1 Robert Moberly and Christopher Raeburn, ‘Mozart's “Figaro”: The Plan of Act III’, Music and Letters, 66 (1965), 134–6 (p. 134).

2 The issue of Da Ponte's competence as a dramatist in arranging La folle journée as an opera libretto was raised in the anonymous satire Anti-Da Ponte (Vienna: Joseph Hraschanzky, 1791). Acting as a witness for the prosecution, Beaumarchais speaks with vehemence: ‘daß er von dem da Ponte gräulich müsse mißhandelt worden seyn’ (‘that he must have been terribly mistreated by Da Ponte’). The librettist, he claimed, had completely misunderstood La folle journée: ‘Er bestand darauf, da Ponte habe sein Stück La folle journée ou le mariage de Figaro gar nicht verstanden, denn sonst würde er keinen solchen Wechselbalg daraus gemacht haben’ (‘He insisted that Da Ponte had completely misunderstood his piece La folle journeé ou le mariage de Figaro, or else he would not have made such a changeling of it’). Anti-Da Ponte, trans. Lisa de Alwis ([n.p.]: Mozart Society of America, 2015), 50.

3 The autograph of Le nozze di Figaro remains in three repositories: Acts 1 and 2 are in the Staatsbibliothek zu Berlin, Preußischer Kulturbesitz, Musikabteilung mit Mendelssohn-Archiv; Acts 3 and 4 are in Cracow, Biblioteka Jagiellońska, Musikabteilung; the recitative ‘Tutto è disposto’ is in Stanford University, Music Library.

4 Alan Tyson, ‘Le nozze di Figaro: Lessons from the Autograph Score’, Mozart: Studies of the Autograph Scores (Cambridge, MA: Harvard University Press, 1987), 114–24 (p. 116).

5 The set of parts is in Vienna, Österreichische Nationalbibliothek, Musiksammlung, Signatur: O.A.295.

6 Dexter Edge, ‘Mozart's Viennese Copyists’ (Ph.D. dissertation, University of Southern California, 2001), 1591.

7 Wolfgang Amadeus Mozart: Le nozze di Figaro, K.492: Facsimile of the Autograph Score, Staatsbibliothek zu Berlin–Preußischer Kulturbesitz, Biblioteka Jagiellońska Kraków (Mus. ms. autogr. W. A. Mozart 492), Stanford University Library, The Juilliard School Library, introductory essay by Norbert Miller, musicological introduction by Dexter Edge, 3 vols. (Los Altos, CA: Packard Humanities Institute, 2007), iii, 18.

8 John Eliot Gardiner, ‘A Better Order for “Figaro”?’, Le nozze di Figaro, CD booklet, Deutsche Grammophon, Archiv 439871 (1994), 13–15 (p. 14).

9 Moberly and Raeburn duly noted the chronological problem caused by the length of time that it takes for Suzanne to report back: ‘It is obvious that Susanna [must] look for the Countess, to tell her what happened during her duet with the Count. […] It is always a shock to hear [her] say, so long afterwards, that [she] has not yet done so.’ Moberly and Raeburn, ‘Mozart's “Figaro”’, 136.

10 Le nozze di Figaro: Comedia per musica tratta dal francese in quattro atti da rappresentarsi nel Teatro di Corte (Vienna: Kurzbek, 1786).

11 The Librettos of Mozart's Operas, ed. Ernest Warburton, 7 vols. (New York and London: Garland, 1992), iv: The Late Works, Translations, and Revisions, pp. xiii–xiv. In Le nozze di Figaro: Kritische Bericht, ed. Ulrich Leisinger, Neue Mozart Ausgabe, II/5/16 (Kassel and Basle: Bärenreiter, 2007; hereafter NMA: KB), 91, it is listed as source N².

12 Several of the unique readings in the German translation relate to scenes at the end of Act 3. In Act 3, scene xiv, the reference to the fandango is left out of the stage direction, which is worded in such a way that Figaro and Bartolo receive Susanna and Marcellina in succession, without the intervention of a dance. This is not mentioned by Dorothea Link in ‘The Fandango Scene in Mozart's Le nozze di Figaro’, Journal of the Royal Musical Association, 133 (2008), 69–92, but it is consistent with her contention that the fandango was included quite late in the day in order to exploit Francesco Benucci's facility as a dancer. An earlier stage instruction (in Act 3, scene xiii) that Figaro should stretch his leg and show that he is able to dance is similarly omitted. In view of this, one wonders whether Rosenberg's refusal (until overruled) to allow the recruitment of professional dancers was itself the reason why Benucci was asked to perform a fandango.

