14 BUREAU OF AMERICAN ETHNOLOGY (BULL, 80 
tribe has its personal characteristics as well as its distinct traditions 
and customs. The principal subject under investigation was the 
music of the Mandan, and it is believed that the songs credited to 
that tribe are old Mandan melodies. A similar number believed to 
be purely Hidatsa was not collected; therefore a detailed comparison 
of the music of the two tribes is not undertaken in this work. It is 
possible, however, to make a tentative comparison of the Mandan > 
with the mixed material through the comparative group analyses on 
pages 16-26. The first 46 songs are believed to be Mandan, and the 
remainder are designated as “Mandan and Hidatsa,” the tribal 
origin of each song, when known, being included in the description 
of the song. In the following consideration the music of the two 
tribes is regarded as a unit. 
The Mandan and Hidatsa songs show in abundance certain char- 
acteristics which were noted with less frequency in the songs previ- 
ously analyzed. Chief among these is the rest which appears in 19 
per cent of these songs. A rest appeared very rarely in the Chip- 
pewa songs, in 10 (less than one-half of 1 per cent) of the Sioux 
songs, and in 13 (more than 11 per cent) of the Ute songs. A rest in 
Indian music is different from a rest in the music of civilization. 
The tribes of Indians thus far studied by the writer, when singing, 
appear to take breath only when they require it, not at stated inter- 
vals, and the taking of breath is almost imperceptible. Rests, when 
they occur, are intentional and part of the song.*® In many in- 
stances it appears that a rest is introduced to give variety to the 
melody, though there are occasional instances in which a rest occurs 
at the end of a phrase. Syncopations (nota legato) occurred in 5 
of the 340 Chippewa. songs, in 3 of the 240 Sioux songs, and in 4 
of 110 Ute songs, but appear in 12 (11 per cent) of the present series. 
A glissando, or sliding from one tone to another, was prominent 
in Ute singing, but not used in any marked degree by the Chippewa 
or Sioux. A downward glissando appeared in the Ute songs of the 
bear dance and was said to imitate the sound made by a bear, and an 
upward glissando was used by them in songs connected with the 
hand game. (See Bull. 75, Bur. Amer. Ethn., pp.58and175.) In the 
present series a glissando appears only in the Mandan songs (see 
analysis of No. 1), and appears chiefly at the end of a phrase, the 
final tone of these songs being followed by a sliding downward of the 
voice, the degree varying from a whole tone to a fourth and being 
the same in all renditions of the song. This glissando, as well as 
%The same peculiarity has been noted in the music of India. ‘“ Rests are seldom 
written (except in order to break up the meter intentionally in a dramatic way) in any 
of their songs, at any rate not, as we should, on account of the words: ... They 
appear to take breath when they want to take it, not at the end of words.” Fox 
Strangways, Music of Hindostan, pp. 192-193. 
