358 HEREDITY IN MAN 



unit-factors or not, is not essential to the present inquiry. We 

 require to be able to give some answer to such questions as 

 whether disease, intellect, and temperament are represented in 

 the germinal constitution or not, and we can find an answer to 

 these questions without concerning ourselves with the ultimate 

 nature of inheritance. 



2. Since studies of inheritance on Mendelian lines do not at 

 present enable us to say what human characters are inherited, 

 we are dependent for our knowledge of this subject chiefly upon 

 biometry. The method pursued by biometricians is as follows. 

 The particulars with respect to any character — say stature — are 

 noted both for the parents and for the children. The average 

 degree of resemblance between parents and children can thus be 

 measured and expressed numerically. If the resemblance was 

 complete, if, that is to say, in respect of any character children 

 exactly resembled their parents, the fact would be expressed by 

 saying that the correlation was equal to unity. When the re- 

 semblance is less, the fact is expressed by representing the degree 

 of resemblance as a fraction. It must be emphasized that this 

 method merely measures the average degree of resemblance ; it 

 does not without further inquiry tell us how far that resemblance 

 is due to inheritance. So far as such degrees of resemblance are 

 found, it is quite possible, supposing that we had no further know- 

 ledge of the subject at all, that they might arise as follows. It 

 might be that all men had the same germinal constitution, and 

 that the degree of resemblance was due to the fact that fathers 

 and their children were brought up under more or less similar 

 surroundings. Though this fact should be borne in mind, such 

 an explanation of the resemblances found in the cases that will 

 be quoted in what follows is as a matter of fact shut out, and we 

 may accept the correlations as a measure of the degree of likeness 

 due to inheritance. 



The correlations between parents and children and between 

 children of the same parents have been determined in respect of 

 many characters — both mental and physical — and found to be 

 about 0-5. Thus, in respect of height the correlation between 

 father and son was found to be 0-514, between father and daughter 

 0-510, between mother and son 0-494, between mother and daughter 

 0-507, between brother and brother 0-511, between sister and sister 

 0-637, between brother and sister 0-553. So too the average parental 



