down on a slip of paper, and he always had a special leather bag for
preserving these sketches, which he carefully guarded. These sketches
differ somewhat in appearance, but generally they contained the melody
or vocal part, together with the bass, and brief indications of the
middle parts, and here and there mention of a special instrument.
This was sufficient subsequently to recall the whole composition to
his memory. In elaborating his scores he hardly ever made any
deviations from the original conception, not even in the
instrumentation; which seems the more remarkable when we reflect that
he was the originator of many new orchestral combinations, the beauty
of which presented itself to his imagination before his ears had ever
heard them in actuality. These new tone-colors, as Jahn remarks,
existed intrinsically in the orchestra as a statue does in the marble;
but it remained for the artist to bring them out; and that Mozart was
bound to have them is shown by the anecdote of a musician who
complained to him of the difficulty of a certain passage, and begged
him to alter it. "Is it possible to play those tones on your
instrument?" Mozart asked; and when he was told it was, he replied,
"Then it is your affair to bring them out."

Beethoven's way of mental composing appears at first sight to differ
widely from Mozart's. But if we had as many specimens of Mozart's
preliminary sketches as we have of Beethoven's, the difference would
perhaps appear less pronounced, and would to a large extent resolve
itself into the fact that Beethoven did not trust his memory so much
as Mozart did, and therefore put more of his _tentative_, or rough
sketches, on paper. He always carried in his pockets a few loose
sheets of music paper, or a number of sheets bound together in a
note-book. If his supply gave out accidentally, he would seize upon
any loose sheet of paper, or even a bill of fare, to note down his