Rossini’s Petite Messe solennelle and Its Several Versions1



The Composition and First Performances of the Petite Messe solenelle

The history of the Petite Messe solennelle has been told many times. In the standard account, most of which is accurate, Rossini first composed the Mass in 1863, during his regular summer sojourn in his villa at Passy, in the suburbs of Paris. The autograph manuscript of the mass, preserved at the Fondazione Rossini of Pesaro, does indeed bear a total of three autograph dedications (one to the Countess Louise Pillet-Will and two addressed to “Bon Dieu”): all three are dated “Passy, 1863.” A more precise indication is found at the end of the Credo movement, which is signed and dated: “G. Rossini / Laus Deo (Passy) / 10 Juin 1863.”

The news that Rossini had completed a major sacred composition during the summer of 1863 caused a sensation. By August notice of the new work had even made its way across the Atlantic to New York, where the “Aesthetic Magazine” Once a Month (“devoted to music, painting, poetry, and kindred arts”) published this account:

Rossini’s Mysterious Mass.---
Rossini has positively finished a grand mass with choruses, but declined to permit the execution of the work when asked by Baron Isidore-Justin-Sévérin Taylor, the President of the Association of Artist-Musicians in Paris. Is it a Requiem, and for whom is it destined?
Time will show.2

Baron Taylor did not have to wait long: he was to be among those attending the first official performance of the Mass some eight months later.3

Conceived for a total of twelve voices (four soloists, who also sing with the chorus, and eight additional choristers), the Mass was scored for two pianos and harmonium. While there is no evidence that it was actually commissioned by Rossini’s bankers and close personal friends, the Count and Countess Alexis and Louise Pillet-Will, the dedication of the Petite Messe solennelle to Countess Louise in the Pesaro autograph is dated “Passy, 1863,” and the work was first performed before an invited public at 10:00 P.M. on Monday, 14 March 1864.4 The occasion celebrated the consecration of a private chapel in the splendid new home constructed by the Count and Countess at 12, rue de Moncey.5

The Count and Countess Pillet-Will and their family were intimate friends of Rossini from very early in his final stay in Paris, where he and his wife Olympe had arrived from Florence toward the end of May 1855. The next year, on his birthday, the composer wrote a lovely piano piece in the autograph album of Alexis’s father (still in the possession of the family), dedicating it: “A M.r Le Comte Frédéric Pillet-Will, son ami devoué, G. Rossini. Paris, ce 29 Février 1856.”6 During his final Parisian sojourn, Rossini’s financial affairs were largely handled by Alexis’s son, Michel-Frédéric Pillet-Will.7

Although the performance of the Petite Messe solennelle on 14 March 1864 was by invitation only, information about it circulated widely in the Parisian press. The female soloists were two of Rossini’s favorite singers, the sisters Carlotta and Barbara Marchisio, respectively a soprano and a contralto, for whom the Paris Opéra had revived Rossini’s Semiramide (in French) in a noteworthy performance of 1860. They were joined by the tenor Italo Gardoni, who had created the role of Carlo in Verdi’s I masnadieri of 1847, and the well-known Belgian bass Louis Agniez (Luigi Agnesi). According to Jacques-Léopold Heugel, the chorus actually consisted of fifteen conservatory students, in addition to the four soloists, so that (despite Rossini’s instructions) nineteen singers were involved in the premiere of the work.8 The ensemble, conducted by Jules Cohen, included Georges Mathias as the principal pianist9 and Andrea Peruzzi on second piano, with the eighteen-year old Albert Lavignac, who was to become an influential theorist and educator at the Conservatoire, playing the “Harmonicorde-Debain.”10

It has long been known that the music performed in 1864 did not include the soprano solo, “O Salutaris” (which appears in Rossini’s autograph manuscript after the Sanctus). It is noteworthy, however, that one contemporary review suggests that Rossini had intended the composition to be placed after the Credo, in the position ultimately filled by the “Prélude religieux pendant l’Offertoire”:

I believe that Rossini originally composed an “O Salutaris” which had been supposed to occupy the place in the score of this piano composition [the “Prélude religieux”]. This “O Salutaris” will not be lost for posterity; but, since one knows that the maestro over the past few years has written a large number of pieces for piano, let us congratulate ourselves on the good fortune that has allowed us to slip a glance at this treasure, and to appreciate the sovereign manner in which the great man has turned himself into a pianist.11

Apart from the absence of “O Salutaris,” however, there has never been any suggestion that the music performed in 1864 was different from the music we know today from Rossini’s autograph manuscript of the Petite Messe solennelle in its original setting for two pianos and harmonium.

