igns which may result from merely repeating one or the
other of the two phrases that make a period, as is shown in Ex. 48.
_All such phrase-clusters as are reducible to two phrases_, because
nothing more than simple repetition has been employed in their
multiplication, should always be classed among ordinary periods; for
two successive phrases, if connected (that is, unless they are
purposely broken asunder by a definite perfect cadence at the end of
the first phrase) always represent the analogy of Question and Answer.

The enlarged form we are at present considering consists of three
_different_ phrases, as a general rule; probably very closely related,
or even distinctly resembling one another; but too independent,
nevertheless, to constitute actual repetition, and therefore to admit
of reduction to two phrases. For this very reason it cannot justly be
called "period" at all, but takes the name of "phrase-group." An
illustration by diagram will make the distinction clear:--

[Illustration: Phrase group diagram.]

Observe that the classification depends upon the number of
phrases,--upon the _melodic_ identity of the phrases,--and upon the
_quality of the cadences_.

No. 1 is illustrated in Ex. 15; No. 2, in Ex. 42 and the first four
measures of Ex. 43 (cadence not perfect, it is true, but same
phrase-melody and _same cadence_); No. 3 is seen in Ex. 44
(phrase-melody similar, but cadences different)--also in Ex. 47; No. 4
is seen in Ex. 48; No. 5 is rare, but an example will be discovered in
Lesson 8; No. 6 is illustrated in the following (Grieg, op. 38, No.
2):--

[Illustration: Example 50. Fragment of Grieg.]

Comparing this sentence with Ex. 48, we discover the following
significant difference: There, no more than two phrases were present;
the whole sentence was _reducible_ to two phrases. Here (Ex. 50),
however, no such reduction is possible; three sufficiently similar--and
sufficiently different--phrases are coherently connected, without
evidence of mere repetition; it is the result of Addition, and the form
is a _phrase-group_. The first cadence is, strictly speaking, a
_perfect_ one; but of that somewhat doubtful rhythmic character, which,
in conjunction with other indications, may diminish its conclusive
effect, and prevent the decided separation which usually attends the
perfect cadence. This is apt to be the case with a perfect cadence _so
near the beginning_ (like this one) that the impressi