afforded a sufficient feeling of variety, and it is worthy of note that
later, when Beethoven used violently contrasting kinds of themes to
express dramatically contrasting feelings, the question of key ceased to
have the same importance. Composers later than Mozart have never
troubled to mark their first key, so that the key of the second subject
might sound like a grateful change and continuation; the stuff of the
themes has been depended on for variety, while for unity the great art
of thematic development has served. So far as Haydn carried this art, we
may note a few of his devices. Double counterpoint, imitation, fugue, or
at least fughetta--these he returned to later. Bits of themes--mere
fragments marking definite rhythms--were used in spinning new melodies,
a rhythm, or perhaps a sufficiently distinctive harmonic progression,
connecting them with what had gone before. This use of a "germ" idea was
chiefly due to Beethoven, who, as in the first movement of the Fifth
Symphony, worked out a gigantic piece of music from four notes. But
Haydn knew well how the value of intervals in a melody might be changed
by the harmony, how a familiar bit of tune, with the simplest harmonies
arranged in a new way, resulted in practically a new melody. This device
he commonly used, sometimes with fine results. The incessant series of
climaxes, leading us on and keeping us in suspense until a certain point
is reached, then releasing the tension for a moment, and preparing to do
the same again--these he employed to an extent, but not as Beethoven
employed them.

All this Mozart perceived, and made instant use of. As for the
mediocrities for whose benefit Haydn is held to have "stereotyped" the
form, what could they learn from him? I will say what they did learn.
They learnt to take themes which did not sound exactly like the
subjects of a fugue; they laid out their first and their second, and
then they did not know what on earth to do, and footled and stumbled