be played satisfactorily by a pianist of slight ability, if only he have the knack of following the singer. But scarcely any of Schumann’s accompaniments can be played satisfactorily except by a performer who is an artist at his instrument. II Schumann’s songs can be easily divided into two groups. The first, comprising about a hundred, was composed in 1840, the year of the composer’s marriage, and of his most fertile and romantic productivity. After writing these, he professed himself satisfied with what he had done (well he might be) and doubted whether he should ever write songs again. His most fervent admirers cannot help wishing that he had kept to his purpose. A dozen or so indifferent songs were produced in the next ten years, and then, from 1850 to 1852, when his mental powers had begun seriously to fail, he turned out more than a hundred. The contrast between the two groups is striking. The first is nearly all genius; the second nearly all mediocrity. At best, two or three songs from this whole later period are worth knowing. Of the first group there is scarcely one which does not show high creative excellence and there are at least fifty of the finest quality or close to it. By far the greater part of the songs of the earlier period are grouped in cycles. Of these, however, there is only one which contains that most distinctive quality of the cycle, a unified narrative--as in Schubert’s _Müllerlieder_. This is the immortal _Frauenliebe und Leben_. Another, the _Dichterliebe_, is closely bound together and primarily intended to be sung as a unit. [Illustration: Robert and Clara Schumann] _After a daguerreotype from life._ Three more cycles, while looser than the one just mentioned, are like it in being set solely to the words of one author. And another, _Myrthen_, has little beyond a common title for unification. A few other songs have no grouping except that of an opus number.