2. The Rossinian Repertoire

© 2022 Massimo Zicari, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0277.02

According to the chronology compiled by Charles Neilson Gattey and Thomas Kaufman, only one Rossinian opera features prominently in Tetrazzini’s repertoire: Il barbiere di Siviglia.1 The first record dates back to 6 March 1894, when she made her appearance as Rosina at the Teatro San Martin of Buenos Aires together with the baritone Pietro Cesari and C. Elias. On reporting how she triumphed in that role, the critic of The Standard wondered why she had insisted ‘on gilding the purest musical gold extant. Was the marvellously florid score not florid enough for you, without taxing your grand voice by adding your fioritura?’2 Tetrazzini adhered to a performance practice that was still shared among prima donnas, however strenuously music critics continued to disapprove of it. After 1894, Tetrazzini would be Rosina in sixty-one productions throughout her career, while her last public appearances in this role were in 1912, in Chicago, Boston and London.3

A second Rossini opera features in her chronology, although in a much more marginal position. On 7 June 1898 Tetrazzini made her appearance in Buenos Aires as Mathilde in Guglielmo Tell together with Francesco Tamagno, Eugenio Giraldoni and R. Ercolani; Leopoldo Mugnone conducted. She would assume this role again only on 2 February 1902, this time in St. Petersburg, with Michele Mariacher, Mattia Battistini and Vittorio Arimondi. The reason why she did not consider singing the role of Mathilde more often has been suggested by Tetrazzini herself, although in general terms. As we have already seen, in How to Sing she raises the question of stylistic appropriateness and suggests that every singer ‘must be familiar also with the varying needs of the different schools of music, with the historical traditions associated with them’.4 Since every repertoire has its own different requirements, in terms of both vocal technique and style, it is unlikely, if not impossible, for an artist to excel in all of them. The consequence is that every singer should confine herself to the class of work more particularly suited to her talent. Tetrazzini herself admitted that, at some point in her career, she had to make a choice and steer away from those works which did not fit her voice and personality.5 If, on the one hand, flexibility was a quality that any singer should cultivate, it was not advisable to push oneself into foreign lands and embrace repertoires that did not fit one’s voice. A case in point is represented by Guglielmo Tell, as the critic of The Standard remarked when reviewing her impersonation of Mathilde on 21 June 1898: ‘Senorita Tetrazzini looked and sang charmingly and it is a pity that in this opera she has so little to show what an excellent soprano leggero she is’.6 To put it bluntly, Mathilde did not offer sufficiently sparkling vocalisation to allow her to show off her voice and upstage the rest of the cast. Unfortunately, Tetrazzini has left no recordings from this opera.

Although no other title from the Rossinian repertoire can be found in the chronology of her live performances, Semiramide, a role she apparently never assumed on stage, appears in Tetrazzini’s recordings. She recorded ‘Bel raggio lusinghier’, Semiramide’s cavatina, on 2 November 1910 in London with Percy Pitt conducting the orchestra. A second recording for the Victor Talking Machine was realised on 11 May 1914 with Walter B. Rogers conducting the orchestra (matrix C14818-1); this has never been issued.

Rosina, the Cunning Girl

It was typical of nineteenth-century Italian operas to give prima donnas and primi uomini a nice cavatina with which to make their entrance and present their character. In this case, Rosina introduces herself by singing an aria that is suggestive of the character’s personality, dramatic role, and vocal quality. Although it is not my intention to elaborate on the extent to which Rosina simply embodies a stereotyped character within the stereotyped framework of a comedy of intrigue, in her cavatina she is showing the audience her double-faced quality. If, on the one hand, she is a sentimental, timid girl whose heart is now touched by the idea that Lindoro may have fallen in love with her, Rosina is also the kind of young woman who can show the determination and wit necessary to overcome any obstacle. This may easily result in designing a hundred tricks and deceiving people. One may even wonder which is her true nature, whether the timid or the cunning, and if she is not just putting on modesty like a social mask, as has been suggested by Janet Johnson.7


Una voce poco fa

A voice a moment ago

a (2+2)

Qui nel cor mi risuonò,

Here in my heart resounded

Il mio cor ferito è già

My heart has been wounded

a’ (2+2)

E Lindor fu che il piagò.

And it is Lindor who wounded it

Sì, Lindoro mio sarà,

Yes, Lindoro will be mine,

b (2+2)

Lo giurai, la vincerò.

I swore, I will win.

b’ (2+2)

Il tutor ricuserà,

The tutor will refuse,

c (2+2)

Io l’ingegno aguzzerò,

I will sharpen my wits,

Alla fin s’accheterà

In the end he will come to his senses

E contenta io resterò.

And I shall remain content.

Sì, Lindoro mio sarà,

Yes, Lindoro will be mine,

b (2+2+1)

Lo giurai, la vincerò.

I swore, I will win


Io sono docile,

I am docile,


a (2+2)

Son rispettosa,

I am respectful,

Sono ubbidiente,

I am obedient,

a’ (2+2)

Dolce, amorosa,

sweet, loving,

Mi lascio reggere,

I let myself be led,

b (2+2)

Mi fo guidar.

I let myself be guided.

Ma se mi toccano

But if they touch me


a (2+2)

Qua nel mio debole,

Here in my weakness,

Sarò una vipera,

I will be a viper,

a’ (2+2)

E cento trappole

And a hundred tricks

b (2+2)

Prima di cedere

Before I give in

b (2+3)

Farò giocar.

I will play.


The lyrics consist of four six-verse stanzas of which the first two present eight-syllable truncated verses, while the second two change into five-syllable verses, thus leading to a clearly perceivable rhythmic acceleration. The music follows this segmentation: the first two stanzas, in which Rosina reflects on her new condition, are set to an Allegro moderato in ¾ time; Lindoro is in love with her and all the difficulties will be overcome. A Moderato in common time follows, which sets to music the second two stanzas; now Rosina moves to a more general description of her personality: respectful but determined, sweet and obedient but cunning and smart if necessary. The opening Allegro moderato consists of a sixteen-bar section followed by a twelve-bar section; it presents a more declamato-like style, with the orchestra supporting the voice mostly with isolated chords. The Moderato section features a more lyric style, and consists of three sections: ABB’. After the first, made up of three four-bar phrases, the aria continues with a livelier section which is repeated twice, to conclude with a short coda.

However used we are to recognising the sweet side of Rosina’s personality in the first melodic idea of the Moderato, and her cunning in the second, more sparkling one, and however strongly we feel the connection between the lyrics and the music Rossini composed to set them to, the initial melodic motive is not original and involves some self-borrowing from previously written and already performed operas. More interestingly, each time it was used this melody was meant to express a different feeling, whether melancholy or joyful, depending on the situation and the dramatic persona. We find it in Arsace’s rondo ‘Non lasciarmi in tal momento’ from Aureliano in Palmira, an opera seria in two acts set to a libretto by Gian Francesco Romanelli, which premiered at the Teatro alla Scala (Milan) on 26 December 1813, and also in Elisabetta’s cavatina ‘Quant’è grato all’alma mia’ from Elisabetta, regina d’Inghilterra, a drama in two acts set to a libretto by Giovanni Federico Schmidt after a play by Carlo Federici based on Sophia Lee’s The Recess. Elisabetta was premiered at the Teatro San Carlo (Naples), on 4 October 1815. In short, when Il barbiere was premiered in Rome on 20 February 1816, it was the third time that Rossini was using, at least in part, the same melodic material, each time for a different theatre and in a different city. The reason for this frequent self-borrowing was clarified by Rossini himself when Ricordi announced plans to publish vocal scores of all of his operas; in 1864 the composer thus expressed his preoccupation to Tito Ricordi: ‘The edition you have undertaken will give rise, justifiably, to much criticism, since the same pieces of music will be found in various operas: the time and money accorded me to compose were so homeopathic that I barely had time to read the so-called poetry to set to music’.8 The composer goes on reminding Ricordi that his only concern in those years was to support his parents and relatives financially. Although one may wonder whether these last apologetic remarks were entirely true, it is clear that the opera production system imposed a very tough working pace not only on composers, but also librettists and singers. Rossini seems to have expressed the same concern to Beethoven, when paying him a visit in Vienna in 1822. A detailed report of the conversation between Rossini and Beethoven has been left by Edmond Michotte, who took notes of the visit Richard Wagner paid to Rossini in Paris in 1860. On that occasion, Rossini told Wagner about his meeting with Beethoven, who had suggested that he confined himself to the composition of comic operas. Rossini told Wagner, as Michotte writes, that he had tried to explain to Beethoven how difficult it was for him to work on the composition of a new opera.9 Librettos were imposed by impresarios, the scenarios were given to him one act at a time, and the composition of the music had to proceed without his even knowing how the story would end. On the other hand, not many operas survived for more than a few seasons and, since opera scores did not circulate in printed form, it was difficult, if not impossible, for critics and audiences to understand whether and to what extent self-borrowing was involved in the composition of a new opera.10

‘Una voce poco fa’ and Nineteenth-Century Performance Practice

A number of reviews, accounts and recollections confirm that during the nineteenth century Rossini’s Barbiere had become not only a favourite among international audiences but also a benchmark against which the value of any prima donna would be assessed. This implied that each emerging young soprano would undergo comparison with her colleagues and predecessors in terms of acting skills, vocal brilliancy, ornamentation and coloratura. Among others, this tradition involved the insertion of newly-designed ornaments, the substitution of new ornaments for those written by the composer, and even the insertion and substitution of entire arias. In particular, in the nineteenth-century Il Barbiere soon came to feature three major manipulations: the insertion of newly-written coloratura in Rosina’s famous cavatina ‘Una voce poco fa’ in the first act, the addition of new ornaments to the soprano part in the duet ‘Dunque io son’, and the substitution of the aria ‘Contro un cor che accende amore’ in the so-called lesson scene in the second act. It is well-known that the insertion of new coloratura had been endorsed by Rossini himself, who wrote different ornamented versions for Rosina’s cavatina: one was dedicated to Matilde Juva Branca in 1852 and is to be found in the Biblioteca Nazionale Braidense (Milan), while others are preserved in Brussels, at the Fonds Edmond Michotte, and in Munich, in the Franz Beyer Collection.11 The intervention involving the addition of new ornaments to the soprano part of ‘Dunque io son’ is also well-documented and Rossini himself wrote an ornamented version for this duet, which he donated to the singer Eugénie Rouget.12 The practice involving the substitution of the aria in the lesson scene arose as soon as Geltrude Righetti-Giorgi made her first appearance in this role in 1816 and continued uninterrupted until the beginning of the twentieth century, with Angiolina Bosio, Pauline Viardot García, and Adelina Patti among its worthiest representatives. Nineteenth-century commentators were not at all lenient towards those interpreters who endorsed these modifications; however, as we will see, they were willing to bestow signs of benevolence upon condition that the cantatrice of the night was of great repute and her talents extraordinary.

Evidence of this long-lasting tradition can be found in a number of contemporary reviews. When Il Barbiere was revived in London on 14 June 1858 at the Royal Italian Opera, the critic of The Times commented on Angiolina Bosio’s delightful rendition of Rosina’s cavatina but could not avoid observing that she would have been much better had she not taken so many liberties in the melody of ‘Una voce poco fa’.

Madame Bosio’s Rosina would be still more delightful if she were to take fewer liberties with the music, and to preserve a little more of the original melody of ‘Una voce poco fa’, which is in itself so beautiful that we cannot help preferring it to all the delicate embroideries with which the most brilliant songstress may exhibit her skill in adorning it. Whatever Madame Bosio attempts she accomplishes to perfection, but she is apt to regard the music of Rosina, too, exceptionally from the ornamental point of view.13

Only a few days later, on 21 June, The Times reviewed another rendition of Barbiere at Drury Lane, this time with Madame Pauline Viardot García as Rosina.

