RouEKTSl ANALYSIS OF SONGS 413 



naturally belonged together and formed one phrase. The rise in 

 pitch which occurs at the end of each measure in the second A phrase 

 affords a pleasing contrast to the level lower ending of the other 

 measures in the first and last two A phrases. 



Tabular Analysis No. 11 



Song No. 11 



Phrases Beats 



a. A 24 



A 24 



A 24 



A 24 



To summarize briefly the points brought out in this study of the 

 structure of the songs, the first that comes to mind is the prominence 

 of the ternary idea. Of the eleven different songs five show it to a 

 marked degree not only, as in some of them, in the larger structure 

 of repetitions of groups of phrases, but in the number of phrases 

 forming a group, and to some extent, but not consistently through- 

 out, in the number of sections within the phrase, or measures within 

 the section, or beats within the measures. Some, like Nos. 3 and 4, 

 which are about the same tune, have two identical parts consisting 

 of three phrases, succeeded by a third short part havmg only one 

 different plirase. Others, like Nos. 6 and 8, have a different ternary 

 form, composed of only two different phrases, one of which is 

 repeated. In one other. No. 5, which is really four-part, only three 

 different phrases are used, one being repeated. Thus eight out of 

 eleven songs have some ternary feature about them and five show 

 such a plan prominently. The remaining three songs are binary in 

 principle, not in the possession of two different phrases which com- 

 plement one another, but in the coupling of two identical or almost 

 identical parts. In Nos. 9 and 10, however, the two sections com- 

 posing the phrases are different. In No. 11 they are identical. It 

 has been seen that the singer was fairly consistent in performance, 

 especially as regards the larger features of the structure, but small 

 melodic, rhythmic, and metric differences characterize all the ren- 

 ditions. Absolute metric regularity was not found in any song, 

 although a few nearly approached it. It is rather surprising that the 

 ternary scheme does not extend more often to the meter, which is 

 always rather obvious in any music, but three-part meters are less 

 common in these songs than several other meters. Apparently there 

 are no metric patterns and the prevailing irregularity of size in metric 

 groupings is to be accounted for probably by the contest for supremacy 

 between word accents and the latent accents in certain melodic 

 situations, as in the songs of most primitive peoples. 



vSome striking rhythms a're noted which are not perhaps entirely 

 due to the natural rhythms of the accompanying words, but they 



