238 Memoirs Bcniicc P. Bishop Museum 



of the lips is recorded as thin, niechuin, or thick. Proonathisin, wliich in a Hvin^- 

 person is a complex and somewhat elusive character, is described as absent, slioht, 

 medimii. or marked. In the ears the development of the lobes (small or large, 

 attached or separate), the roll of the helix (rolled one-third, two-thirds, three- 

 thirds, or flat), and the ])resence or absence of Darwin's tubercle were recorded. 



Particular care was taken to record the form of the up])er incisor teeth with 

 a view to determining the presence or absence of the shovel-shaped incisors. 

 Although ])rimarilv described by tlrdlicka'' as shovel-shaped this condition of the 

 incisor and other teeth has more recently been described by the same author as 

 keilodonty and koilomorph}-. As the fossa is dependent on the formation of the 

 rim, it will be sufficientlv clear and less cumbersome to discuss this condition in 

 terms of rim development or keilodonty. Jn this paper classes of no rim, trace of 

 rim, medium rim, and marked rim correspond to Hrdlicka's classes of no shovel, 

 plain trace, semi-shovel, shovel-shaped. The condition is well described by 

 Hrdlicka:* "The lingual surface of the well developed shovel-shaped incisor is 

 verv striking. The usual moderate concavity from above downward is replaced by 

 a triangular to rtnmded or oblong deep fossa. I'he base of the fossa is formed by 

 the free edge of the tooth, its summit reaches ujiwards near to the gum. The fossa 

 is bounded laterally and generally also distally, hence on all sides by a stout rim 

 of enamel." 



DESCRIPTION OF THE TONGANS 



The results of the seriation and averages are summarized in Tables I, II, 

 and III. In Table I we tind a fairly normal distribution in all anthropometric 

 characters. The number of persons concerned in each character is so small that 

 any departure from the normal cannot be regarded too seriously. Doubling the 

 class interval is usually sufficient to smooth the curve. Even after doing this, 

 however, the distribution of head length and face width present a somewhat skewed 

 distribution. Indications of bimodality are also noticeable in the bigonial diameter, 

 face height, and face index distribution. At present the significance or non-signifi- 

 cance of these facts is not clear. 



^ Hrdlick.i. Ales. Sliovel-shaped teeth: Am. Jour, of Pliys. Antlir., vol. iii, p. 429, 1920. 

 ' Op. cit.. p. 429. 



[8] 



