28 Darbishire, Laws of Heredity. 



And no flaws in the argument of my paper on the 

 supposed antagonism of Mendelian to biometric theories 

 have been pointed out to me ; in fact, Correns* and 

 Giard"f* have expressed their agreement with it. 



I feel most strongly that so long as we confuse 

 physiological with statistical Laws of heredity we are 

 wandering in the dark : we cannot know in what 

 direction our studies are leading us, whether we are 

 establishing correlations among the leaves or are digging 

 among the roots. It may or may not be that what we 

 learn by the former method is all that we shall ever 

 know, and that we shall find nothing by our digging ; but 

 be this as it may, I hold that it is essential to progress 

 in discovery, no less than to clearness of thought, that 

 we should know which of the two we are doing. 



4 {d). Description of a method of dealing %vith the material 

 of a breeding experiment in such a way that the 

 data obtained may be used to test the validity both 

 of Mendel's and Pearson's Laiv. 



When I had finished my last paper on my hybridiza- 

 tion experiment with mice,| I was still of the opinion 

 that Mendelian and biometric Laws of heredity were 

 mutually exclusive, and that if I could discover which of 

 the two was true, I should be making a forward step in 

 our knowledge of heredity. I therefore devised an ex- 

 periment which was destined to settle this question ; and 

 wasted a year in carrying it out. As soon as I discovered 

 the true relation of the two Laws I devised a method 

 of dealing with my experiment, of such a kind that the 

 results could be utilized by the Mendelian or the bio- 



* Correns, :05, p. 43. 

 t Giard, :0S, p. 22. 

 Ij: Darbishire, :04^. 



