Xyz Arend. L. Hagedoorn and A. C. Hagedooru. 



But if another tells us, that this man is six feet long and his 

 brother five, and that therefore they have an average height of five 

 and a half foot, there would certainly be some, who would think that 

 this statement corresponded to something biologically. The tailor of 

 our two brothers would certainly smile if they brought him their 

 carefully tabulated average measure. 



A statistical representation of the facts of inheritance would be 

 justified only in a case in which we would be sure that the inheri- 

 tance were wholly "blending". Cases of true "blending" inheritance 

 have never been found by analytical breeding-experiments worth the 

 name, and probably "blending" inheritance is a fiction, the belief in 

 which has been created by just this statistical method, by making a 

 hash of facts of inheritance. 



Suppose we have a population of birds, in which some are black, 

 and some, because of a lack of one certain gene, are white. We start 

 with equal proportions. We now make an arbitrary scale of colours, 

 beginning with white, calling it one, passing through very light grey 

 and darker gray to black, calling these grades i — lo. The average 

 colour of our population will be one plus ten divided by two, which 

 makes five and a half. Now some influence sets in which favours 

 the survival of the black birds rather than the white ones. For in- 

 stance, the birds are turned out to shift for themselves and the white 

 ones fall more easily a prey to cats and owls. 



Gradually the average shade darkens. It passes from 5,5 to 

 6 and 7 and 8, finally becomes nearly 10, nearly all the white ones 

 having been caught. 



Now this example illustrates several things. In the first place, 

 our average colour will be pushed to one direction, not only in the 

 course of generations, but even during the first winter, when the 

 animals do not breed, but are simply being eaten, the white ones 

 more rapidly than the black ones. 



Secondly, it shows, how the method gives us an alltogether wrong 

 idea of the facts. In reality, in the population there never has been 

 seen an animal of grade 6 or 7, or of any other grade but i or 10. 



And why, if the method proves itself wrong in a case where the 

 difference between the individuals mixed is a striking, "discontinuous" 

 one, should it be less wrong in cases, in which differences are more 

 subtle and smaller, and where it is less easily proven absurd? 



Even if in partially albinistic rats, there did not exist any but 

 self-coloured animals and mutually indistinguishable light Hooded 



