Ferdinand Hérold's 1827 ballet-pantomime La Somnambule, written for the Paris Opéra, is generally remembered today only as a source for Bellini's 1831 opera La sonnambula. However, Hérold's work – which also inspired a series of popular vaudevilles on the same theme – illustrates the strong, voyeuristic appeal of the trance phenomenon at the end of the Bourbon Restoration. It can be viewed as encapsulating wide-ranging contemporary ideas about the relationship between sleepwalking, mesmerism, madness and the supernatural. The aims of this article are twofold. First, it seeks to introduce important nuances into the received and often generalised claims usually made about the containing nature of trance scenes in nineteenth-century theatre, positing an alternative model to that of the unhinged heroine of Italian opera familiar from recent feminist writing on opera. Second, it illuminates the musical practices specific to late Restoration Paris that were so crucial to the aesthetic – and the success – of these sleepwalking heroines. A web of visual and musical allusions conjured up an entranced figure who, although related to the Italian operatic madwoman, has a personality and social implications all her own.

From 1807 to 1864, Parisian music drama was governed by a system of licences that controlled the repertory of its three main lyric theatres: the Opéra (variously Académie Royale, Nationale and Impériale de Musique), the Théâtre-Italien and the Opéra-Comique. Between 1838 and 1840, the Théâtre de la Renaissance gained a licence to put on stage music, and quickly succeeded in establishing a reputation for energetic management, imaginative programming together with artistically and financially successful performances. It could do this only by exploiting what were effectively newly invented types of music drama: vaudeville avec airs nouveaux and opéra de genre. The invented genres however brought the theatre into legal conflict with the Opéra-Comique and Opéra respectively, and opened up a domain of jurisprudence –associated with repertory rather than copyright – hitherto unsuspected.

This study of the ending customarily appended to Giacomo Puccini's unfinished Turandot offers a new perspective on its genesis: that of its principal creator, Franco Alfano. Following Puccini's death in November 1924, the press overstated the amount of music that he had completed for the opera's climactic duet and final scene. In fact, Puccini's manuscripts were so disjointed that Arturo Toscanini, the conductor chosen to lead the première, drafted the reluctant Alfano to fashion them into a viable conclusion. While occupied with this assignment, Alfano spoke with the writer Raymond Roussel about his plans for the opera's completion. This long-forgotten interview, absent from previous studies of Turandot's conclusion, reveals a strategy that would inevitably fall foul of Toscanini's expectations. Rejecting Alfano's first attempt for its extensive original composition, Toscanini forced changes on the conclusion that undermine both its musical coherence and dramatic logic. I assess Alfano's original ending in light of his frustration with Puccini's sketches, as well as the generally deleterious result of Toscanini's interventions. While neither conclusion represents an ideal solution, a judicious conflation of the two versions offers the best chance of reconciling a suitable denouement with the musical character of Puccini's finished score.

This article examines the new finale by Luciano Berio for Puccini's Turandot, commissioned by the Festival de Musica de Gran Canarias and prèmiered in 2002. This finale constitutes a priviledged point of view from which to revisit not only the composer's life-long engagement with musical works of the past, but also his theatrical aesthetics, thrown into sharp relief by the particular challenges of this endeavour. Berio's compositional process is carefully reconstructed, from his interventions on the libretto to his respectful yet creative use of the numerous sketches left behind by Puccini. Particular attention is then paid to Berio's choice of reminiscences from earlier scenes of the opera, as well as to his citations from Wagner's Tristan, Mahler's Seventh Symphony and Schönberg's Gurrelieder. A separate section is devoted to Berio's most original contribution: a long orchestral interlude at the moment when Calaf kisses Turandot. The article concludes with a detailed scrutiny of Berio's orchestration.