either Birde or Singer remained available for that part, while the other could perform Sebastian. The company would have been uncertain of Alleyn’s availability if they anticipated his return at some point during his temporary retirement, but he had not yet firmly committed. Such a situation would not be surprising. Alleyn remained active in the company’s affairs during his temporary retirement, as we have seen, buying play-books for their use and sometimes selling them those books. He had left and returned to the company before, in 1591 and 1594. He would return again in 1600 after his temporary retirement.14
Evidence for a date between March 1598 and July 1600 is found in the presence on the plot of Richard Alleyn and Thomas Hunt. Sometime between those two dates Richard Alleyn and probably Hunt were sharers in the company. They were not sharers on 28 December 1597. On or soon after that date, Henslowe recopied his prior entries for loans to the company going back to 11 October. The last such entry is dated 28 December, and the next, original, entry is 5 January, so Henslowe recopied the list sometime between those dates. Henslowe titled the recopied list: ‘A Juste a cownt of All suche money as I have layd owt for my lord admeralles players […] whose names ar as foloweth borne gabrell shaw Jonnes downten Jube towne synger & the ij geffes’. That is, William Birde, Gabriel Spencer, Robert Shaw, Richard Jones, Thomas Downton, Edward Juby, Thomas Towne, John Singer, and Humphrey and Antony Jeffes. Henslowe does not name Richard Alleyn or Hunt, so they were not sharers on 28 December.15
Nor were they sharers on the following 8 March. Sometime between that date and 13 March, Henslowe summed the total of his loans to the company from the list of recopied entries, plus loans made from 5 January to 8 March. Below the sum, the sharers sign their own names, ‘JSinger’, ‘Thomas Downton’, ‘william Birde’, ‘Robt Shaa’, ‘Richard Jones’, ‘Gabriell Spenser’, ‘Thomas towne’, and ‘Humfry Jeffes’. To the right of that column are added the signatures of ‘Charles massye’ and ‘Samuell Rowlye’. Below all these signatures Henslowe notes: ‘Thes men dothe aknowlege this deat to be dewe by them by seatynge of ther handes to yette’.16
The sharers’ acknowledgement of their debt to Henslowe served a specific contractual purpose. Sharers in a business enterprise were tenants in common, each holding an interest in the enterprise only as long as he remained a sharer. The company’s debts were owed jointly and severally, that is, both by the whole and by each sharer individually. When an individual sharer left the company, a new company was formed and the debts of the prior company became due and payable. If a debt holder agreed, however, the new company formed by the remaining sharers could adopt as their own the prior company’s debt. Such a substituted contract of debt was and is called a novation.
The proceedings here apparently were a novation, as we may infer from Henslowe’s statement of their purpose. The remaining sharers acknowledge the debt to be owed ‘by them’. The Lord Admiral’s sharer who had left was Edward Alleyn. He had ‘leafte [p]laynge’ by 29 December 1597, and Henslowe did not include him among the sharers he named at about the same time at the top of his recopied list of company debts.17 Alleyn thus had stopped performing around 11 October, evidently, but did not formally resign his sharer status until 5 months later. During those 5 months the company accumulated a £46 debt to Henslowe. When Alleyn did formally resign around 8 March, a novation was necessary in order to avoid payment of that debt, payment for which Alleyn would have remained, perhaps unfairly, partly responsible.
Richard Alleyn and Thomas Hunt are not among the novation’s signers, so they still were not sharers on 8 March. Henslowe surely wanted every sharer to sign. Two sharers at the time, Antony Jeffes and Edward Juby, did not sign, but their situations may be distinguished. The ‘ii geffes’ each held a ‘hallffe sheare’, with Humphrey evidently the dominant of the two halves.18 Henslowe probably believed, probably correctly, that in those circumstances Humphrey’s signature sufficed for the pair. Henslowe would have wanted Edward Juby’s signature, but Juby apparently was not available when the other sharers signed and Henslowe neglected to follow up. Henslowe would not, however, have been especially concerned about the absence of Juby’s signature. Signatures merely evidenced assent, and Juby was a longstanding member of the company unlikely to contest his assent. Richard Alleyn and Hunt met no such criterion.
