In the 1784–85 London theatrical season, there was a conjunction between one of the greatest tragedies ever written, Shakespeare's Macbeth, and three of the most dynamic and thrilling artists of the last decades of the eighteenth century: the dancer Geltruda Rossi (d. 1799), the actress Sarah Siddons (1775–1831), and the artist Henry Fuseli (Johann Heinrich Füssli, 1741–1825). Fuseli, engaged with Shakespearean subjects for much of his career, produced in 1783 his Lady Macbeth walking in Her Sleep (Photo 1), a work that circulated starting in January 1784 in the form of an engraving produced by the accomplished printmaker, John Raphael Smith (1752–1812). Sarah Siddons, who had played Lady Macbeth in the provinces, chose the play and the role for her benefit performance in February 1785 at Drury Lane. And for her benefit in March of the same year at the King's Theatre, Geltruda Rossi danced the role of Lady Macbeth in the new ballet of Macbeth. The three artists were thrown into figurative proximity when an anonymous commentator, on viewing Rossi's performance, wrote in the Morning Herald and Daily Advertiser of March 19, 1785: “Madame Rossi, in Lady Macbeth, impresses one more with the recollection of Fusili’s [sic] painting, than of Mrs Siddons’s representation—indeed comparison would be doing an injustice to our critical and admired English performer.” In uniting Rossi, Siddons, and Fuseli, the writer has suggested a particular interpretation of the role by Rossi, one that contrasted with Siddons's apparent conception of the character in 1785 and one that we can try and establish through Fuseli's picture. And while we know much about Siddons, there is little, if any, pictorial representation surviving of Rossi's performances, which makes the Morning Herald’s parallel worth exploring as an example of eighteenth-century London ballet d'action; it is her performance that is the focus of this article.

The story begins with the performances in 1785 of Lady Macbeth by Sarah Siddons, who would come to be considered one greatest actresses of the age; she had played the role in the provinces, but on February 2, performed it on the London stage for the first time. Macbeth was played by William Smith, Banquo by Thomas Hull, and Duncan by John Hayman Packer, with the three witches played by the male actors William Parsons, John Moody, and Robert Baddeley. To take Lady Macbeth onto the stage at Drury Lane was a formidable risk for Siddons, for the interpretation of the role by the much-loved Hannah Pritchard (1711–1768), performed opposite David Garrick's Macbeth, was still the one the public knew well and admired. It was a performance Fuseli had been inspired to capture in “Garrick and Mrs Pritchard as Macbeth and Lady Macbeth: After the Murder of King Duncan” in 1766. As it happens, for Siddons the 1785 performances were, according to the Gazetteer and New Daily Advertiser for February 3, 1785, largely a success:

Photo 1. Lady Macbeth. Henry Fuseli (1741–1825). Oil. Musée du Louvre.

At this stage, then, Siddons played the character in a noble style, with emphasis on her expressive countenance. A later assessment by the author of Siddons's biography in the 1911 Encyclopaedia Britannica suggested that

As this suggests, the role became synonymous with Siddons's skills and style as an actress, so much so, in fact, that she chose it for her farewell performance on June 29, 1812, when her engaged and enthusiastic audience refused to let the play continue past the sleepwalking scene. The 1785 performance included the already well-known music by Richard Leveridge, at this stage in the century thought to be by Matthew Locke, or sometimes even by Henry Purcell, a piece of wishful thinking that linked the country's greatest playwright with its greatest composer.Footnote ¹ On this occasion it was noted by the Gazetteer and New Daily Advertiser for February 3, 1785, that: “The music, which as we think we have already said is not the composition of Purcell, was never so successfully executed, and perhaps never will be so, unless it should be introduced in the Heroic Ballet on the subject of Macbeth, which is projecting at the Opera.” What is clear, then, is that the ballet's preparation at the Opera, London's venue for elite opera and dance, was clearly common knowledge, and there were doubtless some who looked forward to Rossi's benefit with anticipation. As we will see, the composer responsible for both the arrangement and the new music, François Hippolyte Barthélémon (1741–1808), did introduce the Leveridge music, although not as successfully as it seems to have been used in its more usual position in the play (Highfill Reference Highfill, Highfill, Burnim and Langhans1973, 365).Footnote ² Why the commentator thought that the performance of the music would possibly be better “executed” at the Opera is not clear; it may perhaps be a comment on the comparative musical standards between the playhouses on the one hand and the Opera House on the other.