Judged from a musical point of view, Madame-Viardot’s Rosina is a miracle of cleverness; but even more relentlessly than in the instances of some of her renowned predecessors and contemporaries is Rossini sacrificed on the altar of the singer’s vanity. To ‘ornament’ and vary the cavatina ‘Una voce poco fa’ till scarcely a vestige of the original remains—to such excess, in short, that the composer might have spared himself the pains of doing anything more than note down a figured bass, is a privilege apparently claimed by every independent mistress of the florid style; and, however inclined to dispute the theory, it is impossible not to acknowledge the wonderful art with which Madame Viardot embellishes and disguises not only the plain melody, but even the ‘bravura’ of Rossini. With regard to the duet, ‘Dunque io son’, however, neither theory nor practice can be defended. If the lady has a right to embroider, so has her companion; and it would be simply poetical justice were some eager and adventurous Figaro, endowed with the ancient Tamburinian fluency, unexpectedly to contend for vocal supremacy with Rosina herself, answering ‘roulade’ by ‘roulade,’ ‘fioritura’ by ‘fioritura’. Our ‘prima donna’, thus caught in her own trap, the audience would laugh, the barber triumph, and Rossini be avenged. In the lesson scene, in lieu of Rode’s variations, or some other conventional display, Madame Viardot introduced two Spanish airs, accompanying herself on the pianoforte. She sang them with inimitable spirit, and enraptured the audience beyond measure. But then her respectable guardian’s protest against the music preferred by Rosina, and so different from that of his own early predilections,—

‘Ma quest’aria, cospetto! È assai noiosa’

‘la musica a miei tempi era altra cosa’—

lost its point—since in the first place there were two airs instead of one; and in the second, Dr. Bartolo would hardly have applied the epithet ‘tedious’ (nojosa) to the national melody of his own country. And now, having done with criticism, we need only add that the talent of Madame Viardot was never more brilliantly exhibited, and that her performance, from first to last, was received by the audience with unbounded tokens of satisfaction.14

The critic raised three issues of paramount importance. The first regards the cavatina and the way in which interpreters like Bosio and Viardot abused the composer, whose music was constantly sacrificed on the altar of the singer’s vanity. Although no details are provided by the critic, it can be assumed that Viardot’s ornamentation involved both the addition of new ornaments and the substitution of those written by Rossini, as suggested also by Manuel García in his Scuola.15 The nineteenth-century history of the interpretation of this aria is accompanied by endless variants and cadenzas, of which numerous transcriptions exist that can be found reproduced in printed editions. Luigi Ricci’s Variazioni-Cadenze Tradizioni per Canto (1937), Volume I, includes passages transcribed from Adelaide Borghi-Mamo (1829–1901), Maria Malibran García (1808–1836), Barbara Marchisio (1833–1919) and Adelina Patti (1843–1919).16 In her Méthode de chant composée pour ses classes du Conservatoire, published in 1849, Laura Cinti-Damoreau includes, among others, a cadenza for this cavatina.17 In 1900 Mathilde Marchesi wrote a collection of Variantes et points d’orgue dedicated to her students, which includes also a few variants and cadenzas for Rosina’s cavatina.18 The 1943 Book of Coloratura [and] Cadenzas edited by Estelle Liebling includes, among others, variants and cadenzas associated with Lily Pons and Luisa Tetrazzini.19 More recently Austin B. Caswell has edited a volume of Embellished Opera Arias which offers, among others, both Marchisio’s and Laure Cinti-Damoreau’s cadenzas and variants.20 Last but not least Karin and Eugen Ott have edited a volume entirely devoted to the ornamentation of nineteenth-century vocal music in which most of the variants from the operatic tradition have been accurately transcribed for the modern reader to learn and compare.21 What these transcriptions have in common is the consistent choice of ornamentation figures and their position inside the aria; they all support the notion that, against many critics’ better judgement, singers did not hesitate when it came to inserting new coloraturas and making substantial changes in the melodic line. The fact that these cadenzas and variants have come to form a body of published sources is a testament to the importance of this practice, at least in the eyes of the contemporary singers and possibly of the audience.

The second issue regards the duet ‘Dunque io son’ and the strikingly inconsistent use of the ornamentation between the two interpreters: while Pauline Viardot as Rosina deviated from the written text, Cesare Badiali, as Figaro, did not. The duet in question is divided into two sections, the first of which presents a characteristic voice setting where a first melodic idea is sung in turn by either character (Figure 9).

Figure 9 shows the melodic material given to Rosina in the duet ‘Dunque io son’.

In a typical love scene where the soprano and the tenor exchange their vows, the two voices eventually merge into a characteristic pattern of parallel-thirds or parallel-sixths.22 Instead, in this duet, after briefly resuming the first melodic idea, the second part sees each voice assuming a different melodic-harmonic role, more like a concertato situation. The baritone does not take over the melody but supports it with a simple octave-note accompaniment figure (Figure 10).

Figure 10 shows the concertato-like setting in the duet ‘Dunque io son’.

Despite the critic’s perplexity, this modification had long established itself in the tradition. In fact, if we turn our attention to contemporary treatises, we see that this very situation is also contemplated. On addressing the issue of how to add ornaments when it comes to singing a duet, in his Scuola Manuel García suggests that ‘in duets embellishments may be blended in both parts; but in trios, quartettes &c., no change is allowable’;23 the same concepts, although presented in the form of a Socratic dialogue, can be found in García’s Hints on Singing, which were published in 1894, edited by Beata García: ‘in duets singers combine their ornaments; but in concerted music where all parts are of equal importance, no change is ever admissible’.24 At this point, it could be reasonably argued that, given the asymmetrical disposition of the roles, the second part of the duet ‘Dunque io son’ would not lend itself to any ornamentation. Instead, and to our surprise, García offers this very passage as an example to illustrate the way in which ‘changes should correspond to the composer’s idea and present the same effect, but augmented’.25 García offers more than one solution, suggesting that the ornamentation should augment the effect by increasing either the number of notes or their brilliance (Figure 11).

Figure 11 reproduces the variations to ‘Dunque io son’ from Barbiere, present in M. García’s method (1842).

Interestingly, in his Hints, which were published in 1894 as a ‘new and revised edition’, García offers more solutions than in the previous one, making a distinction between a soprano and a mezzo, and offering ornamented variants that may fit either voice compass. As can be observed in Figure 12, the ornamented version for the mezzo soprano voice lies comfortably within the stave, reaching the upper A only twice.

Figure 12 reproduces the variations for mezzo soprano to ‘Dunque io son’ from Barbiere, present in M. García’s Hints (1894).

Instead, the ornamented version for the soprano reaches the upper B and C (Figure 13).

Figure 13 reproduces the variations for soprano to ‘Dunque io son’ from Barbiere, present in M. García’s Hints (1894).

The variants for mezzo soprano share a sixteen-note figure and are characterised by runs and arpeggios, while those for soprano reach the upper part of the voice compass using staccato octave figures. Interestingly, all the variants suggested by García belong to the second part of the piece.

Of course, it is all too easy to speculate on the position and type of ornaments Viardot may have added or substituted. What was her choice? Was she already enriching the first melodic idea while her colleague was adhering strictly to the text? Or should we assume, instead, that she was making changes similar to those we find described in Manuel García’s method? The latter is entirely plausible, especially if we consider the close connection that existed between Pauline Viardot and her brother Manuel García and the fact that they belonged to the same vocal breed, so to speak.

The strength with which this tradition persisted into the twentieth century is testified to by singers belonging to the younger generations. A case in point is represented by Maria Callas’s interpretation of this duet together with Tito Gobbi. Callas and Gobbi recorded this aria twice: a first live recording was made at La Scala Theatre in Milan in 1956,26 with Carlo Maria Giulini conducting, while a second was realised in 1957,27 this time together with Alceo Galliera. In either rendition, when returning to the second melodic figure Callas sings the variations transcribed in Figure 14 while her partner adheres to the written text. Her part already presents some melodic modifications at the beginning of the motive and the variants in the following bars are consistent with, if not similar to, those suggested by García for the soprano voice (Figure 14, second stave).28

Figure 14 presents the ornamented version sung by Maria Callas in 1956 and 1957.

Going back to Viardot-García’s Rosina in 1858, the critic raised a third issue regarding the well-documented substitution in the so-called lesson scene in Act II. The critic makes two points; the first concerns the fact that, instead of Rode’s variations, which were frequently chosen to replace the original aria, Viardot sang two Spanish songs; the second highlights the dramatic inconsistency that, in his opinion, this insertion generated. Bartolo should have liked those two songs, given that he is a Spaniard and the action takes place in Seville.

In this scene, Don Alonso, Count Almaviva in disguise, knocks at Bartolo’s door and introduces himself as professor of music and student of Don Basilio. Since Don Basilio is sick, he will give Rosina her singing lesson. The lesson opens with Don Alonso asking her what she would like to sing, to which she answers ‘Il rondo dell’Inutil precauzione’, the newly-composed opera. The aria Rossini composed for this scene was ‘Contro un cor che accende amore’. As has been already clarified by Hilary Poriss, the substitution of this aria with a virtuoso piece chosen to suit the prima donna’s vocal qualities and musical taste (or lack thereof) dates back to Rossini’s time: the first to make the change was Geltrude Righetti-Giorgi, the very first Rosina, who, already a few months after the 1816 premiere of Barbiere in Rome, performed a different aria in Bologna and yet another in Florence.29 Eventually, not only did it become common practice to substitute this aria with others from either Rossini or any other later composer (including Verdi), but these had come to include mostly theme-and-variation arias, sometimes derived from instrumental pieces adapted for the voice. Among them, Pierre Rode’s Air Varié Op. 10 for violin, which Henriette Sontag introduced in 1826, came to be a favourite.30 Rossini himself wrote a second aria to substitute for the original, setting to music lyrics that reflected the trouble Rosina was in: ‘La mia pace, la mia calma / vo’ cercando e non ritrovo / ogni dì un tormento nuovo / per quel crudo ho da soffrir’ (My peace, my tranquillity / I am in search of, but cannot find / every day I have to suffer / a new torment because of that cruel man). It soon became common practice to interpolate vocal pieces that had nothing to do with the original aria and its dramatic settings. As early as 1819 Giuseppina Ronzi de Begnis used to insert bravura variations to the barcarole ‘La biondina in gondoleta’ setting a precedent that would be followed by every interpreter of repute. Joséphine Fodor Mainvielle sang ‘Di tanti palpiti’ from Tancredi, Maria Malibran’s choice was ’Yo soy contrabandera’ by Cabella, while her sister Pauline Viardot preferred ‘La fiancée du bandit’ which she had composed.31

Of course, young singers of no great repute also endorsed this performance practice in the attempt to keep up with the expectations set by the international divas. This was the case with Russian-born contralto Anna De Belocca (née de Bellokh), who was twenty-one years old when she made her London début as Rosina at Her Majesty’s Theatre in 1875. Her lesson scene included a Russian melody and an aria from Donizetti’s Lucrezia Borgia.