Richard Alleyn became a sharer, however, within a year. He authorized loans by Henslowe to the company on 17 January and 7 April 1599, and again on 6 May 1600. Only sharers could bind the company to contracts, so Henslowe very probably made loans to the company only on the authority of its sharers. The 7 April loan, moreover, covers Alleyn’s and Towne’s expenses ‘to go to the corte’, a function the company presumably delegated only to sharers.19 Hunt also probably was a sharer around the same time. He is identified as a ‘master’ on the backstage plot for Troilus and Cressida, which was prepared around May 1599. The term ‘master’, or rather its abbreviation ‘Mr’, as Greg showed, was consistently applied by the makers of Lord Admiral’s plots only to actors who were sharers in the company. The only clear exception was that the term was never applied to the two Jeffes, presumably because they each held only a half share.20 This specialized use of the term ‘master’ conformed to its legal meaning, designating an individual as one who has authority. Masters in trade companies, for example, were authorized to set up shop for themselves, hire journeymen and train apprentices. In an enterprise, only sharers possessed equivalent authority.
But the tenures of Richard Alleyn and Hunt as sharers were short-lived. The company and Henslowe entered another novation on 10 July 1600. Their accumulated debt had reached the impressive sum of £300. Henslowe records the sum and Shaw writes next: ‘Whiche some of three hundred poundes we whose names are here under written, doe acknowledge our dewe debt & doe promyse payment’. This statement is followed by the individual signatures of Singer, Downton, Shaw, Towne, Birde, Juby, Jones, Massey, Rowley, and both Jeffes. Gabriel Spencer had been killed 2 years before. Otherwise all the company’s known prior sharers signed; except, that is, Richard Alleyn and Thomas Hunt. They apparently were no longer sharers. Because the proceedings were a novation, moreover, some sharer or sharers very probably had left the company. The departing sharers by default were Alleyn and Hunt. Alleyn never again appears in the Diary and Hunt never appears there at all.21
These facts seem to fix Alcazar’s backstage plot between the two dates within which Richard Alleyn, and probably Thomas Hunt, were Lord Admiral’s sharers. The plot names as performers both ‘mr Rich: Allen’ and ‘mr Hunt’. The use of the term ‘master’ denominates them as sharers on the date the plot was prepared. All other actors similarly denominated on the plot are known to be sharers. Several other actors named on the plot are not known to be sharers and are not similarly denominated. Thus, we may reasonably conclude, the Battle of Alcazar plot was prepared sometime between 8 March 1598 and 10 July 1600.
That conclusion is not one hundred per cent certain, of course. Our information that Hunt was a sharer at this time comes solely from the scribe who prepared the backstage plots of both Troilus and Cressida and The Battle of Alcazar. Yet he is correct in ascribing sharer status to Richard Alleyn, if the two plots were prepared at about the same time. We cannot be completely certain that Henslowe did not overlook the absence of Alleyn and Hunt’s signatures on the 1600 novation. Yet some sharer or sharers very probably had left the company, and they are the only candidates. And we cannot be wholly certain that the scribe did not misapply the term ‘master’ in the Alcazar plot to Alleyn and Hunt. Yet they both apparently were sharers between 1598 and 1600, so we need not invent such an anomaly. It seems improbable, moreover, that the two lost their sharer status and nonetheless were hired by the company in 1601 to perform in Alcazar as employee actors.
Between March 1598 and July 1600 a more precise date for the plot, around late November 1598, is suggested by several circumstances. Those circumstances point to a belief by the company that Edward Alleyn might return to the company and to the stage. Alleyn had repaired with his wife to Sussex in September 1598. Henslowe wrote to him there about the killing on 22 September of Gabriel Spencer, who had become a leading member of the Lord Admiral’s. His sudden death must have caused the company serious casting problems. Henslowe implores Alleyn to return to London: ‘I wold fayne have alittell of your cownsell yf I cowld’, he says, regarding Spencer’s loss. Alleyn did return to London; he sold Henslowe copper lace on 8 November.22 Henslowe’s desire for Alleyn’s counsel no doubt was genuine, but we may detect in his appeal a more subtle motivation. The company had been suffering, as we may deduce from Henslowe’s accounts, from Alleyn’s absence on the stage. Henslowe and the company surely hoped that the casting problems created by Spencer’s death could be used as a pretext to persuade Alleyn to return.
