THE MEASUREMENT OP VARIATION. 7 



sary. All that they would demonstrate would be the 

 fact that variation of a similar nature though of a 

 varying degree is present in all organisms, to what- 

 ever class of the Animal or Vegetable Kingdom they 

 belong. Should more evidence of this kind be de- 

 sired, the reader is referred to Wallace's book on " Dar- 

 winism ' (Chapter III). Here an admirable series of 

 diagrams is given, illustrating the variation in several 

 species of lizards, birds, and mammals. The diagram 

 given above is modelled on the plan adopted by Wal- 

 lace, and still earlier by Galton, as the one best adapted 

 for bringing before the eye the facts of individual 

 variability. 



In the above diagram the measurements of only 28 

 different individuals are given, and hence we are not 

 able to gather much as to the distribution of the differ- 

 ent measurements about their means. Supposing that 

 instead of tens, fifties or hundreds of the animals had 

 been measured, what should we expect to find? Would 

 there or would there not be just as many animals with a 

 very long or very short body length, as with a moder- 

 ately long or moderately short one, or as with a nearly 

 average one? Such a question as this is also best an- 

 swered by reproducing the measurements diagram- 

 matically, though in this case they must be arranged on 

 a different system. In the accompanying diagram, Fig. 

 2, 65 measurements of the wing of Sterna hirundo, 

 recorded in the above-mentioned paper of Allen, are 

 plotted out. Here each dot represents one measure- 

 ment, all the measurements between 10.46 and 10.55 

 inches being placed over the number 10.5, and so on. 

 The mean of all the measurements is 10.49 inches, and 



