BIOMETRY . 57 



been due chiefly to the valuable work of Karl Pearson. The 

 underlying idea in biometry is to apply to the study of evolu- 

 tion the precise quantitative methods followed in the study 

 of physics and chemistry with such signal success. 



Biometry is the statistical study of variation and heredity. 

 It deals with masses, not with individuals, differing in this 

 respect from the method of Darwin and Bateson. It seeks 

 to obtain a quantitative estimate, as precise as possible, of 

 variation in one generation, and to compare with this a 

 similar quantitative estimate of the next generation and then 

 by comparing these to learn in what direction evolution is 

 taking place and at what rate. In some cases it has at- 

 tempted to discover the direction of evolution from the 

 character of the variation within a single generation. 



Biometry is best adapted to deal with continuous varia- 

 tion, but it has its uses also in dealing with discontinuous 

 variations. Its ideal, to make biological investigation more 

 accurate and comprehensive, is wholly commendable. But 

 mere collection and compilation of biological statistics will 

 not advance knowledge unless brought into relation with 

 other facts about living things, and it is in this respect chiefly 

 that biometricians have sometimes erred, drawing unwar- 

 ranted conclusions from their statistical data. 



Biometry means literally the measurement of living things. 

 It is obvious that it can deal only with characteristics which 

 are measurable, such as linear dimensions, volume, weight, 

 or number of parts. One of the cases most carefully studied 

 by Galton was human stature. This case illustrates very 

 well the methods and results of biometric study. 



Measurements made at the Harvard gymnasimn of the 

 height and weight of one thousand students of ages eighteen 

 to twenty-five are classified in Table 1. In order that the 

 number of classes may not be too great for convenient sta- 

 tistical treatment, height classes are formed of 3 cm. each. 

 Thus students measuring 155, 156, or 157 cm. are all placed 

 in a common class, whose middle value is 156 cm. In dealing 

 with large numbers, the probability is that each of the three 