13 Memoirs of Lorenzo da Ponte, trans. Elizabeth Abbott (Philadelphia, PA, and London: J. B. Lippincott, 1929), 150. If this description of how composer and librettist worked on Figaro were taken literally, it would imply that sometimes – realistically it can only have happened with recitatives – Mozart set the words even as they were being composed by Da Ponte. From time to time, one does see a recitative (such as ‘Eccovi, o caro amico’) apparently copied in small sections.

14 A four-bar fragment of the start of an alternative to ‘Aprite’ headed ‘|invece del duetto di Susan[n]a e Cherubino.|’ is extant. It is headed scene iii. Salzburg, Internationale Stiftung Mozarteum, Bibliotheca Mozartiana, Autogr 492. NMA: KB, 66; source A^d.

15 When working on the Act 4 finale, Mozart – possibly aware that one or more of the preceding arias might have to be cut – again omitted the scene numbers, and they were never added in.

16 In his Reminiscences (London: [Henry Colburn], 1826), Michael Kelly provides some independent confirmation that the duettino at the start of Act 3 was ready in good time: ‘I called on him one evening; he said to me, “I have just finished a little duet for my opera, you shall hear it.” He sat down to the piano, and we sang it. I was delighted with it, and the musical world will give me credit for being so, when I mention the duet, sung by Count Almaviva and Susan, “Crudel perchè finora farmi languire così”. A more delicious morceau was never penned by man; and it has often been a source of pleasure to me to have been the first who heard it, and to have sung it with its greatly-gifted composer’ (pp. 258–9).

17 Wolfgang Amadeus Mozart: Le nozze di Figaro, K.492: Facsimile of the Autograph Score, iii, [85]. It is described in NMA: KB, 68, as Source A^g. Salzburg, Internationale Stiftung Mozarteum, Bibliotheca Mozartiana, Autogr 492-1.

18 Alan Tyson, ‘Some Problems in the Text of Le nozze di Figaro: Did Mozart Have a Hand in Them?’, Mozart: Studies of the Autograph Scores, 290–327 (pp. 323–4).

19 This interpretation is followed in NMA: KB, 68.

20 In Così fan tutte, Mozart decided to set the words ‘In uomini, in soldati, / sperare fedeltà? / Non vi fate sentir per carità!’, intended by Da Ponte as recitative, as the first section of Despina's aria ‘Di pasta simile’. In this instance, he started the score of the aria's first section on a sheet headed by the end of the previous recitative, ‘Da spacciar queste favole ai bambini’ (‘To sell these tales to children’).

21 The number 18 written at the bottom right-hand corner of the recto of this sheet in red crayon is not in Mozart's hand. NMA: KB, 68. If it was originally intended that ‘Che soave zeffiretto’ should follow the duettino ‘Crudel! perchè finora?’, its number would indeed have been 18, but the position of the mark on the page is somewhat anomalous. Usually a red-crayon indication of a piece's number would be entered only after a pre-existing indication in Mozart's hand.

22 For a detailed analysis of how Da Ponte adapted the Beaumarchais text in this scene and the character of the ensemble as a seduction duet, see Tim Carter, Understanding Italian Opera (Oxford: Oxford University Press, 2015), 128–31.

23 The Court Theatre copy of Acts 3 and 4 is in Vienna, Österreichische Nationalbibliothek, Musiksammlung, Signatur: K.T.315. Arias for Nancy Storace: Mozart's First Susanna, ed. Dorothea Link (Middleton, WI: A-R Editions, 2011), 118–19, gives a useful description of this source.

24 In K.T.315, a chord was added at the start of the accompagnato. This is a common phenomenon in early scores, as whenever an Italian opera was arranged as a Singspiel singers could no longer rely upon being able to pitch their note from the preceding secco recitative.

