ROBERTS) 



ANALYSIS OF SONGS 419 



While discussing this group of scales I should like to call attention 

 to those marked as Group III on page 446. Except that the tones 

 of the songs represented m this group range below the tonic as well 

 as above it, and thus those below it are transposed as regards octave 

 position, they represent the same scale intervals as the scales of 

 Group I, namely, chiefly those of a major diatonic scale, but with 

 some intermediate tones between the tonic and the leading tone just 

 below it, some fluctuations just above and below the subdominant 

 and a number of intermediate tones between the dominant and the 

 submediant. Let us see what these intermediate pitches suggest. 



Song No. 2 (p. 307), the scale of which is the fii-st in Group III, 

 offers more evidence that this singer was influenced in pitchmg his 

 intervals by the general trend of the melody. In the first measure 

 of the firet A the g is slightly sharped, bringing it nearer to the a's 

 on either side of it, between which it stands as a changing note. 

 Thus this little melodic curve has been smoothed, or attenuated. In 

 both sections of the B phrase the c which stands as a changing note 

 between the two a's is flatted, just falling short of what should be 

 the peak of the melodic curve if the true tone were used which occurs 

 in other melodic situations in the song. In the final measure of the 

 C phrase as the melody comes nearer to the low g on which the phrase 

 is to end, the c's are flatted in anticipation of the drop, or in response 

 to the downward trending melodic curve. These modifications of the 

 curves are not consistently carried out for all the repetitions of 

 similar melodic situations but in the first B phrases of Parts II and 

 111 the flatted c's between a's are to be noted. This tendency to 

 reduce the sharpness of upward melodic curves accounts for all the 

 intermediate pitches shown between the diatonic intervals in the 

 scales of Nos. 2, f5, 16, and 17, so that these scales are to be identi- 

 fied with those of Group 1 except that they lie around the tonic, the 

 upper and lower tones being duplicated in difTerent octaves, instead 

 of all tones lying above the tonic. Even in some of the songs of the 

 first group the upper tonic appears as a duplication of the lower tonic 

 in the second octave and the supertonic of the second octave appears 

 when that of the firet does not. 



The other scales of Group 111 (those of No. 10 and its other rendi- 

 tions, Nos. 31, 32, 33, and 34) illustrate a process exactly opposite 

 to tliat of smoothing curves as well as offering some additional 

 examples of such smoothing, although the latter are rare in this song. 

 In the final measure of the first A phrase of No. 10 (p. 379) the two 

 submediants are slightly sharped, altered in the direction away from 

 the trend of the melody. The last measure in No. 10 and corre- 

 sponding notes in the final measures of Nos. 31, 32, and 33 (pp. 

 443, 444) indicate that these tones should have been true 6's. In the 



