50 BUREAU OF AMERICAN ETHNOLOGY [bull. 5:; 



are on the fifth five- toned scale and two on the fourth five- toned 

 scale, tliese comprismg liaK the group. From two of the songs the 

 third is omitted (see analysis of No. 53). None of these songs con- 

 tain an accidental and 87 per cent are purely melodic in structure. 

 Half l)egin with a downward and half witli an upward progression. 

 The minor third is especially prominent in the descending intervals. 

 The average interval is the same as in the war songs, and we note 

 that three of these songs are concerned with a child's game of war 

 and one with war between animals. Half the songs begin on the 

 accented and half on the unaccented part of the measure. Three- 

 fourths begin in 2-4 time and the percentage of songs marked by a 

 change of time is the largest except in the dream songs, the pipe 

 dance songs, and the unclassified songs. Sixty-two per cent contain 

 no rhythmic unit, this proportion being exceeded only in the moc- 

 casin game songs. In the rendering of these songs, as well as in 

 those of the moccasin game, a high degree of excitement prevails. 

 Most of the songs were recorded without the drum; in one song drum 

 and voice show the same metric unit, while in another the drum is 

 slower than the voice. 



The rhythmic units occurring in these songs will be found on 

 page 330. 



UNCLASSIFIED SONGS 



This group comprises the following songs: Nos. 146, 147, 148, 150, 

 186, and 192-196 in Bulletin 45, and Nos. 67, 68, 119-122, 165-169 

 of the present work. These songs present a wide variety of inter- 

 est, including soiigs of the ca'wuno'ga (southern) dance, the divorce 

 ceremony, the friendly visit of one band to another, and a song 

 concerning an historical incident. As the topics of the songs are so 

 diverse it does not seem expedient to consider the group as a unit. 

 The rhythmic units found in the songs are, however, of interest (see 

 pp. 330-332). 



MELODIC AND RHYTHMIC RESEMBLANCES BETWEEN SONG GROUPS 

 (based on TABLES 1-22) 



The preceding analysis suggests connection between the idea of the 

 sono- and its musical form, and also indicates resemblance between 

 groups of songs containing a somewhat similar idea. Tables (pp. 51-58) 

 have been prepared in order that these resemblances may be more con- 

 veniently observed. For instance, it will be noted that the Mide' songs 

 resemble the begging dance songs, the idea common to both being 

 desire for acquirement, in the former for the acquirement of super- 

 natural power and in the latter for gifts of food. Turning to the 

 analysis of the begging dance songs, they are found to be allied to the 

 three groups of songs in which the element of pleasure is strongest — 

 the songs of the woman's dance, the pipe dance, and those for the 