Alleyn apparently did entertain this idea and the company anticipated his return. Those conclusions are suggested by two sets of possibly interrelated circumstances. Before Alleyn left performing, first he acquired ‘for playnge’ both a short velvet cloak embroidered with glass beads, and materials with which to make a velvet jerkin decorated with silver and gold-coloured copper lace. Henslowe later bought those items from Alleyn, the materials with which to make the jerkin still unassembled. By 28 November the company had arranged for the jerkin to be completed. Henslowe sold both the now completed jerkin and the short velvet cloak to them. The cloak and the unassembled materials had lain unused, thus, for a year. The company decided to purchase them and to order assembly of the jerkin, because they now anticipated a previously unanticipated use for them. That use presumably involved the role for which Alleyn had bought the items in the first place, a role which the company had not performed since Alleyn left.23
At about the same time, the company were preparing to perform a series of plays in at least one of which they surely hoped Alleyn would star. The plays concerned France’s multi-decade religious wars of the sixteenth century. The company seem to have hit upon the idea of a series only accidentally. At the end of September 1598 they bought from Dekker and Drayton ‘the firste syvell wares in france’. Note that the play was the ‘first’ of the civil wars, not the first ‘part’. From that beginning the company evidently conceived a series. In early November they bought from the same two playwrights ‘the second prte of the syvell wares of france’, the second ‘part’ that is, which they were preparing to perform late that month. They also commissioned Dekker and Drayton to write ‘the 3 prte of the syvell wares of france’, a play which the playwrights completed at the end of December.24
Any such series of plays, we may plausibly guess, would be built around a revival of Marlowe’s Massacre at Paris, which concerns a middle period in the same wars. Marlowe had written Massacre only 5 years earlier. Even dead he remained the company’s go-to house dramatist, whose plays were often and successfully revived. It seems improbable that the company would pay Dekker and Drayton to write a new play concerning the same material addressed by a Marlowe play to which they already had access. They had access to Massacre because the play was owned by Edward Alleyn.25 So we would have to imagine, with growing scepticism, that they commissioned a new play to duplicate a play owned by one of their most important sponsors.
Dekker and Drayton’s first play, we may guess from its title, covered the years 1540 to 1559 in the French wars. Massacre covers the years 1572 to 1589. So we may conjecture that the second part of ‘the Civil Wars’ covered 1560 to 1571, and the third part the years 1590 through to the Edict of Nantes in 1598 itself. The company orchestrated a similar arrangement half a year later. In April and May 1599 they bought two new plays by Dekker and Chettle, ‘Troyelles & cresseda’ and a play alternately called ‘orestes fures’ and ‘Agamemnon’. The first no doubt concerned the Trojan War’s middle years, the other its aftermath. Those new plays presumably were combined in a sequence with a revival of a play from 1596, ‘troye’. That play concerned the war’s climax, as we may infer from the presence of a ‘great horse with his leages’ in Henslowe’s March 1599 properties inventory.26
A revival of Massacre around early December 1598 seems to be confirmed by Henslowe’s Diary. Henslowe records two loans in late November to William Birde so that he could buy costume items ‘for the gwisse’. When, elsewhere in his Diary, Henslowe refers to ‘the gwisse’ and uses the phrase ‘for the gwisse’, he invariably means the play The Massacre at Paris, in which the leading character is the Duke of Guise. Now, Birde actually was preparing to perform the role of the Duke. Henslowe extended him a third loan for ‘sylke stockens to playe the gwisse in’. If the company had bought Dekker and Drayton’s plays with no thought of Marlowe’s play, one such play necessarily featured the Duke of Guise. So the possibility exists that all three entries concern a role Birde is to perform in a new play by Dekker and Drayton. But if one of Dekker and Drayton’s plays featured the Duke of Guise, Henslowe gives no hint of it. He clearly refers to Dekker and Drayton’s plays no fewer than nine times. He invariably uses ‘the Civil Wars of [or in] France’ formula, never ‘the Guise’.27
