to her first concert. It was to be an experience, an object-lesson. Well—it was. They saw a fine-looking woman with a mediocre voice and a worse method, a method so hopelessly bad that even her undoubted musicianship could not atone for it.

All this goes to prove that a method, to be elastic and adaptable, should be based on a knowledge of the physiology of the voice-producing organs, for such a method naturally adapts itself to physical differences in different individuals. Without doubt Mme. Marchesi's method was admirably adapted to Melba, but not to Eames or to her own daughter.

Bear these circumstances in mind in selecting a teacher. The great singers are not always safe guides in the choice of a teacher, because their own superlative gifts and willingness to slave for the object of their ambition may have been as important factors in their success as the instruction they received. Probably a singer of only fair natural gifts who yet has made a success—which shows that he must have been well taught—can give better advice as to the choice of an instructor than the great artist who owes so much to himself. Moreover, great artists who have studied with the same teacher

will, like Melba and Eames, differ in their estimate of that teacher.

There is, however, one great singer, Lillian Nordica, who knows to whom to give credit for that skill in voice-production which enables her to sing Valentine, Aida and Isolde with equal success. The foundation for her career was laid in this country. Afterward she studied with Mme. Maretzek and in Milan with San Giovanni, but only interpretation. Her voice-production she acquired not from Madame this or Signor that, but from plain John O'Neill, of Boston, "a scholarly man who had made a profound study of the physiology of the voice," and she took good care not to allow any other teacher, however "famous," to undo the work of the man who had taught her voice-production based on correct knowledge of the physiology of the voice-producing organs.

This matter of choosing a teacher is, of course, of the greatest importance, but it barely can be touched on in this book. The selection should be made most cautiously, but, once made, the pupil's parents should not go to the teacher a few weeks later and ask, "Why don't you give Clara some 'pieces'?" They should recall the story of Porpora and Caffarelli which I related in the previous chapter.

"Pieces" are not in order until the voice is prepared for them, and the teacher is the best judge of that. A voice trained on "pieces" soon goes to pieces.

Another mistaken idea is that "any teacher is good enough for a beginner," whereas the beginning is the very time that the foundation of right method or wrong method is laid. Classifying the voice is, of itself, of great importance. Remember that Jean de Reszke's first teacher thought he was a baritone and that he sang as a baritone in opera for five years before a more competent teacher discovered that he was really a tenor. Some voices are so near the dividing line that it requires wide experience and a fine ear for quality on the part of a teacher to determine in what direction they should be developed to greatest advantage. A fine ear may determine that the seeming mezzo is a true soprano, that the notes of the pupil who comes as a baritone have the tenor quality and that his scale safely can be added to, while the would-be tenor has the baritone timbre which will prevent his notes from ever ringing out with the true tenor quality. Yes, this initial task of voice classification is far too important to be entrusted to "any teacher."

There are piano-thumping teachers of voice, who not having voices themselves are obliged to give their pupils the pitch of each note by pounding it out on the pianoforte. Voice quality has nothing in common with pianoforte quality of tone, yet constant thumping of the pianoforte by a singing-teacher in order to give the pupil the pitch, is apt to mix pianoforte color into a pupil's voice and mar its translucent vocal quality. A teacher need not be a fine singer—few vocal teachers are—but, at least, he should be able to give pitch vocally and to suggest with sufficient definition the quality of tone the pupil is to produce.