yh Records of the Indian Museum. [Vol. XXIII, 



Anglo-Indian V = 40672 

 (Selected Weighted Mean 7= 37622 



Anglo-Indian Difference = -3000 

 P.E. of Difference = ± -1245 



D a -1 



.v = — =103 approximately 



(T 



Prom Bioinetric Table II, |(»i + «) = 94 84 49 3 



|(i -0)= -05 15 50 7 



5'i% will be more variable, while 10 2% will differ more 

 from the weighted average than Anglo-Indians. 



Thus even when compared with the weighted Mean, Anglo-Indian 

 Variability is not significantly greater than average Variability. 



We have seen that U.S.A. recruits raise the weighted Mean 

 very considerably. But it is not at all certain that the recruits of 

 the U.S.A. Army are possessed of any great degree of homogeneity. 

 One would surmise rather that they are heterogeneous in character. 

 Let us see the effect of leaving out U.S.A. recruits. 



Omitting U.S.A. recruits we get 



Weighted Mean V= 3 6413 

 Weighted S.D. of V= -2509 

 Weighted P.K. of V= ± -1683 

 Anglo-Indian V = 4-0672 



Weighted Mean V= 36413 



Difference = -4259 + 16-83 



D -4250 

 *=- = -3-22 = r 70 



a 2509 

 From Biometric Table II, |(i+«) = *95 54 34 5 



■J(i -<0 = ' ( H 45 65 5 



4'5% will have greater Variabilities than the Anglo-Indian 

 mple. As regards the Coefficient of Variation, this is the most 

 stringent test we can apply with the non-Indian material at our 

 disposal. We find 



Anglo-Indian Variability is within the limits of probability of 

 homogeneous Variation. Study of the Coefficient of Variation for 

 Stature does not enable us to assert definitely that the present Anglo- 

 Indian sample is heterogeneous in character. 



I shall now consider the whole series of non-Caste samples 

 including the Mediterranian samples, I have omitted the separate 

 y groups for tin- New South Wales Criminals and the Oxford 

 student data. As all these have greater Variability than the 

 .\\ , the stringency of our test will not be diminished by this 



rejection. 1 Another reason why I have omitted the different age- 



■ Jim ussion on p. 72. 