Sometime in October 1598, thus, the company probably conceived the idea to revive Massacre at Paris around early December. They would not have done so unless they expected Edward Alleyn’s cooperation and ideally his participation. Alleyn owned Massacre’s playbook, and ordinarily performed the Duke of Guise. So in November 1598 the company apparently were anticipating Alleyn’s return early the next month. He would wear the costumes for which he had bought the materials and he would play the Duke of Guise. In the event, the company were disappointed. Alleyn may have performed a few times, but the company knew that he was already prepared to desist. Henslowe made the loans to Birde so that he might put together costumes in which he, not Alleyn, would play the Duke of Guise. The company, not Alleyn, bought the cloak and now completed jerkin. They did so presumably because the role for which the items were to be worn was only uncertainly going to be performed by Alleyn.28
IV. The ‘owld Mahemetes head’
Additional evidence suggests that after it was revived beginning in December 1598, Alcazar was still being performed the following March. Both the 1594 quarto and the backstage plot call for an elaborate dumb show before the play’s fourth act. Mahamet, King Sebastian, the Duke of Avero, and Captain Stukley enter to a banquet; the figure of Death enters and presents to them three Furies in a pageant foreshadowing their deaths. The pageant is accompanied by a narrative in which the Presenter points out that Mahamet, as well as the others, awaits ‘bloud and death’. One Fury enters ‘with Dead mens heads in dishes’. In the context those heads must represent the heads of Mahamet and the other three banquet guests.29
With perhaps telling congruence, Henslowe’s March 1599 properties inventory contains an entry for an ‘owld Mahemetes head’. Henslowe’s intention when he wrote ‘owld’ is critical. He might have meant that the character ‘Mahemete’ was elderly, but that seems improbable: no elderly Mahomet appears in any surviving Lord Admiral’s play, nor in any known source for a lost Lord Admiral’s play. Indeed, elderly Moslem men are rarely encountered in all Elizabethan dramatic literature. Moslem men there are virile and vigorous. Henslowe is, on the other hand, describing as ‘owld’ an item kept in storage and not currently in use, so the greater probability is that he means the item has been replaced by a ‘new’ equivalent.
The Lord Admiral’s Battle of Alcazar thus required a property head representing that of Mahamet. An inventory of Lord Admiral’s properties in storage on 10 March 1599 listed a ‘Mahemetes head’. That head was an ‘owld’ one, probably meaning that it had been replaced. A new Mahamet’s head had been fabricated and, as it was not in storage, was currently being used. These circumstances explain themselves if the old head was based on Edward Alleyn’s. Alleyn had played Mahamet in Alcazar’s 1594 revival; Alcazar was now revived again, and Alleyn performed in that revival only briefly or not at all. A new head was necessary based on that of William Birde, whom the plot leaves available to play Mahamet and who substituted for Alleyn as the Duke of Guise. The new head remained in use on 10 March 1599 because the revival continued through that date. Birde was still fairly new to the company in December 1598, but Thomas Towne and Edward Juby, whom the backstage plot names as the performers of Stukley and Avero, and John Singer, whom it leaves to perform Sebastian, probably had been sharers since the company’s 1594 reorganisation.30 If they were reprising their prior roles, the 1594 heads for those characters could be used again.
In this explanation the heads in Death’s pageant were modelled on those of individual actors. That conclusion follows from Henslowe’s entry itself. The head in the entry is specifically that of Mahamet and thus may be distinguished from the heads of all other characters, whether performed by the same or by different actors. The head is an old one, and thus may be distinguished from its replacement, which represents the head of the same character, now performed by a different actor. Henslowe’s properties list also has a disembodied ‘Ierosses head’, which presumably resembled the head of the actor who performed Iris in the play with a ‘raynbowe’. Henslowe’s suits list contains ‘iiij Turckes hedes’. The entry doesn’t show whether those heads were mutually distinguishable, but it does show that they all could be distinguished from the heads of non-Turks.
