s, that, as
thoroughly explained in Chapter III, creates a distinction between the
_melodic_ beginning and the actual vital starting point of the phrase;
or that gives the phrase an apparently shifted location in its measures.

Further (the actual cadence-tone is marked):--

[Illustration: Example 23. Fragments of Beethoven and Mendelssohn.]

[Illustration: Example 23 continued. Fragment of Mendelssohn.]

No. 1 illustrates, again, the absence of preliminary tones in one
phrase, and their presence in the next. In each of these examples
(excepting, perhaps, No. 2) the cadence is so thoroughly disguised that
there is little, if any, evidence left of the "point of repose." In
No. 4, particularly, the cadence-measure is rhythmically the most
active one in the phrase. And yet the presence of a genuine cadence at
each of these places, marked *, is as certain and indisputable as in
Ex. 19. The ear will accept a cadence upon the slightest evidence _in
the right place_,--where a cadence is expected. See, also, Mozart
pianoforte sonata No. 10 (in D major), first 12 measures; measure 8 is
a _cadence-measure_.

Here follow a few more examples which illustrate the most extreme
application of this principle of borrowed tones,--a mode of treatment
very common in the music of Mozart, Haydn, and, in fact, all classic
writers:--

[Illustration: Example 24. Fragments of Mozart.]

[Illustration: Example 24 continued. Fragment of Beethoven.]

It is difficult to believe that in each of these cases the long array
of 16th-notes should not constitute the actual beginning of the phrase,
but are only preliminary; and yet this is the only correct view to take
of it, and it is the view which will simplify all analysis, when
thoroughly comprehended. It must be seen that the cluster of
16th-notes in the cadence-measure (of the preceding phrase) is
_one-sixteenth short of a full measure_, and, therefore, it does not
represent the first measure of the next phrase, because our inviolable
rule is that the first measure of a phrase is its first _full_ measure.
The above examples emphasize the correct manner of counting the
measures; and they simply illustrate possible methods of _disguising
the cadence_.

In some cases it is difficult to determine whether the tones which thus
disturb the "repose" of the cadence-measure belong to the cadence-chord
(that is, to the _present_ phrase), or, as preliminary tones, to the
following phrase.