380 



Oll the Laiüs of I nherUance in Man 



the theory of heredity, it seems well to hold no material back simply becauso oiie 

 kuows it to be untrustworthy. I thereforo givc the frequency distribution für 

 every coefficicnt of parental correlatioii I am awaie of, ncither wcightiüg them 

 with the iHimber of paiis on which the}' arc based, nor rcinarking on the relative 

 reliability of the data, which coveis plants, insects aud aniinals. 



Frequency Distribution of Coi-relation Coefficients of Parental Heredity. 



This distribution is represented graphically in the accoinpanying diagram ; the 

 mean and Standard deviation of the System are "430 + '010 and 'lü? respectively. 

 The fact of the Cluster and its quantitative intensity are thus rendered obvious. 

 The four extreme observations on tlie left are due to the Basset Hound Sires and 

 North AiiiiTicaii liidian Falhers, bnth involviiig doubtfid paternity*. If we omit 



OuoBAM VII. Distribution of Gorrelation Coffficients in 52 Cases of Parental Heredity. 



245 -295 345 -395 -445 



Intensitij of Parental Heredity. 



495 



595 



645 



* I3oth sorics are ftlso very small, 100 to 400, as couipared with the 1000 or morc of luost of the other 

 series. As to their questionable charactcr see R. S. Proc. Vol. 62, p. 414, and Vol. 66, p. 158 and 

 espccially footnotc p. 159. 



