SECTION IX. SUMMARY OF CONCLUSIONS. 



Statistical. 



(i) For stature, with samples of the order of 200, a group- 

 ing unit of 50 mm. is fairly satisfactory. For calculating fre- 

 quency constants the grouping unit should be less than 3*5\/A T 

 (for samples of size N). 



(2) Sheppaid's corrections lead to substantial improvement 

 in the frequency constants and should never be omitted. With 

 small samples finer corrections (e.g. Pairman and Pearson) are use- 

 less. 



(3) The Gauss-Laplacian normal curve is adequate for 50 

 mm. grouping. For proper graduation, i.e. for testing goodness 

 of fit the grouping should be broader than 700/^/N. 



(4) The actual frequency curve belongs to Type IV of Pear- 

 son's Skew family. There is small positive asymmetry with the 

 Mode greater than the Mean, and a slight tendency towards lepto- 

 kurtosis. The general nature of the distribution is similar to other 

 homogeneous distributions. 



(5) There is no definite evidence of statistical heterogeneity. 

 The Anglo-Indian sample may be accepted as a statistically homo- 

 geneous sample. 



Anthropological {Stature). 



(1) The more highly civilised races have greater variabilities 

 than the average. 



(2) This greater variability of more highly civilised races seem 

 to be only moderate in degree and is never excessive, 



(3) Interracially, taller races seem to be more variable than 

 the shorter (both as regards the absolute and the relative variabi- 

 lity). 



(4) Indian Castes and Tribes are significantly less variable 

 than the average. 



(5) Anglo-Indian variability is greater than Indian Caste va- 

 riability but is of the same order as the variability of modern 

 European races. 



(6) The variability of the Anglo-Indian sample though greater 

 than the average is not beyond the range of possibility of homo- 

 geneous variability 



(7) The Anglo-Indians seem to be rather precocious in 

 growth, and there is some indication of the arrest of growth 

 occurring at an earlier age than in the case of European races. 



(8) Variability of the smaller age-groups is distinctly less, 

 showing a decrease of variability with time (or increasing homo- 

 geneity of the younger generation). 



