the romanza-like Andante is palpably a very early composition and that the correspondence in key and measure with the B-flat Concerto might indicate that it was originally designed as a part of that work, a supposition which is strengthened by the fact that the original manuscript is neither dated nor signed. This internal evidence has much in its favor, the more since it is not at all obvious what might have prompted Beethoven to write an independent rondo for concert use. There is no external evidence; if there were, the conception of the B-flat Concerto would have to be set at a much earlier date than has yet been done. The first Vienna sketches for the Concerto, as Nottebohm shows, prove that the present three movements belonged together from the beginning. They were, therefore, surely played at the first performance in 1795. Nottebohm, who repeated Jahn’s surmise in his “Thematisches Verzeichniss,” changed his mind after a study of the sketches and rejected the notion that the rondo had been designed for the Concerto. Only by assuming an earlier date for the rondo can the theory be upheld. Attention may here be called to Wegeler’s statement (“Notizen,” p. 56) that the rondo of the first Concerto (he says, of course, the Concerto in C) was not composed until the second afternoon before the performance. There may possibly have been another. This is not necessarily disproved by the fact that sketches for the present one were in existence. The question is not settled by the evidence now before us, but the probabilities are with Mandyczewski.