Rossi's benefit took place at the King's Theatre on March 17, 1785. The opera on the bill that night was Il curioso indiscreto, a two-act comedy of 1777, that had a story based on an episode in Don Quixote and a libretto possibly by Giovanni Bertati (1735–ca1815), but perhaps by Giuseppe Petrosellini (1727–1799): the 1777 setting was by Pasquale Anfossi (1727–1797). Anfossi was the musical director for the Opera at the King's Theatre from 1782 to 1786, and this was one of five operas of his own he staged there; although it was not a newly composed work, Il curioso was new to London that season. As was usual, there were dances between the acts of the opera: between acts 1 and 2 were two items, a divertissement and a dance entitled The Cossack. At the end of the evening came the “Grand new Heroic Ballet” Macbeth, “composed”—that is, choreographed—by Charles Le Picq (1744–1806). The evening was billed to end with The Caledonian Reel, a dance probably added in an attempt to complement the “Scottishness” of the main ballet of the evening. There was, of course, nothing unusual about the layout of the program; even the inclusion of a ballet on Macbeth seems less surprising in the context of a benefit occasion, one on which we might expect to find a new, and often unusual, work performed. No printed scenario survives for the ballet—few of these survive for eighteenth-century London, in any case—but a trawl through the press suggests that the other dancers included Charles Le Picq, Pietro Angiolini, Monsieur C. Nivelon, Simon Slingsby, Monsieur Frederick, Carolina Angiolini, and Anne Dorival. The ballet included vocal parts, which were performed by Rachele Dorta, Luigi Tacsa, Angelo Franchi, and Vincenzio Bartolini. Charles Le Picq, in the role of Macbeth, partnered Rossi.Footnote ³

It had been Le Picq, then just divorced from dancer Anna Binnetti, who had partnered Rossi in her London debut in November 1782, when she danced a pas de deux at the King's Theatre.Footnote ⁴ Le Picq and Rossi either were, or were soon to be, having a liaison, that produced a daughter; whether he and Rossi ever married is open to doubt, but the Enciclopedia dello spettacolo does record that Le Picq went to Russia in 1786 with “Gertrude” his “new wife.”Footnote ⁵ Until then, Rossi had danced regularly and successfully in a variety of works including Le Tuteur Trompé, Les Epouses Persanes, The Four Ages of Man, and the 1784 Déserteur, in which her superb acting was noted in the Public Advertiser for May 15, 1784: her “fainting Fit, her Agitation preceding it, and her Revival from it,” were matters for comment. She also danced in Jean-Georges Noverre's Les Ruses de I'amour; in his The Four Nations in which Rossi and Le Picq represented the “Cosaque-Polonois”; and in the 1785 London staging of Gluck's opera, Orfeo. Rossi's performances in these works were well-received, and in commenting that the dancers available to the Opera House were “almost equally admirable,” the commentator in the Public Advertiser for December 8, 1783, noted that “if Vestris and Theodore were more surprising, Le Picq and Rossi were more touching, [and] Slingsby more exhibiting.” Rossi giving a “touching” performance suggests the dancer showed the same sensibility reportedly seen in her interpretation of the role in Le Déserteur. She was also regarded by the Morning Herald of August 20, 1783, as a lively dancer, Rosine Simonet's performance being said to blend “the elegance of Baccelli, with the sprightliness of Rossi.” She, too, was, for the writer in the Public Advertiser of March 25, a consummate actress, one whose skills exceeded even those of the great dancers of Vestris and Baccelli. The author of the column in the December 3 Public Advertiser account of her 1782 debut commented: “Her very first step proclaimed her a perfect mistress in the art of pantomime. Her dress was a manteau a la reine, lately imported by our travelling milliners. The body, bones and trimming are puce, the sleeves and lower part of the garment a white gawse.” In fact, the two aspects of the staging mentioned in this short review—the “art of pantomime” and the garment of “white gawse”—have particular significance in the history of dance in London.