25 There are signs of layered copying in K.T.315, suggesting that work on this ensemble was begun before Mozart had added in the wind instruments, except for a fragment in bar 23. Corrections to bars 45 and 51 in some of the performing materials, moving the successive dynamic marks f and p from beats 1–2 to beats 2–3, is a good example of the consequences of Mozart's imprecise vertical alignment of such marks.

26 The join is marked by a number of small inconsistencies in the sources, perhaps best interpreted as signs of haste. The Italian libretto attributes the opening lines – ‘Riconosci in questo amplesso / Una madre amato figlio’ – to the Count, whereas the German translation gives the correct character. In the autograph, Mozart inexplicably first wrote a treble (violin) clef for the Count, correcting the mistake on the spot. The preceding recitative is attacca, indicating the pitch f for the chord of resolution in the first bar of the ensemble, but the first note in the bass line of ‘Riconosci’ is F, an octave lower, and marked forte, rather than sforzando as in the violins. NMA: KB, 182.

27 One explanation for the missing sheet(s) is that this was where the short scene for Cherubino and Barbarina leading to ‘Se così brami’ was originally placed. NMA: KB, 47–8.

28 Ibid., 181.

29 Obviously it is not possible to be sure in every instance what caused a mistake. An unclear case occurs in bar 11, where Mozart initially wrote the clef for Don Curzio but had to correct it to the one for Bartolo.

30 At the end of the first page of ‘Voi che sapete’, for example, he wrote his usual ‘=’ sign in Cherubino's part, denoting a connecting syllable on the next page (‘-mor’ of ‘amor’), but when he turned over to continue writing he skipped a bar and started the second musical phrase. He did not notice this even when putting in the string parts. When he did finally spot the mistake, he drew in staves for an extra bar by hand, and the missing notes were supplied (in a different ink), though not the syllable ‘-mor’.

31 In NMA: KB, 187, Abert/Gerber and Schünemann are suggested as possible sources of these clarifications. Early Viennese copies (such as Modena, Biblioteca Estense Universitaria, Sezione Musicale, Signatur: F 791 (1–2) and New York, The Juilliard School, Lila Acheson Wallace Library, Juilliard Manuscript Collection) have Figaro's last note as a semiquaver but retain the rest, suggesting that the partial correction was made, if not by Mozart himself, at least at an early date.

32 The manner in which this recitative was copied in K.T.315 also betrays clear signs of an amended or unclear original. The two syllables ‘avven-’ are copied well to the right of the end of the musical stave to which they belong, an unusual occurrence in a copyist's score, and the spacing of the note for ‘-for-’ (‘informar’) is unusually constricted.

33 In all three of Mozart's Da Ponte settings, there are moments when the authors debated the choice of an adjective to refer to a stereotypical operatic emotion such as joy or rage. The final phrase of this recitative did not escape careful consideration. In the autograph, the Count's comeuppance is ‘to my relish’ (‘al gusto mio’), while in the libretto it is ‘to my joy’ (‘al gioir mio’). In the German translation it is ‘to my greatest pleasure’ (‘zu meinem gröβten Vergnügen’).

34 In K.T.315, the words of Bartolo's offer were at first mistakenly attributed to Marcellina, yet another sign of an unclear original.

35 One of the most intractable difficulties Da Ponte faced in fashioning a libretto out of the play was the need to provide a ‘multiplicity of musical pieces’ (‘moltoplicità di pezzi musicali’), notably solo arias for all the singers. I have suggested elsewhere that the attempt to insert an ‘arietta’ for Cherubino in Act 3 could imply that one of his other solo pieces had been withdrawn. Ian Woodfield, ‘The Trouble with Cherubino … ’, Mozart Studies 2, ed. Simon Keefe (Cambridge: Cambridge University Press, 2015), 168–94 (pp. 192–4).

36 No music has survived, but since ‘Voi che sapete’ also has quatrains consisting of lines of 5, 4, 5 and 4 syllables, a shorter version of its music could have been adapted. A fully scored reprise of the Act 2 piece beginning ‘Voi che intendete’, alerting the Countess to the approach of Cherubino, was cut from Act 4 before the première. Its loss could have been occasioned by the idea that his arietta should be switched to Act 3, thereby rendering it unfamiliar to the Countess and inappropriate for an Act 4 reprise. A small sign that Mozart started work on the composition of ‘Voi che sapete’ before its text was finalized is that the word ‘ridirò’ (‘I will relate’) is written more boldly than the remainder of the surrounding text, hinting that a blank may have been left while alternative verbs were under consideration. NMA: KB, 138.

