44 
MUSEUM BULLETIN NO. 16. 
represent the rate of variability of any given myth from gener- 
ation to generation . 1 
PSYCHOLOGICAL-LITERARY ELEMENTS IN THE PLOT. 
The preceding discussion was concerned with the probability 
of the origin of the many different versions of myths through 
the free exercise of the author-raconteur’s play instinct. We 
shall now discuss the probability for still further divergence 
through the exercise of the author-raconteur’s literary-psycho- 
logical instinct. 
The Indian author-raconteur must have realized as much as 
any of our own novelists that the effective telling of a story 
depended on a number of devices, the development of a denoue- 
ment, cumulative effect, etc.; also that he was dealing with cer- 
tain psychological situations. It certainly is not presuming 
too much to say that he had sufficient power to analyse his 
myths and separate them in a rough way into the component 
elements of which they were formed. Let us illustrate this by 
examining the plot of the Winnebago myth of “Holy One.” 
What the author-raconteur had in mind was clearly the following 
plot and the following psychological situations. 
Scene I. Holy One’s brother must die. He is warned not 
to take a certain road. 
Scene II. The slayers of Holy One’s brother must be dis- 
covered and punished. Holy One must discover their identity 
from some unwilling agent. He must transform himself, wound 
but not kill the slayers. Then he must find out what has become 
of the slayers, again from an unwilling agent, kill the agent, 
disguise himself in his shape, and finally kill the slayers of his 
brother. 
Scene III. He must be pursued and escape. 
Scene IV. He must attempt to restore his brother to life 
and fail. 
1 1 know particularly of two Winnebago who have the reputation of telling myths remark- 
ably well, but of deviating considerably from the older versions. The older people look askance 
at them, but the younger, especially the members of the family, seem to show no displeasure. 
Under ordinary conditions, we might here have a test case of the perpetuation of markedly 
divergent versions. 