First, the garment of white gauze, the “manteau a la reine,” perhaps better known as a “chemise à la reine” or a “gaulle,” the wearing of which by Rossi was an act that was significant in more ways than one. The garment was French, a fashion newly introduced by Marie Antoinette, who, while at the Petit Trianon, abandoned the uncomfortable and formal robe à la française and adopted the plain and simple chemise à la reine, a piece of clothing unobjectionable in itself, but one that in context was subversive; it meant that courtiers could be mistaken for peasants. The garment became the subject of open scandal in 1783 when Marie Antoinette chose to have her portrait painted wearing it (Photo 2); the outcry was such that the artist, Élisabeth Vigée-Lebrun (1755–1842), in an effort to quell the chorus of disapproval, painted the queen again in the same pose with many of the same details, but in more formal court attire. Such a subversion of courtly norms was interpreted as disloyalty to the French state. Further, not only was the gown made of Indian muslin “exotic,” but because the material was imported, the queen was accused of undermining the French silk industry (Weber Reference Weber2007, 156–75). Some of this background would have been known to members of the Opera House's elite audience, many more would have identified the chemise as a French import, and still more would later adopt it as a style item as it became fashionable. And in the context of theater dance, Rossi's adoption of it was another step in a new “informality” in dance costumes, characterized by a “naturalness” and a growing desire to dress the character in a way that was appropriate to the role and not simply to use all-purpose dance costumes. Noverre, who was ballet master at the King's Theatre in the 1781–82 season, had been very keen on the appropriateness of the dancers’ costumes: “when the scenery shall be adapted to the dresses, and vice versa, the effect of the representation will be complete”; “the artist who designs the dresses … should unite [with the other talents] in giving beauty and perfection to the work”; and “Mademoiselle Clairon … was convinced, that by rejecting our customs, she would come nearer to those of antiquity; that her imitation of the personages she had to represent would become more natural; that her action, being animated, it would be in her power to enliven it still more, when disencumbered from the weight and shackles of a clumsy dress” (Burden and Thorp Reference Burden and Thorp2014, 92, 147, 175–76). Noverre's views circulated in London in the English translation of his works by Parkyns MacMahon, which appeared in 1783, and his control over what the dancers wore was later acknowledged by Anthony A. Le Texier, who complained that (among other occasions) he had “seen a ballet of Furies at Sadler's Wells, wherein the first dancers had red stockings on, and the figurants wore at least black stockings” and had commented that had “Noverre not directed himself the dresses of the women,” the ballets of the Opera would have been in disarray (Le Texier Reference Le Texier1790, 35–36). And we can discount an accidental outing for Rossi's manteau à la reine; Thomas Lupino had been in charge of costumes at the King's Theatre since 1776 and as the Morning Herald for December 13, 1782, shows, his penchant for appropriate costumes was often remarked upon. Siddons's own costume for Lady Macbeth—white, like Rossi's—would also come under scrutiny in the Public Advertiser of February 7, 1782:

A prosaic conclusion perhaps, but also one demonstrating new assumptions that costumes should have a realistic effect even when, as in this case, they have a loaded history. But the appearance of “mad” characters dressed in white on the eighteenth-century London stage was clearly a frequent occurrence. If it had not been, then the satire in R. B. Sheridan's The Critic would have been unsuccessful; the stage direction “Enter Tilburina stark mad in white satin and her confidant stark mad in white linen” assumes an audience's automatic connection between such costumes and insanity (Sheridan Reference Sheridan1781, 95).

Photo 2. Marie Antoinette in a chemise à la reine. Élisabeth Vigée-Lebrun (1755 –1842). Pastel. National Gallery, Washington, DC.

Secondly, the Public Advertiser of December 3, 1782, presented Rossi not just as a skilled dancer, but as a skilled pantomimist: “Her very first step proclaimed her a perfect mistress in the art of pantomime.” The inclusion of “pantomime”—defined by Noverre, who was also Le Picq's master, as “mute dramatic representation”—was essential to a perfect ballet d'action, which was “a picture, drawn from life, of the manners, dresses, ceremonies, and customs of all nations” and which should speak “to the very soul of the spectator.” Noverre saw these “two actions”—dancing and pantomime—as not being “at a very great distance from each other” but also saw a balance between the two as central to an effective ballet. While it is not known whether “the mimed gestures replaced or supplemented the dance steps,” Friedrich Melchior, Baron von Grimm, commented in 1771 that “there is considerably more walking” in Noverre's ballets “than dancing” and that there was dancing “only in the great movements of passion, in decisive moments.”Footnote ⁶ Grimm's comments suggest that a variety of techniques were used to express the drama found in the scenario, and Nye, citing the ballet Il ratto della Sabine in which Rossi danced the role of Ersilia, suggests that by 1780 “it was axiomatic in many people's minds that it should be possible to integrate drama and dance as successfully as Metastasio integrated drama and music,” and that it was a necessity to give the performers “enough stage time to develop psychological plausibility and to sensitise the spectator to their individual fates” (Nye Reference Nye2011, 124). In Il ratto della Sabine, which Le Picq staged when he first arrived in London in 1782, Rossi was described in the Morning Herald of December 13, 1782, as “the most accomplished chef d’œuvre of dumb acting that ever graced heroic pantomime,” and was noted in the role as able to convey “her inward struggles between anger, love, and duty to herself and country.” These were the type of elaborate abstractions—thoughts, feelings, and passions—that had only relatively recently become associated with ballet.