It was in the next act, however, that Mdlle. Belocca set all fears at rest and completed her success. ‘The Lesson scene’ won the sympathies of the whole assembly of connoisseurs. The first song was a Russian melody (‘Solove’) of plaintive character, which must have somewhat astonished Dr. Bartolo, being sung in the Russian language, but it was tuneful and expressive enough to enlist general and hearty sympathy. What followed was the brindisi ‘Il segreto per esser felice’ (Lucrezia Borgia), which, rendered in a characteristic and empathic way, roused general enthusiasm, and was loudly encored. By this time the general conviction was that a new ‘star’ had appeared in the operatic horizon. Such gifts as Mdlle. Belocca is endowed with, such rare personal attractions and such evident aptitude for the profession of her choice was well worth cultivating. The impression created on Saturday night will cause every amateur to watch with interest the future career of the interesting young Russian.32

Her London début was a success and in 1876 she would make her appearance, again as Rosina, at the Academy of Music in New York.33

Rosina was also one of the roles that best fitted Adelina Patti’s voice and personality. According to an anecdote that has been told repeatedly, upon attending her first soirée at Rossini’s place in Paris in 1862, the nineteen-year-old Patti sang ‘Una voce poco fa’ before the composer and the other guests. Camille Saint-Saëns recalled the episode:

Unhappily, I was not present at the soirée during which Patti was heard at Rossini’s for the first time. It is known that when she had performed the aria from Il barbiere, he said to her, after many compliments: ‘By whom is this aria that you just have let us hear?’ I saw him a few days later: he still had not calmed down. ‘I know perfectly well,’ he told me, ‘that my arias must be embroidered; they were made for that. But not to leave a note of what I composed, even in the recitatives—really, that is too much.’34

However, Rossini’s opinion of Patti’s vocal and musical talents would improve over time and the initial irritation was soon forgotten, as Saint-Saëns narrates.

In his irritation he complained that the sopranos persisted in singing this aria which was written for a contralto and did not sing what had been written for the sopranos at all. On the other hand the diva was irritated as well. She thought the matter over and realized that it would be serious to have Rossini for an enemy. So some days later she went to ask his advice. It was well for her that she took it, for her talent, though brilliant and fascinating, was not as yet fully formed. Two months after this incident, Patti sang the arias from La Gazza ladra and Semiramide, with the master as her accompanist. And she combined with her brilliancy the absolute correctness which she always showed afterwards.35

When in 1867 Eduard Hanslick paid a visit to the Pesarese he reported that at some point the name of Patti came up in a conversation: ‘The Maestro’, Hanslick wrote, ‘speaks of the latter Patti with admiring esteem, and always singles her out as an exception when deploring the extinction of truly great singers’.36 Unfortunately, no recording survives that may be a testament to her interpretation not only of this very piece, but also of any other aria from the Rossinian repertoire. However, many contemporary reviews survive that suggest how spirited her impersonation was. When she made her appearance in this role in London in 1863, the critic of The Times described her vocalisation in sparkling, vibrant colours.

Last, not least, the Rosina of Mademoiselle Adelina Patti—one of those impersonations which not merely exhibit to advantage her singular vocal facility, but show her to be a comic actress of the genuine stamp—lively, piquant, full of intelligence and sensibility, in every look, act and gesture original was quite worthy a match with the Almaviva [Mario] and Figaro [Ronconi] of the evening. The profuse embellishments with which she decks out the cavatina (‘Una voce poco fa’), however the stickler for Rossini’s text quand même may find objection (and we still declare ourselves of their party), are excused, if not rendered imperative, by the peculiar range and calibre of her voice. They are, moreover, in some measure, warranted by the precedents of Sontag, Persiani, Bosio, and other renowned singers—not excluding Malibran, who, like her sister, Madame Viardot—might easily have given the music without the alteration of a single note, a restriction to which, nevertheless, those celebrated artists rarely if ever condescended. Mademoiselle Patti’s ornaments and fioriture have the merit of being entirely her own, and the unhesitating manner in which they are delivered obtains acceptance and extorts applause. We are less inclined to submit without a word of protest to the alteration in the duet with Figaro (‘Dunque io son’); nor can we believe they were either expedient or very effective—and this, in spite of the arch ‘bye-play’ and fluent vocalisation of this gifted little lady, which might almost reconcile the testiest amateur to any and every liberty she chose to take. In the lesson scene, Mademoiselle Patti introduced the very difficult ‘valse,’ entitled ‘Di Gioja,’ composed expressly for her by M. Strakosch, her brilliant execution of which elicited so general and spontaneous an encore that she was compelled to repeat it.37

Interestingly, the review touches upon the same issues we have discussed in the review of Viardot’s rendition in 1858.38 The rich and abundant fioriture added by the singer in the cavatina, a tolerated practice that found its justification in a number of worthy predecessors, the questionable alteration in the duet ‘Dunque io son’, and the insertion of a new piece in the lesson scene, this time a waltz composed by Maurice Strakosch, Adelina’s brother-in-law and early manager. In 1895 the critic of The Times reviewed Patti in another performance of Barbiere and commented on the insertion of ‘Bel raggio’ from Rossini’s Semiramide in the lesson scene, which an encore was likely to follow.

Her delivery of ‘Una voce’ may have been a little mannered, but in the rest of the music, whether concerted or solo, she sang as brilliantly as of old. It is true that the quality of some of the higher notes in ‘Bel raggio’—introduced into the ‘lesson scene’—was a little less than beautiful, but the text of the opera was quite delightfully sung. As the aria from Semiramide was the only thing in which any fault was to be detected, it was most rapturously received, and Mme Patti kept to her old traditions of the part by singing ‘Home Sweet Home’ in her usual manner.39

Apparently, it was with Patti that the repertory of coloratura showpieces inserted into the lesson scene came to include Vincenzo Bellini, Giuseppe Verdi, Giacomo Meyerbeer, Charles Gounod, Heinrich Proch, Félicien David, Ambroise Thomas, and Leo Delibes, among others.40 Afterwards, singers who took over this role followed in Patti’s footsteps, as was the case with Tetrazzini: some of these showpieces can be found in her discs, for she recorded them in London in the early 1910s: the ‘Bell Song’ from DelibesLakmé (1907), Benedict’s ‘Carnevale di Venezia’ (1909), the ‘Polonaise’ from Ambroise Thomas’s Mignon, that is to say ‘Io son Titania (1907 and 1911); ‘Charmant oiseau’ from David’s La Perle du Brésil (1911), Proch’s Air and Variations Op. 164 ‘Deh, torna mio bene’ (1911), and Henry Bishop’s ‘Home, Sweet Home’ (1912) are all included in Tetrazzini’s London recordings which were reissued by EMI in digital format in 1992.

It is clear that towards the end of the nineteenth-century the character of Rosina came to feature the three major textual modifications discussed so far, and incorporate them as a stable, if not permanent, component of the opera; no singer would consider a different choice, no music critic would object to that tradition, no audience would give a wince. Quite the opposite; when in 1912 such an authoritative newspaper as The Times advertised the forthcoming opera season at Covent Garden and announced Tetrazzini’s appearance as Rosina in Il barbiere, specific mention was made of the bravura piece that she would insert in the lesson scene: ‘Il barbiere di Siviglia (in Italian) Tonight, at 8.15—Mmes. Tetrazzini, Bérat; MM. McCormack, Marcoux, Malatesta, Zucchi, Sampieri and Sammarco. In the lesson scene Mme. Tetrazzini will sing “Polacca” (Mignon). Conductor, Signor Panizza’.41 To announce the bravura piece chosen by the soprano singer would help market the production and appeal to the audience. At the same time, this confirms the status this interpolation had now achieved, thus suggesting that a body of changes, substitutions and interpolations had come to be part of the performative text.

Tetrazzini’s international career is well documented and a number of reviews accompanied her appearances in this role over many decades, providing evidence of the way in which international audiences showed clear signs of appreciation while critics continued to express doubts about her real talents. In January 1905 Tetrazzini made her first appearance as Rosina in San Francisco where, despite a bad cold, she made a furore. Many a journal recorded the enthusiastic reception and described how the audience ‘screamed, screamed, screamed until, cold or no cold, she had to cast the golden notes again at the end of the lesson scene’.42 On 3 October 1905, when Tetrazzini sang Rosina again in San Francisco, Blanche Partington remarked how striking she had been not only as a songstress but also as an actress.

Tetrazzini, if she lost her voice, could walk round to the Columbia any day right into comedy—if she does not prefer farce. Her Rosina was simply bubbling over with fun from beginning to end. She seemed glad, to get off the grand opera stilts—as glad as we were to have her. There was not the slightest hesitation in sacrificing the voice to the fun, yet she was in lovely voice, though tried a little in the ‘music lesson’ by the heat. But for the ungodly laughter that filled the haunts of grand opera last night Miss Tetrazzini was most largely responsible. Her ‘music lesson’ was encored as usual.43

In 1908 the critic of The Musical Times recorded the successful revival of her Rosina in London and addressed a question concerning the lack of evenness in her voice, a shortcoming that some critics noticed more than once during her career.

A most successful revival of Rossini’s Barber of Seville took place on June 15, Madame Tetrazzini being heard for the first time in England as Rosina. So brilliant was her singing in her first song, the familiar ‘Una voce,’ that the audience interrupted its continuance after the first verse, and from this point the evening was for her a series of triumphs. Madame Patti, who was present and was most generous in her applause, must have felt that her successor could not give the music in the medium part of the voice with the same beauty of tone as she had done, but the higher florid passages were delivered with exquisite finish and a volume of tone and vocal agility that frequently approached the phenomenal.44

Tetrazzini was Rosina again at the end of the year, on 14 November, at the Manhattan Opera in New York City. The Times correspondent in New York reviewed the opera season there and did not fail to mention how successful she had been.

Mme. Tetrazzini’s reappearance last night was the signal for another display of enthusiasm. The Italian singer has firmly established her position as a favourite here, and she essayed the part of Rosina with entire success. Slightly nervous at the outset, she was quite at her best in the ‘Lesson’ scene, in which she introduced Proch’s famous ‘variations,’ with the Bell Song from Lakmé as an encore. She sang both pieces superbly.45

The following year the opera was repeated and the critic from The Times reported briefly on its popular success and how she had fulfilled the general expectations. This time, in the lesson scene Tetrazzini had inserted the Polonaise ‘Io son Titania’ from Ambroise Thomas’s Mignon, followed by Benedict’s Variations de concert sur Le Carnaval de Venise. Although a critical note was expressed with regard to her occasionally imprecise intonation and the signs of ‘effort’ in the polonaise, Tetrazzini’s singing was almost perfect and the whole cast excelled in the vivacity of their acting.

Those who care for Il Barbiere only as it serves to show off the agility of a favourite soprano were given all that they could want last night, for Mme. Tetrazzini was in excellent voice, and from ‘Una voce’ onwards, in the lively scene with Figaro in the second act, and in the famous lesson when she delighted her hearers by her singing of the polonaise from Mignon, followed by the ‘Carnaval de Venise’ variations, she sang everything with the ease and clearness which belongs to her work at its best. A note here and there not quite perfectly in tune could, indeed, be detected, and the trills at the climax of the polonaise showed slight signs of effort, but such things were only small flaws in singing which was nearly perfect of its kind. There are, however, quite other interests in the opera than the vocal skills of the prima donna, and on the whole last night’s company combined well to bring out all the humour of the thing by their acting and especially by their crisp and vivacious treatment of the recitativo secco.46

As we can see, it had become customary for Tetrazzini to introduce a first bravura piece in the lesson scene and then to sing a second one as an encore. The same happened one year later, when she again was in London performing Rosina and other roles. The Times published a positive review suggesting again how the lesson scene had served as an opportunity to showcase the voice and ignite the now predictably enthusiastic response of the audience.

Il Barbiere, with Mime. Tetrazzini as Rosina, has become a regular institution of the Opera season, and the audience, which was a large one, last night gave every sign of enjoying both the vocal gymnastics of the prima donna and the time-honoured jokes of the buffa parts. With Signor Sammarco, as Figaro, Signor Marcoux as Basilio, and Signor Malatesta as Bartolo there was no fear that the fun would flag, and these three carried on the dialogue with all the air of enjoying it themselves, which is the essence of success. Signor Armanini was a moderately successful Almaviva, though at the climaxes his voice was rather forced and thin in quality. In the Lesson Scene Mme. Tetrazzini sang the Polonaise from Mignon, and here, as elsewhere in the opera, she gave us a series of surprises by the beauty of certain notes and phrases contrasting with the hard metallic quality of others. The polonaise brought warm applause, supplemented by a loud ‘Bravissimo’ fired off when the applause had subsided by an enthusiast in the gallery, and when the singer had recovered from the embarrassment caused by this exhibition she responded by singing ‘The Last Rose of Summer’ in a way which showed the best qualities of her voice.47

This time, in addition to the polonaise Tetrazzini sang ‘The Last Rose of Summer’ a traditional Irish song set to lyrics by Thomas More. Very much loved by Adelina Patti, who has left two recordings of the piece,48 this is a very simple song, with a nice melancholy melody which shares nothing with a typical bravura piece.

