10 BUREAU OF AMERICAN ETHNOLOGY [bull. 53 



excitement. The dream songs were undoubtedly composed under 

 abnormal conditions, but no drum was used in their composition and 

 the present study concerns only the manner of their rendition. In this 

 connection it is interesting to note that, according to Beau (1835) and 

 to Barth and Roger (1841),^ the rhythm "of the adult heart, beating 

 60 to 80 and acting normally " is a triple rhythm. The exact rhythm 

 described by these authors is found in two of the Chippewa songs — 

 the song of the war messenger and that of his return (Nos. 11, 12). 

 The \mter has frequently heard this rhythm when the drummers 

 began their performance (see Bulletin 45, p. 6); gradually they 

 changed to that most often recorded on the phonograph, in which 

 the unaccented stroke precedes, instead of follows, the accented stroke. 



In all theMide' songs and in 53 per cent of the war songs there is a 

 drum-rhytlim, of rapid unaccented strokes, two of which are approxi- 

 matel}'' equal to one metric unit of the melody. It is stated that under 

 certain conditions, ''especially a moral emotion or violent ph3"sical 

 exertion," the triple rhytlmi of the heart becomes " allied to a double 

 measure."' The collection of additional data may throw more light 

 on a possible connection between the action of the physical organism 

 and the form assumed by primitive musical expression. 



The next observation concerns the rhythmic unit, or motif (Table 

 19), which appears to constitute the rhythmic nucleus of the song, 

 as the mterval forms its melodic nucleus. As a basis for this classifi- 

 cation, a rhythmic unit w^as defined as ''a group of tones of various 

 lengths, comprising more than one count of a measure, occurring 

 at least twice in a song and having an evident influence on the 

 rhythm of the entire song." According to this basis of classification 

 it was found that 62 per cent of the songs contain a rhythmic unit, 

 while in many other instances the song itself possesses a rhythmic 

 completeness which constitutes it a unit. One hundred and ninety- 

 one songs contain a rhythmic unit, and in 132 songs (69 per cent) 

 the unit occurs in the first measure, showing, as in Table 15, a direct- 

 ness of "attack." 



There are four ways in which a rhythmic unit is used to form 

 a Chippewa song: First, it is continuously and exactly repeated 

 throughout the song (see No. 26) ; second, it is repeated continuously 

 except for a measure or two having a dift'erent rhythm, thus breaking 

 the monotony and gi%'ing character to the rh^rthm of the song as a 



> Dictionnaire dePltysiologie, Richet, Ch., editor, Paris, Tomeiv, 1900, p. 74. "Beau compara un bettement 

 de coeur a une mesure a trois temps, dans laquelle le premier temps serait occupe par le premier bruit, le 

 deuxi&me par Ic deuxieme bruit, le troisieme par le grand silence. . . . D 'apres Barth et Roger le rhythme 

 repr&ente une sorte de mesure a trois temps, dans laqueUe le premier bruit occupe le tiers environ; le petit 

 silence, ii peu pr^s un sixieme; le deuxieme bruit, un sixifeme; et le grand silence, le dernier tiers." 



- Ibid., p. 75 (signed by Laliousse). "Si, au contraire, les battements du coeur sont acci51eres, le silence 

 duninue et Ton n'a plus qu'une mesure qui se rapproche de la mesure a deux temps. . . . C'est surtout 

 quand une emotion morale, ou violent exerciee physique agissent sur le cceur de I'honmie, ou quand il est 

 le sifege de certains etats pathologiques." 