This resolution consistently explains otherwise incoherent artefacts, but it will also inevitably provoke reservations which must be considered. Might not Henslowe have meant ‘head attire’, for example, when he said ‘head’? Probably not. ‘Mahemetes head’ and ‘Ierosses head’ are on the properties list and are unaccompanied there by any obvious item of apparel. Every other head on the list clearly is just a head, and its identity is similarly expressed as a possessive: ‘j bores heade’, ‘j bulles head’, ‘ii lyon heades’, ‘Serberosse iij heades’ and ‘Argosse heade’. The ‘Turckes hedes’ do appear on the suits list, but that list contains several properties.31
The suits list, to extend the point, contains many items that actually are identified as head attire: hats, hoods, and, most relevantly, ‘vj head-tiers’. So Henslowe did say head attire when that is what he meant. When Henslowe associates an item of apparel with a specific character, moreover, he often uses not a possessive but the objective, as in ‘green gown for Maryan’. He uses the objective the only time he apparently associates an item of head attire with a specific character, ‘j hatte for Robin Hoode’. If in this pattern Henslowe’s heads were actually head attires, we should have expected to see at least one entry expressed as ‘head for Mahemete’, ‘head for Ieross’ or ‘iiij hedes for Turckes’. Iris in legend, we may also note, wore a distinctive shimmering or multi-hued gown, but is not associated with any particular head attire.
Another reservation will concern whether patrons at the Rose could recognize the faces of the actors who performed there and, consequently, the faces on any heads representing those of the characters they performed. Several reasons suggest that the patrons could. Human ability to recognize familiar faces is primarily based on the structural characteristics of the faces themselves, rather than superficial embellishments like hairstyle, facial hair, and make-up. Recognition depends in part upon such obvious physical conditions as light and proximity, and also upon psychological factors such as familiarity and context.32
Those conditions favoured recognition of the actors. The Rose was an open theatre where plays began in daylight and usually ended in daylight, at least by twilight. The actors were surprisingly close even to the furthest spectator. For other theatres we have as yet only conjectural dimensions, but the Rose’s actual dimensions are revealed by the 1989 excavation of its foundations. Even after the theatre’s 1592 expansion, the distance between its centre stage and the farthest point in the auditorium was fewer than 20 yards, about 18 metres. The same Lord Admiral’s actors performed 6 days a week at the Rose in a rotating and constantly changing repertory of plays. Regular patrons saw them repeatedly, and the stage at the Rose was a context in which those patrons expected to see those actors. Many probably went to the Rose for that very purpose, knowing what role Edward Alleyn would perform even before they arrived. My own experience at the American Shakespeare Center (ASC) in Staunton, Virginia, suggests that it would be essentially impossible in these circumstances not to recognize the actors. The ASC comprises a stable acting company who perform changing and rotating plays in a replica Elizabethan theatre. Regular patrons almost universally recognize the individual actors. Elizabethan theatre patrons, Katherine Duncan-Jones reminds us in a recent Review of English Studies essay, similarly recognized at least one actor by the appearance of his head alone. Richard Tarlton elicited great laughter merely by peeping out from backstage.33
A third reservation might concern whether acting companies in general depicted on disembodied heads the faces of the relevant actors. Elizabethan plays are well populated by severed heads. In many cases the heads could have been contained in baskets or sacks, and thus have passed unseen by the spectators. In each such case the ‘head’ might only have been a pumpkin. But many play texts require exposed heads. If the acting companies were fully prepared to ask their audiences to accept a non-representational object as a character’s now severed head, they might still have held up a pumpkin. Or they might have offered a generic head, recognisable as human, but not as the individual whose head has been severed. Based on the evidence, however, the balance of the probabilities is that the heads resembled those of the actors who had portrayed the unfortunate characters.
