72 BUREAU OF AMERICAN ETHNOLOGY [BULL. 75 
No. 17. Final Song of Bear Dance (b) (Catalogue No. 693) 
Recorded by StneEr No. 4 
VoIcE ge 144 
MoraAcHE o= 144 
a nea Matty of morache similar to No. 2 
ag Pt get tet ees 
preaes awed Paieed ieeer ce 
Analysis.—This, like the song next preceding, is major in tonality, 
harmonic in structure, has a compass of six tones, and progresses 
only by intervals of the fourth and the major third. The present 
melody contains 15 progressions, 53 per cent of which are downward, 
and the preceding melody contains 67 progressions, 51 per cent of 
which are downward. With these melodic similarities it is interest- 
ing to note the difference between the rhythmic character of the two 
songs. Only one rhythmic unit occurs in the song under analysis, its 
four repetitions comprising the entire melody. The preceding song 
contains three short rhythmic units with many intervening measures. 
The present song contains eight changes of measure lengths, while the 
preceding song is in 2—4 time throughout. 
Piots oF Bear Dance Sones 
In the plots of a large majority of the Bear dance songs we note a 
resemblance to the plots of Sioux songs concerning animals. (See 
Bull. 61, p. 204.) This resemblance consists in the touching of a low 
17 A form of graphic representation, or “plotting,’’ of melodies has been devised by the writer for the pur- 
pose of making the trend of melodies moreapparent. The generalmethod employed is similar to that used in 
showing graphically the course of a moving object. The loci of the object at given periods of time are deter- 
mined and recorded, the several positions being connected by straight lines. In any use of this method the 
interest centers in the several points at which the object is located, it being understood that thelines connect- 
ing these points are used merely as an aid to observation. In the present adaptation of this method the pitch 
of the accented tones in a melody is indicated by dots placed at the intersection of coordinate lines, the 
horizontal coordinates representing scale degrees and the vertical coordinates representing measure lengths. 
These dots are connected by straight lines, though the progress of the melody between the accented tones 
would, in many instances, vary widely from these lines ifit were accurately plotted. The use of accented 
tones exclusively in analyzing these songs has already been employed, the structure of the melodies being 
determined by the pitch of contiguous accented tones. One of the reasons which seem to justify this 
usage is the observation that, when differences appear in the several renditions of an Indian song, these 
differences usually are found to be in unimportant progressions between unaccented tones. Since thesole 
purpose of these plots is to show the trend of the melodies, it seems permissible to omit from the represen- 
tation, not only the unaccented tones occurring in the melody, but also a distinction between whole tones 
and semitones in progressions, and a distinction between double and triple time in measure lengths. It 
is obviously desirable that the graphic representation be as simple as possible, the more detailed observa- 
tion of the melodies being contained in the tabulated and descriptive analyses. 
we 
fae =e —T eae 
