48 



BUEEAU OF AMERICAN ETHNCTLOGY 



[BULL. 61 



Fig. 13. Part of measure on which, song begins. 



TaMe 15 A. — Wo now enter on the consideration of the rhythm of 

 these songs and of the drum with which many of them were accom- 

 panied. Tabic 15A concerns only the rhythm of the voice, and its 

 pm-posc is to determine directness of "attack" in beginning a song. 

 The analysis shows that 61 per cent of the songs begin on the accented 

 part of the measure. Table 2A shows that a majority of the songs 



begin on the tones 

 of the tonic chord. 

 Taken together, these 

 tables suggest a clear- 

 ness of musical con- 

 cept on the part of the 

 Indian. (See fig. 13.) 

 Talle ^^^.— Like 

 the preceding table, 

 this concerns the 

 rhythm of the voice. 

 In the phonograph 

 records of these songs 

 the accented tones usually are unmistakable and clear. As already 

 stated, a measure in the transcription represents the period of time 

 between two accented tones. The division of this time period into 

 two parts is the simplest possible division. In 54 per cent of the 

 songs the first measure is in double time C2-4 or 2-2) and in 40 per cent 

 in triple time (3-4 or 

 3-8), 6 per cent of 

 the songs beginning 

 in combinations of 

 these rhythms. (See 

 fig. 14.) 



Talle 17 A. —K 

 change of time (meas- 

 ure-lengths) is found 

 in 84 percent of these 

 songs (see fig. 15). 

 Such a change in the 

 music of civilized peo- 

 ples usually affects an entire section of a melody or composition. 

 Changes in time in the songs under analysis, on the other hand, 

 commonly affect single measures or only a few measures. These 

 alternations of measure lengths usually find what may be termed 

 their rhythmic explanation, in the rhythmic unit of the song or in 

 the rhythm of the song as a whole. (See fig. 15.) 



O thorwise 

 classified. . . 

 (6 ppr cent) 



First measure 

 in double 

 time 

 (54 per cent) 



Fig. 14. Rhythm of first measure. 



