On Evil in Small Doses 



who are markedly above or below the average are very 

 few. The number, for instance, of individuals in England 

 who are below 4 feet 6 inches or above 6 feet 6 inches 

 in height is exceedingly small. The number who are 

 5 feet or 6 feet in height is very great indeed, but pro- 

 bably insignificant as compared with those who are 

 between 5 feet 5 inches and 5 feet 7 inches. 



This natural variation about the mean has been tested 

 in many ways. Thus in order to get rid of the effect of 

 changes in the environment and the effect of crossing 

 in the ancestry, Johansen established " pure lines " by 

 breeding from a single seed, and endeavoured to trace 

 the character of the variations in its descendants. 2 



The results seem to show the usual variation about a 

 mean character as roughly sketched above.* 



As has been shown in a previous chapter, practical 

 men do, as a matter of fact, obtain valuable varieties by 

 breeding from extreme individuals much above or below 

 the average in one particular character. 



In certain experiments with 40,000 sugar-beets, Messrs. 

 Kuhn at Naarden in Holland found that the amount of 

 sugar varied from 12 per cent, to 19 per cent., with an 

 average of 15.5 per cent. By selection it has been found 

 possible to raise the average quite perceptibly, 3 so also 

 with the percentage of proteids and of oil in Indian corn. 



But by such methods the selectors are just waiting 

 on the plant's whimsies in the way of variations. 



If we could affect or stimulate such tendencies to 

 extremes, even if we could, by artificial means, changes 

 of culture, and the like, set it varying, the benefit would 

 be enormous. The whole process of improvement would 

 be accelerated, and we might incidentally obtain that 

 complete, rounded off, and satisfactory theory of evolu- 

 tion which has not as yet been discovered. 



* It was first clearly stated by De Quetelet. 

 328 



