368 BUREAU OF AMERICAN ETHNOLOGY [Bull. 182 



table 6). Except for size, they also agree fairly well with the much 

 later, northern Muiisee group. The Clarksville crania, on the otlier 

 hand, agree better Avith their local predecessors, the Tollifero group, 

 than with any of the other groups. 



About the only conclusion possible regarding cranial type is a 

 tentative confirmation of the observation made nearly 40 years ago 

 by Hrdlifika (1922, p. 113) and confirmed by Newman and Snow 

 (1942, p. 407), that the older population of the Southeast was doli- 

 chocranic, while later populations were more brachycranic. Whether 

 this change in cranial type is due to intermixture with indigenous 

 southern and eastern brachycephals, as Hrdlicka (1916, 1922) 

 thought ; or to incoming brachycranics, as suggested by Newman and 

 Snow (1942) ; or to evolution in situ, there is not yet enough evidence 

 to decide. Eegarding body size and build, it would seem that the 

 earlier peoples were somewhat shorter than the later, but that increase 

 in body size did not change body proportions. 



POPULATION COMPOSITION AND LENGTH OF LIFE 



The distribution of ages and sexes at the Tollifero and Clarksville 

 sites show a number of differences (table 8). Infant mortality at the 

 two sites is about the same, about 16-17 percent, but the proportion 

 of juvenile deaths at the Clarksville site has nearly doubled. At the 

 Tollifero site, about one-fourth of the persons represented failed to 

 reach adult life; but at the Clarksville site, the juvenile mortality had 

 increased to nearly 45 percent, with a large increase in deaths among 

 young adults. The sex ratio also offers certain peculiarities. At the 

 Clarksville site the sex ratio is about 3 : 2 in favor of the males ; but 

 at the Tollifero site, the ratio is nearly 5:2. At both sites, the males 

 appear to have outlived the females. Although early death was the 

 rule at both sites, in general the adults at the Tollifero site appear 

 to have reached slightly more advanced ages than those from Clarks- 

 ville. Tliis is hardly what one would have expected from the im- 

 proved diet of the later peoples. It is well to recall at this point that 

 these distributions are unlikely to be an artifact of selection by preser- 

 vation, for the population estimates are based on the total number of 

 skeletal remains fomid, including those too fragmentary to retain. 

 The latter, as mentioned above, were described in field notes, and/or 

 in the laboratory before being discarded; although the accuracy of 

 these records could not be checked agamst the specimens, they are 

 unlikely to have introduced differences as marked as those noted. 



It is of course possible that the small size of the samples may have 

 influenced the distribution of ages and sexes at the two sites. Tliis 

 possibility becomes less likely when these distributions are compared 

 with age and sex distributions of the skeletal material from Hiwassee 