It seems that too much pantomime was the Macbeth ballet's undoing; the writer in the Morning Post and Daily Advertiser of April 21 later remarked that “there was, indeed, but little dancing in proportion to the pantomime”; the Public Advertiser of March 18 had already published similarly negative comments:

Le Picq may have been influenced by Noverre, who by the time of Macbeth, had admitted that he was more interested in pantomime than in dancing: “It has happened, indeed, that I have renounced all regard to the mechanical composition of dancing, in order to give weight and brilliancy to the pantomime: the action of the performers should speak; their thoughts should be intelligible from their gestures, and physiognomy.” Noverre went on to acknowledge that there were “beauties of description” that were “inexpressible by pantomime language,” and when that language was required to express “strong passions,” it became “merely a confused stammering.” Perhaps the perceived difficulties in the ballet simply proved that it was impossible to escape from a desire to represent the subtleties and complexities of the play.

In the wider history of ballet in eighteenth-century London, then, this work was at this point an oddity, for it combined singing and dancing; it contained “Grand Choruses” and at least some of the Leveridge Macbeth music,Footnote ⁷ and it had new dance music, here provided by Barthélémon who had compiled the whole score. The only sections of the music we have from the work are the overture and some dance tunes from act 1, music that owes its survival to its publication in keyboard reduction by Longman and Broderip as Macbeth: A Grand Ballet; as it happens, the very fact of publication reflects the ambitious nature of the work, for such music—if it was published at all—usually came out in individual numbers or as individual movements in seasonal collections.Footnote ⁸ Of the fourteen movements included, three are credited as Scottish tunes—The Braes of Ballanden, The last time I came o'er the moor, and Tweed Side—all of which in the ballet are associated with action involving Duncan. All three tunes had already been circulating in England—early London publishing dates for the three tunes are 1770, 1670, and 1725, respectively. Although not necessarily known to members of the audience, they would certainly have been recognizable as folk tunes by style and as Scottish by association; they would have also suggested an appropriate air of antiquity. As can be seen from Table 1, Barthélémon's score mixed these tunes with some new movements, and if the ordering in the Longman and Broderip score is correct, the action of the story filled the first part of the act, culminating in the pastoral movement for the figurants dancing as “Scotch Peasants bringing Fruits and Flowers to Duncan.” This was followed by four untitled longer showpiece movements for the principal dancers that were extraneous to the plot. The act then ended with a return to the action, in a movement in which Macbeth conducted Duncan to the bedchamber, leaving the scenario of the first part of the ballet on an appropriate cliff-hanger, a positioning that suggests that the murder of Duncan took place between acts 1 and 2. As far as the music of the Macbeth-related ballet movements is concerned, Barthélémon's score reflects in some detail action that is implied by titles that appear in the printed source. For example, “The surprise of Macbeth after the witches’ prophecy” opens with two distinct musical gestures both of which are closed off by a fermata over a rest, a musical representation of Macbeth's “startlement” at the weird sisters’ prediction. Similarly, the pauses, articulation, overall tessitura, and above all, the comparatively extreme ranges of the intervals in “Macbeth and Lady Macbeth resolving to murder Duncan” capture their conspiracy in all its horror, the seriousness of their crime being summoned up by the closing calando phrase with its gradual decrease in both tempo and volume. These indications—which we can interpret in the context of the general care used in the preparation of the printed music—allow us some insight into the possible organization of the gestures the dancers employed in the mime. It is impossible to be specific about what these were, but unlike gesture as static representation found in Fuseli's Lady Macbeth (or for that matter, to be found as part of any ballet scenario), the score offers a plan for gesture as movement, even if that plan cannot be realized in its original form with any certainty.

