454 THE GRAMMAR OF SCIENCE 



ing the latter? It has been asserted that the characters 

 of the ovum can be influenced by previous sexual unions 

 of the individual from which it proceeds. Dog breeders 

 assert that if a thoroughbred bitch has cast a litter to a 

 dog of another race, she will not afterwards breed true. 

 Darwin tells us of a mare which after bearing a foal to a 

 quagga bore a zebra-striped foal to a horse. In what way 

 the male element is supposed to influence the later pro- 

 duction of female elements is not defined, and both 

 mechanically and physiologically it seems inconceivable. 

 Such a possible influence is termed telegony. I shall show 

 later that I have been unable to find any quantitative 

 evidence for a steady telegonic influence in man, and 

 quite recently Professor Ewart, repeating the quagga 

 experiment, has attributed markings observed on the 

 second foal to reversion and not to telegony. 



Having now indicated in brief outline the general 

 features of heredity, I will proceed to consider more at 

 length its quantitative treatment. 



^ 7. — On tJie Quantitative Measurement of Heredity 



Let A and B be any pair of relatives, father and son, 

 sister and brother, great uncle and nephew, etc. Let x 

 be any organ of A, and y any of B. Then if x and y 

 be the same organ, we are said to investigate the direct 

 heredity between A and B ; if .r and y be different organs 

 we deal with the cross heredity. Now let the organs of 

 1000 such pairs of relatives be measured, and precisely 

 as we investigated the stature relationship of husband 

 and wife (p. 431), let a correlation table be formed for the 

 organs of the A, B pairs. To illustrate the point, con- 

 sider the following table for the inheritance of stature 

 from father to son. Here the average height of fathers 

 is 69". 1 1, and of sons 6g".2^. Fathers, however, of a given 

 height have not sons all of a giv&n height, but an array 

 of sons of a mean height different from that of the father 

 and nearer to the mean height of sons in general. Thus 

 take fathers of stature 72", the mean height of their 