The Petite Messe solennelle was performed once again at the home of the Pillet-Wills, the next year. Again, a dress rehearsal was held on a Sunday afternoon, in Rossini’s presence,12 with the formal performance following the next evening, Monday, 24 April 1865.13 The vocal soloists, instrumentalists, and conductor were identical to those who performed the work in 1864. According to contemporary reports, in fact, Count Pillet-Will brought the Marchisio sisters back from Florence expressly for the occasion. It is clear from the 1865 printed program, however, that the Mass still did not contain the “O Salutaris,” nor indeed is there any specific reason to suspect that the work performed in 1865 differed in any way from that performed in 1864.

Rossini may well have thought about the possibility of orchestrating the Petite Messe solennelle soon after these performances, indeed there were voices urging him to do so already in 1864,14 but there is no evidence that he had begun this work when, on 23 March 1866, he wrote to his Florentine friend Luigi Crisostomo Ferrucci, Director of the Biblioteca Mediceo-Laurenziana,15 about the possibility of Pope Pius IX permitting women to sing in church:

You may know that I composed a Messa Solenne, performed in a large hall at the home of my friend Count Pillet-Will, which excited much interest. The performance was perfect, the provisional accompaniment was for two pianos and a harmonium (organetto). Despite the entreaties of both the wise and the ignorant, I greatly hesitate about orchestrating it, to permit its performance in a large basilica, because of the lack of soprano and contralto voices (so-called voci bianche), without which it is impossible to sing the glories of the Lord.16

Despite a number of efforts to intervene with Pope Pius IX on this matter, nothing was to come of the initiative.

Nonetheless, Rossini decided to proceed with his orchestration of the Petite Messe solennelle, and contemporary reports indicate that he had completed the work by April 1867.17 It seems likely that this orchestration was largely accomplished during the first months of 1867.18 Having undertaken this effort to ensure that no one else would do it after his death, as he told numerous friends, the composer put away his autograph manuscripts to both versions of the Petite Messe solennelle, the version for two pianos and harmonium and the version with full orchestra, allowing no further performances of either version during his lifetime.

After Rossini’s death, on 13 November 1868, his widow Olympe Pélissier sold performance rights to the Petite Messe solennelle to the impresario Maurice Strakosch, who was responsible for the first performance of the orchestral version, on 24 February 1869, at the Théâtre Italien in Paris. The soloists included Gabrielle Krauss, Marietta Alboni, Ernest Nicolini, and the same Luigi Agnesi who sang in the original performances. Strakosch later arranged for performances throughout Europe and elsewhere.19 A printed full score was issued by the firm of Brandus & Dufour, and a number of closely related scores for one piano, harmonium, and voices were prepared by Brandus & Dufour in Paris, Chappell in London, Ditson in Boston, and Ricordi in Milan. Subsequently, Brandus & Dufour issued a revised edition, which was the basis of a score published by Schott in Mainz, also for one piano, harmonium, and voices.20

With the death of Olympe herself in 1878, the Rossini manuscripts still in her possession were transferred to the Fondazione Rossini of Pesaro. They included the two complete autograph manuscripts of the Petite Messe solennelle: the chamber version for two pianos and harmonium and the orchestral version. It has been commonly affirmed ever since that the work exists in only these two authentic versions. But we now know that this is not true.

