8 BUKEAU OF AMERICAN ETHNOLOGY [bull. 61 



noted moro fn^quently among the Chippewa than among the Sioux. 

 Indian singers occasionally use much freedom in the arrangement of 

 the phras(>s of a song in its repetition. It is therefore considered 

 advisable to analyze^ only the direct and simplest rendition of a song. 

 A "plot" of each melody is also made, as describfid and illustrated 

 on pages 51-54. 



The words of certain songs arc in a ''sacred (esoteric) language/' 

 which disguises their meaning. See words of ^Ilde' songs in Bulletin 

 45; also footnote, page 120 of this work. 



In the present volume the use of cross-references, which are found 

 in Bulletin 53, is discontinued, and in their place will be found ref- 

 erences to the analysis of the first song containing the peculiarity 

 under consideration. Thus if a rest occurs in a song there will be 

 found a reference to the analysis of song No. 79, which contains a 

 list of all songs in which rests appear. The wTiter commends the 

 use of the index of this book and of that of Bulletin 53 to those 

 who wish to study the analyses closely. Songs can be traced also in 

 the following manner: Let us suppose a reader is seeking songs from 

 which the third tone of the octave is lacking. By consulting the 

 index of Bulletin 53 a reference is found to page 5, on which the 

 persistence of the third and fifth is considered; also, to the songs in 

 both Bulletin 45 and Bulletin 53, from which the third is lacking. 

 Fidl treatment of a peculiarity is usually given with the analysis of 

 the first song in which it occurs, but in tliis instance it is given in 

 connection with other peculiarities in Bulletin 53, song No. 53, page 

 140. It was there noted that more than one-half of the entire group 

 were songs concerning or sung by women, but the proportion of women 

 singers is much larger in the present series, while the proportion of 

 songs lacking the third is much smaller (see analysis of song No. 22). 

 Another method of tracing songs is by means of the tabulated analy- 

 ses. Thus a student in search of songs from which the third tone of 

 the octave is lacking would turn to Table 6, pages 21-23 of Bulletin 

 53, note the classes in which such songs are found, and trace them 

 through the tabulated analyses of these several classes. This is the 

 more interesting of the two methods, as it shows the frequency of its 

 occurrence and also suggests a relation between the peculiarity under^ 

 consideration and the class of the song in which it occurs. Having 

 noted that the omission of the third occurs in only 3.5 per cent of 

 the Chippewa songs, the student ascertains in the same way from 

 the present work on Sioux songs that it is absent from only 5, less 

 than 1 per cent of these songs. The tabulated, as well as the descrip- 

 tive, analyses are intended to assist a careful, intelligent observation 

 of Indian music. Both the means used and the results attained 

 should be understood as anticipating a broader as weU as a more 



