Sullivan — 7'ongaii Somatology 255 



EVIDENCES OF MELANESIAN INTERMIXTURE 



Although the differences between Tongans and Saiiioans are very sHght, yet 

 almost without exception they point in the direction of Melanesia. Of the Samoans" 

 I said that considering the group as a unit there seems to be very little Melanesian 

 blood in evidence. On the basis of cultural or linguistic affinities it is common to 

 assume a large amount of Melanesian blood in all Polynesian groups. If such blood 

 exists it should be easily demonstrable. Melanesian intermixture should result in 

 a lower stature, longer heads, broader, shorter noses, shorter ears, more curly, 

 frizzly, or woolly hair, more beard and body hair, a smaller transverse fronto- 

 parietal index, a lower, narrower face, greater prognathism, and a heavier develop- 

 ment of the glabella and supra-orbital region. A large percentage of the difference 

 between Tongans and Samoans is of a nature that from purelv theoretical reasons 

 I suggested might be expected to result from Melanesian mixture. As a matter of 

 fact there are few or no careful and detailed descriptions of those Melanesian 

 groups that are geographically nearest to the Tongans and very meager data from 

 the area as a whole. In order to state with any finality what might be expected 

 from the mixture of Pol3aiesians and Melanesians, in lieu of any absolute data on 

 the question, detailed and accurate descriptions of several living Melanesian groups 

 would at least be necessary. 



Assuming, however, that I have stated with approximate accuracy what 

 might be expected in such a mixture, the analysis can be carried a step further. 

 As a test woolly, frizzly, curly, deeply waved, and, to a less extent, low-waved hair, 

 may be taken to indicate Melanesian physical mixture. If it does indicate Melane- 

 sian mixture and this mixture has taken place on a large scale within fairly recent 

 times, it is to be expected that persons with wavy and curly hair will show other 

 Alelanesian characteristics. By this I do not mean that there is necessarilv any 

 high correlation between the combinations in which physical characters are inher- 

 ited when two races mix but that, purely on the basis of chance, if curly, wavy 

 hair indicates the presence of Melanesian blood, it is reasonable to expect that the 

 curly, wavy-haired group, as a unit, will show a closer approach to the INIelanesian 

 average than will the straight-haired group. 



According!}' I have divided ni}- material on the basis of hair form into three 

 groups. The first group includes the straight-haired persons, the second group 

 those with low waved hair, and the third group includes all with deeply-waved, 

 curly, or frizzly hair. I have compared these three groups with the total series. 

 The averages of the three groups do not necessarily ecjual the averages of the total 

 series since the data for a number of persons who were doubtfully marked "straight 

 to low waves" or "low waves to deep waves" were excluded in making up the 



' Op. cit., p. 96. 



[25] 



