movements in the forms prevalent in his time, and occasionally used vain repetitions of keys like his contemporaries; but when he chose his own lines he produced movements which are perfectly in consonance with modern views. As examples of this the 'Et resurrexit' in the B minor Mass and the last chorus of the Matthew Passion may be taken. In these there is no tautology in the distribution of the modulation, though the extraordinary expanse over which a single key is made to spread, still marks their relationship with other contemporary works. In some of his instrumental works he gives himself more rein, as in fantasias, and preludes, and toccatas, for organ or clavier. In these he not only makes use of the most complicated and elaborate devices in the actual passage from one key to another, but also of closely interwoven transitions in a thoroughly modern fashion. Some of the most wonderful examples are in the Fantasia in G minor for organ (Dörffel 798), and others have been already alluded to.
It is probable that his views on the subject of the relation of keys had considerable influence on the evolution of the specially modern type of instrumental music; as it was chiefly his sons and pupils who worked out and traced in clear and definite outlines the system of key-distribution upon which Haydn and Mozart developed their representative examples of such works.
In the works of these two great composers we find at once the simplest and surest distribution of keys. They are in fact the expositors of the elementary principles which had been arrived at through the speculations and experiments of more that a century and a half of musicians. The vital principle of their art-work is clear and simple tonality; each successive key which is important in the structure of the work is marked by forms both of melody and harmony, which, by the use of the most obvious indicators, state as clearly as possible the tonic to which the particular group of harmonies is to be referred. This is their summary, so to speak, of existing knowledge. But what is most important to this question is that the art did not stop at this point, but composers having arrived at that degree of realisation of the simpler relations of keys, went on at once to build something new upon the foundation. Both Haydn and Mozart—as if perceiving that directly the means of clearly indicating a key were realised, the ease with which it could be grasped would be proportionately increased—began to distribute their modulations more freely and liberally. For certain purposes they both made use of transitions so rapid that the modulations appear to overlap, so that before one key is definitely indicated an ingenious modification of the chord which should have confirmed it leads on to another. The occasions for the use of this device are principally either to obtain a strong contrast to long periods during which single keys have been or are to be maintained; or, where according to the system of form it so happens that a key which has already been employed has soon to be resumed as, for instance, in the recapitulation of the subjects to lead the mind so thoroughly away that the sense of the more permanent key is almost obliterated. Occasionally, when the working-out section is very short, the rapid transitions alluded to are also met with in that position, as in the slow movement of Mozart's E♭ Quartet. The example quoted above from the last movement of his Quartet in C will serve as an example on this point as well as on that for which it was quoted.
A yet more important point in relation to the present question is the use of short breaths of subordinate modulation in the midst of the broader expanses of the principal keys. This is very characteristic of Mozart, and serves happily to indicate the direction in which art was moving at the time. Thus, in the very beginning of his Quartet in G (Köchel 387), he glides out of his principal key into the key of the supertonic, A, and back again in the first four bars. A similar digression, from F to D and back again, may be observed near the beginning of the slow movement of the Jupiter Symphony. But it requires to be carefully noted that the sense of the principal keys is not impaired by these digressions. They are not to be confounded either with the irregular wandering of the composers who immediately succeeded the polyphonic school, nor with the frequent going out and back again of the composers of the early part of the 18th century. This device is really an artificial enlargement of the capacity of a key, and the transitions are generally used to enforce certain notes which are representative and important roots in the original key. A striking example occurs in the first movement of Mozart's symphony in G minor (1st section), where after the key of B♭ has been strongly and clearly pointed out in the first statement of the second subject, he makes a modulatory digression as follows:—
This is in fact a very bold way of enforcing the subdominant note; for though the modulation appears to be to the key of the minor seventh from the tonic, the impression of that key is ingeniously reduced to a minimum, at the same time that the slight flavour that remains of it forms an important element in the effect of the transition.
