28 BUREAU OF AMERICAN ETHNTOLOGY [Bull. 187 



women intone meekly and rarely do they offend the ear with shrill 

 labored soprano. Fannie Stevens, in fact, used to chant towisas in 

 a middle register, lower than the male rendering of new ^skanye 

 songs. In both sexes leading singers show their individuality in their 

 rendering, the men to a much greater extent. The soaring Eagle 

 Dance songs differ from the husky, propulsive Bear Dance songs; 

 the pulsating Feather Dance songs are in no danger of being confused 

 with the panting Husk Face chants, or the crisp Chicken Dance songs 

 with the sustained melodies of Hand-in-hand. Similarly, Chauncey 

 Johnny John's free eccentric huskincss differs from Albert Jones' 

 sonorous sturdiness. Ed Curry has adopted some of Chauncey's 

 ways and blended them with his calm personal style. 



The patterns of the songs will presently be examined more closely. 

 It is easier to observe the great variety from cycle to cycle than to 

 summarize pervading unifying qualities. In no case is the pattern 

 arbitrary. The flow of the melody and its relationship to the 

 instruments is safeguarded by tradition. 



METHOD OF STUDY 



Transcription: 



The most satisfactory agent for transmission of melodies from 

 singers to musical staff remains the phonographic disk, and this has 

 provided the material for most of the enclosed reproductions. A few 

 Scalp songs were written down by ear, and some recent tape recordings 

 have been transcribed directly from a reel. All of these recordings 

 are the work of William N. Fenton in the course of 18 years: 1933, 

 1941, 1945, 1948, and 1951. Tonawanda recordings materialized as 

 a project of Martha Champion Randle with the aid of Dr. Fenton. 



Most of the songs have lent themselves fairly well to conventional 

 notation, with the addition of a few symbols for "blue" notes slightly 

 below or above pitch and for a few other characteristics that did not 

 defy translation into symbols. In the first drafts time signatures 

 were used, but they have now been abandoned as unsuited to the 

 Indians' musical perception. Any metrical divisions are arbitrary 

 and perforce tentative. Certain divisions occur only at the end of 

 phrases and larger sections. Similarly, key signatures and accidentals 

 have been avoided when possible. For better writing and for com- 

 parative purposes, most songs have been transposed downward from 

 a third to a fifth. This seemed feasible because of the Indians' 

 relative rather than absolute sense of pitch. 



The percussion accompaniment appears directly beneath the 

 corresponding melody, in only a few instances in full. As a rule, the 

 characteristic beat appears at its first occurrence. A change of beat 

 during the song or in an ensuing song is similarly indicated. Char- 



