OP THE HUMAN FAMILY. 149 



for the brother of a father and the sister of a mother, must be rcgardtnl in tlio 

 hght of moditications of the primitive form by particuLir usage rather than as 

 deviations from uniformity. 



IV. llehitionships of the Children of a Brother and Sister. It is evident that the 

 relationship of a cousin was unknown in the original system, and that it was an 

 aftergrowth, or further development, designed to remove a blemish. The four 

 different forms in Avhicli the relationships of the children of a brother and sister 

 appear, render it difficult to determine which was the primitive form, only that 

 cousin was not. The principles of the system required that they should stand in a 

 more remote relationshij) than that of brother and sister ; and thus we are led to 

 the inference that it was either that of uncle and nephew, or that of son and 

 father. 



V. Marriage Relationships. There are a number of diversities in these relation- 

 ships, but a sufficient number are constant to establish the unity of the system from 

 this source of evidence alone. 



VI. Mergence of Collateral Lines. In a few of the nations some branches of the 

 collateral lines are more abruptly merged in the lineal than the common form 

 allows ; but of this peculiarity no explanation can be given. 



We are now the better prepared to take up the system of relationship of the 

 Ganowanian family in its several branches ; and by an examination of its structure 

 and details, to verify the preceding propositions, and also to trace this form of the 

 classificatory system to its limits. In no otlicr manner can its remarkable charac- 

 ter, as a domestic institution, be understood or appreciated, or its value estimated 

 for ethnological purposes. 