Despite his obvious competence, today's judgments on Barthélémon's music have not been positive; his “… compositions exhibit considerable charm, but lack originality or a clearly developed personal style … [which] never evolved much beyond the mid-century style of which his colleague J. C. Bach was perhaps the most celebrated proponent” (Zaslaw and McViegh Reference Zaslaw, McViegh and Sadie2001, 2: 782–84). But as a performer, Barthélémon was said in the Journal de musique of May 1770 to have “exquisite taste” and was renowned for his realization of adagio movements, interpretations described by Burney as “truly vocal” (Burney Reference Burney1789, 4: 682). Burney's daughter, the novelist Fanny Burney, took her heroine, Evelina, to Vauxhall Gardens, where she was inspired to describe Barthélémon as a “player of exquisite fancy, feeling and variety” (Burney Reference Burney1794, 2: 55). There is no doubt that Barthélémon's playing was capable of every nuance suggested by the scores that seem to us to lie rather limp on the page (Burden Reference Burden2010). Despite this, Macbeth was, in general, poorly received, the writer in the Morning Herald of March 19 commenting: “Mr Barthelemon has made a wretched assemblage of airs. Matthew Locke's venerable mantle, by being trimmed with the plaid of light Scotch Melodies, and trifling jigs, appears sadly disgraced.” Other fundamental problems with the work, both in its structure and in its performance, were identified by others, such as the author of the following remarks from the Public Advertiser of March 18:

Not for the first time, artists’ tangling with the Scottish Play came to grief; the “Locke” music was too well-known and too widely accepted to be placed alongside the lighter music of late eighteenth-century France and Italy. The witches were not only singing in the music of Italian recitative, but doing so with “foreign-sounding” pronunciation. And above all, the Opera House (as opposed to the playhouse) was meddling with Shakespeare. These problems had in fact been foreshadowed in the General Advertiser of February 28 in which one writer had remarked that the intended ballet had to be “executed with the greatest nicety, or the attempt will fail. One laugh destroys the whole.” Ultimately, though, the main problem was that there was just not enough dancing; as the Morning Post of March 18 remarked: “On the next performance of this ballet we would recommend a greater variety in the dances, and, if it is beneath the dignity of Macbeth and his Lady to exhibit themselves, that some of their subjects and courtiers may be permitted to enliven the action with a larger mixture of comic dancing.” Even allowing that “comic dancing” may simply be non-serious, such a suggestion seems counter to everything that was written about the style of this “Grand new Heroic Ballet”; it does seem that the press wanted more comédie—or dramatic action-based dancing—rather than these lengthy Noverre-style mime scenes.

Macbeth was not, though, the only unusual work staged at the Opera House that season. The new opera, Alina, regina di Golconda, had been staged on March 18, 1784; the new libretto, attributed to Antonio Andrei, was set by Venanzio Rauzzini (1746–1810). The opera had been advertised with ballets between the acts; those between acts 1 and 2 appear to have been the usual plotless assemblages of dance routines, while the last one, after act 3 (which also closed the evening), was an elaborate new ballet by Theodore D'Auberval, entitled The Four Ages of Man, which was cast appropriately to present childhood, youth, manhood, and old age.Footnote ⁹ Like the Macbeth ballet, the first performance was for a benefit, this one for D'Auberval as composer of the dances. However, the advertisement in the Morning Chronicle and London Advertiser claimed that the opera would be “an entirely new species of entertainment, after the French style” while the libretto claimed that the opera will be performed “With Dances analogous to the DRAMA, Interwoven with CHORUSSES and SONGS.” “Analogous” because they addressed the subject of the opera and were not introduced “without bearing the least relation to the Subject of the Drama, where, therefore, they are far from answering the Purpose” (Lockman Reference Lockman1747, v). The opera had seven distinct sequences of dances and choruses integrated into the three acts; the act 1, scene 4 sequence, for example, had a chorus “With dancing and cheerful voices” followed by the chorus “Let the mariner repose in safety” during which the Indians and sailors dance and which is then interrupted by an air (Andrei Reference Andrei1784, 14–17). The London audience kept the opera alive for eleven performances, but after that, the opera was shelved. Shelved it may have been, but the ballet was not entirely forgotten: as late as the novel Marston published in 1835, Captain Norris chose gambling over being “wrapped in ecstacy at some entrechats of Madamoiselle Rossi in the beautiful ballet of Macbeth” (St Clair Reference St Clair1835, 1: 29).

