10 BUREAU OF AMERICAN" ETHNOLOGY [Brix. et 



thoir first ocoiirronco, in Nos. 5, 6, 8, 12, 19, 64. In Nos. 12 and 13 

 tho metric unit of drum and voice are in the ratio 2 to 1, suggesting 

 points of coincidence, but the occasional prolonging or shortening of 

 tones by the voice is such as to prevent a mechanical relation between 

 the two. In many instances the tempo of the drum appears entirely 

 independent of that of the voice. The illustration on page 1 10 shows 

 the seeming lack of relation between drum and voice in No. 8, but in 

 this, as in the simple ratio, the slight variability of the voice should 

 be taken into account. In No. 125 the drum was struck with a 

 clearness which made it possible to transcribe its beats throughout 

 the song. In tliis instance the drum and voice coincided on the first 

 of the measure, but the drum-rhythm was broken. In a few songs 

 the drum and voice were at variance during the song until the closing 

 measures, in which they coincided. These were songs in which the 

 drumbeat was in quarter-note values, and this "swinging together" 

 of voice and drum is noted in the descriptive analyses. 



At any gathering there is one man who acts as leader of the singei's, 

 who sit around the drum. The number at the drum varies with the 

 size of the gathering ; if a large number are dancing the singers sit as 

 close as possible around the drum, each man beating the drum as ho 

 sings. Sometimes as many as 10 men can "sit at the drum." A 

 singer of recognized ability may, if he likes, bring a decorated drum- 

 stick of his own, but the common custom is for the drummers to use 

 ordinary drumsticks consisting of sticks wound at the end with cloth. 

 If a singer at the drum becomes weary he lays down his drumstick, 

 whereupon someone who has been dancing, or sitting with the spec- 

 tators takes his place. It was said that "the leader starts every song, 

 and if it is an easy song the others commence right away, but if it is 

 new and hard they begin more carefully." Most of the songs may be 

 sung an indefinite number of times, the leader giving a signal for the 

 close by two sharp taps on the drum, after which the song is sung 

 only once. 



Sioux women usually sing with the men during the dancing songs ; 

 this is not the custom among the Chippewa. The Sioux women sit 

 on the ground, forming a circle back of tlie drummers, and sing in a 

 high falsetto, an octave above the men. In three instances (Nos. 201, 

 203, and a duplication of 173) the part sung by the women is shown 

 in the transcription. Several other songs were recorded in this 

 manner, but tho transcription of each is from a subsequent rendition 

 by a man singing alone. 



Among the Sioux as well as among the Chippewa, variations in 

 either time or intonation are found more often in unimportant than 

 in important parts of a song. 



It was said that there were ''different ways of ending songs," and 

 that "a man could tell the kind of song bv the way it ended." The 