Although no one has ever suggested that the music preserved in Rossini’s autograph manuscript of the Petite Messe solennelle in its chamber setting might not have been identical to the music performed at the Pillet-Wills in 1864 and 1865 (with the exception of the later addition of the “O Salutaris” movement), several important contributions over the past twenty years to the history of the Petite Messe have complicated the matter. In 1980, Angelo Coan, editing the score for Edizioni musicali OTOS in Florence, pointed out that there had never been a complete edition containing both pianos and harmonium. Worse, the Ricordi arrangement circulating since the 1880s, with one piano and harmonium, did not reproduce Rossini’s manuscript: it was actually a new keyboard reduction of the orchestral score. Only with Coan’s edition, then, did it finally become possible to perform Rossini’s chamber version according to the readings of the autograph manuscript in Pesaro.21

At about the same time, Nancy P. Fleming began investigating all available sources more fully. She was surprised to learn that most of the early printed scores with accompaniment for one piano and harmonium (the first Brandus & Dufour edition, Chappell, Ditson) actually differed in a number of significant ways from the Pesaro manuscript. She noticed, in particular, that among other “idiosyncracies,” all these editions make some noteworthy cuts. She mentions “a bar at the end of the ‘Domine Deus,’ the first seven measures and one at the end of the ‘Qui Tollis,’ nineteen bars of postlude to the ‘Quoniam’.” But she draws the wrong conclusion from these observations: she assumes that the cuts were imposed by the editors, and that “such amputation must have facilitated excerpting the movements to sell as individual numbers.”22

Work on the Péchés de vieillesse, music of Rossini’s last Parisian period, for the Edizione critica delle opere di Gioachino Rossini has taught us to be wary of such assumptions. Whereas the autograph manuscripts of the music Rossini composed during his years in the theater are remarkably clean, rarely revealing significant compositional changes, autograph manuscripts of the Péchés de vieillesse are miniature battlefields. These compositions were the product of a process of careful and, often, continuous revision, carried out over an extended period of time. The sixth setting of “Mi lagnerò tacendo” in the Musique anodine, for example, was drafted as early as 29 February 1852, but did not assume its final state until 1857.23 The “Chœur de Chasseurs Démocrates” from the Album français, for which a considerable number of sources survive, can be shown to have undergone at least eight independent layers of corrections, each associated with work on specific manuscripts or performing parts.24

Even a casual glance at the autograph of the “original” version of the Petite Messe solennelle reveals that Rossini made innumerable alterations during the course of his compositional labors. In many cases his work involved scratching out earlier layers and writing his revisions directly over the canceled material. In other cases sections with heavy revisions are immediately followed or preceded by an absolutely pristine passage on a separate sheet of paper: when his revisions grew so numerous as to threaten the legibility of the later version, in other words, Rossini replaced an earlier page or pages with new paper on which he would write the new music.

When did he make these revisions in the “original” version? Did they all predate the performances of 1864 and therefore belong to the pre-history of the Petite Messe solennelle? Or did some of them occur after the work was originally performed? Could they have been made between the performances of 1864 and those of 1865? Or were they entered after April 1865, the last performance during Rossini’s lifetime, as the composer contemplated and then brought to fulfilment his plan to orchestrate the Mass? In that case, were they made together, as part of a final revision, or on a number of different occasions? How many discrete stages, in short, can be identified in the history of the “original” version of the Petite Messe solennelle?

Only the survival of multiple sources for the Mass, as in the case of the “Chœur de Chasseurs Démocrates” from the Album français mentioned above, could help us to clarify such questions. Until now, however, the Pesaro autograph manuscript and the posthumous editions for one piano, harmonium, and voices were the only known sources. Yet logic tells us that there must once have existed a rich array of contemporary sources: performing parts used by the soloists, chorus, and instrumentalists in 1864 and 1865; a master copy employed by the conductor Jules Cohen (unless he conducted from Rossini’s autograph, which seems most unlikely); and copies of the “original” version that served for the Brandus & Dufour edition of 1869 (which was not merely a reduction of the orchestral score).25