Both the ballet of Macbeth and the opera of Alina, regina di Golconda offer different versions of the same narrative: the London audience wanted their dance without singing and their opera without dancing, even if they did require the dancing as a bonbon at the end of the evening of opera, serious or comic.Footnote ¹⁰ But at least one later report, that in the Morning Chronicle and London Advertiser of March 21, 1785, which praised Barthélémon's music for the Macbeth ballet, wondered why the score (from which it was suggested that “the hacked English Makabet, &c” could be omitted) had been given up after only one benefit performance: with alterations “there can be no doubt of the Ballet acting with great attraction for several nights to come.” After all, the columnist concluded, what “greater praise” could be offered a foreigner than to say that his music for “this subject of Shakespeare's” was some of the most pleasing and appropriate he had composed since his arrival in London?

In all this critical noise, however, it was Rossi, as Lady Macbeth, who clearly gave the most memorable performance of the evening, the performance that impressed the writer in the Morning Herald of March 19, “with the recollection of Fusili’s [sic] painting.” We first hear of Henry Fuseli's Lady Macbeth painting—which seems to have been begun in 1781—in a letter dated September 1783 in which the artist remarked: “It is now nearly finished and I would wish [to know if Your mind is altered since] Smith has just finished the Lady Macbeth. In some parts, considered as a whole of Light and Shade it exceeds my expectation, and the expression of the waiting woman only excepted (I think he has done me more than Engravers justice)” (Weinglass Reference Weinglass1982, 21).Footnote ¹¹ He later refers to the print as a “very large mezzotinto … from a large Picture of mine representing Lady Macbeth walking in her Sleep” (Weinglass Reference Weinglass1982, 22). The engraver, John Smith, of whom Fuseli seemed to approve, published his print on January 6, 1784. Fuseli's subject is taken from the play's act 5, scene 1, and the print carries the legend: “one, two why then, ’tis time to't.” The quotation—given Fuseli's comment above, we can assume that the artist agreed that it fitted his conception of the picture—occurs immediately after the more famous phrase, “Out damned spot!” regarding the imagined stain of Banquo's blood: “Out, damned spot! Out, I say! — One, two. Why, then, ’tis time to do't. Hell is murky!—Fie, my lord, fie! A soldier, and afeard?” Lady Macbeth is, of course, sleepwalking, a condition acknowledged in the conversation between her gentlewoman and the doctor, in which this text is embedded: the doctor on arriving sees that her “eyes are open,” but the gentlewoman, with the perception of the more intelligent servant, responds “Ay, but their sense is shut.” The doctor and the gentlewoman are both seen in the background of Fuseli's picture. Lady Macbeth occupies the foreground, and holds aloft the taper that she has commanded be “by her continually.” These figures, together with the background, appear almost supernatural, for the light is thrown onto Lady Macbeth's face at the center of the canvas and across her shoulders to the tip of her left-hand thumb and forefinger. Fuseli's nonspecific model for Lady Macbeth suggests that the gestures in the picture cannot be attributed to any individual actress, but his acquaintance with the theater indicates that they can be given a specific theatrical interpretation. One source, Gilbert Austin's Reference Austin1806 gesture compendium Chironomia, is particularly useful in this context, for it is not a simple instruction manual but presents gesture in a historical continuum, making it possible to consider Fuseli's gestures within its framework. Here, Lady Macbeth's left-hand thumb and forefinger create a gesture in which the fingers point upward, with “the other fingers turned inwards contracted with force according to the energy of the speaker” (Austin Reference Austin1806, 337 and illustrations 45, 66, and 67). At the same time, this is combined with the hand being held “perpendicular to the horizon, with the fingers pointing upwards” (Austin Reference Austin1806, 338 and illustrations 64 and 74). Austin records this as a composite gesture of both aversion and accusation, and its use by Fuseli suggests that he is representing Lady Macbeth's aversion to the killing of Duncan which she has effected, and an accusation directed at Macbeth for the subsequent murders he has undertaken to secure the throne (Austin Reference Austin1806, illustrations 100 and 101). Her pose is not dissimilar to that found in the gladiator drawing with which Fuseli's earlier study of the sleepwalking scene has been associated (Tomory Reference Tomory1972, 79 and plate 29). The dislocation experienced by the viewer is partly caused by the figure thrust forward, its slipper missing, its flowing hair, and wild costume, but the dislocation may also be the result of what Marguerite Tassi has identified as the transfer of the heroic; the heroic character is here is the deranged Lady Macbeth, not a tragic male lead (Tassi Reference Tassi2011, 64).

