348 
ANTHROPOLOGY: R. H. LOWIE 
in English are classed together through the extensive use of reciprocal 
terms, members of a related pair addressing each other by a common 
term. But these systems are 'classificatory' only in an etymological 
sense of the word, the basis of the classification being wholly distinct 
from that which moulds the collateral-lineal terminology of the systems 
customarily designated as 'classificatory.' 
Summing up, we may say that there is a very high degree of correla- 
tion between the practice of exogamy and the ranging in a single cate- 
gory of collateral and lineal kin. The aberrant cases are relatively 
few and some of them are readily intelligible as the result of cultural 
influences from neighboring tribes. It is, of course, possible that the 
correlation may ultimately turn out to have an unexpected meaning, 
for example, that the emphasis on exogamy is misplaced and should be 
on definite organizations of any kind, whether exogamous or not. Such 
an interpretation might perhaps eliminate some of the at present anom- 
alous instances of non-exogamous tribes sharing the nomenclature of 
exogamous peoples. At all events, the North American data furnish 
strong evidence in support of the Tylor-Rivers theory. 
That the classification of kin by certain tribes is a function of the 
exogamous grouping, may be demonstrated in more rigorous fashion. 
Within the Siouan family there are tribes with kinship systems that not 
only fail to distinguish between collateral and lineal relatives but also 
class together members of distinct generations, which is contrary to the 
usual form of ' classificatory ' nomenclature. Rivers has suggested ex- 
ceptional forms of marriage to account for the exceptional mode of classi- 
fication; as a matter of fact it may be shown that the apparent exceptions 
are merely the result of an unusually consistent application of the exog- 
amous principle of grouping. 
The following are the facts. The Crow and Hidatsa, Siouan tribes 
tracing descent through the mother, class the father's sister's son with 
the father; the father's sister's daughter, father's sister's daughter's 
daughter and all her female descendants through females with the 
father's sister. It is to be proved that these classifications are con- 
nected with the exogamous social grouping. 
The facts in the case may be summed up by saying that a single term 
is appHed to male members of the father's clan regardless of generation, 
and a single term to female members of the father's clan who belong 
to his own and all descending generations. If this objective statement 
also represents the psychological basis of the grouping, the terminology 
should be modified as soon as we pass outside the clan. We pass outside 
the clan when we take not the father's sister's daughter's daughter, but 