It is in this context that the survival of a manuscript of the Petite Messe solennelle in a copyist’s hand, presented to the work’s dedicatee, Countess Louise Pillet-Will, must be understood. We do not know when Rossini made his gift. It is likely that the title page would have borne an autograph dedication and perhaps a date, but it is precisely that title page, together with the first seven bars of the Kyrie (surely written on the verso of the title page), that are missing today. Otherwise the manuscript is complete. Unlike Rossini’s autograph manuscript, in which the principal score contains only the vocal parts and the first piano, with the harmonium part and the second piano part in separate fascicles, the Pillet-Will copy includes the vocal parts, first piano, and harmonium in the principal score; only the second piano part is in a separate fascicle.26

Although we do not know precisely when the Pillet-Will manuscript was copied, it unquestionably represents what Rossini believed to be a completed version of the Petite Messe solennelle. It seems unlikely that the composer would have presented the manuscript to the Countess before the dedication of the chapel in March 1864. Nor is it likely that the presentation significantly postdates this performance. The music copied into the Pillet-Will manuscript, in short, probably presents the score as it was first performed, and the manuscript itself was probably completed no later than the summer of 1864.27

The Pillet-Will copy differs in numerous respects from the music of the Pesaro autograph. In many cases its readings can still be discerned in the autograph, where they have been subsequently scratched away and replaced. In other cases its readings could not have been reconstructed from Rossini’s autograph, since the composer himself removed the relevant pages from his manuscript and replaced them entirely. While the differences are not enormous, and the most significant ones pertain to the instrumental passages that open and close individual sections, they have a cumulative effect: the readings of the Pillet-Will manuscript give the vocal lines a simpler frame, one even more appropriate for its original setting as a salon work than the version preserved in the Pesaro autograph.

Here is a summary of the major differences in the Petite Messe solennelle as it is preserved in the Pillet-Will manuscript.

The movement is essentially the same: there are only a few differences of detail in the spacing of chords and rhythm in the piano and harmonium parts.

The introductory section ("Gloria in exclesis Deo
) has a somewhat simpler accompaniment. The end of the Et in terra paxsection also has a simpler accompaniment.

The end of the section has a less rhetorically pointed accompaniment.

Do-of Domineat the end of the section is held for one less measure (two instead of three in Pesaro autograph).

The first seven, highly chromatic measures of the Pesaro autograph are not present: the music begins immediately with the F minor harmony. In addition, the five concluding measures of the Pesaro autograph are shorter by two measures and are simpler in the Pillet-Will copy.

The twenty-bar instrumental postlude of the Pesaro autograph, with its elaborate modulation to the new key of the
Cum sancto spiritu,is absent: in its place we find only two measures on the tonic.

The introductory section has a somewhat simpler accompaniment (as in the opening of the
Gloria). The choral parts in this passage are significantly different.

There are some differences in the harmonium part at the end of the

The second piano part is diverse at the beginning of the
Et resurrexit. In the cadential section of the fugue the instrumental bass is significantly different for eight measures (a passage subsequently repeated). An important choral cadential phrase, also played twice, is less extended: it is only eight measures long (rather than the twelve measures of the Pesaro autograph). Most important, the concluding measures of the Credo are less extended and less rhetorically emphatic than in the Pesaro autograph.

In the Pesaro autograph, Rossini rewrote the vocal parts near the conclusion and introduced a more frequent alternation of soloists and chorus than in the Pillet-Will manuscript.

This movement did not figure in the Pillet-Will performances of either 1864 or 1865, and it is not present in the Pillet-Will manuscript.

Only the final measures are different in the Pillet-Will manuscript. The modifications in the Pesaro autograph once again provide a more rhetorically emphatic conclusion.


In practically every case the Pillet-Will manuscript offers readings that are less rhetorically emphatic than the Pesaro autograph. Why would Rossini have made these changes and when? On the basis of the sources currently at our disposal, it seems most likely that the changes were introduced only after the 1864 and 1865 performances, and that they (like the addition of the “O Salutaris” movement) were intended to enrich the work for its eventual orchestration. There is, in fact, some evidence that the process was undertaken in two stages: the Brandus & Dufour edition for one piano, harmonium, and voices (and with it the Chappell and Ditson editions) agrees in some places with the Pillet-Will manuscript and in some places with the Pesaro autograph. This strongly suggests that the work, in its “chamber” version, was copied again after a first stage of revisions (perhaps in 1866?), and that this copy formed the basis for the Brandus & Dufour edition. Rossini would then have made a final set of revisions in what we know today as the Pesaro autograph while he orchestrated the Petite Messe solennelle.