As mentioned above, Fuseli did not use a particular actress as a model for his painting, and no identifiable performance appears to have inspired it (Smith Reference Smith2010, 30–40); it is a product of the distinction made “between paintings derived from the literary text and those that originated in the theatre.” This distinction has been dated to the early 1780s when “Shakespearean subjects began to be painted in some quantity,” and as Altick notes, “the former bore the more honoured credentials” (Altick Reference Altick1985, 256). The artist Robert Edge Pine drew the distinction clearly in commenting that his series of pictures is “not meant to be representations of stage scenes; but will be treated with the more unconfined liberty of painting.”Footnote ¹² And Fuseli's work, such as the witches from Macbeth, is said be “dramatic, but it did not simply seek to draw on, or re-enact theatrical traditions” (Postle Reference Postle and Cannon-Brookes1991, 13).Footnote ¹³ Nevertheless, Fuseli does tie his painting to the play by the use of the epigraph, which was attached to the print already in circulation when, in April 1784, the picture, obviously now not an image new to the public, was exhibited at the Royal Academy as item 66 (Weinglass Reference Weinglass1994, 74–75). A report in the Gazetteer and New Daily Advertiser of April 27, 1784, noted that: “Mr Fusali [sic] laudably perseveres in the pursuit of the higher departments of the art. His genius employs itself as heretofore on the turbulent passions of the soul. And in his Lady Macbeth he demonstrates very striking faculties.” The “passions of the soul” may have moved some members of the audience, but not Horace Walpole, who hated it: “execrable” was the adjective he used to describe it (Whitley Reference Whitley1928, 2: 377).

The strangeness of the picture echoed the strangeness of its painter. Born in Zurich, Switzerland, Fuseli was the son of the collector and writer on art, Johann Caspar Füssli (1706–1782). His early career was that of poet, and when he visited England first from 1764 to 1770 with the British chargé d'affaires in Berlin, Andrew Mitchell (1708–1771), it was with the intention of promoting German poetry. His departure for Italy was with the encouragement of Joshua Reynolds whom he had met in 1768 and who persuaded him to follow a career as an artist. One success led to another, and Fuseli was elected to the Royal Academy in 1790 and was professor of painting there in 1799; he was described by one acquaintance as “an extraordinary and very entertaining character” (Weinglass Reference Weinglass1982, 40). Benjamin Robert Haydon found his character to be “the most grotesque mixture of literature, art, scepticism, indelicacy, profanity, and kindness” (Paston Reference Paston1905, 18).Footnote ¹⁴ Fuseli's pictures include Satan Starting at the Touch of Ithuriel's Spear, The Death of Dido, and The Mandrake: A Charm, the last described by Horace Walpole as “shockingly mad, madder than ever, quite mad” (Whitley Reference Whitley1928, 2: 377). But it was a madness that appealed to the public; his Nightmare took London by storm.

However, it was Fuseli's encounter with Shakespeare's works that was to provide a profound moment in the artist's career. His literary ambitions had introduced him to the plays on the page, and when he first arrived in London, it was the performances influenced by David Garrick that were on offer. Fuseli was a devotee of the theater—Weinglass recounts that in “later life he often dismissed his classes at the Royal Academy early in order to indulge his passion for the playhouse”—and Shakespeare was, of course, a constant in the repertory (Weinglass Reference Weinglass, Matthew and Harrison2004, 21: 209 and Tomory Reference Tomory1972, 71). In seeking to describe the effect on Fuseli of seeing Garrick playing Shakespeare, Tomory claims that it was similar to Jacques Louis David's response in seeing Naples for the first time: the result was akin to seeing the world after “an operation for a cataract” (Tomory Reference Tomory1972, 15).Footnote ¹⁵

Fuseli had encountered the play Macbeth early in his life and had translated it while still in Zurich (Tomory Reference Tomory1972, 15). From his years in Rome in the 1770s, there is a reference to the now-missing painting A Scene from Macbeth Footnote ¹⁶ and to some designs for a Shakespeare Room in the manner of the Sistine Chapel.Footnote ¹⁷ And from the mid-1770s comes an earlier drawing, a Roman-style depiction of Lady Macbeth Sleepwalking that includes two watching figures that Tomory traces to a mosaic by Dioscorides (Tomory Reference Tomory1972, 79 and plate 30). The drawing shows Lady Macbeth striding forward in a manner that can best be described as “concentrated raving,” and although it is in no way preparatory for Fuseli's 1785 painting, it nonetheless presents his subject in an already recognizable interpretation. But it was in 1783 that he exhibited The Weird Sisters at the Royal Academy, a painting of which the commentator in the Morning Post of April 29, 1783, wrote: “We have already traced this Artist's pencil with success in pieces of deformed stature. In the present picture he is very successful, as the Weird Sisters answer every description given of them by the creative imagination of Shakespeare.” Fuseli, as a purveyor of the monstrous, is here credited with following Shakespeare's text, but also following his own now already well-trodden path in depicting deformity.