37 Changes within recitatives, even major ones, did not usually need to be indicated for the orchestral players, but in this case a cue at the start of ‘E Susanna non vien!’ had to be revised so that the player responsible for bringing in the orchestra would do so upon hearing the final words (‘al gusto mio’) of the recitative ‘Eccovi, o caro amico’. The original cue was crossed out in red and the new cue inserted: ‘al gusto mio – in cadenza’. Later markings in heavy pencil and blue crayon confirmed the reading. The viola part has the same cue (also with ‘in cadenza’) very boldly in red crayon (although the original words are not crossed out), but there is no sign of the change in the violin 2 or basso parts. Although the picture is inconsistent, the cancellation of the cue in the violin 1 part implies that the cut was at least being considered. In K.T.315, there were several changes of mind at the end of ‘Eccovi, o caro amico’. The words ‘al gusto mio’ were crossed out (along with the remainder of the a4 passage). The word ‘volti’ was added in red crayon, followed by an unclear instruction, and then ‘alla scena 9^ma’, rubbed out but still clearly visible. If an attacca to ‘E Susanna non vien!’ was intended, the concluding measures of ‘Eccovi, o caro amico’ would have required revision to make the join work with a V–Ib cadence.

38 Le nozze di Figaro: Commedia per musica da rappresentarsi nel Teatro di Monza (Milan: Giovanni Batista Bianchi, 1787); Woodfield, ‘The Trouble with Cherubino … ’, 188.

39 Bayreuther Zeitung, 149 (14 December 1787), Anhang, 1099: ‘Schreiben aus Wien, vom 8. Dec. […] Se. Majestät haben in höchster Person einige Verkürzungen anbefohlen, die diesem Meisterwerke nichts von seinem Werthe benehmen’ (‘His Majesty has personally ordered some abbreviations, which do not detract from the value of this masterpiece’).

40 Bayreuther Zeitung, 61 (22 May 1786), Anhang, 400: ‘Wien, vom 16. May […] Die zur Oper gemachte Nozza del Figaro von Hrn. Mozzart, gefällt nur halb, und die Kenner haben verschiedenes daran ausgesezt, welches der geschickte Herr Compositor zu verbessern sich vorgenommen hat’ (‘Mozart's opera Le nozze di Figaro has only half pleased, and connoisseurs have identified various things that the skilful composer has set about improving’). Dexter Edge, ‘Report on the Reception of Le nozze di Figaro (Bayreuther Zeitung)’, Mozart: New Documents, <https://sites.google.com/site/mozartdocuments/documents/1786-05-16> (accessed 16 June 2014).

41 Daniel Heartz, ‘Constructing Le nozze di Figaro’, Journal of the Royal Musical Association, 112 (1986), 77–98 (p. 79).

42 The libretto – Die Hochzeit des Figaro: Ein komisches Singspiel in zwey Aufzügen (Brno: Johann Sylvester Siedler, 1789) – will be discussed in Ian Woodfield, Cabals and Satires: Mozart's Comic Operas in Vienna, Chapter 3 (Oxford: Oxford University Press, forthcoming).

43 Gardiner, ‘A Better Order for “Figaro”?’, 13. It is not likely that Acts 2 and 3 would both have started with a soliloquy for the Countess. The very late positioning of ‘Porgi amor’ at the start of Act 2 could indicate that a similar opening for Act 3 had yet to be ruled out.

44 In her brief soliloquy at the end of Act 2 of La folle journée, the Countess dreams of her dalliance with Chérubin, but Da Ponte rigorously excised all such expressions. In order that she should be seen as an honourable woman (an absolute requirement for the Vienna stage), he needed to depict her as being profoundly wronged but entirely blameless.

45 Frits Noske, The Signifier and the Signified: Studies in the Operas of Mozart and Verdi (The Hague: Nijhoff, 1977), 24: ‘Viewed in the light of the preceding recitative, the dramatic significance of “Dove sono” has been curiously undervalued.’

