BOAS] HANDBOOK OF AMERICAN INDIAN LANGUAGES 51 



nated only a single time, and were diffused by the influence of 

 one people upon another; or it may be that tlie}^ are due to an 

 independent origin in many parts of the world. 



Tliis alternative is present in the explanation of all ethnic phe- 

 nomena, and is one of the fundamental questions in regard to which 

 the ethnologist, as well as the investigator of languages, must be 

 clear. In the older considerations of the position of the American 

 race among the races of man, for instance, it has always been assumed 

 that occurrence of similar phenomena among the peoples of the 

 Old World and of the New proved genetic relationship. It is 

 obvious that this method of proving relationship assumes that, 

 wherever similarities occur, they must have been carried by the 

 same people over dift'erent parts of the world, and that therefore 

 they may be considered as proof of common descent. The method 

 thus applied does not take into consideration the possibility of a grad- 

 ual diffusion of cultural elements from one people to another, and 

 the other more fundamental one of a parallel but independent 

 development of similar phenomena among different races in remote 

 parts of the world. Since such development is a logical possibil- 

 ity, proofs of genetic relationship must not be based on the occur- 

 rence of sporadic resemblances alone. 



A final decision of this vexed problem can be given only l)y historical 

 evidence, which is hardly ever available, and for this reason the 

 systematic treatment of the question must always proceed with the 

 greatest caution. 



The cases in which isolated similarities of ethnic phenomena in re- 

 mote parts of the world have been recorded are numerous, and many 

 of these are of such a character that transmission cannot be proved at 

 all. If, for instance, the Indians of South America use sacred 

 musical instruments, which must not be seen by women, and if 

 apparently the same custom prevails among the Australian aborigines, 

 it is inadmissible to assume the occurrence of what seems to be 

 the same custom in these two remote districts as due to transmission. 

 It is perfectly intelligible that the custom may have developed inde- 

 pendently in each continent. On the other hand, there are many cases 

 in which certain peculiar and complex customs are distributed over 

 large continuous areas, and where transmission over large portions of 

 this area is plausible. In this case, even if independent origin had 

 taken place in different parts of the district in question, the present 



