222 



The Law of Äncestral Heredity 



coiifidfiit tliat parcntal inhuritanci; fi«r man lies bctwcen 'iö and 5, and is quite 

 iiiconipatible wilh a valuc of J*. 



I now tum to the probli'Ui i»f Iidw lar the data can be represented by a 

 geometrical series. I fitted the best geoinetrical series fiist to the horee only, 

 then to man, aml fiimlly to both with the following results. 



Here we have : 



a=4921 and »• = •6428. 



Lastly, if" wc take a = "5, /• = '^ as close round niunbers, we have : 

 D. Horso, observed Man, observed Close Series 



•52 -49 50 



•30 32 33 



•19 19 ^22 



•15 — l'' 



Withiii the eiTors of observalinn this close series represents e.xcellently the 

 observed re.siilts fur either eye-coloiir in man nr coat-cülour in thi- horse. In othcr 

 words : 



* As long as the only data for a measurablc character in mau — Mr Galton's stature data — gave a 

 parental correlation of about ä it seemed ncodful to eniidiasi-sc the distinctiou between the results for 

 eye-colour and stature. In the light of nijprcsent knowlcdm», the distinction between the two classes of 

 charactcrs seems not so clear or necdful. A large series of podigree dogs at prosent in band gives a 

 parental correlation almost idcntical with tliat of horse and man. 



