ROBERTS) ANALYSIS OF SONGS 423 



soon to come or it may be a case of sharpening the curve between 

 (l.y and/ as No. 27 would indicate. At any rate it is another example 

 of a modified subdominant, which in No. 28 in relation to the at> imme- 

 diatelj' following it becomes a tone which would be the mediant 

 with d\} as the tonic. The raised (l\> in the first measure of the C 

 phrase of No. 8 is probablj* a case of smoothing the downward curve; 

 the din's of the second measure again are a clear example of the 

 sharping tendency. In No. 27 the first note of the third measure 

 probably illustrates the smoothing process but the sharped c's are 

 comparable to the d's of No. 8, just discussed. The instability of the 

 principal tones of the scale affects judgment regarding the ephemeral 

 tones. I can only recommend a comparison of the songs, which will 

 give some idea of the difficulties of the problem. 



No. 6 appears to be a more stable version of the song than No. 8 

 and its other renditions, 27 and 28. Here the tendency to acuate 

 curves is evident throughout. It is clearer from No. 6 and its other 

 two renditions that the distinctions between major and minor thirds 

 which appear continually at different levels and in dift'erent situations 

 throughout the song is very difficult for this singer. The second 

 measure of No. 23, showing the depressed peak of the curve between 

 the two ^'s, probably explains the c:'s of the first measure in Nos. 8 

 and 28. 



In No. 5 (p. 353) the depressed eb in the third measure of the C 

 phrase appears to have been affected by the long series of lower 

 notes preceding it and the drop to the Jb's in the next measure. In 

 other words, the peak of true e\> seems to have been gained with 

 difficulty and the tendency to attenuation of the curve is observed, 

 as also in the fourth measure of the B phrase. These two inter- 

 mediate pitches are therefore probably to be discounted as main 

 scale tones in the table where they appear as lowered g and lowered/. 

 Both of the sharped pitches in the final measure of the A phrases of 

 No. 21 (the other rendition of No. 5, p. 353) are probably due to 

 smoothing downward melodic curves. 



Comparable to Group II, except that the principal tone occupies a 

 middle position just as in the songs of Group III, are the scales of 

 Group IV which are derived from Songs 3 and 4 and their other 

 renditions, Nos. 18, 19 and 20. They are more lunited in tonal con- 

 tent than those of Group II, although their range covers an octave. 

 The presence of eb as a prominent or structural tone forming a minor 

 third with the tonic, rather than the prominence of elq which would 

 create a major third, makes them minor, although the major third 

 is not entirely lost sight of, as its infrequent use proves. It is inter- 

 esting to observe the low beginnings of the songs the scales of which 

 form Group IV in contrast to the higli initial notes m the songs of 

 Group III where the tonic also occupies a middle position in the scale. 



