Of this form there are numerous examples in Chopin, as in the latter part of the Ballade in A♭, and in the Prelude in the same key (No. 17). Beethoven makes use of successions of thirds in the same way; of which the most remarkable example is the Largo which precedes the fugue in the Sonata in B♭, op. 106. In this there are fully eighteen successive steps of thirds downwards, most of them minor. This instance also points to a feature which is important to note. The successions are not perfectly symmetrical, but are purposely distributed with a certain amount of irregularity so as to relieve them from the obviousness which is often ruinous to the effect of earlier examples. The divisions represented by each step are severally variable in length, but the sum total is a complete impression based upon an appreciable system; and this result is far more artistic than the examples where the form is so obvious that it might almost have been measured out with a pair of compasses. This point leads to the consideration of another striking device of Beethoven's, namely, the use of a cæsura in modulation, which serves a similar purpose to the irregular distribution of successive modulations. A most striking example is that in the Prestissimo of the Sonata in E major, op. 109, in bars 104 and 105, where he leaps from the major chord of the supertonic to the minor of the tonic, evidently cutting short the ordinary process of supertonic, dominant and tonic; and the effect of this sudden irruption of the original key and subject before the ordinary and expected progressions are concluded is most remarkable. In the slow movement of Schumann's sonata in G minor there is a passage which has a similar happy effect, where the leap is made from the dominant seventh of the key of D♭ to the tonic chord of C to resume the first subject, as follows:—
In the study of the art of music it is important to have a clear idea of the manner in which the function and resources of modulation have been gradually realised. It will be best therefore, at the risk of going occasionally over the same ground twice, to give a short consecutive review of the aspect it presents along the stream of constant production.
To a modern ear of any musical capacity moduation appears a very simple and easy matter, but when harmonic music was only beginning to be felt, the force even of a single key was but doubtfully realised, and the relation of different keys to one another was almost out of the range of human conception. Musicians of those days no doubt had some glimmering sense of a field being open before them, but they did not know what the problems were which they had to solve. It is true that even some time before the beginning of the seventeenth century they must have had a tolerably good idea of the distribution of notes which we call a key, but they probably did not regard it as an important matter, and looked rather to the laws and devices of counterpoint, after the old polyphonic manner, as the chief means by which music was to go on as it had done before. Hence in those great polyphonic times of Palestrina and Lasso, and even later in some quarters, there was no such thing as modulation in our sense of the word. They were gradually absorbing into their material certain accidentals which the greater masters found out how to use with effect; and these being incorporated with the intervals which the old church modes afforded them, gave rise to successions and passages in which they appear to us to wander with uncertain steps from one nearly related key to another; whereas in reality they were only using the actual notes which appeared to them to be available for artistic purposes, without considering whether their combinations were related to a common tonic in the sense which we recognise, or not. Nevertheless this process of introducing accidentals irregularly was the ultimate means through which the art of modulation was developed. For the musical sense of these composers, being very acute, would lead them to consider the relations of the new chords which contained notes thus modified, and to surround them with larger and larger groups of chords which in our sense would be considered to be tonally related; and the very smoothness and softness of the combinations to which they were accustomed would ensure a gradual approach to consistent tonality, though the direction into which their accidentals turned them was rather uncertain and irregular, and not so much governed by any feeling of the effects of modulation as by the constitution of the ecclesiastical scales. Examples of this are given in the article Harmony; and reference may also be made to a Pavin and a Fantasia by our great master, Orlando Gibbons, in the Parthenia, which has lately been republished by Mr. Pauer. In these there are remarkably fine and strong effects produced by means of accidentals; but the transitions are to modern ideas singularly irregular. Gibbons