46 The German translation opts for ‘the’ (‘den’).

47 Anti-Da Ponte, 50; trans. de Alwis, 52.

48 Noske, The Signifier and the Signified, 24: ‘Its preceding recitativo accompagnato shows how much the Countess feels the intended travesti as a real humiliation.’

49 The new join would have required minor reworking. Act 3, scene i now ends with a V in D, while the new start at the word ‘Signor’ is in E♭. Edge, ‘Mozart's Viennese Copyists’, 1638–9.

50 An example of this in Act 3 comes at the end of the recitative ‘Queste sono’, where Figaro's words ‘mai quel che non so’ were cut, necessitating a new cue (‘Io non impugno’) in all the orchestral parts.

51 Edge, ‘Mozart's Viennese Copyists’, 1608–16.

52 Wye J. Allanbrook, ‘Pro Marcellina: The Shape of “Figaro”, Act IV’, Music and Letters, 63 (1982), 69–84.

53 Kristi Montesano-Brown, Understanding the Women of Mozart's Operas (Berkeley, CA: University of California Press, 2007), 177.

54 This consideration did not trouble a later reviser in K.T.315 when it became necessary to transpose the Countess's aria down a tone into B♭. The amended recitative ends with a prosaic I–V–I cadence.

55 Jane Glover, Mozart's Women: His Family, his Friends, his Music (London: Macmillan, 2005), 255.

56 Mozart omitted the stage direction ‘parte’ given in the Italian libretto and the German translation (‘gehet ab’).

57 It is interesting to see some recognition in K.T.315 that its content connects to the moment later in Act 3 when Antonio enters ‘pian piano’ to reveal Cherubino. A small but distinctive red triangle at the start of ‘Io vi dico’ has a counterpart at the start of Act 3, scene xi (‘Eh! Cospetaccio!’).

58 The Monza libretto (see above, note 38), which makes all the cuts to the Cherubino–Barbarina subplot marked in K.T.315, retains this recitative.

59 Dittersdorf placed the Countess's aria (with no equivalent for the accompagnato) at the start of the letter-writing scene (Act 2, scene xi) with the direction: ‘Gräfin, und hernach Susanna’.

60 A bifoliation number was added at the start of a pair of (usually conjoined) sheets, that is, one number for every four pages (sides).

61 In his analysis of this juncture in the plot, Tim Carter wondered whether the unexplained continuity direction (‘Dopo il duettino’) at the start of the abandoned particella could refer to a second duettino for the Countess and Susanna on the subject of this brief disagreement. Although this would mirror Beaumarchais, it seems unlikely that there would have been two successive ensembles for the same set of characters so late in an act – it usually happens only in an introduzione. Tim Carter, W. A. Mozart: Le nozze di Figaro (Cambridge: Cambridge University Press, 1987), 153.

62 The use of the augmented-sixth chord to end a secco recitative is rare in Mozart. Another example is the attacca link in Così fan tutte from ‘Eccoli’ to Don Alfonso's small aria of feigned anguish ‘Vorrei dir’. In this case, the augmented-sixth chord of the recitative resolves directly onto a second-inversion tonic chord of F minor. The melodic expression, however, is relatively understated, with an augmented fourth quite low in the singer's range setting the tremulous exclamation ‘L'idol mio … ’.

63 Edge, ‘Mozart's Viennese Copyists’, 1580.

64 In NMA: KB, 193, the hand is identified as that of the person who copied the recitative ‘Dunque voi’ in the autograph.

65 This reading was also transmitted in later scores. NMA: KB, 193.

66 Woodfield, ‘The Trouble with Cherubino … ’, 173–4.

67 Bruce Alan Brown, ‘Beaumarchais, Mozart and the Vaudeville: Two Examples from “The Marriage of Figaro”’, Musical Times, 127 (1986), 261–5.

68 There was little time to proofread this section. An aria or ensemble would normally be distinguished from recitative through the use of line indentations, but there is no sign of this feature. Only the rhyme scheme betrays the presence of a concerted piece.

