206 
There were several reasons for this. In ritual songs handed down 
from one generation to another expressions became archaic and finally 
meaningless. Then there was a frequent transmission of songs from 
group to group and from tribe to tribe even when the latter spoke 
unrelated languages, and in such transfers both the words and the 
music underwent modification; so some of the ritual songs of the 
western Carrier contained what were undoubtedly Tsimshian words, 
although so greatly transformed that they were no longer intelligible. 
Often again the original compositions bore only catchwords whose 
explanations soon became lost. But perhaps the principal reason 
was that the emotions found their outlet not in the words^ but in 
the music, the metre, the more or less variable rhythm, and the 
motor activities that accompanied the singing, just as we ourselves 
give little heed to the words of a dance song provided the music 
is pleasing and both rhythm and metre appropriate. 
There was no part-singing in the music of the aborigines; men 
merely sang the melodies an octave lower because of the lower range 
of their voices. The songs were much more irregular than ours, for 
not only did the music modify the words (e.g., by changing their 
accents), but the words affected the melodies far more than we 
permit in our sophisticated art where the music is so often divorced 
from any text and developed for its own sake entirely. Tliere were 
rhythmic patterns corresponding sometimes to our verses and refrains, 
but they were seldoin uniform throughout a composition. The 
measures were generally in two, three, or four-part time, but some- 
times also in five and even seven part; five-part time was fairly 
common on the Pacific coast.- Yet the natives seldom adhered to 
any strict measures, but altered the time continually, sometimes from 
exigencies of breath, sometimes to suit the words and their domin- 
ating accents. It is quite common, in fact, to find long, rhythmic 
units without any recognizable measures at all. 
The octave naturally provided the basis for the musical scale, 
though most Indian songs ranged less than that interval and a play 
of melodies on octaves (apart from the differences in men’s and 
women’s voices) was rare. Its divisions, however, did not coincide 
exactly with the tones and semitones familiar to Europeans. The 
1 No doubt tlie words of new sfings aroused some emotion, especially in their composers. 
2 Boas, F.: “Primitive Art,” p. 340, Oslo, 1927. 
