This page has been proofread, but needs to be validated.
190
Fugue.
[Chap. XI.

After the explanations of the preceding exposition, few words are needful concerning this one. If the student will examine it, he will see that a codetta is introduced before each new pair of entries. The treble, which led with the second subject, makes its final appearance at bar 27 with the first, showing the inversion of the two subjects, just as they were shown at bar 14 of the example by Hummel.

378. Another important point is illustrated by the two expositions just quoted. It will be seen that though both are for four voices, only three are present at any one time, while a considerable part of the exposition in § 375 is for two voices only. We often find this in the exposition of a double fugue; the reason is that it is absolutely necessary that both the subjects shall be clearly and easily distinguished, and this end is attained by leaving them either without any other counterpoint, or with only one added part. Clearness is, if possible, of even more importance in a double fugue than in a fugue with only one subject.

379. The general form of such double fugues as those we are now describing mostly follows the plan shown in Chapter IX. But there are some differences of detail often to be met with that must be mentioned. In the first place we sometimes find less episode in a double than in a simple fugue. For instance, the 'Kyrie' of Mozart's 'Requiem,' one of the finest double fugues ever written, contains only one episode, and that is but a bar and a half in length. When there is so little episode, its place is usually taken by developments of one of the two subjects without the other.

380. In general, both subjects should be heard together in each group of middle entries. Occasionally, we find such an entry for one subject alone; but this is far more exceptional than the entry of a subject in a simple fugue without its countersubject.

381. It is seldom practicable in a double fugue to write a stretto in which both the subjects shall take part. We have already seen with simple fugues that if there is much stretto, there is generally either no countersubject, or, if there be one, it is omitted in the stretti. The obvious reason for this is, that