With the recovery of the Pillet-Will manuscript, it has become possible for the first time for modern audiences to hear the Petite Messe solennelle as Rossini conceived and performed it for the dedication of the private chapel of the Count and Countess Pillet-Will in 1864. While the differences are not vast, they enhance the chamber-music quality of the original conception, and provide a fascinating insight into the art of Rossini’s last period.

CIAO home | CIAO Essays contents





1 This essay was originally published in Music Observed: Studies in Memory of William C. Holmes, ed. by Colleen Reardon and Susan Parisi (Warren, Michigan, 2004), 139-146.
No one knew better than Bill Holmes the pleasure of tracing an early version of a major work: I will never forget his delight at having identified Verdi’s first ideas for the St. Petersburg version of La forza del destino. This essay relates a similar experience that would have pleased Bill greatly. It derives from the time that my wife and I spent together at the home of Count Jacques Pillet-Will and his wife Elizabeth during the Fall of 1994, when they generously invited us to their home at the Chateau d’Offémont in Saint-Crépin-aux-Bois. Those days together were an experience we will never forget. Jacques, himself an accomplished musician, shared with us his deep love for music, his justified pride in his family’s ties to Rossini, and the manuscript materials that made possible these discoveries. To my deep regret Jacques did not live to hear the first performances of the earliest version of Rossini’s Petite Messe solennelle at the Rossini Opera Festival of Pesaro during the summer of 1997: he passed away in the Fall of 1996. An earlier version of this essay was written for the program book accompanying those performances in Pesaro.

2 Once a Month, ed. by Richard Storrs Willis, August 1863 (Vol. 2, No. VIII).

3 The Baron’s name figures among guests mentioned in an article about the premiere that appeared in Le Ménéstrel on 20 March 1864.

4 Although Rossini did not attend this semi-public performance, he participated in the dress rehearsal, held the previous afternoon in the Pillet-Will home, in the presence of a very small group of friends and musicians.

5 The home was apparently completed in 1863. For further information about the family and various interesting documents, see the catalogue, edited by Jean-Marie Bruson, of the exhibition organized at the Musée Carnavalet of Paris (27 October - 31 December 1992): Rossini à Paris, 147-9, 164-5. The property at rue du Moncey was later acquired by a foreign embassy, and the home of the Pillet-Wills was razed.

6 According to Giuseppe Radiciotti, Gioacchino Rossini: vita documentata, opere ed influenza su l’arte, 3 vols. (Tivoli: Arti Grafiche Majella di Aldo Chicca, 1927-29), 2:436n, Frédéric, who died in February 1860, “was a magnificent protector of musical art and a musician himself; he formed friendships with the most famous composers and artists of his time; among the most assiduous of those frequenting the musical evenings he gave in his own home, was Rossini.”

7 See Bruson, Rossini à Paris, 164. I have found no documentary evidence suggesting that Rossini was intimately connected with the family before his return to Paris in 1855, but it seems likely that they (perhaps together with the Rothchilds) had already been representing his financial interests in Paris.

8 The founder of the major French music publishing house, Heugel claims to have followed the composition of the Mass at Passy during the summer of 1863. His account of the first performance appeared in the pages of his firm’s magazine, Le Ménéstrel, on 20 March 1864.

9 Mathias was a professor of piano at the Conservatoire, among whose students figured Erik Satie. The lineage is suggestive.

10 Alexandre François Debain was the chief builder of harmoniums and similar instruments in France from 1840 through his death in 1877. Notice that the instrument here specified is a “Harmonicorde,” or sustaining piano, and not a “Harmonium,” even though in his autograph score Rossini refers generically to a “Harmonium.” It would be interesting to learn more about these distinctions.