As this survey suggests, Tetrazzini shows how she had embraced a long and uninterrupted set of conventions and brought it to its zenith. It is possible that her personal acquaintance with Adelina Patti might have offered her the opportunity to come into contact with a tradition going back to Rossini himself. Furthermore, she might have found the role of Rosina particularly suited to her sparkling personality and dramatic skills, especially in moments like the lesson scene, where she could break the fourth wall with the audience and show off herself, not only her voice.

Rosina in Tetrazzini’s Recordings

The reviews published year after year during Tetrazzini’s career do not tell us much more of the way in which she may have interpreted this opera, nor do they reveal the extent to which she may have adhered to the tradition of which Patti still was the most authoritative living representative. With the exception of the accounts discussed so far, not much can be gleaned from the reviews on how she approached this opera. Instead, if we turn our attention to Tetrazzini’s discography and listen to the recordings she realised in the early 1910s, we can form an opinion on how she sang the famous cavatina.

Tetrazzini recorded ‘Una voce poco fa’ three times in total. The first recording was realised on 8 September 1904 (matrix number 3513, Zonophone catalogue number 2501). It lasts 2’20’’, includes the first part only (Allegro moderato) and the first measures in the introduction have been cut, possibly owing to time constraints. A digital version was issued by Nimbus Records in 1998: Tetrazzini, ‘Prima Voce’, Volume 2 (NI 7891).

The second recording, with Percy Pitt conducting the orchestra, was realised in London in December 1907 by the Gramophone Company of London (matrix 2178f, Gramophone 053146; Victor 92020). This recording has been included in the three-CD box-set issued by EMI Classics in 1992 (CDH 7 63803 2). Although technological developments in 1911 allowed for longer recording time, two small cuts are still present in the orchestra (the first in the introduction and the second in the final coda) and the B section is not present at all; Tetrazzini moves directly from section A to section B’, where she performs her repertoire of variants.

The third recording was realised on 17 March 1911 by the Victor Talking Machine Company of Camden, New Jersey (US). The original audio (matrix C10071-1, Victor 88301, 6337; HMV 2-053046, DB 690) was transferred and reissued by Nimbus Records in 1990 (Tetrazzini, ‘Prima Voce’, NI 7808). In 1911 Walter Rogers, who was then music director of the Victor Company in Camden and took it upon himself to prepare the necessary orchestral adaptations, conducted the orchestra. This last recording presents cuts similar to those made in the 1907 recording, with a few more interventions made in order to shorten the orchestral introductions, both in the Andante and the Moderato (Figure 15).

Figure 15. The chart shows the cuts in Tetrazzini’s recordings of Rosina’s cavatina.

As we have seen, after the difficult reception of its 1816 première in Rome, Il barbiere di Siviglia soon entered the international operatic repertoire and continued to score one success after the other, thus becoming a must in the playbills of opera theatres all over the world. As a consequence, and contrary to most of Rossini’s other operas that were long neglected or completely forgotten, a richly-documented history of the performance practice of this opera survives, to which Tetrazzini’s phonographic evidence makes an important contribution.49

Tetrazzini is consistent with the tradition of which she was said to be a worthy representative, and if we take a look at her ornamentation and cadenzas, we can see how old and new are often blended together in making the final result. The first visible addition in the Andante occurs at bar 28, where a sparkling cadenza concludes the first section (Figure 16).

Figure 16 shows the cadenza at bar 28 of the Andante.

This cadenza can be found in Ricci’s Variazioni-Cadenze, which collects a number of variants belonging to Borghi-Mamo, Malibran and Patti.50 The same can be said of the cadenza with which the Allegro moderato concludes; this can be found in both Ricci’s Variazioni51 and Mathilde Marchesi’s Variantes et points d’orgue52 (Figure 17).

Figure 17. The closing cadenza of the Andante.

At this point a doubt arises concerning the reliability of these sources. Considering that Ricci’s first volume of collected cadenzas was published in 1937, when both Marchesi’s Variants and Tetrazzini’s recordings had long ago made their appearance, why did he fail to mention either singer? Should we doubt Ricci’s sincerity when he attributes to Borghi-Mamo, Malibran and Patti a number of ornamentations that Tetrazzini had sung on a regular basis and that Marchesi had already published in 1900?53 Or is Marchesi lying when she entitles her volume Variants et points d’orgue composés pour les principaux airs du repertoire par Mathilde Marchesi pour les élèves de ses classes de chant? Was she pretending she had composed these variants and cadenzas for her pupils, while instead she had taken them from her predecessors, be it Borghi-Mamo, Malibran or even Patti?54 Going back to Tetrazzini, it is at least possible to argue that in her 1904 recording she had incorporated in her cadenza a passage that was already to be found in Marchesi’s 1900 published volume. Whether this passage came from one of the predecessors Ricci mentions in his later collection, or from Marchesi, is another question. A plausible answer to these questions could be that Tetrazzini may have incorporated some coloratura passages among those Marchesi had written down in her volume, while Ricci’s transcriptions, published in 1937 with Tetrazzini still alive, hide the name of the diva in order to avoid legal or copyright issues. In any case, one may still wonder whether Ricci had a reason for omitting either singer’s name.

If we now turn our attention to those variants from Cinti-Damoreau and Marchisio that can be found collected in Caswell’s volume, the difference between them and Tetrazzini is striking. When compared to the richly flourished, dashing cadenzas these two champions used to sing, Tetrazzini’s coloraturas look discreet and hesitant, if not cautious. Much more noteworthy is the fact that only a few minor changes are present in Tetrazzini’s recorded renditions, which, as we will see, is the case in all her recordings.55 This seems to support the notion that, once she had learned a role, she was unlikely to go back to the relevant coloratura, add or modify the embellishments, and memorise and rehearse them anew. This theory finds confirmation in Tetrazzini’s own words in 1908, when interviewed by The Sun.

I do not practice at all during the season, except, of course, when I am going over a new role. When I go on the stage, the orchestra says ‘la-la’ and I respond ‘la-la,’ and the conductor nods that it is all right and I sing away. That is all the rehearsing I have.56

Consistently, a memory slip may explain why the passages B-F# and A#-C# appear to have been shortened in the second and third recording (Figure 17).

Moving to the Moderato, the cut of the B section involves a dramatic loss of information for, as a consequence of that cut, no evidence survives of the cadenza that was traditionally associated with the suspension leading to the repeat (B’). We do not have information on whether, and which, new melodic material may have been sung by Tetrazzini in a live performance (Figure 18).57

Figure 18 shows the cadenza traditionally sung before the repeat (B’) in the Moderato section.

There seems to be sufficient evidence to suggest that not only the repeat (B’), but also the B section, was varied. This is clear from the sources that have been reproduced by Caswell58 (Barbara Marchisio), and Ricci.59 Something similar happens in the rest of the aria; Tetrazzini’s ornamentation bears more than a resemblance to some of the variants reproduced by Ricci, regardless of the singer who originally sang them. Instead, strong differences can be observed when her ornaments are compared to those reproduced by Caswell, especially the versions transposed into F major, sung by Jenny Lind and Josefa Gassier (1821–1866). Lind’s and Gassier’s renditions are much more flourished; their interventions in the written music are much more conspicuous, if not invasive.

In the final measures Tetrazzini sings the variants Rossini himself wrote for Matilde Juva Branca but, in the 1911 recording, she concludes with the top E, a feat that kept the international audience breathless, before they erupted in frantic applause (Figure 19).60

Figure 19 shows the final measures of the aria, where Tetrazzini sings the variants Rossini wrote for Matilde Juva; in the 1911 recording she concludes with an accented trill leading to the top E. Her cadenza was transcribed by Ricci in his Variazioni-Cadenze.

This last example draws attention to the trill and the different ways in which it could be executed. The accents on the trill at the final B, which can be clearly heard in Tetrazzini’s recording, seem to be consistent with what has been suggested by García in reference to this embellishment. Commenting on how new habits have widened the palette of possible solutions, García also includes a trill with a number of inflections and an ascending tremolo (Figure 20).61

Figure 20 illustrates how a trill could be executed according to the new, modern habits discussed in García’s Scuola.

As can be seen, Tetrazzini’s recordings confirm how strongly her interpretations were rooted in a tradition in which some of the most widely shared interpretative devices of the nineteenth century can be recognised together with the influence that such preeminent personalities as Patti exerted on her.

The Lesson Scene

As already discussed, it was typical of Tetrazzini to substitute a showpiece for Rossini’s aria ‘Contro un cor che accende amore’ in what is generally known as the lesson scene in Act II. This was consistent with a long-lasting performance tradition that dates back to 1816, and had in Adelina Patti a turning point; in fact, it was her who turned the lesson scene into a miniature concert involving a first interpolated coloratura showpiece followed by an ‘encore’.

‘Io son Titania’

The evidence presented so far suggests that Tetrazzini often substituted the ‘Polonaise’ from Ambroise ThomasMignon for the original aria. We are talking of the Récit et Polonaise ‘Je suis Titania la blonde’ (‘Io son Titania’ as the Italian translation sung by Tetrazzini reads) which Philine sings towards the end of Act II in Ambroise Thomas’s Mignon (1866), an opéra comique based on Goethe’s novel Wilhelm Meisters Lehrjahre and set to a French libretto written by Michel Carré and Jules Barbier. This is the moment in which Philine, an actress (the role was created by the Belgian coloratura soprano Marie Cabel) portrays the light-hearted Titania, from Shakespeare’s A Midsummer Night’s Dream. During her career Tetrazzini made her appearance in Thomas’s Mignon on many an occasion, although this opera did not feature prominently in her repertoire. The first time was in Buenos Aires, at the Teatro San Martin, in 1895, then on 23 February 1899 and on 1 January 1900 in St. Petersburg, when Tetrazzini sang Philine with Sigrid Arnoldson as Mignon. She performed it yet again in 1902 (Tbilisi), 1903 (St. Petersburg, with Arnoldson), 1904 (Mexico City, with Livia Berlendi) 1905 (Guadalajara, Mexico), 1911 (Boston, USA for the first time, with Fely Dereyne as Mignon), and 1913 (Chicago, with Mabel Riegelman).62

Tetrazzini left four recordings of the aria: December 1907 (Matrix 2171f, Gramophone 053142; Victor 92015, 15-1001), August 1908 (Gramophone Matrix 2574f, unpublished), 17 March 1911 (Matrix C19973-1, Victor 88296, 6342), 14 July 1911 (Matrix ai5181f, HMV 2-053058, DB 540).

Tetrazzini sings the Italian translation prepared by Giuseppe Zaffira and published by Heugel in Paris shortly after the publication of the original score:63

Io son Titania la bionda

I’m now Titania, fair, entrancing,

Son Titania figlia del sol

Airy daughter of the morning light,

Vo’ pel mondo ognor balda e gioconda,

Through the world I go, e’er gaily dancing,

Più lieve dell’angel che l’aer fende a vol.

More swiftly than the bird that upward takes his flight

Io son Titania la bionda,

I’am now Titania, fair, entrancing,

Ah! Corro ognor gioconda etc.

Ever gaily dancing,

Mille folletti intorno a me si,

A thousand fairies, gay and light,

danzando van con agil pie’

Night and day take round my car their bright flight!

