6 Darbispiire, Laws of Heredity. 



and that half of its grandparents are tabbies, a quarter 

 whites, and a quarter blacks, you are enabled to predict 

 by Galton's Law the proportions in which these three 

 kinds of cats will occur in that generation. Pearson's Law 

 does not enable you to do this : it is of an entirely different 

 kind. For Pearson's Law to be true it is not necessary 



FATHERS. 



Black. 



Dark 

 Grey. 



Grey. 



Pale 

 Grev. 



White. Totals. 



Purple. 



Bluish 

 1^ Purple. 



K Purplish 

 P Blue. 



I— I 



() Blue. 



Pale 

 Blue. 



Totals. 



The degree of correlation in the above Table is that between the 

 number of dice exhibiting 4 or more than four points (or "pips") upper- 

 most, in first throws and the number exhibiting those faces uppermost in 

 second throws in a series of 300 double tiirows. The two throws are 

 correlated by leaving half the first throw on the table, so that the second 

 throw has half the dice lying exactly as they fell in the first throw. For 

 details of this process the reader is referred to Weldon :o6, p. 100. 



The degree of correlation in my imaginary Table is "54 ; for calculating 

 which I am indebted to Mr. Udny Yule. The "coefficient of parental 

 heredity," therefore, in this case is identical with that for the inheritance of 

 deafness, of which Pearson's Law is true, recently worked out by Schuster 

 (:o6, p. 478). Yet, in my imaginary case, none of the children could be 

 mistaken for any of the parents. 



