52 BUEEAU OF AMERICAN ETHNOLOGY [bull. 40 



distribution is fully explained by the assumption of extended dissem- 

 ination. 



It is true, for instance, in the case of similar traditions which are 

 found distributed over large districts. An example of this is the 

 story of two girls who noticed two stars, a bright one and a small one, 

 and wished these stars for their husbands. The following morning 

 they found themselves in the sky, married to the stars, and later on 

 tried to return to the earth by letting themselves down through a 

 hole in the sky. This rather complex tale is found distributed over 

 the American continent in an area extending from Nova Scotia to the 

 mouth of the Mississippi river and westward to the Rocky mountains, 

 and in places even on the Pacific ocean, for instance, in Alaska and in 

 the state of Washington. It would seem difficult to assume, in a case 

 of this kind, the possibilit}'' of an independent invention of the tale at 

 a number of distinct points; but it must be assumed that, after the 

 tale had once attained its present form, it spread by dissemination 

 over that part of the continent where it is now found. 



In extreme cases the conclusions drawn from these two types of ex- 

 planation seem quite unassailable; but there are naturally a very 

 large number of others in which the phenomenon in question is neither 

 sufficiently complex, nor distributed over a sufficiently large contin- 

 uous area, to lead with certainty to the conclusion of an origin by dis- 

 semination; and there are others where the sporadic distributions seem 

 curiously arranged, and where vague possibilities of contact occur. 

 Thus it happens often that a satisfactory conclusion cannot be 

 reached. 



We must also bear in mind that in many cases a continuous distri- 

 bution may once have existed, but may have become discontinuous, 

 owing to the disappearance of the phenomena in question in inter- 

 mediate regions. If, however, we want to follow a safe method, we 

 must not admit such causes for sporadic distribution, unless they can 

 be definitely proved by other evidence; otherwise, the way is open to 

 attempts to bring into contact practically every part of the world with 

 all others. 



The general occurrence of similar ethnic phenomena in remote 

 parts of the world admits also of the explanation of the existence 

 of a certain number of customs and habits that were common to 

 large parts of mankind at a very early period, and which have main- 

 tained , themselves here and there up to the present time. It can 



