ar,
of course, in the original,--to which the student is expected to refer.

[Illustration: Example 54. Fragment of Beethoven.]

[Illustration: Example 54 continued.]

[Illustration: Example 54 continued.]

This is a genuine First Rondo-form. All the factors of which it is
composed, Phrases, Parts and Themes, are so closely interlinked that
the continuity, cohesion and _unity_ of the whole is complete. The
variety of contents which these factors exhibit (greatest, naturally,
between the two themes), does not disturb the impression that the whole
movement is a unit. This is due, at least partly, to the manner in
which the perfect cadences are disguised; each one is passed over with
the least possible check of rhythmic movement (measures 8, 19, etc.),
thus snugly dove-tailing the structural factors. The coda is elaborate
and unusually long; it consists of several "sections," as follows (see
the original): from measure 1 (the last measure in Ex. 54) to measure
4, a phrase, derived from the second Part of the Principal theme;
measures 5-7, an abbreviated repetition; measures 8-14, a phrase,
derived from the Principal theme; measures 15-17, a transitional
passage; measures 18-25, a period, closely resembling Part I of the
Principal theme; measures 26-30, final phrase.

LESSON 13.--Analyze the following examples. They are not classified;
the student must determine whether the form is pure First Rondo, or an
intermediate grade between Rondo and "Song with Trio." One of the
examples is a genuine Song with Trio; and one is a _Three-Part
Song-form_; with reasonable vigilance the student will detect these
"catches." To distinguish these three designs from each other,
recollect--

That the Three-Part Song-form consists of three _single Parts_, fairly
similar in character, fairly small in form, and severed either by a
firm cadence, or by unmistakable proof of new "beginning;"

That in the first Rondo-form, at least one of the themes (if not both)
contains _two_ (or three) Parts; and,

That in the Song with Trio, the two "Songs" are more independent of
each other, and more decisively separated, than are the "themes" of the
Rondo-form.

With reference to all uncertain cases, it must be remembered that _the
more doubtful a distinction is, the less important is its decision_.
These designs naturally merge one in another, and at times it is folly
to impose a definite analysis upon them.

The analysis should be as