E notte e dì, di mia corte ognor

Behind me all my courtiers move,

Va cantando i fasti dell’amor.

Rove and sing of pleasure and love.

Mille folletti intorno a me si,

A thousand fairies, gay and light,

danzando van con agil pie’

Nightly take round my car their bright flight,

Ma fuggon di Cinzia all’apparir!

Until the dawn breaks on their sight.

Per entro i fior che l’aurora

Among the flowers then unclosing,

Fa sbocciare

Safe reposing,

Per declivi adorni ognor

And ‘mid grasses of the field,

d’erbe e fior, …

Lie concealed. …

E dell’onde sulle spume

On the waves so whitely foaming,

Fra le brume,

In the gloaming,

Godo ognor con agil pie’

You may see me turning round,


Lightly bound…

Con agil pie’, fra l’erbette ed i fior

With footsteps light, thro’ the forests at night,

E nelle brume, godo ognor saltellar,

‘Mid shades profound my little footsteps are found,

Godo ognor saltellar

Where I trip over the ground.

Io son Titania la bionda

Behold, Fairy queen am I

Son Titania figlia del sol

I am now Titania, fairy child of the air



It is clear that the connection between the original lyrics and the newly-inserted aria is non-existent and all the concerns for the plot and its dramatic consistency were simply put aside.64 While in the original text Rosina reflects on her condition and the boundless power of true love, thus anticipating the happy ending of the opera, the words uttered by Titania, the fairy queen, depict the character’s flirtatious, light-hearted nature, surrounded as she is by fairies and birds. The music follows an A-B-A’ form, with the first section, in B flat major, sparkling and vivacious, and the second, in B major, sweet and melodic. After the reprise the piece concludes with a cadenza that stretches up to a top E flat before concluding with a long trill on the leading tone.

In terms of musical contents and interpretative choices the three recordings are identical; Tetrazzini sings this bravura piece as written. The small modifications present in her renditions fall into the space we now understand as rightfully belonging to the interpreter: here and there she changes the articulation (staccato instead of legato), and gives special emphasis to specific passages by adding fermatas and indulging in a sustained note in the high register. A few bars are cut in the short staccato passage that leads to the central section, possibly because of the limited capacity of the disc. In all the recordings she makes a small modification to the final cadenza, perhaps to emphasise the moment where the climax is reached: she approaches the top E flat through a small grace then descends chromatically to the low F. The published score presents a simplified variant, with the C substituting for the top E, should the singer not possess it, and suggests also that the most challenging measures could be left out (Figure 21). The brilliance of the aria, with its sparkling melodic figures and vivacious polonaise rhythm, offered Tetrazzini a wonderful opportunity for vocal display.

Figure 21. The cadenza concluding ‘Io son Tatiana’, which Tetrazzini recorded in 1907.

In sum, the recordings are identical, were it not for their different length: the third (4’24’’) is significantly longer than both the first (4’00’’) and the second (4’06’’). This suggests a relatively fast pace in the earlier recording, which eases up in the second and the third, thus approximating the average length of later discs.

Proch’s Variations

On other occasions, Tetrazzini inserted the Air and Variations Op. 164 by Heinrich Proch, which she recorded twice: on 18 March 1911 with Clement Barone playing the flute (Matrix C10077-1, Victor 88307, 6336; HMV 2-053045, DB 523) and on 14 July 1911 (Matrix ai5182f, HMV 2-053065).65 Again, we have a showpiece that offers itself as a splendid opportunity for the singer to show off their voice and win the applause of the audience, no matter how inconsistent it is with the dramatic situation and the scene. The lyrics, in Italian, were published in English translation by White, Smith & Co. in their Artists’ Vocal Album of 1887.66

Deh, torna mio bene,

Ah whence comes this longing

mio tenero amor,

My heart why so sad

dà tregua alle pene

Why come tear-drops thronging

del povero cor.

That pain and make glad.

Per te questo sen

Tis love that alone

più pace non ha,

Such joy can impart

sol teco mio ben

Tis love that alone

beato sarà

Makes bright my heart.

As can be observed, the text would lend itself to any dramatic situation, upon condition that a couple of young lovers be involved and their love hindered by the intervention of a greedy father or a heartless suitor. The melodic material of this aria di baule (trunk aria) consists of a simple theme in D flat major that unfolds through a double eight-bar structure, to which three variations follow: the first and the third are more brilliant while the second is slower and characterised by long trills and small grace notes.

Tetrazzini sings a shortened version of this piece: the first eight bars of the orchestral introduction to the theme are cut, the whole second variation is skipped and a few bars are cut from the third. Of special interest is this last variation, in which Tetrazzini modifies some passages in order to reach the high D and E flat on concluding a repeated and brilliantly sung arpeggio figure.

Figure 22. Tetrazzini adds some modifications to Proch’s third variation.

The aria concludes with a short cadenza with the flute, as we can see from the piano reduction published in the Artists’ Vocal Album in 1887 (Figure 23).67

Figure 23. The cadenza with which Proch’s Variations conclude.

In Tetrazzini’s recording the cadenza, aside from some differences in the use of staccato, is enriched by the interpolation of an extra passage before the close, which, of course, allows her to indulge once again in the top E flat before resting on the high D flat (Figure 24).

Figure 24. Tetrazzini’s cadenza to Proch’s Variations.

Tetrazzini was not alone in the habit of inserting additional passages into the written cadenza; at the beginning of the century other performers used to do the same and took a further opportunity to show off the voice in a longer vocal-instrumental arabesque. This is the case with Regina Pacini, who in 1905 recorded a version of this piece (Milan, Società Italiana di Fonotipia, matrix XPh 297, catalogue N° 39232), which includes the theme and the third variation only (3’15’’ ca); although she is accompanied by the piano, the flute intervenes in the cadenza, which is slightly longer and more elaborate than the original. In November 1907 Maria Galvany recorded the Proch Variations (theme, first and third variation) with Carlo Sabajno conducting the orchestra for the Gramophone Company (Matrix 10894b, Gramophone 53526, DA 494, VA 46, Victor 87058); the recording does not include the flute, although the singer enriches the cadenza by adding new and original passages. Nor can the flute be heard in the recording realised by Frieda Hempel in 1908 with Friedrich Kark conducting the orchestra (Matrix xB 4330/xB 4331, Odeon 99217/99218, O-5532).68 Tetrazzini recorded the Variations together with Clement Barone playing the flute on 18 March 1911 (Camden, matrix C10077-1, Victor 88307, 6336; HMV 2-053045, DB 523) and later at Hayes with an unnamed flautist on 14 July 1911 (matrix ai5182f, HMV 2-053065), and a few years later it was Amelita Galli-Curci’s turn to commit them to disc. Again, Clement Barone played the flute in a new and slightly more elaborate cadenza (Matrix C20663-2, Victor 74557, 6134; HMV 2-053133, DB 265). In 1940 Estelle Liebling published a piano reduction of these Variations,69 and included a first optional cadenza with the flute together with the transcription of what Tetrazzini and Galli-Curci used to sing, thus marking the end of a long period of performance practice and leading to its final canonisation in the form of a written, teachable, text-based document.


As we have observed, the showpiece that was generally used to substitute for the original aria in the lesson scene triggered the enthusiastic reactions of the audience, a circumstance that more often than not led to the diva singing a second aria as an encore. At least three pieces appear among those Tetrazzini chose in order to thank her devotees: ‘The Bell Song’ from Léo Delibes’s Lakmé (recorded in 1907 and 1911), Julius Benedict’s Variations de concert sur Le Carnaval de Venise (recorded in 1909 and 1911), and ‘The Last Rose of Summer’ (recorded in 1911). Tetrazzini made her appearance in Delibes’s Lakmé six times in total over her entire career, starting in 1904 (Mexico City), then in 1910 (New York City, Philadelphia, London), and finally in 1911 (London, Chicago). The so-called ‘Bell Song’ is in reality the last section of the ‘Scène et Légende de la Fille du Paria: “Où va la jeune Hindoue?”’ sung by Lakmé, the title role, in Act II.70 The role was created by Marie Van Zandt in 1883 and, in spite of many extremely difficult passages of staccato notes in the high register, this aria seems to defy our common understanding of showpieces, in that it does not offer so many opportunities for vocal display. The ‘Légende’ is mostly in a recitative style and only the Allegro moderato lends itself to a more brilliant vocalisation, with the characteristic octave-note figures imitating the bells (imitant la clochette) and leading to the final trill and top E to gain the final applause (the aria is in three sections, A-B-A’, but the recordings are cut). The same can be said of ‘The Last Rose of Summer’, a traditional Irish song set to lyrics by Thomas Moore and characterised by a nice, melancholy melody far from typical bravura pieces. Tetrazzini, who sang this song on a number of occasions, recorded it on 15 March 1911 (Victor 88308, 6343 HMV 03241, DB 527). Instead, Benedict’s Variations de concert sur Le Carnaval de Venise set to music a simple text that deals with a young lady waiting for her beloved one, whom she will be able to meet again after crossing the Laguna in a gondola. Here again, the variations offer a nice opportunity for vocal display and in fact Tetrazzini’s recording features an elaborate cadenza as early as at the end of the first part, Andante con moto, which the well-known theme and variations follow.

As already suggested, Luisa Tetrazzini was following in Patti’s footsteps: once the first substitution aria triggered the enthusiastic response of the audience, she indulged them and sang a second piece of a different character, maybe in a different language. The same had been done by Melba, Sembrich and Galli-Curci, the same would be done by other divas until the decline of this performance tradition between the 1920s and 1950s. By the time Tetrazzini sang and recorded them, both the showpiece and the encore had come to form a set of fixed and predictable components of this scene; each individual singer had a select repertoire of substitution arias to be used depending on the situation.

Semiramide, Queen of Babylon

Semiramide, a ‘melodramma tragico’ in two acts set to a libretto by Gaetano Rossi, was premiered at La Fenice theatre in Venice on 3 February 1823 and its length and musical style must have perplexed most members of the audience. The story of the Queen of Babylon was not new to nineteenth-century opera composers, and its first adaptation for the operatic stage dates back to the seventeenth century, with La Semiramide in India, performed in Venice in 1648. Of particular importance for the definition of the main character was Voltaire’s Sémiramis, first shown at the Comédie Française in Paris in 1748, which at the end of the century inspired a number of operatic libretti. Rossi also drew on Voltaire and, having made the necessary adaptations, turned his drama into a new libretto. The plot revolves around Semiramide, Queen of Babylon, who becomes a powerful and licentious monarch upon her husband’s death. Many versions of this legend survive and in Rossi’s libretto Semiramide poisons her husband, the legendary Assyrian King Nino, helped by Assur, Prince of the Blood of Baal. Semiramide now chooses Arsace as the new king and her husband. But, Arsace turns out to be her own son Ninia, who, long thought dead, upon learning of his real identity and the reasons behind his father’s death, is now ready to take revenge. It is Nino’s ghost who announces that Arsace will become king only when his death will have been vindicated.71 In the final scene Arsace tries to stab Assur to death but accidentally kills his mother instead. Since the three main characters, Semiramide, Arsace and Assur are a soprano, a mezzosoprano en travesti, and a bass, Rossi had to introduce the character of Idreno, a tenor, to complete the typical operatic vocal cast.72

Rossini was undoubtedly familiar with this subject and must have had in mind the adaptation of Voltaire’s text that Simeone Antonio Sografi prepared for a Semiramide set to music by Sebastiano Nasolini, which was first performed in Padua in 1790. In fact, Isabella Colbran, Rossini’s wife and first Semiramide in 1823 Venice, had appeared in the title role of Nasolini’s opera when it was produced in Naples in 1815 as La morte di Semiramide.73 As had not been possible with most of his earlier operas, Rossini took his time to discuss the subject with the librettist and worked in a leisurely manner on the composition of the music. We can assume that he shaped the vocal part of the title role around Colbran’s talent, although in 1823 her voice was not what it once had been. The fact that, when reviewing the opera in Venice on 6 February 1823 the critic of the Gazzetta privilegiata di Venezia failed to mention her cavatina may be read as a form of politeness towards a singer who could no longer keep up with her reputation. In his Musical Reminiscences, Richard Mount-Edgcumbe devotes a long section to Madame Colbran, suggesting how disappointing her performances had been when visiting London in 1824: ‘She is entirely passée, and her powers are so diminished that she is unable to produce any effect on the stage, where she gave little satisfaction: but her taste was acknowledged to be excellent, and she was much admired in private concerts’.74 Despite a difficult début, in Venice the opera was a success and would be performed twenty-eight times at La Fenice by 10 March.75 In the course of the century Semiramide gradually disappeared from theatres, with occasional reprises here and there. Rarely staged at the beginning of the last century, the modern revival of Semiramide dates back to 1940 when it was staged in Florence, and then in 1962 at La Scala. The limited interest in this opera should not come as a surprise, since at the beginning of the twentieth century Rossini’s comic works were more popular than his serious or tragic ones, and only a few remained in the regular repertoire.

