[Illustration: Example 13 continued. Fragments of Mendelssohn and Mozart.]
In No. 1, the motive begins squarely with the measure, upon the accented beat. In No. 2, the same motive is enlarged by two tones at the outset, which locates its beginning upon the fourth 8th—the second half of the second beat. In No. 3 the motive begins upon an accented beat, but it is the lighter (secondary) accent of the 3d beat. The various conditions of unaccented beginnings in Nos. 4, 5 and 6 are easily recognizable. In No. 7 quite a large fraction of a measure precedes the first accent (at the beginning of the full measure). Examine, also, all the preceding examples, and note the different accented or unaccented locations of the first tone, in each figure and motive.
When a figure or motive starts at the accented beat, it begins, so to speak, in the right place; any tone or tones which precede the accent are merely preliminary or introductory tones. While they are very desirable and necessary, in the fulfilment of certain purposes, they are not an essential part of the motive; they appear to represent the ornamental rather than the stable element of the melodic sentence, and their employment is therefore a matter of option and taste rather than of absolute necessity. The accent indicates the point where the body of the motive begins; the accent is the point where the stake is driven; all that goes before is simply preparatory,—the changeable material which flutters about the fixed center. Therefore the preliminary tones do not indicate the essential or actual beginning of the motive, but its apparent or conditional beginning only; or what might be called its melodic beginning. For this reason, also, the actual "first measure" of a motive or phrase or sentence of any kind is always the first FULL measure,—the measure which contains the first primary accent; that is to say, the preliminary tone or tones do not count as first measure. For this reason, further, it is evident that preliminary tones are invariably to be regarded as borrowed from the final measure of the preceding motive or phrase; they must be accounted for in someway,—must derive their metric pulse from some group,—and as they cannot be a part of the first measure, they obviously form a borrowed portion of the (preceding) last measure. This will be better understood by reference to Ex. 14, No. 3; the two 16ths at the end of the 4th measure (preliminary tones of the following phrase) are borrowed from the f which precedes,—the final tone of the first phrase, that would, but for this reduction, have been the full half-note necessary to complete the four measures (like the final g).
Perhaps the most striking feature of this rule of preliminary tones is the absolute freedom of its application. It is always wholly optional with the composer to begin his figure or motive at whatever part of the measure he may elect; at the accent or not; with or without preliminary tones; to borrow beats from the preceding ending or not, as his judgment or taste, or possibly some indirect requirement, may decide. So valid is this license, that it is by no means unusual to find consecutive members of the same phrase beginning at different points in the measure. This results, apparently, in motives of irregular, unsymmetric lengths; but no confusion is possible if the student will recollect and apply the rule that the objective point (the heart, so to speak) of each motive is the first primary accent it contains; counting from these points, all irregularities of melodic extent become purely accidental and harmless. For illustration (the preliminary tones are marked a):—
