548 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1941 



The following list shows only a portion of the work, as many 

 hundreds of songs have been recorded and not transcribed. The 

 summary comprises work from September 1907 to November 30, 1941. 



1. Transcribed records submitted to the Bureau of American Ethnology 



and transferred in 1940 to the National Archives for permanent 

 preservation 2, 237 



2. Transcribed records submitted to the Bureau after the collection was 



transferred to the Archives 150 



3. Transcribed records in possession of Southwest Museum, Los Angeles 



(copies of 33 of these included in item 1) 205 



4. Transcribed records in possession of North Dakota Historical Society 40 



Total 2, 632 



CONCLUSION 



The two principal obsei'vations made by those who have listened to 

 the singing of Indians are that it is chiefly rhythmic and that it is 

 minor in character. The rhythm of Indian singing appears first 

 because of its prominence and insistence. The songs heard by a 

 casual observer are generally the songs of dances, but a study of the 

 recorded melodies shows that the rhythm of important Indian songs is 

 more elaborate than the rhythm of corresponding songs in our own race. 

 A desire to check these and other impressions prompted my analysis 

 of recorded Indian songs. It was not difficult to assign a keynote to 

 most of the melodies by the test of the ear, and the songs were divided 

 into two groups, major and minor, according to the interval of the 

 third and sixth tones above this apparent keynote. The term "key" 

 was avoided and the term "tonality" decided upon, partly at the 

 suggestion of CJiarles K. Wead, examiner, United States Patent Office, 

 about 1909. It was found that more than three-fifths of 180 Chippewa 

 songs under analysis were major in tonality. In subsequent analyses 

 of larger groups of songs it was found that the minor third was the 

 most frequent interval except the whole tone, which is generally a 

 passing tone. The prominence of this interval had given the impres- 

 sion that the songs were minor in tonality, according to our musical 

 system. Continuing this investigation, all the intervals in large groups 

 of songs were expressed in terms of a semitone, and the average progres- 

 sion was found to contain approximately a tone and a half which is a 

 minor third. This table of analysis was last used in my "Yuman and 

 Yaqui Music" (Bur. Amer. Ethnol. Bull. 110, table 13, p. 34, 1932) 

 wliich shows that the average interval in a cumulative analysis of 1,343 

 songs contains 3.03 semitones. 



The first tabulated analyses used in my work were nine in number, 

 contained in my first book, "Chippewa Music" (Bur. Amer. Ethnol. 