Semiramide’s cavatina ‘Bel raggio lusinghier’ adheres to the same two-part structure we have seen in Barbiere, with a first Andante grazioso (6/8) in A major followed by an Allegretto in common time (4/4).76 The scene opens with a female chorus, with young citarists and damsels in various groups trying to cheer her up: Arsace is back and love is in the air. Semiramide’s words echo the chorus and express her sense of relief: having trembled and sobbed for so long, Arsace’s return gives her new hope.

Andante grazioso

Bel raggio lusinghier,

Beautiful flattering ray

A major

a (2+2+2)

Di speme e di piacer

Of hope and of pleasure

Alfin per me brillò:

finally shone for me:

Arsace ritornò,—Sì, a me verrà.

Arsace returned,—Yes, he will come to me.


Quest’alma che sinor

This soul that until now

b (2+3)

Gemè, tremò, languì…

Moaned, trembled, languished…

Oh! come respirò!—

Oh, how it breathed!—

Ogni mio duol sparì,

all my grief disappeared,

c (2+2+2)

Dal cor, dal mio pensier,

from my heart, from my thoughts,

Si dileguò il terror… Si,

The terror vanished… Yes,

Bel raggio lusinghier,

Beautiful flattering ray,

A major


Di speme e di piacer

Of hope and pleasure

Alfin per me brillò:

Finally shone for me:

la calma questo cor

Calm this heart


Arsace renderà

Arsace will return

Arsace ritornò,—Sì, a me verrà.

Arsace returned,—Yes, it will come to me.

The chorus intervenes again shortly in a pertichini77 section, to proceed to the Allegretto, which makes up what is soon to become known as a typical cabaletta in A-A’ form, with one more choral intervention before the reprise. The sweet thought of Arsace’s love prevails, grief and sadness fade away, and the sparkling music conveys Semiramide’s sense of cheerful anticipation.


Dolce pensiero,

Sweet thought,



Di quell’istante,

Of that moment,

A te sorride

On you it smiles,

L’amante cor.

The loving heart.

Come più caro,

How much dearer,


Dopo il tormento,

After the sorrow,

È il bel momento

Is the beautiful moment

Di pace, e amor.

Of peace and love.

c(4+4+1) fermata


Dolce pensiero, …


The Andante grazioso section features a typical declamato-like vocal style, with a first six-bar (2+2+2) richly flourished phrase resting briefly on a dominant chord. The next four bars develop into a more assertive chordal figure that sets to music the words ‘Arsace ritornò, sì a me verrà’ (Arsace is back, yes, he will come to me), to which a more doleful phrase follows in the relative minor key of F# minor, to express the past trembling and sobbing. Having resumed the original key, the melody moves towards a more lyrical concluding motif. The Allegretto is characterised by an anacrusis figure that develops into a number of florid passages, which, according to the practice, singers used to change, as soon as they would present themselves for a second time.

‘Bel raggio lusinghier’ and Nineteenth-Century Performance Practice

When Semiramide was first performed in London in 1824, Rossini himself held the baton. The critic of The Times expressed profound scepticism and thought that the opera would never be popular.

The plot possesses more interest than is usually found in the subject dramatized by the modern Italian poets, who seem to think that if they produce a given number of verses, no matter how trite and trifling, their work is done; the rest is left to the ingenuity of the composer, whose pleasant duty it becomes to clothe these poetical abortions in the rich robes of tasteful harmony. […] The music of this opera is of an elevated and heroic character. Rossini felt that the personages introduced were ‘the honourable of the earth’—princesses, princes, and warriors; and he has endeavoured very successfully to give them strains befitting their proud and lofty fortunes. We think, however, that the opera never will be popular. It is from beginning to end too abstrusely and elaborately scientific to please the million. To those who are well acquainted with the science of harmony, many of the marches and choruses will afford great delight; but there is little in the opera that can give pleasure to the lover of pure melody.78

About Giuditta Pasta, who made her appearance in the title role, the critic had little to say: ‘Madame Pasta represented Semiramide with admirable effect. She wore her royal robes with dignity and looked “every inch a queen”’. Remorini, who sustained the character of Assur, was a worthy partner: ‘We have rarely heard a singer with a voice of equal depth, who could introduce, with so much success, so many ornaments as this gentleman does’.79 Interestingly, the critic refers to the bass role introducing new ornaments but fails to consider the soprano and the tenor; García’s impersonation was said not to have been striking, although both arias for Idreno were maintained.80

Similar remarks were made by the critic of The Harmonicon who accused Rossini of blatantly borrowing from himself as well as many other composers, and to such an extent that his last opera may easily have been called a pasticcio.81 To start with, for the opening symphony Rossini was ‘indebted to the Pastoral Symphony of Beethoven, and the first subject of the last movement is very like a popular melody of Mozart’. The second part of the aria given to Idreno, ‘Ah dov’è il cimento?’ was said to be an inversion of Michele Carafa’s favourite aria ‘Fra tante angoscie’: the cavatina for Semiramide had not the slightest pretence of quality; as for the ghost scene, the part was an imitation of ‘the last incomparable scene in Don Giovanni’ with passages copied from Mozart’s music; the ‘Preghiera’, although ‘its great beauty and pathos are undeniable, and the sentiment is most judiciously expressed’, draws on ‘E amore un ladroncello’ from Così fan tutte, and on a cavatina in Winter’s Proserpina.82 The opera lacked originality, was far too long and showed clearly that the composer had out-Germanised the Germans. In Semiramide the orchestra was louder and heavier than any other predictably heavy German operatic work.83 Not much was said of the singers and their interpretation.

Having in another department of this work spoken at large of the musical merits of this opera, we have nothing to add upon that subject here. Madame Pasta is the main support of the piece, and is ably seconded by Madame Vestris. Signor García has little that is worthy of him to do, and Signor Remorini, who sustains his part well, gains no attention.84

Nor do we learn much from the review that The Times published in 1828, when Pasta was Semiramide again at the King’s Theatre: ‘Madame Pasta, as the Babylonian Queen, was on Saturday as powerfully effective as on all former occasions. Such is the peculiar influence exercised by her intense conception of great characters, that her latest efforts always increase the previous impression made on her audience’.85

In 1829 it was Maria Malibran’s turn to sustain the role of Semiramide and it was inevitable for the critics to draw a comparison between her and her predecessor.

The vocal part of the character was given by Madame Malibran with great accuracy and effect. We might have said as much of her acting, had it not frequently forced comparison with the ablest personation of the character which has ever been witnessed on any stage,—a personation, no doubt, as fresh in the recollection of every opera frequenter who has seen it within the last five years at this theatre, as it is in our own. […] Madame Malibran gave much dramatic force to the address in the 12th scene, wherein the nation is convoked to swear obedience to the new King about to be proclaimed; but it wanted the majestic dignity and the gracefulness of deportment which made it so impressive and interesting a scene in the hands of Pasta; and her slender and juvenile appearance was completely destructive of the required illusion.86

In the following years, more comparisons would be drawn between Pasta and Malibran, and the very fact that a comparison was possible meant paying the latter a great compliment. We have to wait until Giulia Grisi’s impersonation in 1843 to have a first hint at those moments in the opera where the singer’s interpretative decisions involving changes and substitutions may have made a difference.

Grisi’s Semiramide may stand by her Norma; it is one of those grand representations which this artist can give, seeming as if she merely obeyed the impulses of her own nature. A splendid performance was her scena ‘Bel raggio lusinghier’, with its nice expression, its brilliant embellishments, its delicate shading. Really tragic was ‘La forza primiera’. It won a burst of energy directed against Assur,—we felt that it should have withered him; and yet how admirably did the vocal artist remain in spite of the impassioned actress. Seemingly hurried on by the rage of the moment, Grisi varied her embellishments when the air was repeated, and the effect was immense.87

Although it is not clear whether the critic, when talking of the cavatina, refers to the embellishments originally written by Rossini or those added by Grisi, in his review he draws attention to the cabaletta in the duet between Semiramide and Assur in Act II, ‘La forza primiera’. When in 1843 this duet had to be repeated, the changes and new embellishments belonged to the interpreters. Similar comments accompanied Grisi’s impersonation of Semiramide in 1846, when she was made the object of enthusiastic comments regarding her voice and dramatic impersonation: ‘there is no other that can bring that weight of tragedy, that grandeur of emotion, that irresistible torrent of passion, which Grisi can throw into her characters of the highest walk’.88 Grisi excelled in the brilliant ‘Bel raggio’ with the sparkling ‘Dolce pensiero’, and the conclusion of the duet with Assur, ‘La forza primiera’, was one of her most magnificent displays; however, no mention was made of the passages she may have inserted. Again, in 1847 The Times paid Grisi the usual compliments with regard to both the cavatina and the duet but no information can be gleaned from the review as to the addition of further embellishments.89 In the following decade Grisi would be Semiramide in London almost every year, and only in 1851 The Times made short mention of the cavatina: ‘The difficult cavatina “Bel raggio” was executed with remarkable fluency and one or two hardly perceptible waverings in the ornamental fioriture of the cabaletta, or quick movement, were overlooked in the fervour and animation of the whole’.90 Again, from these words it is not possible to ascertain whether the ornamental fioriture the critic was referring to are those written by Rossini. We can only assume that, according to the performance practice of the time, the most proficient interpreters were both able and willing to interpolate new passages and add new embellishments.

After Grisi retired it was Thérèse Tietjens who had to take over the difficult responsibility of being as credible a Semiramide as Grisi had been. Still, no reference can be found in the columns of The Times to any possible modification in the cavatina when in 1871 Tietjens and Zelia Trebelli-Bettini were pronounced the worthy successors of Grisi and Alboni in the roles of Semiramide and Arsace respectively.91 We have to wait for Adelina Patti to find some more explicit reference to the addition of embellishments and interpolation of passages in the cavatina, this time thanks to the intervention of the composer himself. According to Hermann Klein, when Adelina Patti decided to add Semiramide to her repertoire and sing it at the Kursaal Theatre in Bad Homburg on 22 August 1866, she managed to have Rossini himself write a new set of coloraturas and cadenzas.

It is worthy of mention, if only for the fact that during this Homburg visit she made her first essay in the part of Semiramide. It was more or less of an experiment, but Rossini wished her to try it, and provided her with three entirely new cadenzas written expressly for the occasion. Despite her success in this role, she did not sing it in London until ten years later. It had long been associated there with the names of two glorious tragic artists, Grisi and Tietjens (the latter now at her best), and the public naturally regarded it as belonging exclusively to the repertory of a dramatic soprano. Rossini thought otherwise.92

This information finds confirmation in other contemporary sources, also suggesting that Rossini himself had written some new coloraturas for Patti. In 1868 the Gazzetta musicale di Milano reported briefly on Patti’s appearance in Bad Homburg and mentioned the composer’s intervention with regard to the cavatina and the duets between Semiramide and Arsace.