The commentator mentioned above, in whom Rossi's performance inspired a “recollection of Fusili’s [sic] painting,” remarked on its contrast to that of Mrs. Siddons, “our critical and admired English performer.” The unspecific reports of Siddons's interpretation suggest that in these early performances, her style was more like that of Hannah Pritchard as painted in the role by Fuseli and Johann Zoffany. These two portraits date from around 1760–66 and 1768, respectively, and as Shearer West has pointed out, in both cases: “Pritchard's Lady Macbeth is in full command of the situation—not displaying the after-effects of that moment of weakness that Montagu felt could be glimpsed in the character prior to the murder” (West Reference West and Eger2013, 170).” Siddons would later describe her interpretation of the role, an interpretation that comes closer to the manner suggested by Fuseli's picture. But Siddons's biographer, Thomas Campbell, claims that Siddons had not fully developed her ideas about the character until about 1815, ideas that ultimately led to her “Remarks on the Character of Lady Macbeth” published by Campbell in a chapter of his Siddons biography in 1834 (Campbell Reference Campbell1834, 3–39).Footnote ¹⁸

Fuseli's Lady Macbeth walking in her Sleep represents what is a key scene for both the dancer and the actress. For Siddons, her own account of what happened backstage with R. B. Sheridan in her dressing room on the night of that benefit performance, indicates that not only did her interpretation deliberately differ from that of Pritchard, but did so in one essential detail:

It is true that Siddons is later shown holding the candle in the engraving by Thornthwaite (Photo 3) and again in a drawing by the amateur caricaturist, George Murgatroyd Woodward (1760–1809),Footnote ¹⁹ but it is possible, given Siddons's emphasis on this point, that both artists added the detail of the candle based on their experiences and assumptions of the role as played by Pritchard. Siddons is not pictured with it again, even in the major 1814 painting by George Henry Harlow, in which she is wringing her hands, without a candlestick in sight. This light source is, though, a key element of Fuseli's picture, where, far from being the Wee Willie Winkie candlestick of Thornthwaite, it is a large-scale, taper-like candle held aloft that emphasizes the wild, primitive, and even supernatural nature of Lady Macbeth's madness (Sillars Reference Sillars2006, 224–25).

Photo 3. M^rs SIDDONS in LADY MACBETH. Out Damned Spot! Richard Westall (1765–1836). Engraved by John Thornthwaite (b. 1740?). Engraving. Victoria and Albert Museum, S.2383–2013.

Despite Siddons's gloss on her own performance in the sleepwalking scene, one of the most competent reviews—that in the London Chronicle of February 1–3, 1785—was critical of her staging of this passage: “In the taper scene, she was defective: her enunciation was too confined, and her looks not exactly accommodated to what one must supposed to be her feelings at the time; the faces she made were horrid, even ugly, without being strictly just or expressive.” The “too confined” enunciation suggests a rather stately performance. Rossi's danced interpretation seems, in contrast, to have been extravagant, one probably with gestures in the Italian mode and more like Lady Macbeth as conceived by Fuseli: George Bell commented that Fuseli thought it an error in Shakespeare that when the cry of women marks the death of Lady Macbeth, she did not appear; “he would have had her struggling in death among white sheets” (Bell Reference Bell1870, 126–27).

We can see through Bell's sarcasm to Fuseli's “struggling” character, one Rossi clearly played like a woman possessed: tortured, histrionic, and potentially out of control. And her performance came off well even when compared with that of Siddons, a comparison the author made in the review in the Morning Post of March 18, 1785: “In the banquet scene, Madame Rossi fell short of Mrs. Siddons, but in some of the others, was equal at least, if not superior to that admired actress.” As Siddons grew into the role, it seems by her own account that her performance of Lady Macbeth became more like the image painted by Fuseli. But in 1785, Siddons's interpretation had a “solemn dignity of demeanor,” a “brilliant beauty,” and an “expressive countenance,” and it was Geltruda Rossi who “accomplished chef d’œuvre of dumb acting,, who dancing the role as devised by Le Picq, gave a deeply “touching” performance and brought a dramatic “sprightliness” to the part, à la Fuseli.