11 The remark is found in the review published in Le Pays: Journal de l’Empire of 20 March 1864.

12 As several newspaper accounts point out, that same evening the dress rehesarsal of Meyerbeer’s L’Africaine was held at the Opéra. (Its first performance took place on 28 April 1865.) But Meyerbeer, who was present at the first performance of the Petite Messe solennelle in March 1864, had died on 2 May 1864.

13 As in 1864, Rossini chose not to attend. On 26 April 1865 La France reported: “Rossini, who all eyes sought, once again withheld himself from the ovations that awaited him. Little concerned with new triumphs, he remained quietly at home; but Mme Rossini attended the soirée given by Count Pillet-Will.”

14 See, for example, the review appearing in the Courrier des Alpes, 29 March 1864.

15 Ferrucci seemed to Rossini a reasonable person with whom to open this conversation, since he had passed several years at the Biblioteca Vaticana in Rome and was well connected in the papal court. See the fundamental article by Stefano Alberici, “Rossini e Pio IX alla luce di documenti inediti dell’Archivio Segreto Vaticano,” in Bollettino del centro rossiniano di studi 1-2 (1977): 5-35.

16 Alberici, “Rossini e Pio IX,”13. Rossini goes on to explain that at present the situation was particularly difficult, since there were currently available neither sopranisti (castration was no longer permitted) nor women, and he abhorred the sound of boys whose voices had not yet changed. Alberici also reprints Rossini’s correspondence with Franz Liszt on the same subject.

17 Describing his visit to Rossini in that month, Emil Naumann refers to “the still-wet manuscript.” Naumann’s report is cited in Radiciotti, Gioacchino Rossini: Vita, 2:463-5 and Herbert Weinstock, Rossini: A Biography (New York, 1968), 345-6.

18 Thus, it would seem unlikely that the bills for copies pertaining to the Petite Messe solennelle mentioned in Bruson, Rossini à Paris, 166 (#269 e 270), pertain to the orchestral version. But neither is there any specific evidence that these bills, drawn up almost a year apart (13 July 1865 and 3 October 1866, respectively), are related to the presentation copy of the work destined for Countess Louise Pillet-Will, to be discussed below.

19 Weinstock, Rossini, 326, refers to other performances, including one in Moscow, featuring the Marchisio sisters, under the direction of Anton Rubinstein, and another in Australia.

20 The history of these early editions is discussed in detail by Nancy P. Fleming in her D.M.A. dissertation for the University of Illinois, Rossini’s “Petite Messe solennelle” (Urbana, 1986). See, too, her article by the same name in Choral Journal 30, 7 (February 1990): 15-21. Fleming’s performing edition of the Petite Messe solennelle (1991) is available from Oxford University Press.

21 Herbert Handt had recognized some of these problems already in the late 1960s. See his article “A proposito del Sanctus della Petite Messe” in Bollettino del centro rossiniano di studi 2 (1972): 5-9. His speculations about the order of pieces, however, find no support in the Pillet-Will manuscript to be discussed below.

22 Fleming, “Rossini’s Petite messe solennelle,” Choral Journal, 19.

23 See Musique anodine, Album italiano, edited by Marvin Tartak, in Edizione critica delle opere di Gioachino Rossini, Sezione VII, Vol. 1 (Pesaro, 1995), 29-34 and 39-44.

24 See Album français, Morceaux réservés, a cura di Rossana Dalmonte, in Edizione critica delle opere di Gioachino Rossini, Sezione VII, Vol. 2 (Pesaro, 1989). The discussion of the sources and their complex history, worked out by Patricia B. Brauner, is on pp. 350-2.

25 Only a close collation of the other early editions can clarify whether they were derived directly from the Brandus & Dufour edition or from still other manuscript copies of the original version.

26 It seems reasonable to suppose that the manuscript used by Brandus & Dufour for their edition was constructed in the same way.

27 It cannot be excluded, of course, that Rossini did not present the manuscript to the Countess until after the performances of April 1865, although the long delay would have been uncharacteristically impolite. Still, if that were the case, the copying bill of 13 July 1865 mentioned above could indeed pertain to the Pillet-Will copy of the Petite Messe solennelle.