69 He was perhaps also accommodating varying levels of background knowledge in his audience: ‘Da Ponte's inclusion of the word “zeffiretto” in his made up timbre encourages ambiguous interpretations, as if he were purposely accommodating spectators both with and without knowledge of the vaudeville practice.’ Bruce Alan Brown, ‘Lo specchio francese: Viennese Opera Buffa and the Legacy of French Theatre’, Opera Buffa in Mozart's Vienna, ed. Mary Hunter and James Webster (Cambridge: Cambridge University Press, 1997), 50–81 (p. 70).

70 The editor of the Prague libretto – Le nozze di Figaro, o sia La folle giornata: Comedia per musica tratta dal Francese in quattro atti da rappresentarsi nei Teatri di Praga (Prague: Giuseppe Emanuele Diesbach, 1786) – systematically restored Da Ponte's readings in the recitatives (in cases of dispute), but went with Mozart's choice here. Ian Woodfield, ‘Werktreue in the Prague Productions of Le nozze di Figaro (1786) and Così fan tutte (1791)’, Mozart Jahrbuch 2012 (2014), 245–66. The Prague libretto is the only contemporary printed text which comes close to matching Mozart's setting and is evidence of the care with which the editor made the revision.

71 Wienerblättchen, 28 February 1785, 238: ‘Susanne muβ dem Grafen einen Brief schreiben, und in demselben versprechen, sich Abends, wenn sie der Dunkelheit Schatten deckt, in seinem Garten unter der Allee Kastanienbäumen einzufinden’ (‘Susanna must write a letter to the Count, promising that in the evening, under cover of the shadow of darkness, she will be in his garden at the avenue of the chestnuts’).

72 Dittersdorf opted for a Linden arbour, but his letter scene is a send-up, probably intended to satirize Storace, who was notoriously a plain speaker. The signed missive is brusque in tone: ‘In the face of your impertinence, Count, I must finally give way. On the dot of ten o'clock tonight you will find in the Linden arbour: your Susanna Malabini’ (‘Zu ihrer Zudringlichkeit, Graf, muβ ich endlich nachgeben. Mit dem Schlag zehn Uhr heute Abends werden sie in der Lindenlaube finden. Ihre Susanna Malabini’).

73 Montesano-Brown, Understanding the Women, 186, suggests this nuance as the reason for Susanna's hesitation: under the pines? Perhaps she has misheard? Or perhaps she understands only too well.

74 Anti-Da Ponte, trans. de Alwis, 54–5, 62–3, 71.

75 Possibly Rautenstrauch wished to thumb his nose at the censor, who had deprived him of his fee as a result of the ban on the stage performance.

76 A clear indication of how late work on the duettino continued is its muddled presentation in the German text. Susanna does not repeat the first phrase of the song (‘Welch angenehme Zephyre’), even though the stage instruction for her to sit and write (‘Sus. sitzt nieder und schreibt’) comes earlier. She enters the conversation only to repeat the conclusion to the Countess's next phrase: ‘werden auf dem Abende wehen / auf dem Abende wehen’, in that respect following Beaumarchais. A part label for the Countess must have been omitted, as otherwise Susanna would make up the final line (‘unter den Fichten im Busche’) herself as well as pronouncing the verdict: ‘Das übrige wird er ohnehin verstehen.’

77 There are other occasions in Act 3 where additional stage directions in the German libretto sharpen up the delivery of the banter. In Act 3, scene iv, the Count, responding to Figaro's attempt to wriggle out of his commitment to Marcellina on the grounds that he would need the consent of his parents, asks: Where are they? Who are they? He is to deliver these lines ‘in a scornful tone’ (‘in einem hämischen Ton’). Bartolo, too, has to ask his question as to whether Figaro was a foundling ‘sneeringly’ (‘höhnisch’).

78 In K.T.315, the final instruction reads ‘leggendo insieme lo scritto’.

79 A case in point in Don Giovanni is Leporello's ‘Ah pietà’, where a significant number of stage instructions appear only in the autograph.

80 Jessica Waldoff and James Webster, ‘Operatic Plotting in Le nozze di Figaro’, Wolfgang Amadè Mozart: Essays on his Life and his Music, ed. Stanley Sadie (Oxford: Oxford University Press, 1996), 250–95 (pp. 269–74).

