Iiiflueiice of Selection 203 



a sufficient number of individuals had been examined. It is 

 true that these extreme individuals may be more common after 

 careful selection through many generations, but the species can- 

 not be carried as a whole beyond or much beyond the upper limit 

 to which it sometimes attains without selection. Thus it appears 

 impossible by the process of selection of fluctuating variations 

 to transform the species into anything new or different from what 

 it was. 



There is a further point of especial importance in this connec- 

 tion. If vigorous selection of the extreme individuals is not 

 carried out in every generation, there is a quick return to medi- 

 ocrity, and what has been laboriously gained is quickly lost, — 

 the species left to itself swings back to its more stable condition. 

 The effects of selection are only temporary, and nothing perma- 

 nent can be acquired in this way. Consequently we are justified, 

 I think, in denying that through natural selection of indi- 

 vidual differences the process of evolution could have taken 

 place. 



We cannot leave the subject without referring to Galton's im- 

 portant law of ancestral inheritance. This law states that each 

 individual inherits on the average from his two parents together 

 50 per cent, or one half of the whole inheritance ; from the four 

 grandparents 25 per cent, or one fourth; from the eight great 

 grand-parents one eighth of the whole, etc., the total inheritance 

 being 100 per cent. It is probable that this rule ^ applies only 

 to fluctuating variations, and not to cases of sudden or discon- 

 tinuous variation to be discerned later. 



Finally, correlated variation is often observed between differ- 

 ent parts of the same individual, which vary together. For 

 example, height in man is correlated with the length of the 

 humerus. There are, however, other organs in man which do 

 not seem to be correlated with height; for instance, the length 

 of the clavicle. 



^ Pearson's law of inheritance is different (see Darbeshire), the correlation 

 between successive generations being expressed by the fractions .3, .15, .075, 

 .0375, etc. 



