130 



BLASTOGENIC VAEIATIONS. 



parents, the variability of the race does not become less 

 and less for each generation, and so finally be reduced 

 to zero. Why this is not the case is perhaps most 

 readily grasped by examining a statistical table of the 

 relations between parent and offspring in respect of 

 some character. The table here given is reduced from 

 a larger one given by Galton in his " Natural Inherit- 

 ance " (p. 208), and represents the numbers of adult 

 children of various heights born of 205 mid-parents of 

 various heights. For instance, we see that the 17 mid- 



parents 71 to 73 inches in height had between them 62 

 children, of whom the most frequently occurring were 

 also 71 to 73 inches, or the same height as their parents. 

 Children of from 69 to 71 inches were, however, nearly 

 as frequent, so that, on the whole, it is obvious that the 

 stature of the children was more mediocre than that of 

 the parents. The median, or middle height of all these 

 62 values, was, in fact, only 70.6 inches, or 1.4 inch 

 less than the median height of the mid-parents. The 

 median of the children of mid-parents 69 to 71 inches 

 in height, was .9 inch less than their median; whilst in 

 the children of mid-parents varying from 67 to 69 

 inches it was only .1 inch less, for the median of these 