81 The chaotic legacy of this change of mind can be seen in an early score in Copenhagen, Det Kongelige Bibliotek, C. I. 280 ms 7502.0131. On the brace, the wind instruments are given as ‘oboe solo’, ‘flautto solo’ and ‘flauto solo’. In bar 37, the flute entry is written as though for bassoon and even has a tenor clef. A cautionary indication – ‘flauto’ – was added, but the notes were not corrected.

82 Wiener Zeitung, 53 (5 July 1786), 1597.

83 Quoted in Link, The National Court Theatre, 339.

84 Quoted ibid., 355.

85 Quoted ibid., 367.

86 L'ape musicale rinnuovata: Comedia per musica in tre atti (Vienna: Kurzbek, 1791); Marina Maymone Siniscalchi, L'ape musicale di Lorenzo Da Ponte (Rome: Il Ventaglia, 1988). In K.T.315, ‘Che soave zeffiretto’ is headed by a large red-crayon marking ‘Partitura’ (score). This probably relates to the copying of the letter duettino for L'ape musicale rinnuovata. Its score (if there was one) has not survived.

87 Siniscalchi, L'ape musicale, 240–1

88 Ibid.

89 It is quite likely that the performance of ‘Che soave zeffiretto’ in the pasticcio would have featured the version seemingly mandated by Mozart in K.T.315, in which the Countess takes the higher line in the final couplet. Second thoughts as to which singer would be better suited to the top part became evident during the composition of Acts 1 and 2. By the time that Act 3 was under way, Susanna (Storace) had been allocated a part above the Countess (Laschi) in ensembles in which they appear together. At some point, however, this decision was reversed, and K.T.315 was marked up to effect the change. Edge, ‘Mozart's Viennese Copyists’, 1570–1.

90 Siniscalchi, L'ape musicale, 240–1.

91 The Trieste libretto – L'ape musicale ossia Il poeta impresario: Commedia per musica in due atti de Signor Abate da Ponte (Trieste: […] Stamperia Governiale, 1792) – gives the text of the duettino more accurately including its vaudeville title: ‘Canzonetta sull'aria: / che soave Zeffiretto / questa sera spirerà – / Sotto i pini del boschetto … / Ei gia il resto capirà: / certo certo il capirà.’ Siniscalchi, L'ape musicale, 293–4.

92 Montesano-Brown, Understanding the Women, 193.

93 Wye J. Allanbrook, Rhythmic Gesture in Mozart: Le nozze di Figaro and Don Giovanni (Chicago, IL, and London: University of Chicago Press, 1983), 147.

94 Mozart himself preferred the more specific image of a healing balm when in Zerlina's aria of consolation he altered Da Ponte's ‘È certo antidoto’ to ‘È un certo balsamo’.

95 The periodic removal of ‘La’ from ‘Contessa’ seems less significant, although Mozart does preserve equality in his segue indication for the duettino: ‘Segue Duettino di Susan[n]a e Contessa’. Da Ponte had to stick to ‘Il Con.’ and ‘La Con.’ to avoid confusion. Mozart's choices seem quite random: ‘Contessa’ in the Act 2 finale scene braces; ‘La Contessa’ in the equivalent Act 4 indications.

96 In 2014, I was offered the opportunity to participate in just such an experiment. At the suggestion of Kenneth Baird, managing director of the European Opera Centre, I reconstructed a version of Così fan tutte on the basis of my study Mozart's Così fan tutte: A Compositional History (Woodbridge: Boydell Press, 2008). Its central hypothesis was that Ferrando and Guglielmo were originally to have returned in disguise, each to serenade his own lover, but that there was a late change of plan, with the Act 2 pairings being switched so that each man seduces his comrade's woman. Signs of a drastic rethink along these lines include a series of unexplained oddities in the text of the opera, akin to those that led Moberly and Raeburn to consider a late reordering of Act 3 of Le nozze di Figaro. Three performances of a production by Bernard Rozet were given by the Royal Liverpool Philharmonic Orchestra at St George's Hall Concert Room, conducted by Laurent Pillot. The project's avowed aim was to gain an improved understanding of the problems faced by the authors as they struggled to construct a coherent drama.