It is a pleasure to recall certain picturesque details which interested me intensely during the preliminary rehearsals. First of all, the curious audacity of Lucy Arbell, our La Belle Dulcinée, in wanting to accompany herself on the guitar in the song in the fourth act. In a remarkably short time, she made herself a virtuoso on the instrument with which they accompany popular songs in Spain, Italy, and even in Russia. It was a charming innovation. She relieved us of that banality of the artist pretending to play a guitar, while a real instrumentalist plays in the wings, thus making a discord between the gestures of the singer and the music. None of the other Dulcinées have been able to achieve this tour de force of the creatrix. I recall, too, that knowing her vocal abilities I brightened the rôle with daring vocalizations which afterwards surprised more than one interpreter; and yet a contralto ought to know how to vocalize as well as a soprano. _Le Prophète_ and _The Barber of Seville_ prove this. The staging of the windmill scene, so ingeniously invented by Raoul Gunsbourg, was more complicated at the Gaîté, although they kept the effect produced at Monte Carlo. A change of horses, cleverly hidden from the audience, made them think that Don Quixote and the dummy were one and the same man! Gunsbourg's inspiration in staging the fifth act was also a happy chance. Any artist, even though he is the first in the world, in a scene of agony wants to die lying on the ground. With a flash of genius Gunsbourg cried, "A knight should die standing!" And our Don Quixote (then Chaliapine) leaned against a great tree in the forest and so gave up his proud and love lorn soul. CHAPTER XXVII A SOIRÉE In the spring of 1910 my health was somewhat uncertain. _Roma_ had been engraved long before and was available material; _Panurge_ was finished and I felt--a rare thing for me--the imperative need of resting for some months. But it was impossible for me to do absolutely nothing, to give myself up completely to _dolce farniente_, delightful as that might be. I looked around and found an occupation which would weary neither my mind nor heart. I have told you that in May, 1891, when the house of Hartmann went under, I entrusted to a friend the scores of _Werther_ and _Amadis_. I am speaking now only of _Amadis_. I went to my friend who opened his strong box and brought out, not banknotes, but seven hundred pages (the rough draft of the orchestration) which formed the score of _Amadis_ and which had been composed at the end of 1889 and during 1890. The work had waited there in silence for twenty-one years!