Tlia Groioth of Children. 



05 



Tablk No. 24. 



ShoiruKj Reldtion of Height to Weight in Groiviiig Girls. (Weiglit 

 given in pounds.) 



HEIGHT, iiuhes. 



42, 

 43, 

 44, 

 45, 

 46, 

 47, 

 48, 

 49, 

 50, 

 51, 

 52, 

 53, 

 54, 

 55, 

 56, 

 57, 

 58, 



61, 

 62, 



The following conclusions may be drawn from an examina- 

 tion of the tables : — 



because it rests upon the assumption that the averaj;e weight for a given age is the same 

 as the average weight of all individuals, without regard to age, whose height is equal to 

 the average height for that age. This assumption clearly involves a trifling error, for, 

 since the weights of growing children increase approximately as the 2.7 powers of the 

 heights, it is evident that at any given age the weight of those children who are above the 

 average height will tend to raise the average weight for that age more than the weights of 

 the children below the average height will tend to lower it, supposing the observations to 

 be uniformly distributed on both sides of the average according to the binomial curve of 

 Quetelet ; conse<iuently, the average weight for a given age will be somewhat greater than 

 the average weight of all the individuals, regardless of age, whose height is equal to the 

 average height for that age. Notwithstanding these defects, the method has been adopted, 

 first, because it is believed that the errors involved are so small as to be of no practical 

 importance ; and, secondly, because relative rather than absolute values were sought, and 

 a comparison between several sets of observations is not prevented by a small constant 

 error running through them all. 