Hombourg. Adelina Patti sang Semiramide for the first time and with brilliant success. It is well known that Rossini has adapted this cavatina expressly for her, together with the duets with the contralto. We can guarantee that the music of this opera fits wonderfully the voice of the famous songstress.93

When in 1878 Adelina Patti was Semiramide in London for the first time, singing at the Royal Italian Opera, she was said to excel in both dramatic skills and vocal style.94 Again, The Times made explicit mention of Rossini’s newly-written coloratura in the by-now celebrated cavatina:

That Madame Patti, in adding this essentially Rossinian opera to her already varied and extensive repertory, has earned fresh laurels cannot admit of a doubt. […] The great vocal triumphs of the evening were the cavatina, ‘Bel raggio lusinghier’, and the two duets with Arsace—‘Serbami ognor sì fido’ and ‘Ebbene a te ferisci’. The first is profusely embellished by Madame Patti with extraneous passages and cadenzas; but as all the changes and fioriture were composed expressly for her by the master himself, the use of them cannot be regarded as otherwise than legitimate.95

This information is reported also in Hermann Klein’s account of Patti’s career:

I remember the night well, more especially for two things—Patti’s magnificent singing of ‘Bel raggio’ with the new Rossini changes and cadenzas; and the extraordinary effect that she created with Scalchi in the famous duet, ‘Giorno d’orrore’. I thought the audience would bring the roof down.96

The evidence discussed so far is strongly suggestive of Patti’s intention not only to follow in Grisi’s and Tietjens’ footsteps but to surpass them all, although it is not clear how and when she may have asked Rossini to write those new embellishments.97 After her appearances in this role in 1866 and 1868 at Bad Homburg, in Germany, she sang Semiramide on many an occasion in the next decades, especially from 1880 to 1890. After London (Covent Garden), she would be Semiramide in New York (Academy of Music), Philadelphia, Chicago, St. Louis, Cincinnati, Boston etc.98 Unfortunately, Patti never recorded the cavatina, nor do these variants appear in the critical edition of Rossini’s Semiramide (Ricordi 2001). Instead, Patti seems to have taught Marcella Sembrich her Rossinian fioriture, and we should be able to find them incorporated, at least in part, in the recording the latter realised in 1908.99

Among Tetrazzini’s immediate predecessors we find also Nellie Melba. She was Semiramide on 12 January 1894 and again in 1895 at the Metropolitan Opera House in New York but never recorded the cavatina.100 Mathilde Marchesi includes this aria in her Variantes et points d’orgue (pp. 71–75) and offers a number of solutions. The same can be said of Ricci, whose transcriptions of embellishments from different singers include both the cavatina and the duets; unfortunately, it is not possible to ascertain to whom these variants belonged, since Ricci is not specific. Estelle Liebling also offers a few solutions in her Coloratura Cadenzas (1943), but, once again, she does not indicate whom these ornaments and variants come from.

‘Bel raggio lusinghier’ in Tetrazzini’s Recording

As already mentioned, Tetrazzini never performed Semiramide on stage during her career, even though she twice recorded the cavatina. The first recording was made in London on 2 November 1910 with Percy Pitt conducting the orchestra (matrix 4578f, Gramophone 2-053034, HMV DB 537, VB 15); the second, which has never been issued, was made on 11 May 1914, this time in Camden with W. Rogers conducting the orchestra (matrix C14818-1). Tetrazzini’s only available recording of ‘Bel raggio lusinghier’ is 4’18’’ long and presents a few substantial cuts (Figure 25).

Figure 25 shows the cuts present in Tetrazzini’s 1910 recording of ‘Bel raggio lusinghier’.

The recording opens with the orchestra playing the final introductory chords before Semiramide’s first entrance when the Andante grazioso is almost complete, at least as far as the vocal part is concerned. Only the final choral pertichini and orchestral coda are cut, to move immediately to the Allegretto. The fermata on the final orchestral chord represents a clear invitation for the interpreter to insert a cadenza, an instance of which, likely to have belonged to Marchisio, can be found in Ricci’s volume (Figure 26).101

Figure 26 presents the cadenza reproduced by Ricci that concludes the Andante grazioso.

Unfortunately, no trace of this cadenza can be found in Tetrazzini’s recording. The Allegretto is literally cut by half, and only the second part, A’, is present, a circumstance that deprives us of the cadenza with which the A section concludes. For this cadenza many solutions can be found in the sources we have taken into consideration so far, that is to say Ricci, Marchesi, etc. In the second part of the cabaletta three bars are also cut when the vocal line approaches its close, and the recording ends on the chord that supports the singer’s last note.

A closer look at Tetrazzini’s and Sembrich’s recordings (2 November 1908, matrix C6573-1, Victor 88141, 6356; Gramophone 053217, HMV DB 433) should allow us to check whether and to what extent any relationship between the two divas exists and, if so, whether it is correct to assume that what Rossini wrote for Patti in the late 1860s and Patti handed over to Sembrich ever reached Tetrazzini. Again, the results show a mixture of old and new, with Tetrazzini featuring a number of passages that seem to have originated from personal choice rather than the model offered by Sembrich.

As can be observed in the opening bars, Tetrazzini’s newly-inserted passages follow Sembrich’s only to a limited extent; while the initial triplets at bar 3 are the same, the next long passage is clearly different from the one sung by the Polish soprano (Figure 27). Moreover, were the triplets at bar 3 from Rossini? It is impossible to say.

Figure 27 shows Sembrich’s and Tetrazzini’s recorded renditions of the initial measures of ‘Bel raggio lusinghier’.

Similar solutions are present in the following bars, tentatively suggesting a line of continuity between the two interpreters (Figure 28).

Figure 28 shows Sembrich’s and Tetrazzini’s recorded renditions of ‘Bel raggio lusinghier’ (measures 4 to 9).

However, this line of continuity is blurred by a number of differences, as is strongly suggested by the close at bar 13 (Figure 29). Here, not only do Tetrazzini and Sembrich differ from each other, but their recordings also bear evidence of a marked individuality when compared to what both Ricci102 and Marchesi103 report.

Figure 29 shows Sembrich’s and Tetrazzini’s recorded renditions of ‘Bel raggio lusinghier’ (measures 11–15).

The closing phrase leads to a cadenza-like suspension, where all sources seem to converge into a similar solution, despite a few minor differences (Figure 30).

Figure 30 shows Sembrich’s and Tetrazzini’s recorded renditions of ‘Bel raggio lusinghier’.

As already observed, both Tetrazzini’s and Sembrich’s recordings are cut, and no trace of the long final cadenza reported by Ricci can be found in either interpretation. A second cut concerning the Allegretto is present in each recording: section A is not present and the cadenza that bridges it with the reprise is also missing. Of this cadenza a number of examples can be found reproduced in both Ricci104 and Marchesi105 (Figure 31). Similarly, small differences appear in the Allegretto, which seem to suggest only a faint line of continuity between the two interpreters.

Figure 31 shows the traditional cadenzas before the reprise of ‘Dolce pensiero’.

Tetrazzini’s rendition includes a distinctive element in the final bars, when she inserts a diatonic descending-ascending passage to prepare the trill and lead to the final high A. This is not present in Sembrich’s recording (Figure 32).

Figure 32 shows Sembrich’s and Tetrazzini’s recorded renditions of the final cadenza of ‘Dolce pensiero’.

In conclusion, the idea that by drawing a comparison between Tetrazzini and Sembrich it would be possible to measure the distance between the former and a tradition that dated back to Rossini himself is only in part supported by the phonographic evidence discussed so far. While a few similarities may suggest a connection between the two divas and their renditions, many discrepancies indicate the opposite, and suggest that Tetrazzini may have followed a different, more personal path. Nor is it possible to ascertain whether and to what extent Sembrich’s 1908 recording may have incorporated the Rossinian variants Patti may have passed over to her. Furthermore, many similarities in the choice of coloraturas, passages and cadenzas may easily stem from a consistent reading of the score, and a similar understanding of the harmonic language and the melodic contour among different musicians trained in the same tradition.

As already seen, more substitutions came to be part of Semiramide’s performance practice during the second half of the century; these can be found also in later sources; for instance, passages for the duet in the second act ‘Se la vita ancor t’è cara… La forza primiera’, can be found in Ricci,106 who reproduces a number of variants belonging to the Marchisio sisters and to Bottesini. Unfortunately, no trace of these can be found in Tetrazzini’s recordings.

The material discussed so far suggests how strong the connection between Tetrazzini and the nineteenth-century tradition was and, at the same time, how difficult it is to trace and identify the models that may lie behind her own interpretative choices. Barbiere must have offered Tetrazzini a wonderful opportunity not only for vocal display, but also to show off her skills as a delightful comedian. Her Rosina was said to be ‘bubbling over with fun from beginning to end’ and she showed no hesitation in sacrificing the beauty of her voice to the fun.107 On the other hand, the cavatina of Semiramide, which she never sang on stage, should be considered as one of those many operatic old time favourites which found in the newly-born recording studio the springboard for reviving their success, despite the fact that the opera had virtually disappeared from the repertoire until the end of the nineteenth century.

1 Gattey, Luisa Tetrazzini, p. 280.

2 Ibid., pp. 15–16.

3 This number represents an approximation, since each production of an opera could imply a different number of performances. For instance, we are informed of six performances of Il Barbiere when it was produced in London in 1908: 15 and 23 June, 4, 9, 17 and 22 July. See ibid., p. 102.

4 Luisa Tetrazzini, How to Sing, p. 95.

5 Ibid.

6 Gattey, Luisa Tetrazzini, p. 25.

7 See Janet Johnson, ‘Il barbiere di Siviglia’, in The Cambridge Compaioni to Rossini, ed. by Senici, pp. 170–74.

8 ‘L’edizione da voi intrapresa darà luogo (con fondamento) a molte critiche, poiché si troveranno in diverse opere gli stessi pezzi di musica: il tempo e il denaro che mi si accordava per comporre era sì omeopatico che appena avevo io il tempo di leggere la cosiddetta poesia da musicare’. Lettere inedite e rare di Gioacchino Rossini, ed. by Giuseppe Mazzatinti (Imola: Ignazio Galeati e figlio, 1892), pp. 174–75. The English translation can be found in The Cambridge Companion to Rossini, ed. by Senici, p. 81.

9 See Herbert Weinstock, Rossini (New York: Lilelight Editions, 1987), pp. 120–21. The full transcription by Michotte can be found reproduced in Luigi Rognoni, Gioacchino Rossini (Turin: Einaudi, 1977; first edition 1968), pp. 385–426.

10 See Philip Gossett, ‘Compositional Methods’, in The Cambridge Companion to Rossini, ed. by Senici, pp. 80–81.

11 These manuscripts have been transcribed and are reproduced in both the critical edition published by Ricordi in 2014 (edited by Alberto Zedda) and the one published by Bärenreiter in 2008 (edited by Patricia B. Brauner). See also Philip Gossett, Divas and Scholars, Performing Italian Opera (Chicago and London: University of Chicago Press, 2006), p. 315.

12 Again, see both the Ricordi 2014 and Bärenreiter 2008 critical editions.

13 The Times, 14 June 1858, p. 12.

14 The Times, 21 June 1858, p. 12.

15 Talking of Viardot-García’s Rosina in 1847, Robert Schumann wrote in his operatic notes from Dresden that ‘Viardot makes a great variation of the opera: not one melody is left untouched’. Robert Schumann, La Musica Romantica, edited by Luigi Ronga (Turin: Einaudi, 1942), p. 231.