Shakespeare discusses the very problem in Measure for Measure. Angelo has ordered the provost to deliver Claudio’s severed head and the disguised Duke asks the provost to substitute the head of Bernadine, a criminal also to be beheaded. The provost demurs. ‘Angelo’, he notes, ‘hath seen them both, and will discover the favour’. The Duke argues that ‘death’s a great disguiser’, that the provost might blur the difference further by shaving Bernadine’s head and tying his beard, and that if this fails he in his friar’s role will, ‘by the saint whom I profess’, plead the provost’s case. That oath surely rang hollow in Elizabethan protestant ears. And the provost, who is in a position to know better, remains unconvinced. Then another prisoner, Ragozine, dies unexpectedly. He was ‘a man of Claudio’s years; his beard and head just of his colour’, and his ‘visage […] more like to Claudio’. His head is substituted for Claudio’s, and when the Duke sees the changeling, he opines: ‘Convenient it is’. The head could be in a basket, unseen by the audience. If, however, they were shown Ragozine’s head when the provost announces ‘here is the head’, should they not have expected, after all the prior discussion, to see a head resembling that of Claudio?34
In Dekker’s Sir Thomas Wyatt the severed head unquestionably is exposed, and there again the text suggests that the head resembled that of the character. Lady Jane Grey has been onstage for some time when the headsman appears. Dekker then calls attention to her face. Her husband, Dudley, asks the headsman: ‘Hast thou the heart to kill a face so faire?’ The headsman returns, according to the stage directions, ‘with Janes head’. Dudley, about to be executed himself, observes:
The focus on her facial features and expression suggests that this was not a stock male head, but a head crafted specifically to represent a virtuous Lady Jane Grey. Similarly, in Peele’s Old Wives Tale, an old conjurer has exchanged appearances with Erestus, a comely young knight. The conjurer dies at the end of the play and his spells are released. His head is brought in:
|Eumenides:||How now Jack, what hast thou there?|
|Jack:||Mary maister, the head of the conjurer.|
|Eumenides:||Why Jack that is impossible, he was a young man.|
|Eumenides:||How now Jack, what hast thou there?|
|Jack:||Mary maister, the head of the conjurer.|
|Eumenides:||Why Jack that is impossible, he was a young man.|
Again, the dialogue suggests that the head resembled that of the spellbound Erestus, not that of a generic male.35
In two plays stage characters act on their beliefs that false heads are authentic. The Shah seeks to discover in The Travels of Three English Brothers, by Day, Rowley and Wilkins, whether his niece has illicitly loved Robert Sherley. He orders Sherley removed and his niece brought back. Then an officer brings in, according to the stage directions, ‘a counterfeit head like Sherley’s’. The niece is completely deceived by this ruse, kisses the head, condemns the beheading, and clarifies the relationship as one purely of mutual respect. In the B-Text of Marlowe’s Doctor Faustus, Benvolio, with two friends, plans to avenge an insult by killing Faustus. Then ‘Enter Faustus with the false head’, a shorter actor wearing a false head of Faustus, and one of the conspirators observes ‘see, he comes’. They wrestle Faustus down and Benvolio strikes the head off. One of the conspirators holds the head up and they discuss what to do with its beard, eyes and tongue; they are terrified when the headless Faustus then rises.36
As in the cited examples, most stage directions identify severed heads by name. Little purpose was served by that identification if the head brought in was always the same one. The authors of these directions, whether playwrights or book holders, presumably knew the practices of the companies for whom they were working. A head’s identity, as expressed in the stage directions, was especially important when the dialogue failed to name the head. In such cases the play depended upon the head itself to inform the audience. In 2 Henry VI, for example, Suffolk is beheaded, and after two intervening scenes and 202 lines the Queen enters ‘with Suffolk’s head’. She grieves silently and in asides over the head, commenting upon ‘this lovely face’, yet the head isn’t orally identified. If it were a merely generic head, spectators must have been puzzled.37