16 Luigi Ricci, Variazioni-Cadenze Tradizioni (Milan: Ricordi, 1937), vol. I.

17 Laura Cinti-Damoreau, Méthode de chant composée pour ses classes du Conservatoire (Paris: au Ménestrel, 1849), p. 93.

18 Mathilde Marchesi, Variantes et points d’orgue (Paris: Huguel, 1900).

19 Estelle Liebling, The Estelle Liebling Book of Coloratura Cadenzas containing Traditional and New Cadenzas, Cuts, Technical Exercises, and Suggested Concert Programs (New York: Schirmer, 1943).

20 Embellished Opera Arias, ed. by Austin Caswell (Madison, WI: A-R Editions, 1989), vols VII–VIII.

21 Karin und Eugen Ott, Handbuch der Verzierungskunst in der Musik, Band 4, Die Vokalmusik im 19. Jahrhundert (Munich: Ricordi, 1999).

22 Scott L. Balthazar, ‘The Primo Ottocento Duet and the Transformation of the Rossinian Code’, Journal of Musicology, 7.4 (1989), 471–97.

23 Manuel García, The Art of Singing (Boston: Ditson, [n.d.]), p. 59. This passage is to be found in the Italian translation by Alberto Mazzucato and published by Ricordi as early as 1842. ‘Nei duetti, le due parti possono combinar le loro fioriture; ma nei pezzi concertati é proibito il più lieve cangiamento’, p. 39.

24 Manuel García, Hints on Singing by Manuel Garcia, Translated from the French by Beata Garcia (New York: Ascherberg, 1894), p. 64.

25 García, L’Arte del Canto (Milan: Ricordi, 1842), p. 40.

28 These examples can be found in García’s Hints on Singing (London: Ascherbg, 1894), p. 65. See also Caswell, Embellished Opera Arias, p. 96.

29 See Hilary Poriss, Changing the Score: Arias, Prima Donnas, and the Authority of Performance (Oxford: Oxford University Press, 2009), p. 135.

30 Ibid., p. 151.

31 Tutti i libretti di Rossini, ed. by Marco Beghelli and Nicola Gallino (Turin: UTET, 1995), p. 380.

32 ‘Her Majesty’s Opera’, The Times, 26 April 1875, p. 9.

33 On 13 April 1876 The New York Times remarked that she had completed her musical education with Maurice Strakosch, whose role, as we will see, was of pivotal importance in launching Adelina Patti’s early career.

34 Herbert Weinstock, Rossini (New York: Limelight, 1987), p. 276 (first edition New York, Knopf, 1968, p. 66). Also cited in John Fr. Cone, Adelina Patti Queen of Hearts (Portland, Oregon: Amadeus Press, 1993), p. 56.

35 Camille Saint-Saëns, Musical Memories, translated by Edwin Gile Rich (Boston: Small, Maynard & Co. 1919), p. 207.

36 Weinstock, Rossini, p. 349.

37 The Times, 11 May 1863, p. 12.

38 In the 1850s and 1860s James William Davison was chief music critic of The Times. Although reviews were published anonymously at that time, it is possible to argue that the same person was behind the two reviews taken in consideration, possibly Davison himself.

39 The Times, 20 June 1895, p. 6. Interestingly, and to the critic’s deepest disappointment, on the same occasion, a vocal waltz by Tito Mattei was inserted into the final scene of the opera.

40 Poriss, Changing the Score, pp. 160–61.

41 The Times, 1 June 1912, p. 8.

42 San Francisco Bulletin, cited by Charles Neilson Gattey, Luisa Tetrazzini, p. 46.

43 Blanche Partington, ‘Throws Tivoli into Giggles’, The San Francisco Call, 4 October 1905, p. 2.

44 ‘The Opera’, The Musical Times, 1 July 1908, pp. 456–57.

45 ‘New York Opera Season’, The Times, 17 November 1908, p. 7.

46 ‘Royal Opera’, The Times, 14 May 1910, p. 12.

47 ‘Royal Opera’, The Times, Thursday, 1 June 1911, p. 10.

48 See Cone, Adelina Patti, p. 315.

49 The so-called Rossini Renaissance dates back to the late 1960s when an increasingly larger number of operas was staged and performed. See Charles S. Brauner, ‘The Rossini Renaissance’, in The Cambridge Companion to Rossini, ed. by Senici, pp. 37–38.

50 See Luigi Ricci, Variazioni-cadenze tradizioni (Milan: Ricordi, 1937), I, pp. 3–9. For an in-depth analysis of the intepretative tradition of Rossini’s repertoire see Michael Aspinall, ‘Il cantante nelle interpretazioni delle opere rossiniane’, Bollettino del centro rossiniano di studi, 1 (1970), 11–21, and, more recently, Rosina e la ‘phonographic evidence’: analisi di alcune variazioni vocali presenti nelle registrazioni storiche, ‘Opera Click’, http://www.operaclick.com/print/speciali/rosina-e-la-%E2%80%9Cphonographic-evidence%E2%80%9D-analisi-di-alcune-variazioni-vocali-presenti-nelle-regi#_ftn13.

51 Ricci, Variazioni-Cadenze, p. 4.

52 Marchesi, Variantes et points d’orgue, p. 13.

53 Ricci reports also some variants for contralto and indicates Marchisio’s name (p. 5).

54 It is my personal opinion that Ricci is not entirely trustworthy. Many of Tetrazzini’s cadenzas can be found reproduced in his volumes although he does not mention her name. On the other hand, Tetrazzini, whose recording had been long available on disc, was still alive (she passed away in 1940) when Ricci’s collections were published. Michael Aspinall reports that Ricci’s students confirmed that Ricci transcribed many variations from discs that were commercially available, and among them those left by Tetrazzini, which remained in the catalogue of His Master’s Voice until 1956.

55 The same can be said of other singers whose repeated recordings of the same arias show evidence of the consistency of their interpretative choices over time. Marcella Sembrich and Nellie Melba are two cases in point.

56 ‘Making of a Great Singer’, The Sun, 8 March 1908, p. 6.

57 Of course, Tetrazzini was adhering to a tradition that involved a number of cuts, and it is only by examining further evidence from the music scores used by her and her conductors that we can shed further light on this question.

58 Caswell, Embellished Opera Arias, p. 113.

59 Ricci, Variazioni-Cadenze, p. 6

60 See Appendice I, ‘Varianti vocali autografe’ of the new critical edition of Rossini’s Il barbiere di Siviglia, edited by Alberto Zedda (Milan: Ricordi, 2014), p. 570. See also Philip Gossett, Divas and Scholars, Performing Italian Opera (Chicago: The University of Chicago Press, 2006), p. 329.

61 García, Trattato di canto, Part II, p. 46.

62 See Charles N. Gattey and Thomas G. Kaufman, ‘A Chronology of Tetrazzini’s Appearances’, in Gattey, Luisa Tetrazzini, pp. 280–326.

63 Ambroise Thomas, Mignon, Opéra in Trois Actes, Paroles de MM. Michel Carré et Jules Barbier, Traduzione di Giuseppe Zaffira (Paris: Heugel). The Italian translation of the libretto was published separately in 1912 by Casa Editrice Madella in Sesto San Giovanni (Milan) and presents an almost identical reading of this aria (p. 24).

64 See Poriss, Changing the Score, pp. 156–65.

65 A third recording was realised on 11 July 1911 (Matrix ac5162f) but never issued.

66 Artists’ Vocal Album (Boston: White, Smith & Co., 1887), p. 70.

67 In this collection, the aria is in B flat major.

68 Again, the second variation is missing, but the aria is split into the two sides of the disc and is ca. 4’38’’ long. https://archive.org/details/1HempelProchVariationen/2+Hempel+Proch+Variationen.wav.

69 Estelle Liebling, Arrangements and Editings for Coloratura Voice, Series I (New York: G. Schirmer, 1940), pp. 3–14.

70 Lèo Delibes, Lakmé, Opéra en Trois Actes, Poème de MM. Edmond Gondinet et Philippe Gille (Paris: Heugel, [between 1883 and 1889]).

71 Heather Hadlock, Tancredi and Semiramide’, in The Cambridge Companion to Rossini, ed. by Senici, p. 142.

72 See Tutti i libretti di Rossini, ed. by Marco Beghelli and Nicola Gallino (Turin: UTET, 1995), pp. 735–56.

73 See the preface to Gioacchino Rossini, Semiramide, critical edition edited by Philip Gossett and Alberto Zedda (Milan: Ricordi, 2015), p. xxxiii.

74 Richard Mount-Edgcumbe, Musical Reminiscences, Containing an Account of the Italian Opera in England from 1773 (London: J. Andrews, 1834), p. 153.

75 Weinstock, Rossini, p. 129.

76 The critical edition includes a first draft of the cavatina, which Rossini abandoned before orchestrating it, where the aria follows a single-section structure. See Appendix II in Rossini, Semiramide, Critical Edition (Milan: Ricordi, 2015).

77 The expression con pertichini (with pertichini) was used to indicate the intervention of secondary characters or the chorus, whose comments were interpolated in the main aria or in the recitativo.

78 The Times, 19 July 1824, p. 2.

79 Ibid.

80 See Rossini, Semiramide, Critical Edition, p. xxxix.

81 The Harmonicon, XX, August 1824, pp. 162–64.

82 Ibid. We can fairly assume that the author refers to Il ratto di Proserpina, a serious opera in two acts by Peter von Winter, set to a libretto by Lorenzo Da Ponte.

83 The Harmonicon, XX, August 1824, pp. 162–64 (p. 164).

84 Ibid., p. 167.

85 The Times, 21 April 1828, p. 2.

86 The Times, 3 June 1829, p. 3.

87 The Times, 21 April 1843, p. 4.

88 The Times, 3 July 1846,

89 The Times, 7 July 1847.

90 The Times, 4 April 1851.

91 The Times, 6 November 1871.

92 Herman Klein, The Reign of Patti (London: T. Fisher Unwin, 1920), p. 173.

93 ‘Hombourg. Adelina Patti cantò per la prima volta, e con brillante successo, la Semiramide. È noto che Rossini ha espressamente ridotto per lei la cavatina e i due duetti col contralto. Assicurasi che la musica di quest’opera convenga mirabilmente alla voce della rinomata cantante’. Gazzetta musicale di Milano, 6 September 1868, p. 295.

94 The Times, 12 July 1878.

95 The Times, 23 July 1878.

96 Klein, The Reign of Patti, p. 194.

97 See also Henry Sutherland Edwards, The Prima Donna: Her History and Surroundings from the Seventeenth to the Nineteenth Century (London: Remington and Co., 1888), p. 123. See https://archive.org/details/primadonnaherhis02edwa?q=adelina+patti.

98 See John Frederick Cone, Adelina Patti: Queen of Hearts (Portland, OR: Amadeus Press, 1993), pp. 305–83.

99 See Giancarlo Landini, ‘Patti, Adela Juana Maria’, Dizionario Biografico degli Italiani—Volume 81, Treccani (2014) http://www.treccani.it/enciclopedia/patti-adela-juana-maria-detta-adelina_(Dizionario-Biografico)/. See also Michael Aspinall, ‘Rossini, il bel canto e Adelina Patti’, Musica, 119 (2000), 58–61.

100 Thérèse Radic, Melba the Voice of Australia (St. Louis, MO: Magnum Music-Baton, 1986), pp. 189–98.

101 Ricci, Variazioni, vol. I, p. 72.

102 Ibid.

103 Marchesi, Variantes et points d’orgue, p. 71.

104 Ricci, Variazioni, p. 72

105 Marchesi, Variantes et points d’orgue, p. 73.

106 Ricci, Variazioni, vol. I, pp. 73–77.

107 ‘Throws Tivoli into Giggles’, The San Francisco Call, 4 October 1905, p. 2.

Powered by Epublius