One play, The Revenger’s Tragedy, does raise doubts. Ambitioso and Supervacuo have conspired to have Lussurioso beheaded. They are discussing Ambitioso’s plan to spring their younger brother from prison when an officer arrives at the palace ‘to present you with the yet bleeding head’. The two brothers assume that the head is Lussurioso’s, but it turns out to be Junior’s. Could this be a macabre joke about the interchangeability of stage heads? Several facts indicate, instead, that the officer has brought the head in a sack. The brothers see the officer and speculate that he brings ‘desired news’, but they do not realize that he is carrying a head until he says so. A ‘yet bleeding’ bare head might have been difficult to rig; pig’s blood might easily have been put in a sack. When he learns of the switch Supervacuo blurts: ‘Villain, I’ll brain thee with it.’ He uses the head to pummel the fleeing officer, blows that could most easily be managed by grasping the top of a sack.38
A final reservation might concern the ability of acting companies to have heads fabricated to individual specifications. But there is no doubt that acting companies could access some source for multiple heads. Several plays in addition to Alcazar call for multiple bare heads on stage at once. Greene’s Alphonsus of Aragon requires three, each mounted at the corner of a canopy. 2 Henry VI requires at least two, when those of Saye and Crowmer are mounted on poles and made to kiss each other. The backstage plot of Tamar Cham calls for the heads of three noblemen rebels to be carried in sequentially, all three on stage at the scene’s end. The noblemen are not named in the plot but were performed by William Cartwright, Thomas Marbeck and William Parr. In these plays the different heads surely had varying features. The heads of Saye and Crowmer kissing each other would have been inexplicable if they looked alike, and the point in Tamar Cham of the heads’ sequential appearances would have been lost if they all looked the same. And if the heads had varying features, we might reasonably suppose that the features resembled those of the characters whose heads were represented, and of the actors who had performed the roles, Cartwright, Marbeck and Parr in the case of Tamar Cham.39
No course of historical events will be fully revealed in surviving documents, a point especially true of events surrounding Elizabethan dramas. Thus, despite the unusual number of documents related to Alcazar, uncertainties remain. A course of events did occur, nevertheless, and the documents allow reasonable inferences as to what those events were. If the inferences made here are correct, the documents offer a detailed chronology of one Elizabethan play’s performance history. We lack such a detailed chronology for any other period play. Alcazar probably was first performed beginning in 1589, after Peele composed it around late 1588. It was revived for a run of eight performances from August 1594 to February 1595, as we may infer from Henslowe’s Diary; it was revived again around December 1598 through at least March 1599, as suggested by the backstage plot and Henlowe’s properties inventory; and it was revived for the last time in the Elizabethan era beginning in August 1601, as indicated again by the Diary.
The evidence we have examined also hints at a previously unknown aspect in the biography of one of the period’s most prominent actors. Edward Alleyn may briefly have resumed acting with the Lord Admiral’s, in the middle of his temporary retirement, late in 1598. When in Poetaster Jonson elaborately parodies performances of Mahamet by Alleyn, he may be corroborating that hint. Jonson assumes that his audiences will remember those performances. Jonson cannot have written Poetaster after late August 1601, when Alcazar was once again revived––the chronology tells us so. Dekker wrote at least half of Satiromastix after he had seen Poetaster performed, and Satiromastix was registered on 11 November. Even so speedy a writer as Dekker needed several weeks to write half a play, especially a half as clever as his burlesque on Jonson. Jonson, by his own report, took 15 weeks to write Poetaster. Allowing for rehearsal, Jonson cannot have undertaken Poetaster later than June 1601, and probably undertook the play around January, as he has ‘a player’ refer to the prior season as ‘this winter’. Because Jonson believes his audiences will recognize performances of Mahamet by Alleyn, they may well have seen performances more recent than Alcazar’s prior revival ending in February 1595.40
A prior version of this article was reviewed by Lawrence Manley, Andrew Gurr, George W. Williams and Paul W. White, each of whose suggestions have enhanced the revision. Martin Wiggins generously provided extensive commentary and many suggestions which have profitably been incorporated. I am grateful to each, especially Dr Wiggins.